United States
Office of Air
Planning and
Protection
Agency
i Triangle Park, NC 27711
EPA-454/R-93-025
JULY 1993
AIR
SEPA
Summary of Public Comments and EPA Responses
on the Draft Report
The Role of Ozone Precursors in Tropospheric
Ozone Formation and Control
-------
> SUMMARY OF PUBLIC COMMENTS AND EPA RESPONSES
on the Draft Report
THE ROLE OF OZONE PRECURSORS IN TROPOSPHERIC OZONE FORMATION AND
CONTROL
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
>• Environm.*.-A
•:- 5,
/ T \t( t ~
July 1993
-------
DISCLAIMER
This report has been reviewed by the Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, and
has been approved for publication. Any mention of trade names or
commercial products is not intended to constitute endorsement or
recommendation for use.
ii
-------
SECTION 1 - GENERAL COMMENTS AND RESPONSES
Introduction and Overview
The draft Section 185B Report entitled The Role of Ozone
Precursors in Tropospheric Ozone Formation and Control underwent
a 30 day public review and comment period as announced in the
February 26, 1993 Federal Register. Comments were submitted by
Trade organizations, major corporations, the Department of Energy
and one member of the National Academy of Science's (NAS)
Committee on Tropospheric Ozone. Those who submitted comments
are identified below in Section 2.
The EPA appreciates the contributions to the Section 185B
Report provided by the commenters. Ta^fen together, they reflect
a deep understanding of the complex issues underlying the ozone
control issue, and help to foster a useful dialogue directed
toward developing the most effective precursor control
strategies.
The commenters agree in principle with the basic tenets in
the EPA draft Report, but have requested more specifics in
response to their concerns and an expansion of the EPA Report
beyond one of capsulating the NAS Report (Attachment 2). Five
general comments which represent most of the concerns raised are
presented with a subsequent response immediately below in this
Section. Section 2 provides more specific responses to
individual comments.
-------
Commentl: The commenters request information on how the
agency proposes to incorporate emerging scientific
findings and data into the regulatory process.
Response; The policy and scientific programs within the ozone
program area share similar objectives (e.g., the abatement of
ambient ozone), but often do not operate on coincident time
scales. Regulatory time scales are governed by discrete
legislative mandates which generally are on faster tracks than
research programs. For example, the November, 1994 SIP submittal
date (CAA Section 182c) is a statutory requirement dictating
completion of technical analyses in 1994. There is no date
earmarked for which "quality science" arrives. Scientific
information emerges on a continuum which is incorporated in the
regulatory process in a stepwise manner.
Policy and regulatory decisions will be based on the best
information available, consistent with the intent of Congress as
reflected in the 1990 CAA requirements.
There will always be a lag in incorporating new science into
regulatory policy. The challenge is to minimize this lag and
accelerate the incorporation of quality science and information
in the policy making process. The five steps outlined below are
being pursued by the agency to the extent resources and
legislative mandates permit. These steps are designed to improve
the scientific underpinnings of the ozone control program in a
manner that effectively incorporates new information without
regulatory delay:
1. Recognizing that the 1994 SIPs reflect the most
credible applications within existing resource and time
and scientific constraints, we must focus on future
efforts to assess accuracy of the SIPs and provide for
appropriate adjustments.
2. Establish adequate ambient monitoring programs for
evaluating the success of VOC and NOx control program
effectiveness (see comment 2).
3. Perform special intensive studies in highly polluted
areas lacking such analyses, such as the Northeast U.S.
4. Because public sector resources are lacking, areas have
developed "private/public partnerships" to optimize
available resources and pool support for advanced field
and modeling studies. Such efforts should be
encouraged and expanded.
5. Encourage justified "mid-course" corrections to the SIP
for ozone nonattainment areas.
-------
Comment 2: The EPA Report acknowledges many concerns and
outlines programs/steps/recommendations toward
resolving problems without sufficient evidence of
commitment.
Response; The Section 185B Report includes several explicit
program initiatives addressing the shortcomings noted by the NAS
Report. These include, among others, the Photochemical
Assessment Measurement Station (PAMS) program, the 1990 emission
inventory development, and the reassessment of the ozone research
program. While acknowledging that additional efforts are needed
(e.g., securing support for PAMS and initiating special intensive
studies), the EPA is constrained by resource limitations and
CAA-mandated time frames. One strategy to deal with these
constraints is discussed above within the context of mid-course
corrections and private/public partnerships. Recommendations
discussed in the Report have been or currently are being
transformed into several initiatives, examples including:
1. EPA commitment to fund PAMS has increased from $4
million to $6 million for FY 94. However, funding for
full implementation of the network will cost well over
$20 million/year after 1994. Funding commitments
beyond FY-94 levels have yet to be secured. PAMS is
noted throughout the 185B Report as a major element of
EPA's response to shortcomings identified by the NAS.
2. start of the public/private cooperative COAST Study
along the Texas Gulf Coast,
3. participation of industry, academia and States to
assist EPA in developing the Coordinated North American
Research Strategy for Tropospheric Ozone,
4. EPA participation in the industry-funded Regional Model
Evaluation Study evaluating performance and behavior of
current grid models on several model domains,
5. exploring development of private/public partnerships to
enhance the PAMS system and initiate intensive study in
the Northeast, and
6. EPA's $1 million ozone nonattainment research
initiative for FY-93.
-------
Comment 3: By failing to incorporate specific policy
recommendations, the EPA Report does not
adequately extend or complement the NAS Study.
Response: The Section 185B provision in the CAA provided explicit
direction on the scope of topics to be addressed in the Report.
Those topics are technical in nature, and the Report reflects a
synthesis of the technical perspectives and program activities
addressing tropospheric ozone control.
In other sections of the CAA, Congress directed the EPA to
develop policy and technical guidance to address related topics.
For example, Section 182(c) and Section 182(f) require
development of EPA guidance to address NOx substitution and
exemption demonstrations, respectively. These documents either
are under development or available. The EPA also has issued the
Title I General Preamble .and NOx Preamble supplement to provide
direction on NOx/ozone control policy issues. Several additional
NOX/ozone guidance documents, including direction on available
control technology (ACT), are and will be available to address
policy issues.
-------
Comment 4: The EPA Report concurs with most NAS Findings, yet
fails to justify the limited, deterministic use of
models to address the probabilistic form of the
ozone standard.
Response; The justification for using gridded photochemical
models to demonstrate attainment is based on the Section
182(c)(2)(A) CAA provision requiring such application. Air
quality models are used in this way to satisfy requirements of
the CAA. The EPA acknowledges the difficulties associated with
applying deterministic techniques to stochastic events. In
addition to dealing with a probabilistic standard, the governing
photochemical processes (i.e., meteorological and emissions
events) behave stochastically as well. Thus, are we applying the
modeling tools in an appropriate manner? The EPA UAM regulatory
application guidance (EPA, 1991) requires that the simulated
ground-level ozone fields have values of 0.12 ppm or less in
every grid cell for every selected episode day. The National
Ambient Air Quality Standard (NAAQS) for ozone is violated when
observed ozone exceeds .12 ppm on average more than once per
year. The modeling requirement appears more restrictive because
of the resource constraints inherent in complex model
applications. Generally, a handful of episode days are simulated
(the guidance requires a minimum of three) whereas the
observational set encompasses days within several ozone seasons.
Thus, a statistical approach is constrained by the number of days
simulated and a conservative, deterministic method is followed.
The deterministic approach is intended to provide a "best
estimate" given currently available data bases, models and
resources. The procedure includes provision for diagnostic
analyses and model sensitivity tests. These tests can be used to
help prioritize future data collection and model development
efforts. Use of results to help define regulatory measures
provides needed incentives for the public and private sector to
invest in improving data bases and modeling techniques.
Further, the EPA guidance allows alternative demonstration
proposals. Conceivably, those willing and capable of modeling a
set of days suitably robust for statistical attainment
demonstrations are not discouraged from doing so. On the other
hand, the EPA believes that the resource requirements to model a
statistically adequate set of days are beyond the capabilities of
most State and local agencies and therefore sets a reasonable
minimum number of required modeling days.
The explicit use of models for demonstrating attainment is
questioned by some in the modeling community. The model/data
uncertainty concerns and the form of the standard are substantive
issues. In addition, the strong focus on the attainment
demonstration might inadvertently prevent using the full
capabilities of the modeling system. Nonetheless, the model can
-------
be used to assess the relative benefits of various control
scenarios, perform exposure and cost/benefits analyses, assess
other impact on other pollutant species and averaging times, and
a multitude of other analyses. Again, unrestricted use of the
model is not discouraged; although, in being responsive to the
CAA provisions, the EPA's focus has necessarily been on the
attainment demonstration. In part because of the amended CAA's
requirements, gridded photochemical model applications have
increased ten fold or more. The technical community at large is
challenged to take advantage of this fact in performing the
analyses and attendant interpretations to broaden both the scope
of information development and knowledge transfer to decision
makers.
-------
Comment 5; Given the delay between release of the NAS Report
(December, 1991) and the Report to Congress
(1993), the EPA Report should include recent ozone
trends data from 1989 - 1992 which show increasing
progress toward attainment.
Response: The ozone data from 1989 - 1991 do indeed show reduced
levels, especially with respect to 1988. However, the three year
period from 1989 - 1991 is too short to discern any long term,
meaningful trend linking air quality improvement with emissions
reductions. Thus, it is premature to associate "progress towards
attainment" with the recent ozone data. Nevertheless, we agree
that these recent data should be included and have modified
Section 3.2 of the EPA Report accordingly. Ozone trends data
through 1991 are reported in EPA's recent Air Quality Trends
Report (EPA, 1992).
-------
SECTION 2 - INDIVIDUAL COMMENTS AND RESPONSES
Several commenters indicated favorable agreement with most
of the themes and statements in the 185B Report. For example,
the commenters agreed with the area-specific approach, emphasized
throughout the Report, for determining the most effective mix of
NOx and VOC controls. Only those comments requiring responses
are addressed in this Section. The following sequence of
comments reflects the chronological order of submissions received
by the EPA.
Distilled Spirits Council of the U.S.
Comment: The report does not endorse strongly enough a
shift away from VOC controls.
Response; The Report purposely provides no endorsement of
unilateral precursor control applied everywhere, and advocates
area-specific analyses to determine the best mix of precursor
controls. Recent evidence indicates that NOx reductions may be
of greater use than believed previously in the design of control
programs. VOC reductions also are expected to reduce ozone in
many areas. The CAA requires all areas classified moderate and
above to meet the RFP VOC reduction requirements. These
requirements allow substitution of NOx controls for VOC
reductions to meet RFP goals after 1996.
Comment; The report does not endorse the incremental
reactivity scheme.
Response: Reactivity is considered explicitly in the
modeling demonstrations, limited by the extent to which the
existing chemical mechanisms resolve individual compound reaction
characteristics. Thus, a modeling demonstration focusing on-
"more reactive" VOC emissions conceivably could demonstrate the
same benefits as a demonstration requiring more total VOC
reductions applied uniformly over all VOC classes. The major
advantage of this approach is that the modeling exercise takes
account of environmental conditions affecting the relative
reactivity of various classes of compounds.
-------
Gas Research Institute(GRI)
Comment: Corroborative monitoring techniques should be
pursued to provide independent checks of models
and control strategy decisions.
Response; Corroborative monitoring techniques are being
pursued through various avenues. These avenues include the PAMS
program; and other initiatives under the Southern Oxidant Study
(SOS), in cooperation with the GRI, internal EPA research
programs, and others. Currently, the EPA is exploring the
feasibility of applying derivatives of the Australian Smog
Production Algorithms (SPA) (Johnson and Quigley, 1989) to
ambient precursor and ozone data to develop spatial mappings of
an area's relative ozone sensitivity to NOx or VOC reductions.
Deterministic grid-models are the best tool for 1)
characterizing the physical and chemical processes underlying
ozone formation and control, and 2) simulating the effects on air
quality from future-year precursor control strategies. The
recent efforts from the SOS and SPA program are extending the use
of ambient data to provide additional indicators of an area's
ozone sensitivity to NOx or VOC control. Although the ambient
approaches are qualitative in nature, they have the potential for
improving confidence that the model's predicted precursor control
direction is correct. Furthermore, the ambient techniques may
ultimately be easy to apply and may provide a fast screening
assessment to determine whether or not a NOx- or VOC-weighted
control strategy is appropriate. The ambient approaches under
development are promising, but the conditions under which they
can be applied and limitations understood must be established.
-------
Chevron
Comment: The need for case-by-case analyses and better data
bases should be restated in the Executive Summary.
Response: The Report does not include a standard Executive
Summary, but contains overview (Section 2) and end-of-text
Summary Sections where these thoughts are expressed.
Comment: The Report oversimplifies complex issues (i.e.,
NOx and VOC chemistry).
Response: The NAS Report (Attachment 2) includes detail
beyond the EPA Report on relevant atmospheric chemistry
processes. The EPA overview in Section 2 also has been
criticized as too complex. The intent of the overview is to
present a concise set of facts as a lead-in to the controversy
underlying the NOx/VOC issue.
Comment: We agree in principle with many of the themes and
text of the Report; however, it provides no
specifics on program limitations.
Response: Currently, sufficient funding does not exist
within EPA to improve the technical underpinnings of the ozone
program to the level implied by the NAS Report. We agree with
many NAS findings, encourage activities such as private/public
partnerships, but do not commit to actions that can not be backed
by available or foreseeable resources.
Explicit recognition of program limitations often is
overlooked. This is unfortunate since knowledge of what programs
cannot accomplish facilitates future improvements. The Report
emphasizes the positive expectations from agency programs and
does not dwell on limitations. Again, the 185B Report includes
the NAS Report, which points out numerous limitations of model
and data base systems which are intrinsic to the ozone programs.
The EPA Report discusses major efforts like PAMS, ROM and UAM
modeling and emission inventory programs geared to address many
of the concerns raised. However, current routine regulatory
efforts are not designed to:
* develop research-grade episodic emission inventories
for air quality model evaluation,
* provide adequate horizontal and vertical spatial
coverage of ozone and precursor concentrations to fully
characterize boundary conditions required for model
input,
10
-------
* provide precise, low concentration NOx and NOy data
which are important for characterizing ambient
chemistry (with respect to strategy analysis) in rural
and some urban areas,
* push development of advanced modeling techniques which
address physical and chemical limitations of current
models, and
* characterize uncertainty in the model application
process.
These and other limitations are best addressed through
tightly-focused special intensive programs which can develop the
necessary resources. The EPA encourages and participates
intensive program initiatives like the Southern Oxidant Study
(SOS), Lake Michigan Ozone Study (LMOS) and the COAST study.
Although EPA's base programs lack the resources to commit to
additional intensive efforts, an approach for developing
additional initiatives is described above in Section 1.
Comment: The PAMS program is insufficient to meet stated
objectives and lacks flexibility; individual
scoping studies are needed for each area prior to
network establishment.
Response: The PAMS program will operate as a routine
nationally-based program collecting long term measurements for
the primary purpose of tracking ozone precursor reductions and
monitoring the success of emission control measures. Some, but
not all, modeling objectives will be met. Type 2 sites will be
phased in first; they are designed to provide data reflecting
representative area-wide emission trends slightly downwind of
center city locations. A priority is placed on establishing the
type 2 sites to develop a base line reference frame before
implementation of programs such as Reformulated Fuels. Scoping
studies are recommended, indeed necessary, precursors of special
'intensive studies complementing modeling. However, the PAMS
program can not satisfy the modeling objectives that intensive
programs typically address. Those gaps should be met with
special studies in areas. Given these concerns, at this time
scoping studies are not required for PAMS. The EPA is concerned
about PAMS support•beyond 1994. Full implementation of PAMS will
require more than $20 Million per year. While the base $6
Million to establish Type 2 sices in several cities appears firm
for FY-1994, continued funding at this base level and beyond for
cost:-1994 is far from solid.
11
-------
Alternative PAMS network plan submittals are allowed under
the enhanced ozone monitoring regulation. In developing the
basic PAMS requirements and in reviewing alternative Plans, EPA
will consider the following when reviewing alternative network
submittals:
* the potential for providing a robust statistical data
base for trends analysis,
* the ability to characterize atmospheric chemistry
changes due to future strategies and programs, and
* comprehensive speciated and temporal data over the
initial 1 or 2 years to justify possible subsequent
scaling down of sampling and measurement frequencies.
Comment: More cooperation with transportation agencies is
required to interface transportation models with
air quality models.
Response: Transportation demand models often are used to
derive more resolved vehicle miles travelled (VMT) and speed data
estimates in the modeling application. Unfortunately, travel
demand models generally are not designed with emphasis on meeting
air quality modeling objectives. Some progress has been made in
carbon monoxide (CO) modeling through the EPA/Federal Highway
Administration (FHA) CO intersection Modeling workgroup, and the
joint participation of EPA and FHA in the NAS intersection air
quality modeling Panel. Activities like those should be
established for ozone modeling purposes.
Comment: Comments alluding to the "high cost of intensive
studies" should be countered with comments
discussing-the "high costs of misplaced controls."
Response; Congress, through the 1990 CAA, set a time table
.for regulatory action requiring the 1994 SIPs to be based on
photochemical grid models utilizing the best available
information. Potentially large economic benefits through control
strategy optimization might be derived from comprehensive
intensive studies. This factor is one of the primary reasons
that EPA encourages such efforts. The prospect for misplaced
controls is a strong motivator for improving the monitoring and
modeling efforts.
12
-------
. Comment: The grid models like EPA's regulatory Urban
Airshed Model (UAM) are not appropriate for
demonstrating attainment.
Response: The grid models are the best available tools to
conduct ozone strategy analyses. As discussed above in Section 1
(Comment 4), models are limited in their accuracy for
demonstrating attainment, but no superior alternative exists.
The issue is the concept of demonstrating attainment, not the use
of gridded models. In addition, the use of gridded photochemical
models is required by Section 182(c) of the CAA.
Comment: Delete the ROMNET discussion to avoid further
acceptance of the Regional Oxidant Model (ROM) as
a regulatory tool.
Response: The Section 185B Report addresses the science and
technical efforts underlying ozone regulatory programs.
Accordingly, the ROMNET study plays an important role in
summarizing hypothesized regional-scale effects of NOx and VOC
reductions on ozone. Furthermore, the NAS panel made it clear
that they relied heavily on ROM results in the discussion of NOx
and VOC control effectiveness. However, only urban-scale models
such as the UAM have the required resolution of emission sources
to adequately assess NOx and VOC control effects for particular
urban areas.
Comment: The State Implementation Plan (SIP) time frames
are not compatible with time required to improve
the underlying science and data bases.
Response: Congress through the 1990 CAA has set a time table
for regulatory action via the 1994 SIPs that will be based on the
best available information. This comment is addressed above in
Section l, comment 1.
Comment: The Report should discuss phasing of controls to
reduce total exposure.
Response: The Report discusses in Section 3 use of the grid
models for conducting exposure analyses and sequential phasing of
controls. The model formulation allows for an assortment of
analyses. The phasing issue is related directly to the
discussion on the basic effects on ozone from reducing NOx and/or
VOC in Section 2 of the EPA Reoort. NCx reductions have trhe
13
-------
potential to accelerate1 the ozone formation process in regions
of elevated NOx (e.g., central urban areas and in vicinity of
major NOx sources). Thus, the greatest benefits from NOx
control are likely downwind, with relatively modest decreases, or
even.increases in ozone occurring near major sources of NOx.
Consequently, NOx controls have the potential to impart
disproportionate benefits from an exposure assessment
perspective. Proper phasing of controls potentially can lessen
accumulated population exposure over time. For example,
implementation of VOC controls initially might reduce near-field
(i.e., close to emission sources or central downtown locations)
ozone so that subsequent NOx reductions produce less cumulative
exposure. As discussed above in Section 1, the EPA guidance only
requires the model results to be analyzed in the context of
reducing peak ozone to 0.12 ppm or less by the attainment year.
Model users are not discouraged from performing analyses
addressing exposure and phased control.
1 The accelerated ozone formation is partly explained by an
initial reduction in hydroxyl radical (which drives much of the
hydrocarbon related ozone formation) removal through reaction
with nitrogen dioxide, as well a loss of NO to scavenge ozone
directly.
14
-------
Comment; Policy should be integrated into the Report.
Response; This comment is addressed above in Section 1.
Comment: How will evolving science be incorporated into the
regulatory program?
Response; This comment is addressed above in Section 1.
Comment: Private/Public partnerships should be developed to
improve the information base.
Response: Private/Public partnerships are discussed above,
in Section 1. The EPA encourages these partnerships. Although
the EPA is exploring various opportunities, serious commitments
beyond project committee participation2 and information exchange,
with the exception of the Southern Oxidant Study, have yet to be
made. The EPA is examining technical, legal and program
implementation issues related to use of public/private
partnerships to improve technical data bases. These preliminary
investigations need to be completed to define the scope and
limitations in using public/private partnerships prior to
developing a large scale commitment to this approach.
Comment: There is no justification for assuming ROM or the
UAM are credible when applied with routine data.
Response: EPA has devoted considerable time and attention to
the development of emission inventories. Current inventory
efforts are discussed at length in Section 3 of the EPA Report.
Recognizing the shortage of ambient air quality data for model
input use, the Regional Oxidant Modeling (ROM) program is
designed to provide air quality inputs to the UAM. A distinction
exists between "credible" and "accurate." The EPA believes the
.ROM and UAM applications with available data are highly credible.
The model data bases and forthcoming results are and will be
subject to large uncertainty which can translate to inaccuracies
in both model performance and prediction of control strategy
impacts. Nevertheless, the CAA requires application of models
using current data bases for the 1994 SIPs. The "credibility"
concern magnifies the importance of programs like PAMS and
special intensive studies.
: For example, the Regional Ozone Modeling for Northeast
Transport-2 (ROMNET2) External Review Workgroup, the industry
funded Regional Model Evaluation Study and industry participation
in assisting EPA in developing a long term ozone research plan.
15
-------
Comment; The development/application of corroborative
ambient techniques is encouraged.
Response: This comment is addressed above at the beginning
of Section 2.
Comment; The ozone regulatory program relies too heavily on
peak ozone as a determining metric.
Response; This comment is discussed above in Section 1 under
Comment 4 and in the Section 2 responses to Chevron.
Comment; The potential for NOx reductions with alternative
fuels is speculative.
Response; The EPA presents a balanced and brief summary of
an unclear issue. It simply states that "...there exist
possibilities with alternative fuels for reductions of both VOC
and NOx that do not exist with gasoline." Thus, the current
uncertainty about whether or not alternative fuel use will result
in NOx reductions is clearly conveyed.
The API states that CNG and alcohol vehicles normally have
NOx emissions equal to or higher than gasoline vehicles.
Actually, the existing data are highly variable, with some
vehicles showing lower NOx as well. In fact, Chrysler
Corporation recently certified a CNG van at 0.02 grams/mile NOx
with the California Air Resources Board (Geiss and Burkmyre,
1992) . In addition, a 1976 study using an experimental single-
cylinder engine demonstrated that extremely low NOx emissions are
possible with methanol when the extended lean misfire limit of
methanol is fully utilized (Koenig et al., 1976).
Comment; More coordination with other Federal Offices is
needed to address multi-faceted technical issues.
Response: A related comment regarding cross-fertilization
with transportation agencies is discussed above in the Chevron
comments. Certainly other bodies of information exist in
agencies which would reduce duplication and enhance the analysis
efforts. The model application process itself relies on a broad
assortment of demographic and economic information, in addition
to the basic emissions and aerometric data. Actually, the
coordination at the State and local agency level is probably more
important than Federal Office coordination to tailor needs to
specific areas.
16
-------
Comment: The specification of "minimum data bases" for
model applications is not addressed.
Response: EPA has devoted considerable time and attention to
the development of emission inventories. Current inventory
efforts are discussed at length in Section 3 of the EPA Report.
Recognizing the shortage of ambient air quality data for model
input use, the Regional Oxidant Modeling (ROM) program is
designed to provide air quality inputs to the UAM. However,
minimum data bases are not specified in EPA guidance nor in the
Section 185B Report because the requirement for performing timely
SIP submittals based on model demonstrations supersedes the
status/quality of available data bases.
17
-------
General Motors
Comment; It is a misconception that past programs required
only VOC controls; motor vehicle NOx reductions
should be acknowledged.
Response; The commenter is correct. Fleet average motor
vehicle NOX emissions were reduced about 55% from 1975 to 1990.
Additional reductions are anticipated in response to the mobile
source provisions of the 1990 CAA. Tier I light duty vehicle
standards will reduce new car NOx emissions from 1.0 g/mile to
0.4 g/mile beginning with 1994 models, and Tier II (if adopted)
to 0.2 g/mile beginning with 2004 models. New heavy duty diesel
engine NOx emissions were reduced from 6 to 5 g/bhp-hr in 1991,
and will be reduced to 4 g/bhp-hr in 1998. Other programs
expected to impart NOx emission benefits include enhanced/high
technology Inspection and Maintenance (with associated mechanic
training), transportation controls and onboard diagnostics.
Despite progress in motor vehicle NOx emissions, total NOx
emissions from all sources generally were not reduced (because of
increased growth and activity) over the last several years.
Comment: Evidence exists suggesting that VOC controls are
effective in reducing ozone as recent data show
decreasing ozone with decreasing ambient VOC
levels.
Response: The ozone data from 1989 - 1991 show reduced
levels, especially with respect to 1988. Ozone trends data
through 1991 are reported in EPA's recent Air Quality Trends
Report (EPA, 1992). The recent ozone data open up a wealth of
interesting speculation relevant to the NOx versus VOC reduction
issue. For instance, analysis of limited ambient VOC data and
combined VOC/NOx ratios in the Northeast indicate a downward
trend over the same period (Wolff and Korsog, 1992; Zealewsky e.t
al., 1992). These reduced ozone levels coincide with
implementation of reduced vapor pressure (RVP) reductions
starting in 1989. Lower RVP leads to a reduction in evaporative
VOC emissions from motor vehicles and gasoline storage and
transfer operations.
However, little if any confirmatory evidence supports the
view that RVP reductions effected lower ozone levels in the early
1990's. The period 1988 - 1991 is too short to discern any
meaningful cause-effect relation among emission reductions and
air qualify. The effects of meteorology are not accounted for 'in
a cursory look at recent ozone trends. The shorter the trend
period, the greater the chance that meteorology has a large
impact on ozone trends data. In particular, the summer of 1988
is on record as one of the driest and hottest, prime conditions
for producing high ozone levels. The recent summers have been
18
-------
different in character, either wetter, cooler or both, from 1988.
In fact, if 1988 were removed from the ozone trends data (page 3-
5, EPA Report), only slow progress toward attainment can be seen
over the 10 year record from 1982 - 1991. Thus, it is premature
to associate recent RVP reductions with' reduced ozone levels in a
cause-effect context.
The EPA welcomes confirmation that RVP reductions produced
measurable benefits. The current round of SIP modeling should
provide additional insight on the benefits of the RVP programs.
In addition to the environmental benefit, evidence that VOC
reductions are beneficial supports the Report's thesis lacking
endorsement of unilateral (NOx or VOC only) precursor control
applied everywhere. Further, the fact that an area responds to
VOC control does not preclude NOx control being equally or more
effective in the same area.
Comment: The Report does not extend beyond the NAS study.
Response; This issue is discussed above in Section 1 under
Comment 3.
Comment: The Report must acknowledge recent ozone trends
and improvement toward attainment.
Response: This issue is discussed above in Section 1 under
Comment 5.
Comment; The grid models are not appropriate for
demonstrating attainment.
Response: This issue is discussed above in Sections 1
(comment 4) and 2 (response to Chevron).
19
-------
Aerodyne Research (Charles Kolb of NAS Committee)
Comment: The Research Section (3.10) of the EPA Report is
filled with protestaions of good faith which
hopefully are indicative of more effective EPA
actions in the tropospheric ozone area. A main
concern is whether EPA will actually close the gap
between what is promised and future deeds in the
research area.
Response; EPA's Office of Research and Development (ORD) is
sensitive to the concerns raised about our research program and,
as a result, we convened a group of 140 senior scientists
representing a cross-section of disciplines and public and
private organizations to develop a coordinated, national research
program. The cost of such a program ranges from $600 million to
$1 billion over the next-10 years.
To ensure that the highest priority research initiatives are
undertaken, public and private sector partnerships are essential
to supply the necessary funding. Only with this type of
cooperation can the gap be closed between the work that is needed
to address the ozone problem and the research that is actually
undertaken.
Comment: EPA has not pursued research on analytical
techniques for speciated VOC and claims too much
contribution to measurement science.
Response: In the area of ambient pollutant measurement, EPA
has recognized two distinct needs: a measurement data base for
studying and assessing air pollution problems, and a need for
accuracy and precision of the requisite measurement methods.
Although there has been an emphasis on the first, significant
research and development efforts in the ambient VOC measurement
and VOC emission analysis areas have been conducted or sponsored
by EPA. Examples include: the effort conducted by the
Hydrocarbon Measurement, Technology, and Standards Group within
the EPA-sponsored Southern Oxidants Study, and additional efforts
being commenced in reaction to the stimulus raised by the NRC
Report. Recognizing that the on-going research activities are
not adequate for achieving the goals identified by the NRC
Committee, EPA is developing a coordinated North American
research strategy for tropospheric ozone, which addresses among
other critical issues the need for more definitive VOC
methodology and measurements. This strategy currently is
undergoing Agency and external review and, contingent upon
funding, the reviewer's concerns will be addressed.
20
-------
Comment; Only superficial evaluation of grid models has
been conducted by EPA, and the models require more
thorough development of supporting data bases to
provide even directionally correct ozone responses
to precursor control.
Response; The EPA has sponsored numerous model evaluations
with urban scale grid-models, especially the Urban Airshed Model
(UAM) over the past 20 years. A comprehensive model evaluation
was performed on the UAM using the RAPS-St. Louis urban field
study database during the late 1970's. This database represented
the most complete set of urban air quality measurements for ozone
and its precursors for the time. Subsequent operational model
evaluations for ozone for urban areas, such as Tulsa, Denver,
Philadelphia, and New York were performed using the routine AIRS
ozone monitoring data, sometimes supplemented with locally-
obtained data. The EPA is now participating in applying the UAM
to the most contemporary urban field study databases obtained
during the more recent SCAQS (Los Angeles) and LMOS (Chicago)
field studies. These databases contain observations not only for
ozone, but also for individual nitrogen and hydrocarbon species,
as well as comprehensive meteorological measurements, and will
enable thorough diagnostic testing of the UAM in a manner that
has been difficult to achieve previously.
The Regional Oxidant Model (ROM) also has been subjected to
on-going model evaluations from its inception in the early
1980's. The first truly regional field experiment to obtain
model evaluation databases was performed in the eastern U.S.
during the summers of 1979 and 1980 (NEROS/NECRMP project). The
ROM was diagnostically analyzed with ozone, nitrogen oxides, and
hydrocarbon data from this database. Subsequent operational
model evaluations were conducted using AIRS ozone data for
regional episodes during the summer of 1985, the base year of the
_last NAPAP emissions inventory, and 1988, the last year of
widespread chronic high ozone concentrations across the region.
Currently the ROM is undergoing diagnostic evaluation exercises
using the LMOS Lake Michigan airshed database from the summer of
1991 which contains ozone and precursor data at the surface as
well as aloft from aircraft observations. Emerging data from the
EPA-sponsored Southern Oxidants Study (1991-1995) will also be
used for diagnostic ROM evaluations in the southeast U.S.
21
-------
Coastal Corporation
Comment: Less complex, easy-to-use models need to be
developed for the regulated community.
Response: Less complex models offer the advantages of
performing multiple simulations with reduced resources. Prior to
the 1990 CAAA, the physically simplistic EKMA approach was
the most widely applied ozone regulatory model. The move toward
using grid models in the regulatory program generally is
considered a major improvement. Models like EKMA and derivatives
of grid-based models with simplified chemistry are valuable tools
for performing an array of investigative analyses. In addition,
various plume models which treat individual sources and
incorporate photochemistry have been developed. Access to an
assortment of less complex photochemical models is available
through EPA or private contractors. These "screening" approaches
have limitations relative to complex area-wide models (like the
UAM) which the EPA requires for most regulatory applications.
Comment: The report should address policy aspects.
Response; This issue is discussed above in Section 1,
comment 3.
Comment: Reasonable timing should be provided for control
implementation with phasing of controls on major
sources first.
Response; A State submitting a SIP needs to demonstrate
attainment for specific attainment years. This approach does not
preclude phased implementation of controls up to the attainment
year.
Comment: The declassification process is too time
consuming.
Response: This topic is beyond the technical scope of the
185B Report.
Comment: Additional monitoring efforts are encouraged.
Response: This topic is addressed throughout the Report and
this Attachment.
Comment: Poor timing exists between SIP and research needs.
Response: This issue is addressed in Section 1, comment 1
above.
22
-------
American Gas Association
(see comments and responses for Coastal, above)
Interstate Natural Gas Association
(see comments and responses for Coastal, above)
Electric Power Research Institute
Comment; Screening models can be developed to provide
comprehensive cost/benefit analyses.
Response: The use of screening approaches and the expansion
of analyses to address cost/benefits, and other measures, is
discussed above in Section 1 (comment 4) and Section 2 (response
to Coastal).
23
-------
Pacific Gas and Electric
Comment: The ozone trends in CA indicate that ozone levels
generally are higher during weekend periods. We
suspect that this is due to a reduction of NOx
emissions on weekends, particularly the loss of
concentrated NOx during morning and afternoon rush
hours.
Response: Assuming these trends are statistically
significant, this observed "weekend" effect is indicative of the
complexities associated with NOx control issues. Evidently, the
rate of progress in reducing ozone in certain California
locations has been greater for weekends than weekdays. The
elevated weekend ozone may be due to a variety of changes in the
temporal and spatial distribution of precursors. Moderate
reductions in NOx emissions can increase ozone in some locations;
this is one of several plausible explanations for the observed
increases. An evaluation of weekend effects is best conducted
through modeling by accounting for weekend specific emission
changes. The existence of this weekend effect offers an
opportunity to conduct population exposure impact assessments, to
the extent that weekend and weekday activity patterns differ.
24
-------
The economic risks of proceeding with uncertain
models/data bases merit development of a more
realistic SIP timetable.
Response; This issue has been addressed above throughout
Attachment 3. Congress, through the 1990 CAA, was clear in
setting the current SIP time schedules. The schedule requires
SIP submittals in most areas by November, 1994.
Comment: Where is the 182(f) guidance describing the
exemption demonstration procedures?
Response: The NOx exemption guidance is being developed
separately from the Section 185B Report. This guidance is
discussed above in Section 1 (comment 3).
Comment: The SIP schedule is not practical given poor data
bases, and need for additional research.
Response: see comments throughout thrs Attachment.
Comment: The 1989-1992 ozone data for trends should be
added.
Response: A discussion of ozone trends data through 1991 has
been added to Section 3.2 of the EPA Report.
Comment: The overall summary of technology is accurate;
however, additional results are available that are
not presented in Attachment 2.
Response: Due to time and budget restraints, Attachment 2
will not be updated. More current information and data are
contained in the alternative control techniques (ACT) documents
.that were or will be issued for various source categories from
1991 to September 1993.
Comment: Generation of NOX is exponentially proportional to
temperature and square root dependent on oxygen
concentration.
Response: This detailed information is included in the
various ACT documents referred co above.
Comment: Section 2.1.4.1 does not address the different:
selective catalytic reduction configurations.
Response: This information is included in the various ACT
documents.
25
-------
Comment: Burners out of service and biased burner firing
should be considered as NOX controls for coal
fired utility boilers.
Response: These controls are evaluated in the utility
boilers ACT document.
Comment; Peak values of 55 to 65% shown for low NOX burners
with overfire air or advanced overfire air are
optimistic. 50% is a practical maximum relative
to a properly tuned boiler. Operating at lower
emission levels lead to operational problems such
as boiler corrosion and increases in unburnt
carbon and carbon monoxide emissions.
Response: Based on limited data, the emission reduction for
low NOX burners with overfire air varies from 30-65% (Attachment
2, page 3-9). The specific emission reduction is heavily
dependent on site specific parameters such as boiler design and
operating characteristics such as firing configuration, furnace
size and heat release rate, type of fuel used, boiler capacity
factor and condition of existing equipment.
Comment:
Response:
document.
Various data are available for retrofits that had
no problems with unburned carbon, carbon monoxide
or boiler slagging.
These data are included in the utility boiler ACT
Comment: Water or steam injection is an attractive
alternative for industrial, commercial and
institutional boilers.
Response; This technique is included in the non-utility
boiler ACT document.
Comment: NOX generation from the transportation sector is
not discussed.
Response: As noted in the introduction, Attachment 2 only
addresses NOX controls for stationary sources. A brief
discussion of motor vehicle NOx control techniques is provided in
Section 4 of the primary Report.
Comment: Advanced combustion techniques such as AFBC, PFBC
and coal gasification are not discussed.
Response; These processes are included in the utility boiler
ACT document.
26
-------
Comment; Combined NOX-S02 controls are largely overlooked.
Response; The scope of the Report was limited to basic
availability and extent of NOx control technologies to address
the specific CAA Section 185B requirements.
Comment: No cost information is provided. Integrating
reduction capability over boiler population to
estimate current extent of controls or to project
potential control levels is not done.
Response: The scope of the Report was limited to basic
availability and extent of NOx control technologies to address
the specific CAA Section 185B requirements. These subject areas
are covered in detail in the various ACT documents.
27
-------
REFERENCES
Carter, W.P.L., 1991. Development of ozone reactivity scales for
Volatile Organic Compounds. EPA 600/3-91-050, U.S.
Environmental Protection Agency, Research Triangle Park, NC,
27711.
EPA (Environmental Protection Agency). 1991. Guideline for
Regulatory Application of the Urban Airshed Model. EPA-
450/4-91-013. Environmental Protection Agency, Office of
Air Quality Planning and Standards, Research Triangle Park,
N.C.
EPA (Environmental Protection Agency). 1992. National Air
Quality and Emissions Trends Report, 1991. 450-R-92-001.
Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, N.C.
Geiss, R.O, W.M. Burkmyre and J.W. Lanigan. 1992. Technical
Highlights of the Dodge Compressed Natural Gas RAM-
Van/wagon. SAE paper no. 921551.
Koenig, W.E, W. Lee and W. Bernhart, 1976. Technical and
economical aspects of methanol as an automotiove fuel. SAE
paper no. 760545.
Johnson, G.M. and S.M. Quigley, 1989, "A universal monitor for
photochemical smog.", AWMA 82nd Annual Meeting, Anaheim, CA.
Paper No. 89-29.8.
Wolff, G.T. and P.E. Korsog, 1992, "Ozone control strategies
based on the ratio of volatile organic compounds to nitrogen
oxides.", J. Air Waste Manage. Assoc. 42:1173-1177.
Zealewsky E., G. Sistla, R. Henry and S.T. Rao, 1992, "Trends in
ozone and its precursors in the Northeastern U.S.",
presented at the AWMA ozone specialty conference in Boston,
MA.
28
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. RETORT NO.
EPA-454/R-93-025
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
SUMMARY OF PUBLIC COMMENTS AND EPA RESPONSES
on the Draft Report:
THE ROLE OF OZONE PRECURSORS IN TROPOSPHERIC OZONE
FORMATION AND CONTROL
6. REPORT DATE
July 30, 1993
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Technical Support Division (MD-14)
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
16. SUPPLEMENTARY NOTES
16 ABSTRACT
This Report is Attachment 3 of the CAA Section 185B Report entitled, The Role of Ozone Predcursors in
Tropospheric Ozone Formation and Control. It includes a summary of comments and EPA responses based on a
30 day review of the draft Section 185B Report.
7 KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Ozone, oxides of nitrogen
volatile organic compounds
photochemical modeling
control strategies
4 DISTRIBUTION STATEMENT
Release Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS IRepom
Unclassified
20. SECURITY CLASS (Pagal
Unclassified
c COSATI Field/ Group
21 NO OF PAGES
32
22 PRICE
* Form 2220-1 IRav. 4-771 PREVIOUS EDITION IS OBSELETE . , >".-,'
Reg'-on 5. -;;.-; .. , '^ 12th Rocr
77 West Jaciw-- : „;,. -;'u>
Chicago, IL 60oo4-ooy
------- |