Air Pollution: EPA Should Improve Oversight of Emissions	 
Reporting by Large Facilities (06-APR-01, GAO-01-46).		 
								 
The Environmental Protection Agency (EPA) performs limited	 
oversight of states' processes for verifying the accuracy of	 
large industrial facilities' emissions reports. EPA's data show  
that most emissions determinations from large sources are based  
on generic emissions factors. Although EPA allows facilities to  
estimate their emissions in this manner, EPA officials generally 
consider direct methods to be more reliable. The accuracy of	 
these reports is important because they influence (1) the	 
financing of states' regulatory programs through fees and (2) the
development of emissions inventories, which, in turn, help	 
regulators to develop control strategies and establish permit	 
limits. Furthermore, steps taken to assess the accuracy of these 
reports, such as more thoroughly reviewing the supporting	 
information, could improve compliance with Clean Air Act	 
requirements. For example, a more thorough review of the	 
information underlying a facility's emission reports, or a more  
systematic comparison of these reports over time, could identify 
increased emissions. Such indications could, in turn, trigger a  
review of compliance with New Source Review requirements, an area
in which EPA found widespread noncompliance in four industries.  
In the four states that GAO reviewed, the states that had	 
performed the most detailed reviews found widespread		 
inaccuracies. Although it is taking steps to improve its overall 
compliance monitoring strategy, EPA does not plan to evaluate	 
states' processes for verifying emissions reports from large	 
facilities.							 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-01-46						        
    ACCNO:   A00762						        
  TITLE:     Air Pollution: EPA Should Improve Oversight of Emissions 
             Reporting by Large Facilities                                    
     DATE:   04/06/2001 
  SUBJECT:   Air pollution control				 
	     Environmental law					 
	     Environmental policies				 
	     Federal/state relations				 
	     Industrial facilities				 
	     Industrial pollution				 
	     Noncompliance					 
	     Pollution monitoring				 
	     Reporting requirements				 
	     Kentucky						 
	     North Carolina					 
	     Pennsylvania					 
	     Virginia						 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-01-46
     
GAO- 01- 46

Report to the Ranking Minority Member, Committee on Government Reform, House
of Representatives

United States General Accounting Office

GAO

April 2001 AIR POLLUTION EPA Should Improve Oversight of Emissions Reporting
by Large Facilities

Page i GAO- 01- 46 Air Emissions Reporting Letter 1

Appendix I Development and Reliability of Air Emissions Factors 25

Appendix II Comments From the Environmental Protection Agency 27

Tables

Table 1: Violations Found Through EPA?s Intensive Investigations in Two
Industries 11 Table 2: Emissions Factor Ratings 26

Figure

Figure 1: Large Stationary Sources as a Source of Selected Air Pollutants,
1998 6

Abbreviations

EPA Environmental Protection Agency GAO General Accounting Office Contents

Page 1 GAO- 01- 46 Air Emissions Reporting

April 6, 2001 The Honorable Henry A. Waxman Ranking Minority Member
Committee on Government Reform House of Representatives

Dear Mr. Waxman: In 1999, 62 million Americans lived in areas that did not
meet federal air quality standards. Under the Clean Air Act, the U. S.
Environmental Protection Agency (EPA) is responsible for overseeing the
states? air quality programs and for ensuring that the states appropriately
implement the act?s requirements. Title V of the Clean Air Act Amendments of
1990 requires nearly 20, 000 major stationary sources (such as factories
that emit more than specified levels of pollutants) to obtain and comply
with the conditions of a permit that consolidates all applicable air
pollution control requirements. 1 Generally, the permits contain provisions
that limit emissions of pollutants as well as requirements for record
keeping and monitoring. In addition, most major sources must report their
aggregate annual emissions to their state air quality agency and pay fees
based partly or entirely on their level of emissions. 2 While these reports
are not intended to demonstrate compliance with Clean Air Act requirements,
their accuracy is important for determining the amount of permit fees that
major sources must pay and for developing state and national emission
inventories, which are used to develop air pollution control strategies.

1 Under title V of the Clean Air Act, sources emitting pollutants above
certain thresholds are classified as ?major sources? and must obtain title V
operating permits. According to an EPA air quality official, as of January
2001, about 19, 880 major sources had already received or could expect to
receive a title V permit. The official told us that almost all facilities
that had not yet received their title V permit already had some form of
state- issued air quality permit. Sources that emit pollutants below major
source thresholds are called ?minor sources? and do not have to obtain a
title V permit. Some minor sources, called ?synthetic minors,? have the
potential to emit pollutants at major source levels but choose to limit
their operations and emit below these thresholds. For simplicity, we refer
to major sources and synthetic minors as ?large sources,? and distinguish
between the two where necessary. 2 EPA has authorized all 50 states and 63
local governmental units to act as clean air

regulatory agencies and issue title V permits. For simplicity, in this
report, we refer to all of them as states.

United States General Accounting Office Washington, DC 20548

Page 2 GAO- 01- 46 Air Emissions Reporting

Concerned about the air quality consequences of emissions that exceed
allowable levels, you asked us to provide information on (1) the steps that
EPA and state regulators take to verify that large sources comply with their
title V or state permit and the extent of compliance found; (2) the steps
that regulators take to verify the accuracy of emissions reports submitted
by large industrial sources and the extent of errors found; and (3) the
steps that EPA is taking, if any, to improve its oversight of these
processes. To address these questions, we performed audit work at EPA?s
headquarters, 2 of EPA?s 10 regional offices, and the air quality offices of
four states (two states within each of two of EPA?s regions). 3 We selected
these regions and states because of their large numbers of stationary
sources and their levels of recent regulatory activity. Within each region,
we accompanied state air quality program officials on inspections of large
facilities. For information on the extent of compliance and the extent of
errors in emissions reports, we used data for fiscal years 1998 and 1999,
the most recent years for which data were available at the time of our
review.

EPA requires major sources to certify at least annually that they comply
with all applicable clean air regulations and to provide information on
deviations from permit conditions. Furthermore, EPA and states routinely
inspect large sources to monitor their compliance with their title V or
state permit. Federal and state regulators performed about 17, 800 routine
inspections each year in fiscal years 1998 and 1999 and found that about 88
to 89 percent of the facilities complied with their permit. However,
according to an EPA air enforcement official, routine inspections do not
necessarily identify instances in which the facilities have made physical or
operating changes that could increase emissions and require revising their
permits. Recognizing this shortcoming, EPA has undertaken intensive
investigations, targeted at selected facilities in four industries: electric
utilities, petroleum refining, pulp and paper mills, and wood products.
These investigations have found widespread noncompliance with certain air
pollution control requirements. For example, EPA found that 76 percent of
wood products facilities that it investigated had made operational changes
without revising their permits. Moreover, EPA?s investigations in the
refinery industry found widespread underreporting of emissions from leaking
valves and other equipment.

3 The states were Pennsylvania and Virginia (EPA Region III) and Kentucky
and North Carolina (EPA Region IV). Results in Brief

Page 3 GAO- 01- 46 Air Emissions Reporting

All four of the states included in our review generally check the
calculations in emissions reports submitted by major sources, but the states
vary in the extent to which they seek to verify the accuracy of the
supporting information. In preparing emissions reports, which are not
intended to demonstrate compliance with Clean Air Act requirements, large
facilities rely primarily on estimates and extrapolation instead of directly
measuring their pollutant emissions. To estimate their annual emissions of
each pollutant, most facilities use industry- and pollutantspecific
emissions factors-- average emissions rates that have been calculated for
various combinations of industrial processes, raw materials, and types of
pollution control devices. EPA?s data show that, nationally, emissions
factors are used for about 80 percent of emissions determinations. Although
these states do not maintain data on the number and severity of the problems
identified with the emissions reports, officials in three of the four states
told us they often found reports questionable enough to require contacting
the facility. For example, officials in one state estimated that one- third
or more of the reports had to be resubmitted.

EPA has taken three steps to improve its oversight of facilities? compliance
with the Clean Air Act but does not plan to enhance its oversight of the
states? processes for reviewing large facilities? emissions reports. First,
EPA is training and encouraging personnel in its regional offices and the
states to conduct intensive investigations. Second, EPA is revising its
strategy for monitoring facilities? compliance with the Clean Air Act?s
requirements. While not yet completed, the preliminary strategy calls for an
increase in direct measurements of pollutant emissions. EPA is currently
negotiating its proposed strategy with state agency officials and plans to
issue the document in April 2001. Third, in September 1998, EPA issued
guidance encouraging large facilities to use more reliable methods, such as
continuous emissions monitors and source tests, to support certifications of
compliance with operating permits. This guidance, however, was set aside by
an April 2000 court decision, which found that EPA did not comply with
necessary rule- making procedures. EPA did not appeal the decision and is
currently evaluating other regulatory options that would achieve the same
objective. EPA performs limited oversight of states? efforts to verify large
facilities? emissions reports. Although it has encouraged its regional
offices to evaluate states? emissions fee programs for major sources, EPA
has not asked them to evaluate the processes used to verify emissions
reports. We are making a recommendation to the EPA Administrator designed to
improve EPA?s oversight of the states? review of emissions reports by
evaluating the adequacy of these reviews and, if necessary, strengthening
them.

Page 4 GAO- 01- 46 Air Emissions Reporting

Under the Clean Air Act, EPA has set air quality standards for six principal
pollutants- the so- called ?criteria pollutants?- to protect public health.
These are carbon monoxide, lead, nitrogen dioxide, particulate matter, and
sulfur dioxide, as well as ground- level ozone. The latter is not directly
emitted by stationary sources but is formed by the airborne reaction of heat
and sunlight with nitrogen oxides and volatile organic compounds, which are
emitted by the sources. For the criteria pollutants, EPA sets limits- called
?national ambient air quality standards?- on the acceptable levels in the
ambient air anywhere in the United States. These limits are intended to
ensure that all Americans have the same basic health and environmental
protections.

In addition to the six principal pollutants, EPA regulates 188 hazardous
pollutants known as ?air toxics.? People exposed to toxic air pollution-
which can be highly localized near industrial sources- have an increased
chance of getting cancer and experiencing other serious health effects.
Under the Clean Air Act, EPA specifies a limit on emissions of air toxics
that is based on the achievable control technology. EPA is also to evaluate
the residual risk to human health after the adoption of the technology
standards and, if necessary, establish more stringent health- based
standards. 4

Industrial facilities emit over 100 million tons of pollutants into the air
of the United States each year. Many of these sources are large stationary
sources, such as chemical manufacturers and electric utilities. For the
purpose of the Clean Air Act, these facilities fall into two categories:
major sources and minor sources. Generally, under the Clean Air Act, major
sources are facilities that annually emit or have the potential to annually
emit (1) 10 or more tons of any one toxic air pollutant, (2) 25 tons of any
combination of toxic air pollutants, or (3) 100 or more tons of any of the
six principal pollutants. 5 Minor sources are those facilities that emit
below these thresholds. Some minor sources are referred to as ?synthetic
minors.? While not defined in the act, synthetic minors are facilities that,

4 As we reported in April 2000, EPA had met 117 of 221 Clean Air Act
requirements relating to air toxics, although 102 of these requirements were
met late. (See Air Pollution: Status of Implementation and Issues of the
Clean Air Act Amendments of 1990 [GAO/ RCED- 00- 72;

Apr. 17, 2000]). 5 The definition of a major source also depends on the air
quality in its geographic area. For

example, sources emitting as little as 10 tons of volatile organic compounds
a year may be classified as major sources in areas with the poorest air
quality. Background

Large Stationary Sources Emit Significant Amounts of Regulated Pollutants

Page 5 GAO- 01- 46 Air Emissions Reporting

according to EPA, have the potential to emit pollutants at the same levels
as major sources but choose to limit their operations, thus reducing their
emissions to levels below those of major sources. According to an EPA air
quality official, as of January 2001, about 19,880 major sources had already
received or expected to receive title V permits. According to EPA officials,
the agency does not maintain information on the total number of minor
sources.

As shown in figure 1, large stationary sources accounted for varying
portions of the nation?s emissions of certain regulated air pollutants in
1998. They accounted for 86 percent of the sulfur dioxide emissions and 38
percent of nitrogen oxide emissions in 1998. On the other hand, large
stationary sources emitted only 4 percent of particulate matter, 6 percent
of carbon monoxide, and 12 percent of volatile organic compound emissions in
1998. 6 In addition, major sources accounted for nearly 24 percent of the
emissions of toxic air pollutants in 1996 (the most recent year for which
EPA has data). 7

6 These data were obtained from EPA?s National Emission Trends database.
According to an EPA official, for purposes of the database, large stationary
sources include most major sources as well as some smaller ones. This
database does not contain information on lead emissions. Data on particulate
matter emissions are for particles equal to or less than 10 microns in size.
7 These data were obtained from EPA?s 1996 National Toxics Inventory.

Page 6 GAO- 01- 46 Air Emissions Reporting

Figure 1: Large Stationary Sources as a Source of Selected Air Pollutants,
1998

Source: EPA.

In monitoring air quality, regulators look at the levels of pollutant
emissions- the amounts being emitted into the air- as well as measured
concentrations- the levels detected in ambient air. According to EPA?s data,
national emissions of carbon monoxide, lead, sulfur dioxide, and volatile
organic compounds decreased from 1990 through 1998, while emissions of
nitrogen oxides and particulate matter increased. According to data from
EPA?s national network of air quality monitors, the concentrations of all
six pollutants, on an aggregate national basis, decreased from 1990 through
1999. These improvements ranged from a 4- percent decrease in ground- level
ozone to a 60- percent decrease in lead concentrations. Despite these
improvements, in 1999, approximately 23 percent of Americans (62 million)
lived in areas that did not meet federal ambient air quality standards for
at least one of the six principal pollutants.

Subject to EPA?s oversight, state agencies are responsible for administering
air quality programs. The Clean Air Act Amendments of 1990 provide for these
agencies to issue title V permits to major stationary sources within their
jurisdictions. According to an EPA official who manages the title V permit
program, all 113 state and local agencies have federally approved title V
permit programs. EPA and States Share

Responsibility for Administering Clean Air Program

Particulate matter

Carbon monoxide

Volatile organic compounds

Nitrogen oxides

Large stationary sources Mobile and other sources Sulfur

dioxide

4% 96% 6% 94%

12% 88% 38% 62%

86% 14%

Page 7 GAO- 01- 46 Air Emissions Reporting

Title V permits contain emissions- related, record- keeping, and monitoring
requirements. Emissions- related requirements can include limitations on
emissions per year or per hour, or on total production levels. In addition,
some facilities have no limits on the amount of total pollution they emit
but, instead, have efficiency standards that require them to remove a
certain proportion of the pollution they generate by using specific
pollution control equipment. For example, a facility might be required to
operate control equipment that removes 90 percent of the pollution generated
by a particular process or production line.

Record- keeping and monitoring requirements specify activities that
facilities must perform to demonstrate compliance with their title V or
state permit. For example, a permit may require a facility to maintain
information on its operating conditions, such as the amount of raw materials
used or outputs produced.

According to EPA officials responsible for overseeing states? permit
programs, these agencies finance their title V permit programs (including
permits for and inspections of major sources) with fees paid by regulated
facilities. An EPA official responsible for overseeing states? title V
permit programs stated that, as of December 2000, the national average fee
paid by major sources was $28 per ton of pollution emitted. According to EPA
officials responsible for developing emissions inventories and air quality
policies, in addition to serving as the basis for fees, the emissions
reports are used in developing emissions inventories. These officials also
explained that the inventories inform regulatory decision- making at the
local, state, and federal levels. For example, regulators use them to
develop control strategies and to establish permit requirements.

Each year, EPA and states perform thousands of inspections at large
facilities to monitor their compliance with requirements contained in title
V and state permits. However, because of the limitations of routine
inspections and suspected noncompliance, EPA initiated intensive
investigations within four industries. These intensive investigations found
indications of significant noncompliance with provisions of the Clean Air
Act that routine inspections do not generally address.

While inspections and intensive investigations are used to monitor
compliance, title V permits hold company officials at major sources
accountable for compliance by requiring those officials to submit at least
once a year a statement certifying as to their compliance status with all
applicable clean air requirements. Major sources must also report every 6
Routine Inspections

Generally Found Compliance, but Intensive Investigations Found Widespread
Noncompliance

Page 8 GAO- 01- 46 Air Emissions Reporting

months on all deviations from the permit?s requirements. A company official
must attest to the truth, accuracy, and completeness of the statements. 8

Routine inspections, which focus on compliance with the permits, address
emissions- related, record- keeping, and monitoring requirements. According
to EPA?s guidance, a routine inspection must include the following three
components:

 Observing visible emissions. This inspection technique can indicate
whether the measures to control certain pollutants emitted from a facility
are being properly operated and maintained.

 Observing and recording data on control devices and operating conditions.
This enables inspectors to compare observed operating conditions (such as
temperature or production levels) with those specified in a facility?s
permit.

 Reviewing records and log books on the facility?s operations. These
records provide information on a facility?s operating conditions during
times when inspectors are not present at the facility.

Nationally, EPA and state officials conducted routine and other inspections
at 17,812 large sources in 1999, according to EPA?s data. Of these, 88
percent (15, 618 facilities) were found to be in compliance with their
permit, while 12 percent (2,194 facilities) did not fully comply. EPA?s data
for 1998 show similar national results: 89 percent (15, 805 facilities) were
in compliance, and 11 percent (1, 997 facilities) were not. EPA does not
maintain data on the extent to which facilities found in noncompliance
directly violated emissions- related requirements rather than recordkeeping
or monitoring requirements. An EPA Air Enforcement Division official told us
that administrative and record- keeping violations sometimes conceal
emissions- related violations. For example, the official said that if a
facility has a limit on its emissions per unit of production but fails to
maintain production records, an inspector might not be able to determine if
excess emissions had occurred. In such a case, the inspector might cite the
facility for noncompliance with a record- keeping provision.

8 EPA?s compliance data administrator told us that the agency could not
provide information on the number of cases where major sources self- report
noncompliance or deviations from permit requirements because state agencies
are not required to provide EPA with this information. Routine Inspections
May

Not Detect Emissions Violations

Page 9 GAO- 01- 46 Air Emissions Reporting

The six inspections we observed (at a chemical manufacturer, a diesel engine
manufacturer, a fiberboard manufacturer, a municipal waste incinerator, an
absorbent material manufacturer, and a steel mini mill) illustrate the
limitations of routine inspections. The permits for the six facilities
contained various limits on emissions and production levels. All six
facilities had production- based and hourly emissions limits for at least
some of their production lines. The permits for five of the six facilities
imposed aggregate annual limits for some or all of the pollutants that the
facilities emit. For example, the permit for the diesel engine manufacturer
specified an annual limit on benzene (a toxic air pollutant) emissions from
on- site generators and diesel engine test booths, while the permit for the
absorbent material manufacturer had limits on hourly emissions, total annual
emissions, and annual production levels. In contrast, the fiberboard
manufacturer had no annual emissions limits.

The inspectors whom we accompanied checked for visible emissions, reviewed
facility records, and checked equipment at all six facilities. At the waste
incinerator, the inspector observed a source test conducted in accordance
with a state- approved sampling plan. 9 The inspectors determined that while
four facilities were in compliance, the other two facilities were out of
compliance.

 At the steel mini mill, the inspector noted that the facility

 lacked records on the readings of visible pollutants;

 lacked records on the functioning of various control devices for a 22month
period;

 lacked information on the sulfur content in certain raw materials used;
and

 did not maintain a certain emissions capture system, which caused visible
emissions to leak from gaps and holes in the building.

The inspector also noted an apparent disparity between the amount of nitrous
oxides emitted annually from one furnace. For permit purposes in 1998, the
facility estimated the amount was 31 tons. But on the basis of a 1999 source
test, the amount was 94 tons.

9 Source tests (also called ?stack tests?) involve the use of procedures to
measure emissions for a short period of time (usually a few hours or more).
Data derived from these tests are used to determine compliance with
applicable emissions limits.

Page 10 GAO- 01- 46 Air Emissions Reporting

 At the fiberboard manufacturing facility, the inspector noted a number of
problems indicating the improper operation and maintenance of control
equipment, including

 one improperly sealed vent and

 excess emissions from 7 of the facility?s 20 emissions control filters. In
addition to observing six inspections, we reviewed state regulatory records
for 23 large facilities to identify other methods used by regulators to
identify excess emissions. For example, we noted that, at a metal-
canmanufacturing facility, a state inspector was concerned about the amount
of volatile organic compounds in the coating used to seal the cans. He
arranged for samples of the coating to be analyzed at a state laboratory and
allowed the facility to contract for its own analysis. Both the state lab
and the company?s lab found that the facility was exceeding its permit limit
for the volatile organic compound content of the coating by about 10
percent.

In recent years, EPA has performed intensive investigations in four
industries and identified widespread noncompliance. In three industries-
electric utilities, pulp and paper mills, and wood products- these
investigations focused primarily on compliance with New Source Review
requirements. Under the Clean Air Act, facilities must obtain a New Source
Review permit for new construction or major modifications that increase a
facility?s emissions of certain regulated air pollutants. According to an
EPA air enforcement official, routine inspections do not necessarily
identify instances in which facilities have made physical or operating
changes that could increase emissions and require them to revise their
existing permits or obtain New Source Review permits. In the fourth
industry- petroleum refining- EPA investigated compliance with both New
Source Review requirements and regulations that require the monitoring of
?fugitive emissions? leaking from valves, pumps, and other equipment. In the
pulp and paper and wood products industries, EPA found widespread
noncompliance. In the electric utility and petroleum refining industries,
many of the companies investigated agreed to take remedial actions on the
basis of EPA?s preliminary findings rather than actual findings of
noncompliance. However, because EPA targeted facilities that were determined
to be most likely to have violated their permits, the results of these
intensive investigations may not represent conditions at other facilities.
Intensive Investigations

Found Widespread Noncompliance

Page 11 GAO- 01- 46 Air Emissions Reporting

To identify industries on which EPA should focus its intensive
investigations, agency staff analyzed industry- by- industry information on
production levels, profits, and other factors that could help them identify
industries with facilities that had increased production but may not have
applied for new construction permits. Next, the EPA staff considered which
facilities within those industries to focus on. They gathered and analyzed
industry journals and other publicly available information about companies,
as well as information in state agency files.

The intensive investigations generally consisted of visiting the facility
for 3 days to identify equipment, determine when it was installed, and
evaluate the history of physical or other changes in the use of that
equipment. EPA staff also obtained financial data for the facility to
identify expenditures that may indicate an increase in production capacity.
Afterwards, they spent from several months to a year analyzing this
information to determine whether the facility violated Clean Air Act
requirements.

In the petroleum refinery industry, EPA also performed investigations to
determine if facilities accurately reported the number of emissions leaks
from valves, pumps, compressors, and other equipment. Federal regulations
require refineries to monitor equipment for leaks on a routine basis and to
fix leaking equipment. The failure to identify and fix these leaks can
result in excess fugitive emissions of volatile organic compounds and other
hazardous air pollutants.

In the pulp and paper and wood products industries, EPA found widespread
noncompliance. As shown in table 1, of the 96 facilities where EPA has
completed investigations, 75 (about 78 percent) were not in compliance.

Table 1: Violations Found Through EPA?s Intensive Investigations in Two
Industries Industry Number of facilities

investigated Number of facilities not in compliance Proportion not in

compliance

Pulp and paper 12 11 92 percent Wood products 84 64 76 percent

Total 96 75 78 percent Source: EPA.

Page 12 GAO- 01- 46 Air Emissions Reporting

Common types of violations included the failure to

 install pollution control devices (both industries),

 obtain New Source Review permits required by the Clean Air Act (both
industries),

 meet emissions limits (pulp and paper), and

 perform required testing (pulp and paper). According to an EPA Air
Enforcement Division official, EPA took a different approach in the electric
utility and refining industries. The official told us that EPA initiated
investigations of specific facilities and then met with company officials to
present its preliminary findings. This official also told us that, in many
cases, the companies agreed to take such actions as installing pollution
control equipment at one or more of their facilities on the basis of EPA?s
preliminary findings rather than risk an actual finding of noncompliance.

As of February 2001, EPA had reached three agreements covering 20 facilities
in the electric utility industry and three agreements covering 19 facilities
in the petroleum refining industry. According to an EPA air enforcement
official, all of these facilities agreed to pay fines and install the
pollution control equipment they would have been required to install had EPA
formally found them in noncompliance. In return, according to the official,
EPA agreed to resolve possible past violations at the facilities. EPA
estimates that a recent settlement with one electric utility company will
require the company to install control equipment and take other steps to
reduce its emissions of sulfur dioxide and nitrogen oxides by 400, 000 and
100,000 tons per year, respectively.

At 17 refineries investigated for leaks of volatile organic compounds, EPA
found a larger proportion of leaking emissions points and a larger volume of
leaks than the companies reported. Specifically, whereas the companies
reported finding leaks in 1. 3 percent of the potential emissions points,
EPA?s investigators found leaks in 5 percent. EPA estimated that annual
fugitive emissions from the 17 refineries investigated could be more than
6,000 tons per year greater than previously believed. By extrapolating these
findings, EPA estimated that refineries may be emitting an additional 40,000
tons of volatile organic compounds each year because leaks are not properly
identified and repaired promptly.

Page 13 GAO- 01- 46 Air Emissions Reporting

According to EPA officials who oversee state permit programs, because most
title V permit programs assess emissions fees, at least in part, on the
basis of the total tonnage of pollutants emitted by major sources, most
major sources are required to submit annual reports listing their total
emissions. In the four states included in our review, all major sources are
required to report annually on their total emissions in the previous
calendar year; in addition, the states require synthetic minors to report
periodically on their total emissions. While many of the largest emitters,
such as coal- powered electric utilities, must continuously measure their
emissions of certain pollutants, most facilities rely primarily on estimates
or extrapolations from source tests to determine their emissions. 10 All
four states in our study generally reviewed the facilities? emissions
reports for arithmetic errors but varied in the extent to which they
verified the accuracy of data on which the facilities based their
calculations. While the states did not track the extent to which they
discovered errors, officials in one state that performed detailed reviews
estimated that between onethird and one- half of all reports had to be
resubmitted.

According to EPA officials, the method used by a facility in determining its
emissions depends on a number of factors, including the type of facility and
the raw materials used. 11 Methods range from direct measures of emissions
to estimates based on emissions factors, as outlined below:

 Under the 1990 Clean Air Act Amendments, certain types of facilities must
directly measure their emissions using continuous- emissions- monitoring
systems (hereafter called ?monitors?). Monitors constantly measure
pollutants released by a single point, such as a smokestack within a
facility. For example, EPA requires most coal- burning electric utilities
and certain other types of facilities to use monitors to measure their
emissions of certain pollutants. State regulators also have discretion to
require other air pollution sources to use monitors. For example, a
Pennsylvania state agency official told us that Pennsylvania has required
the use of 445

10 According to an EPA Air Enforcement Division official, emissions factors
cannot be used to demonstrate compliance with emissions limits but can be
used in developing emissions reports or determining whether a facility emits
above or below major source thresholds. The official also said that
facilities cannot avoid major source requirements if it is subsequently
proved that they emit pollutants in excess of these thresholds. 11 For
simplicity, we use the term ?determination? to describe all methods that
facilities use

in preparing their emissions reports, including direct measurement,
emissions factors, and extrapolations. Four States? Reviews

of Emissions Reports Varied

Large Facilities Rely Primarily on Indirect Methods to Determine Their Level
of Emissions

Page 14 GAO- 01- 46 Air Emissions Reporting

monitors in addition to 327 monitors required by federal regulations. EPA
officials consider monitors to be the most reliable method for determining
annual emissions.

 According to an EPA official, extrapolations from the short- term data
derived from source tests can, in some cases, be used to estimate longterm
emissions from the tested facility or from similar facilities. EPA officials
told us that short- term source tests are considered less reliable than
monitors for determining long- term emissions. The limitations of source
tests include their short duration and facilities? common practice of
performing the tests under optimal conditions, such as shortly after
purchasing or servicing control equipment.

 Emissions factors are broad averages of the emissions of pollutants that
can be expected, given the processes and/ or pollution control equipment
generally used in an industry. Facilities using emissions factors estimate
their volume of emissions by multiplying their activity rate by the
appropriate factor. For example, a facility that wants to estimate its
carbon monoxide emissions from burning distillate oil in an industrial
boiler would multiply the emission factor for that process (5 pounds of
carbon monoxide for each thousand gallons of oil burned) by the quantity of
fuel consumed. If the facility burned 3,000 gallons a day, its estimated
carbon monoxide emissions would total 15 pounds a day. Because emissions
factors represent average emissions, the level of emissions from some
sources using them may be higher than the factor, while others may be lower.
(App. I provides additional information on the development and reliability
of emissions factors.)

According to EPA and state agency officials, facilities use emissions
factors to make most emissions determinations for the purpose of emissions
reports. EPA?s nationwide data on emissions determinations made by both
large and small facilities show that indirect methods that do not involve
site- specific direct measurement were used in about 96 percent of all
determinations, while direct measures, such as monitors and source tests,
were used for about 4 percent of all determinations. Of the 96 percent
involving indirect methods, emissions factors accounted for about 80 percent
and other methods accounted for 16 percent.

Similarly, most facilities in the states we visited relied on indirect
methods. For example, about 63 percent of the emissions determinations from
large industrial facilities located in North Carolina relied on EPA?s rated
emissions factors. In Virginia, about 71 percent of the large facilities
used emissions factors to determine emissions from at least one of their

Page 15 GAO- 01- 46 Air Emissions Reporting

emission sources, while about 10 percent relied on monitors for at least one
emission source.

The percentage of emissions determinations made by a certain method may not
equal the percentage of the total emissions that were quantified by that
method. EPA does not track the quantities of emissions determined by each
quantification method, but emissions data for electric utilities that must
use monitors show that such facilities account for a large percentage of the
emissions of certain pollutants. For example, while monitors are used for
less than 5 percent of all emissions determinations nationwide, EPA?s data
show that in 1998, electric utilities required to use monitors to measure
their emissions of nitrogen oxides and sulfur dioxide accounted for about 24
percent of total national nitrogen oxide emissions and about 65 percent of
total national sulfur dioxide emissions.

Each of the four states included in our study assesses major sources? fees,
at least in part, on the basis of the number of tons of pollution they emit.
Each of the states requires similar information from facilities. The
facilities typically provide detailed information on emissions from
production lines or processes that are regulated in their permits. For
example, one state requires facilities to provide, among other things,
information on the raw materials they use, their operating schedule, the
sulfur and energy content of fuels, the efficiency of pollution control
devices, the method used to calculate emissions, and the tons of pollution
emitted. In addition to providing the information described above, each
state requires a company official to certify, under penalty of law, the
report?s truth, accuracy, and completeness.

Each of the four states uses the information contained in the reports to
independently calculate each facility?s emissions and, in the three states
where facilities provided estimates of total emissions, to compare the
agency?s calculations of total emissions with those provided by the
facility. (One state does not require facilities to estimate total
emissions; instead, according to a state official, agency personnel use the
information provided by the facilities to perform the calculations
themselves.) All four states routinely compared the reports with those
submitted in previous years to identify noteworthy changes that might
indicate inaccurate reporting.

While we found similarities in the states? procedures for verifying the
emissions reports in the four states, we also found variations, as shown
below. States? Methods of

Verifying Emissions Reports Varied

Page 16 GAO- 01- 46 Air Emissions Reporting

 In two states, the field inspector who performs the compliance inspection
of a facility typically also reviews that facility?s emissions report. An
official in one of these states explained that having the inspector with the
greatest understanding of each facility review the report maximizes the
agency?s ability to identify questionable data. In contrast, a third state
assigns all reports for a certain facility type to one inspector; thus, the
inspector reviewing the emissions report for a facility may not be the
person who performed compliance inspections at that facility. In the fourth
state, the personnel responsible for developing the state?s emissions
inventory, rather than those who inspect facilities, review the reports.

 The state agencies also vary in the extent to which they seek to verify
the data that facilities submit on the material they used or the removal
efficiency of their control equipment. For example, officials in one state
told us that they typically check for the use of appropriate emissions
factors and pollution control efficiencies and review previous inspection
reports and other relevant documents to ensure that facilities account for
all emissions points. Alternatively, regulators in another state told us
that they simply rely on facilities to provide accurate data.

None of the four states maintain data on the type or number of inaccuracies
found during their efforts to verify emissions reports. Regulators in all
four states told us that those responsible for reviewing the reports contact
the facility directly to resolve any problems or inaccuracies identified
through the verification process. After resolving any questions about the
report, the facility revises its statement as necessary. For example,
officials in one state told us that they consider problems with the reports
to be inadvertent, and that the inspector performing the review works with
the facility to resolve the differences.

Because state agency officials were unable to provide comprehensive data on
the type or number of inaccuracies found, we asked them to estimate the
proportion of all the reports submitted that had significant problems. One
state provided a statewide estimate. Officials in this state, which
performed detailed reviews of the reports, said that one- third to one- half
of all its reports required corrections and resubmittal of the report by the
facility. Officials in another state said that the agency?s regional offices
verified the reports and that the thoroughness of the reviews varied across
the regional offices. The regional office performing the most detailed
reviews estimated that 80 percent or more of the reports had problems that
required additional consultation with the facility, while the regional
office performing the least thorough reviews found such problems with 10
percent of the reports. An official in the third state told us that there
are

Page 17 GAO- 01- 46 Air Emissions Reporting

few problems or missing data in reports from facilities that had reported
previously, but that almost all reports from facilities reporting for the
first or second time required follow- up because of incomplete data.
Officials in the fourth state said that they relied on facilities to provide
accurate data.

EPA has undertaken or is planning three initiatives to improve its oversight
of compliance with the Clean Air Act but does not plan to enhance its
oversight of state processes for verifying the accuracy of emissions
reports. With respect to compliance, first, EPA developed and issued
guidance to state regulators on the types of information that major sources
must maintain to demonstrate their compliance with permits. Second, EPA is
revising its compliance- monitoring strategy, which will grant states
greater flexibility in their approaches to inspections and will encourage
regulators to obtain more site- specific emissions data through the
increased use of direct measurements via source tests. Third, EPA is
training regional office staff and states to conduct intensive
investigations. With respect to the emissions reports, EPA officials in
headquarters and the two regions we visited all told us that EPA relies on
the states to review these reports. At the same time, EPA has encouraged its
regions to audit state programs for calculating emissions fees, which often
depend in part on the amounts of emissions, but has not asked its regions
specifically to evaluate states? processes for verifying emissions reports.
The two EPA regional offices we visited perform little oversight of their
states? verification processes.

EPA?s first initiative, in September 1998, was issuing guidance on the type
of information that major sources must periodically gather and maintain to
demonstrate their compliance with applicable air regulations. EPA sought to
clarify its policies on self- monitoring by facilities and to encourage
state agencies to consistently interpret these policies. According to EPA,
the definition of ?adequate monitoring? had been subject to interpretation,
and the level and type of monitoring that state authorities required were
not consistent.

EPA?s guidance document states that facilities must maintain reliable,
timely, and representative data on the status of their compliance. The
document further states that the use of an emissions factor does not
constitute adequate monitoring unless the factor was developed directly from
the unit in question. In addition, the guidance encourages state authorities
to require the use of monitors and indirect monitoring derived from periodic
source tests. EPA Plans to Improve

Oversight of Compliance but Not Verification of Emissions Reports

EPA Is Working to Improve Data Quality and Facility Monitoring

Page 18 GAO- 01- 46 Air Emissions Reporting

The implementation of EPA?s guidance has been suspended because of an April
14, 2000, ruling by the U. S. Court of Appeals. 12 The court held that, in
issuing the guidance, EPA, in effect, amended its monitoring regulation
without complying with the necessary rule- making procedures. EPA did not
appeal the decision and is currently evaluating other regulatory options to
meet the same objectives.

EPA?s second initiative, in March 2000, was issuing a draft national policy
for state regulators to use in ensuring compliance with the act. This policy
was developed in response to two reports that found problems with EPA?s air
enforcement program. EPA?s Inspector General reported in 1998 that no one
within the enforcement program was responsible for the oversight and
implementation of the agency?s Clean Air Act compliance- monitoring program.
13 The report described inconsistent implementation and disregard for agency
directives as diminishing the effectiveness of the air enforcement program.
The report also described cases where inspections conducted by state
regulators did not meet EPA?s definition of a ?routine

inspection? or were documented poorly. In addition, a 1999 study
commissioned by EPA found that most EPA regional offices did not adhere to
the agency?s compliance- monitoring strategy. 14

EPA?s draft policy states that it would, among other things, provide
regulators with increased flexibility in the types of inspections they
conduct and require sources with no better means of determining their
emissions rates to conduct source tests. EPA?s compliance data administrator
told us that the draft policy would also require states to provide EPA with
information on annual compliance certifications and semiannual compliance-
monitoring reports that are submitted by major sources. EPA?s Air
Enforcement Division officials said that they were working with
representatives of state agencies to revise the draft and that the agencies
have expressed concerns over the document?s provisions for the increased use
of source tests. 15 EPA has revised the document to

12 Appalachian Power Co. v. EPA, 208 F. 3d 1015 (D. C. Cir. 2000). 13 See
Consolidated Report on OECA?s Oversight of Regional and State Air
Enforcement Programs, Office of Inspector General, EPA (EIGAE7- 03- 0045-
8100244, Sept. 25, 1998).

14 See A Review of the Compliance Monitoring Strategy, Perrin- Quarles
Associates, prepared for EPA?s Office of Enforcement and Compliance
Assurance (July 26, 1999). 15 EPA?s Office of Inspector General recently
evaluated EPA?s oversight of state source

testing programs. (See Report of EPA?s Oversight of State Stack Testing
Programs, Office of Inspector General, EPA [2000- P- 00019, Sept. 11,
2000]).

Page 19 GAO- 01- 46 Air Emissions Reporting

address these concerns, but the issue remains unresolved. In addition, EPA?s
Air Enforcement Division officials said that they have not determined
whether the final strategy will take the form of guidance (as originally
proposed) or an administrative rule. They told us they plan to issue the
document in April 2001.

Finally, EPA?s Air Enforcement Division officials noted that EPA is
encouraging personnel in its regional offices and the states to conduct
intensive investigations to ensure compliance with New Source Review
requirements.

An EPA official who oversees state permit programs stated that the agency
has not taken or proposed actions specifically intended to improve the
accuracy of emissions reports from major sources, although one initiative
has the potential to provide information on states? review processes. In
1998, EPA encouraged its regional offices to review state permit authorities
to determine whether, among other things, they were correctly implementing
their fee programs and collecting sufficient fees to cover the costs of
administering their title V permit programs. EPA developed and distributed
to the regions an audit protocol for evaluating state programs. Although the
audit protocol does not ask regions to determine whether permit programs
have adequate controls in place to verify emissions reports, it does ask
them to examine the documentation of how the annual fees are determined and
to audit pollution sources? bills, which most permit authorities- including
those in all four of the states where we worked- based, at least in part, on
each facility?s reported level of total emissions. An EPA official who
oversees state permit programs told us that regions have full discretion in
determining whether they use the audit protocol in evaluating the permit
programs.

An official in the Air Protection Division of EPA?s Philadelphia office
stated that the regional office has used the audit protocol to review three
permit programs in that region. One of these reviews found that the state
audited did not verify emissions reports. Officials in the Environmental
Accountability Division of EPA?s Atlanta office told us that they were not
using the audit protocol in reviewing programs in their region or seeking to
evaluate the processes in place for verifying emissions reports.

While EPA does not plan to evaluate states? processes for verifying
emissions reports, it does check the quality of emissions data submitted by
states for developing emissions inventories. This includes checking for data
errors that could have affected emissions values, as well as, in some EPA
Does Not Plan to

Evaluate States? Processes for Verifying Emissions Reports

Page 20 GAO- 01- 46 Air Emissions Reporting

cases, comparing estimates with those submitted in previous years and with
those from other facilities in the same industry. In addition, EPA posts
facility- specific emissions data on the Internet for review by outside
parties.

EPA performs limited oversight of states? processes for verifying the
accuracy of emissions reports submitted by major sources. EPA?s data show
that most emissions determinations are based on generic emissions factors.
While EPA allows facilities to estimate their emissions in this manner, EPA
officials generally consider direct methods to be more reliable. The
accuracy of these reports is important because they influence (1) the
financing of states? regulatory programs through fees and (2) the
development of emissions inventories, which, in turn, assist regulators in
developing control strategies and establishing permit limits.

Furthermore, steps taken to assess the accuracy of these reports- such as
more thoroughly reviewing the supporting information- could provide benefits
in terms of compliance with Clean Air Act requirements. For example, a more
thorough review of the information underlying a facility?s emissions reports
or a more systematic comparison of these reports over a period of time could
identify indications of increased emissions. Such indications could, in
turn, trigger a review of compliance with New Source Review requirements, an
area where EPA found widespread noncompliance in four industries.

In the four states included in our review, the approaches taken to verify
the accuracy of the reports varied significantly. The state that performed
the most detailed reviews found widespread inaccuracies. However, EPA?s
oversight of these processes is limited; the agency had audited only three
permit authorities in the two EPA regions we visited and found that one of
the three authorities had no process in place for verifying the accuracy of
the emissions reports. While taking steps to improve its overall compliance-
monitoring strategy, EPA does not plan to evaluate state processes for
verifying emissions reports from large facilities.

To help ensure the accuracy of large facilities? emissions reports, we
recommend that the Administrator of EPA evaluate states? programs to
determine whether they have adequate mechanisms in place for verifying the
accuracy of emissions reports. If the results of these reviews identify
inadequacies, the Administrator should work with the states to improve
Conclusion

Recommendation for Executive Action

Page 21 GAO- 01- 46 Air Emissions Reporting

their processes in order to provide reasonable assurance that facility
reports are subject to thorough review.

We provided EPA with a draft of this report for review and comment and
received a letter from the Acting Assistant Administrator for Air and
Radiation. (App. II contains the text of his letter, along with our detailed
responses; in addition, EPA provided us with several clarifications, which
we incorporated where appropriate.)

The Acting Assistant Administrator questioned the intent of our
recommendation, stating that

if the intent is to improve the accuracy of emissions reports to ensure the
sufficiency of fees that states collect to support their title V permit
programs, EPA disagrees and believes the recommendation is unnecessary
because the states can simply raise the fee rate (the fee per ton of
emissions) if fee revenues prove insufficient;

if the intent is to improve the emissions inventories used in state planning
and in developing national inventories, EPA concurs; and

if the intent is to improve compliance with applicable permit requirements,
EPA disagrees because emissions reports are not intended to determine
compliance with permit requirements.

The intent of our recommendation, as stated in the draft report, is to help
ensure the accuracy of emissions reports because of the role that the
reports actually play or can play in all three areas: (1) setting fees to
cover the costs of state programs; (2) developing state and national
inventories and, concomitantly, strategies for further controlling
emissions; and (3) potentially alerting state regulators to emissions levels
that suggest noncompliance with operating permits or other air quality
requirements.

We agree that states facing a shortfall in fee revenues could simply
increase the rate applied to all sources to raise aggregate fee revenue, but
we do not agree that the accuracy of emissions reports used for fees is a
secondary concern. Increasing the fees levied on facilities that accurately
report their emissions as well as on those that underreport (who would
continue to pay proportionately less than warranted on the basis of their
relative contribution to total emissions) could lead to inequitable results.
While states have latitude in their approach to collecting fees, most of
them rely, at least in part, on each facility's level of reported emissions
in Agency Comments

Page 22 GAO- 01- 46 Air Emissions Reporting

calculating fees. Thus, a facility that reports more emissions will
generally pay more in fees. Especially in the absence of state oversight,
some facilities could view this system as an incentive to underreport their
emissions and thus pay lower fees. Inconsistent or limited review of
emissions reports reduces regulators? ability to identify underreporting and
sends the signal that facilities face little chance of detection if they
choose to underreport. To the extent that any facility underreports its
emissions and thus pays less than its fair share of title V fees, other
facilities will pay more than their fair share. In the short run, this
raises questions about the equity of the fees being charged. In the long
run, this possibility- unless counteracted- could lead to more widespread
underreporting and undermine the system of emissions reporting.

In addition to helping ensure that emissions fees are collected equitably,
more thorough state reviews could also help improve emissions inventories at
the state and national levels.

Finally, more thorough reviews could help EPA and state compliance efforts.
Specifically, through its lengthy and resource- intensive investigations,
EPA identified widespread potential violations of New Source Review
requirements in all four of the industries it reviewed. We believe that more
thorough reviews of facilities' emissions reports might have provided
indications of such problems much earlier and at much less cost.
Furthermore, emissions reports often contain information not only on total
emissions but also on levels of production and raw material use. Many of the
title V permits we reviewed had provisions that limit production levels as a
surrogate for total emissions. EPA and state enforcement officials told us
that reviewing this information would help inspectors evaluate a facility's
overall compliance status. For example, as noted in our report, two of the
states assign the same field inspector responsible for inspecting the
facility for compliance to review the facility?s emissions report. This
practice enhances the potential that any discrepancies between emissions
reports and the results of compliance inspections will be detected.

To fulfill our objectives, we interviewed officials from, and reviewed
studies and other documents prepared by, EPA?s headquarters and regional
offices and four states. The EPA headquarters offices were the Office of Air
Quality Planning and Standards and the Office of Enforcement and Compliance
Assurance. The two EPA regional offices were Region III (headquartered in
Philadelphia), which generally covers the mid- Atlantic region, and Region
IV (headquartered in Atlanta), which generally covers the Southeast. The
states were Pennsylvania and Virginia Scope and

Methodology

Page 23 GAO- 01- 46 Air Emissions Reporting

in EPA?s Region III and Kentucky and North Carolina in EPA?s Region IV. The
conditions in these two regional offices and four states may not represent
the conditions in other regional offices and states.

In addition, we accompanied EPA or state officials on their routine
inspections of six facilities representing different industries- a chemical
manufacturer, a diesel engine manufacturer, a fiberboard- manufacturing
plant, a municipal waste incinerator, an absorbent material maker, and a
steel mini mill. The conditions at these six facilities may not represent
the conditions at other regulated facilities. In addition, as agreed with
your office, we do not name the facilities in our report. We did not
independently validate the data provided by EPA or the states. We conducted
our review from November 1999 through March 2001 in accordance with
generally accepted government auditing standards.

As arranged with your office, we plan no further distribution of this report
for 30 days from the date of the report unless you publicly announce its
contents earlier. At that time, we will send copies to Senator Robert C.
Smith and Senator Harry Reid in their respective capacities as Chairman and
Ranking Member, Senate Committee on Environment and Public Works;
Representative W. J. Tauzin and Representative John D. Dingell in their
respective capacities as Chairman and Ranking Minority Member, House
Committee on Energy and Commerce; Representative Dan Burton, Chairman of the
House Committee on Government Reform; other interested Members of Congress;
the Honorable Christine Todd Whitman, Administrator of EPA; the Honorable
Mitchell E. Daniels, Jr., Director of the Office of Management and Budget;
the governors of the four states we visited; and other interested parties.
We will make copies available to others upon request.

Page 24 GAO- 01- 46 Air Emissions Reporting

If you have any questions about this report, please contact me or David
Marwick at (202) 512- 3841. Key contributors to this report were Philip L.
Bartholomew, James R. Beusse, Michael Hix, Karen Keegan, and William F.
McGee.

Sincerely yours, David G. Wood Director, Natural Resources

and Environment

Appendix I: Development and Reliability of Air Emissions Factors

Page 25 GAO- 01- 46 Air Emissions Reporting

Regulators and regulated facilities use air emissions factors to estimate
emissions from a variety of sources. Emissions factors are averages of the
amount of emissions produced from a given process with given inputs, for
example, the quantity of carbon monoxide generated per unit of oil burned in
an industrial boiler.

The Environmental Protection Agency (EPA) publishes information on air
emissions factors. Regulators and industry use air emissions factors to
assist in developing emissions inventories and control strategies, and for
other purposes. For example, an EPA official told us that facilities use
emissions factors to determine whether their estimated annual emissions
place them in the major source category.

The reliability of the emissions factors varies widely. EPA rates the
reliability of emissions factors on a scale of A (excellent) to E (poor).
These ratings, in turn, reflect four underlying criteria:

 the estimated reliability of the test data used,

 the randomness of the facilities from which the data were derived,

 the variability of emissions levels across the sources tested, and

 the number of facilities for which test data are available. Thus, the
highest (A- rated) factors are those derived from high- quality data taken
from many randomly chosen facilities with low variability among the sources.
Conversely, the lowest (E- rated) factors are those derived from low-
quality test data, when doubts exist regarding the randomness of the test
facilities used, and when there is wide variability among the sources
tested.

As of October 1999, EPA had rated 12,390 factors in its compilation of
emissions factors. As shown in table 2, 20 percent of the factors were rated
?above average? or ?excellent,? while 46 percent were rated ?below

average? or ?poor.? Along with the rated factors, EPA maintains information
on approximately 4,200 unrated factors (25 percent). Appendix I: Development
and Reliability of

Air Emissions Factors

Appendix I: Development and Reliability of Air Emissions Factors

Page 26 GAO- 01- 46 Air Emissions Reporting

Table 2: Emissions Factor Ratings Emissions factors as of Oct. 1999 Letter
grade Description Number Percentage

A Excellent 1,613 10 B Above average 1,626 10 C Average 1,563 9 D Below
average 3,347 20 E Poor 4,241 26 Unrated 4, 209 25

Total 16,599 100

Source: EPA.

In its compilation of emissions factors, EPA describes problems with the use
of such factors to estimate emissions for individual facilities. Each factor
is generally assumed to represent the long- term average for all facilities
in a source category but may not reflect the variations within a category
because of different processes and control systems used. The underlying data
from which emissions factors are derived can vary by an order of magnitude
or more. For example, the emission factor for petroleum conversion at oil
refineries- 45 pounds of particulate matter per thousand barrels of
feedstock- is based on test results ranging from 7 to 150 pounds. EPA
assigned this factor a B (above average) rating. Thus, facilities? actual
emissions can, and do, vary substantially from the published factors.

Appendix II: Comments From the Environmental Protection Agency

Page 27 GAO- 01- 46 Air Emissions Reporting

Appendix II: Comments From the Environmental Protection Agency

See comment 1. Note: GAO?s comments

supplementing those in the report?s text appear at the end of this

Appendix II: Comments From the Environmental Protection Agency

Page 28 GAO- 01- 46 Air Emissions Reporting

See comment 2.

Appendix II: Comments From the Environmental Protection Agency

Page 29 GAO- 01- 46 Air Emissions Reporting

See comment 3. See comment 4. See comment 5.

Appendix II: Comments From the Environmental Protection Agency

Page 30 GAO- 01- 46 Air Emissions Reporting

The following are GAO?s comments on the Environmental Protection Agency?s
letter dated March 9, 2001.

1. EPA notes that permit authorities are required to charge fees that cover
the costs of their regulatory programs and describes the accuracy of
emissions reports used in that process as a secondary concern because state
agencies could correct a shortfall in fees that results from the
underreporting of emissions by raising the fee per ton of emissions. We
believe that the accuracy of the emissions reports is integral, initially,
to establishing an equitable fee structure and, later, to ensuring that each
regulated entity is charged only its fair share of the overall fees.
Emissions reports support both processes, and the steps we recommend are
intended to help ensure the accuracy of these emissions reports.

While EPA asserts that states should have discretion in how they assess
emissions data to establish fees, it has already initiated oversight of
these state efforts. In 1998, EPA distributed to its regional offices an
?audit protocol? that they could use to monitor whether the permit agencies
in their region had, among other things, established a proper fee structure
and were submitting appropriate bills to regulated entities. As of fall
2000, only one of the two regions we visited had chosen to use the protocol.
When EPA regional offices use the protocol to evaluate state programs, they
could implement our recommendation by amending the protocol to include
evaluating the states? processes for verifying emissions reports.

2. We recognize that EPA performs quality assurance on data provided by
states and have revised the report to acknowledge this. However, we continue
to believe that a more thorough review of these reports at the state level
could lead to more reliable local, state, and national emissions
inventories.

3. We believe that thorough reviews of the reports could improve EPA?s and
the states? ability to identify noncompliance with New Source Review
requirements and the terms of title V permits. While EPA states that these
reports are seldom, if ever, used for determining compliance with title V
permits, we believe that they contain information that could assist in doing
so. Each state we visited said that the reports contain information on
production levels and, in most cases, total emissions. Many of the title V
permits we reviewed have provisions that limit production levels.
Furthermore, the EPA and state GAO?s Comments

Appendix II: Comments From the Environmental Protection Agency

Page 31 GAO- 01- 46 Air Emissions Reporting

enforcement officials we spoke with said that reviewing the emissions
reports could help in evaluating a facility?s compliance status. Also, as
our report notes, two of the states provide for a review of a facility?s
emission report by the field inspector responsible for that facility. In
that way, the review can be performed by the individual most knowledgeable
of the facility and therefore best positioned to identify any irregularity
in the report.

4. EPA misstates our conclusion. We concluded that more thorough reviews of
the supporting information contained in emissions reports

? could provide benefits in terms of compliance with Clean Air Act
requirements,? not that ?improving the accuracy of emissions reports will
improve compliance with Clean Air Act requirements? (emphasis added). We
view this as an important distinction.

5. As stated in comment 3, we believe that reviewing the information
contained in the emissions reports could assist in identifying noncompliance
with New Source Review and title V permits.

(160504)

The first copy of each GAO report is free. Additional copies of reports are
$2 each. A check or money order should be made out to the Superintendent of
Documents. VISA and MasterCard credit cards are also accepted.

Orders for 100 or more copies to be mailed to a single address are
discounted 25 percent.

Orders by mail:

U. S. General Accounting Office P. O. Box 37050 Washington, DC 20013

Orders by visiting:

Room 1100 700 4 th St., NW (corner of 4 th and G Sts. NW) Washington, DC
20013

Orders by phone:

(202) 512- 6000 fax: (202) 512- 6061 TDD (202) 512- 2537

Each day, GAO issues a list of newly available reports and testimony. To
receive facsimile copies of the daily list or any list from the past 30
days, please call (202) 512- 6000 using a touchtone phone. A recorded menu
will provide information on how to obtain these lists.

Orders by Internet

For information on how to access GAO reports on the Internet, send an email
message with ?info? in the body to:

Info@ www. gao. gov or visit GAO?s World Wide Web home page at: http:// www.
gao. gov

Contact one:

 Web site: http:// www. gao. gov/ fraudnet/ fraudnet. htm

 E- mail: fraudnet@ gao. gov

 1- 800- 424- 5454 (automated answering system) Ordering Information

To Report Fraud, Waste, and Abuse in Federal Programs
*** End of document. ***