.

EPA’s Boiler MACT: Controlling Emissions of
Hazardous Air Pollutants

James E. McCarthy
Specialist in Environmental Policy
August 4, 2011
Congressional Research Service
7-5700
www.crs.gov
R41459
CRS Report for Congress
P
repared for Members and Committees of Congress
c11173008

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Summary
On May 16, 2011, the U.S. Environmental Protection Agency (EPA) announced that it was
delaying the effective date of new Maximum Achievable Control Technology standards for
boilers (the “Boiler MACT”), in order to take additional public comment and complete
reconsideration of the rule. The standards were promulgated February 21, 2011, to meet the
requirements of Section 112 of the Clean Air Act. There is widespread interest in the rule’s
requirements and their potential effects, because boilers are used as power sources throughout
industry, and for power or heat by large commercial establishments and institutions as well.
EPA developed the regulations because it has found, based on emissions data, that boilers
(including coal-, biomass-, and liquid-fired boilers) are major sources of hazardous air pollutants
(HAPs). The Clean Air Act defines a major source as any facility that emits 10 tons or more of a
single listed HAP or 25 tons of any combination of HAPs annually. The HAPs themselves (187
substances) were listed by Congress in the 1990 Clean Air Act Amendments.
The rule promulgated in February would replace a 2004 version of the rule that was vacated and
remanded to EPA by the D.C. Circuit Court of Appeals. EPA was under a court order to
promulgate a replacement by January 2011. In early December 2010, the agency asked that the
promulgation deadline be postponed to April 2012 so that it could revise the proposed rule based
on new information it had received. The court refused EPA’s request, but did grant it five extra
weeks. In order to obtain the time it wanted, EPA announced on the same day that it promulgated
the rule that it would reconsider it. The agency subsequently asked for comment on 14 different
aspects of the rule, and stated that it will propose any changes to the rule by October, and make
final decisions by April 2012.
As promulgated in February, the MACT would affect 13,840 boilers and process heaters, with
capital costs of $5.1 billion, according to the agency; annualized costs, which spread the costs of
capital over the expected life of the equipment and include operating and maintenance expenses,
were estimated at $1.4 billion per year. Most of these costs would be borne by boilers that burn
coal, biomass, or liquid fuels, which together account for only 12.7% of all the units covered by
the rule. Most (83%) of the boilers affected by the rule are fueled by natural gas or refinery gases.
These boilers would not have to install pollution control equipment and would experience cost
savings under the rule, according to EPA. For the rule as a whole, EPA estimated that benefits—
including the avoidance of 2,500 to 6,500 premature deaths annually—would outweigh costs by
at least $20 billion per year.
Affected industries and many in Congress have raised objections to the rule both as proposed and
as promulgated, and bills have been introduced in both the House and Senate (H.R. 2250 and S.
1392) to alter the rule’s requirements and delay its implementation. In response to comments, the
rule promulgated in February 2011 had already reduced the number of units expected to require
controls, and made the emissions standards much less stringent than those in the proposed rule,
reducing the agency’s estimate of annualized control costs from $2.9 billion to $1.4 billion.
This report also discusses three related rules that EPA promulgated at the same time as the Boiler
MACT, dealing with smaller “area source” boilers and with commercial and industrial boilers that
burn solid waste (the “CISWI rule”). These rules have been less controversial, but the CISWI rule
is also being stayed to allow its reconsideration.
Congressional Research Service

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Contents
Introduction ................................................................................................................................ 1
Why Is EPA Considering Regulating These Sources? .................................................................. 2
Reconsideration of the Rule ........................................................................................................ 2
The Promulgated Standards......................................................................................................... 3
Standards for Existing Coal, Biomass, and Liquid Units ........................................................ 3
Standards for Existing Gas-Fired Units.................................................................................. 6
Standards for New Boilers..................................................................................................... 7
EPA’s Estimates of the Boiler MACT’s Costs and Benefits .......................................................... 7
EPA’s Projected Costs ........................................................................................................... 8
EPA’s Projected Benefits ..................................................................................................... 10
Other Cost Estimates................................................................................................................. 11
The CIBO Study ................................................................................................................. 11
The NACAA Critique ......................................................................................................... 13
The AF&PA Study .............................................................................................................. 14
Some General Thoughts on Cost and Economic Impact ....................................................... 14
Should EPA Have Set Health-Based Standards Under Section 112(d)(4)? .................................. 15
Smaller (Area) Sources ............................................................................................................. 17
Related Rules on Solid Waste Incineration ................................................................................ 19
Defining Solid Waste .......................................................................................................... 19
Used Oil ....................................................................................................................... 20
The CISWI Rule ................................................................................................................. 22
Conclusion................................................................................................................................ 22

Tables
Table 1. Change in Emission Standards for Existing Major Source Boilers, Promulgated
Boiler MACT vs. Proposed Rule .............................................................................................. 5
Table 2. Estimated Costs to Existing Boilers for Compliance with EPA’s Promulgated
Boiler MACT........................................................................................................................... 9
Table 3. Proposed and Promulgated Emission Limits for HCl Emissions from Existing
Boilers ................................................................................................................................... 17
Table 4. Annualized Compliance Costs for Area Source Boiler Rule ......................................... 18

Contacts
Author Contact Information ...................................................................................................... 23

Congressional Research Service

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Introduction
This report provides information concerning EPA’s Maximum Achievable Control Technology
standards for boilers (the Boiler MACT), which were finalized February 21, 2011, and appeared
in the March 21, 2011, Federal Register.1 The report also discusses three related rules that were
promulgated at the same time. On May 16, 2011, EPA announced that it was delaying the
effective dates2 of the Boiler MACT and one of the related rules, in order to take additional public
comment and complete reconsideration of the rules, leaving in doubt both the final form of the
rules and when the standards might go into effect. Because boilers are used as power sources
throughout industry, and for power or heat by large commercial establishments and institutions,
there has been widespread interest in the rules’ requirements and their potential effects.
Even before EPA’s May 16 decision to delay the rules’ effective date, the agency had announced
its intention to reconsider the promulgated rules. The Notice of Reconsideration, which appeared
in the March 21 Federal Register, 3 did not include a schedule, and it listed 14 provisions
(including the subcategories the agency used to set standards and the emission standards
themselves) for which the agency thought additional opportunity for public review and comment
should be obtained. It stated that the agency might seek public comment on other aspects of the
rules, as well.
The May 16 notice4 established a new 60-day public comment period, which ran through July 15,
and it delayed the effective date of the rules until EPA completes its reconsideration of the rules
or until proceedings for judicial review of the rules are complete, whichever is earlier. EPA
subsequently announced that it expects to propose any changes to the rules by October 31, 2011,
and finalize its decisions by April 30, 2012.
Given the rules’ potential impacts and the uncertainty of their final form, there continues to be
concern among stakeholders as to the final requirements. EPA estimates that, as promulgated in
February 2011, the Boiler MACT would affect 13,840 boilers and process heaters.5 In order to
reduce emissions of a wide array of hazardous air pollutants, about 13% of the affected units
would be required to install pollution control equipment. The 13% include coal-fired, biomass-
fired, and liquid-fired boilers. The agency estimated the capital costs associated with the rule at
$5.1 billion to meet the compliance deadline in 2014; annualized costs, which spread the costs of
capital over the expected life of the equipment and include operating and maintenance costs as
well, were estimated at $1.4 billion per year. These cost estimates were about half the estimated
cost of EPA’s proposed version of the rule, for reasons that we will explore below.

1 U.S. EPA, “National Emission Standards for Hazardous Air Pollutants for Major Sources: Industrial, Commercial,
and Institutional Boilers and Process Heaters; Final Rule,” 76 Federal Register 15608, March 21, 2011.
2 The rules, as promulgated, had effective dates of May 20, 2011, although compliance with their emission standards
would not have been required until at least three years later.
3 U.S. EPA, “National Emission Standards for Hazardous Air Pollutants; Notice of Reconsideration,” 76 Federal
Register
15266, March 21, 2011.
4 http://www.epa.gov/airquality/combustion/docs/20110516nextstep.pdf.
5 The data in this paragraph are from U.S. EPA, Regulatory Impact Analysis: National Emission Standards for
Hazardous Air Pollutants for Industrial, Commercial, and Institutional Boilers and Process Heaters, February 2011, p.
3-3, at http://www.epa.gov/ttn/ecas/regdata/RIAs/boilersriafinal110221_psg.pdf. Hereinafter, “Regulatory Impact
Analysis.”
Congressional Research Service
1

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Most boilers—83% of those affected by the rule—are fueled by natural gas or similar gases such
as refinery gas, according to EPA. These gas-powered boilers would incur capital costs averaging
a little less than $6,500 per unit, according to the agency. Through fuel savings, the agency
expects a reduction in operating costs to more than compensate for the capital expenditures of
most gas-powered units.
Why Is EPA Considering Regulating These Sources?
EPA has developed regulations addressing boiler emissions because it has found, based on
emissions data, that the coal-fired, biomass-fired, and liquid-fired units are major sources of
hazardous air pollutants (HAPs). Section 112 of the Clean Air Act, which requires controls on
major sources of HAPs, defines a major source as any facility that emits 10 tons or more of a
single listed HAP or 25 tons of any combination of HAPs annually. The HAPs themselves (187
substances) were listed by Congress in the 1990 Clean Air Act Amendments.
Boilers emit at least 20 of the listed HAPs, including mercury, arsenic, chromium, cadmium,
selenium, nickel, lead, manganese, phosphorous, antimony, beryllium, polycyclic organic matter,
benzene, formaldehyde, acetaldehyde, dioxins, furans, hydrogen chloride, hydrogen cyanide, and
hydrogen fluoride. Six of these 20 are classified as known or probable human carcinogens. Others
affect the lungs, skin, central nervous system (including adverse developmental effects), and/or
kidneys.6 By controlling boiler emissions, EPA expects to avoid 2,500 to 6,500 premature deaths
annually, as well as many other health effects, including 4,000 nonfatal heart attacks annually.7
The Boiler MACT replaces a rule promulgated on September 13, 2004, and subsequently vacated
and remanded to EPA by the D.C. Circuit Court of Appeals.8 The court vacated the rule in 2007,
saying EPA had wrongly excluded many industrial boilers from the definition of solid waste
incinerators, which have more stringent emissions requirements under the Clean Air Act. EPA
was under a court order to promulgate replacement rules by February 21, 2011.
Reconsideration of the Rule
In early December 2010, the agency petitioned the District Court for the District of Columbia for
up to 15 months of additional time to complete the rulemaking. The agency argued that in light of
the extensive comments it received on the proposed rules, “EPA believes that the overall public
interest is best served by allowing EPA to re-propose the rules so that [it] will be able to issue
emission standards that are based upon a thorough consideration of all available data and reduce

6 U.S. EPA, “National Emission Standards for Hazardous Air Pollutants for Major Sources: Industrial, Commercial,
and Institutional Boilers and Process Heaters; Proposed Rule,” 75 Federal Register 32048, June 4, 2010. Also, see
Comments of Clean Air Task Force, Earthjustice, Natural Resources Defense Council, and the Sierra Club on National
Emission Standards for Major Sources: Industrial, Commercial, and Institutional Boilers and Process Heaters: Proposed
Rule, Docket ID No. EPA-HQ-OAR-2002-0058, August 23, 2010, p. 3.
7 U.S. EPA, “Final Air Toxics Standards for Industrial, Commercial, and Institutional Boilers and Process Heaters at
Major Source Facilities,” Fact Sheet, pp. 2-3, at http://www.epa.gov/airquality/combustion/docs/
20110221mboilersfs.pdf. Hereinafter, “EPA Fact Sheet.”
8 Natural Resources Defense Council v. EPA, 489 F. 3d 1250, 1257–61 (DC Cir. 2007).
Congressional Research Service
2

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

potential litigation risks.”9 The court had issued a summary judgment against the agency in 2006
for failure to discharge fully its duty to promulgate standards for emissions of hazardous air
pollutants.10 On March 31, 2006, the court imposed a schedule under which EPA was to have
discharged all of the statutory duties at issue by June 15, 2009. That deadline was subsequently
extended by more than a year and a half.
On January 20, 2011, the court denied EPA’s request for a further 15-month extension, concluding
that EPA has engaged in discretionary delay in the face of a congressional directive (i.e., the 1990
Clean Air Act Amendments, under which the rules were to have been promulgated by November
2000); the court gave the agency one month to issue final rules.11 Having been denied the
extension it sought, the agency issued a statement saying, “The standards will be significantly
different than what EPA proposed…. The agency believes these changes still deserve further
public review and comment and expects to solicit further comment through a reconsideration of
the rules.”12
True to its word, the agency issued a Notice of Reconsideration at the same time that it
promulgated the final rule. As noted earlier, the notice listed 14 provisions for which the agency
thinks additional opportunity for public review and comment should be obtained, and it stated
that the agency may seek public comment on other aspects of the rules. The 14 provisions
included such basic elements as the subcategories used to set standards in the boiler rule, the
emission standards themselves, and the monitoring requirements. This left numerous questions
concerning not only the substance of the rule, but the schedule for implementation. Following
promulgation, existing facilities would normally have three years to comply with the standards,
but if the agency is reconsidering key aspects of the rules, one is left to wonder how regulated
entities should determine what standards they will ultimately be required to comply with, and on
what schedule. By delaying the effective date of the standards, the agency has addressed this
uncertainty, effectively giving itself and regulated entities an extension of time similar to what the
court had denied them. There remains some uncertainty as to the effective date of the final
standards; but by delaying the effective date until it completes its reconsideration of the rules, or
until proceedings for judicial review of the rules are complete (whichever is earlier), the agency
assures the regulated community that it will have three years to comply with the standards once
they are final.
The Promulgated Standards
Standards for Existing Coal, Biomass, and Liquid Units
In order to understand the standards that EPA promulgated in March and the controversy
surrounding them, it helps to begin with the the agency’s June 4, 2010, proposal. In the proposal,

9 Sierra Club v. Jackson, No. 01-1537, 2001 Westlaw 181097, at *7 (D.D.C. Jan. 20, 2011).
10 Sierra Club v. Johnson, 444 F. Supp. 2d 46, 47 (D.D.C. 2006). This case is separate from the vacatur and remand of
the Boiler MACT in the 2007 NRDC v. EPA decision. In effect, there are two courts in two separate proceedings
overseeing EPA’s boiler rules.
11 Sierra Club, 2001 Westlaw 181097.
12 U.S. EPA, “EPA Announces Next Steps on Emissions Standards for Boilers, Certain Incinerators,” Press Release,
January 20, 2011, at http://yosemite.epa.gov/opa/admpress.nsf/6424ac1caa800aab85257359003f5337/
58f5bee5e13c61228525781e007e9881!OpenDocument.
Congressional Research Service
3

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

EPA divided boilers into 11 subcategories, with separate emission limits for new and existing
units in 9 of the 11.13 The nine subcategories included three types of coal-fired boilers and four
types of biomass-fired boilers.14
The proposed emission limits covered five substances (or groups of substances): mercury;
dioxins/furans; particulate matter (as a surrogate for non-mercury metals); hydrogen chloride (as
a surrogate for all acid gases); and carbon monoxide (as a surrogate for non-dioxin organic air
toxics, including formaldehyde).15
The Clean Air Act requires that MACT emission standards be based on the emission control
achieved by the best controlled similar sources. Thus, the emission limits proposed for the five
groups of pollutants were based on monitoring data obtained from facilities in each of the nine
subcategories of existing boilers.16
• For new sources, the statute requires (in Section 112(d)(3)) that standards be
based on the emission control achieved by the best controlled similar source.
• For existing sources, on the other hand, the same subsection of the statute
requires standards no less stringent than the average emission limitation achieved
by the best performing 12% of existing sources. The performance of the best
12% is generally referred to as the “MACT floor,” since it sets the minimum
requirements for MACT standards. The MACT floor is based solely on the
performance of existing facilities in the category or subcategory of sources, with
no consideration of the cost or economic impacts thereof. The Administrator is
only allowed to take costs, health, energy, and environmental factors into
consideration to the extent that she considers setting standards that go beyond the
floor.
Given the statutorily required methodology for identifying the MACT floor, the number of
subcategories the agency identifies is an important factor in determining how stringent the
standards will be: the more subcategories EPA identifies, the more variation there will be in the

13 The other two subcategories were only required to meet work practice standards. Work practice standards require
certain actions from the regulated entity (e.g., the boiler must be tuned up annually, or the owner must conduct an
assessment to identify energy conservation opportunities), but they don’t set numeric standards for emissions.
14 The three types of coal-fired boiler identified in the proposed rule were coal stoker, coal fluidized bed, and
pulverized coal. The four types of biomass-fired boilers were biomass stoker, biomass fluidized bed, biomass
suspension burner/Dutch oven, and biomass fuel cells. In addition, the agency proposed emission limits for liquid-
fueled boilers, and gas-fired boilers using “other process gases.”
15 Hydrogen chloride is often referred to as hydrochloric acid, because when the gas encounters water in the
atmosphere it forms an acidic solution of hydrochloric acid. The specific emission limits EPA proposed for each of the
five pollutants can be found in the June 4, 2010, Federal Register at p. 32012, Table 1.The final standards are at 76
Federal Register 15612, Table 1, March 21, 2011.
16 EPA has acknowledged that it did not have as much data as it might have wished to use in establishing the boiler
subcategories and the proposed MACT standards. In a September 28, 2010, letter, the Administrator stated: “In an
effort to establish subcategories wherever appropriate, and to calculate accurately the standards for each subcategory,
EPA asked the affected companies and institutions for technical data about their facilities long before the court-ordered
deadline for publishing a proposal. As is often the case in Section 112 rulemaking efforts, however, EPA did not
receive many data. While the agency was not left entirely lacking in relevant information, the limited response from
affected businesses and institutions did make it difficult for EPA to delineate subcategories and calculate standards that
fully reflected operational reality. The agency nevertheless was legally required to publish proposed subcategories and
standards based on the information it had at the time.” Letter of EPA Administrator Lisa P. Jackson to Senator Mary L.
Landrieu, September 28, 2010, p. 2.
Congressional Research Service
4

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

MACT floor for each, and thus the more flexibility the agency will have in setting different,
potentially less stringent emission standards for different boiler types. If, because of
subcategorization, the Administrator decided that a subcategory’s MACT floor did not provide
sufficient protection for human health or the environment, she would still have the authority to set
“beyond the floor” standards for a subcategory: in doing so, however, she could consider the cost
of the standards and other factors. Thus, one issue raised by commenters on the proposed rule was
whether EPA’s subcategorization of the boiler universe appropriately considered the differences in
size, fuels, boiler design, location, etc., or whether the subcategories should be modified from
those proposed.
In the final rule, EPA responded to the comments it received by modifying its subcategories.
Instead of 11 types of boiler, the final rule identifies 15, including an additional biomass
subcategory, a limited-use subcategory,17 and a subcategory for “non-continental” (i.e., Hawaiian,
Puerto Rican, and U.S. island territories) liquid-fueled boilers.18 Based on new data provided by
industry and on some corrections to the data it had used earlier, the emission standards for almost
every subcategory were modified, mostly making the standards less stringent. In many cases, the
change was substantial: for example, as shown in Table 1, existing coal-fired boilers will be
allowed 95% more particulate matter (PM), 75% more hydrogen chloride (HCl), 53% more
mercury, and 77%-440% more carbon monoxide emissions. The increase in allowable emissions
is even greater for most of the pollutants emitted by existing biomass units: they will be allowed
to emit almost six times as much HCl and five times as much mercury as they would have under
the proposed standards, in addition to 95% more PM. The standards for liquid-fueled units also
changed, allowing 10 times as much carbon monoxide and 200 times as much dioxin and furans.
Table 1. Change in Emission Standards for Existing Major Source Boilers,
Promulgated Boiler MACT vs. Proposed Rule
Estimated Particulate
Hydrogen
Number
matter
chloride
Carbon
Fuel type
of Units
(PM)
(HCl) Mercury
monoxide Dioxin/furans
Coal 540
+95%
+75%
+53%
+77%
to
no change
+440%
Biomass 474
+95%
+483%
+411%
-53%
to
no change to
+247%
+19,900%
Liquid 713
+87%
-63%
-12%

+900%
+19,990%
Other process
118 -14%
+56,566%
+6,400%
+800% +789%
gases
Source: U.S. EPA, with calculations performed by CRS.
Notes: Some values are expressed as a range to reflect the fact that there are multiple subcategories among
boilers burning the particular type of fuel. Does not include non-continental liquid-fueled units.

17 The agency defined limited use units as those having a federally enforceable limit of no more than 876 hours per year
of operation (i.e., operable at most 10% of the year). 76 Federal Register 15684, March 21, 2011.
18 In the rule’s preamble, EPA stated that there was a “need for a noncontinental liquid fuel subcategory for island units
that have limited fuel options and other unique circumstances.” 76 Federal Register 15633, March 21, 2011. See also
discussion at p. 15635.
Congressional Research Service
5

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

A second issue raised by critics of the agency’s proposal had to do with the way that EPA
identified the best performers within the subcategories. As it has done previously for other
categories of sources, EPA averaged the emissions performance of the top 12% of existing units
separately for each of the five pollutants subject to emission limits. Critics who believed the
proposed standards were too stringent argued that by considering the pollutants separately, the
agency was, in effect, cherry-picking the best performers and setting a combined standard for the
five pollutants that no existing facility may actually meet.
This question—whether one identifies the best-performing sources pollutant-by-pollutant or for
all the pollutants as a group—was addressed in regard to another standard, the
Hospital/Medical/Infectious Waste Incinerator rule, which EPA promulgated in October 2009.19 In
promulgating that rule, the agency stated:
There is no reason not to consider emissions data and controls in use at sources that may be
the best performers from some pollutants but not for other pollutants. The MACT floor
controls applicable for one pollutant do not preclude the use of MACT floor controls for
another pollutant. Therefore, it is appropriate to consider controls at sources employing
MACT controls for some pollutants, but not all.20
EPA acknowledged in the preamble to that rule that “there appears … to be a substantial
ambiguity in the statutory language about whether the MACT floor is to be based on the
performance of an entire source or on the performance achieved in controlling particular
hazardous air pollutants.”21 But the agency noted that commenters in the past have not objected to
the use of the pollutant-by-pollutant approach. They also noted that the D.C. Circuit Court of
Appeals has reviewed MACT floor determinations made on a pollutant-by-pollutant basis without
finding error in the approach.22 Thus, the agency believes the best reading of the act is that the
standards are to be set on a pollutant-by-pollutant basis—the only exception being if there is
reason to believe that control of one pollutant will lead to increased emissions of another.
In the preamble to the final Boiler MACT, EPA provides a similar discussion, concluding that,
although the language of Section 112(d)(3) is ambiguous, “EPA’s HAP-by-HAP approach fulfills
the evident statutory purpose and is supported by the most pertinent legislative history.”23
Standards for Existing Gas-Fired Units
In the final rule, three subcategories are not subject to emission limits: natural gas/refinery
gas/clean gas (a subcategory that EPA calls Gas 1), metal process furnaces, and limited-use units
(defined as units that operate less than 10% of the time). For these three, the agency set only a
work practice standard, requiring that boilers be tuned up annually and that the owners submit
reports to EPA setting forth specific information from the tune-up procedure. The Administrator
has authority to substitute a work practice standard for emission standards when, in her judgment,
it is not feasible to prescribe or enforce an emission standard. As noted earlier, 83% of existing

19 Medical Waste Institute et al. v. EPA, No. 09–1297 (D.C. Cir.).
20 U.S. EPA, “Standards of Performance for New Stationary Sources and Emissions Guidelines for Existing Sources:
Hospital/Medical/Infectious Waste Incinerators; Final Rule,” October 6, 2009, 74 Federal Register 51381.
21 Ibid.
22 Ibid. The case in question was Sierra Club v. EPA, 167 F.3d 658, 660 (D.C.Cir. 1999).
23 The full discussion is found in Section V.A. of the Preamble, at 76 Federal Register 15621-15623, March 21, 2011.
Congressional Research Service
6

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

boilers fall into the natural gas/refinery gas/other clean gas subcategory, and thus are only subject
to the tune-up requirements. Limited use boilers that burn liquids would also qualify for work
practice standards, requiring a tune-up every two years.
All boilers would also be required to perform a one-time energy assessment to identify cost-
effective energy conservation measures.
Standards for New Boilers
EPA also promulgated MACT standards for new (as opposed to existing) major source boilers.
These standards are, in all cases, more stringent than the standards for existing units—in many
cases, substantially so.24 The agency assumes, however, that no new coal- or biomass-fired major
source boilers (and very few major source boilers of any kind) will be built in the next three
years. The agency has stated that the projected type and number of new boilers comes from the
Energy Information Administration at the Department of Energy and is not based on the Boiler
MACT.25
Of the estimated 47 new units, the agency expects 34 to be powered by natural gas, with
annualized costs of compliance averaging -$588 apiece (i.e., a savings of $588, due to fuel
savings, compared to the absence of regulation).26
EPA’s Estimates of the Boiler MACT’s Costs and
Benefits

Among the boilers affected by the Boiler MACT rule, there are an estimated 713 units that burn
liquids and 1,014 units that burn solids (nearly 600 of them coal-fired and the rest biomass-fired).
The rule also applies to other types of boilers, but these 1,727 units, which account for 12.5% of
the affected units, account for 95% of the compliance cost.
In general, the promulgated emission limits apply to boilers that have a designed heat input
capacity of 10 million Btu per hour or greater. How big is this? A coal-fired boiler subject to the
MACT would be one that is capable of burning roughly 1,000 pounds (a half-ton) of coal per
hour.27 Wood has less energy per pound than coal: a biomass-fired boiler burning wood might
require as much as 1,500 pounds of wood per hour to produce 10 million Btus. A boiler burning
fuel oil would need to burn about 70 gallons per hour. Many of the boilers to be regulated are
substantially larger than this. An analysis released by the Council of Industrial Boiler Owners

24 76 Federal Register 15612, Table 1.
25 Letter of EPA Administrator Lisa P. Jackson to Senator Mary L. Landrieu, September 28, 2010, p. 2.
26 Boiler MACT Rule, Preamble, Table 3, 76 Federal Register 15651, March 21, 2011. The other 13 units would be
powered by liquids, according to EPA, and would face annualized costs of $6.1 million, an average of $469 thousand
each.
27 A rough rule of thumb for coal is that it contains about 10,000 Btus of energy per pound. To be more precise, the
heating value ranges from 6,500 to 13,000 Btus per pound, depending on rank (i.e., type of coal), with bituminous coal
containing more than 10,000 and subbituminous and lignite less.
Congressional Research Service
7

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

(CIBO), for example, used a 250 million Btu/hour boiler as the base for its cost estimates.28 For a
boiler burning fuel oil, this would mean burning 1,750 gallons per hour.
In order to comply with the rule’s emission limits, these facilities may need to install fabric filters
(also known as baghouses) to achieve PM and mercury control; wet scrubbers to meet the
hydrochloric acid limits; replacement burners, tune-ups, and combustion controls for carbon
monoxide and organic HAPs; and carbon injection for mercury, dioxins, and furans. These are the
available technologies for maximum control of the relevant emissions.
Some observers maintain that, because EPA weakened the standards, the vast majority of facilities
won’t have to install these technologies. The National Association of Clean Air Agencies
(NACAA), the association that represents state air pollution control officials, surveyed its
members in 2008 to determine what should be defined as MACT. Using the data it obtained from
state officials, NACAA concludes that EPA’s final mercury emission standard for coal-fired
boilers is almost 16 times higher than the average of the best performing 12%; the carbon
monoxide standard is 213 times what the MACT floor should be, according to NACAA. The
problem according to NACAA’s Executive Director, is that “Compliance test results provided by
state and local permitting officials were not used [in setting the MACT standards]; instead EPA
relied on industry data.”29
EPA explains that some of the data that NACAA provided could not be used, because the test
reports were incomplete. The agency also notes that its process for setting a standard is more
complicated than simply averaging the best test results. Specifically, the agency subjects the
emissions data to what is called a “variability analysis.” This type of analysis attempts to
recognize that operating conditions and resulting emissions vary over time, yet facilities need to
be in compliance with emissions limits at all times. Emissions can change for several reasons:
there is variation in the amount of contaminants in fuel, for example; the boiler will sometimes be
operating at less than full load; and statistical tests applied to the data are used to set the actual
standard. The agency first identifies the best 12% by ranking the units based on their best test
results. In the next step they add all available test results for those units. Finally, using a statistical
test, they calculate a standard that these units can meet 99% of the time, despite variability in
operating conditions. This results in standards that are less stringent than the straight average of
the best 12% test results.30
EPA’s Projected Costs
As shown in Table 2, EPA estimates the capital costs of the promulgated rule to be nearly $5.1
billion, with annualized costs of nearly $1.4 billion. These costs fall almost entirely on units
burning solids (coal or biomass) or liquids. Most boilers, which are fueled by natural gas, will

28 IHS Global Insight for CIBO, The Economic Impact of Proposed EPA Boiler/Process Heater MACT Rule on
Industrial, Commercial, and Institutional Boiler and Process Heater Operators
, August 2010, Appendix A, p. 28.
29 Bill Becker, “Clean Air Issues Facing States and Localities: Regulatory Update,” presentation, Institute of Clean Air
Companies, April 27, 2011.
30 For a further discussion of variability analysis, see Amanda Singleton, ERG (Eastern Research Group, Inc.),
“Revised MACT Floor Analysis (2011) for the Industrial, Commercial, and Institutional Boilers and Process Heaters
National Emission Standards for Hazardous Air Pollutants – Major Source,” Memorandum to Brian Shrager, U.S. EPA,
January 4, 2011, at http://www.epa.gov/ttn/atw/boiler/rev_mact_floor_analysis_major_boilers_process_heaters.pdf.
Congressional Research Service
8

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

experience a reduction in operating costs that more than compensates for any capital costs,
according to EPA.
Despite the clear advantage that the promulgated rule would give to natural-gas-fired boilers, EPA
did not consider fuel-switching as a potential compliance strategy, for a variety of reasons. In the
preamble to the proposed rule, the agency stated: “This decision was based on the overall effect
of fuel switching on HAP emissions, technical and design considerations discussed previously in
this preamble, and concerns about fuel availability.”31 Although switching from solid to gaseous
fuels “would decrease PM and some metals emissions, emissions of some organic HAP (e.g.,
formaldehyde) would increase,”32 according to the agency’s analysis. Further, the agency
maintained, natural gas may be unavailable:
Natural gas pipelines are not available in all regions of the U.S., and natural gas is simply not
available as a fuel for many industrial, commercial, and institutional boilers and process
heaters. Moreover, even where pipelines provide access to natural gas, supplies of natural
gas may not be adequate.33
Table 2. Estimated Costs to Existing Boilers for Compliance with EPA’s
Promulgated Boiler MACT
Estimated
Annualized Cost
Number of
Capital Costs
Annualized Cost
per Unit
Subcategory
Affected Units
($ million)
($ million)
($ thousand)
Solid
units 1,014 $2,182 $846 $834
Liquid units
713
$2,656
$828
$1,161
Non-continental
liquid units
27
$86
$21
$777
Gas 1 (natural gas
/ refinery gas /
other clean gas)
units 10,797
$70
($325)
($30)
Gas 1 metallurgical
furnaces 694 $4.5
($6)
($9)
Gas (other) units
118
$79
$37
$314
Limited use
477
$3.1
($25)
($52)
Total
13,840
$5,080.6
$1,376

Source: U.S. EPA, Preamble to Boiler MACT Final Rule, Table 3, 76 Federal Register 15651.
Notes: (1) Parentheses indicate cost savings, resulting from fuel savings. (2) Per unit cost was calculated by
CRS. Some of the difference in unit costs could be accounted for by differences in boiler size.
Nevertheless, if the cost of compliance is sufficiently great, the incentive to explore fuel-
switching would seem substantial, particularly for facilities not burning a byproduct of the plant’s
operation. Recent accounts of the substantial increases in gas reserves as shale gas resources are

31 U.S. EPA, Proposed Boiler MACT, 75 Federal Register 32019, June 4, 2010.
32 Ibid.
33 Ibid.
Congressional Research Service
9

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

developed could ease some of the natural gas availability concerns, and might bear further
analysis.34
EPA used a multi-market partial equilibrium model developed for its Office of Air Quality
Planning and Standards to determine how stakeholders in 100 U.S. industries might respond to
the promulgated rule. The model found no U.S. industry in which production would decline by
more than 0.05%.35
EPA’s Projected Benefits
EPA estimates that implementation of the Boiler MACT, as promulgated, would reduce
nationwide emissions from major source boilers and process heaters by:
• 1.4 tons per year (tpy) of mercury,
• 2,700 tpy of non-mercury metals,
• 30,000 tpy of hydrogen chloride,
• 47,000 tpy of particulate matter (PM),
• 7,000 tpy of volatile organic compounds, and
• 440,000 tpy of sulfur dioxide.36
For most of these pollutants, the expected reductions are similar to those of the proposed rule.
This is not the case for mercury, however. The proposed version of the rule was estimated to
reduce mercury emissions by 7.5 tons, more than five times as much as the final version. Boilers
are currently thought to be the fourth-largest stationary source of mercury, yet other categories of
sources have been required to reduce mercury emissions to a greater extent than will be required
by the promulgated Boiler MACT rule.37
According to EPA, beginning in 2014, emission reductions resulting from the rule would lead to
important health benefits, including the annual avoidance of:
• 2,500 to 6,500 premature deaths,
• 1,600 cases of chronic bronchitis,
• 4,000 nonfatal heart attacks,
• 4,300 hospital and emergency room visits,

34 See, for example, the U.S. Energy Information Administration’s Annual Energy Outlook 2010 with Projections to
2035
, at http://www.eia.doe.gov/oiaf/aeo/gas.html: “A 4-fold increase in shale gas production from 2008 to 2035 more
than offsets a 31-percent decline in other lower 48 onshore natural gas production in the AEO2010 Reference case.
Significant increases in shale gas production are expected in the Northeast, Gulf Coast, and Midcontinent regions….”
Also, CRS Report R41543, Global Natural Gas: A Growing Resource, by Michael Ratner.
35 U.S. EPA, Regulatory Impact Analysis, Appendices A and B.
36 EPA Fact Sheet, p. 2.
37 EPA National Emissions Inventory, “Trends in Mercury Air Emissions Between 1990 and 2005,” June 28, 2010, and
U.S. EPA, “Emissions Overview: Hazardous Air Pollutants in Support of the Proposed Toxics Rule,” March 15, 2011,
available on request..
Congressional Research Service
10

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

• 3,700 cases of acute bronchitis,
• 310,000 days when people miss work,
• 41,000 cases of aggravated asthma,
• 78,000 cases of respiratory symptoms, and
• 1.9 million days when people must restrict their activities.38
EPA estimates the annual value of these benefits to range from $22 billion to $54 billion in
2014—outweighing the annualized costs by at least $20 billion. In its Regulatory Impact
Analysis, the agency states that this is only a partial estimate of the value of the rule’s benefits:
We were unable to monetize the direct benefits associated with reducing HAPs in this
analysis. In Section 7.5.5 of this RIA, we provide a full qualitative discussion of the direct
health benefits associated with the reductions in emissions of HAPs anticipated by these
rules, including a full discussion of the complexity associated with monetizing HAP benefits.
We also provide maps of reduced mercury deposition in that section. Therefore, all
monetized benefits provided in this analysis only reflect improvements in ambient PM2.5 and
ozone concentrations. Thus, the monetized benefits estimate is an underestimate of the total
benefits. The extent of this underestimate, whether small or large, is unknown.39
Other Cost Estimates
Not surprisingly for a rule of this size, EPA’s cost estimate is not the only one available. Industry-
funded studies of the proposed rule, including one from the Council of Industrial Boiler Owners
(CIBO), placed the costs of the rule substantially higher than EPA’s estimate, while an analysis by
the National Association of Clean Air Agencies concluded that CIBO’s study exaggerated the
potential costs.
The CIBO Study
CIBO’s study concluded that capital costs of the proposed Boiler MACT would be $20.7 billion,
more than double EPA’s estimate of the rule as proposed, and four times EPA’s estimated cost of
the rule as promulgated.40 CIBO estimated the cost of carbon monoxide controls at $2.7 billion,

38 EPA Fact Sheet, pp. 2-3.
39 Regulatory Impact Analysis, p. 7-3. In the qualitative discussion of the benefits of reducing HAP emissions, the RIA
states that the effects of exposure to HAP emissions can include neurological, cardiovascular, liver, kidney, and
respiratory effects as well as effects on the immune and reproductive systems and cancer. Reducing emissions may
reduce these effects, but none of these benefits were quantified in the benefit estimates. As summarized by the agency’s
Science Advisory Board, “(T)he challenges for assessing progress in health improvement as a result of reductions in
emissions of hazardous air pollutants (HAPs) are daunting ... due to a lack of exposure-response functions,
uncertainties in emissions inventories and background levels, the difficulty of extrapolating risk estimates to low doses
and the challenges of tracking health progress for diseases, such as cancer, that have long latency periods.” (RIA, p. 7-
43) As a result, the agency concluded, “Large reductions in HAP emissions may not necessarily translate into
significant reductions in health risk because toxicity varies by pollutant and whether or not there are exposures at or
above levels of concern is not known.” (RIA, p. 7-41)
40 IHS Global Insight for CIBO, The Economic Impact of Proposed EPA Boiler/Process Heater MACT Rule on
Industrial, Commercial, and Institutional Boiler and Process Heater Operators
, August 2010, pp. 29-30. Hereinafter,
“CIBO Study.”
Congressional Research Service
11

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

200 times EPA’s estimate; and the report estimated the cost of carbon injection at $1.7 billion, 180
times the EPA amount. CIBO’s estimate for hydrogen chloride controls was only three times as
expensive as EPA’s estimate; but, because EPA’s estimate was already more than $3 billion, the
difference would add $6 billion to the total cost of the rule.
CIBO identified six factors that accounted for most of the difference. According to the CIBO
study:
• EPA used outdated control cost estimates;
• to achieve the proposed carbon monoxide (CO) limits, it would have been
necessary to implement combustion controls, fuel feed system improvements, or
install a CO catalyst, at far greater cost than EPA’s conclusion that a tune-up or
burner replacement would be adequate to achieve the CO limits;
• activated carbon injection, in addition to fabric filters, would be required to
achieve the proposed rule’s mercury limits at a far higher number of boilers, and
EPA underestimated the cost of this technology by a factor of 15;
• PM emission controls would require fabric filters, which are more expensive than
EPA’s assumption that electrostatic precipitators (ESP) would be adequate to
meet the standard;
• more expensive scrubbers than EPA identified would be required for hydrogen
chloride control; and
• more facilities than EPA estimated would need to control dioxin/furan emissions.
Using an estimate of 16,000 jobs “at risk” of being eliminated for each billion dollars of cost,
CIBO concluded that 337,702 jobs would be at risk from implementation of the proposed rule.41
Nearly 70,000 of these would be in the directly affected industries; the rest would be in supplier
industries or spread across local economies through reductions in spending by workers directly
and indirectly affected.
Like the cost estimates, CIBO’s jobs “at risk” were much higher than EPA’s estimates. EPA found
that the employment impact of the proposed rule would range from a loss of 6,000 jobs to a gain
of 12,000.42 For the promulgated rule, EPA estimated employment impacts as ranging from a loss
of 3,100 to a gain of 6,500.43
Several factors help explain why CIBO’s estimates were so much higher than EPA’s. The most
important was the choice of model used to estimate the economic impacts. CIBO used an input-
output (I-O) model. I-O models are deterministic, in the sense that they start with an assumption
regarding a change in one industry’s output and track the effect of the change on other industries’
output and employment. An assumed loss in one industry translates directly to bigger losses in the
economy as a whole.

41 CIBO Study, pp. 9-10. The “at risk” jobs include direct impacts on jobs at the facilities that have to install pollution
controls, indirect impacts on jobs at suppliers of those firms, and induced impacts due to reductions in spending by
employees in the direct and indirect categories.
42 Draft Regulatory Impact Analysis, pp. 1-2 and 4-6 to 4-9.
43 Final RIA, p. 1-2.
Congressional Research Service
12

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

In the CIBO study, the assumption was that the expenditure of $20.7 billion on pollution controls
would be equivalent to a reduction of output by that amount.44 In fact, spending on pollution
control does not cause an equal reduction in output. Rather, changes in output caused by pollution
control expenditures would include increases in some industries along with declines in others.
• Output declines occur in cases where industries increase prices to cover their
higher costs, and consumers respond by demanding less of the affected product.
In these cases, the higher costs would generally reflect “annualized costs,” not
the full capital cost used by CIBO.
• Meanwhile, output in other industries will usually increase. As EPA noted, in the
Administrator’s September 28 letter to Senator Landrieu, expenditures on
pollution control are not simply a loss to the economy: they stimulate demand
and provide jobs in the pollution control sector.45
Thus, the assumption that output would decline by $20.7 billion at the base of CIBO’s analysis is
flawed. As a result, little credence can be placed in CIBO’s estimate of job losses.
The NACAA Critique
An analysis by the National Association of Clean Air Agencies46 cited other flaws in the CIBO
study. In addition to echoing the critiques above—that CIBO included no estimates of economic
or health benefits and treated one-time costs as recurring expenses—NACAA raised two other
major points. First, NACAA maintained, CIBO exaggerated the cost of the proposed rules by
overestimating the number of sources that must be controlled: CIBO’s estimate of the number of
sources that must be controlled is “grossly in error,”47 according to NACAA, because CIBO
assumed that any source for which there were no emissions data would have to install controls.
NACAA collected stack test data for boilers in 2008 in order to develop a model Boiler MACT
permit for use by the states, and thus is the source of much existing emissions information. Using
its existing data sets, NACAA looked at coal-fired boilers and found that 87% of the 39 units for
which there were emissions data already met EPA’s proposed standard for carbon monoxide. In
developing its cost estimates for the coal-fired units, however, CIBO assumed that none of the
remaining 146 (untested) units in the subcategory would meet the proposed standard, and thus
they would all have to install additional controls. Second, NACAA concluded that the available
data “reveal a continuum of emissions performance where there are substantial numbers of units
whose emissions are within 10 to 40 percent of the proposed standard….” For many of these units
“it should be anticipated that minor changes, such as blending in small amounts of clean fuel, will
suffice in lieu of major capital projects.”48

44 CIBO study, p. 11.
45 Letter of EPA Administrator Lisa P. Jackson to Senator Mary L. Landrieu, September 28, 2010, p. 3. Others have
also discussed the CIBO study’s failure to account for the increased output in the pollution control industry, along with
other criticisms of the study. See, for example, Laurie Johnson, “Boiler Industry’s Junk Economic Analysis of
Proposed EPA Toxic Emission Standards for Industrial Boilers,” NRDC Switchboard, http://switchboard.nrdc.org/
blogs/ljohnson/boiler_industrys_junk_analysis.html.
46 National Association of Clean Air Agencies, EPA’s Proposed Regulations on Hazardous Air Pollutants from
Boilers
, December 8, 2010, at http://www.4cleanair.org/Documents/
EVALUATIONOFCIBOSTUDYFINAL12082010.pdf. Hereinafter, the “NACAA study.”
47 NACAA study, p. 5.
48 Ibid.
Congressional Research Service
13

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Furthermore, according to NACAA, EPA data show a large number of sources that are capable of
burning a variety of fuels. This creates “significant opportunities to reduce capital costs by
shifting the mix of fuels combusted,” according to NACAA, a lower-cost compliance strategy
that both CIBO and EPA did not consider. NACAA provided numerous examples of how such
strategies might be used by different types of boilers.49
The AF&PA Study
The American Forest and Paper Association (AF&PA) also sponsored a report on the rule’s
potential impacts, although the AF&PA report combined the potential impacts of the Boiler
MACT with those of three other pending air rules.50 The AF&PA study concluded that the Boiler
MACT alone would cost pulp and paper mills $4.6 billion in capital costs, plus $560 million in
operating costs, and would place 30 mills with 16,888 employees at risk of closure.
Some General Thoughts on Cost and Economic Impact
In general, over the last 40 years, Clean Air Act rules have proven less expensive than both EPA
and industry estimates have projected before they were promulgated. As the EPA Administrator
noted in a September 2010 speech, after recounting examples of exaggerated projections of the
consequences of proposed rules, “… the Clean Air Act has not only reduced harmful pollution—it
has also been particularly effective at proving lobbyists wrong.”51 The National Association of
Clean Air Agencies reached the same conclusion: “Pre-regulation estimates by industry sources
have historically overstated the cost of compliance with proposed regulations, often by substantial
amounts.” The NACAA report cites the Clean Air Act’s acid rain program, catalytic converters on
automobiles, the removal of lead additives from gasoline, the replacement of ozone-depleting
substances in air conditioners, and the impacts of the 1997 National Ambient Air Quality
Standards for ozone and particulate matter as examples of major regulatory programs whose costs
were overestimated.52
More to the point, as noted earlier, EPA legally cannot take cost or economic impact into
consideration in identifying the MACT floor, and the standards for 11 of the 15 identified
subcategories are based on the MACT floors for each.
But the agency can distinguish among classes, types, and sizes of sources within categories or
subcategories. This could lead to less stringent standards if the agency identifies additional
subcategories from within the boiler universe. This is one way in which the agency revised the
proposed rule before promulgating it. In response to a September 24, 2010, letter sent by 41
Senators, the Administrator stated that it was the agency’s intent to “… focus on making the
regulatory subcategories appropriately reflect industrial variation in the real world, and on
aligning the standards in each subcategory with the performance that real-word conditions prove

49 Ibid., pp. 9-10.
50 Fisher International Inc. for the American Forest and Paper Association, “Economic Impact of Pending Air
Regulations on the U.S. Pulp and Paper Industry,” August 2010, 8 p. Hereinafter, “AF&PA Study.”
51 Administrator Lisa P. Jackson, “Remarks on the 40th Anniversary of the Clean Air Act, As Prepared,” September 14,
2010, at http://yosemite.epa.gov/opa/admpress.nsf/8d49f7ad4bbcf4ef852573590040b7f6/
7769a6b1f0a5bc9a8525779e005ade13!OpenDocument.
52 NACAA study, p. 4.
Congressional Research Service
14

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

are already achievable.”53 The Administrator explained that this would be possible because the
affected companies and institutions had provided additional information in response to the EPA
proposal.
The agency should not expect an entirely free hand in setting additional subcategories (or
perhaps, even in promulgating standards based on the many subcategories it originally proposed).
In comments on the proposed rule, a group of four environmental organizations that frequently
have challenged EPA regulations, objected to EPA’s proposed subcategorization, calling it
“unlawful, arbitrary, and unsupported by the record.”54 They note that while the act provides that
the Administrator may distinguish among classes, types, and sizes of sources within a category,
such subcategorization is not required: “… the plain text of the Act demonstrates that Congress
intended EPA to creates [sic] categories and subcategories as a step towards establishing
emissions standards, … not as part of a scheme to provide incentives for existing sources to avoid
standards. And yet, that is the effect of EPA’s subcategorization scheme.”55
NACAA’s comments also argue that EPA set several of the standards at levels that were more
lenient than the MACT floor. The NACAA study provided details on two of these, the carbon
monoxide and mercury standards for coal-fired boilers.56
Should EPA Have Set Health-Based Standards
Under Section 112(d)(4)?

According to EPA, “... emissions data collected during development of the proposed rule show
that hydrogen chloride [HCl] emissions represent the predominant HAP emitted by industrial,
commercial, and institutional (ICI) boilers, accounting for 61 percent of the total HAP
emissions.”57 Given the importance of HCl emissions, one of the key issues in considering EPA’s
proposal was whether the agency should have exercised its authority to set standards for HCl and
other acid gases under Section 112(d)(4), which gives the Administrator flexibility to set
standards less stringent than MACT for HAPs that have a health threshold (i.e., substances that
are not harmful to people exposed to levels below some threshold).
In developing and promulgating other regulations, including the vacated 2004 MACT standard
for boilers, EPA established that HCl has a health threshold, that it is not classified as a human
carcinogen, and that there is limited health risk associated with HCl emissions from discrete units.
Nevertheless, in the June 2010 proposal, the Administrator decided not to exercise her discretion
to set less stringent standards for HCl emissions for several reasons, including:

53 Letter of EPA Administrator Lisa P. Jackson to Senator Mary L. Landrieu, September 28, 2010, p. 3.
54 Comments of Clean Air Task Force, Earthjustice, Natural Resources Defense Council, and the Sierra Club on
National Emission Standards for Major Sources: Industrial, Commercial, and Institutional Boilers and Process Heaters:
Proposed Rule, Docket ID No. EPA-HQ-OAR-2002-0058, August 23, 2010, p. 3.
55 Ibid.
56 NACAA study, pp. 5-8.
57 Boiler MACT proposal, 75 Federal Register 32011, June 4, 2010.
Congressional Research Service
15

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

1. the agency lacked information on the peak short-term emissions of HCl from
boilers and thus could not determine whether acute exposures will pose health
concerns;
2. HCl emissions from boilers mix with other emissions that are respiratory
irritants, and EPA has no studies explicitly addressing the toxicity of these
mixtures;
3. in considering whether to exercise her discretion under Section 112(d)(4), the
Administrator must determine that a health-based standard in lieu of a MACT
will not result in adverse environmental effects. HCl gas forms an acidic solution
in the atmosphere and could exacerbate the impacts of acid deposition from
sulfur and nitrogen oxides;
4. the agency had limited information on facility-specific emissions that it would
need to set a health-based standard;
5. the agency would have needed to decide whether it would be appropriate to set
112(d)(4) standards for each acid gas emitted by boilers, or a single standard as a
surrogate for them all; and
6. as proposed (and as promulgated), the MACT standard would result in significant
reductions in emissions of other pollutants, most notably sulfur dioxide,
particulate matter, other acid gases, mercury, and other metals. These reductions
would provide substantial public health benefits that would be foregone if the
agency set a less stringent standard.58
Whether the agency should have set standards for HCl under Section 112(d)(4) was one of the
key points raised in comments, including those submitted by 41 Senators in a letter to the
Administrator, on September 24, and by 105 Members of the House in a letter submitted August
2. As the Senate letter stated:
To help reduce the burden of the rule in a manner that does not compromise public health
and safety, … we ask that you carefully consider the extensive record that supported the
Agency’s determination to include health-based emissions limitations for hydrogen chloride
and manganese in the previous Boiler MACT rulemaking that was set aside by the reviewing
court on wholly unrelated grounds.59
In the final rule, the agency did not change its mind on the use of Section 112(d)(4), but it did
significantly change the hydrogen chloride standards, presumably based on new data supplied by
affected entities. The changes are summarized in Table 3.

58 For more information on the 112(d)(4) issue, see the discussion in the Boiler MACT proposal at 75 Federal Register
32030-32033, June 4, 2010.
59 Letter of Senator Mary L. Landrieu at al. to EPA Administrator Lisa Jackson, September 24, 2010, p. 2.
Congressional Research Service
16

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Table 3. Proposed and Promulgated Emission Limits for HCl Emissions from
Existing Boilers
(lbs./million Btu)
Subcategory
Proposed Limit
Promulgated Limit
% Difference
Coal stoker
0.02
0.035
+75%
Coal fluidized bed
0.02
0.035
+75%
Pulverized coal
0.02
0.035
+75%
Biomass stoker / other
0.006
0.035
+483%
Biomass fluidized bed
0.006
0.035
+483%
Biomass Dutch oven /
0.006 0.035 +483%
suspension burner
Biomass fuel cel s
0.006
0.035
+483%
Biomass suspension /
0.006 0.035 +483%
grate
Liquid 0.0009
0.00033
-63%
Gas 2 (other process
0.000003 0.0017 +56,566%
gases)
Non-continental liquid
0.0009
0.00033
-63%
Source: U.S. EPA, Proposed and Final Boiler MACT Federal Register notices, Table 1.
Notes: Although “biomass suspension / grate” was not a proposed subcategory, al biomass burners had
proposed HCl limits of 0.006 lbs./million Btu.
Smaller (Area) Sources
Smaller boilers (those at facilities that emit less than 10 tons of an individual HAP and less than
25 tons of all HAPS combined) face regulations as well, but for the most part the Clean Air Act
allows them to meet a less stringent standard, termed “Generally Available Control Technology”
(GACT). A separate rule setting standards for these “area sources” was promulgated the same
day as the MACT standards.60 EPA is also reconsidering this rule, but it did not extend the rule’s
effective date.
The area source rule distinguishes boilers that have a heat input capacity of 10 million Btu per
hour or more from those that are smaller. The smaller units make up the overwhelming majority
of the units covered by the area source rule; they would be subject to GACT. Under GACT, these
units would not be required to meet emission limits. Rather, they would be required to meet a
work practice standard by performing a boiler tune-up every two years. According to EPA, “By
improving the combustion efficiency of the boiler, fuel usage can be reduced and losses from
combustion imperfections can be minimized. Minimizing and optimizing fuel use will reduce
emissions of mercury and all other air toxics.”61

60 The area source rule is at 76 Federal Register 15554, March 21, 2011.
61 U.S. EPA, “Final Air Toxics Standards for Industrial, Commercial, and Institutional Boilers at Area Source
Facilities,” Fact Sheet, p. 2, at http://www.epa.gov/airquality/combustion/docs/20110221aboilersfs.pdf.
Congressional Research Service
17

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Some units under the area source rule would be subject to MACT for at least some pollutants.
These are the coal-fired units that have a heat input capacity of 10 million Btu per hour or more,
but are at facilities that don’t meet the major source definition because, even counting their boiler
emissions, they emit less than 10 tons of any individual HAP and less than 25 tons of any
combination of them. According to the agency, these larger boiler units at area sources would
need to meet standards based on MACT for some of the pollutants they emit: “The final standards
for existing and new coal-fired boilers at area sources are based on MACT for mercury and CO,
and on GACT for PM. The final standards for existing and new biomass boilers and existing and
new oil-fired boilers at area sources are based on GACT.”62
The area source rule would affect approximately 183,000 existing boilers powered by oil,
biomass, and coal, located at 92,000 facilities. It would impose annualized costs of $535 million
in 2014, according to EPA’s Regulatory Impact Analysis. After considering fuel savings from
efficiency improvements that would result from the tune-ups required by the rule, the estimated
annualized cost is reduced to $487 million.63 EPA also estimates that 6,779 new boilers will be
constructed at area sources in the next three years; net costs for meeting the area source standards
at these facilities are estimated by EPA to be $48 million annually. After accounting for fuel
savings from improvements in combustion efficiency, EPA estimates that new sources will
experience cost savings of $3.6 million annually rather than incurring compliance costs.64 EPA’s
estimate of costs at area source boilers is summarized in Table 4.
Table 4. Annualized Compliance Costs for Area Source Boiler Rule
Estimated
Annualized Cost
Number of
Annualized Cost
per Unit
Source Subcategory
Affected Units
($ million)
($ thousand)
Existing Units
Coal 3,710
$37
$10
Biomass
10,958
$24
$2
Oil
168,003
$374
$2
New Units
Coal 155
$0.4
$3
Biomass
200
$2.6
$13
Oil 6,424
$45
$7
Facility Energy Audit
Al 189,450
$52
$0.3
Source: U.S. EPA, Regulatory Impact Analysis, Table 3-2.
Notes: Does not include fuel savings from improving combustion efficiency. Per unit cost was calculated by CRS.
Gas-fired boilers, of which EPA estimates there are 1.3 million, would not be affected by the area
source rule.
Because the costs of compliance are substantially less than for the MACT rule, the area source
rule has not been particularly controversial.

62 Ibid., p. 5. The actual standards can be found at 76 Federal Register 15559, Table 1. These standards do not address
most of the pollutants covered by the major source MACT. Compared to the major source MACT, they are also less
stringent for the pollutants that they do address.
63 Regulatory Impact Analysis, p. 3-6.
64 76 Federal Register 15579, March 21, 2011.
Congressional Research Service
18

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

Related Rules on Solid Waste Incineration
The Boiler MACT and the Area Source Rule were two of four related rules that EPA promulgated
the same day. The other two rules are projected to have less impact than the Boiler MACT, but
they address the issues that were at the heart of the court decision that overturned and remanded
the boiler rules in 2007. As noted earlier, the U.S. Court of Appeals for the D.C. Circuit, in
Natural Resources Defense Council v. EPA, found that EPA had wrongly excluded many
industrial boilers from the definition of solid waste incinerators, which have more stringent
emissions requirements under Section 129 of the Clean Air Act.65 Thus, the agency promulgated
two rules on March 21 dealing with incineration of solid waste in boilers, in addition to the Boiler
MACT and Area Source rules: first, a rule on the Identification of Non-Hazardous Secondary
Materials that Are Solid Waste,66 and second, a rule that would set emission standards for
Commercial/Industrial Solid Waste Incinerators (the “CISWI Rule”). The first rule identifies solid
waste, and the second sets emission standards for the facilities that burn it.
Defining Solid Waste
The purpose of this rule is to clarify which materials are considered solid waste when burned in
combustion units and which are not. To be considered solid waste, the basic criterion is whether
the material has been discarded. Discarded materials are generally considered solid waste; other
materials are not. But some discarded materials can still avoid classification as waste if they meet
a number of what EPA calls “legitimacy criteria”:
1. if the material is managed as a valuable commodity;
2. if the material has meaningful heating value (or, for a material considered an
ingredient, if it makes a useful contribution to the production or manufacturing
process); and
3. if the material contains contaminants at levels comparable to or lower than
traditional fuels or ingredients.
Non-hazardous secondary materials that meet the legitimacy criteria, such as the following,
would not be considered solid waste under the rule :
• material used as a fuel that remains within the control of the generator (whether
at the site of generation or another site the generator has control over);
• scrap tires removed from vehicles and managed under established tire collection
programs;
• resinated wood residuals, provided they have not been discarded and are used as
fuel either by the generator or outside the generator’s control;
• material used as an ingredient in a manufacturing process (whether by the
generator or a third party);

65 Natural Resources Defense Council v. EPA, 489 F. 3d 1250, 1257–61 (DC Cir. 2007).
66 76 Federal Register 15456, March 21, 2011.
Congressional Research Service
19

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

• material that has been sufficiently processed to produce a fuel or ingredient
product; and
• material that has been determined through a case-by-case petition process to not
have been discarded and to be indistinguishable in all relevant aspects from a fuel
product.67
Controversy over this rule centered on how EPA would interpret these criteria for certain
recovered materials that are commonly used as fuel, particularly “off-spec” used oil and whole
scrap tires. The proposed rule did not specifically identify these materials as solid waste.
However, in the preamble to the proposal, EPA did identify these materials as solid waste even
when they are used as fuel.
Used Oil
EPA defines used oil as either complying with limits for contaminants of concern (“on-spec”) or
not (“off-spec”). On-spec oil is exempt from waste management regulations, because the
contaminants in it are either at the same concentration or at a lower concentration than in virgin
refined fuel oil. Off-spec used oil, on the other hand, even if it is managed within the control of
the generator, contains contaminants at levels that are not comparable to traditional fuels, and
thus would not be considered a non-waste fuel under the legitimacy criteria described above.
Under previous regulations promulgated under the Resource Conservation and Recovery Act
(RCRA, 40 CFR part 279), off-spec used oil could be burned in used-oil-fired space heaters,
provided that, in EPA’s words:
(1) The heater burns only used oil that the owner or operator generates or used oil received
from household do-it-yourself used oil generators; (2) the heater is designed to have a
maximum capacity of not more than 0.5 million Btu per hour; and (3) the combustion gases
from the heater are vented to the ambient air. The RCRA used oil regulations base this
provision on a finding that uncontrolled emissions from these sources do not pose a
significant threat to human health and the environment. (Used Oil Final Rule, 50 FR 49194
(November 29, 1985).) However, consistent with our determination that off-spec used oil be
considered a solid waste when burned as a fuel, we believe that off-spec used oil managed
within the control of the generator would not qualify for the generator controlled exclusion
when burned in a used oil fired-space heater, since contaminant levels are not comparable to
traditional fuels. Therefore, we are proposing that off-spec used oil combusted at a unit that
is within the control of the generator would be solid waste.68
If the used oil is classified as solid waste, then the space heater would have to meet the “CISWI”
incinerator standards described below, which no space heater is likely to meet.
Most used oil is considered on-spec, but many of those who commented on the proposal argued
that unless there is a general exclusion such as that written into the existing RCRA regulations, it
would be necessary to test the oil and determine that it is on-spec before burning it in a space
heater. Doing so would be costly and impractical. Thus, the proposed regulations, in the minds of

67 Ibid., pp. 15459-15460.
68 U.S. EPA, “Identification of Non-Hazardous Secondary Materials that Are Solid Waste; Proposed Rule,” 75 Federal
Register
31865, June 4, 2010.
Congressional Research Service
20

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

many commenters, would have had the practical effect of banning the use of waste oil in space
heaters.
The proposal also appeared to contradict the existing RCRA regulations, but did not specifically
repeal them. As a result, the Code of Federal Regulations might have contained two conflicting
sets of rules applicable to the combustion of used oil.
The final rule clarifies these issues:
EPA is specifically clarifying in this final rule that used oil combusted in an oil-fired space
heater that meets the provisions of 40 CFR 279.23 [i.e., the existing RCRA regulations] need
not be tested to establish whether or not such oil is on or off-spec. This includes used oil
generated by small facilities such as auto repair shops and machine shops that have such
units, and used oil-generated by homeowners who change their own oil (referred to as ‘‘do-
it-yourself’’ or ‘‘DIY’’ oil) that are burned in such units. This is because the CISWI
regulations promulgated elsewhere in the Federal Register today do not establish emissions
limits for such units, and therefore the concerns of the commenters that such units would
have to comply with CAA Section 129 standards have been addressed for this population of
combustion units.69
Scrap Tires
The proposed rule also would have imposed new restrictions on the use of scrap tires as fuel:
… whole used tires (even if collected from tire dealerships and automotive shops and
overseen by a state tire collection oversight program) are initially abandoned and thus meet
the plain meaning of discard. As a result, whole used tires that are not processed into a
legitimate fuel or ingredient (e.g., shredded/chipped with steel belts removed) would be
considered a solid waste. We acknowledge that whole tires can be legitimately burned as
fuel, but because they have been discarded, whole tires would be considered solid wastes and
subject to the CAA section 129 requirements unless processed into a non-waste fuel
product.70
This would have been a change from current policy and would have affected the use of scrap tires
as fuel.
EPA reversed itself in the final rule:
After careful consideration of the comments and all the material in the rulemaking record,
including documents cited in the ANPRM [Advance Notice of Proposed Rulemaking] and
the preamble to the proposed rule, the Agency agrees that a system where scrap tires are
removed from vehicles and are collected and managed under the oversight of established tire
collection programs are not “discarded in the first instance.” Such tires (including both whole
tires and tires that have been shredded—with or without metal removal) are non-waste when
used as a fuel in combustion units. These programs ensure that the tires are not discarded en
route to the combustor for use as fuel and are handled as a valuable commodity as required in
the legitimacy criterion in today’s rule at § 241.3(d)(1)(i).71

69 76 Federal Register 15502, March 21, 2011.
70 U.S. EPA, “Identification of Non-Hazardous Secondary Materials that Are Solid Waste; Proposed Rule,” 75 Federal
Register
31864, June 4, 2010.
71 76 Federal Register 15491-15492, March 21, 2010 [footnotes omitted].
Congressional Research Service
21

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

The CISWI Rule
The Commercial/Industrial Solid Waste Incinerator (CISWI) rule promulgated on March 21
would set emission standards for commercial and industrial facilities that burn materials
determined to be solid waste (i.e., materials that do not meet the above criteria).72 CISWI’s
emission standards are required to be set under Section 129 of the Clean Air Act, which has more
stringent requirements than Section 112. In addition to the five groups of pollutants addressed by
the Boiler MACT, the CISWI rule sets emission limits for lead, cadmium, sulfur dioxide, and
nitrogen oxides. Section 129 also makes no distinction between major sources and area sources,
thus setting the more stringent MACT standards for smaller facilities.
EPA has identified five subcategories of CISWI facilities: incinerators, energy recovery units for
solids, energy recovery units for liquids and gases, waste burning kilns, and small remote
incinerators—a total of 88 existing facilities. The agency estimates that three of the units already
meet the promulgated emission limits. The total nationwide annualized costs of compliance for
the remaining 85 units were estimated to be $232 million. EPA estimates the benefits of the final
rule, including the avoidance of 40-100 premature deaths annually, at $360 million-$780
million.73
Controversy over the proposed CISWI standards focused on a category called “burn-off ovens.”
Burn-off ovens, as defined by EPA in the proposed rule, are units that combust residual materials
off racks, parts, drums or hooks so that those items can be re-used in various production
processes. Operators of such facilities stated that there are more than 15,000 such units (EPA had
identified 36), and they maintained that the units should not be characterized as incinerators, but
should be considered boilers, subject to either the Boiler MACT or the Area Source rule. In the
final CISWI rule, EPA concluded that it didn’t have sufficient data for burn-off ovens, and
removed them and several other types of units from the incinerator definition.
In order to consider additional data and public comments, on May 16, EPA announced that it
would delay the effective date of the CISWI rule and would accept additional comment until July
15, 2011.
Conclusion
EPA’s Boiler MACT remains controversial. The promulgated version of the rule is much less
stringent than the rule as first proposed, and it may change further as a result of a reconsideration
process now underway. EPA accepted additional data and information until July 15, 2011, and the
agency is now considering changes. The agency expects to propose any changes by October 31,
and make final decisions by April 30, 2012.
Members of Congress have been active participants in EPA’s public comment process (more than
100 Members of the House and more than 40 Senators wrote EPA regarding the proposed rule)

72 U.S. EPA, “Standards of Performance for New Stationary Sources and Emission Guidelines for Existing Sources:
Commercial and Industrial Solid Waste Incineration Units; Final Rule,” 76 Federal Register 15704, March 21, 2011.
73 U.S. EPA, “Final Amendments to New Source Performance Standards and Emission Guidelines for Commercial and
Industrial Solid Waste Incineration Units,” Fact Sheet, February 21, 2011, pp. 2-3, at http://www.epa.gov/airquality/
combustion/docs/20110221ciswifs.pdf.
Congressional Research Service
22

.
EPA’s Boiler MACT: Controlling Emissions of Hazardous Air Pollutants

and they remain interested in these rulemakings. The House Energy and Commerce Committee’s
Subcommittee on Energy and Power held a hearing on the promulgated rules, April 15, 2011.
Bills have been introduced in both the House and Senate to change the Clean Air Act
requirements for these rules and the deadlines for implementation.
On May 11, 2011, Representative Carter introduced resolutions to disapprove both the MACT
and area source rules under the Congressional Review Act (H.J.Res. 58 and H.J.Res. 59). Under
the Congressional Review Act, if Congress approves a joint resolution of disapproval and the
resolution becomes law, the rule cannot take effect or continue in effect. Also, the agency may not
reissue either that rule or any substantially similar one, except under authority of a subsequently
enacted law.
Two other bills (H.R. 2250 and S. 1392) would provide additional time for implementation of
standards and would change key aspects of the Section 112 requirements as they apply to boilers
and CISWI units. They would prohibit modification or repromulgation of the Boiler MACT and
related rules until at least 15 months after the date of the bills’ enactment; and they would require
EPA to set a compliance date no earlier than five years after the date of promulgation. At a
minimum, this would give the affected units almost three years of additional time to comply with
MACT standards. The bills would remove the requirements that currently apply in the absence of
EPA regulation—what are called the “MACT hammers”: under current law, permits issued in the
absence of MACT regulations are required to include MACT emission limits determined on a
case-by-case basis. The bills define certain sources currently considered as “new” to be “existing”
sources, which would be subject to less stringent requirements. And they set less stringent
requirements for the standards themselves, requiring EPA to choose the “least burdensome”
regulatory alternative, and requiring that standards can be met “consistently and concurrently
with emission standards for all other air pollutants,” which presumably would prohibit EPA’s use
of the “pollutant by pollutant” approach that it used in setting the currently promulgated
standards.

Author Contact Information

James E. McCarthy

Specialist in Environmental Policy
jmccarthy@crs.loc.gov, 7-7225


Congressional Research Service
23