PHOTOCHEMICAL REACTIVITY WORKSHOP
U.S. Environmental Protection Agency
May 12-14,1998
PROCEEDINGS
Regal University Hotel
2800 Campus Walk Avenue
Durham, North Carolina 27705
-------
950R98003
TABLE OF CONTENTS
PRESENTATIONS
SESSION I
Introductions 1-1
Moderators: Jake Hales, NARSTO International. ENVAIR
Howard Feldman, American Petroleum Institute
Welcoming Remarks 1-3
Gary Foley, National Exposure Research Laboratory, EPA
Sally Shaver, Air Quality Strategies and Standards Division, EPA
The Public-Private Partnership Process 1-4
Ron Patterson, NARSTO International. National Exposure Research Laboratory, EPA
Current EPA Regulatory Viewpoint on Reactivity 1-7
Bill Johnson, Ozone Policy and Strategies Group, EPA
Current EPA Research Viewpoint on Reactivity 1-12
Basil Dimitriades, National Exposure Research Laboratory, EPA
California's Hydrocarbon Reactivity Program 1-17
Randy Pasek, California Air Resources Board
The NAS/NRC Project on Reactivity (no materials available)
William Chameides, Georgia Institute of Technology
VOC Reactivity - Beyond Ozone 1-29
Alan Hansen, EPRI
SESSION II
Current Status of VOC Reactivity Research 2-1
William Carter, University of California at Riverside
VOC Reactivity Quantification: Approaches, Uncertainties, and Variabilities 2-13
Ted Russell, Georgia Institute of Technology
Quantification of Uncertainties in Reactivity Estimates for Volatile Organic Compounds . . 2-27
Jana Milford, University of Colorado at Boulder
Comparison ofPOCP and MIR Scales 2-28
Richard Derwent, Meteorological Office, United Kingdom
EPA's MODELS3 Framework and the Community Multi-Scale Air Quality Model 2-42
Robin Dennis, National Exposure Research Laboratory, EPA
Establishing a Community Modeling Capability 2-63
Kenneth Galluppi, University of North Carolina at Chapel Hill
Emissions Modeling Issues for Reactivity Calculations: State and Status of the Sparse Matrix
Operator Kernel Emissions (SMOKE) Modeling System 2-74
Neil Wheeler, Microelectronics Center of North Carolina-NC Supercomputing
Desirable Scientific and Operational Criteria for Use ofEulerian Model to Compute VOC
Reactivity 2-85
Harvey Jeffries, University of North Carolina at Chapel Hill
-------
SESSION III
Dunn-Edwards Proposed NARSTO Research on Ozone Formation Potential ofVOC Emissions
from Architectural Coatings ->~
Edward Edwards, Dunn-Edwards Corporation
JO
CMA Research Initiatives -3"0
Jonathan Kurland, Union Carbide Corporation
CSMA Position on the Importance of Relative Reactivity 3-17
Doug Fratz, Chemical Specialties Manufacturers Association
Reactivity Concerns 3-31
Phil Ostrowski, Occidental Chemical Corporation
Categorization of Low Reactivity Compounds 3-41
John Owens, 3M Company
Impact of a Molar Ethane Standard on the Number and Type ofVOC-Exemptible Compounds:
Practical and Environmental Implications 3-45
Daniel Pourreau, ARCO Chemical Company
General Industry Concerns with the Process 3-54
Donna Carvalho, Pennzoil Company/Magie Brothers Company
SESSION IV
A Global 3-D Radiative-Dynamical-Chemical Model for Determining Large-Scale Impacts of
Atmospheric Ozone Precursors 4-1
Eduardo Olaguer, Dow Chemical Company
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism 4-2
William Stockwell, Fraunhofer Institute
Hydrocarbon Reactivity and Ozone Production in Urban Pollution According to the Stockwell
el. al. (1990) Reaction Mechanism 4-15
Chris Walcek, State University of NY at Albany
Multi-day Impacts from Low Reactivity Compounds 4-24
Gary Whitten, Systems Application International, Inc.
Computing Volatile Organic Compound Reactivities with a 3-DAQM 4-29
Zion Wang, University of North Carolina at Chapel Hill
The Use of NAPS Data to Generate Sensitivities of Ozone Production Towards Changes in NOx
and VOCs 4_30
Paul Makar, Atmospheric Environment Service
Temperature Dependence of Ozone Chemiluminescent Reactions with Organics: Potential
Screening Method for VOC Reactivities 4.42
Jeffrey Gaffney, Argonne National Laboratory
VOC Receptor Modeling as an Aid to Evaluating the Effect of Reactivity Changes on Ozone
Formation 4.44
Donna Kenski, US Environmental Protection Agency
The Impact ofBiogenic VOC Emission Modeling on the Simulation of a Long-term Ozone Time
Series 4_59
John Sherwell, Maryland Department of Natural Resources
11
-------
TV A's Research Efforts in Tropospheric Ozone Formation and the Contribution of Natural
Hydrocarbons to the Reactivity of Summertime VOCs in the Rural Southeastern US 4-69
Roger Tanner, Tennessee Valley Authority
Computational Studies ofOxidant Reactions of Volatile Organic Compounds Relevant to the
Formation of Tropospheric Ozone 4-70
David Dixon, Pacific NW National Laboratory
SESSION V
Atmospheric Chemistry of Organic Compounds 5-1
Roger Atkinson, University of California at Riverside
Atmospheric Chemistry of Oxygenated Organic Compounds 5-14
Ray Wells, US Air Force
Multicomponent Aerosol Generation System (MAGS) for the Study of Fine Particulates on
Photochemical Reactivity ofOrganics 5-28
Shri Kulkarni, Kultech Incorporated, M. B. Ranade, Particle Technology, Inc.
Numerical Study of the Development of Ozone Episodes in Germany; Relation of Anthropogenic
and Biogenic Hydrocarbons 5-29
Franz Fiedler, University of Kalsruhe, Germany
European Studies on the Photooxidation Mechanisms of Aromatic Hydrocarbons and
Oxygenates: Reactivity Implications 5-34
Ian Barnes, Bergische University at Wuppertal, Germany
QUESTION/DISCUSSION SUMMARIES
PRESENTATION SUMMARIES 6-1
FREE FORUM SUMMARY 6-9
POLICY AND SCIENCE QUESTIONS SUMMARY 6-15
PUBLIC/PRIVATE PARTNERSHIP SUMMARY 6-19
AD-HOC OPERATOR TASK FORCE SUMMARY 6-20
APPENDICES
Appendix A - Attendees List A-l
Appendix B - Ad-Hoc Task Force - Minutes of Initial Meeting B-l
in
-------
PRESENTATIONS
-------
Reactivity Workshop
Meeting Objectives
1. Obtain participant input on important reactivity-related issues.
( Incorporate finalized list in Proceedings Report)
2. Obtain participant input on reactivity-related research needs.
(Incorporate finalized list in Proceedings Report)
3. Establish a volunteer Reactivity Research Planning Group to
develop a responsive research program plan.
4. Identify follow-up action items and an associated time table.
5. Determine type of partnership forum to be used in planning
and implementing the reactivity research program
J M Hales. May
Some Design Features
of this Meeting
Focus on scientific aspocts of reactivity concepts"Policy-relevant,
but not policy-driven science;" (the NARSTO paradigm).
Structured around "science questions" and related "policy
questions." (Initial sets to be modified by group process).
Define science issues associated with policy issues; don't
attempt to resolve policy issues.
No pre-conceptions; no loaded agendas; a level playing field.
This is an information-gathering and distillation process.
"Everybody wins" environment. (Including the citizen and
the taxpayer). .'
J M Hales. M.iy I I'l'f
1-1
-------
Having Said This . . .
. . There are some "complicating realities:"
The term "reactivity" is used, but not defined explicitly in the 1990 CAAA.
Any attempt to define this term explicitly, including adoption of any
particular reactivity scale, has immediate policy implications.
Associated legal action is contemplated and/or in progress.
J M Hales. May
Therefore . . .
... If s critically important that we maintain an
objective and scientifically oriented atmosphere
at this workshop.
Howard
&
Jake ^,-V
Science
J M Hales MJV J'.
Policy
Legal
1-2
-------
Welcoming Remarks
Dr. Gary Foley, Director
National Exposure Research Laboratory
US Environmental Protection Agency
Office of Research and Development
Welcoming remarks were offered on behalf of EPA by Dr. Gary Foley, Director, National Exposure
Research Laboratory, EPA Office of Research and Development, and Ms. Sally Shaver, Director,
Air Quality Strategies and Standards Division, EPA Office of Air Quality Planning and Standards.
Dr. Foley welcomed the idea of incorporating reactivity research into the North America Research
Strategy for Tropospheric Ozone (NARSTO) program. He stressed that the NARSTO partnership
between government and the private sector has been very effective in investigating the
photochemical ozone problem and that he is confident that the reactivity research effort will be
equally successful.
Sally ShaVer, Director
Air Quality Strategies and Standards Division
Office of Air Quality Planning and Standards
US Environmental Protection Agency
Sally Shaver made welcoming remarks on behalf of EPA's Office of Air Quality Planning and
Standards. She stated that it is important that EPA's policies reflect the best science, and she said
EPA looked forward to hearing the information which participants would provide.
1-3
-------
A Public-Private Partnership
for
Photochemical Reactivity Research
Ronald K. Patterson
Associate Management Coordinator
NARSTO International
U.S. Environmental Protection Agency
Office of Research and Development
National Exposure Research Laboratory
ABSTRACT
The decision of the U.S. Environmental Protection Agency to re-evaluate its policy on photochemical
reactivity in an open forum, where regulated industries, the science community, and other stakeholders
can participate, sets the stage for forming a public-private partnership to resolve the science and policy
issues identified by the process. The partnership approach to complex environmental issues normally
provides an opportunity for joint research planning, and a platform that the stakeholder community can
use to coordinate research activities, limit research gaps and duplication, leverage resources, and share
results.
NARSTO is a model for this type of partnership. NARSTO was established, under a non-binding
Charter in 1995, as a science-driven, tri-national, public-private partnership. To date, membership
consists of science agencies, regulatory agencies, regulated industries, academic institutions and public
interest groups in Canada, Mexico, and the United States. NARSTO facilitates, plans, and coordinates
policy-relevant research on atmospheric processes in the troposphere over North America. Ozone and
particulate matter are the current focus of the NARSTO program. However, the research emphasis is
placed on atmospheric chemistry, modeling, emissions, monitoring, meteorology, methods development,
and integrated analysis and assessment. NARSTO provides quality assurance, data management, data
archival, and data accessibility guidelines and services to its members. Peer review of major NARSTO
outputs is provided by the National Research Council.
Photochemical reactivity is an integral part of the NARSTO research agenda, and this area of research
could be expanded under the NARSTO Modeling and Chemistry Team. If this workshop selects the
NARSTO partnership as its forum for conducting future reactivity research, then a NARSTO
subcommittee could be formed from volunteers identified in this audience. The subcommittee would
develop a strategic research plan for implementation by the member organizations. Each member
organization would accomplish its portion of the work using their individual institutional planning and
funding mechanisms. Under the NARSTO scenario, each organization involved must sign the NARSTO
Charter as a sponsoring or participating partner. Participating partners are usually academic and
contractor institutions. Many of the organizations attending this workshop are already NARSTO
partners.
This workshop will be declared a success, if those in attendance can (1) identify the major policy and
science issues, (2) agree upon & partnership arrangement for planning and implementing future reactivity
research, (3) commit to participation in the planning process, and (4) commit to funding the research
agenda.
1-4
-------
A PUBLIC-PRIVATE PARTNERSHIP
FOR
PHOTOCHEMICAL REACTIVITY RESEARCH
Ronald K. Patterson
Associate Management Coordinator
NARSTO International
Purpose of this Reactivity Workshop
i Examine the policy issues
i Evaluate the state-of-the-science
i Commit to forming a partnership to resolve the
science and policy issues
i Commit to funding research under the
partnership
Public-Private Partnership
Involves all stakeholders
Provides a forum for joint research planning
Leverages resources v
Coordinates research agenda
Limits research gaps and duplication
Shares data and information
The NARSTO Forum
Science driven, tri-national, public-private partnership
Established by a non-binding Charter in 1995
Facilitates, plans, and coordinates policy relevant research
on atmospheric processes in the troposphere over North
America
Provides quality assurance, data management, data
archival, and data accessibility guidelines and services
Provides NRC peer review of major products
The NARSTO Organization
i Members
Science Agencies
Regulatory Agencies
Regulated Industries
Academic Institutions
United States, Canada, Mexico, (Europe)
Membership Types
Sponsoring Partners (54)
Participating Partners (20)
Arfllliated Partners (2)
EUROTRAC
IGAC
(See Poster: 'A New Approach to Complex Environmental
Problems in the Continental Troposphere')
Reactivity - 05/07/98
1-5
The NARSTO Program
Programmatic focus on Ozone and PM
Research emphasis on:
atmospheric chemistry
modeling
emissions
monitoring
meteorology
methods development
Integrated analysis and assessment
-------
The NARSTO Process
(if selected as your partnership forum)
Success for this Workshop
Organized as a Subcommittee under the Modeling and
Chemistry Science Team
Planned and Coordinated by the Subcommittee
Implemented through the normal planning and funding
mechanisms of the sponsoring Institutions
Quality assured, peer reviewed, archived, and assimilated
using existing NARSTO systems
Identify the major Policy and Science Issues
Agree Upon a Research Partnership Arrangement
Commit to Participation in the Strategic Planning
Process
Commit to Funding the Research Plan
COMMITMENT QUESTIONNAIRE
ii in lavor of a public
-------
EPA Current
Regulatory Viewpoint
on Reactivity
Bill Johnson
Photochemical Reactivity Workshop
May 12-14,1998
What is the History of...
EPA's VOC Reactivity Policy?
EPA announced its reactivity policy on July 8,1977.
This policy classified VOC's into categories ...
* Reactive
Negligibly Reactive
Four compounds were originally classified Negligibly
Reactive...
* Methane
Ethane
Methyl chloroform
« Freon 113
Today EPA still uses this 1977 Reactivity Policy.
1-7
-------
What VOC's have been ...
Classified as Negligibly Reactive ?
Early additions to the Negligibly Reactive list tended
to be chlorofluorocarbons which possessed Sow
reactivity and which could serve as replacements for
stratospheric ozone depleters.
Later Ethane was used as the standard cut-off
comparing the reactivity of compounds. Those
compounds with reactivities below ethane might be
considered for Exemption
Since 1977 more than 42 additional compounds or
classes of compounds have been classified
Negligibly Reactive and added to the Exempt list.
How is the Reactivity of...
a Compound Determined?
Most Exemptions were determined using the kOH value
(the reaction rate constant for the reaction of a
compound with the OH hydroxyl radical), expressed in
units of cm3/moiecule-sec, and compared to the kOH
value of Ethane.
In 1993, EPA began receiving VOC Exemption Petitions
based on the Maximum incremental Reactivity (MIR)
scale developed by Dr. William Carter at the University
of California at Riverside. MIR values are expressed in
units of grams ozone produced per gram of compound
reacted.
Acetone was the first compound evaluated for
Exemption using MIR values. This evaluation was made
on a per gram basis as stated in the Federal Register.
1-8
-------
Questions Raised in using ...
Values to Compare Reactivity
The MIR values raised a number of questions.
> Are MIR values better to use than kOH values for
comparing reactivity?
> Should a per gram or per mole be used for
comparing compounds to ethane?
> Could the ranking of compounds in order of
reactivity by MIR value be used in some kind of
substitution scheme of control? How would such a
scheme work?
A review of the original 1977 experimental work used to
select ethane as the Exempt cut-off showed that the
experiments were done on a mole basis. Comparisons
to ethane on a per gram basis may not be valid.
Qu£sti|>ns Retitjpns Raised ab^ut;.,
Consideration of Collateral Effects
Recent Petitions have raised questions about considering
collateral effects in granting exemptions to compounds.
> Should special consideration be given to exempting a
compound slightly more reactive than ethane, but may
displace more reactive and/or toxic compounds such
as xylene? Would this be a positive environmental
move, even if unlimited amounts of the exempt
compound could then be used?
> Should very toxic compounds or stratospheric ozone
depleters be exempted if they are of low reactivity?
The 1977 policy indicates;EPA should consider such
environmental impacts in making exemption decisions (e.g.
we have never exempted benzene even though of low
reactivity.)
1-9
-------
What is the Current Status of ...
Submitted to EPA?
EPA has received 14 petitions on which action is
pending. Most of these compounds are:
> Not less reactive than ethane on a per
mole basis or
*- Hazardous air pollutants (under section
112) or are stratospheric ozone depleters.
Text for vapor pressure figure
EPA's definition of VOC does not include a vapor pressure cut off. At one time in
the late 1970's, EPA recommended a 0.1 mm Hg vapor pressure cut off. In 1987,
EPA asked States to remove this from the VOC definition in their regulations.
The reason for removing the 0.1 mm Hg cut off can be seen from Figure 1.
The 0.1 mm Hg at 20 °C cutoff would only control compounds with carbon
number lower than about C12. The older Los Angles Rule 66 cut point (0.5 mm
Hg at 104°C) was actually more strict and would control compounds up to about
C18.
Studies have shown that compounds
-------
FIGURE
Vapor Pressure and Temperature in Relation to
VOC Definition
Vapor pressure (mm Hg)
1,000 F
100
10
0. 1
0 0 i
Rule 66 cutoff - 0.5 mm Hg at 104 degrees C.
0. 1 mm Hg ot 20 degrees C.
0 100 200 300
Temperature (degrees C)
1-11
-------
CURRENT EPA RESEARCH VIEWPOINT ON REACTIVITY
by
Basil Dimitriades
EPA/ORD/NERI7HEASD/ACPB
RTP,NC
EPA's main reason for wanting to update its reactivity policy, is the simple fact that the
"now science", that is, the scientific understanding we now have about the reactivity property of
organic compounds, is considerably different than the "then science", that is the scientific
understanding we had some two and a half decades ago, during the early 1970s, when the current
policy was conceived and formulated. Therefore, it would be useful to include in this workshop
this and other presentations that would help convey to this audience some understanding of the
changes in reactivity science that took place during the past two and a half decades.
In this presentation, I will give you some historical information on the development of the
current policy, and after that, I will describe the scientific bases of the policy and point out
whatever weaknesses were revealed by the recent scientific evidence. This presentation and the
presentations later by the active researchers in this area, should make it possible to judge the
significance of the reactivity science changes that occurred during the last two and a half decades.
I should mention that nearly all of the material I will be discussing today is in a report I submitted
for publication in the Journal of the Air and Waste Management Association, some 3 months
ago.
The initial version of the reactivity policy now in effect was developed in 1971, as part of
EPA's guidance to the States for preparation of State Implementation Plans for ozone attainment.
(Appendix B in EPA's 1971 Guidance...). In that version, EPA emphasized reduction of total
mass of organic emissions , but it also did take into account reactivity. Specifically, it allowed
for substitution of a less reactive for a more reactive organic emission, if it could be shown that
such substitution would result in a reduction of the total reactivity of the emissions mixture. This
latter concession encouraged States to develop organic emission substitution regulations, such as
LA's Rule 66, which allowed a large number of organic emission species and emission mixtures
to be exempted from the ozone control regulations on grounds that they had if not negligible, at
least tolerably high reactivities.
A few years after that, the researchers came out with the finding that pollutant transport
conditions in the atmosphere enhance ozone formation so as to make many of those organics that
were previously thought to be unreactive, to act as significant ozone producers. This led EPA to
rethink its reactivity policy, and in 1977 it issued its next policy, under the title "Recommended
Policy on Control of Volatile Organic Compounds", which is the policy version now in effect,
and which policy, in contrast to the preceding one, it went from an extremely tolerant policy to an
extremely conservative one. Specifically, by the 1977 policy only four organics were accepted as
negligibly reactive: methane, ethane, 1,1,1-trichloroethane (methyl chloroform), and
Trichlorotrifluoroethane (Freon 113). All other organic were assumed to be reactive, and, of
course, subject to the ozone regulations. The policy, however, was flexible in that it allowed for
reactive organics to be re-classified as negligibly reactive if and when new scientific evidence is
1-12
-------
produced and shown to support the reclassification. It should be noted at this point that at that
time EPA had not had standard test methods issued for determining whether an organic
compound is negligibly reactive or not the burden for making this determination rested outside
the Agency. The Agency did support, however, development of such methods, and eventually
accepted them and offered them, unofficially at least, for public use. These methods, known as
the kOH and MIR methods, have been used routinely in the recent years to produce reactivity
evidence for many previously unstudied organics, and, as a result, several tens of organic
compounds have been re-classified and placed in the list of negligibly reactive or exempt
organics.
Lastly, in 1992, EPA restated its reactivity policy, except that this time it used the term
"VOC" to denote organic compounds with significant potential for producing ozone, and
declared all organic emission species to be VOCs except those that were determined by EPA to
be non-VOCs.
So much about the history of the EPA reactivity policy development. Next, I will talk
about the scientific basis of the existing policy and its weaknesses as?we see them now.
There are five key components or elements that constitute the existing reactivity policy.
Subjects of these components are:
(1) Exemption of Organic Emissions on Reactivity Bases.
(2) Use of the, ethane reactivity as the "bright line" separating VOCs from non-VOCs.
(3) Reactivity Classification Guideline Methods.
(4) Assumed universal validity of reactivity scales
(5) Consideration of Emission Volatility.
(1) "Exemption policy". This policy element mandates that non-VOCs must be exempted
from the ozone-related control and inventory requirements. The bases for exempting the non-
VOCs are two judgments made by EPA at that time:
- While all organic compounds are capable of producing more or less ozone in the
atmosphere, not exempting those that produce only negligible amounts of ozone would be
impractical.
- The other judgment was based on the perception at the time that there is a significant
number of organic compounds that have negligible reactivity ~ so, it would be worth the
effort to classify organics into VOCs and non-VOCs.
And so, in 1975 EPA decided to develop for official use a two-class reactivity
classification scale. To develop the requisite scale, EPA conducted a smog chamber study in
1-13
-------
which several organic compounds were irradiated under a standard set of extremely favorable
conditions, and resultant ozone yields were compared with the ozone air quality standard which
at the time was 0.08 ppm O3. The conditions used were 4ppm in moles of organic reactant -- a
concentration representative of the most polluted urban atmospheres at the time in the US --
0.2ppm in moles of NOX (giving a VOC-to-NOx ratio of 20:1, extremely favorable for ozone
formation), radiation comparable to that of natural sunlight in Los Angeles during summer, and
the test mixture was irradiated until the ozone concentration peaked out. Results showed that
only four organics produced less than 0.08 ppm ozone: methane, ethane, trichloroethane, and the
trichlorotrifluoroethane, and these were, therefore, the only entries in the table of negligibly
reactive organics included in the 1977 policy.
Obviously, that reasoning behind the 1977 reactivity classification was judged by EPA to
be consistent with the scientific evidence available and the best that could be thought of at the
time. Today, however, we have several reasons for questioning the validity of that evidence and
thinking. One specific and strong objection we now have is to exempting the non-VOCs from
the emission inventory requirement. Non-VOCs exempted from control will accumulate in the
atmosphere due to growth, and their ambient concentration will eventually reach the point at
which they will contribute to ambient ozone significantly. At that point, of course, they should
be taken into account in calculating control requirements. This would be impossible, however,
because the models used to compute control requirements require emission inventory data. Thus,
this "Non-VOC exemption" part of the EPA policy, clearly, must be re-considered.
(2) Ethane reactivity bright line. According to this policy element, organic emissions with
reactivity at or below the, ethane reactivity bright line shall be exempted from the ozone
regulations as being non-VOCs. This policy element was never issued by EPA officially. It
acquired the policy status when EPA began to use the comparison with ethane as the basis for
judging whether an organic is a non-VOC, and the rationale of this basis was, of course, the fact
that ethane is the most reactive species of those identified by EPA in 1977 as being negligibly
reactive. Regarding the validity of this policy element, there are several questions at issue. I will
discuss the most important ones, namely:
(a) Is the choice of ethane as the boundary reactivity species an appropriate one?
(b) To determine VOC or non-VOC nature of an organic, should the comparison of the
organic with ethane be made on a per-unit-weight or on a per-mole basis?
(c) Is the distinction between VOCs and non-VOCs really necessary?
To answer the first question, of whether the ethane reactivity is the appropriate boundary
separating VOCs from non-VOCs, we need to go back to the reasoning EPA used in 1977 to
classify organics into VOCs and non-VOCs. In that study there were some 20 low reactivity
organic compounds tested, and of those, four were found to produce ozone less than 0.08 ppm.
Of those four, ethane was the most reactive one and on those bases, ethane was taken,
unofficially, to be the "boundary" reactivity species. Are these bases valid? Today we don't
think so, for several reasons. First, we believe today that ethane, if allowed to accumulate in the
1-14
-------
atmosphere at higher than 4 ppm levels, will produce more ozone than 0.08 ppm. Second, we
also know today that testing organics using modern smog chamber methodology gives much
higher ozone concentration results, the main reason being that nowadays the test organic is not
irradiated in the smog chamber alone, as was done then; instead, it is irradiated in the presence of
other organics, and this, of course, is much more realistic and it results in much more ozone.
The second question, whether organics should be compared with ethane on a per-unit-
weight basis or a per-mole basis, is raised because it makes a difference whether we use the one
or the other basis. Relative to the per-mole basis, comparison with ethane on a per-unit-weight
basis tends to cause reactive, high molecular weight organics to have artificially reduced
reactivity values and be classified as negligibly reactive, in conflict with the per-mole
comparison. Which is correct? To answer this question, once again we need to go back to the
EPA smog chamber study of the mid-1970s, that led to the selection of ethane as the boundary
species. If you recall, the comparison of the organics in that smog chamber study was done for
equimolar concentrations. Based on that, and for consistency shake, the comparison with ethane
must be made on a per-mole basis. Needless to say, we could have compared the test organics in
the EPA study for equal weight, rather than mole concentrations, but then the boundary
reactivity species would have been not ethane but some other species. Comparing organics with
that species on a per-unit-weight basis would alleviate but not eliminate the problem. In fact, our
conclusion is that regardless of how we compare the reactivities of organics to the boundary
level, it is not possible to avoid this problem completely, and this constitutes a conceptual
weakness of this policy element that calls for classifying organics into VOCs and non-VOCs
based on comparison with a given organic compound's reactivity.
V
The final question, whether the VOC-vs-nonVOC distinction is necessary is to some
extent of policy nature. The need for making such a distinction depends on whether the Agency
wants to stay with the existing policy of exempting only those organics which are shown to be
negligibly reactive, or decides to adopt, another policy instead, for example, a policy of
exempting organics that have non-negligible reactivity but are to be used as substitutes for other
much more reactive organics.
3. Reactivity Classification Guideline Methods Existing EPA guidelines for such
methods are unofficial. The kOH and MIR methods acquired the status of guideline methods
because data obtained with such methods have been accepted by EPA in processing petitions for
exempting organic emissions from ozone regulations. What do we think today about these
methods?
We now believe that the kOH method has very limited utility because it represents how
fast the organic reacts with the OH radical ~ the first reaction step in the ozone forming process
but it tells us very little about the follow-up chemistry that results in formation of ozone. The
MIR method has greater validity in that it does provide a direct measure of ozone production, but
it is based on use of the EKMA model, a model now thought to be outdated. We still consider
the MIR method very useful, but we believe that current science allows for development of other
modeling methods which, while much more complex, can provide more accurate reactivity
estimates. [More about these methods in subsequent presentations]
1-15
-------
4. Universal validity of reactivity scales. The existing policy is based on the assumption
that the current classification of organics into VOCs and non-VOCs is universally valid,
independent of ambient conditions. We now know, however, that the ozone formation chemistry
is such that reactivity varies with conditions such as organic-to-NOx ratio, ambient organic
composition, sun light intensity, and pollutant transport conditions. This, obviously, raises
questions about the practicality of a reactivity scale that is not universally valid. We need to
think about this and find some way of using the reactivity concept in a way that the effect of this
problem is minimized. For example, consideration should be given to the fact that the
uncertainty introduced by this ambient condition variability factor is much smaller for relative
reactivities than for absolute reactivities. This means, of course, that policies that place the
emphasis on relative reactivity data, e.g., emission substitution, may be more reliable in so far as
this problem is concerned than policies that place the emphasis on absolute reactivity data, as the
existing policy is.
5. Emission Volatility. The existing EPA policy distinguishes between VOCs and non-
VOCs and cites reactivity as the basis for the distinction but says nothing about volatility. There
are guideline methods for determining whether an organic is negligibly reactive or not, but there
are no such methods for determining whether an organic is negligibly volatile or not. Thus, the
reactivity and volatility properties of organic emissions have been receiving different regulatory
treatments, apparently without scientific justification, and this inconsistency should perhaps be
rethought.
CONCLUSIONS
In contrast to the mid-1970s thinking, our current thinking suggests that
1. exemption of non-VOCs from the inventory requirement is unjustifiable;
2. use of the ethane reactivity as the boundary level separating VOCs from non-VOCs has
both validity and operational problems (basis of comparison);
3. current science allows for development of more reliable guideline methods for
classifying organics into "reactive" and negligibly reactive.
4. policies that place the emphasis on relative reactivities are probably-subject to less
uncertainty relative to those that rely on use of absolute reactivity data; and
5. it is unjustifiable for the current policy to treat the reactivity and volatility properties of
organic emissions differently with respect to the photochemical ozone pollution problem.
1-16
-------
California's
Hydrocarbon
Reactivity Program
By
Randy Pasek Ph.D.
California Air Resources Board
at
U.S. EPA Reactivity
Workshop
May 12-14, 1998
California's Reactivity
Beginnings
+ In 1987 ARE formed advisory
board on fuels.
+ Needed way to compare
alternatively fueled vehicle's
emissions.
+ Use of reactivity of vehicle
exhaust for comparisons.
+ Adopted LEV/CF regulations
in late 1990 (RAF).
+ VOC exemptions
-------
oo
Other Possible
Reactivity Based
Programs
+ Future Possibilities
- Aerosol Coatings
- Consumer Products
- Emissions Trading
- Aerosol Forming Reactivity
- Motor Vehicles
Challenges
Encountered
+ Reactivity Scale
- MIR
Dual control program in CA
MIR scale complements NOx
controls
+ Uncertainty
- Alternative fuels program
(LEV/CF)
MIRs -30 - 70%
RAFs ~5 - 15%
- Consumer Products
Most-used compounds well
characterized
-------
Challenges
(continued)
+ Speciation Profiles
- Uncertainty
- Inventory and modeling
+ Unstudied Compounds
- Alternative fuels program
(LEV/CF)
Most compounds some data
- Consumer Products
Some compounds ~ no data
Upper limit MIR Estimation
Challenges
(continued)
+ Regulation Development
- Good science
- Flexibility
- Simple
- Equivalent ozone benefits
- Enforceable
- Cost effectiveness
-------
to
o
Challenges California's Reactivity
(continued) Program
+ How to incorporate exemptions
into reactivity regulations. ^
+ Research
- VOC exemptions
CA has own review + Reactivity Team
Methane or ethane
- Other policy exemptions + External Assistance
-------
Research
+ Uncertainty
- Chamber parameters' effect on
MIR values
+ Reactivity Value Estimates
- Consumer product compounds
- Auto exhaust
- Other important mixtures
+ Improvement of Reactivity
Value Estimates
- Lower cost
- Lower uncertainty
Research
(continued)
+ Atmospheric Chemistry
- C10 alkanes
+ Speciation
- Architectural & Aerosol coatings
- "Complete picture profiles"
- Surveys to obtain formulation
data
+ Model Improvement
- Chemical mechanisms
- Regional Scale
- Aerosols
- Photolysis
-------
Feature*
Chemistry
Physics
Emissions
Melric
MIR (Carter)
- SAPRC-90
- SAPRC-97 (ongoing)
- Simple box model
- Represents urban airshed
- No transport
- Simple "meteorology"
- EKMA type
- Represents 39 cities
- Maximum change in O,
concentration
Previous Grid-Based Simulations
(McNairtt al., and Bergin el al.)
- Lurman, Coyner and Carter (LCC)
Complex 3-D grid-based domain
- Represents urban airshed
- No long-range NO, transport
- Diagnostic meteorology
- Carnegie/California Inslitulte of
Technology (CIT) with Diagnostic
Wind Model
- Detailed grid-based urban and rural
emissions from most source
categories
August 30 -September 1 , 1 982 (SC)
- August 27-29. I987(SC)
- Maximum change in peak O,
concentration(l-hr)
- Exposure above National and State
AAQS
ARB'i Grid-Based Simulations
- SAPRC-97
- possible aqueous-phase
chemistry in the near future
- Complex 3-D grid-based domain
- Represents regional airshed
- Long-range NO, transport
- Prognostic meteorology
- SAQM with MM5
- Detailed grid-based urban and
rural emissions from most
source categories
Maximum change in peak O,
concentration (1-hr, 8-hr)
- Exposure above National and
State AAQS
-Others?
Reactivity Team
+ Airshed Modeling
+ Uncertainty Analysis
- Perform overall analysis for
stationary sources
Includes uncertainties in speciation
and reactivity estimates
+ Exploring Aerosol forming
potential
- Determine implications for
different control strategies
1-22
-------
K)
UJ
External Assistance
+ Reactivity Research Advisory
Committee
+ Reactivity Scientific Advisory
Committee
Reactivity Research
Advisory Committee
+ Over 20 representatives from
industry.
+ Provide technical assistance on
reactivity related issues
consumer products and
coatings.
+ Forum to coordinate research
activities between industry and
government.
-------
to
Reactivity Scientific
Advisory Committee
Six independent, respected
scientists
- Professors John Seinfeld,
Roger Atkinson, Jack Calvert,
Harvey Jeffries, Jana Milford, and
Armistead Russell
Offer recommendations on
reactivity related scientific
issues
Development of a
Reactivity Program for
Consumer Products in
California
-------
Current Qzone Control Strategy for
Consumer Products (Mass Based)
Antiperspirant
& Deodorant
Phases I, II, III
Aerosol
Coatings
80
Categories
Regulated
VOC Limits
Innovative Product
Provision
(product-lo-product comparison)
Alternative
Control Plan
(emissions averaging)
Using Photochemical
Reactivity as a VOC Control
Strategy
+ Formed Workgroup with Industry
July 1995
+ Draw on Experience from LEV Program
+ MIR Scale Developed by Dr. Carter
+ Voluntary Alternative
+ Ensure Equivalent Ozone Reductions
1-25
-------
Main Goal: Flexibility While Reducing
Ozone Formation Potential More Efficiently
Existing Program: Decreased VOC, Decreased O3
Reformulation Costs
Water-Based System
Apparent Change
Lower Performance?
Non-Complying Complying
Reactivity Program: Decreased O3, Little/No Change in VOC
Reformulation Costs
Solvent-Based System
Transparent Switch
Higher Performance?
Non-Complying
Complying
Proposed Draft Voluntary
Regulation for Aerosol Coatings
+ Law Requires 60% Reduction in VOC
Emissions (by end of 1999)
+ Goal to Establish Equivalent Reactivity
Limits to Existing Mass-based Limits
+ Speciated Data are Necessary to Establish
Limits
1-26
-------
Hypothetical Paint Formula and
Calculation of Weighted Reactivity
Weight MIR Weighted
Contents Percent (Relative) Reactivity
acetone
toluene
propane
xylene
butane
solids
30%
20%
20%
10%
10%
10%
0
1.26
0.14
2.09
0.29
0
0.00
0.25
0.03
0.21
0.03
0.00
Total 100% 0.52
Product MIRabs = 0.52 x 4.06 = 2.11 g O3/g product
Aerosol Coatings
Schedule
Workshop, 5/19/98 | Draft Reactivity Reg. for
-<.
Aerosol Coatings
Workshop, 7/98
Workshop, Late
Summer 1998
November 1998
Draft reactivity-based
VOC limits for Aerosol
Coatings
Revised draft reactivity
reg. (if necessary)
Board Hearing
1-27
-------
O3 Equivalence to Percent
Reduction
Steps Example
(1) Determine % reduction of the* Reduce emissions by 50%
VOC limit
(2) Calculate abs. SWA-MIR for-*. MIRcat = 2.0 (g O3/ g VOC)
Aerosol Paint Category
(3) Apply % reduction to MIR of-*- Reduce MIR by 50%
category = 0.5 x 2.0 = 1.0 (g O3/ g VOC)
(4) Result = abs. wtd-MIR of -*- Wtd-MIR Limit
category equal in percent = 1.0 ( g O / g VOC)
ozone reduction from VOC 3
limit
Summary
+ ARB has successfully used
reactivity in regulations since
1990
+ ARB has an integrated program
to address challenges associated
with using reactivity in
regulations
- Research
- Internal Expertise
- External expertise
1-28
-------
VOC Reactivity - Beyond Ozone
D. Alan Hansen
EPRI
Photochemical Reactivity Workshop
Durham, NC
12-14 May 1998
1-29
-------
VOC Reactivity - Beyond Ozone
Outline
1. Purpose of talk: In keeping with the principle of integrating assessments of emissions
management across air quality issues as promoted by the FACA subcommittee on
ozone, PM and regional haze, I want to emphasize the point that considerations of VOC
reactivity in managing tropospherk ozone should also take fine particles into account.
Fine particles, of course, are largely responsible for the optical effects associated with
regional haze.
2. I will briefly summarize the interplay between VOCs and NOx in the photochemical
production of ozone and fine particles.
3. I wfll then show some modeling results that demonstrate, assuming that the models
capture the essence of the precursor chemical interactions among themselves and with
meteorology, the complexity of the responses of ozone and selected fine particle
constituents to reductions in precursor emissions.
4. I wfll finish with some issues related to VOC reactivity that should be resolved if we are
to manage tropospheric ozone and fine particles (and, by extension, regional haze)
effectively.
Chemistry
Referring to the chemical mechanism schematic (taken from the as yet unpublished
NARSTO Critical Review paper, "Oxidant Production and Fine Particles: Issues and
Needs" by Hidy, Hales, Roth and Scheffc):
1. Note precursors :JVOx, VOCs, SOj.
2. Note route to ozone: Daytime, O3->O1 D->OH->RO2->NO2->O3;
3. Note routes to fine PM:
Daytime: OH+VOC-> PM ; O3+VOC-> PM ; OH+SO2 (NH3)->PM ; OH+NO2
(NH3)->PM
Nightime: O3+olefins->OH etc.; NO3+VOO PM; N2O5 (H2O, NH3)->PM
4. Note VOCs and NOx play prominent roles in both ozone and PM formation chemistry
So what happens if we change the concentrations of VOC and NOx through emissions
changes? The picture is sufficiently complex that answering this question can rely on
environmental chamber experiments. However, these cannot account for meteorological
effects and usually cannot be conducted at precursor concentrations as low as those
prevailing in real urban and regional atmospheres. To explore how VOC reactivity
influences ozone and PM behavior under realistic conditions, we must rely on modeling.
Model Results
The modeling results shown in the four tables are taken from runs made with UAM-AERO
simulating SCAQS episodes in June and December 1987. The may not be representative of
results obtained with other models or from other geographical locations or episodes. They
1-30
-------
have been selected only to demonstrate the complex interplay between changes in
precursor NOx and VOC emissions and changes in ozone and various components of PM
2.5. In the tables PM2.5 OM denotes the "organic material" component of fine particles.
The values shown for PM2.5 and it components are averaged over the two days of each
episode. The discussion below focuses on the Percentage Change tables:
Points to note for the June 1987 episode:
1. With 50% NOx reduction:
O3 can increase substantially (>20%) or decrease (<15%)
PM2.5 NO3 decreases substantially (20-46%)
PM2.5 OM increases slightly (3-6%)
PM2.5 SO4 is relatively unaffected (±2%)
PM2.5 Mass decreases slightly (6-12%)
Question: What if the material of interest from a health effects perspective was in
the OM component, which did not decrease with a 50% NOx reduction?
2. With a 50% VOC reduction:
O3 decreases substantially (20-42%)
PM2.5 NO3 decreases slightly (6-13%
PM2.5 OM decreases slightly (8-10%)
PM2.5 SO4 is relatively unaffected (+2%)
PM2.5 Mass decreases slightly (4-9%)
3. With 50% NOx and VOC reduction:
O3 increases less than with NOx reduction only and decreases less than with VOC
reduction only.
PM2.5 NO3 decreases about the same as with NOx reduction only.
PM2.5 OM decreases less than with VOC reduction only and with no increases.
PM2.5 Mass decreases the same or slightly more than with either NOx or VOC
reduction alone.
Bottom line:
4. With 50% NOx reduction:
O3 can go up or down
PM2.5 NO3 and Mass go down
PM2.5 OM goes up
PM2.5 SO4 is relatively unaffected.
5. With 50% VOC reduction:
Everything goes down.
However, PM2.5 OM is relatively insensitive to both NOx and VOC reductions.
Points to note for the December 1987 episode:
1. With 50% NOx reduction:
O3 increases 60-138% (even the highest value, 122 ppb, more than doubles)
1-31
-------
PM2.5 NO3 increases (counterintuitively) 10-42%
PM2.5 OM increases 4-9%
PM2.5 SO4 is relatively unaffected, increasing up to 3%.
PM2.5 Mass increases 4-17%.
2. With 50% VOC reduction:
O3 decreases 0-37% (again a counterintuitive lack of response in some parts of the
domain)
PM2.5 NO3 decreases 27-42% (about the same as in summer)
PM2.5 OM decreases 5-9% (about the same as in summer)
PM2.5 SO4 is relatively unaffected, decreasing 2-5%, but with no small increases as
in summer.
PM2.5 Mass decreases 8-13%, slightly more than in summer.
3. With 50% NOx and VOC reduction:
O3 generally increases, but less than with NOx reduction alone.
PM2.5 NO3 decreases about the same as with VOC reduction alone, except the
domain max, which decreases half as much as with VOC reduction alone.
PM2.5 OM decreases slightly (3-5%).
PM2.5 SO4 is essentially unchanged.
Bottom line:
NOx reduction increases everything.
VOC reduction decreases everything.
NOx and VOC reduction decreases PM 2.5, but not O3.
V
Outstanding Questions
If higher reactivity VOCs prompt ozone formation and lower reactivity VOCs delay
ozone formation, how does the relative amount of ozone formed per carbon atom
compare?
Is there a relationship between VOC reactivity and amount of PM formed?
Does the chemical mechanism accurately reflect the SVOC produced from oxidation of
HC*s? From unsaturated oxygenates?
« Do current lumping schemes hi chemical mechanisms accurately reflect the nuances of
reactivity with respect to ozone aerosol production as well as the role of reactive
intermediates in the process?
How will the composition and rate of production of SVOC respond to changes in VOC
reactivity?
1-32
-------
How will rate of production and yield of nitrate and sulfate respond to changes in VOC
reactivity?
How will deposition of N and S respond to changes in VOC reactivity?
1-33
-------
NighlthitNSOS
Chemtrtry
Radical Pool
HO2-; RO2-
SOx
Clouds/Aqueous
Figure 9. Process diagram illustrating tropospheric chemistry pathways linking oxidant and secondary
PMi.s formation. The gas phase reactions leading to atmospheric oxidant formation, including O3, are
generally to the left and the top right of the diagram. The aerosol particle formation processes are linked
with the oxidant forming cycle, and are indicated in the middle and lower right of diagram.
f
1-34
-------
ABSOLUTE CHANGE
Jun-87
Baseline
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
DOMAIN MAX
DAY 175 DAY 176 Day 175 or 176 PM
O3 O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
404.8
451.2
272.0
337.8
353.5
300.5
230.1
261.2
48.1
33.5
42.0
33.7
19.1
19.6
17.4
17.8
32.7
32.2
32.4
32.2
100.8
93.8
91.7
91.7
HIGHEST STATION
DAY 175 DAY 176 Day 175 or 176 PM
O3 O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
Baseline
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
152.2
183.1
88.9
146.6
199.1
185.0
118.2
164.1
26.8
14.4
25.1
15.9
8.7
9.2
7.8
8.1
7.4
7.4
7.4
7.4
68.0
60.0
63.9
58.8
AVERAGE OF STATIONS
Day 175 Day 176 Day 175 or 176 PM
03 O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
Baseline
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
90.6
103.7
72.2
89.4
85.5
103.8
66.3
86.3
16.8
13.4
15.5
13.0
6.4
6.7
5.9
6.1
8.1
8.2
8.0
8.1
53.1
50.1
50.8
48.7
ABSOLUTE CHANGE
Dec-87
Baseline
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
DOMAIN MAX
Day 344 Day 345
O3
Day 344 or 345
PM
O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
90.0
185.0
90.0
104.0
122.0
280.0
80.0
108.0
61.4
87.0
35.6
49.0
37.5
40.7
34.3
35.6
26.0
26.0
25.4
25.6
190.0
223.0
166.0
167.0
HIGHEST STATION
Day 344 Day 345
Baseline
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
O3
Day 344 or 345
PM
O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
40.8
73.0
33.0
51.5
39.7
62.8
34.1
49.8
20.3
24.5
14.7
14.5
23.7
24.8
22.6
22.9
3.9
3.9
3.7
3.8
116.0
123.0
107.0
107.0
AVERAGE OF STATIONS
Day 344 Day 345
Day 344 or 345
PM
Baseline
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
03
O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
34.7
75.8
21.9
45.3
30.4
< 72.1
19.2
42.5
18.7
20.5
13.6
13.3
18.0
18.8
17.1
17.4
2.9
3.0
2.8
2.9
86.0
89.0
79.0
79.0
1-35
-------
PERCENT CHANGE
Jun-87
Baseline, ppb or ug/m3
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
DOMAIN MAX
DAY 175 DAY 176 Day 175 or 176
PM
O3 Q3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
11.5
-32.8
-16.6
-15.0
-34.9
-26.1
"-30.4
-12.7
-29.9
2.6
-8.9
-6.8
-1.5
-0.9
-1.5
-6.9
-9.0
-9.0
HIGHEST STATION
DAY 175 DAY 176 Day 175 or 176
PM
O3 O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
Baseline, ppb or ug/m3
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
20.3
-41.6
-3.7
^^^^w
-7.1
-40.6
-17.6
-46.3
-6.3
-40.7
5.7
-10.3
-6.9
0.0
0.0
0.0
-11.8
-6.0
-13.5
AVERAGE OF STATIONS
Day 175 Day 176 Day 175 or 176
PM
PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
Baseline, ppb or ug/m3
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
PERCENT CHANGE
Dec-87
Baseline, ppb or ug/m3
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
DOMAIN MAX
Day 344 Day 345
Day 344 or 345
PM
O3 O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
nrifflMBrifflffl^^ '111 ~
105.6 129.5 41.7 8.5 0.0 17.4
0.0 -34.4 -42.0 -8.5 -2.3 -12.6
15.6 -11.5 -20.2 -5.1 -1.5 -12.1
HIGHEST STATION
Day 344 Day 345 Day 344 or 345
PM
Baseline, ppb or ug/m3
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
O3
m
78.9
-19.1
26.2
O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
58.2
-14.1
25.4
20.7
-27.6
-28.6
4.6
-4.6
-3.4
0.0
-5.1
-2.6
6.0
-7.8
-7.8
Baseline, ppb or ug/m3
50% Red. NOx
50% Red. VOC
50% Red. NOX&VOC
AVERAGE OF STATIONS
Day 344 Day 345 Day 344 or 345 PM
O3 O3 PM2.5NO3 PM2.5OM PM2.5SO4 2.5 Mass
||jB|y2||jjUH^JiJR[fdH^H|^^^^^^^^^^^^^^^^IIj5g~0
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^WKli^'MtfV
118.4 137.2 9.6 4.4 3.4 3.5
-36.9 -36.8 -27.3 -5.0 -3.4 -8.1
30.5 39.8 -28.9 -3.3 0.0 -8.1
1-36
-------
Current Status of VOC
Reactivity Research
William Carter
University of California
at Riverside
Presented at:
Photochemical Reactivity Workshop
Durham, North Carolina
May 12-14, 1998
CURRENT STATUS OF REACTIVITY RESEARCH
BACKGROUND
DEFINITION OF REACTIVITY
FACTORS AFFECTING REACTIVITY
MEASUREMENT OR CALCULATION OF REACTIVITY
RESEARCH AREAS AND UNCERTAINTIES
CHEMICAL MECHANISM
STATUS OF MECHANISM DEVELOPMENT
DATA NEEDS FOR MECHANISM AND REACTIVITY
EVALUATION
UNCERTAINTY ANALYSIS
AIRSHED MODEL UNCERTAINTY
DEPENDENCE OF REACTIVITY ON ENVIRONMENTAL
CONDITIONS AND OZONE QUANTIFICAITON METHOD
-------
INCREMENTAL REACTIVITY
FACTORS AFFECTING INCREMENTAL REACTIVITY
to
NJ
Lim
OZONE
FORMED
WHEN VOC
ADDED TO
EPISODE
-
OZONE '.
FORMED
IN
EPISODE
[VOC]-» 0
[VOC ADDED]
INCREMENTAL
REACTIVITY
OF A VOC IN
AN EPISODE
NOT AN INTRINSIC PROPERTY OF THE MOLECULE.
DEPENDS ON THE EPISODE AS WELL AS THE VOC.
THIS IS THE MOST DIRECTLY RELEVANT REACTIVITY
MEASURE FOR APPLICATION TO CONTROL STRATEGIES:
CAN BE MEASURED EXPERIMENTALLY IN SMOG CHAMBERS
OR CALCULATED FOR POLLUTION EPISODES USING
AIRSHED MODELS.
NOT SAME AS OZONE PRODUCTIVITY, THE AMOUNT OF O3
ATTRIBUTABLE TO NO TO N02 CONVERSIONS CAUSED BY
PEROXY RADICALS FORMED FROM THE VOC.
INCREMENTAL
REACTIVITY
MECHANISTIC
REACTIVITY
KINETIC y
REACTIVITY
DIRECT
MECHANISTIC +
REACTIVITY
MECHANISTIC
REACTIVITY
INDIRECT
MECHANISTIC
REACTIVITY
KINETIC REACTIVITY: FRACTION OF EMITTED MOLECULE
WHICH REACTS.
PROPORTIONAL TO REACTION RATE FOR SLOWLY
REACTING COMPOUNDS
INDEPENDENT OF REACTION RATE (APPROACHES 1.0)
FOR RAPIDLY REACTING COMPOUNDS
MECHANISTIC REACTIVITY: AMOUNT OF OZONE FORMED
PER MOLECULE REACTING
DIRECT REACTIVITY ("PRODUCTIVITY"): 03 FORMED
FROM THE PEROXY RADICALS FROM THE VOC.
INDIRECT REACTIVITY: CHANGE IN O3 FORMED FROM
PEROXY RADICALS FROM THE OTHER VOCs PRESENT.
EFFECTS ON RADICAL LEVELS AFFECTS HOW
MUCH THE OTHER VOCs REACT.
EFFECTS ON NO, CONSUMPTION AFFECTS HOW
MUCH 03 IS FORMED FROM A PEROXY RADICAL.
-------
ENVIRONMENTAL FACTORS WHICH
AFFECT INCREMENTAL REACTIVITY
MEASUREMENT OR CALCULATION
OF ATMOSPHERIC INCREMENTAL REACTIVITY
K)
UJ
NO, AVAILABILITY IS MOST IMPORTANT SINGLE FACTOR
AFFECTING MECHANISTIC REACTIVITIES.
03 MOST SENSITIVE TO VOCs WHEN NO, IS HIGH, NOT
SENSITIVITY TO VOCs WHEN NO. LOW.
REACTIVITIES AT HIGH NO, ARE SENSITIVE TO
MECHANISTIC FACTORS WHICH AFFECT RATES OF O3
FORMATION (E.G. RADICAL INITIATION/TERMINATION).
REACTIVITIES AT LOW NO, ARE SENSITIVE TO
FACTORS WHICH AFFECT RATES OF NO, REMOVAL.
DURATION OF SCENARIO AND RADICAL LEVELS AFFECTS
REACTIVITIES OF SLOWLY REACTING COMPOUNDS.
SENSITIVITY TO RADICAL INITIATION/TERMINATION IS
AFFECTED BY LEVELS OF OTHER RADICAL INITIATORS.
OTHER FACTORS (E.G., SUNLIGHT AND TEMPERATURE)
AFFECT DEPENDENCE OF REACTIVITY ON NO,
REACTIVITY CAN BE MEASURED IN ENVIRONMENTAL
CHAMBER EXPERIMENTS. BUT THE RESULTS ARE NOT THE
SAME AS REACTIVITY IN THE ATMOSPHERE.
NOT PRACTICAL TO EXPERIMENTALLY DUPLICATE ALL
ATMOSPHERIC CONDITIONS AFFECTING REACTIVITY
CHAMBER EXPERIMENTS HAVE WALL EFFECTS,
USUALLY HIGHER LEVELS OF NO, AND ADDED TEST
VOC, STATIC CONDITIONS, ETC.
ATMOSPHERIC REACTIVITY MUST BE CALCULATED USING
COMPUTER AIRSHED MODELS, GIVEN:
MODELS FOR AIRSHED CONDITIONS
CHEMICAL MECHANISMS FOR THE VOC's
ATMOSPHERIC REACTIONS
CALCULATIONS OF ATMOSPHERIC REACTIVITY CAN BE NO
MORE RELIABLE THAN THE CHEMICAL MECHANISM USED.
ENVIRONMENTAL CHAMBER EXPERIMENTS ARE
NECESSARY TO TEST THE RELIABILITY OF A MECHANISM
TO PREDICT ATMOSPHERIC REACTIVITY.
-------
to
MECHANISMS FOR REACTIVITY ASSESSMENT
SAPRC-90 MECHANISM
REFLECTS KNOWLEDGE AS OF 1989.
03 PREDICTION EVALUATED AGAINST CHAMBER DATA FOR
REPRESENTATIVES OF MAJOR VOC CLASSES
HIGHLY SIMPLIFIED REPRESENTATION OF LOW-NO,
CHEMISTRY.
HIGHLY SIMPLIFIED REPRESENTATION OF HIGHER
OXYGENATED PRODUCTS.
OVER > 100 TYPES OF VOCs REPRESENTED
A FEW SIMPLE COMPOUNDS (FORMALDEHYDE,
ACETALDEHYDE, ETC.) REPRESENTED EXPLICITLY.
AROMATICS REACTIONS BASED ON PARAMETERIZED
MECHANISMS ADJUSTED TO FIT CHAMBER DATA.
ALKANE MECHANISMS GENERATED BY A COMPUTER
PROGRAM USING PUBLISHED ESTIMATION METHODS
MECHANISMS FOR MANY TYPES OF COMPOUNDS
HIGHLY APPROXIMATE AND UNTESTED.
USED TO DERIVE VARIOUS REACTIVITY SCALES FOR > 100
VOCs, INCLUDING THE WIDELY-USED MIR SCALE.
SAPRC MECHANISM UPDATES
SAPRC-93 MECHANISM
CHANGES TO PAN KINETICS CAUSED HIGHER ABSOLUTE
REACTIVITIES FOR ALMOST ALL VOCs.
ALKENE MECHANISMS CHANGED TO REFLECT NEW DATA
ON O3 + ALKENE REACTIONS
MECHANISMS FOR MTBE AND A FEW OTHER VOCs
MODIFIED BASED ON AVAILABLE DATA.
UPDATED ISOPRENE CHEMISTRY ADDED
NOW BEING USED IN SEVERAL RESEARCH-GRADE AIRSHED
MODELS
SAPRC-97 MECHANISM
AROMATICS MECHANISMS MODIFIED TO FIT NEW
CHAMBER DATA AND TO ACCOUNT FOR ISOMERIC
DIFFERENCES. MOST MORE REACTIVE.
MECHANISMS FOR A NUMBER OF VOCs UPDATED BASED
ON ONGOING REACTIVITY STUDIES
USED TO DERIVE REACTIVITY DATA AND UNCERTAINTY
SUMMARY RECENTLY PREPARED FOR THE CARB.
CURRENT WORKING MECHANISM AVAILABLE ON THE
INTERNET.
-------
SAPRC MECHANISM UPDATES
SAPRC-98 MECHANISM (UNDER DEVELOPMENT)
BASE MECHANISM HAS BEEN COMPLETELY UPDATED.
MANY SMALL CHANGES.
THE IMPORTANT OH + N02 RATE CONSTANT FOUND TO
BE HIGHLY UNCERTAIN BUT WAS NOT CHANGED.
MORE DETAILED REPRESENTATION OF LOW NOX ORGANIC
REACTIONS. CHANGES IN PRODUCT DISTRIBUTION AT
LOW NO. CAN NOW BE PREDICTED.
ESTIMATED MECHANISMS FOR ALKANES, ALKENES, AND
to MANY OXYGENATES ARE GENERATED AS FOLLOWS:
I
COMPUTERIZED ESTIMATION PROCEDURE GENERATES
EXPLICIT MECHANISMS WHICH ARE USED TO DERIVE
PRODUCT YIELD PARAMETERS FOR THE MODEL.
PROCEDURE USES ESTIMATED OR ASSIGNED RATE
CONSTANTS FOR THE COMPETING REACTIONS.
REPRESENTATION OF ORGANIC PRODUCTS BEING
UPDATED USING PREDICTED PRODUCT DISTRIBUTIONS.
STILL NECESSARY TO USE PARAMETERIZED MECHANISMS
FOR AROMATICS ADJUSTED TO FIT CHAMBER DATA.
MECHANISM IS INCORPORATING RESULTS OF RECENT
STUDIES OF CONSUMER PRODUCT AND OTHER VOCs.
STATUS OF MECHANISM DEVELOPMENT
AND UNCERTAINTIES BY VOC CLASS
ALKANES
MECHANISMS FOR LOWER ALKANES WELL ESTABLISHED,
ESTIMATION METHODS USED FOR HIGHER ALKANES.
THE CB+ N-ALKANE MECHANISMS WHICH FIT CHAMBER
DATA HAVE UNREASONABLE ASSUMPTIONS.
MINERAL SPIRITS DATA SUGGEST REACTIVITIES FOR C10,
BRANCHED AND CYCLIC ALKANES ARE OVERESTIMATED.
ALKENES
SAPRC-98 EVALUATION SHOW MORE PROBLEMS WITH
ALKENE MECHANISMS THAN PREVIOUS SUSPECTED.
MODELS USING ACCEPTED OH YIELDS FOR O3 REACTIONS
GREATLY OVERPREDICT REACTIVITIES OF C4< 1-ALKENES.
UNCERTAIN 0(3P) REACTIONS AFFECT MECHANISM
ADJUSTMENTS FOR PROPENE, BUTENES, AND ISOPRENE.
EXTENT TO WHICH MECHANISMS MODIFICATIONS WILL
AFFECT ALKENE REACTIVITY IS UNCERTAIN.
-------
STATUS OF MECHANISM DEVELOPMENT
AND UNCERTAINTIES BY VOC CLASS
AROMATICS HYDROCARBONS
STILL NECESSARY TO USE PARAMETERIZED MECHANISMS.
YIELDS AND PHOTOLYSIS RATES OF UNCHARACTERIZED
PRODUCTS CANNOT BE UNAMBIGUOUSLY DETERMINED.
NO MECHANISM CAN SATISFACTORILY FIT ALL CHAMBER
DATA FOR BENZENE.
NO OBVIOUS EXPLANATION FOR LOWER MECHANISTIC
REACTIVITY FOR ETHYLBENZENE COMPARED TO TOLUENE.
CURRENT MECHANISMS PROBABLY ARE INCONSISTENT
WITH PRODUCT DATA FOR REACTIONS OF PHENOLS
UNCERTAIN WHETHER PARAMETERIZED MECHANISMS
EXTRAPOLATE CORRECTLY TO LOW NO, CONDITIONS.
HIGHER OXYGENATES (HIGHER KETONES.
ETHERS, ESTERS. GLYCOLS, ETC.)
EXPERIMENTAL REACTIVITY DATA ARE BECOMING
AVAILABLE, SIGNIFICANTLY REDUCING UNCERTAINTIES.
CURRENT ESTIMATION METHODS OFTEN PERFORM POORLY
IN SIMULATING CHAMBER DATA PRIOR TO ADJUSTMENTS.
ATTEMPTS TO IMPROVE PERFORMANCE OF ESTIMATION
METHODS ARE UNDERWAY.
STATUS OF MECHANISM DEVELOPMENT
AND UNCERTAINTIES BY VOC CLASS
HALOGENATED COMPOUNDS
REACTIVITY DATA ONLY AVAILABLE FOR CHLOROPICRIN
(CCI3NO2), TRICHLOROETHYLENE, AND ALKYL BROMIDES.
NO REASONABLE MECHANISM SATISFACTORILY FITS ALL
CHAMBER DATA FOR TCE AND ALKYL BROMIDES.
STUDIES ARE NEEDED ON SIMPLER SYSTEMS.
NITROGEN-CONTAINING COMPOUNDS
REACTIVITY DATA LIMITED TO N-METHYL PYRROLIDINONE
(NMP) AND SEVERAL AROMATIC ISOCYANATES.
NMP IS UNUSUAL IN THAT NO3 REACTIONS CONTRIBUTE
TO ITS REACTIVITY.
THE AROMATIC ISOCYANATES STUDIED DO NOT PROMOTE
OZONE FORMATION. MECHANISM UNKNOWN.
SILOXANES
CHAMBER DATA SHOW THAT THESE ARE O3 INHIBITORS,
BUT MECHANISMS WHICH FIT CHAMBER DATA ARE NOT
CONSISTENT WITH RESULTS OF PRODUCT STUDIES.
-------
to
MECHANISM UNCERTAINTY ANALYSIS
REACTIVITY-BASED CONTROL STRATEGIES WILL PROBABLY
NEED TO TAKE INTO ACCOUNT VARYING LEVELS OF
UNCERTAINTIES FOR DIFFERENT VOCs.
PROPOSALS TO USE ADJUSTMENT FACTORS OR UPPER
LIMITS FOR UNCERTAIN VOCs IN REACTIVITY-BASED VOC
CONTENT REGULATIONS.
UNCERTAINTY ANALYSIS APPROACHES
FORMAL UNCERTAINTY ANALYSIS
ULTIMATELY THE BEST APPROACH, BUT HAS ITS OWN
UNCERTAINTIES.
RELIES ON SUBJECTIVE UNCERTAINTIES FOR INPUT
DATA. INCONSISTENCES AMONG EVALUATORS.
DIFFICULT TO TREAT POSSIBILITIES OF FOR "MISSING-
REACTIONS OR INCORRECT PARAMETERIZATIONS.
NOT PRACTICAL TO DO FOR ALL TYPES OF VOCs IN
USEFUL TIME FRAME.
NEAR-TERM UTILITY IS TO AID EVALUATION OF
SUBJECTIVE OR CATEGORIZATION APPROACHES.
PROJECT UNDERWAY TO ANALYZE UNCERTAINTIES IN
MECHANISMS ADJUSTED TO FIT CHAMBER DATA.
UNCERTAINTY ANALYSIS APPROACHES
(CONTINUED)
CATEGORIZATION BASED ON EXPERT ASSESSMENT OF
QUALITY OF MECHANISM AND EXTENT TO WHICH
MECHANISM EVALUATED.
PRELIMINARY CATEGORIZATION HAS BEEN DONE FOR
ALL VOCs IN THE SAPRC-97 MECHANISM.
NEED TO BE UPDATED AND PEER-REVIEWED BEFORE
INCORPORATED IN ANY REGULATIONS.
DOES NOT GIVE NUMERICAL UNCERTAINTIES.
UPPER AND LOWER LIMIT REACTIVITY ANALYSIS
CAN BE USED FOR QUANTIFYING UNCERTAINTIES FOR
ALL VOCs.
RELATIVELY STRAIGHTFORWARD TO ESTIMATE UPPER
LIMIT REACTIVITIES FOR A GIVEN SCALE. PROPOSED
APPROACH HAS BEEN DEVELOPED.
LOWER LIMIT REACTIVITIES FOR VOCs OF UNKNOWN
MECHANISM IS ZERO, SINCE THEY MAY INHIBIT 03.
THIS METHOD GIVES HIGH UNCERTAINTY RANGES.
MAY NOT BE ACCEPTABLE FOR REGULATORY USE.
UNCERTAINTY RANGES FOR SOME VOC CLASSES CAN
BE NARROWED BY MECHANISTIC CONSIDERATIONS.
-------
to
oo
CHEMICAL MECHANISM UNCERTAINTIES
BASE MECHANISM (INORGANIC, COMMON PRODUCT
REACTIONS) HAS NON-NEGLIGIBLE UNCERTAINTIES. '
REACTIVITIES VOCs WITH LARGE INDIRECT
REACTIVITIES (E.G., INITIATORS/INHIBITORS)
SENSITIVE TO BASE MECHANISM CHANGES.
REACTIVITY UNCERTAINTIES FOR WELL-STUDIED
VOCs ESTIMATED TO BE -30%
UNCERTAINTIES IN MECHANISMS FOR INDIVIDUAL VOCs
CAN BE MUCH GREATER IF VOC INADEQUATELY STUDIED.
ONGOING RESEARCH IS REDUCING NUMBER OF VOC
CLASSES WITH INADEQUATE DATA.
UNSTUDIED VOCs MORE OF A CONCERN FOR
STATIONARY SOURCES THAN MOBILE SOURCES.
REACTIVITY CHANGES DUE TO UPDATING MECHANISM
GIVE AN INDICATION OF UNCERTAINTIES
REFLECTS RESULTS OF ONGOING RESEARCH.
CHANGES FOR WELL-STUDIED CONSISTENT WITH
-30% MINIMUM UNCERTAINTY ESTIMATE.
TYPES OF ENVIRONMENTAL CHAMBER
EXPERIMENTS CURRENTLY USED
TO TEST CHEMICAL MECHANISMS
SINGLE VOC-NO.-AIR RUNS:
MOST STRAIGHTFORWARD TEST OF A VOC's
MECHANISM, THOUGH ONLY USEFUL FOR VOCs WITH
RADICAL SOURCES.
NOT A "REALISTIC" ENVIRONMENT. CORRELATES
POORLY WITH REACTIVITY.
COMPLEX MIXTURE-NO.-AIR RUNS:
TESTS MECHANISMS' ABILITY TO SIMULATE O3
FORMATION UNDER REALISTIC CONDITIONS
NOT USEFUL FOR MECHANISM DEVELOPMENT
REACTIVITY EXPERIMENTS (MIXTURE-NO.-AIR COMBINED
WITH MIXTURE-NOX-AIR RUNS WITH TEST VOC ADDED):
CAN TEST MECHANISMS OF SINGLE VOCs UNDER
REALISTIC CONDITIONS
BEST TEST OF MECHANISM'S ABILITY TO PREDICT
INCREMENTAL REACTIVITY
NOT SAME AS ATMOSPHERIC REACTIVITY.
-------
MS
DATA NEEDS FOR MECHANISM EVALUATION
(NEAR TERM)
MECHANISM EVALUATION DATA NEEDED FOR CLASSES OF
COMPOUNDS NOT PREVIOUSLY STUDIED.
GOOD PROGRESS BEING MADE FOR SOLVENT SPECIES
SUCH AS ESTERS, GLYCOLS, ETC.
BUT EPA EXEMPTION POLICY HAS CAUSED FOCUS OF
RESEARCH TO BE ON LOW-REACTIVITY COMPOUNDS.
REACTIVITY-BASED CONTROLS WILL ENCOURAGE
RESEARCH ON COMPOUNDS OF ALL REACTIVITIES.
BETTER METHODS NEEDED TO EVALUATE REACTIVITY
HIGH COST OF OBTAINING REACTIVITY DATA LIMITS
ACCEPTABILITY OF REACTIVITY-BASED CONTROLS.
CURRENTLY NO WAY TO ASSESS REACTIVITIES OF
VERY LOW VOLATILITY COMPOUNDS.
ONLY A FEW LABORATORIES ARE PRESENTLY
CAPABLE OF GENERATING REACTIVITY DATA.
DEVELOPMENT OF NEW REACTIVITY
MEASUREMENT METHODS
PROJECT UNDERWAY TO DEVELOP NEW REACTIVITY
MEASUREMENT METHODS.
INITIAL FOCUS IS ON USE OF HONO/VOC STIRRED FLOW
SYSTEM. CALCULATIONS INDICATE THIS CAN GIVE
USEFUL DATA ON FACTORS AFFECTING REACTIVITY.
LOW TO MODERATE VOC TO HONO RATIOS: DATA
SENSITIVE TO kOH AND NO TO NO2 CONVERSIONS.
HIGH VOC TO HONO: ALSO SENSITIVE TO RADICAL
TERMINATION EFFECTS.
POTENTIALLY LOWER COST WAY TO OBTAIN DATA FOR
MECHANISM EVALUATION, REACTIVITY SCREENING,
DERIVING EMPIRICAL REACTIVITY-RELATED PARAMETERS.
FLOW SYSTEM POTENTIALLY ADAPTABLE TO VERY LOW
VOLATILITY COMPOUNDS
CLEAN HONO GENERATION SYSTEM HAS BEEN
CONSTRUCTED. FLOW SYSTEM BEING CONSTRUCTED FOR
INITIAL TESTING WITH PROPANE.
-------
N)
O
DATA NEEDS FOR MECHANISM EVALUATION
(LONGER TERM)
MAJOR INVESTMENT IN CHAMBER FACILITIES NEEDED TCI
IMPROVE EVALUATION OF EXISTING MECHANISMS
MECHANISMS INADEQUATELY EVALUATED FOR LOW
NO, (REGIONAL OR NEAR-ATTAINMENT) CONDITIONS.
CHAMBERS CURRENTLY USED FOR MECHANISM
EVALUATION UNSUITABLE FOR LOW NO. STUDIES.
ANALYTICAL CAPABILITIES AT OPERATING CHAMBER
FACILITIES NOT ADEQUATE FOR FULL MECHANISM
EVALUATION OR DETERMINING ALL VOC IMPACTS.
TEMPERATURE EFFECTS UNCERTAIN. CURRENT
CHAMBERS INADEQUATE TO STUDY THIS.
LARGE TEMPERATURE-CONTROLLED INDOOR
CHAMBER NEEDED TO STUDY PARTICULATE
FORMATION UNDER CONTROLLED CONDITIONS.
STUDIES FOCUSED ON SPECIFIC COMPOUNDS CANNOT BE
USED TO FUND THE NEEDED FACILITY IMPROVEMENTS.
AIRSHED MODEL UNCERTAINTIES
UNCERTAINTIES IN REPRESENTATION OF A GIVEN
SCENARIO. (EMISSION UNCERTAINTIES, ETC.)
LESS OF A PROBLEM FOR GENERAL SCALES
REPRESENTING A RANGE OF CONDITIONS
USE OF SIMPLIFIED PHYSICAL SCENARIOS (EKMA MODELS)
FOR COMPUTATIONAL TRACTABILITY
LESS OF A PROBLEM FOR GENERAL SCALES
REPRESENTING A RANGE OF CONDITIONS
STUDIES SUGGEST NOT A MAJOR PROBLEM WHEN
PREDICTING REACTIVITIES RELATIVE TO O3 EXPOSURE
UNCERTAINTIES IN DISTRIBUTION OF CONDITIONS
RELEVANT TO ASSESSING OZONE CONTROL
NOT ADEQUATELY STUDIED. EPA SCENARIOS USED
BY CARTER (1994) WERE NOT DEVELOPED FOR
REACTIVITY ASSESSMENT.
MAJOR PROBLEM FOR DEVELOPING GENERAL SCALES
REPRESENTING A RANGE OF CONDITIONS.
LACK OF ADEQUATE STUDIES OF INCREMENTAL
REACTIVITIES IN REGIONAL SCALE MODELS
IMPORTANT WHEN ASSESSING WHAT IS "NEGLIGIBLE-
REACTIVITY.
-------
APPROACHES FOR DEALING WITH DEPENDENCE
OF REACTIVITY ON AIRSHED CONDITIONS
AND OZONE QUANTIFICATION METHOD
USE A "REPRESENTATIVE" OR "WORST CASE" EPISODE.
MAY NOT BE OPTIMUM FOR ALL CONDITIONS.
BASE THE SCALE ON CONDITIONS WHERE VOCs HAVE
MAXIMUM INCREMENTAL REACTIVITIES (MIR SCALE).
REFLECTS CONDITIONS MOST SENSITIVE TO VOCs
AND CORRELATES WITH EFFECTS ON 03 EXPOSURE.
BUT DOES NOT REPRESENT CONDITIONS WHERE
HIGHEST OZONE CONCENTRATIONS ARE FORMED.
USE MULTIPLE SCALES REPRESENTING THE RANGE OF
APPLICABLE CONDITIONS.
ALLOWS ASSESSMENT OF EFFECTS OF VARIABILITY
BUT NOT USEFUL WHEN SINGLE SCALE REQUIRED.
USE A SCALE OPTIMIZED FOR A RANGE OF CONDITIONS.
REQUIRES IMPROVED ASSESSMENT OF RANGE OF
CONDITIONS RELEVANT TO OZONE FORMATION
REQUIRES AN OBJECTIVE DEFINITION OF "OPTIMUM"
HAS NOT RECEIVED ADEQUATE ATTENTION TO DATE.
APPROACHES FOR DEALING WITH DEPENDENCE
OF REACTIVITY ON CONDITIONS
(CONTINUED)
CARB VEHICLE REGULATIONS USE THE MIR SCALE, BASED
ON PEAK 03 IN EKMA SCENARIOS WITH NO, ADJUSTED TO
GIVE MAXIMUM SENSITIVITY OF 03 TO VOCs.
VOC EXEMPTION PROPOSALS HAVE USED DISTRIBUTIONS
OF INTEGRATED AND PEAK 03 REACTIVITIES IN THE 1-DAY
EKMA SCENARIOS, AND OTHER CONSIDERATIONS.
IF THESE METHODS CHANGE, IT MAY CHANGE REACTIVITY
SCALES MORE THAN UPDATES IN MECHANISM OR MODELS
POLICY ISSUES
HOW SHOULD OZONE IMPACTS BE QUANTIFIED?
WHAT CRITERIA SHOULD BE USED TO DETERMINE
WHAT IS AN OPTIMUM REACTIVITY SCALE?
WHAT ARE THE MOST APPROPRIATE ENVIRONMENTAL
CONDITIONS TO USE WHEN ASSESSING REACTIVITY?
SCIENTIFIC CHALLENGE IS TO DERIVE SCENARIOS, MODELS
AND PROTOCOLS BEST ADDRESSING POLICY PRIORITIES.
THE MIR SCALE HAS BECOME THE DEFAULT. IF NOTHING
IS DONE, IT WILL CONTINUE TO BE USED.
-------
INFORMATION AVAILABLE ON THE INTERNET
REACTIVITY TABULATIONS AND UNCERTAINTY
CLASSIFICATIONS:
http://cert.ucr.edu/ ~ carter/rcttab.htm
REPORTS ON RECENT REACTIVITY AND CHAMBER STUDIES
AND SAPRC-97 MECHANISM DEVELOPMENT:
http://cert.ucr.edu/ ~ carter/bycarter.htm
K> SAPRC-97 MECHANISM:
t»
to
http://cert.ucr.edu/~carter/saprc97.htm
CHAMBER DATA BASE FOR MECHANISM EVALUATION
(THROUGH 1995):
ftp://cert.ucr.edu/pub/carter/chdata/
SOFTWARE FOR REACTIVITY CALCULATION AND
MECHANISM EVALUATION
ftp://cert.ucr.edu/pub/carter/model/
-------
VOC Reactivity Quantification
Methods, Uncertainties and
Variabilities
Jim Wilkinson, YJ Yang, M. Kahn, Lewis Qi, Ted
Russell and others
School of Civil and Environmental Engineering
Georgia Institute of Technology
trussell@pollution.ce.gatech.edu
Georgia Institute of Technology
Issues/Outline
* What are we trying to do and why?
* What have we done
* Available/future methods
* Uncertainties
* Variabilities
Georgia Institute of Technology
2-13
-------
What we are trying to do
* Quantify reactivity of VOCs
* Understand and quantify uncertainties
* Understand variabilities
* Address related issues
Georgia Institute of Technology
V
* Save money
- Provide incentives to save even more
* Protect human health
- Relevant reactivity measures
Georgia Institute of Technology
2-14
-------
Approach(es)
Air quality modeling
- Box modeling
- Three-dimensional
Reactivity quantification
- Brute force
- Direct sensitivity analysis
* DDM-3D
Uncertainty assessment
- Monte Carlo and other
Variability analysis
- Multiple domains
- Multiple periods
- Multiple endpoints
50-200
Air Quality
Model
Atmospheric Diffusion Equation
Discretize
~^ + L(x,Oc-f(x.t)
I
^ Operator splitting
c(t-v2Al) -L,;At)L.. (At! Ls/;2.A'-i I.yi All Lt(Ai
Georgia fnsttiute of Technology
AQM Reactivity Quantification
Methods
* Brute force
- Run base case
- Perturb inventory
» Add dE, of species in emissions
* Find d[O3]
* Reactivity: 1^ = d[03]/dEj
» Problems
- Tedious
- Numerical errors
* Direct sensitivity analysis
- Find d[O3]/dEi directly using DDM
- Multiple reactivities simultaneously
- Not as prone to numerical errors
Georgia InsiUule of Technology
2-15
-------
Brjrte Force
O^t.x.y.z)
NO(t,x,y,z)
VOCjt,x,y,z)
Oj(t,x.y.z)
NO'(t,x,y,z)
N02(l,x,y,z)
Georgia Institute of Technology
Sensitivity (DDM-3D)
03(t,x,y,z)
N0(t,x,y,z)
|f N02(t,x,y,z)
-'
Georgia Institute of Technology
2-16
-------
Sadies
1
* California LEV/Clean Fuels Assessment
- Assess MIR reactivity weighting of exhaust emissions
* Auto/Oil & NREL reactivity quantification and uncertainty
assessment
- Compare box and airshed model reactivities
- 3-D modeling of reactivity and 1&3 D modeling of uncertainties
* National Aerosol Association relative reactivity study
- Economic assessment of using reactivity in control strategies
* National Science Foundation
- Developed and applied DDM-3D, multi-domain analyses-
- 3D Monte Carlo reactivity uncertainty assessment
< Others
- Solvent studies, variability analyses, regional domains
Georgia Institute of Technology
Arjshed Vs. Box Model Reactivities
^p|s$$:"'
* Compared alternative fuel MIRs, box
model and L.A.-based airshed modeling
Results usually
similar, but aromatics,
in particular, can differ.
- , . wn .
1
^^ ..^
J
m i
QL A. Peak
Box
8 t 8 § I .8
5 ° £ a 3
1 ! ! s
Compound
Georgia Institute of Technology ^~
2-17
-------
Mjj&ric Differences
* Compared peak ozone reactivity Vs.
exposure-based reactivities
Compound
Georgia Institute of Technology
* Slightly negative
Dijnain Differences
* Compared Peak 1-hr reactivities calculated
for L.A., Swiss Plateau, Mexico City
0.01
Georgia InsliluU of Technology
1 I 1 I I I
PS S? £ >- f,
I a s
8 o- <
Compounds
*Slightly negative
2-18
-------
lities : Box Model Analysis
* Compared net reactivity with relative exhaust
reactivity (RAF, in each domain)
- Little variability in RAF, lots in net reactivity
Absolute Reactivity
Relative Reactivity
f § "
o E J-°
03 01 X
IE
Gtorgia Institute of Technology
MS5 LPG Phiist 2
CNO E85 RPA
1.0
LL
CC
MS5 CNG LPC E85 Pluist :
Fl'EL
Uncertainties
;;"
:» Conducted uncertainty analysis using 3-D
airshed (L.A. case)
* Perturbed rate constants and product splits
* Re-normalized reactivities
* Found about a 15-30% uncertainty
- Species dependent
Georgia Institute of Technology
2-19
-------
a
o
u
OS
35-
3-
as--
2
1.5
Figure 2c. Btect of 2o Rate Constant Pwtuibatlont on Normalized
Reactivities Bawd on Peak Ozone
o NO2»OH
x N02 Pholdyiit
A Aldehyde Photdysls
x Peroxyacylt + NO
x PeroxyacyU * NO2
» O3» NO
- PAN, PPN Decanpoilllvi
3D Monte Carlo Reactivity
^Uncertainty Assessment
* 3D Monte Carlo assessment of
uncertainties
- Los Angeles basin
- Chemical mechanism and emissions
uncertainties
- Direct sensitivity
- Found reactivity uncertainties
* Species and metric dependent
* Spatial variations in reactivities and uncertainties
Georgia Institute of Technology
2-20
-------
\ Methods
\
* Urban/Regional air quality models
- Advanced chemistry (SAPRC97+)
* Direct sensitivity analysis
- Faster, more accurate
* Uncertainty analysis approaches
- Guided Monte Carlo and others
Georgia Institute of Technology
Direct Sensitivity Vs. Brute
Force
* Direct sensitivity provides a more rapid,
accurate approach
DDM Reactivities Vs. Brute Force
Georgia Institute of Technology
Fuel .-Species
2-21
-------
Economic Assessment (Back to Why)
^
* Compared costs of reactivity and mass-based
control strategies
- Mixed integer, non-linear programming of control
cost effectiveness in Los Angeles
- Understates potential savings (reformulations)
I.i
o
O
Reactivity-based
ings
Georgia Institute of Technology % flf "
02006
; Midway Summary: Use of Reactivity
~--<
* Scientifically compelling
- Significant differences in ozone impacts
- Still some issues to resolve
* Economically compelling
- Significant economic benefits
- More to be found if incentives are provided
Georgia Institute of Technology
2-22
-------
Gaps in our understanding
* Regional reactivity assessments
- Little work on eastern U.S., Texas, etc.
- Assess variabilities
* Episodic reactivity Vs. longer term impact
* More comprehensive uncertainty analysis
Georgia Institute of Technology
Looming Issues
* Metric(s) of importance
- Peak 8-hr: Standard
- Human exposure: Health
- Regional exposure: welfare
* Most likely of less relevance
- NOx limitations
Georgia Institute of Technology
2-23
-------
Regional Reactivity Assessment
Texas-Mexico Border
Sensitivity of ozone to butane
(preliminary study underway)
Eastern U.S.
(To be done)
Georgia Institute of Technology
Related Issues
* Secondary PM formation
- Can we develop a PM-formation potential
scale
- Will it be similar to ozone formation potential
(reactivity) scale(s)
* No
Georgia Institute of Technology
2-24
-------
^ Exemptions
\
Bad Good
* Mistakes have been made
* Traffic can go both ways
- Highly reactive ==> less reactive exempt Reactivity
- Less reactive => marginally reactive exempt
: Does not account for change in mass emissions
- Solvent/propellant changes can impact mass emissions
» Twice as much of something half as reactive is not a good
deal
* Reactivity scale scientifically more sound
* Less policy inertia
- Do not have to "unexempt" a compound
* Regulatory uncertainty can be expensive
Georgia Institute of Technology
Research Needs
ff
* Regional reactivity assessments
* Further sensitivity/uncertainty analysis
- Regional scale
- Updated chemistry (e.g., SAPRC98?)
* Further species mechanism development
* Assess dependence on metric
- Peak Vs. exposure
- Episodic Vs. long term
* PM impact scale
Georgia Institute of Technology
2-25
-------
\ Summary
V.
* Investigated reactivity quantification issues:
- Fuels, Solvents, Propellants
- Methods
* Air quality models and direct sensitivity analysis
- Uncertainties
- Variabilities
* Uncertainties reasonably small
30% by species (depending on metric), less by source
* Variabilities decrease using relative reactivities
* Scientifically and economically compelling
* Brightline approach is flawed
Georgia Institute of Technology
Acknowledgements
\
* Jana Milford, Michelle Bergin, Bart Croes,
Bill Carter, Basil Dimitriades and Lauri
McNair
* California Air Resources Board, National
Science Foundation, National Aerosol
Association, Occidental Chemical
Company, Coordinating Research Council,
Georgia Power, CONACyT, NREL
Georgia Institute of Technology
2-26
-------
Quantification of Uncertainties in Reactivity Estimates
for Volatile Organic Compounds
Presented by
Jana B.Milford
Co-authors
Michelle Bergin and Lihua Wang
Department of Mechanical Engineering
University of Colorado, Boulder 80309
Abstract
The research to be presented quantifies uncertainties in air quality model-based estimates of
absolute and relative reactivities for volatile organic compounds (VOCs). Monte Carlo
techniques have been used to propagate estimates of uncertainty in model inputs and parameters
V
to generate confidence intervals for estimates of VOC reactivity. In this presentation, previously
published estimates of reactivity uncertainties due to chemical parameters alone (Yang et al.3
1995; 1996a;b; Bergin ct al., 1998) will be compared to new estimates that also account for
uncertainties in emissions and meteorological conditions for a specific time and location, namely
August 27-28, 1987, in California's South Coast Air Basin. For selected aromatic compounds,
new estimates of uncertainties attributable to smog chamber-derived reaction parameters will
also be presented. The results suggest priorities for future research to reduce modeling
uncertainties, but also indicate that the effect of existing uncertainties can be minimized by
formulating reactivity policies in terms of relative reactivity estimates, as opposed to absolute
reactivities. Implications of modeling uncertainties for a prospective photochemical reactivity
policy, and recommendations for additional research, will be discussed.
2-27
-------
COMPARISON OF PHOTOCHEMICAL OZONE CREATION POTENTIALS
CALCULATED USING A MASTER CHEMICAL MECHANISM WITH THE MIR
REACTIVITY VALUES FOR UP TO 120 ORGANIC COMPOUNDS
Dick Derwent
Atmospheric Processes Research
Meteorological Office
Bracknell
Berkshire
United Kingdom
ABSTRACT
Photochemical ozone creation potentials POCPs for 120 organic
compounds have been calculated with a photochemical trajectory
model for realistic European conditions. The model employs a
Master Chemcial Mechanism (Jenkin et al. 1997) containing 2410
chemical-species and over 7100 chemical reactions. POCPs provide
an estimate of the likely contribution to European regional scale
ozone formation over a five day timescale from unit mass emission
of each organic compound relative to ethylene. Photochemical PAN
creation potentials have also been estimated. The POCP values
have been carefully compared with the corresponding MIR
incremental reactivities of Carter et al. (1995), developed for
the single photochemical day situation appropriate to the Los
Angeles airshed. For the vast majority of organic compounds,
there is a clbse functional relationship between POCP and MIR
values. Some differences are apparent between the single day and
multi-day indices and they provide important insights into the
quantification of reactivity. Organic compounds which generate
specific highly unreactive organic compounds in their degradation
schemes tend to show lower POCPs relative to their MIR values.
Such unreactive compounds include: acetone, alkyl nitrates and
formate esters. Furthermore, aldehydes, and in particular
formaldehyde, show lower POCPs in multi-day situations where OH-
degradation predominates over photolysis compared with single day
or smog chamber situations. Work is in hand with the Master
Chemical Mechanism to improve the representation of the
degradation schemes for aromatic hydrocarbons and incorporate the
recent mechanistic work by Jeffries et al. (1997) to improve the
reliability of the POCP estimates for this important group of
organic compounds.
2-28
-------
to
tb
MODELLING OZONE FORMATION
FOR POLICY FORMULATION
Models are required to
* quantify transbbundary transport
* define what is needed to meet environmental criteria
* define the balance between emission controls in the UK
and beyond
* assess the rotes of VOC and NOx controls
* examine the different VOCs and VOC emitting sectors
MO PHOTOCHEMICAL TRAJKTORY MODEL
-------
EMEP
1991
UKNAEI
1W1
10km x10km
FIGURE 2. A diacruamibc representation of the nesting between the emissions
Ifnds employed in the photochemical trajectory model
ORGANIC COMPOUNDS DEGRADED
MASTER CHEMICAL MECHANISM
ALKANES
ALKENES
DIALKENES
ALKYNES
AROMATICS
ALDEHYDES
KETONES
ALCOHOLS
ETHERS
ESTERS
ORGANIC ACIDS
CHLOROCARBONS
total
22
15
2
1
18
6
10
17
10
8
3
8
C1-C12
C2-C6
C4-C5
C2
C6-C11
C1-C5
C3-C6
C1-C6
C2-C7
C2-C6
C1-C3
C1-C2
120 organic compourKJs
2-30
-------
DEVELOPMENT OF MASTER CHEMICAL
MECHANISM
to
7000 chemical reactions
2500 chemical species
120 emitted organic species
Master Chemical Mechanism development uses:
* ACCORD for EXCEL as a chemical spreadsheet
* FACSIMILE as a variable order Gear's method
For example, the butane scheme contains 510 chemical
reactions and 186 chemical species (of which 20 are
primary emitted species)
Following processes initiate ozone production from the
organic compounds:
OH radical attack
* ozone reactions
N03 radical attack
* photolysis
* Q atom attack
Available on WWW web page:
TION
* 1
PHOTOLYSIS
Gtrtonyli
ROOM RC(O)OOH
ndRONO!
OH rtaclion
MVOCm)
oiyg«nil*0 produdi
+
«,«.
r^,rr,
1
OXY(BO)
rSr
L
X
PEROXY (ROi)
nu wjj r«jj
HC>2 Rt),
i
DO r»»ct*on
pnxJoct*
EicMOCRlEGEE
(RRtX)0*)
\ SUMlUIIW
KwxxiipoMie
C*itnny«
RC(O)OOH R
m
PANl.
XDH ROH
C(O)OH CO
'
(RR-COO)
SOj NO NOj
CO,
FIGURE 1. Summary of chemistry of organic species considered in the
mechanism construction protocol (JenMn et al., 1997).
WWWhttp://chem.leeds.ac.uk:80/Atmospheric/MCM
-------
DESCRIPTION OF CHAMBER-INDEPENDENT
ATMOSPHERIC CHEMISTRY MECHANISMS FROM
SMOG CHAMBER DATA
U)
SMOG CHAMBER DATABASES
1. Dual outdoor smog chamber data (University of North
Carolina)
2. Indoor teflon chamber (SAPRC)
3. Indoor evacuable chamber (SAPRC)
4. Dual outdoor chamber (CSIRO)
INPUT DATA REQUIRED
a. concentration measurements
b. chamber-dependent photolysis rates
c. smog chamber auxiliary mechanism
d. evaluated mechanistic data
OUPTUT ATMOSPHERIC CHEMISTRY MECHANISMS
i Carbon Bond Mechanisms (CBM-IV)
ii. CAL, SAPRC-90
iii. Generic Reaction Set
-------
140-1
120-
100-
COMPARISON OF THE MCM WITH CBM-IV
MECHANISM
It is concluded that the differences between CBM-IV and
the MCM for the estimation of daytime photochemical
ozone production are small, within +-6 ppb in about 100
ppb.
There is a significant difference in the treatment of tow-
NOx nighttime chemistry wtthin the MCM.
This dose correspondence may not follow for all the
secondary products and free radicate. However, this is
the reason why the MCM has been developed.
Comparison between the MCM and the smog chamber database
Travel time, hours
FIGURE 2. Comparison between the explicit chemkal mechanisms: MCM and
DJ&S and the smog chamber mechanisms: CBM-IV and CAL, and the ozone
concentrations produced along the five day trajectory.
2-33
-------
Rtductlort In Oiont for 35% reductloi In NO« and HC millions plotted coalnit OVNOx
Hydrocarbons »nd Irantporl of ozone and PAN
1669
Fig. 2(b). The itme development of ozone in the F.R.G.-Rcpublic of Ireland trajectory case.
2-34
-------
1 AHU. 1 Itx:)1 and I'PCI' value* for I20 orpiimc orxniKuniV Uolcrmini-U wiih ihr MCM nnd tho UK Photochemical Trnj<«clory Mixlol
Organic compound
AlxanM
methane
ethane
propane
n-buLane
i-bulane
n-pentane
i-penlane
neopenlane
n-hexane
2-melhylpenlane
3-methylpentane
2 2-djmethylbutane
2,3-djmethylbuume
n-heptane
2-methylhexane
3-methylhexane
n-octane
n-nonane
n-decane
n-undecane
n-dodecane
TOC I1
06
123
176
352
307
395
405
17.3
482
120
479
24 1
54 1
494
41 1
36.4
463
41 4
384
38.4
367
PPCF
09
173
137
31 4
117
297
427
67
448
294
666
163
634
61.9
31.9
40.1
42.9
349
2S.S
29 1
372
Organic compound
DiaLXonw
1.3-buladiene
isoprene
Alkvnes
acetylene
Aromalica
benrene
toluene
oxylene
m-xylene
p-xylene
ethyloenzene
propylbenrene
i-propylbenzene
1,2,3-lrimethyIbeniene
1.2,4-tnmethylbeniene
1.3,6-lnmethylb«niene
o-ethyltoluene
m-ethyltoluene
p-ethyltoluene
3, 6-dun ethy lethy Iben Mne
3,6-dielhylloKiene
POCP
8!> 1
1002
86
21.8
63.7
106.3
110.8
101 0
73.0
63.6
50.0
126.7
127.8
138 1
898
101.9
90.6
132.0
129.5
PPCP
20 H
774
22
4.5
478
960
946
922
44.9
34.8
162
119.1
1185
122.4
71.8
80.8
73.2
108.9
99.8
Organic compound
Ketonai
acetone
methylethylketone
methyl-i-butylketone
methylprop;lk«ton«
dielhyUt«ton«
mslhyl-i-propylketone
h«xan-2
-------
IDENTIFICATION OF THE IMPORTANT PARAMETERS
WHICH DETERMINE OZONE FORMING POTENTIAL
to
U)
ov
THESE FOUR FACTORS ARE:
MASS EMISSION RATE
MOLECULAR WEIGHT
OH RATE COEFFICIENT
CHEMICAL STRUCTURE
HOW DOES CHEMICAL STRUCTURE
INFLUENCE REACTIVITY VALUES SUCH AS
POCPs
Mainly through the formation of intermediate
compounds which degrade much more slovdy
than the parent compounds.
Examples of unreactive intermediates include
* carbon monoxide
* alkyl nitrates
* acetone
* formate esters
-------
FIGURE 4. The POCPa for alkanes, alkenes and aromatic compounds with NOx
emissions halved, standard and doubled.
NJ
ii
d>
C
CL
O
O
0_
180-1
160-
140-
120-
100-
80-
60-
40-
20-
LOW REACTIVITY ORGANIC COMPOUNDS
ON A BY MASS BASIS
2-methylbutan-2-ol 14.2
styrene 14.2
i-propanol 14.0
methanol 13.1
ethane 12.3
t-butanol 12.3
acetic acid 9.7
acetone 9.4
acetylene 8.5
methylene dichloride 6.8
t-butyl acetate 6.5
methyl acetate 4.6
methyl formate 3.3
formic acid 3.2
tetrachloroethylene 2.9
chloroform 2.3
methylchloroform 0.9
methane 0.6
methyl chloride 0.5
benzaldehyde -9.2
1 1 1 1 1 T
Halved Standard Doubled
NOx NOx NOX
-------
"T
COMPARISON OF POCP AND MIR
REACTIVITY VALUES
POCPs address regional scale ozone
formation over the 1-5 day timescale
appropriate to Europe.
MIR values address urban scale ozone
formation.
For the majority of organic compounds the
scales are in excellent agreement.
There are some differences:
* role of formaldehyde
* 1,3-butadiene
* butylene
* ethyl t-butyl ether
FIGURED. Comparison between the MIR and POCP reactivity scales for up to 70
organic compounds.
X fomultfehyd*
I *
O A
Xolluft
2-38
-------
WHAT NEXT WITH POCPs
Expansion of Master Chemical Mechanism to
include latest aromatic compound degradation
pathway studies of Jeffries and co-workers and
Barnes and co-workers.
Include an additional 20 organic compounds
mainly CIQ aromatic compounds making 140
in all.
Expanding the range of oxygenated organic
compounds.
Deriving estimation procedures for POCPs.
Derive equivalent of POCPs for large industrial
emission sources of organic compounds for
controlling downwind ozone formation.
FIGURE 5. Comparison of ozone-formation indices for the reactive alknnes,
\
alkenes and dienes.
300
° .2 *
t 5
ID
T!>
to
CJ
JT
>.
«
oi
$jS£SIS3lS||3SI1lfII
"11 ^-^i^-I^S^^l^.-
>.
>,
.c
Q)
CM
c £" c
I I £
c
0)
c
0)
a.
a> CD
c c
a>
c
0)
Q.
O
PJ
VI
C
Organic compound
2-39
-------
UTILITY OF REACTIVITY VALUES
INCLUDING POCPs
In the exact limit, no two ozone footprints from
different organic compounds can be
superimposed.
Single number reactivity values must involve
some form of approximation and assumption.
A reactivity value is not a geophysical quantity
such as a rate coefficient though it may
depend on one or many.
Reactivity values are user-oriented constructs
whose calculation depend on understanding of
a few environmental processes but also on
some policy-oriented choices, such as the
spatial scale of interest.
Reactivity values are not subject to observation
and testing but are best judged by the insights
they give into the role of each organic
compound in forming ozone in real situations
and their usefulness to policy-makers.
TOP TEN HYDROCARBONS ACCORDING
TO INVENTORIES AND POCPs
1
2
3
4
5
6
7
8
9
10
toluene
n-butane
ethyiene
m-xylene
p-xylene
o-xylene
i-pentane
ethyl alcohol
propyteoe
Together these account for 49% of the ozone
forming potential of UK emissions
-------
Ozone reduction In ppb p«r thousand tonn«s p«r year abated
010
t
VCX: emitting sector
-------
Reactivity: 05/12/98
EPA's Models-3 Framework and the
Community Multi-scale Air Quality Model (CMAQ)
Robin L. Dennis
Atmospheric Modeling Division
National Exposure Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC
Photochemical Reactivity Workshop
Durham, NC
May 12-14, 1998
-------
RMetfvtty OV12/VC
EPA/On-Site CMAQ SCIENCE TEAM
Science Team Leader: Daewon Byun
Motivation
K)
MM5/MCIP
Al Bourgeois
Hao Jin
Jon Pleim
Tanya Spero
Ruen Tang
MEPPS
Bill Benjey
Chris Maxwell
Nick Moghari
Tom Pierce
Project Management
Jason Ching
Robin Dennis
Joan Novak
Ken Schere
CMAQ
Frank Binkowski
Jerry Gipson
Jim Godowitch
Sharon LeDuc
Sang-MiLee
Shawn Rose lie
Jeff Young
The complexity of environmental or air quality
prediction has increased because we must deal with
secondary pollutants, as well as primary ones.
The multi-pollutant nature of the atmosphere is being
recognized, also increasing prediction complexity.
Multi-pollutant interactions tell us that we should be
thinking and modeling increasingly from a one
atmosphere perspective. They also tell us that we need
to be thinking multi-scale.
There is greater dependency on the realism of the
simulations. We recognize that we need to predict
outside today's mix of chemical species to future, very
different mixes under conditions of complexity and
nonlinearity.
Our Modeling Should Have:
Increased Reliability Entailing Improved Realism in
an Expanded Scope, a One Atmosphere Scope
-------
K)
We Came to the conclusion that this necessitated
models that:
Are as first principles as possible or feasible
Include a full set of interconnected physical
and chemical process descriptions
Incorporate a full marriage with prognostic
meteorological modeling
We also came to the conclusion that:
Current Systems Not Expected to Cope
Incrementalism Not Expected to "Get Us There"
No Single Group Can Do It All (Nor should be
expected to)
Three Pillars of Requirements for a 3rd Generation Modeling
System
Science
Increased realism and adaptability
System Framework
Increased Modeling capability
Computing Infrastructure
Increased compute power and flexibility
Umbrella Concept of Community Modeling Is To
Permeate
Our Response: Develop a 3rd Generation Modeling System.
This new modeling system is composed of a 3rd
Generation Modeling Framework, termed Models-3,
and a 3rd Generation Air Quality Model, termed the
Community Multi-scale Air Quality Model, CMAQ or
Models-3/CMAQ.
-------
IJ
Up to date science with better or easier integration
of new science
(Keep up with the best)
> More complete multldisciplinarity; One atmosphere
scope
(Integration of "complete" set of physical and
chemical processes)
> Better evaluation
(Diagnosis of model processes; diagnostic
evaluation)
> More robust and adaptable model structure
(CTM adaptable to different dynamic driver
configurations)
> Better modularity to support community modeling
Increased Modeling Capability (SYSTEM FRAMEWORK)
4 Support for the different levels of modularity.
Full system available to dispersed scientific
community for process study, model/module
development, and incorporation of advances.
(Support community modeling at the science level.)
+ Full system available to operational user
community for assessment, "easy," controlled
execution.
> Support for levels of interoperability and
intercommunication needed to support community
modeling.
4- Full execution analysis: verification, visualization,
output analysis, process analysis, model
evaluation.
-------
RMctMty O5/12/W)
SCIENCE
Set of Three Models: CMAQ/ MM5v2/ MEPPS
Designed to be Multi-pollutant and Multi-scale
O Ozone, Acidic Deposition (S and N), Nutrients
(N), Fine Particles (primary and secondary:
sulfate, nitrate, organics), and Visibility
O From Continental to Urban, with embedded
Plume-in-Grid
Annual Releases of the Model Set Are Expected for the Next
Several Years.
Public Release of CMAQvl.O
June 1998
Major Update and Expansion of Selected Science
Options (Given Below) of all Three Models, With a
Focus on CMAQ
June 1999 Release
Subsequent Release Dates to be Determined
MCIP
Emission
Processor
(MEPPS)
Meteorology
(MM5)
Data Flow
Principal Non-framework Components
Models-3 Interfaces
Analysis Tools
System Instructions
-------
R**ctMty 05/12/90
Ructivfcy 05/12/98
NJ
Meteorological Model: MM5v2
Basic Pedigree
Penn State/NCAR Mesoscale Model Version 5
State-of-the-science prognostic meteorological
model
Started in 1970's; on-going development today;
MM5 is contributed to extensively by the scientific
community
National and international use for operations and
research
MM5 Operation
We are using MM5 Version2, the latest release
It is non-hydrostatic (to be able to go down to small
grid sizes)
Tested, most applicable set of physics options
invoked
One-way nesting (we were first to debug this
option) at 108-, 36-, 12-, and 4-km resolutions
4-Dimensional data assimilation (Analysis Nudging)
to recreate past meteorology as closely as possible
for 108-36-12 resolution
Augmented, adapted output to better serve air
quality modeling
MM5 Future Development by Our Group
Higher resolution land-use and soil data (Vegeland)
More advanced PBL, microphysics, radiation
(Pleim-Xiu with Vegeland)
Implementation on workstations and massively
parallel computer - CrayTSE
10
-------
Daewon W. Byun
Chemical Transport Model: CMAQ
Model design
MEPSE
CMAQ Adaptability
Chemistry - Transport model
Chemistry _ Advection Diffusion
2-48
-------
RudMty OS/12/M
NJ
Science Features - CMAQvl.O for June 1998 Release
Generalized coordinate system internal to CTM (to
work with any map projection and meteorological
driver)
Generalized chemical mechanism reader
Gas-phase chemical mechanisms and solvers
RADM-2+ (Carter isoprene)
CB-IV
QSSA for workstation
SMVGEAR for Cray (vector machine)
Piecewise Parabolic Method (PBM) for advection
Vertical diffusion: Kv (eddy dlffuslvity)
Horizontal diffusion: KH (resolution dependent)
Clouds (and precipitation)
Large-scale: grid-resolved at all1 resolutions
Convective: sub-grid parameterizations at 36-
and 12-km
Precipitating
Non-precipitating
Aqueous chemistry (RADM)
CMAQvl.O for June 1998 Release (cont.)
Particulate Matter
Modal dynamics
3 size ranges
PM-fine Secondary (S, N, Organics)
Primary (emissions
inventory)
PM-Coarse Primary (emissions
inventory)
Chemical speciation tracked
Size dependent dry deposition
Regional Haze: Light extinction; Deciviews
Plume in Grid
Photolysis rates: Look-up table, with cloud
attenuation
Implemented on PC/NT; Sun Ultra, Dec Alpha, and
SGI (close); and Cray vector supercomputer
12
13
-------
Hn «| asnvm
K)
i
(^i
O
Additional Science Features Expected - CMAQ, June
1999 Release
Chemistry: SAPRC
Advection: Bott, ASD (spectral), YAM
Diffusion: Asymmetric Convective Model (ACM)
Surface PBL: Vegeland Pleim-Xiu (PX)
PM: Improved production of organics
Incorporation of sea salt
IC/BC: Stratospheric background (top)
Photolysis: 4-D implementation
Implementation on massively parallel computer -
CrayT3E
MCIP, Meteorology-Chemistry Interface Processor
Features - MCIP for June 1998 Release
Generate coordinate dependent meteorological
data (Jacobian) for generalized CTM simulation
(traditionally treated in CTM)
> Maintains modularity of CMAQ regardless
of coordinates
> Allows consistent links to many
meteorological models
> Provides meteorologically consistent
interpolation methods
Process meteorological data
> Window to CMAQ domain
> Compute or pass through surface and
PBL parameters
> Diagnose cloud parameters
> Compute species-specific dry deposition
velocities (gases)
> Output meteorological data in Models-3
I/O API format
14
15
-------
NJ
Additional Features Expected - MCIP, June 1999
Release
Link with RAMS (by end of Calendar 1998)
Deposition: CMAQ method (linked to
Vegeland_PX land-surface model)
Plans for Future Improvements: CMAQ and MCIP
Morphecule Chemical Mechanism
Links to other meteorological models (ETA, RUC,
ARPS, WRF)
Additional diffusion options: Hybrid (local/non-
local), Transilient
Greater option consistency between MM5 and CTM
Explicit simulation of aqueous phase chemistry in
clouds
Particulate modeling: External particle mixtures
(particles with same size but different chemistry);
better representation of blowing dust.
Improved methods for mass conservation (mass vs
mixing ratio)
Deal with mass tracking, source
apportionment
Sensitivity analysis packages incorporated
Emissions Model: MEPPS
Models-3 Emissions Processing and Projection System
INPRO/IDA (Inventory Data Analyzer)
QC of source/emissions inventory emissions data
Format conversion to Models-3 I/O API
EMPRO
Modified GEMAP/EMS-95 for point and area
sources
SAS-based system, incorporating ARC/lnfo
geographic information system to accomplish
spatial allocation of emissions data.
MobileSa mobile source emissions model
BEIS2 biogenic emissions model
Speciated emissions for RADM-2 and CB-IV
mechanisms
ECIP
Linked to MCIP for meteorological data
Calculate plume rise for major point sources
Linked with Plume-in-Grid
16
17
-------
r
K)
dn
NJ
V)
I
a
f
?
3-
P
3
KADM-1
70
!
MEPRO
Project base year emission data using source-
category-specific Economic Growth Analysis
System (EGAS) factors for input to EMPRO
Additional Features Expected - MEPPS June 1999
Release
MobileSa with PM-fine emissions
BEIS3 biogenic emissions model
Speciation for SAPRC (fixed stochiometry)
Future Improvements for Emissions
SMOKE emissions processor
Generalized speciation
Link to Morphecule chemical mechanism
18
-------
RwctMty W12»«
to
Modeling Scales
Horizontal
Nests of 108-km, 36-km, 12-km, and 4-km
Windowing to subdomains from continental to
regional
layers from the surface to the top of the free
troposphere
30 layers collapsed to 21 layers, converting the top
18 layers to 9 layers
predictions for 24-hour simulation
segments; typically for 5-day simulation penods
meteorological inputs at 36- and 12-km
15 minute meteorological inputs at 4-km
Aggregation (statistically weighted average) of 44
synoptic flow patterns for seasonal and annual
averages at the continental scale
MO If 130 W 120 If HOW 100 y BOW BO If 70 If 60 IT 60 If 40
30 N
110 W ir
too w
60 W
70 W
IT
SO N
40 N
20 H
19
-------
121
RADM VERTICAL DOMAIN
.40* 130 W 120 y 110 W 100 W 90 W BOW 70 W 00 W 60 ₯ 40 W
K)
i .
50 N
110
100 W
90 IT
60 W
70 W
200 220 240 260 280
Temperature (K)
-------
R.KItvtty OS-1MH
CMAQ MODEL EVALUATION
Near Term
Purpose: Acceptance by Regulatory Community
Traditional Operational Evaluation - NARSTO NE '95
> O, predicted versus measured
> Daily bias & gross error aggregated over all sites
> Accuracy of peak predictions
Time-Space Disaggregated O, Statistics - NARSTO NE
'95
> Space: type of grid cell based on photochemical
Processes
> Time:
* early morning: titration & inversion breakup
# daytime: mixing height & dilution; O3 production
# nighttime: deposition, surface layer loss
20
RuctKHty-OS'12/M
CMAQ MODEL EVALUATION
Long Term
Purpose: Acceptance by Scientific Community
Process understanding and scientific uncertainties
Determine best model configuration to reduce
uncertainties
Assess value of new measurement information
Diagnostic Evaluation - Nashville SOS '95
> Insight into processes generating O3
* OH & NOX cycle Interactions with resulting O,
production
* Integrated reaction rate/ mass balance
* Indicator species
> Sensitivity : CB4 versus RADM2+ chemistry
Sensitivity Analysis - Nashville SOS '95
> Characterize importance of process differences
> Interpretive analysis related to appropriate use of
CMAQ
21
-------
SYSTEM FRAMEWORK
Enhanced Modeling Capability
* Support for the different levels of modularity
* Full system available to dispersed scientific
community for process study, model/module
development, and incorporation of advances.
(Support community modeling at the science level)
: Full system available to operational user
community for assessment, "easy," controlled
execution.
< Support for levels of interoperability and
intercommunication needed to facilitate community
modeling
* Full execution analysis: verification, visualization,
output analysis, process analysis, model
evaluation.
An Advanced Computer-based Problem Solving and
Modeling Environment or Framework With An Effective
Human-Computer Interface for Environmental Modeling and
Assessment That is Adaptable to a Changing Computing
Infrastructure.
Assist Environmental Analysis and Model Development
> Facilitate execution of air quality simulation
modeling systems, especially air quality models,
and the visualization and analysis of their results.
> Minimize the tedium and chance of error
associated with modification of rigid model
execution scripts.
> Provide comprehensive data management to assist
in storing, accessing, tracking, identifying, and
capturing processing history of numerous datasets
associated with modeling studies.
> Manage and organize large collections of model
executions and associated data.
> Provide cross-platform (a variety of computing
platforms) computing of complex modeling studies
with distributed data management.
22
23
-------
R««c«vty 05/12/90
K)
Assist Environmental Analysis and Model Development
(cont.)
> Aid the assembly, testing, and evaluation of
science process components by facilitating the
interchange of process modules and minimizing
the chance of incompatible assumptions.
> Facilitate the tailored execution of the modeling
system, including customized process analysis.
> Provide the flexibility to change key "global" model
specifications such as grid resolution, map
projection, or chemical mechanism without
rewriting code, thus minimizing error.
Functionality Achieved Via the Following Management
Components Incorporated in the Modeling Framework:
Science Manager
Model Builder
Program Manager
Data Manager
Strategy Manager
Study Planner
Tool Manager
Source Code Manager
Framework Administrator
Flexibility for Future Change Achieved by Architectural
Layering
24
25
-------
ARCHITECTURAL LAYERING:
FLEXIBILITY FOR FUTURE CHANGE
USER
INTERFACE
USER INTERFACE
Management Layer:
Data Manager Science Manager
Study Planner Model Builder
Strategy Manager Source Code Manager
Tool Manaaer Framework Administrator
Environment Layer: OS, System "Personality"
Computational Layer: Programs: models, analysis, visualization,...
Data Access Layer: I/O Applications Programming Interface
Data Structure/Representation: netCDF, XDR
Data Storage: File systems & databases
Physical Device Layer: Disks, networks, printers, machines
s
o
b,
7
3'
t!
<3)B
i *
I
I
I
I
5>
E
2-58
-------
CMAQ Adaptability
Chemistry - Transport model
Chemistry Advection Diffusion
to
HI
00
CL
o
o:
i
UJ
I
u.
CO
CO
UJ
a
o
.2 c
re .2
E |
o o
*£ <"
£ c>
0)
01
re
re
0)
o
.£
'o
CO
en
c
o
0)
Jr CO
a a,
a. 'o
re u
E °-
08 ^
= =
a)
re .E
£ re
(/) £
To o
.Q "D
£ 0)
O) "O
si
1A
od £ o
>, o
w S
o ^
I TO
.E o
"S1
o o
O £
O O
A A
k.
0)
'5
m
"3
TJ
O
JC
+J
0
*"*
o
'o
0>
a
10
"3
o
0
E
re
o
to
k_
o
o
4-J
CO
0
C
o
Q.
IE
ojS
"3 2
CO Q.
TO
"TO
o
"TO
13
0
OT
TO
>
co
"5
o
o
E
0)
u
c
.2?
'o
(0
E
re
Q.
O)
s
TO
O
u
a;
c
0)
C
0)
c
o
a.
E
o
a
o
E
N «T
II
w "
3 TO
a)
O)
re
c
TO
E
2
O)
2
a.
JC
o
0)
E
TO
*-
2
"c
E
2
D)
O
a.
T3
4-t
TO
|
re
_o>
JQ
TO
C
0)
TO 2 S>
O) WE
O O ra
QL EQ[
O C
3
U
0)
X
0)
V)
'5)
V
C£
"TO
O TO
o c
22
2-59
-------
R««c«vtty O5/12JW
to
MODELS-3 FRAMEWORK CAPABILITIES (cont.)
Data Manager
Data access from any networked system
No need to convert data between computers
Manage dataset history information
Uses Federal Geospatial Metadata Standard
Strategy Manager
Evaluate alternative emissions control options and
future years
Study Planner
Automated multiple-platform execution of a series of
interlinked programs/ model/ independent modules
User control over input parameters
MODELS-3 FRAMEWORK CAPABILITIES (cont.-2)
Tool Manager
Prepare emission inputs compatible with selected
model
Invoke visualization & analysis tools
> VIS5D - 3-D analysis & animation
> PAVE - 2-D analysis & remote viewing
> IBM DX - specialized analysis tools
V SAS
Source Code Manager
Version control of science code ensures documented
software evolution & replication of previous executions
Facilitates automated building of models from
components
Framework Administrator
Controls integrity of official model versions
Controls security & access to data/ code
Provides system maintenance capabilities
27
28
-------
Seamless Computing & Data Management
From PC to Scalable Parallel Computer
State/Regional Offices
Public Access
Jylaster
Metadata
& Global Info
Object Oriented DBMS
Federal Agencies
HI
D
UJ
z
o
I
o
8
(±
UJ
H
UJ
LU
O
LU
^
<
Q
» CD
i. 2.
E O
o o
o zi-
o
m
T3
Q)
t c
0)
c
re
I .=
u
a>
5"
O
vi re
W t3
g 2
o a>
re 5
s *
I/)
0)
o> .£
u <£
...c "5.
ow E
J=£ w
o w 2
o 3 re
£"E °>
; c
S
in
'En
0)
o:
o
o
UJ
LU
a
* £
E x
O G,
o _
c
o
o
c
re
o
« a
a>
tf *
«
re
Q
c
o
o
2 t
o
0)
o
z
exte
v>
0)
0)
e
u
s
a;
O)
c
E
E
2
O)
o
.2 JE 0-
HI
2 '5 S
.
a
G> *i
"! a
CDF
Q_Q
c i
A A A
D
re
c
re
W
re
re
re
4^
i
o
a
o
c
re
"5.
E
o
u
S
'n
o
re
I
2-61
-------
RtcclMly 03/17/M
NJ
0\
K)
KEY MANAGEMENT LAYER COMPONENT DETAIL
STUDY PLANNER
Automated multiple-platform execution of a series of
interlinked programs / models
Automatic registration of generated output files
Execution "History" information written on each output
file
> Link to compile time information
* science module versions
* configuration file, compile environment & date
> Runtime control information
* environmental variables, namelist input,
command line
* study/plan identification
* link to input/output file metadata
SOME FEATURES OF THE MODELS-3/CMAQ SYSTEM THAT
SHOULD AID THE SUPPORT OF REACTIVITY
CALCULATIONS
Chemical Mechanism Reader
Customizable Process Analysis
Study Planner
Data Manager
Multi-pollutant CTM
Responsibility for Evaluation, Including Diagnostic
Evaluation
Meteorological Case Variety and Availability
Expected to be Extensive
> runtime environment
* execution date & time
* hardware platform, OS version
30
31
-------
Establishing a Community Modeling Capability
Kenneth Galluppi
Univeristy of North Carolina at Chapel Hill
Taken From: Air Quality Community Modeling and Analysis System
Attributes and Implementation
Draft 09/15/97
I. Introduction
Over the past four years, there has been a tremendous growth in the community's utilization
of air quality models in the development of emission strategies related to Clean Air Act
compliance. For example, the Ozone Transport Assessment Group, OTAG, witnessed model
applications growing beyond the four "official" centers to extend to many States, private
industries, consulting engineers, and environmental groups. With this, exposure, came
concerns of public access and consistency in model evaluation, analysis, interpretation, and
scientific integrity of the systems and their application. One approach to reducing these
concerns is to implement a community modeling and analysis system.
From August 27-29, 1997, a workshop was held in Research Triangle Park to discuss the
purposes and needs for models, the benefits of a community modeling approach, and the
attributes of and obstacles to developing and implementing; such a system. The workshop
had fifty participants whp represented federal and state governments, industry, and university
researchers. This paper is a summary of the background, findings and recommendations
from the workshop.
II. Models, Analysis, and Their Applications
There are many approaches to modeling ranging from statistical to comprehensive models
based on governing equations of physics and chemistry. Each approach has its strengths and
weaknesses in terms of reliability, predictability, and cost. A diversity of modeling
approaches serves to check our models for consistency resulting in greater confidence in their
application for emission strategy development.
-
^Untangling the physical-chemical relationships required to understand the cause-effects of
f pollution problems is a daunting task. Over the past decade, the air quality community has
been engaged in various developments and applications of comprehensive models in
conjunction with complex analysis of observations. Each model or analysis improvement
has opened the door to new sets of unexplained observations and hypotheses. Improving our
scientific understanding is, and always will be a dynamic process, but remains at the pinnacle
of developing sound environmental practices.
It is assumed that the community is committed to the development of scientifically credible
models and in building skill in their use for guidance, to discern probable cause-effects and in
2-63
-------
develop mitigation and prevention strategies of air quality problems. This commitment
stems from the increasing costs of assigning culpability for pollution abatement and
-prevention. Policies are establishing who pays and how much which opens questions to how
reliable is the scientific guidance used for this judgement. The question remains, how does
the community develop the best simulation tools for use by both scientists and managers
whose common purpose is achieving a sustainable environment?
While the scientific community is engaged in developing new insights into science, the
management community is engaged in making decisions based upon the best available
science. Most would agree that these are complementary objectives. However, history has
shown us that scientific knowledge and new analysis methods are difficult to transfer to a
large, diverse, and often an divided management community whose time and funding
constraints may inhibit the use of "best" science or its proper application and interpretation.
The development of numerical models and their application to real world problems has
enabled mediocre transfers of scientific knowledge and technology utilization. Howeyer, the
rising costs of environmental protection demands the quicker development of better models
and their transfer for use in scientific and management practices. In order to develop low-
cost, equitable policies, that the community has confidence in, we need to capitalize on a
shared goal to develop the best scientific formulations and learn to apply models in a
consistent manner.
The question explored in the workshop is whether the community can enhance these efforts
through common, integrated development and application efforts. If so, does this warrant the
defining, development afod implementation of a community-based modeling and analysis
system?
HI. What is a Community Modeling and Analysis System (CMAS)?
A Community Modeling and Analysis System (CMAS) is an approach to model
development, application and analysis that leverages the community's complementary talents
and resources in order to set new standards for quality of science and reliability of
application of air quality models. The resulting comprehensive system forms the foundation
which the community, including governments, industry, academia and other stakeholders,
participates in the examination of issues and the subsequent development of strategies that
meet societal challenges of environmental protection.
A community modeling system is a computerized framework and intellectual process that
integrates the research and development findings, and application experiences into a common
set of tools and knowledge base. The information and tools in the CMAS are open and
readily accessible to everyone. It is called a system because the integration is organized into
a wide range of computerized information modules. It is the desire of the CMAS approach to
increase productivity and reduce the cost of examining issues and developing alternate
scenarios. The cornerstone for accomplishing this is.the framework that enables quicker
integration of science and techniques and for easier transfer of knowledge and experiences to
2-64
-------
the community. The result is peer reviewed, process of providing the best simulation and
analysis tools for use in a regulatory setting.
A CMAS is not a single "mother of all models", but an integrated collection of science
modules which can be linked together to form multiple model configurations. The
underlying premise is that there is no single group of modules that can claim to be the "best"
model. Rather, by changing specific modules one can gain insight into the validity of
modeled results. This does not preclude the regulatory community from assigning and
locking a prescribed set of modules into being the "regulatory model" for consistency
purposes. In fact, there is an advantage to having several "regulatory" models for routine
scientific comparisons.
For example, the National Weather Service in its earlier days of weather forecasting, utilized
three models for daily comparison of predictions: the complex Limited Fine Mesh (LFM),
the 2 variable Baroclinic, and the 1 variable Barotropic models. Although formulated
differently, they were compared for consistency. The "simpler" models aided the
meteorologist to get a feel for the numerical forecast generated by theJLFM, the model of
choice. The models are more sophisticated now, but are still cross-checked against models of
similar complexity.
How does this freedom of module use potentially impact the regulatory process? Which
modules can be utilized and when? Are we adding more confusion to the analysis than
before? How can do we know when we are getting a "better" simulation". How do we
prevent model "calibration"? These are but a few of the more imposing questions that will
inevitably arise and many of which were addressed at the workshop.
IV. Benefits of a Community Modeling System
Increasing productivity, raising confidence in results, interpretation and use, increasing
stakeholder buy-in, and reducing costs are all worthy goals of any modeling approach. The
CMAS is intended to maximize these benefits through the reduction of overlapping
developments, peer review, and ease of development and application. Given a robust and
streamlined system design and a peer review process by which the CMAS can be used for
development and application, the benefits of a CMAS. The benefits should easily justify the
focusing of resources and change in process by which models, tools and techniques are
currently developed in the community. There are many benefits to the community. These
include:
a) Implementing scientific advancements into models can be difficult. A modular
community system allows for process and formulation changes easier and with less
resource. This will enable the community to take advantage of new approaches sooner
than ever before.
b) Model versions and revision levels can be more readily controlled. Further, time to create
new revisions will be significantly reduced. A CMAS inherently will have far better
2-65
-------
quality assurance procedures when many more are involved in establishing a new
version.
c) With a common framework, model formulations and their applications will be far easier
to compare.
d) New scientists and engineers can be trained in specific areas of expertise since only one
framework needs to be learned. Further, over time, there will be a cadre of experts who
can share their knowledge with new personnel. This will lead to a dramatic reduction in
costs associated with training and continuing education.
e) Scientists and computer scientists will complement one another without having to learn
the other's field. Development groups will be able to better utilize the skills and training
of its personnel.
f) The science and application talents and knowledge will be added to a common base and
made available to all. This will enable all problems to be addressed at the best available
skill level.
g) Databases that are utilized by the models and for analysis will be more easily shared
which should lead to more thorough quality control, analysis and consistent
interpretation.
h) Analysis techniques and methods can be more readily transferred to the community. This
should enable more revealing techniques to be utilized and understood by a larger pool of
experts.
i) A common system will have several levels of use including the management decision
maker, scientist/engineer, and model practitioner. A robust CMAS enables each of these
specialists to maintain the appropriate level of detail needed to keep their understanding
within context, and while at the same time enable each group to communicate more
effectively to one another.
j) User groups and other transfer methods would be implemented to develop a knowledge
base.
k) A peer review process will assure that quality science, application, and analysis are being
implemented and put to best use by the community.
V. Attributes of a CMAS
The workshop breakout groups looked at six areas of attributes and implementation issues:
science, environmental management, education, support and maintenance, intellectual
property, and model application and evaluation. In this draft, each group's listing of
attributes is given. In later drafts, the groups' replication of comments will be eliminated.
The attributes are characteristics that describe a community modeling and analysis system.
In other words, what must a CMAS be able to do to meet the specific perspectives of
modeling support?
Group A - Science Workgroup
a) General enough to minimally integrate the processes required to simulate regional and
urban scaled problems.
b) Incorporates the physical-chemical processes to enable utility for examination of multiple
pollutant (ozone, aerosol, acid deposition), air quality problems. Eventually, it must
2-66
-------
extend to cross-media where chemical transport crosses media types, for example to
water.
c) Able to produce three dimensional concentration and deposition fields of key chemical
and physical species that enable a diagnosis of the cause-effects relationships.
d) Must be usable by scientists, environmental managers, and stakeholders within the
effected communities and regulated industries.
e) Brings together a working community to examine issues such as the NAAQS, NSR, and
deposition.
f) The science modules must be peer reviewed, relevant processes required for multi-
pollutant assessments. Further, the CMAS must be extensible to include alternate
formulations of known processes and the inclusion of new science when it becomes
available.
Group B - Environmental Management Workgroup
a) Able to assist the managers by modeling input variations consistent with varying
environmental strategies.
b) Have a high level of quality control to raise confidence of use.
c) Should be linked to and assist in risk assessment and decision making techniques
including social-economics and health risk.
d) Be able to expose and explain sensitivities and uncertainties in the formulation and their
impacts on results, interpretations and utilities.
e) Responsive to time constraints and adjust formulation accordingly
f) Must be able to facilitate communication to multiple levels of users from managers,
practitioners and public.
g) Must be a ble to summarize and explain input scenarios and impacts on results through
the use of multiple analysis techniques.
h) Must be well documented and understandable by multiple levels of users.
i) Must be reasonable cost to not be prohibitive for any stakeholder to utilize.
j) The CMAS must be open for all participants to examine. The process to include new
modules must be open, peer-reviewed, and well understood.
Group C - Education Workgroup
a) The community requires a diverse set of expertise to be available to educate and be
trained in. This includes: emission engineering, meteorology, computer science,
atmospheric chemistry and physics, data analysis, and program and operations
management.
b) Additionally, the community needs to be educated into the impacts on economics, control
; technologies, risk assessment, population exposure, community and industrial planning,
?! and other environmental concerns.
c) Good practices in computer science need to be brought to the training program. This
includes: systems engineering, database management, programming,
graphics/visualization, and system administration.
d) A thorough program management training must be undertaken that includes: planning
and scheduling, science and technology appreciation, resource allocation,
communications, and negotiation skill. J>
2-67
-------
e) Multiple education and training methods must be deployed. Examples include: hands-
on, internships, web-based, print, video, satellite, workshops, user groups, and chat
rooms.
f) Various sectors within the community must have better dialog and cross training. These
include the scientific, management and practitioner groups, as well as government,
industry, environmental, and academia groups.
g) Guidelines are needed to help establish consistency in personnel skill and training.
h) Education programs need to be able to expand quickly and be effective to train many
users in a relatively short time period.
Group D - Support and Maintenance Workgroup
a) The CMAS is a common platform for model development and application for both the
scientific and regulatory communities.
b) The system automates as many of the computer operations as practical and makes it easy
for code compatibility and code re-use. To the extent possible, CMAS is modular and
standardized.
c) The CMAS is self-documenting and readily available
d) Acceptability criteriaTor any change to CMAS will be established. This includes new
modules as well as for using in scientific and official regulatory studies.
e) The CMAS will be supported by a core maintenance group. This group also has
responsible for maintaining key databases and datasets that are utilized by the
community.
f) The CMAS and data are open and readily available through several distribution channels.
g) The core group assigned to maintain the CMAS, implements the procedures for updates,
testing and distribution. This include science and computer updates.
h) The support group is established and overseen by a governing board. The Governing
Board is responsible for establishing official policies regarding CMAS and establishing
funding support.
i) Procedures are established by the Board for creating updates to the system and releasing
versions for scientific and regulatory use. A proposed schematic linking the institutions
involved with CMAS and the procedures that they follow is shown in Figure 1.
Group E - Intellectual Property Workgroup
a) There are three approaches to making software available that need to be considered: It is
accepted that public domain or shareware concept are acceptable only if the code is fully
open and accessible.
Public Domain: Creator relinquishes all rights to intellectual property and makes
software freely available. The code is open.
Shareware: Author maintains a copyright on the intellectual property and code but
makes it available free or at low cost to promote sharing. The code may be open or
closed.
Proprietary: The author retains all rights and generally charges a fee for use. The
code is usually closed.
b) The source code for software that affects model results should be available for free. This
promotes understanding, review, and trust of the^science and its implementation. Related
components, such as interfaces and graphical tools, that do not affect the model or
2-68
-------
analysis results must be readily available for free or low cost but do not necessarily have
to be open.
c) CMAS tools could also include third party software, such as SAS or Arc/Info, which
would remain proprietary. However, regulatory procedures should give a low cost, low
resource, alternative.
d) Value added modules for analysis, graphics, etc. can have license fees, but must be open
to scrutiny.
e) Most credit will be given by reputation and publishing. However, contributed modules
should be documented for credit and given proper credit by the users.
f) There may be liability issues related to deficiencies in best science, bad science, and
coding errors. The Governing Board will need to establish procedures for investigating
these issues and establishing limits to liability, if any.
Group F - Model Application and Evaluation
a) The CMAS is managed by a central organization and overseen by a Board of Governors
that represents the air quality community stakeholders.
b) The CMAS center supports, facilitates and maintains the development of a low-cost, open
modeling and analysis system. The center does not apply the system.
c) The CMAS does not preclude outside model developments but serves to promote more
involvement in development.
d) Research intersts should be linked to regulatory needs, including time and funding
schedules.
e) All aspects of CMAS must be peer reviewed.
f) CMAS must be linked to data bases that are readily available, quality controlled and
documented. The CMAS management organization can act as a data clearinghouse.
g) The management center must have sustained support, remain viable, reliable and growing
in competence.
h) In order for applications to be successful, CMAS must provide the following types of
tools: a flexible framework, analysis tools, tutorials, technical guidance, and transfer
mechanisms such as workshops.
i) In order for management practices to be improved via a CMAS the following needs must
be met: assurance of the best tools, sanctioned regulatory configurations, detailed
guidance for use and interpretation, and technical support for all levels.
j) To improve model evaluation, the CMAS must provide a range of open diagnostic tools
that facilitate collaboration. The data and modeled results must be easy to access and
with proper guidance, easy to analyze and interpret. Performance evaluation should be
understood.
rVI. Issues With Implementing a CMAS
There are many obstacles that would prohibit an effective community approach from being
implemented, the most obvious being funding support. However, there are many other
pressing issues, each of which could inhibit the benefits of a Community Modeling and
Analysis System. Whereas, the workshop groups were able to define attributes along six
different sets of attributes, the implementation obstacles were much more ubiquitous during
2-69
-------
the breakout and plenary discussions. Because of overlap of issues, a consolidated view of
the workshops can be presented.
a) A sustainable funding is critical to achieving a CMAS. To this end, the benefits of a
CMAS must be clear and communicated to stakeholders in government, industry,a nd the
public sector. Long term commitments to funding must be identified and put in place.
This will be difficult due to funding mechanisms, mix or public and private dollars, and
accountability.
b) A consensus view of the attributes needs to be developed. This will not be straight
forward as funding constraints will prevent all attributes from receiving the same
attention.
c) Defining and promoting standards within the community will be difficult. This covers
coding and module integration standards.
d) An agreed upon management structure must be put in place. This includes a Governing
Board as well as the CMAS center administration and technical staff.
e) Guidance and modeling protocols need to be established as to how to utilize a CMAS.
This includes guidance for modeling evaluation and peer reviewed acceptance of
modules and science.
f) Setting up a CMAS, management structure, databases, and codes will take time.
Additionally, the community will need to change how it performs modeling currently.
These changes will take time as well as funds. It is questionable to some whether this
resource cost is justified.
g) The community is accustomed to using certain models and analysis techniques.
Establishing credibility for a flexible system will be difficult. The community would
need to establish an evaluation program that meets a concensus view of acceptance.
VII. Recommendations
There is general agreement that the benefits of a community modeling system are significant
enough to warrant further study and defining. The main recommendation is to form a group
of writers to put together a first proposal. This initial draft proposal needs to be outline how
a Community Modeling and Analysis System can be funded, managed, and leads to
improved science and decision practices. The proposal needs to clarify the following points.
a) Clearly define the CMAS for all to understand.
b) Clarify the benefits to the various stakeholder groups.
c) Propose a detailed funding proposal that shows costs to implement and sustain a CMAS
Center and future developments. This includes the number and type of positions required
and their reporting structure.
d) Clarify how the development cycles of the regulatory and scientific versions of the
community system will work together to assure the best science availability on regulatory
timeframes.
e) Outline how the CMAS would be applied within guidelines, peer reviewed procedures,
that enable the best application practices.
2-70
-------
f) Clarify the peer review checks for the science, the CMAS developments, regulatory
recommendations, and application guidance.
g) Clarify potential legal issues involving intellectual property and liability.
It is highly recommended that progress towards defining a CMAS be made quickly and
systematically. To this end, the following recommendation are made:
(a) Groups that have responsibility for achieving progress with Clean Air Act compliance
must be briefed about the CMAS proposal. Their buy-in to the concept is extremely
important. These institutions include the EPA's Office of Air Quality Planning and
Standards (OAQPS), industry and other stakeholder communities, State environmental
groups, consultant groups, and academia. These groups must be made aware of the
benefits of a community approach and how it would work.
(b) The Federal Advisory Committee (FACA) looking at new approaches to achieving
national air quality standards needs to be briefed about CMAS. FACA should provide
guidance as to the needs for a CMAS.
(c) An outline of the CMAS proposal needs to be circulated as soon a§ practical to get a
wider acceptance to the proposal needs of the community. Accompanying this outline
should be a proposal for gaining community buy-in.
(d) A full proposal outlined above must to be circulated within a wide community audience.
The community should have time to comment to the proposal whereby and final proposal
to establish a CMAS be written.
(e) A workshop should be considered for the spring of 1998 that will examine and debate
the merits of a CMAS, its feasibility, and funding. The outcome of the workshop would
be targeted at gaining a community go or no-go decision and commitments.
2-71
-------
Establishing a Community
Modeling Capability
Kenneth Galluppi
University of North Carolina at Chapel Hill
Summary of the Air Quality Community
Modeling and Analyze System Workshop
August 27-29. 1997
What is a Community Model and
Analysis System (CMAS)?
Leverage community-wide talents for
development and application of modeling
and analysis.
CMAS is computes-framework for
integrating "best" tools and knowledge.
Available to all for operational and research
needs.
Benefits of CMAS
Less resources to develop and maintain and
evaluate.
Regulatory and research needs are satisfied
and complement one another.
Operational needs for maintenance and
training are a priority.
Consistency.
Peer reviewed for all purposes.
K)
CMAS Workshop Groups
Science
Environmental Management
Education and Training
Support and Maintenance
Intellectual Property
Model Applications and Evaluation
Science Workgroup
Regional and urban scaled problems, for
multiple pollutant (O3, aerosol, acid dep.)
Usable by science, management and
stakeholders.
Peer reviewed.
Processes that get to the causes.
Accommodates robust analysis and
alternate formulations.
Environmental Management
Consistency of formulation and use.
Decision enhancing analysis.
High level of quality control.
Linked to risk-assessment, and other
decision making parameters.
Responsive to regulatory time constraints.
Facilitate communications.
Documented, open, no (low)-cost.
-------
Education Workgroup
Diverse levels of education/training needed:
management to practitioner, cross-
disciplines including computers.
Many forms of training to meet all needs.
Guidelines for training to establish
consistency in knowledge and application.
Timely and affordable.
Education to link to decisions and impacts.
Intellectual Property Workgroup
Give proper credit where credit is due.
Open, peer reviewed and
Free if it impacts science results, may
charge for value added modules, but open.
If community "owned," are there liability
issues?
Support and Maintenance
Modular, standardized, oriented for code re-
use. This includes inputs and analysis tools.
Well maintained and documented.
Procedures for making changes and
acceptance for scientific and regulatory use.
Core support group is maintained and
overseen by a board.
to
Application and Evaluation
Central organization overseen by a Board.
Maintenance, QA, and training are critical.
Peer review everything in appli. and eval.
Facilitate application and developments of
an open system but remain neutral.
Assure best tools, sanctioned versions,
guidance for use and interpretation.
Good set of tools and facilitated collab.
Difficult Issues
Sustainable funding for support and maint.
Consesus view of attributes given funding.
Design and promoting standards is hard.
Management structure.
Guidance and modeling protocols a MUST.
Setting up CMAS takes time and money.
Establish credibility for flexible approach.
Recommendation for CMAS
Clarify CMAS and its benefits.
Detail funding requirements.
Clarify development and "locking" of
science and regulatory versions.
Guidance procedures for application, peer
review, and evaluation.
Clarify legal issues of a "community"
model
-------
ABSTRACT
Emissions Modeling Issues for Reactivity Calculations: State and Status of
the Sparse Matrix Operator Kernel Emissions (SMOKE) Modeling
System.
Neil Wheeler
Environmental Programs
North Carolina Supercomputing Center
There are a number of scientific and operational criteria for emissions modeling systems
to support the use of photochemical grid models to calculate VOC reactivity. These
include:
1. The flexibility to support high-resolution chemical mechanisms that are
continually changing.
2. The ability to validate that the emissions modeling system is the same as used by
other parties.
3. The ability to validate that new mechanisms are properly represented in the
emissions modeling system.
4. Sufficient processing speed to allow for the development of multiple emission
scenarios, fortaioderate to long episodes, and potentially complex (large)
chemical mechanisms.
5. Designed to be integrated with state-of-the-science meteorological models,
photochemical grid models, and analysis tools.
6. Be reasonably easy to use and have a base of technical support for maintenance,
technology transfer, and user training.
In this presentation these criteria are discussed in further detail in the context of the
Sparse Matrix Operator Kernel Emissions (SMOKE) modeling system. SMOKE was
developed as a research prototype for high performance emissions processing under a
USEPA cooperative agreement. It has evolved into an operational emissions modeling
system under funding from various state and federal agencies. The current state of
SMOKE and other currently used emissions modeling/processing systems, and what is
needed for each of these systems to support reactivity calculation will be discussed.
2-74
-------
. Environmental Programs
Emissions Modeling Issues for
Reactivity Calculations
State and Status of the Sparse Matrix
Operator Kernel Emissions (SMOKE)
Modeling System
Neil Wheeler
North Carolina Supercomputing Center
Environmental Programs
Research Triangle Park, NC
North Carolina Supercomputlng Center
Emissions Modeling Issues
Flexibility
Validation
Speed
Integration
Ease of Use
Technical Support
. Environmental Programs
North Carolina Supercomputlng Center
2-75
-------
. Environmental Programs
Flexibility
The flexibility to support high-resolution
chemical mechanisms that are
continually changing
. North Carolina Supercomputing Center
. Environmental Programs
System Validation
The ability to validate that the emissions
modeling system being used is the same
as used by other parties
North Carolina Supercomputing Center
2-76
-------
. Environmental Programs
Mechanism Validation
The ability to validate that new
mechanisms are properly represented in
the emissions modeling system
. North Carolina Supercomputing Center
. Environmental Programs
Processing Speed
The need for sufficient processing speed
to allow for the development of multiple
scenarios, for moderate to long
episodes, and potentially complex
(large) chemical mechanisms
North Carolina Supercomputing Center
2-77
-------
. Environmental Programs
Integration
Designed to be integrated with state-of-
the-science meteorological models,
photochemical grid models, and
analysis tools
North Carolina Supercomputlng Center
. Environmental Programs
Ease of Use and Support
The system must be reasonably easy to
use and have a base of technical
support for maintenance, technology
transfer, and user training
North Carolina Supercomputlng Center
2-78
-------
. Environmental Programs
Sparse Matrix Operator Kernel
Emissions (SMOKE)
Modeling System
. North Carolina Supercomputlng Center
. Environmental Programs
SMOKE - Concepts
Traditional Emissions Processing Paradigms
- Self-contained records describing sources
- Admirably suited to 1970's-vlntage machines with
minuscule available memories and tape-only storage
- Passing of redundant data
- No exposed parallelism
Factor based tasks - linear matrix operations
Sparse matrices
Re-arrange the order of multiplications to
avoid redundant computations
Uses the Models-3/EDSS I/O API
Integrated within EDSS
^ North Carolina Supercomputlng Center
2-79
-------
. Environmental Programs
SMOKE Sparse Matrices
Transformed
Emissions
OUTi
OUTa
OUTE
Each column =
one model species, one grid Inventory
cell, or one control factor Emissions
Si
Sa
SN
North Carolina Supercomputing Center
. Environmental Programs
EPS 2.0 Processing Paradigm
SMOKE Processing Paradigm
North Carolina Supercomputing Center
2-80
-------
En
SMOKE Submodels and Dataf
A
?v:ONC
Land Use
Data
Meteorology
Data
Emissions
Inventories
L
Biogenic :
Submodel
/
MoMe /
Submodel ^
L
Area M
Submodeti /
//
Submodel J
I Time-Stepped
yfl Emissions \
( \
fe Matrices -^
7, 7
Time-Stepped ^
Layer Fractions
vironmental Programs
lows
Merge
oeesscir::::
\
Model-Ready
Emissions
North Carolina Supercomputing Center
Structure of SMOKE Submodels
. Environmental Programs
Inventory
Data
Ifet*
Structuring
/
Vleteorology
Data
Profiles &
Xrefs
Source 1
Database |
MC.fJC
North Carolina Supercomputing Center
2-81
-------
Environmental Programs
SMOKE Features
Inputs: EPS 2.0 or EMS-95
Outputs:
- Models-3 (CMAQ) / EDSS (MAQSIP)
- UAM-IV / UAM-V / CAMx
- SAQM with converter
Machine Independent I/O API (M-3/EDSS)
Multi-day runs
Approximately 30 times faster than EPS or
EMS-95
70% less disk space than EMS-95
__ North Carolina Supercomputing Center
. Environmental Programs
SMOKE
Speed
Disk Space
SMOKE vs. EMS-95 on OTAG 1890
IBM RS-6000 CPU and Wan-dock Time,
I
SMOKE vs. EMS-95 on OTAG 1990
Disk Storage
CD
o
EMS-95
OSMOKE
P« PMdl) 604J
KWwrio efts*
. North Carolina Supercomputing Center
2-82
-------
. Environmental Programs
Status of SMOKE
Operational Prototype
Validation
- EMS-95(OTAG)
EPS 2.0 (North Carolina)
BEIS2
Current Applications
Season Model for Regional Air Quality (SMRAQ)
North Carolina Regional and Urban Modeling
Availability
Publicly released February 1998
http://envpro.ncsc.org/products/smoke
North Carolina Supercomputing Center
Environmental Programs
SMOKE - Future Needs
Enhanced support for new species
Enhanced sorting of profiles/cross references
Run-Time Dimensioning
Control-Related Enhancements
Enhanced Error Handling
Integration with a Chemical Mechansim
Reader
Enhanced Quality Assurance
Improved User Interface
Improved input format
Integration with inventory development
. North Carolina Supetvomputlng Center
2-83
-------
. Environmental Programs
Work Plans and Level of Effort
. North Carolina Supercomputlng Center
. Environmental Programs
Further Information
Neil Wheeler
(919)248-1819
wheeler ©ncsc.org
http://envpro.ncsc.org
http://envpro.ncsc.org/EDSS/
http://envpro.ncsc.org/products/smoke/
North Carolina Supercomputlng Center
2-84
-------
Scientific and Operational Criteria
for the
Use of Eulerian Models to Compute
Reactivities
May 12, 1998
U.S. EPA
Photochemical Reactivity Workshop
Regal University Hotel
Durham. NC
Harvey Jeffries
Environmental Sciences and Engineering
University of North Carolina
Chapel Hill, NC
-< Operational Ov.-v;
Background
"Reactivity" is a sensitivity, or change in a system's
state relative to a change in the system's input.
It has been clearly established from theory and
experimental evidence that:
The reactivity of an organic compound in the
atmosphere is a strong function of both the
compound's unique properties and the
conditions of the ambient environment in which
it is reacting.
Therefore, there is not a single reactivity per
compound, but instead a continuum of reactivities.
This same phenomenon prevents the direct
application of "smog chamber" reactivities to the
atmosphere.
2-85
-------
Peak Ozone Sensitivity
11.2-H
| m
I 0.8
TJ
«3
00.6
Q.
Q.
«0.4-
o
I 0.2
TO
6 to 1 HC to NOx
4 to 1 HC to NOx
ni'«priki
PAR
ETH
OLE TOL XYL FORM ALD2 MIX
CB4 Model Species
Harvey Jeffries
All VOCs Contribute To Os
o
£
Q.
Q.
Q.
O
I
0.11
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00
-0.01
-0.02
| 6 to 1 HC to NQ,
g3 4 to 1 HC to NQ,
Trajectory Simulation
jcoi
For 2% change in TOL
\ ' Incr. Reactivity
Sum all of TOL
organic i ^~-
chemistry :
TOL
': PAR
CO | PAR I OLE I XYL I ALD
CH4 EtH T^L HCHO
I Iriit | Dil"~|
Inor Emis Loss
Harvey Jeffries
2-86
-------
Cf.-v. -<*: .md Operational Cnu>n.-
Background, cont.
Such understanding raises questions about the use of
a "bright-line" test or about a "national exemption".
Basil Dimitriades described this problem as,
All these methods of using the reactivity
concept in regulatory organic emissions
programs have varying merits and difficulties.
Scientific issues associated with obtaining a
valid, acceptably accurate measure of an
organic compound's reactivity presents a
common difficulty.
-, .< :«?;-- .jrto Operational CHU . >.*
Background, cont.
Given the newest national standards and goals,
focusing only on role of VOCs in ozone formation is
not acceptable.
Instead, VOC reactivity needs to be expaned to
include the atmospheric compositional effects that a
VOC would have on:
Peak one-hour ozone and 8-hour ozone
PM2 5 formation
Regional Haze
HAPS formation
Nitrate deposition
Conceptual Model of VOC reactivity shows complexity
2-87
-------
Ci !
Background, cont.
Previous efforts to deal with this complexity have
centered around trying to find "standard conditions"
which would ideally produce a "universal reactivity
scale", for example, manipulate ambient NOX
conditions, to produce maximum ozone response (the
MIR scale). To use these scales for policy, one must
argue their relevance to natural-use environments.
An alternative method is to assess all impacts of a
real-world VOC emissions change scenario for a set
of well-simulated regional and urban conditions and
to include the VOC's projected use conditions.
Modern, community-based Eulerian Air Quality
Models can meet these requirements.
Different Mechanisms.
i^*^-> * f\. *_» *»
toluene IR Compared by Mecli
CB4 TOLD
[ I CAL86 TOLD
SAPRC90 ARO1
E '
Harvey Jeffries
2-i
-------
M85 Less Reactive
Than lnd.Avg.Gaso5ine
0.50
0.50
oo
04-Sep-91 .
8 9 10 11 12 13 14 15 16 17 18 19
L.L.I.', I . I . I , I . I , I , I , I . I . I .
6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS,EOT
Harvey Jeffries
...But, Not
o.so
0.40
£ °'30
5
£ 0.20
b
0.10
0.00
I I I I I
I ' I ' I ' I ' I ' I ' I ' I
. 2.04 ppmC SynUrb+HCHO / SynM85 03-Sep-91
,0.50
........2.04 ppmC SynUrb+HCHO / SynlAG
- J.40
NO
0.00
567
10 11 12 13 14 15 16 17 18 19
HOURS, EOT
100
E
>-,
sf
I
S 6 7 8 9 10 11 12 13 14 15 16 17 18 19
HOURS, EOT
Harvey Jeffries
-------
Conceptual Model of Organic Trace Gas Reactivity
-' time '
/**-, -
C Property/
/ Xsect
X Q Yield
Organic Trace Gases
Abstract Property
Process Rates
Concentrations
.'Property,'
Purpose
To identify and describe the attributes of a modeling
system that can compute a VOC's (or a mixture of
VOCs') impacts on atmospheric composition.
To compare these attributes with those of existing
modeling systems.
To identify significant work that needs to be done to
use this modeling system to evaluate possible
strategies that incorporate VOC reactivity.
2-90
-------
im! Operational Criirna
Components
Simulate Target VOC's Chemistry
Simulate Target VOC's Emissions
(also Replaced VOC's Emissions)
Simulate Environments including
Target VOC
Evaluate Atm. Compositional
Changes
I will discuss each of these in detail.
!;!(' six! Operational Criteria
Simulate VOC Chemistry
Ail reactivity calculations require the use a chemical
simulation model to establish the radical and
chemical environment in which the VOC reacts.
If the reaction mechanism and reaction parameters for
the target VOC are not available, a kinetics and
chamber study must be undertaken to produce these
items. This would be the responsibility of the party
wanting to assess the reactivity of the target VOC.
This target VOC chemistry must be integrated with a
chemical reaction mechanism for the regional and
urban atmospheric environment.
Particle and HAPS formation are new data
requirements.
- -.'.! ' ietr-t-jit
2-91
-------
:JM-', anil Operational Cniet ia
Simulate VOC Emissions
To have accurate estimates of a target VOC's effects
on atmospheric composition, we must have a
accurate simulation of the emissions of the VOC,
resolved in both space and time within the test
domains. This would also be the responsibility of the
party wanting to assess the reactivity of the target
VOC.
Further, the expected growth in emissions over the
requested deferment period must be provided. In
addition, the decrease in any displaced emissions
already in the existing base and future case model
inventories must be specified.
'. H-jstiJft. and Operational Crilrrta
Simulate Environments
Eulerian models permit our most advanced and most
accurate simulations of urban, regional, and global
atmospheric chemical environments.
For reactivity assessments, we want to perform "best
operational practice" model simulations of 3 to 5
"well-simulated" and "well-understood" regional and
urban domain test cases. Ideally, a test case that
represents a significant market for the target VOC
would be included.
The responsibility for the creation and maintenance of
these test cases rests with the EPA and the States.
States and industries share responsibility to run tests.
2-92
-------
rational C>
Evaluate Simulation Results
A variety of ways exist to assess atmospheric
compositional impacts. I prefer those that indicate
the entire range of results, e.g., changes in frequency
of predicted concentrations. (Need some example
results to test meaningful ways of analysis).
The impact could be classified as:
always negative impact outcomes
non-detectable outcomes
always positive impact outcomes, and
mixture of outcomes.
UAM Surface Cell Os Frequency Distribution
One-hour at Os peak, 625 total cells
100% Std Gasoline Vehicles
o o o o o
0000
9CDCMIDO900CMI0O
SO) O O. »-»-»- CM CM O
P T T *"^ *"!*";
oooooo'ooo'ooo
'? Ozone interval, ppm
2-93
-------
UAM Surface Cell Oa Frequency Distribution
eo JB One-hour at O3peak7625 total cells
100% Nat Gas Vehicles
100% Std Gasoline Vehicles
ooooooooooo
Ozone Interval, ppm
nj
0)
£
C
-------
;.(: attona! C- < >
Evaluate Results, cont.
For always negative outcomes, the "proposed use"
should not be permitted and the VOC should be
controlled.
For non-detectable and always positive outcomes, the
"proposed use" should be permitted for a renewable,
fixed-time period.
For a mixture of outcomes, the "proposed use" might
be permitted for a renewable, fixed-time period, if the
adverse effects are "tolerably small" compared to the
costs and other societal benefits.
Community Modeling System for Reactivity
EPA/States
Research
EPA/Slates
onirounify
Modeling's
ISystem
cnemlcal
reaction
echanism, uses
chemistry ( EPA/StateS
mechanism
Research
Results
2-95
-------
! . an,!- Operalional
Work Needed
Major Research Topics
Fundamental VOC chemistry, target VOC chemistry, analytical
methods, synthesis of products for further testing, better
coordinated observations for challenging models, ambient
measurement of VOC products.
Major Modeling System Improvements
Rapid emissions processing system, sensitivity computational
subsystem, reactivity results display and analysis system.
Major Modeling Data Set Needs
Nested-grid, regional and urban emissions inventories, nested-grid
regional and urban meteorological scenarios.
Education and Technology Transfer Needs
Full educational and reactivity modeling technology transfer
package, support for a Community Modeling User's Group.
£
o
MSIS-I^U-SI >
*amjji2jiij
«
-------
"INCREMENTAL REACTIVITY FOR THE PAINT
INDUSTRY"
PRESENTATION TO
EPA PHOTOCHEMICAL REACTIVITY WORKSHOP
May 12-14, 1998
Durham, North Carolina
By
Fdward D Edwards. Ownership
Robert Wendoll. Director of Environmental A/fairs
DUNN-EDWARDS CORPORATION
CLEAN AIR ACT 1990
SECTION 183(e)
(2) STUDY AND REPORT
(A) STUDY.The Administrator shall conduct a
study of the emissions of volatile organic compounds into the
ambient air from consumer and commercial products (or any
combination thereof) in order to
(i) determine their potential to contribute to ozone
levels which violate the national ambient air quality standard
for ozone; and...
REQUIREMENTS OF 183(e
1. Ambient air validity
2. Contribution to ozone levels that violate the
iNAAQS
A. Includes from 80 ppbv O3 and up
B. Includes ambient availability, not just
content
3. Determine potential
-------
(3)
CLEAN AIR ACT 1990
SECTION 183(e)
REGULATIONS TO REQUIRE EMISSION
REDUCTIONS
to
(A) IN.GENERAL.Upon submission of the final
report under paragraph (2), the Administrator shall list those
categories of consumer or commercial products that the
Administrator determines, based on the study, account for at
least 80 percent of the VOC emissions, on a reactivity-adjusted
basis, from consumer or commercial products in areas that
violate the NAAQS for ozone.
COMMENTS:
This section requires analysis of regulations on the
same reactivity adjusted basis that matches the study.
OZONE FORMATION POTENTIAL OF
VARIOUS VOCs
A useful definition of reactivity is that of incremental reactivity,
defined as the amount of ozone formed per unit amount (as
carbon) of VOC added to a VOC mixture representative of
conditions in urban and rural areas in a given air mass (Dodge,
1984; Carter and Atkinson, 1987, 1989b; Carter, 1991),
Incremental reactivity = &[ozone]
AfVOCJ
where Afozone] is the change in the amount of ozone formed as
a result of the change in the amount of organic present, A[VOC]
(note that Carter and Atkinson [1989b] used the quantity
A[ozone]-[NO]) rather than A[ozone] under conditions where
the maximum ozone was not attained and NO was not fully
consumed). This concept of incremental reactivity corresponds
closely to control strategy conditions, in that the effects of
reducing the emission of a VOC or group of VOCs, or of
replacing a VOC or group of VOCs with other VOCs, on the
ozone-forming potential of complex mixture of VOC emissions
are simulated.
Source: Rethinking the Ozone Problem in Urban and Regional Air Pollution,
A Report to Congress, 1992, The National Academy of Sciences, pps.
153-154
-------
ATMOSPHERIC CHEMISTRY
155
TABLE 5-4 Calculated Incremental Reactivities of CO and Selected VOCs
as a Function of the VOC/NO, Ratio for an Eight-Component VOC Mix
and Low-Dilution Conditions
+ 11.0-,
VOC/NO,, ppbC/ppb
Compound
Carbon monoxide
Ethane
n-Butane
/i-Octane
Ethene
Propene
omj-2-Butene
Benzene
Toluene
m-Xylene
Formaldehyde
Acetaldehyde
Benzaldehyde
Methanol
Ethanol
Urban mix*
4
0.011
0.024
0.10
0.063
0.85
1.23
1.42
0.033
0.26
0.98
2.42
134
-0.11
0.12
0.18
0.41
8
0.022
0.041
0.16
0.12
0.90
1.03
0.97
0.033
0.16
0.63
1.20
0.33
-0.27
0.17
0.22
032
16
0.012
0.013
0.069
0.027
0.33
0.39
031
-0.002
-0.036
0.091
032
0.29
-0.40
0.066
0.065
0.08S
40
0.005
0.007
0.019
-0.031
0.14
0.14
0054
-0.002
-0.051
-0.025
0.051
0.098
-0.40
0.029
0.006
0.011
"Eight-component VOC mix used to simulate VOC emissions in an urban
area in the calculations. Surrogate composition, in units of ppb compound per
ppbC surrogate, was ethcne, 0.025; propene, 0.0167; /i-butane, 0.0375; n-pcn-
tane, 0.0400; isooctane, 0.01S8; toluene, 0.0179; m-xylene, 0.0156; formalde-
hyde, 0.0375; and inert constituents, 0.113.
Source: Adapted from Carter and Atkinson (1989b).
fl
7-
.6-
AVOC
.3-
2-
Calculated Incremental Reactivities
AO,
AVOC
at 40 ppbV NOX (in mole units)
Ethene
Toluene ------
m-Xylene =
Ethanol
Urban mix
10
20
VOC
NOX
Source Adapted from Carter and Atkinson (1989)
From Rethinking the Ozone Problem in Urban and Regional Air Pollution
-------
Isopleth Lines - Ozone Production Rate In ppbv/nr
Production of O3 as function of Nox for 3 levels of VOC's
Slaflelbach and Neflel [1997]
>s adapted from StUlmtn el aJ [1990]
U)
0
10
NOx [ppbV]
15
20
For Incremental Reactivity,
We need to look at both slopes and the transition area
Species data must be from ambient air shed chamber work
This yields production and loss as incremental reactivity,
not mechanistic reactivity
0.01
1 10
Propy-Equiv (ppbC)
1.0CO
Fig. 16. Summary of hydrocarbon/NO^ regimes. The four rect-
angles indicate the typical total [Propy-Equiv]surf and NO, concen-
tration ranges observed at I. urban/suburban sites; II, rural sites in
the eastern United States; III. remote sites in the tropical forests of
Brazil; and IV, remote sites in the marine boundary layer. Isopleth
lines are used to indicate model-calculated net rates of ozone
photochemical production (in units of ppbv per hour) at midday as a
function of assumed NO, and propylene concentrations. The cal-
culations were carried out using the photochemical box model and
methodology described by Chameides tl al. [1987] with rate con-
stants from Gery el al. [1989] and DeMore et al. [1990]. The model
adopted a temperature of 290 K, a dew point of 285 K, a solar zenith
angle of 30°, a CO concentration of 85 ppbv, a CH4 concentration of
1.85 ppmv, and an Oj concentration of 25 ppbv. The shaded area in
the figure denotes the concentration regions for which photochem-
ical processes were calculated to produce a net loss of ozone.
-------
Suggested NOX levels for Incremental Reactivities
Chamber Studies Simulating Ambient Air
001
0.00
050
1.50
2 00
250
VOC (ppmC)
1 The dotted lines indicate the following NO, levels: .001, 005. 01. 02, & 04
INCREMENTAL REACTIVITY
1. Can reasonably simulate ambient air with appropriate
chamber work.
2. Is based upon measured ozone forming potential of VOCs
in simulated ambient air.
3. Determines a VOC's "contribution to levels that violate the
NAAQS" for all ozone levels required by NAAQS starting
at 80 ppbv (8-hour average).
4. Is speciated.
5. Can validate the slopes of ozone isopleths generated by SIP
model.
6. Can be used to validate a SIP modeling process that
simulates attainment at two points rather than the current
one point method.
7. Can distinguish between naturally clean air and ozone
attainment that results from VOC reductions only
(important to determining ozone transport potential).
8. Provides BRIGHT LINE regulatory distinction by VOC
species by air shed.
Source FW. Lurmann. Sonoma Technology. Santa Rosa. Calif. (1990)
From Rethinking the Ozone Problem in Urban and Regional Air Pollution. Figure 6-2 page 171
-------
o\
9. Provides PALE LINE regulatory distinction by VOC
species by air shed to reduce overall reactivity.
10. Provides improved graphical basis for decision between
VOC + NOX controls vs. NOX only controls.
11. Can be validated for all air sheds.
12. Provides for seasonal control strategies through
temperature and sunlight variables.
13. Provides for speciated VOC regulation that declines as
reactivities reach zero or become negative.
14. Reduces the discrepancy between container VOC content
and reactive emissions to allow one adjustment factor by
product type.
15. Fits into a CTG framework.
16. Is better than a preemptive Federal approach.
PROPOSAL AND REQUIREMENTS
1. Requires a new large chamber available to
Government, Industry and Academia
2. Chamber to be built that meets the requirements of the
CAA of 1990 Section 183(e)
3. A reactivity engine and a peak 1 hr/ 8 hr simulator that
are in public domain and the best currently available
4. Support by one or more major regional or local Air
Management Districts
5. Support by Paint Companies affected by the District(s)
6. A peer reviewed process developed under NARSTO
7. Joint Funding by Districts, Government and Industry
8. EPA incorporate Incremental Reactivity into the SIP
process
-------
CLEAN AIR ACT 1990
SECTION 183(e)
(3) REGULATIONS TO REQUIRE EMISSION
REDUCTIONS.
(C) USE OF CTGS.For any consumer or
commercial product the Administer may issue control
techniques guidelines under this Act in lieu of regulations
required under subparagraph (A) if the Administrator
determines that such guidance will be substantially as effective
as regulations in reducing emissions of volatile organic
compounds which contribute to ozone levels in areas which
violate the national ambient air quality standard for ozone.
COMMENTS:
Incremental reactivity, with a CTG is far more
effective in reducing ozone for each air shed than a national
mass-based regulation.
-------
EPA Photochemical Reactivity Workshop
May 12-14, 1998
Durham, North Carolina
Presenter
The Solvents Council of the Chemical Manufacturers Association. Jonathan Kurland of
Union Carbide Corporation will make the presentation on behalf of the Solvents Council.
Title of Presentation
CMA Research Initiatives
Abstract
The Chemical Manufacturers Association (CMA) is a Washington D.C. based trade
association with over 200 member companies that collectively produce approximately 90
percent of the basic industrial chemicals produced in the United States. The CMA Solvents
Council represents the major U.S. manufacturers of hydrocarbon and oxygenated organic
sol vents. -
The Solvents Council recognizes (1) the need to better understand the role that individual
ozone precursors (including solvents) play in forming ground-level ozone and (2) the
potential benefits of designing regulatory systems that take into account differences in the
ozone-forming potential of different precursors. For this reason, the Council has been
involved in both research and public policy discussions about the concept of photochemical
reactivity.
This presentation will briefly summarize the activities of the Council in both the research and
public policy arenas, including the following:
The work being sponsored by the CMA Ethylene Glycol Ether and Propylene Glycol
Ether Panels to examine the kinetics and photochemical oxidation mechanism of
glycol ethers.
The Council's support for the HONO work being conducted by Dr. William Carter.
The Council's perspective on the Reactivity Research Advisory Committee (RRAC)
established by the California Air Resources Board (CARB).
The approach developed by the Council and submitted to CARB for taking uncertainty
into account in reactivity-based programs.
The following companies are members of the Solvents Council: ARCO Chemical
Company; BP Chemicals, Inc.; Celanese Ltd; The Dow Chemical Company; Eastman
Chemical Company; Exxon Chemical Company; Phillips Chemical Company; Shell
Chemical Company; Sun Company; Union Carbide Corporation.
3-8
-------
CMA Research Initiatives
Barbara Francis
Chemical Manufacturers Association
Jonathan Kurland
Union Carbide Corporation
EPA Photochemical Reactivity Workshop
May 13, 1998
Chemical Manufacturers Association
(CMA)
Trade Association
Over 200 member companies
Collectively produce approximately 90% of
the basic industrial chemicals produced in
the U.S.
3-9
-------
Solvents Council
Represents the major U. S. producers of
hydrocarbon and oxygenated solvents
Addresses environmental issues that
affect both users and producers
Solvents Council Members
ARCO Chemical Company
BP Chemicals, Inc.
Celanese Ltd.
Dow Chemical Company
Eastman Chemical Company
Exxon Chemical Company
Phillips Petroleum Company
Shell Chemical Company
Sun Company
Union Carbide Corporation
3-10
-------
Atmospheric Reactivity Task Group
Includes both scientists and policy
specialists
Active in:
Policy issues
Research
Policy Activities
*
Participates in developing a regulatory
framework for reactivity
Provided comments to CARB for dealing
with the uncertainty in MIR values in
regulations
Presented an alternative model for
categorizing complex hydrocarbons, such as
mineral spirits, naphthas, Stoddard Solvent,
etc.
3-11
-------
Research Interests
GARB Reactivity Research Advisory
Committee
Chemical Mechanism Studies
Reactivity studies on glycol ethers
Support for better reactivity estimation
methods
Environmental fate
Reactivity Research Advisory Committee
Participation on RRAC of the California EPA Air
Resources Board (CARB)
Choice of significant compounds for MIR
determination by Dr. Carter under his
contract with CARB
Advice on prioritization of CARB research
Support for MONO work by Dr. Carter
3-12
-------
Ethylene Glycol Ether and Propylene Glycol Ether Panels
Research Activity
Kinetics and Mechanism of Photochemical Oxidation
of Glycol Ethers
2-Butoxyethanol and 1-Methoxy-2-propanol
Work by Dr. Roger Atkinson
- Kinetics and products of oxidation
- koH and products of the reaction of the ethers with HO- in
air with NOx
- Soon to be published
Work by Dr. William P. L Carter
- Smog chamber runs
- Mechanism and kinetics derived from Atkinson's studies
- Calculation of MIRs
Principal Factors Determining the MIR
The rate of reaction with HO-
If the compound is unreactive this is the dominant term. Very
reactive compounds will be completely consumed.
Direct reactivity
The more oxidations of NO to NO2 by peroxy radicals, the
more ozone production
Effect on HO- concentration
Compounds that generate free radicals by photolysis will
increase the concentration.
Reactions of alkylperoxy radicals with NO to make alkyl
nitrates instead of NO2 and RO- reduce the concentration.
The effect on NOx concentration
Formation of alkyl nitrates removes NOx
3-13
-------
Ethylene Glycol Ether and Propylene Glycol Ether Panels
Research Activity
Determination of Nitrate Yields in Photo oxidation
Work to be done by Dr. Paul Shepson at Purdue University
Oxidation of ppm levels of glycol ethers in air with NOx
PTFE-coated chamber
Controlled hydroxyl radical source by photochemical initiation
Determination of nitrate yields by
gas chromatography
pyrolysis of RONO2 to RO and NO2
determination of N02 by chemiluminescence
- no calibration needed to quantify molar yield of nitrate
- speciation requires further work
GC/MS or synthesis of authentic samples
HONO Work by Dr. William P. L. Carter
Purpose: To support better estimation of MIRs when direct
experimental data is lacking, and avoid use of arbitrary
default values
Objective: Obtain direct reactivities to make possible a
scheme for estimating MIR values that considers by SAR
or other means koH, direct reactivity and indirect reactivity.
Direct reactivity is the NO to NO2 conversions during
oxidation of a compound. The current Carter scheme for
upper limit MIRs lumps the direct and indirect reactivity.
HONO yields HO and NO upon photolysis.
In a flow system, radical generation loses its importance and
the direct reactivity can be determined.
3-14
-------
Related Work
Other CMA Panels
sponsor chemical-specific work
Individual members of CMA acting
independently
atmospheric chemistry and kinetics of
specific chemicals
Other Interests Areas of CMA
Environmental Fate
Low-vapor-pressure (LVP) Compounds
Volatility as well as reactivity influences
ozone formation.
Down-the-drain Factor
The EPA inventory of consumer product
emissions recognizes that some VOC
largely go into wastewater and are
biodegraded at a POTW.
3-15
-------
Summary
CMA is committed to involvement in these
issues.
CMA is active in reactivity research and
policy development.
CMA is interested in related issues.
We want to play an active role in future
activities.
Acknowledgments
CMA Staff
Atmospheric Reactivity Task Group
3-16
-------
CSMA Position on the
Importance of Relative
Reactivity
D. Douglas Fratz
Director of Scientific Affairs
Chemical Specialties
Manufacturers Association
Washington, DC
Presented at (lie
EPA Photochemical Reactivity Workshop
May 12-14,1998
Durham, North Carolina
Good morning. I am Doug Fratz, Director of Scientific Affairs for the Chemical
Specialties Manufacturers Association in Washington. DC. CSMA represents
manufacturers of formulated chemical products for household, institutional.
commercial and industrial consumers. These products include cleaners.
disinfectants, pesticides, polishes, automotive products, and aerosols of all types
I'd like to talk to you today about CSMA's positions on the importance ol
relative reactivity to ozone attainment strategies.
Importance of Reactivity
« History of CMSA Support
** Importance to Ozone Attainment
Strategies
> CSMA Research on Reactivity
** Future Research Needs
In presenting our current positions. I'll also review the history of our industry's
support tor reactivity: the many ways in which reactivity is important to cost-
effective ozone attainment; some studies CSMA has sponsored over the past 10
years, and some concepts for future research that we believe is still needed
-------
oo
History of CSMA Support
?- 1988 - CARB Consumer Products
Program
»* 1990 - 183(e) of Clean Air Act
» 1991 - CARB/South Coast Conference,
Irvine
>*- 1993 - Support for VOC Exemptions
>* 1994 - California SIP Hearings
We began urging the consideration of reactivity in 19X8 in relation to ihe
California ARB regulations on consumer products, which began wnh underarm
products. We also helped develop Section I83(e) of the Clean Air Act of I WO
on Consumer & Commercial Products, a section which contains requirements lor
EPA lo consider reactivity in relation to those products. 1 gave a talk at the 1991
Irvine conference on reactivity, supporting the consideration of reactivity for
consumer products. Around 1993, we began supporting EPA exemptions for
negligibly reactive VOCs. urging that standard criteria and protocols be
promulgated. In 1994, we were successful in getting the California Air
Resources Board to include the potential to consider reactivity in the consumer
products element of the State Implementation Plan.
History of CSMA Support
» 1995 - EPA Report to Congress -
Chapter 3
» 1996 - Reactivity Principles to CARB
» 1997 - Proposed CARB Reactivity-
Based Compliance
» 1998 - Proposed CARB Regulatory
Language
In 1995. EPA's Report to congress on Consumer & Commercial Products was
finally completed. We worked closely with EPA to assure that they met their
statutory requirements lo consider reactivity under IK3(e). Chapter 3 of the
report and the regulatory prioritization process EPA conducted met those
mandates. By 1996. we were working again with California on reactivity.
putting forth the principles we believed were essential to establishing reactivity-
based compliance options for consumer products, as allowed in the SIP. Since
1997, we have also participated in ARB's Reactivity Research Advisory
Committee. In 1997. we outlined the specific elements of a consumer products
reactivity program, and this year we have developed specific draft regulatory
language upon which a voluntary reactivity compliance option can be based.
-------
CSMA Policy Positions
»-EPA VOC Exemptions Program
Ethane Criteria
Mass-Based Criteria
Standard Criteria & Procedures
Expeditious Review
Now. I'd like in review our current positions on leaciivnv related r,sues VW
support retaining ethane as the standard lor exernritm;' VOCs as /i<-i'li;ribly
reactive There is no need to reproduce the 1977 exprimii'iii ih.it (iiovides iln-
basis lor choosing ethane We may need to inventory some i-.vm|iii-d \'()('> loi
use in modeling, bui there is no need in.subject idem toconiiols Wi- lavoi
ma.vs-ba.sed, as opposed to mole-based, comparisons lor VfX' exemptions All
ozone attainment inventories, controls and strategies to date have hern mass-
based. We support EPA developing standard criteria and procedures lor
exemptions. And we support expeditious reviews of exemption pennons Our
industry ha.s relied on these exemptions to meet regulatory siandanlv and they
arr i-ssential to our meeting the o/.onr ri-diK non goals brmj' pl.i> >-il upon us
CSMA Policy Positions
<* Consideration of Relative Reactivity in
Ozone Attainment
Alternative to Mass-Reduction
Regulatory Prioritization
Increased Cost-Effectiveness
Tost Effectiveness Based on Cost Per
Ozone Reduction Benefits
We- belirvr ili.n consideration ol relative reactivity is essential in mjny other
aspects .is well In particular, u is an important alternative 10 simple mass basic
V(K' rrdui nons We support allowing regulated parties to use reactivity
reductions in painally or lully meet their VOC mass reduction goals Tins
would result in increased cusi-ettectiveness in moving toward ozone reduction
goals Finally, we .support all regulatory analyses of ozone attainment options
being based on cost per ozone reductions, not precursor reductions This would
[mi all V(X.' controls, and even NOx controls, on an even playing lield
-------
u>
rO
O
CSMA Policy Positions
Reactivity Positions for
Consumer Products
. Principles
- Optional
- Sound Science/Stable/Flexible
- Cost Effective/Efficacy
Maintained
- Existing Provisions/Enforceable
These are the principles we proposed in California in 1996. Reactivity
considerations must be optional, to allow companies to determine whether
reactivity or mass-reductions are more cost-effective. Our VOOs already are of
low reactivity. Mandatory reductions would not make sense. The reactivity
values assigned to VOCs must be based on sound science, and be both stable (so
manufacturers will use them) and flexible (to encourage continued research).
The program must encourage cost-effectiveness, and allow manufacturers to
make products that are both safe and effective. We believe that all of the
provisions related to the mass-based standards, such as the various exemptions.
must be maintained. Obviously, the program must also be enforceable.
CSMA Policy Positions
>- Reactivity Positions for
Consumer Products
Policy Position
Optional
Product-By-Product Ozone Equivalence
Maintain Exemptions
Scientifically Determined MIRs
All Products
We reiierated these same positions in our updated policy positions over the past
year, and added a few more specific ones. In particular, we believe the most
sound program would assure product-by-product equivalence of ozone impact.
or equivalence by groups of products. We would like to see a program that
could benefit all consumer products, and be adapted for other industries as well.
-------
to
CSMA Consumer Products
- Reactivity-Based Compliance Option
and Credit Program
Equivalent Ozone Impact on Product-By-
Product Basis
Official List of Reliable Ml Rs
Over-Reduction Credits Transferable
between Products
This year, we developed a specific proposal tor a reactiviiy-based. voluniary
compliance option for consumer products in California. Our program is a
relatively simple one. It is designed to assure ozone equivalence on a product-
by-product basis. It would be based on a list of VOCs with reliable MIR
values. VOCs without reliable MIRs would not be allowed to be traded, and
have to meet mass-reduction goals. We are proposing, however, that excess
reductions credits for one product be usable for products which can't meet their
reduction goals. This adds further flexibility. We are not proposing any
complex schemes to address uncertainty. Only reliable MIRs would be used.
What is often overlooked is that mass-based reductions also demonstrate high
uncertainties regarding ozone reduction. In consumer products, we already have
generally minimized VOC content for cost reasons. Mass reduction
reformulations usually require different VOCs. These might be 3 times more
reactive, or 3 times less reactive, than those they replace. The We uncertainix
for reactivity is small when compared to this 3(X)"r level nt uncertainty. Bui
even the.se larte uncertainties will even out over Hills ot products.
CSMA Research on Reactivity
1988 - UAM Study on Underarm
Products in California
- 1993 - UAM Study on Pesticides,
Disinfectants, Air Fresheners in New
York
1997 - Impact of Consumer Products on
California Air Quality
Now. I'd like to review some of the reactivity studies we've done over the past
decade All have been modeling studies on the effects of consumer product
emissions on ozone formation. The three studies are: a 19X8 study on underarm
products in California; a 1993 study on pesticides, disinfectants and air fresheners
in New York: and. a study completed just last year entitled. "Impact of Consumer
Products on California Air Quality" All of the modeling in these studies was
done by Dr. Gary Whitten.
-------
UJ
K)
ro
1988 UAM Study
» Underarm Products in
California South Coast AQMD
. 0.06% VOC Inventory
No Measurable Impact
0.2ppb
One Third as Reactive
The first study, way back in I98X, looked a! the impact of underarm
untiperspirants and deodorants in the Los Angeles basin. These products
represent only about 0.06% of their VOC inventory, and are mostly composed ol
ethanol. simple hydrocarbons, and other low-reactivity VOCs. The modeling
showed that even eliminating all emissions would have no measurable impact on
peak ozoneless than 0.2ppb, primarily because of their low reactivity. This
study may have been done before its time. It was spectacularly unsuccessful in
convincing California not to regulate. Underarm products remain targeted for an
80% reduction that our industry still hasn't figured out how to meet.
1993 UAM Study
>* Pesticides, Disinfectants, Air Fresheners
New York City Metropolitan Area
. 0.1% Ozone-1990
. 0.25% Ozone -1997
No Measurable Impact
One Third as Reactive
We are not un industry that gives up. however, and our second study was more
successful with similar findings. This time, we looked at household pesticides.
disinfectant and air fresheners in New York City. The ozone contribution here
was O.\c'f in 1990. 0.25% in 2(X)7. Again, the low reactivity contributed to the
unmeasurable ozone impact. This study convinced New York to promulgate
only reasonable standards for these products.
-------
U)
NJ
1997 Study - Sierra Research
'-"Impact of Consumer Products on
California's Air Quality"
Inventory
MIR-Weighted Inventory
UAM Sensitivity Runs
Our final study, completed last year by Sierra Research, was a multi-phase study
looking at the inventories and reduction goals of the IW4 C.ililorma SIP The
study included correcting the VOC emissions inventory tor consumer products;
creating a reactivity-weighted inventory for aj] VOC sources: and. dome Urban
Airshed Modeling to determine the ozone benefits of the additional VOC
reductions for which consumer products were targeted in the SIP. Both South
Coast and Sacramento UAM runs were done.
1997 Sierra Research Study
> Consumer Products Inventory
corrections
. Non-emitted VOCs
Survey Errors
Industrial/Agricultural Products
The inventory corrections for consumer products included removing non-
emitted VOCs. most of which were those biodegraded in waste water treatment.
This correction was made by EPA in its consumer and commercial products
inventory, but has never been made by California. The study also corrected
errors in consumer products survey data, mostly non-VOC ingredients that were
misreported as VOCs. And, products that were industrial, agricultural, or
otherwise double-counted in other VOC inventories were removed. The result
was a change in the totaJ VOC emissions inventory for consumer products in
California from 265 tons/day to less than 215 tons/day.
-------
1997 Sierra Research Study
> MIR-Weighting of Inventory
Specialization of Inventory
MIR Estimates
Less Than One Half as Reactive
MIR Values for Consumer
Products and Other VOC
Sources
2.6 1.3
D AitraKt fur
SutTumtniti H/II
Consumer Products
DAifr»K« for South
Coast »/« Consumer
Products
QSitrra Kstinmlr
Consumer Products
The reactivity-weighting was done tor all VOC .sources in the IW-4 SIP
inventory. MIR estimates were used to create an MIR-wea'hied inventory. The
consumer products VOC inventory was found to be less than halt'as reactive as
the overall inventory.
This slide shows how consumer product reactivity compares to the overall 1990
VOC inventory in California. Consumer products emissions average an MIR ot
1.3. The average for other emissions was found to be 2.6 in South Coast and 3.0
in Sacramento.
-------
1990 Sacramento VOC
Emission Inventory
A3
45
40
22
li
12
II
10
J
4'
* ',..
'. -A
&..
«# ,
->'
Toul - 2 1 y loru/diy
On-RMdExhuut
D On-Rued Ev»por»»k
$o)<*MUn
OffRtmd
OUIYodAMktf
OCeaimiHQo
OCoMQnMr Product*
QMlKProtnc
This slide, which I hope you can read, shows the 1990 Sacramento inventory by
VOC tonnage. The emissions inventory is dominated by on-mad exluust. on-
road evaporation, solvent use. and off-road emissions. Smaller, but still
seemingly significant contributions occur form oil production, combustion.
consumer products, miscellaneous processes, and industrial processes
1990 Sacramento Ozone
Forming Potential Inventory
273
HX
X4
71
54
32
IK
15
5
lOn-Koid
odl c 648 lons/diy
aorr-Raid
On-Koid EMporilion
Sohf nl I ir
Q M he Proctisri
OOil Prod & Mkle
OConsumrr Products
Dlndo>lri.l I'rncr"
Here is lhat same Sacramento inventory that is MIR-weighted On-road exhaust
now dominates the ozone production inventory, with oft-road also increasing its
contribution All of the other sources, including consumer products, now
represent a much smaller portion of the inventory, and can be seen to be
makjng a much smaller contribution to ozone formation.
-------
u>
to
1990 South Coast VOC
Emissions Inventory
43V
293
261
ISO
110
84
84
S3
19
^
On-Koud KxhmiM
On-Roiicl Kvuporutioii
OfTKuttd
QMiw: TrtHTt-ss
DCoiviunter I'ruducLs
OlnHiisirUIIViHirxs
D C!omhiistiiin
TuUl« 1493 lons/d.y
Here is the same inventory, this time for South Coast. Once again, on a VOC
tonnage basis, on-road exhaust is closely followed by solvent use. on-roud
evaporation, and off-road, with others making smaller contributions to VOC
emissions.
1990 South Coast Ozone
Forming Potential Inventory
1710
MO
5(13
441
12V
111
102
HO
A3
On-Rimd Klh.uil
OOffRo.d
On-Rowi Ev.purL.llon
Solvrnl U«r
Oil Prod 4 Mkm
B Curnumrr Product*
D Mlsc Protrrsscs
D Indu*lri»l Procev.
D ComhiBlion
Toul - 3749 lons/d.y
And. once again, in the MIR-weighted inventory, we see the actual ozone
contribution of on-road exhaust and off-road increase significantly, while the
contributions of other sources, including consumer products, go down
proportionately. Consumer products here contributed only less than 3% of
ozone formed. We believe that this type of analysis could provide a valuable
tool in prioritizing ozone precursor reductions. The MIR-weighting technique.
however, can only look at VOCs; the use of airshed modeling could allow this
consideration to include NOx as well.
-------
u>
to
1997 Sierra Research Study
» UAM Sensitivity Runs
SCAQMD and Sacramento
30% and 85%/38% Reductions
Peak Ozone Remained within 1-Hour
Standard
UAM Runs
Effects of Consumer Product Controls
South Coast - 2010
» 85% Controls
>« 30% Controls
One-Hour Peak Ozone
* 122.3
* 124.9
>» One-Hour Standard »» 125
Our study also included UAM sensitivity runs by Dr. Whitten. These runs were
using the uncorrccted VOC inventories in the SIP. even though we knew
consumer product emissions were overestimated. Both South Coasi and
Sacramento runs were done, with the base case being the 30<7
-------
to
oo
UAM Runs
Effects of Consumer Product Controls
Sacramento - 2005
'» 38% Controls
'30% Controls
One-Hour Peak Ozone
>» 124.2
* 124.5
'» One-Hour Standard » 125
Likewise, here are the data on peak ozone in the Sacramento run. In this ca.se.
the additional consumer product reductions in 2 UAM Sensitivity Runs
Reasons for Low Peak Ozone Impact
Low Reactivity
Emissions Geography
So in both South Coast and Sacramento, no additional controls on consumer
prodycts still result in levels less than 125ppb in 2010 and 2005. respectively, if
all other VOC and NOx reductions in the SIP are made. The two reasons why
this result occurred were determined to be that, first, the low reactivity of the
consumer product emissions, and. second, where the emissions occur
geographically. They are distributed in the grids of the model by population.
which does not allow them to contribute ozone formation proportionally in the
areas of peak ozone formation.
-------
to
Reactivity - Future Directions
Continued VOC Exemptions
Reactivity-Weighted Trading
Cost-Effectiveness on Ozone Basis
These are our recommendations for future directions in ozone attainment policy.
First, we support continued exemptions for negligibly reactive VOCs. Second,
we support the establishment of broad-based reactivity-weighted trading
programs. Finally, we support the use of cost-effectiveness analyses based on
ozone reductions instead of precursor reductions in looking at regulatory
options.
Future Research Needs
>* Identification of Negligibly Reactive VOC
»» Improved Precursor Inventories
Tonnage
Speciatfon
>- Chamber Studies to Establish Additional,
Reliable MIRs
'* Chemical Mechanisms Studies
» Reactivity Estimation Protocols
»» UAM Sensitivity Runs
'» UAM Cost-Effectiveness Optimization Runs
Right now our scientific knowledge is many years ahead of our regulatory
policies, but we do have suggestions for further research. Research is needed for
identifying more negligibly-reactive VOCs. Research is needed to improve
precursor inventories, both in terms of tonnage and speciation. Chamber studies
are needed to establish additional MIR values. Chemical mechanisms studies are
needed to improve both MIR determinations and the computer models.
Sensitivitity runs, like those we did in the Sierra Research study: could help to
evaluate regulatory options. In addition, similar computer modeling of regulatory
options should be run routinely to optimize the cost-effectiveness of the
regulatory options being considered for ozone controls, both those for VOCs and
NOx.
-------
Conclusions
>*- Reactivity Consideration Essential
?* CSMA Supports Reactivity-Based
Compliance Options for Consumer
Products
> Additional Research Needed
In conclusion, CSMA and the consumer products industry continue to believe
that the consideration of reactivity is essential to ozone attainment policy. We
support continued VOC exemptions as well as cost-effective, voluntary.
reactivity-based, compliance options. We also believe that continued research is
needed. The best incentive for encouraging our industry to fund research,
however, would be to see ozone attainment policies change in response to the
many years of research that has already been accomplished. Thank you.
-------
Reactivity Concerns
Philip J. Ostrowski
Occidental Chemical
Photochemical Reactivity Workshop
May 13, 1998
5/11/98
Summary
Regulations based on multiple day
reactivity should be used for the best
long term scientific approach
More data is needed to properly
implement reactivity regulations
5/11/98
3-31
-------
Summary
For the interim
- Old VOC exemptions must continue to
be used
- No new exemptions of marginally
reactive compounds
- Use the reactivity of methane as a cut
off for new exemptions to avoid long
term mistakes
5/11/98
Summary
There are practical concerns on the
enforcement side of reactivity
regulations
5/11/98
3-32
-------
Current Two Tier VOC System
VOCs are treated equally
- VOCs differ significantly in O3 formation
potential
- Sources are not encouraged to use
VOCs with low reactivity
Current Two Tier VOC System
Exempt compounds are treated
equally
- Exempt compounds differ significantly
in O3 formation potential
- Sources are not encouraged to use
exempt compounds with low reactivity
5/11/98
3-33
-------
Reactivity Regulations
Reactivity based VOC regulations
make sense for the long term
- Sources will be encouraged to use
materials which have low reactivity
- Reactivity must be based on weight
since most emissions are measured by
weight
- Air quality should improve
Reactivity Regulations
Reactivity regulations must be
flexible.
- Relative ranking for the one day box
model has in some cases changed
significantly over time
- Need flexibility to accommodate better
data (up or down)
5/11/98
3-34
-------
Reactivity Regulations
Reactivity regulations must have
scientific input
- Need scientific community consensus
for values of reactivity
- Uncertainty must be addressed
- Need periodic review as more
information becomes available
5/11/98
Current Reactivity Thinking
Box model
One day episodes
Absolute reactivity differs in different
air shed
Relative ranking is about the same in
all airsheds so regulations are
possible
5/11/98 10
3-35
-------
Current Reactivity Thinking
What makes sense in California may
not work for other parts of the
country
- LA has little influence from upwind
sources
- LA has unusual weather and geography
Weather includes lots of sunshine and a
persistent high that causes inversions
Mountains add to the trapping of pollutants
5/11/98 11
National Reactivity Regulations
Need regional models East of
Mississippi
Cities in the Northeast corridor are
affected by transport of ozone and
precursors from upwind sources
5/11/98 12
3-36
-------
National Reactivity Regulations
1 Use multiple day reactivity,
compared on an integrated O3 scale
-The relative reactivity ranking of some
compounds in multiple day episodes
shows significant differences versus
the one day box model
- Incorrect reactivity could encourage
use of the wrong chemical
/11/98 13
National Reactivity Regulations
Second highest one hour average O3
is not reliable for measuring trends
- Compounds should be compared on an
integrated O3 scale
Relative reactivity scales must be
developed for multiday events using
an integrated O3 scale
More Data is Needed
5/11/98 14
3-37
-------
For the Interim
Old VOC exemptions must continue
to be used
-Without the old VOC exemptions non
compliance would soar
- Sources may revert back to old highly
reactive compounds
-Time must be allocated to switch from
two tier to reactivity systems
5/11/98 15
For the Interim
No new exemptions of marginally
reactive compounds
- Mistakes could be made with excess
use of the wrong compound
- Multiday reactivity will increase or
decrease the relative reactivity of
compounds
5/11/98 16
3-38
-------
For the Interim
No new exemptions continued
- Compounds in the EPA VOC exemption
petition queue may have different
relative rankings in multiday reactivity
evaluations
- Some of these compounds will have
increased relative reactivity
17
For the Interim
No new exemptions continued
- More VOC exemptions could encourage
widespread emissions of a compound
which may adversely impact air quality
- Marginally reactive exempt compounds
can be used to replace exempt
compounds with much lower reactivity
5/11/98
18
3-39
-------
For the Interim
Use the reactivity of methane as a
cut off
-This approach should eliminate
mistakes since methane has very low
reactivity
5/M/9S 19
Enforcement Policy
Complex analytical issues must be
dealt with
- Laboratory Methods
- GC Mass Spec
- Method 24
-Variability issues
-Theoretical composition versus actual
Production Records use
5/11/98
20
3-40
-------
Categorization of Low Reactivity Compounds
John G. Owens
3M Chemicals
St. Paul, MN USA
It is recognized that all organic compounds which volatilize into the atmosphere do not
contribute to the formation of tropospheric smog. Some compounds have low reactivity
with respect to common atmospheric removal mechanisms such as photolysis and reaction
with hydroxyl radicals. These low reactivity compounds are stable enough to become well
dispersed throughout the troposphere prior to the onset of their decomposition. As a
result, these organic materials are incapable of contributing to the production of ground
level ozone.
As compounds are considered for exemption from VOC regulations, it would be useful to
first categorize them based upon their reactivity. Those which are shown to have low
reactivity will not contribute significantly to ground level ozone regardless of their ozone
yield during decomposition. These materials could be considered for exemption from
VOC regulations without need for further information. Compounds of higher reactivity
have the potential to add to ground level ozone during their decomposition. These higher
reactivity compounds could be selected for further study to determine the ozone
production from their atmospheric oxidation. Such categorization could streamline the
exemption process and focus resources on the study of compounds which have the
greatest potential to impact smog formation.
3-41
-------
Categorization of Low Reactivity
Compounds
John G. Owens
3M Chemicals
Criterion for VOC Exemption
(x) < kOH (C2H6)
(compound x is atmospherically longer lived than ethane)
> k OH (C2H6) - 0.24 x 10-12 cm3/molecule-s
which translates into an atmospheric lifetime of
approximately 0.24 years
> Additional criteria being being developed such as
ozone production during tropospheric decomposition
> Reactivity of compounds in recent exemption requests
have been very close to and in some cases higher
than ethane
3-42
-------
Very Low Reactivity Compounds
> Some compounds are very low in reactivity
koH (x) « koH (C2H6)
and do not photolyze in the lower atmosphere
e.g. saturated, halogenated compounds
> Compounds with sufficiently low reactivity will
be well dispersed throughout troposphere during
decomposition
> Oxidation of these compounds does not
contribute measurably to ground level smog
regardless of O3 yield.
Categorization of Low Reactivity Compounds
> Useful to establish a criterion which distinguishes
between :
I. Compounds with
reactivity relatively close
to ethane
i.e. more likely to
contribute to smog
formation and smog
chamber studies
necessary
Vs.
II. Compounds which are
significantly less
reactive than ethane
i.e. clearly will not
contribute to smog
formation and k0H
and photolysis data is
sufficient
3-43
-------
Transport Times for
Chemical Species in the Atmosphere
The longer a compound survives in the atmosphere, the greater the
proportion transported to high altitudes by diffusion and convection.
30km
1 to 3 years into stratosphere
10km
1km
days to weeks to reach tropopause
hours to reach planetary boundary layer
Categorization of Low Reactivity Compounds
> further clarify VOC definitions
> focus resources on evaluation of compounds which
are more likely to be contributors to smog formation
> streamline exemption process
> provide benefit to chemical users
3-44
-------
EPA Photochemical Reactivity Workshop, May 12-14, 1998, Durham, NC
Presenter:
Daniel B. Pourreau, Ph.D., Coatings Development Manager, ARCO Chemical
Company.
Title of Presentation:
Impact of a Molar Ethane Standard on the Number and Type of VOC-Exemptible
Compounds; Practical and Environmental Implications.
Abstract:
Since 1977, the USEPA has granted several industry petitions to exempt specific
organic compounds on the grounds of "negligible photochemical reactivity". Most of the
early exemptions were granted on the basis of kinetic reactivity data, that is the rate of
hydrogen abstraction by atmospheric OH radicals. The compound's hydrogen-
abstraction rate constant, kOH, expressed in molar units was typically compared to that
of ethane. Compounds with kOH constants lower than ethane were considered
"negligibly reactive" and appropriately exempted from VOC regulations.
More recent petitions have been granted on the basis of both kinetic and mechanistic
data. The reason for relying on mechanistic reactivity was the realization that several
compounds with negligible kinetic reactivity had significant ozone yields when irradiated
in the presence of other more reactive gases and NOx pollutants. Mechanistic reactivity
is defined as the incremental amount of ozone formed when a compound is added to a
polluted atmosphere under well defined conditions.
The compounds' maximum incremental reactivity, or MIR, expressed on a weight basis
has been the published standard since the EPA's Report to Congress in 1995 and was
the basis for the exemption of Acetone from VOC regulations. Here, the "cutoff
between reactive and "negligibly reactive" compounds was the MIR of ethane on a per
gram basis.
Since then, the EPA has received several petitions from industry to exempt other
-compounds based on MIR data. Because of this and concerns about possible future
.petitions, EPA is now considering a tightening of the MIR standard by requiring that
^compounds be less reactive than ethane on a per mole basis.
We will present evidence that shows that the number of useful compounds that meet
the current gram-based MIR standard is limited and that the proposed mole-based
standard would severely limit the number of useful VOC-exempt compounds available
to industry. The impact this policy change would have on the industry's ability to meet
current and future VOC regulations wilPalso be discussed.
3-45
-------
Introduction - Title Slide
Good Morning. First, I'd like to thank the EPA for calling this workshop and giving
Industry the opportunity to participate. ARCO Chemical also has a petition before the
EPA to exempt a new solvent from VOC regulations. I will, however, not discuss this
petition oday, only to the extent that it illustrates how such low reactivity materials can
provide immediate and substantial environmental benefits by replacing more reactive
VOCs still in use today.
The main thesis of my presentation is that the EPA's proposal to adopt a new, stricter
standard based on the photochemical reactivity of ethane on a per mole basis would
virtually eliminate all viable substitution candidates. The impact of such a decision
would be two-lold:
1. It would strengthen the Industry's opposition to further mass-based VOC limitations
and,
2. it would favor substitution to environmentally persistent halocarbons.
Neither would achieve anywhere near the environmental benefits which would result
from the exemption of a handful of solvents which meet the EPA's current gram-based
exemption criterion.
Slide 1
t
Let me start by illustrating the magnitude of the challenge we face today. Based on
recent Industry analyses, the US Coatings, inks, and adhesives industries alone
consume close to 5 billion pounds of solvent per year, despite all the regulatory efforts
to decrease their usage. As you'll see later on, there are many reasons why solvents
continue to be popular tools in these industries.
Slide 2
Low VOC technologies such as water and powder have made significant strides in the
past few years but are still not suitable for many applications and often lag in
performance compared to solvent-based systems. For example, you still cannot
powder coat a bridge nor can you repaint it with water in cold and damp weather. These
are realities we have to deal with.
Abatement technologies have also been very helpful in reducing the amount of solvent
emitted into the atmosphere. But solvent recovery is not often practical and incineration
generates NOx.
For these reasons, the EPA should continue to encourage the development of low
reactivity solvents and their use in place of many of the more reactive and toxic solvents
still in use today.
3-46
-------
Slide 3
By far the most popular solvents in these industries are aliphatic and aromatics
hydrocarbons and oxygenated solvents such as esters, ketones, and alcohols.
Halogenated solvents, which are still popular in the cleaning industries because of their
high solvency and low flammability account for less than 1% of solvent usage in the
coatings industry.
Slide 4
The reasons for choosing hydrocarbon and oxygenated solvents are numerous.
The best solvents have intermediate solvency. Strong enough that they reduce resin
viscosities effectively but not so strong that they strip primer coatings or attack the
substrate. This is one of the reasons strong solvents such as methylene chloride are
almost never used in the coatings industry.
Waterborne coatings also require solvents to stabilize the paint formulation and improve
film formation. Here, the best solvents are oxygenated solvents such as glycol ethers
and N-Methyl-pyrrolidone. Hydrocarbons and halogenated solvents are not useful here
because they are essentially insoluble in water.
Different coating operations also require different drying or evaporation rates. Fast
solvents are used in ak-dried systems, slower solvents under bake conditions. Most
often, blends of solvents are used to tailor the dry time to the specific operation and
optimize the performance and appearance of the coating.
Another reason for choosing solvent-borne systems is that their dry times are
independent of environmental conditions. This is not the case with waterborne systems
that dry much slower under cold and damp conditions.
Most solvents used today have relatively low toxicities. Solvents such as benzene are
no longer used and those solvents listed as Hazardous Air Pollutants are strictly
regulated by OSHA in the workplace and by State and Federal Environmental
Protection Agencies. Many companies have already reformulated their products with
non-HAP solvents such as Cypars, P-series glycol ethers, alcohols, and esters.
These solvents are also inexpensive. Because paint is sold by the gallon and solvents by the
pound, low density is a distinct benefit. Halogenated solvents, in contrast, have high densities
and are relatively expensive on a volume basis.
Another advantage of non-halogenated solvents is that emissions can be incinerated.
They typically have good fuel value and no corrosion issues related to halo-acid
formation. The coil coating industry, for example, is effectively using solvent
incineration as an abatement and energy producing tool.
3-47
-------
Finally, non-halogenated solvents typically have low environmental persistence. Since
they do not contain halogens, they have zero ozone depleting potential and low acid
rain contributions. They rapidly oxidize in the atmosphere to water and carbon dioxide
which has relatively low global warming potential.
Slide 5
I'd like to expand a little more on this point with this table which compares the ozone
depleting and global warming potentials of the major halogenated solvents and the
average of several halogenated and non-halogenated solvents. As you can see, the
atmospheric lifetimes of halogenated solvents are typically measured in months
whereas non-halogenated solvents typically oxidize in the matter of days.
Since the global warming potential of a solvent is a function of its atmospheric lifetime,
its ozone depleting potential, and its total infrared absorbance relative to carbon dioxide
it is easy to see why halogenated solvents are likely to have a greater impact on global
warmr.g than their non-halogenated counterparts.
Slide 6
Which brings me back to the challenge we face today:
Replacing close to 5 billion pounds of solvent per year with substitutes that generate
less tropospheric ozone yet do not have a lot of health and environmental baggage
attached to them. v
This graph provides a clue as to how we might achieve that. Of the top 20 solvents
used in the coatings industry today, only a handful contribute significantly more than
their actual emissions to tropospheric ozone formation. These are the aromatics, and
to a lesser extent, some of the higher members of the ketone, alcohol, and glycol ether
families.
Clearly, one way to significantly reduce ozone formation would be to develop policies
that encourage substitution of these highly reactive VOCs with less reactive ones. And
to a certain extent, the EPA's current policy does that.
Slide 7
By exempting VOCs with incremental reactivities less than ethane on a per gram basis,
the EPA has given Industry the incentive to develop and use low reactivity solvents
such as acetone and PCBTF. Petitions for several other solvents with similar or lower
reactivities are before the EPA today. Granting these petitions could result in
immediate and substantial environmental benefits which are illustrated here.
With additional exempt solvents, current ozone Jevels could potentially, with time, be
reduced by 88%, assuming that all exempt solvents had reactivities equal to ethane on
3-48
-------
a per gram basis. In fact, some of the petitioned solvents have reactivities less than
half that of ethane so ozone reduction opportunities of over 90% are conceivable.
Which brings us to this key question:
If such substantial benefits are achievable using ethane on a per gram basis as the
cutoff between exempt solvents and VOCs, why not make the standard stricter? Why
not go to ethane on a per mole basis?
Slide 8
The reason you do not benefit from using a molar ethane cutoff is that you drastically
reduce the number of viable exempt substitutes for more reactive VOCs. As this graph
illustrates, the number of practical non-halogenated solvents which meet the current
MIR standard on a per gram basis is approximately ten. Going to a mole standard
reduces this number to one.
Given the wide range of properties required from today's coating formulations, having
10 viable substitutes for high reactivity VOCs is a minimum. Having one is essentially
useless.
Another potential impact of this new proposed standard is that it would favor
halogenated solvents. As I illustrated earlier, halocarbons have limited use today in
coatings, inks, and adhesives and have relatively high atmospheric persistence. A
stricter ethane standard would have relatively little impact on the number of exemptible
halogenated solvents and could drive Industry to turn to these solvents to meet the new
and more stringent mass-based VOC limits. This could have a negative impact on
worker health, acid rain, global warming, and stratospheric ozone depletion.
Slide 9
I'd like to leave you with a real life example of how low reactivity solvents can be used
to replace more reactive and toxic ones. What we have done here, and for several
other coating formulations, is taken a conventional high solids formulation and replaced
the solvents with lower reactivity and non-HAP alternatives. We then calculated the
.ozone impact of each formulation on a pounds ozone per pound of solids applied basis.
Finally, we compared these solvent-based formulations to a standard waterborne
formulation.
Slide 10
This slide graphically illustrates that reformulating conventional solvent-borne systems
with low reactivity and non-HAP solvents can significantly reduce the ozone impact and
toxicity of these formulations. In this particular case, we were even able to lower the
ozone yield and HAP content of the solvent-borne system below that of the waterborne
3-49
-------
system.
You will notice, however, that the low ozone formulation has the same solvent content
as the conventional system. Without VOC exemptions, there would be no incentive for
industry to go to this type of formulation.
Worse yet, if the EPA goes to a mole based ethane standard or suspends further
exemptions, Industry will have the incentive but no tools to reformulate with.
Slide 11
In conclusion, we urge the EPA to continue exempting VOCs based on their reactivity
relative to ethane on a per gram basis. The status of exempt solvents can always be
revisited at a later date in light of newer and better science. It would counterproductive,
however, to wait for the outcome of what will likely be a lengthy and complicated
process to reap the benefits that can be achieved today, with the current policy.
Thank you for your time.
3-50
-------
Impact of a Molar Ethane Standard
on the Number and Type of
VOC-Exemptible Compounds;
Practical and Environmental
Implications.
EPA Photochemical Reactivity Workshop
May 12-14, 1998
Durham, NC
Daniel B. Pourreau, Ph.D.
Coatings Development Manager
ARCO Chemical Company
1996 US Solvent Usage in Coatings by End Use
Special Purpose
OEM
Architectural
3.2 Billion Ibs
Kusumgar, Nerfli & Growney. 1&97
. Powder limited to larger volume OEM operations
. Waterborne coatings have performance limitations
. Abatement not practical in architectural & special purpose sectors
. Incineration generates NOx
1997 US Solvent Usage by the
Coatings, Inks, and Adhesives Markets
Million Pounds
4000 ,
3000
2000
1000
3,240
Coatings
Inks
BOO
Adhesives
. Solvents continue to be widely used
Hydrocarbons and Oxygenated Solvents Popular
,l! Aromatics23%
Ketones 14%
Aliphatics 32%
Glycol ethers7%
Esters 11%
Alcohols 11%
3.2 Billion Ibs
Kusumgaf. Nerfli A Growney. 1997
, Hydrocarbons and oxygenated solvents account for >99% of usage
, Halogenated solvents account for <1% of coating solvent usage
-------
Reasons Why Hydrocarbon and Oxygenated
Solvents are Popular Coating Solvents
. Physical and Solubility Properties
. Intermediate solvency for coating resins
. Some can stabilize resins in water
. Range of evaporation rales
Dry times independent of conditions
. Low cost per gallon
. Environmental, Health, & Safety
. Can be incinerated
. Relatively low toxicities
. Low environmental persistence
. No ozone depleting potential
. Low acid rain contribution
. Low global warming potential
to
Top 20 Coating Solvents: Volume &
Tropospheric Ozone Impact
Mineral Spirits
Xyiene
VMiP naphta
Toluen*
MEK
Butyl Acetate
Higher aromatic*
Acetone
EB
Butanola
Propunolt
Ethanol
Lacquer solvents
Ethyl acetate
MIBK
Ethylene Glycol
, PM Acetate
1 PM
OB
Methanol
Billions of Pounds
) 05 1 15 2 25 3
' ^
!=3
5=>
i
b | Solvent Used Q Ozone Equivalents
. Top 20 coating solvents account for 92% of usage
. Aromatics generate the most tropospheric ozone
Based on Cartel MIR data & KN&G Solvent study
Impact of Atmospheric Lifetime
on the Global Warming and
Ozone Depleting Potential of Solvents
Major Halocarbon Solvents
Methylene Chloride
Chloroform
Carbon Tetrachloride
1,1,1-Trlchloroethane
Perchloroethylene (PERC)
Other Halocarbon Solvents
(Avg. of 24)
Non-halocarbon (Avg. of 13)
References
Average
Atmospheric
Lifetime,
days
131
200
50 years
6 years
130
229
26
Pounds
CO2 per
Pound
Solvent
0.53
0.37
0.29
0.66
0.53
0.84
2.60
Ozone
Depletion
vs. CFC-11
> 0
> 0
1.1
0.1
>0
>0
none
Global
Warming
Relative
to CO2
28
15
> 2,000
> 360
unknown
>15I
~K5
1 1994 Report of the Scientific Assessment Wonting Group of Intergovernmental Panel on Climate Change
2 Kirk Othmer Encyclopedia of Science and Technology. 4th Ed
'Based on Ihe relative lifetime and CO2 equivalents ot chloroform
Potential Environmental Benefits with
Ethane Cutoff on a Weight Basis
Today
w/ Exempt solvents
. >88% reduction in ozone possible
. Major HAP reductions also achievable
Assumes replacement of lop 20 solvents with noo-HAP solvents with MIRs equal to ethane on wofcht basis
-------
Impact of a Molar Ethane MIR Cutoff on the Type
and Number of Available Exempt Solvents
150 200
Halocarbon
Non-Halocarbon
Q Meet Molar standard
H Meet Weight standard
Solvents Evaluated
148
> Stricter standard would drastically limit the number of
exempt hydrocarbon and oxygenated solvents
. Substitutions would be limited to exempt halocarbons
Based on Cartel MIR data
Wood Coatings
Two-component Urethanes vs Waterborne Lacquer
Ibs/lbs solids
onventional 2K PU
TBAc-Based 2K PU
L30.3
mm.
Waterborne Lacquer
. Ozone Impact of TBAc-based wood coatings lower than water-borne
. HAP content lower than waterbome & conventional
. Ease of use and durability superior to waterbome
Wood Clearcoats
Rohm & Han CL-204 Wood Clear Formulation
Potential Environmental Benefits of
Current EPA Policy
Today
w/ Exempt solvents
. >88% reduction in ozone possible
. Major HAP reductions also achievable
Assumes replacement of top 20 solvents with non-HAP solvents with MIRs equal to ethane on a weight basis
-------
General Industry Concerns With the Process
Remarks by Donna Carvalho
Pennzoil Products Company
Good Morning, my name is Donna Carvalho. I am here representing Pennzoil Products Company.
I am here today to offer an industry perspective on the need for further photochemical reactivity
research. Pennzoil Products Company, through its subsidiary, Magie Brothers Oil Company,
currently has a delisting petition before the Agency. I will not be discussing that petition except as
it relates to why clear science is needed.
Pennzoil applauds this EPA effort to identify research needs and partners. We also welcome the
opportunity to have this forum for policy discussions. As EPA seeks to determine what research is
necessary, we offer the following suggestions:
First, we ask EPA to focus whatever process is adopted for photochemical reactivity decisions to that
purpose only. This process does not need to be a substitute program for new source review,
prever ion of significant deterioration, global warming or the hazardous air pollution program. We
recognize that the final process may impact each of these other programs; however, we believe that
any negative impact will be very small while the positive impact could be significant as state and
federal agencies and industry rightfully focus on controlling those VOC emissions that are most
volatile and reactive.
Second, we suggest that EPA recognize that it does not have to have 100% surety to delist a
chemical. Whatever scientific approach is adopted should combine scientific excellence with
realistic and timely policy making. States inherently recognize the tension between what knowledge
is available and what can and should be controlled when they exempt certain materials from control
requirements. Your counterparts in EPA will be doing the same when they finalize the VOC
consumer product rule. As proposed, this rule exempts materials with a particular vapor pressure,
or where the vapor pressure is not known, with a carbon-number cutoff. These various rules
combine what the regulator knows about the materials being regulated with what can be realistically
and cost-effectively controlled.
Third, as members of industry, we suggest that EPA develop and adopt an easily understood
"cookbook" type of approach to making photochemical reactivity decisions. As discussed in more
detail later, our preference is that this approach would include giving deference to volatility issues.
Other alternatives might include a carbon number cutoff, use constraints, and/or magnitude or
volume of use considerations.
Finally, Pennzoil offers the following specific comments. First, we would prefer that photochemical
assessments be made on a per-gram basis rather than a per-mole basis. Second, we would also
recommend a reactivity scale that looks at ozone formation over a period of time.
I will now discuss each of these suggestions and recommendations in more detail. As noted initially,
the research and final decision-making process that is developed out of these workshops should be
focused solely on defining photochemical reactivity. The final method for determining
photochemical reactivity should be consistent with other programs to the greatest extent possible,
3-54
-------
but should not replace other programs. Trying to address other pollution concerns with a process
for determining whether a chemical is photochemically reactive or not will only make an already
complicated process more complex and time-consuming. There are already numerous other EPA
programs which have been developed to control hazardous and non-hazardous VOC emissions. The
goal of the research coming out of the workshop should be to determine which chemicals participate
in ozone formation and which do not. Once this group has established the process that answers that
question and that question only, then other EPA programs fill the gap to determine what and how
emissions are to be controlled.
Having an overly broad process is a real concern for us. I mentioned earlier that Pennzoil has a
delisting petition before the Agency. This petition is the culmination of nearly 20 years of tests and
research. One of the initial tests done in 1982 indicated that the product which is the subject of the
petition had essentially the same photochemical reactivity as ethane. However, at that time, EPA
expressed to us reluctance to act on the findings because of the uncertainty regarding what Congress
would be doing with hazardous air pollutants within the context of the then proposed Clean Air Act.
That question took another eight years to answer. Recently, EPA has said in its perc delisting
whether the product is hazardous or not is not a factor to be considered when making its VOC
delisting decisions. Even though EPA's current position would not help or hinder our petition, we
think this position is the correct one and should be maintained. Other air pollution concerns should
not play a role in determining if the material is photochemically reactive or not. Rather, the other
EPA programs in place will address other concerns.
As noted earlier, we do not believe that limiting the scope .of the program will negatively impact
these other programs. Instead, controls can be directed to those VOCs which actually merit control.
Pennzoil also believes that the Agency should identify realistic research goals and decision-making
frameworks. This is what combining scientific excellence with realistic and timely policy making
means. The Agency must decide what method or methods will reasonably satisfy it. As part of this
decision-making, EPA should determine which method or methods get the best information it can
have for the most reasonable cost. Pennzoil has tested its materials several different ways. Other
companies who do not have our resources may not be able to make a twenty-year investment in
evaluating their products under changing tests or reactivity scales.
Our preference is that EPA adopt a relatively simple, easy to understand, "cookbook" method. Our
history with the Agency shows that we have tested our materials in various manners and conditions.
^When it became apparent that the Agency preferred one scale over another, we ensured that our
testing results included those scales. However, industry should not have to continue to try and hit
a moving target. One company should not be evaluated under one set assumptions one day and
another company evaluated under different assumptions the next. The Agency's goal today should
be to develop a simple method that can be understood and easily performed.
In this light, Pennzoil would urge the Agency to reconsider looking at volatility as a surrogate for
photochemical reactivity and/or as an initial screening measure. After all, you are determining if
something is a "volatile organic compound". It seems counterintuitive that non-volatile or negligibly
volatile chemicals are nonetheless "volatile organic compounds"
3-55
-------
Unlike reactivity scales which are developed after time-consuming and sometimes costly chamber
tests are performed, volatility is easy to determine. Further, there is already substantial EPA
precedence for using volatility to determine control requirements. As noted earlier, your
counterparts in EPA are expected to issue any day now a consumer products rule that will exempt
products with a volatility of less than 0.1 mm/Hg at 20 degrees C. Where vapor pressure is unknown
the Agency will exempt products with more than 12 carbons. Similarly, most states exempt the
storage and/or use of low volatility products from control requirements in EPA-approved SIPs.
Given the industry familiarity with these types of tests and controls, EPA should determine whether
it can build and/or improve upon a process which uses volatility. Neither EPA nor industry should
have to spend time and effort determining if something is "photochemically reactive" when in some
cases, the emissions will be exempted from control requirements anyway. Unfortunately, even if
there are no control requirements, there are costs associated with identifying and quantifying the low
volatility emissions. Further, the facilities may be unnecessarily paying to obtain and maintain a
Title V permit.
Another option for EPA to consider is using volatility as the screening tool. Other easily understood,
uniform tests (whether chamber tests or modeling) could be required if one cannot pass the screen.
Finally, to the extent that EPA adopts a uniform process where chamber tests are conducted and
results are modeled, we offer the following specific recommendations. First, such tests should
compare the tested material with ethane on a per-gram basis. The basic reason for determining if
a compound is photochemically reactive is to determine if it is a "volatile organic compound" which
must be controlled. Under all the regulatory programs of which we are aware, VOC emissions are
controlled on a weight basis (usually pounds per hour or tons per year). For paints, coatings, and
consumer products, where solvent substitution is often the most effective approach for reducing
VOC emissions, VOC limits are set on a grams per liter basis. Where solvent substitution is used
as a control strategy, the substitutions are made on a volume basis, which is fairly close to a weight
basis and has nothing to do with the molecular weight of the compounds. The VOC control program
and the process for determining if a compound is or is not a VOC should be consistent. EPA has
publicly announced that, in making decisions about VOC exemptions, it will compare compounds
to ethane on a gram-basis. EPA should continue to follow this policy.
Similarly, whatever reactivity scale is finally decided to be appropriate should model ozone
formation over a period of time and not just look to peak occurrences. Again, we suggest this for
consistency with the new ozone control requirements. The systems for determining ozone
attainment now take into account the fact that ozone develops over a period of time and is affected
by weather and other local conditions. When comparing tested materials to ethane, one should look
to see what happens to both materials over time and differing conditions. If the results are basically
the same, the material should be exempted.
Pennzoil hopes that these remarks will give you an industry perspective on these issues. We too
want to see a process developed where all the players will know what is expected to make
enlightened and accurate photochemical reactivity decisions. Thank you for this opportunity to
speak.
3-56
-------
TITLE: A GLOBAL 3-D RADIATIVE-DYNAMICAL-CHEMICAL
MODEL FOR DETERMINING LARGE-SCALE IMPACTS
OF ATMOSPHERIC OZONE PRECURSORS
PRESENTER: Dr. Eduardo P. Olaguer
The Dow Chemical Co.
ABSTRACT
The Dow Chemistry-Climate Model (DOWCCM) is a new 3-D modeling tool that utilizes
sophisticated methods for simulating radiative transfer, photochemistry, and geophysical
fluid dynamics in order to compute large-scale atmospheric impacts of ozone precursors.
These large-scale impacts include global and regional ozone formation potentials, global
warming potentials, ultraviolet actinic flux changes, tropospheric oxidation capacity
changes, and changes in global atmospheric circulation. DOWCCM combines an 11-
wave, spectral meteorological model with a grid, chemical transport model to enable the
simultaneous prediction of the general circulation and chemical composition of the
atmosphere from 0 to 79 km. The current photochemical scheme incorporates up to 136
gas phase and heterogeneous reactions involving about 40 species, including those
pertaining to methane oxidation. Work is now in progress to expand the tropospheric
chemical mechanism to include non-methane hydrocarbons and oxygenated species such
as acetone. The DOWCCM employs a fourth-order, positive-definite Bott scheme similar
to that used in Models-3 to simulate tracer advection. Also incorporated in the model is a
parameterization for convective venting of tracers from the boundary layer to the free
troposphere. The DOWCCM is very computationally efficient (approximately 1 CPU
minute is required per model day on a Cray T90), yet it successfully simulates the basic
features of the general circulation and of total column ozone, and precisely predicts the
atmospheric lifetime of methyl chloroform. Future versions of DOWCCM may contain
nested regional models within a global framework.
4-1
-------
Reactivity Calculations with the
Regional Atmospheric Chemistry Mechanism
William R. Stockwell
Fraunhofer Institute for Atmospheric Environmental Research (IFU), Kreuzeckbahnstr.
19, D-82467 Garmisch-Partenkirchen, Germany, e-mail: stockwel@ifu.fhg.de
The gas-phase chemical mechanism is one of the most important components of an air
quality model. The Regional Acid Deposition Mechanism, version 2 (RADM2)
(Stockwell et al., 1990) is used in a number of air quality models including the
MODELS3/CMAQ modeling system. Many new measurements of mechanism
parameters have become available after the RACM2 mechanism was completed 8 years
ago. We have used these measurements to create a successor to the RADM2 mechanism,
the Regional Atmospheric Chemistry Mechanism (RACM) (Stockwell et al., 1997).
The RACM mechanism has a reasonably complete set of explicit inorganic reactions that
include 21 chemical species. The revisions to the RADM2 inorganic chemistry for the
RACM mechanism were relatively small. The RACM organic chemistry was highly
revised from RADM2. The most important revisions included a reevaluation of the yields
of aldehydes and ketones from alkanes, the yield of HO from the ozonolysis of alkenes,
revised branching ratios for the reactions of acetyl peroxy radicals with NO and NO2, a
revised aromatic oxidation scheme, new oxidation schemes for isoprene and terpenes and
the addition of thd reactions of NO3 radical with organic peroxy radicals. The reactions
of organic peroxy radicals with NOs radical and the revised branching ratios for the
reactions of acetyl peroxy radicals with NO and NO2 lead to predicted PAN
concentrations by RACM that nearly 40% lower than those predicted by RADM2 under
similar conditions. The RADM and RACM mechanisms have been tested against
environmental chamber data and the agreement is good for 03, NOX and hydrocarbons.
We have applied the RACM mechanism to ozone reactivity calculations for biogenic
emissions and highly oxygenated compounds for rural European conditions. We believe
that the new scientific data included in the RACM mechanism make it a better mechanism
for the determination of incremental reactivities than the RADM2 mechanism. The new
RACM mechanism should replace the RADM2 mechanism in any future development of
comprehensive Eulerian air quality models.
References
Stockwell, W.R., F. Kirchner, M. Kuhn, and S. Seefeld, A New Mechanism for
Regional Atmospheric Chemistry Modeling, J. Geophys. Res., 102, 25847-25879,
Stockwell, W.R., P. Middleton, J.S. Chang and X. Tang, The Second Generation
Regional Acid Deposition Model Chemical Mechanism for Regional Air Quality
Modeling, J. Geophys. Res., 95, 16343-16367, 1990.
4-2
-------
-fx
OJ
Requirements for Chemical
Mechanisms for
Eulerian 3-D Regional
Atmospheric Chemistry Models
1. Predict concentrations of
H2O2, ROOM, PAN, HNO3, H2SO4..
2. Mechanism must give accurate
predictions over chemical
concentrations ranging from
clean to moderately polluted.
Peroxy radical reactions are
important
3. Mechanisms must be valid for
multiday simulations.
Nighttime chemical species
such as NOa are important.
Less reactive species which
are subject to long range
transport are important.
The Regional Atmospheric
Chemistry Mechanism (RACM)
A completely revised version of the RADM2
mechanism of Stockwell et al. [1990]
Mechanism Includes:
237 reactions
17 stable inorganic species
4 inorganic intermediates
32 stable organic species
(4 are primarily of biogenic origin)
24 organic Intermediates
William R Stockwell
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunholer Institute for Atmosphenc Environmental Research (IFU)
William R Stockwell
Reactivity Calculations wuh the Regional Atmosphenc Chemistry Mechanism
Fraunhofer Institute for Atmosphenc Environmental Research I IFU)
-------
The Regional Atmospheric
Chemistry Mechanism
1996
Includes:
Detailed and explicit inorganic
chemistry
Lumped organic chemistry
Type
Number
Alkanes
Alkenes (including biogenics)
Aromatics
Carbonyls
Organic Peroxides
Organic Acids
Organic Nitrate and PANs
5
7
3
9
3
2
3
William R. Slockwell
Reacliviiy Calculations with
-------
Statewide Air Pollution Research Center
Smog Chamber Experiments for
Testing of RADM Mechanism
Description
Propane
n-Butane
Ethene
Acetaldehyde
Toluene
Toluene + n-Butane
m-Xylene
Multi-Component
Identification Numbers
EC216
EC178, EC305, EC306
EC142, EC143
EC254
EC340
EC331
EC344, EC345
EC231, EC232, EC233,
EC237, EC238, EC241,
EC242, EC243, EC245,
EC246
Statewide Air Pollution Research Center
Multi-Component Smog Chamber
Experiments
Components
NOX
n-Butane
2,3 Dimethylbutane
Ethene
Propene
t-2-Butene
Toluene
m-Xylene
HCHO
CO
William R. Stockwell
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofer Institute for Atmospheric Environmental Research (IFU|
William R Stockwell
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofer Institute for Atmospheric Environmental Research lIFU)
-------
NO.
SAPRC environmental chamber
experiment EC-237
o\
o.
O.
2
**
c
Ol
u
c
o
U
0
0
100 200
Time, Min
300
400
O Experimental values of NO2
A Experimental values of NO
RACM simulations; dotted lines
RADM2 simulations
William R. Stockwell
Reaciivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofer Institute for Atmospheric Environmental Research (IFU)
Ozone
SAPRC environmental chamber
experiment EC-237
E
o.
o.
C
o»
u
o
U
0
0
400
O Experimental values of 03
RACM simulations; dotted lines
RADM2 simulations
William R. Stockwell
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofer Insmuie for Almosphenc Environmental Research (ffTJ)
-------
Comparison of Simulation and Smog-Chamber Run for Isoprene
Comparison of Simulation and Smog-Chamber Run for d-Limonene
I °H
*
S o.4H
i
OJ (expcmncMil dau)
03 (KADM2)
O3 (»cw mcduniun)
03 (eipcrimonl dm)
03 (RADM2)
O3 (a«w mcrtartm)
; «!>
s §
i §.,_..> a § §
a»-
5 04-
?
Q O3 (u^rimal dju)
O3 (RADMJ)
O3 (ne" meditniim)
0 8 § i,miB) § |
William R~ Siockwell
1 § § I §
l(min)
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofer Institute for Aimosphenc Environmental Research (CFU)
i(min)
the experimental data are given by the following symbols: squares represent
ozone, circles NO and triangles the sume of NO2 and nitrates
the lines represent the simulation results for the corresponding species but
the NO2 line represents only NO2 (without nitrates)
WiUiam R. Siockwell
Reacuvity CoJculauon.s with the Regional Atmospheric Cherrustry Mechanism
Fraunhofer Institute for Atmospheric Environmental Research (IFTJi
-------
Comparison of Simulation and Smog-Chamber Run for a-Pinene
the experimental data are given by the following symbols: squares represent
ozone, circles NO and triangles (he sume of NO2 and nitrates
the lines represent the simulation results for the corresponding species bui
the NO2 line represents only NO2 (without nitrates)
Maximum ozone concentrations
predicted by RACM and RADM2
mechanisms plotted against
SAPRC experimental values.
* RADM2
O RACM
0.0
William R. Slockwell
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Experimental Maximum Ozone, ppm
William R. Slockwell
Reactivity Calculations with ihe Regional Atmospheric Chemistry Mechanism
Fraunhofer Institute for Atmospheric Environmental Research I [FU)
Reactivity Calculations with the Regional Alnmphenc CnemiMry Mechanism
Fraunholer Institute for Atmospheric Environmental Research lll-lh
-------
Time of the maximum of the ozone
concentrations predicted by
RACM and RADM2 mechanisms
plotted against
SAPRC experimental values.
600
E
s
E
I
H
"8
I
ea
"a
E
500-
400-
300-
200-
100-
0
X RADM2
O RACM
0 100 200 300 400 500 600
Experimental Time of Ozone Maximum, Min
Maximum NO2 concentrations
predicted by RACM and RADM2
mechanisms plotted against
SAPRC experimental values.
X RADM2
O RACM
0.0 0.2 0.4 0.6 0.8 1.0
Experimental Maximum NO2, ppm
William R. Stockwell
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofer Institute for Atmospheric Environmental Research I [FU)
William R Stockwell
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofer Institute for Atmospheric Environmental Research (IFU)
-------
CM
si 111
3^B CO "
_ Q.CM 0)0>
*X C Q TO 'Z
(0 (O ^ Q)
E
'u
a»
a
x
jw 'uinuiixei\
s
II
II
II
y Q.
Is
1<
I Is
" C-5
* >-ji
i 11
I s =
> QCU.
ff I
E = E
= E. P-
I ^ S
3 s.
^i
l!
Effect of Revisions on
Ozone and PAN
40
234
Time, days
iiiir
234
Time, days
4-10
-------
njc <
III
7^ JO
In v,
c
IK *
is =
??
f I
Fate of RCO3 Radicals
in RACM
0.8
£0.6-
Q.
0)
15 0.4 ^
cc
c
«0.2-
0.0-
00
30 £
>»
o
0)
cr
20 2
-10
00 6:00 12:00 18:00 24:00
Time, Day 5 of Simulation
w
_>.
o
CM
O
- J(N02)
RCO3 + RO2
RCO3 + CH3O2
RCO3 + HO2
2 x (RCO3 + RCO3)
E3 RCO3 + NO3
E3 RCO3*NO
[Villenave etal., submitted]
T3 4)
C £
O
CM
X
c
> (0
o
0)
: E
I
cc
ol
« (qdd)
HOVb
o<
§°-
U o>
II
i_ 0)
£8
>.*
X «
(0
£o
II
a
a
o z
= 3
o X
10 CM
0)
8'
000
o
o
in
o o o
ooo
n CM »
o n
o r-i
o
o
(qdd) Eo
(qdd)
HOVH
a:
E
ft
t
~c
1 ss
S =5
US.
4-11
-------
CM c mo>
5§0£
O eZ~
<£xc
^o
,,2cc
;-oC>
; « w o**
£ W t
(0
JR e O C fl>
O E o « w
0)
Q.
O
0)
0)
'o
0)
Q.
co CD r- co CD
co o in o> co
O> O> O) O> O>
o o o d d
O) «3- o> in o
1- t*- O "T-
CD CNI O O O
d o d d
oo ^ co T- i-
co o ^t t !
!"
STS
ft
I!
p*
1?
l§.
i!
"8
Initial Conditions for
Incremental Reactivity Calculations
Start Time (Local Hour) 3:00
End Time (Local Hour) 22:00
Temperature (K) 298.15
Pressure (mbar) 1013.25
Photolysis Frequencies July 01, Latitude 45*
Initial Concentrations
Species (ppb)
O3 50
2.0
0.2
0.5
0.1
1.0
200
1700
500
H202
NO
NO2
HN03
HCHO
CO
CH4
H2
H20
02
N2
1.0
20.9
78.1
Emission Rates
Species
NOX
SO2
CO
Ethane
Low Reactive Alkane
Medium Reactive Alkane
Highly Reactive Alkane
Ethene
Internal Alkene
Terminal Alkene
Toluene
Xylene
HCHO
Aldehyde
Ketone
Emissions
(ppb mln'1)
Varied
5.18 x1(H
5.65 x 10-3
2.41 x 10-4
2.94 x 10-3
7.70 x 10-4
4.52 x 1(H
4.56 x 10-»
1.88x10-*
2.19 x
5.72 x
5.19 x1(H
1.39X10-4
3.62 X10-5
5.02 x1(H
4-12
-------
-
if |
II
S?
|
~
80
75
70
n
O
« 65
60
55
RACM Mechanism
NOX Variation for PLUME/2 Case, Day 1
EBIR
MOIR
\
MIR
5 10 15
Initial + Emitted NOX (ppb)
7.5
O
5.0
0.0
20
a
a
t.
o
2.5 O
a
a.
u
CO
Max O3 (ppb)
IR NOX
IR VOC * 10
IR VOC
July 01 Latitude 45
Start Time = 3:00
Stop Time = 22:00
Total Time = 19 h
-2.5 S
0)
£
u
-5.0 £
^ ^
?
1 1-
|l
*%
c"' =>
= =
5 ^
ll
n
d
t/5
8
W
s
r>
~"
0
o
>
0
Q.
Q.
r>
O
jQ
a
a
i
5n
i
RACM Mechanism
Incremental Reactivities for PLUME/2 Case
1.5-
1.0-
0.5-
On
-
I
O
^^
I
o
r-
W/S/////A
i
9 9
C C
|
0
^
i
1
a a>
c c
- > e.
x < i2 < -g x §
~ M
D MIR
H MOIR
~l(]ftmr>r>n.n.
0)00)50)0)0)0) O CM O
0) 0) Q) «S (Q CO O CQ CO JA ^"^
C 3 C Cat ^^ ^^ *^ ^^ «C
9"5E <<5
cc
Species
o
o
O)
X
4-13
-------
Maximum Incremental
Reactivity (MIR) vs kHO
10
DC
0.1-
HCHO
Terminal
Alkene
Internal
Alkene
Ethenex.
__^_«
Aldehyde /£> \ Isoprene
Toluene 0 / Xylene « K
Low
Reactive
Alkane \
/
Ketone
Ethane
CO
.SO
d-Limonene
cc-Pinene
*Middle Reactive Alkane
*\
Highly Reactive Alkane
0.01
1.0E-13
1.0E-12 1.0E-11 1.0E-10
(cm^molecule"1s"1)
1.0E-09
Conclusions
We strongly recommend the
RACM mechanism for use in
atmospheric chemistry models
over the RADM2 because the
RACM chemistry is based upon
more recent and reliable data.
Compared with the RADM2 the
RACM predicts concentrations:
- somewhat greater for 03,
H2SO4 and HNO3
- almost the same H2O2
- significantly less PAN.
The extended RACM has been
applied to ozone incremental
reactivity calculations for rural
European conditions.
We have estimated incremental
reactivities of isoprene, terpenes
and dimethoxymethane.
William R. Siockwell
Reactivity Calculations wilh ihe Regional Atmospheric Chemistry Mechanis
Fraunhofer Institute for Atmospheric Environmental Research (IFU)
William R. Stockwell
Reactivity Calculations with the Regional Atmospheric Chemistry Mechanism
Fraunhofet Institute for Atmospheric Environmental Research (UFU)
-------
Hydrocarbon reactivity and ozone production in urban pollution according to the
Stockwell et al., (1990) reaction mechanism
Chris J. Walcek
State University of New York at Albany
Abstract- A method for ranking the ozone production potential for various classes of reactive
hydrocarbons is presented. Using the Stockwell et al.,(1990) chemical reaction mechanism
("RADM2"), ozone production efficiencies for 14 classes of emitted hydrocarbons included in
the mechanism are quantified over a wide range of background NOX and hydrocarbon
concentration regimes. 63 production efficiencies are calculated by running a box model
initialized with specified concentrations for a 2-day period under fixed sunlight and
meteorological conditions, after which NOX is oxidized and ozone production ceases. Individual
organic compound concentrations are then perturbed by 1 ppb, and the resulting changes in 63
after 2 days are compared with the base simulation. For some ranges of NOX and organic
compound concentrations, the additional ppb of 03 produced from each additional ppb of organic
compound is somewhat constant, but there are some compounds under some chemical conditions
for which the additional ozone production potential is highly variable. Despite these variations,
net ozone production from each class of organics is approximately correlated with the
corresponding reactivity of the organic compound with HO radical, although there are some
broad violations of this correlation. HO reactivity may only crudely be an indicator of ozone
production potential under many conditions for some classes of organic compounds.
4-15
-------
o\
Hydrocarbon reactivity and ozone production in
urban pollution according to the
Stockwell et al., (1990) reaction mechanism
Chris Walcek
Slate University o( New York at Albany
1. Simple method for calculating
"ozone productionefficiency"
2. Detailed description of a couple "cases"
3. Generalized results for wide range of
NOx and hydrocarbon concentrations
4. Conclusions
Ozone concentration change in
air parcel, calculated using
Stockwelletal., (1990)
mechanism
33.0
11.40 1145 11:50 11:55 1200 12:05 12:10
Local 11m* (hrmln)
noon conditions,
50% clear-sky photolysis rates
20 'C, 50% Rh; 1.8 ppb isoprene
-------
Ozone formation rates vs
concentrations of NOx and
hydrocarbon concentrations
Ozone concentration (ppb) vs time
0.1 1 10 100 1000 10000
non mathana, r»actlv* organic concentration (ppb C)
noon conditions
April, 40'north
50% clear-sky photolysis rates
20'C
50% Rh
1.8 ppb isoprene
a
Q.
a.
c
o
Initial N0x= 10 ppb
NMHC=100ppbC
Base Case
+ 1 ppb Acetaldehyde
Initial NOx= 1 ppb
NMHC= 10ppbC
time (hours)
-------
oo
NOx and Organic concentration vs. time
Q.
a.
c
o
u
c
o
u
OJ
O
f
O
01
jQ
a.
a.
S i
§
o
0)
I
ai
2
-------
(qdd) uojjejjuaouoD auozo
Perturbation ozone concentration induced by 1
ppb increase in various hydrocarbon
'. Initial NOx=10ppb; Initial NMHC-100ppbC
Carbonyls
Alkanes
ETH
HC3
HC5
HC8
Alkenes
Aromatics
TOL
XYL
CSL
-------
Aromatics Change in accumulated ozone (ppb) after
two days due to change in initial specified
organic by 1 ppb
Toluene
(& less reactive)
TOL
001 -
0.1 1 10 100 1000 10000
non methane, reactive organic concentration (ppb C)
>' 10OO
0.01
0.1 1 10 1OO 10OO 10000
non methane, reecllve organk concentretlon (ppb C)
1000
$ 100
I
Cresol | 10
(& other Z
hydroxy-substituted) g ,
CSL
o.oi
0.1 1 10 100 1000 10000
non methane, reactive organic concentration (ppb C)
Olefins (Alkenes)
Change in accumulated ozone (ppb) after
two days due to change in initial specified
organic by 1 ppb
Ethene OL2
Terminal alkenes OLT
01 1 10 100 1000 10000 0.1 1 10 100 IOOO 10000
non methane, reactive organic concentration (ppb C) non methane, reactive organic concentretlon (ppb C)
Internal alkenes OLI
Isoprene ISO
0.1 1 10 100 1000 10000
non methane, reactive organic concentration (ppb C)
_!_//£__
0.1 1 10 100 1000 100OO
non methane, reactive organic concentration (ppb C)
-------
Carbobvls
/aldehydes)
v ueiiyue*/
Change in Accumulated ozone (ppb) after
two days due to Cnan9e in initial specified
organic by 1 ppb
1000
Formaldehyde
HCHO
0.01
0.1 1 10 tOO 1000 10000
non methane, reactive organic concentration (ppb C)
100
_^ Acetaldehyde | 10
' (& higher aid) |
I- ALD I ,
o.i
0.1 1 10 1OO 1000 10000
non methane, reactive orgenic concentration (ppb C)
Ketones
KET
1 10 100 1000 10000
non methane, reactive organic concentration {ppb C)
Alkanes
Change in accumulated ozone (ppb) after
two days due to change in initial specified
organic by 1 ppb
Ethane ETH
Slow reactive HC3
10 100 1000 10000
non methane, reecllve orgenic concentration (ppb C)
Intermediate reactive HC5
_J
0.1 1 10 100 1000 10000
non methane, reactive organic concentration {ppb C)
Fast reactive HC8
o.oi
0.1 1 10 100 10OO 10000 0.1 1 10 100 1000 10000
non methane, reactlva organic concentration (ppb C) non methane, reactive organic concentration (ppb C)
-------
Change in Accumulated ozone (ppb) after
two days due to 1 ppb change in initial NOx
Ozone "production efficiency" from hydrocarbons
in the Stockwell et al., (1990) mechanism
initial conditions: 10 ppb NOx, 100 ppb(C) organic
after 48 hours, continuous noon conditions
NOx
(NO + NO,)
to
S)
0.01
0.1 1 10 100 100O 10000
non m«lh«n«, rucllv* orginlc concentration (ppb C)
y - 2076.7 * 10*(0.94024x) R*2 - 0.7«S
-T~-_
1 ISO
v-r-v?;-'.-Tway
-05
00 0.5 1.0 1.S 2.0
ppbO3 produced p«r MtdNlonil ppb orginlc «dd*d
-------
Correlation coefficient (r) between
(a) change in accumulated ozone (ppb)
due to 1 ppb change in all organics
vs.
(b) log (rate coefficient for reaction of
each organic with HO)
to
0.01
0.1 1 10 100 1000 10OOO
non m»th«n«, rwctlv* organic concentration (ppb C)
Joint probability distribution:
NO» and nonmethane hydrocarbon concentrations
% probability of observing concentration per 1/3 logio
concentration range
900 - 950 mb. Northeast U. S.. 21-24 April 1981
(RADM 35x38 domain)
1000
i 10 100 1000
Non mathan* organlca (ppb)
10000
-------
MULTI-DAY OZONE FROM
LOW-REACTIVITY VOC's
by
Gary Z. Whitten
Systems Applications International
Overview
I Concern for downwind areas
I New technique based on UAM
I Compares candidate VOC to ethane
I Provides incremental impacts
4-24
-------
UAM-BASED TECHNIQUE
VOC increment added to upwind cells
Uses "back" trajectory from final peak cell
100 tons ethane for base case
Spread over 9 cells over 1 hour
Use explicit chemistry for ethane and
candidate VOC
Example Candidate VOC
I l-bromopropane
I Same k^ by weight as ethane
I Molecular koH 4 times faster
I Chemistry assumed to be like propane
I Molecular k^ same as propane
4-25
-------
UAM Results
I 1st day both nearly 14 ppb impact
I Not at main peak (cloud not there yet)
I 2nd day impacts differ
I ethane 4 ppb (on peak of 190)
I candidate only 2.3 ppb
I 2nd (off peak impacts)
I ethane 7.8 ppb
I candidate 4.8 ppb
LEVEL I Ozon* (ppb)
Time 4-2
-------
LEVEL I Oioni (ppb)
Time 0-2400 August 27. 1987
UAXIUl'M 1.1 9 pj.h
UINIUUU « 04 ppl.
275 325 375 425 475 525 575
401 i i i i i i i i | | ii i I I I i I I | I i I I I I I I i | i i i i i i i i i | i i i i i i i I I | I I i I M II i| i l-r'1
7JO
3670
Difference in Maximum Simulated Ozone Concentrations
August 27, 1967 (Simulation 100 ton ethane minus base)
LEVEL 1 Ozone (ppb)
Time .0-2-100 August 27 1987
75 325 375
I I I I I I I I I I I I I I I I I I I I I
30
20
10
* MAXIUl M . 139 pph
MINIMUM -0 fl pph
425
475 525 575 _,D_
i I i i i i i i i i i [ i i i i i i i i i I i i i i j 38 i I)
3770
3720
10
20
30
40
50
60
3670
Difference in Maximum Simulated Ozone Concentrations
August 27. 1987 (Simulation 100 ton bromopropane minus base)
4-27
-------
LEVEL I Otone (ppb)
Tim. 0-2400 AugUJl 28. 1967
MAXIMUM 7 6 ppli
MINIMUM -0 J pph
575
r r 3H70
60
Difference in Maximum Simulated Ozone Concentrations
August ZB. 1967 (Simulation 100 ton ethane minus base)
LEVEL 1 Ozon* (ppb)
Tim^ 0-2JOO August 28 1987
.275 325 375
JUi i ' ' i i i i i i I i ' i i i i i i i i i i
20
10
* MAXIMUM < 8 ppb
MINIMUM -0 I pph
425 475
I I | I I I I 1 I I I I | T I
525
575
Q1 ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' I I 1 I ' I I I I I ' I Nil
10
30
30
40
50
Difference in Maximum Simulated Ozone Concentrations
August 26. 1987 (Simulation 100 ton bromopropane minus base)
< 38 .'I I
\
60
3770
3720
3670
4-28
-------
Abstract
Computing Volatile Organic Compound Reactivities with a 3-D AQM
Zion Wang
University of North Carolina
at Chapel Hill
In many urban areas, selective VOC control on reactive VOCs is much more advantageous over
indiscriminate control. This raises the need for identifying reactive VOCs. One of the currently
used methods to quantify VOC reactivity is by measuring how changes in VOC emissions in an
airshed affect ozone formation in the same airshed with the EKMA modeling method. However,
due to simplifications in the dispersion component of the model and in the ambient conditions and
emissions inputs, the use of three-dimensional photochemical models to obtain reactivity data is
desirable. This study attempts to use a three-dimensional photochemical model to compute the
reactivity data for a few VOC species. The study also examines how different parameterization
techniques impact reactivity calculations.
4-29
-------
Box Modelling of NOx and VOCs to Determine Emission Reduction Strategies
PA. Makar
Atmospheric Environment Service
Environment Canada, 4905 Dufferin Street
Downsview, Ontario, Canada, M3H 5T4
paul.makar@ec.gc.ca
T.Dann
Environment Canada, River Road Lab, Ottawa
D. Albin
MYDA Consulting
May 25, 1998
Abstract:
Measurement data from the Canadian National Air Pollution Surveillance
monitoring network was used to provide initial conditions for a series of sensitivity runs
of a box model of local chemistry. The sensitivity runs were used to determine the
factors having the greatest impact on ozone concentrations at 15 sites in urban centres in
Canada. Sensitivities to NOX were either zero or negative, indicating that the
measurement sites were likely VOC limited and subject to NOX titration of ozone.
Decreases in model Nox at these sites led to increases in ozone production by the model.
Sensitivities to total VOC and ten major unoxygenated VOC classes were positive,
indicating that reductions in VOCS would result in decreases in ozone concentrations.
Specific VOC classes had a much greater effect than others, with internal-bond alkenes
and higher aromatics having the greatest impact on ozone production.
Model Description:
The photochemical model employed had three main components:
(1) A gas-phase chemical mechanism (Makar et al, 1998) used as input for the model
chemical calculations. The mechanism employed has been under development for
several years at AES, and is intended as a replacement for the regional model
mechanism of ADOM. The major revisions to the species of the previous mechanism
include:
the inclusion of three additional PAN-like species
the inclusion of CO and C2H6 as advected, non-constant variables of the system
the separation of higher alkanes into C4-C5 and C6-C8 species
the separation of higher alkenes into terminal and internal double bond species
the separation of higher aromatics into di and tri substituted species
the inclusion of species-specific RO2s and R(O)O2s
the inclusion of six (previously one) higher carbonyls
the inclusion of formic acid, acetic acid, and C1-C3 alcohols
the inclusion of HNO4 as an advected species.
The new mechanism has 251 reactions, compared to the ADOM mechanism's
114. Some of the more important revisions include:
4-30
-------
All reactions retained from the ADOM model have rate constants updated
according to the most recent data for those reactions.
Inclusion of specific RO2+RO2 reactions and their products
NO3 + RO2, NO3 + R(O)O2 reactions included
Much greater detail in the formation of oxygenated products from the original
nonoxygenated VOCs. Aside from the greater speciation, the formation of
products of higher alkanes, alkenes, and aromatics has been altered to incorporate
recent laboratory studies on these species. The aromatic mechanism now includes
the highly reactive broken ring products and dicarbonyls known to form after the
initial oxidation by OH, O3 and NO3.
A more detailed isoprene mechanism, including the formation of MACR and
MVK and MPAN has been included.
Higher Terpenes have been included as a separate species.
All photolysis rates have been updated according to the most recent information
available.
(2) A one-dimensional radiative transfer subroutine. This model (Yung, 1976) was
used to calculate the intensities of light as a function of wavelength at each site for
which the model was applied. The radiative transfer model made use of the US
Standard Atmosphere (1976) for ozone and total column number density information.
Solar zenith angles were calculated using the latitude, longitude, local time and time
zone of each site for which calculations were performed. The resulting solar
intensities were used to calculate photolysis rates for the chemical model.
(3) A numerical solver to solve the system of differential equations resulting from the
chemical mechanism. The solver used here was that of Kahaner et al (1989), a
variation on the predictor-corrector code of Gear (1971). The same solver was set up
by the author of the current work for use as the numerical driver in the AES
CREAMS box model.
As input data, the model made use of the NAPS database. The National Air
Pollution Surveillance database includes time coincident NOX and VOC measurements
made at several sites across Canada, from 1986 until the present. A subset of 15 sites was
used in the current study. VOC measurements include standard testing for 175 different
species. These were lumped into the model speciation using the reactivity weighting
method of Middleton et al, (1990).
Unfortunately, until 1996, the NOX data was usually only reported to the nearest
10 ppbv. In addition, the records contain only a total NOX expressed as NOi. The lack
of more detailed NOX data presents an important confounding factor to the conclusions of
the study, as is noted below.
4-31
-------
Methodology:
A model run of an individual record would take place through the following
stages:
(1) A measurement record, consisting of the site latitude, longitude, time
(year,month,day, start hour, end hour, time zone), NO2, NO, NOX, O3, CO, SO2, and
VOCs was read in by the model.
(2) Most of the NAPS records did not have NO reported as a separate variable, and the
NOx and NO2 concentrations were usually equal. Initial concentrations of NO2 and
NO were calculated using the following stages:
(a) If NOx was reported, and NOx=NO2 in the record, then it was assumed that
the NO concentrations were less than the 5 ppbv, and the model mechanism
was used to generate an initial NO concentration by assuming a steady state
with the other model variables. The same process was used if only NO, was
reported (a smaller number of stations reported NOX, and zero NO2 and NO).
(b) If both NO2 and NO values were reported, those values were used as initial
conditions.
(3) Following initialization of NOX, the model was run forward in time for one hour to
initialize the other variables. If the site was located in a city, the concentrations of
NO, NOa, CO, SOa and the unoxygenated VOCs were held constant during this
initialization, simulating the replenishment of these variables by emissions. If the site
was considered to be rural, then all variables aside from methane, water vapour,
oxygen and the tqtal number density were allowed to vary with time. Table 1 gives a
listing of the stations, and their categorization as rural or urban. This procedure
allowed the generation of initial concentrations for the other model species (eg.
oxygenated hydrocarbons, organic peroxides, hydrogen peroxides, radicals, etc.).
These were used as initial values these species in the sensitivity runs which followed.
The idea here was to allow the chemical model to "spin up" slightly, to avoid the
sensitivity calculations being affected by the initial conditions for the unmeasured
species.
(4) Twenty-five sensitivity runs were then performed on each record. For urban runs, the
concentrations of NO and the ten unoxygenated VOCs {ethane (C2H6), propane
(C3H8), C4-5 alkanes (C4AK), C6-8 alkanes (C7AK), ethene (C2H4), terminal bond
alkenes (as propene; PRPE), internally bonded alkenes (as trans-2-butene; BUTE),
toluene (TOLU), di-substituted aromatics (DARO) and tri-substituted aromatics
(TARO)} were held constant at the measured (or calculated, as was often the case for
NO) values, once again in an attempt to mimic the emissions dominated regions. For
rural runs, all species were allowed to vary with time, the above species having their
initial concentrations taken from the measurements.
The twenty-five runs consisted of
A Base run: 1 hour integration as described above.
Two NOx runs: NO increased and decreased by 25%, relative to the base
run.
4-32
-------
Two Total VOC runs: the hydrocarbons listed above all increased or
decreased by 25%, relative to the base run.
Two individual VOC runs for each of the ten VOCs listed above, with
increases and decreases of 25%, relative to the Base run.
(5) The output data from the runs were used to calculate sensitivities of ozone
concentrations to the given perturbation by fitting the three data points (+25%, base, -
25%) values to a parabola, then calculating the resulting derivative of ozone with
respect to percent change in the particular variable).
The concentrations of biogenic hydrocarbons were not affected by the sensitivity
runs; isoprene and alpha-pinene were treated like the other unoxygenated VOCs, but no
sensitivities were calculated.
The resulting output was a list of sensitivities of ozone towards each of the
perturbed variables, in units of change of ozone concentration per percent change in the
parameter from its measured concentration (AO3 / A variable; units (ppbv/%)). Positive
sensitivities indicate that a decrease in the variable will result in a decrease in ambient
ozone concentrations. Negative sensitivities indicate that a decrease in the variable will
result in an increase in ozone concentrations.
Results:
The records were analyzed in two groups: cases for which ozone concentrations
were greater than 70 ppbv (G70) and cases for which ozone concentrations were less than
50 ppbv (L50). Here, only the summer (July and August) O3 > 70 ppbv cases will be
examined in detail.
The results for these records are shown in Figures 1 to 7. Figure 1 shows the
sensitivity of model ozone towards changes in NO. Regional differences are apparent in
these stations. Stations in Windsor, Toronto, Hamilton and Sarnia all show that decreases
in local NO concentrations would lead to increases in ozone concentrations. This
probably indicates the effects of local ozone titration; if the NO concentration was
decreased, then ozone concentrations close to the emissions sources would increase due
to a reduction in importance of the NO + Os removal pathway. In Simcoe and
Stouffville, this effect is less apparent. In the three west coast stations (Coquitlam,
Surrey, Langley), changes in the NO concentration had little effect on the ozone
concentration.
Figure 2 shows the effect of changes in total VOCs. For all sites, sensitivities
are positive; VOC reductions result in ozone decreases. The greatest reductions as a
function of percent change in the local VOC concentration are for Coquitlam, Sarnia and
Windsor (note: the Coquitlam value is the result of only three records, and must be
considered less statistically significant than the other cases). Figure 2 shows that a 25%
reduction in total VOCs would result in a reduction in ozone concentrations during
episodes of about 15 ppbv (0.6 ppbv/% x 25%).
Figures 3 and 4 show die sensitivities towards the two VOCs which had the
greatest impact on ozone concentrations out of the ten for which sensitivities were
calculated. They show the effects of targeting particular VOCs for emissions reduction.
4-33
-------
Figure 3 shows the sensitivity to BUTE; the internally bonded alkenes. Urban
Ontario values are between 0.05 and 0.25 ppbv/%. Simcoe, Stouffville, Surrey and
Langley have virtually no change. Coquitlam values are very sensitive to changes in
BUTE.
Figure 4 shows Tri-subsituted aromatics, with sensitivities of about 0.03 ppbv/%
being typical.
The remaining species have progressively smaller effects on the ozone
concentration. In decreasing order of importance, they are: Toluene, Di-substituted
aromatics, C6-8 Alkanes, terminal-bond alkenes, ethene, C4-5 Alkanes, Propane and
Ethane.
These results suggest that targeted reductions of internal-bond alkenes, followed
by higher aromatics would have the biggest effect on local ozone concentrations in most
of southern Ontario. Certain VOCs (those with both high reactivity and high
concentrations) result in much of the ROi formation leading to ozone production.
Figures 5 - 7 show the sensitivities grouped according to region and plotted
rel uive to each other. The greatest impacts of VOC controls are seen in the urban regions
of Ontario, Sarnia and Windsor in particular, followed by Toronto. Impacts of reductions
are smaller in Simcoe, Stouffville and the Vancouver sites other than Coquitlam. The
latter has a high sensitivity to BUTE as in the Ontario sites, but this may be due to small
sample size (three records).
In addition to the summer ozone episode cases, sensitivities were also calculated
for spring episodes ^April to May). The relative results were similar to the summer
episodes for most stations, but the magnitudes of the sensitivities were smaller. For
example, the Junction Triangle station's sensitivity to total VOCs was about 0.075
ppbv/% in the spring versus 0.25 ppbv/% in the summer. NO sensitivities at the same
site were also lower in spring versus summer, -0.25 ppbv/% versus -0.4 ppbv/%. The
same pattern of sensitivities for individual VOCs was noted as for the summer cases;
sensitivities tend to be highest for internally bonded alkenes, with aromatics following in
importance.
The spring data included a single record from Edmonton. Although statistically
insignificant, it is interesting to note that the NO sensitivity was still negative, and that
the pattern of VOC sensitivities has changed. Di-substituted aromatics have the greatest
impact on ozone concentrations, followed by toluene, propene and the C6-C8 alkanes.
Further episode measurements would be required to determined whether this reflects true
regional differences in the ozone production due to hydrocarbons.
A large number of records with ozone less than 50 ppbv were examined to see if
the sensitivities of ozone production due to NOX or VOCs differed between ozone
episodes and non-episode scenarios. The sites show the same pattern as for the summer
cases, with negative NO sensitivities, positive VOC sensitivities, with butene followed by
the higher aromatics leading the VOC sensitivity magnitudes. The magnitudes are
smaller than for ozone (>70 ppbv) episodes (eg. Windsor VOC sensitivity 0.25 versus 0.5
ppbv/%); VOC controls will have a smaller impact on non-episode situations than during
episodes. Figure 22 shows the sensitivities for the Rocky Point site at Coquitlam, BC,
4-34
-------
with negative NO sensitivities and positive VOC sensitivities, internally bonded alkenes
(BUTE) being the most important of the latter.
Discussion:
The results from these tests suggest the following:
(1) At the sites for which ozone episode data was available, decreasing local NO
concentrations would lead to increases in ozone. Provided that this was not a
result of model setup (see below), this would suggest that NOX reduction
strategies in the vicinity of urban areas would actually lead to local increases in
the ozone concentrations.
(2) At all sites, reductions in the total VOC loading resulted in immediate local
ozone decreases. These reductions in ozone concentrations were greatest for a
total, across the board VOC cut, but more detailed work showed that specific
VOCs (in order of precedence: internally bonded alkenes, tri-substituted
aromatics, di-substituted aromatics, and toluene) probably make up most of the
VOC effects. These VOCs are both sufficiently reactive and have sufficiently
high concentrations to have a significant local impact on ozone production.
(3) The direction (if not the magnitude) of the sensitivities was the same for cases
in which the ozone concentration was less than 50 ppbv. This would suggest
that a strategy in which different components of the reactive mix are targeted
for reductions at different times seems unnecessary. The same VOC reduction
strategy m#y be used regardless of whether an ozone episode is taking place;
VOC reductions during low ozone days will not have adverse effects.
Two confounding factors should be noted at this point.
NOX concentrations, and NOX sensitivities. As was mentioned above, the
NO2 and NO concentrations in the NAPS database were usually reported only to
the nearest 10 ppbv, and usually total NO* was reported as NO2. The strategy of
determining NO concentrations from the use of steady-state may have led to errors
in the sensitivities in two ways; through lack of accuracy in the original
measurements, and through the use of steady-state to generate NO concentrations,
followed by perturbations from that steady-state to generate sensitivities.
Assumption of local emissions via constant NO and VOCs. Another
source of uncertainty (for the urban sites) is the assumption of a local emission
source (ie. the concentrations of NO and unoxygenated VOCs being held
constant; the level being changed for sensitivity calculations). The effect of these
uncertainties can only be resolved with more accurate measurement data and
further model runs, as discussed below.
Conclusions and Plans for Future Work:
The work performed here indicates that the NAPS sites studied have positive
sensitivities of ozone production with respect to VOCs, and negative sensitivities with
respect to NOX. NOX reductions at the given sites could lead to increases in local ozone.
4-35
-------
VOC reductions will reduce local ozone, and specific VOCs (internally bonded alkenes,
followed by aromatic compounds) can be targeted as having the greatest local impact on
ozone production.
It should be noted that significant uncertainties exist in the NOX conclusions, due
to the limitations on the measurement accuracy and the assumptions in the use of the
measurements. The following steps are recommended for future work to resolve these
uncertainties:
(1) The use of more accurate NOX data (ie. both NO and NO2 resolved, with ppbv or
better accuracy) from measurement intensives (eg. Pacific 93, NARSTO). These
can be used to test the effect of the NAPS data being reported in lOppbv intervals;
the intensive data can be degraded in the same fashion as the NAPS data, and the
resulting model output compared to that resulting from the true NOX initial
conditions.
(2) The effect of the "emissions" boundary condition can easily be tested with the
current data set and model; all species can be made time variables with the
measurement data providing an initial condition only.
References
Gear, C.W., 1971: Numerical Methods for Initial Value Problems in Ordinary
Differential Equations. Prentice-Hall, Englewood Cliffs, New Jersey.
Kahaner, D., C. Moler, and S. Nash, 1989: Numerical Methods and Software, Prentice-
Hill, Englewood Cliffs, New Jersey.
Makar, P.A., S-M. U, P.B. Shepson and J. Bottenheim, 1998: The AES Gas-Phase
Mechanism for Tropospheric Chemistry: Theoretical Formulation: AES Internal
Report, Atmospheric Environment Service, Downsview, Ontario (In preparation).
Middleton, P., W.R. Stockwell, and W.P.L. Carter, 1990: Aggregation and analysis of
volatile organic compound emissions for regional modelling. Atm. Env., 24A, pp
1107-1133.
Yung, Y.L., 1976: Numerical Method for Calculating mean intensity in an
inhomogeneous Rayleigh-scattering atmosphere. J. Quant. Spec. 16, pp755-761.
Tables:
1. NAPS stations included in this study.
Figures:
1. Sensitivity of model ozone with respect to changes in NO, Summer O3 > 70 ppbv episodes.
2. Sensitivity of model ozone with respect to changes in Total VOCs, Summer Oj > 70 ppbv
episodes.
3. Sensitivity of model ozone with respect to changes in BUTE; internally bonded alkenes,
Summer Os > 70 ppbv episodes.
4. Sensitivity of model ozone with respect to changes in TARO; tri-substituted aromatics,
Summer Os > 70 ppbv episodes.
5. Sensitivities of model ozone with respect to NOx and VOCs, Southern Ontario Stations
outside Toronto, Summer O3 > 70 ppbv episodes.
4-36
-------
6. Sensitivities of model ozone with respect to NOx and VOCs, Toronto stations, Summer O3 >
70 ppbv episodes.
7. Sensitivities of model ozone with respect to NOx and VOCs, West Coast Stations, Summer
O3 > 70 ppbv episodes.
Table 1.
NAPS#
30118
60204
60403
60413
60418
60422
60424
60512
61004
62601
63201
90130
100111
100127
101301
ST. NAME
Roy Building
UIC Building
Evans & Arnold
Elmcrest Rd.
Junction Triangle
33 Edgar Ave.
Bay and Grosvenor
Beasley Park
Centennial Park
Experimental Farm
Hwy 47 & Hwy 48
10255-104* st.
Rocky Pt. Park
Surrey East
Langley Central
NEAREST
CITY
Dartmouth
Windsor
Toronto
Toronto
Toronto
Toronto
Toronto
Hamilton
Sarnia
Simcoe
Stouffville
Edmonton
Coquitlam
Surrey
Langley
PROV.
N. Scotia
Ontario
Ontario
Ontario
Ontario
Ontario
Ontario
Ontario
Ontario
Ontario
Ontario
Alberta
B.C.
B.C.
B.C.
CITY/
RURAL
(OR)
C
C
C
C
C
C
C
C
C
R
R
R
C
R
C
4-37
-------
1.2
0.8
s °-<
.2
a
-n 00
Figure 1.
Sensitivity of Model O3 to NO vs Collection Station
Summer Ozone Episodes Greater Than 70 ppbv
0.
Q.
CO
O
-04
-0.8
-1 2
IE NorvOutUw MlK
NorvOunlM Min
D Me*»n
O Outliers (>1 Sstdin)
* Eritemes (>3 sld err)
O
[Di [Dj
I _]_
-o-
60204 60413 60422 60512 62601 100111 101301
Windsor 60403 Toronto 60418 Toronto 60424 Hamilton 610M Simcoe 63201 Coquitlam 100127 Langley
Toronto Toronto Toronto Sana Stoutfville Surrey
STATION
2.2
2.0
1 8
1.6
1.4
1.2
1.0
0.8
06
04
02
00
-0.2
Figure 2.
Sensitivity of Model O3 to VOC vs Collection Station
Summer Ozone Episodes Greater Than 70 ppbv
I
m
O
O
H; Non-Outlier Mu t
NorvOutliarUin
D Medun
O Outlets (>l 5 sld en;.
VK Exlrames (>3 std err)
..T,
I-
60204 60413 60422 60512 62601 100111 101301
Windsor 60403 Toronto 60418 Toronto 60424 Hamitor, 61004 Simcoe 63201 Coquiflam 100127 Langley
Toronto Toronto Toronto Samia StouffviBe Surrey
STATION
4-38
-------
Figure 3.
Sensitivity of Model O3 to BUTE vs Collection Station
Summer Ozone Episodes Greater Than 70 ppbv
0.8
,*»
£ °-6
1^_
5
< °4
X
f}
a 0.2
0
o
00
.n 3
*
III Non-Outlier Max »
O
O
D
! * [=!
In! ** hp
^ o
8
o
$
f°l El
- ' . Non-ouoierMIr
O Median
O Cutlers C1 5 Bid err)
; * Extremes (»3 std err)
L :
|D[ :
a
LJ 41-1 o
-o- -o- «- -o-
60204 60413 60422 60512 62601 100111 101301
Windsor 60403 Toronto 60418 Toronto 60424 Hamilton 61004 Simcoe 63201 Coquitlam 100127 langley
Toronto Toronto Toronto Sarnia Stouffville Surrey
STATION
Figure 4.
Sensitivity of Model O3 to TARO vs Collection Station
Summer Ozone Episodes Greater Than 70 ppbv
0.20
# 0.16
O
f* 0.12
"a rj.08
'
O
< 0.04
0.00
o
o
...
, : *
r~i T
L
ii'.Q'T
-
r.
:
.1C Non-Outlier Max 1
Non-Outlier Mm
CD 75% to 25%
D Median
O Outliers (>i. 6 std err)
* Extremes (>3 sld err)
i ~'~ .
\ n
^ j_ ' ^
i~i _*_
n T' o1 n
o r1! £j Lpl .
U ISJ i i ^TP ^ .-0- «?
60204 60413 90422 60512 62601 100111 101301
Windsor 60403 Toronto 60418 Toronto 60424 Hamilton 61004 Simcoe 63201 Coquidam 100127 Langley
Toronto Toronto Toronto Samia Stouffville Surrey
STATION
4-39
-------
Figure 5.
Sensitivities of OS Production to NOx and VOC
Summer Ozone Episodes Greater Than 70 ppbv
1.8
g 1-2
3 ctd oa&oo : {jj&ooffiaaaaaoo
61004 62601 63201
Sarnia Simcoe Stoufrvilla
D NO
A VOC
« C2H6
o C3H8
o C4AK
* C7AK
o C2H4
o PRPE
a BUTE
A TOLU
o DARO
o TARO
Southern Ontario NAPS Stations
Figure 6.
Sensitivities of O3 Production to NOx and VOC
V
Summer Ozone Episodes Greater Than 70 ppbv
1.2
3- 0.8
0)
.a
g 0.4
ra
I"
8 -0-4
*"
-0.8
0
$
l*»&6 ^&*5 .U
M - H
J
:
."XNon-OullierMaxt
Non-Outlier Min
D Median
O Outliers (>1. 5 stdeir)
* Exl/empt (>3slderr)
?
(Xjio6io56 ^ooaXx^^Oo -oocwoo^aoo
D '.?
y
60403 60413 60418 60422 60424
Evans&Amold BmcrestRd Junction Trfa 33 Edgar Ave Bay&Grosvenor
n NO
A VOC
« C2H6
o C3H8
a C4AK
A C7AK
o C2H4
o PRPE
n BUTE
A TOLU
o DARO
o TARO
Toronto NAPS Stations
4-40
-------
Comparing the full time series
Correlation between
full observed and modeled time series
1
4^
| 0.8
£
V
00.6
-2 0.4
.2
4)
S02
O
10 50 10/70 V1 V3
30 70 30/70 V2
Biogenic Scenario
Comparing the intra-annual perturbation
-------
And the synoptic perturbation
6s
" ] - ;
1 ""! " i ! ! 1 ~!~~ i
, .j 1_ ._j. J..._ , - -... ,-
1M IM W*
Comparative statistics...
Correlation Analysis of Synoptic Term
10 50 10/70 V1 V3
30 70 30/70 V2 UAM-V
Biogcnic Scenario
-------
ON
Error Analysis
10
70 30/70 V2 UAM-V
50 10/70 V1 V3
Biogenic Scenario
Comparative Statistics
50
a* 40
Q.
a.
£30
(U
5
75 20
3
C
< 10
0
Obs 30 70 30/70 V2
10 50 10/70 V1 V3
Biogenic Scenario
100
80
60
CJ
05
40 *
20 5
0
-------
Conclusions
[ Shows promise as an approach for developing a
"modeled ozone climatology"
* applications in weight-of-evidence
component in SIP
.fx
^ r* Model performance seems good
00 *» need to ensure that we're not "getting
the right answers for the wrong
reasons"
Demonstrates the importance of modeling the
time dependence of biogenic VOC emissions
*» gives an indication of the temperature
dependence that is well known for
ozone in the Northeast
-------
Abstract
TVA's Research In Tropospheric Ozone Mitigation
and Contribution of Natural Hydrocarbons to VOC Reactivity
Roger L. Tanner, Ph.D.
Atmospheric Sciences & Environmental Assessments
Environmental Research and Services
Tennessee Valley Authority
Muscle Shoals, AL 35662-1010
The Tennessee Valley Authority has, through its partnerships with OtNr federal and state agencies,
universities, and private sector participants, Contributed substantially to the knowledge base
concerning tropospheric ozone. Particularly through participation in the Southern Oxidants Study
(SOS), ;ts field studies and related modeling efforts, new information concerning the formation of
ozone in plumes from urban areas and point sources of precursor NOX and VOC emissions has been
developed. This information is critical in diagnosing whether reductions in NOX emissions, VOC
emissions, or both are most effective in reducing the likelihood of exceedances of the NAAQS for
ozone. These efforts have also provided new insight on whether emissions of NO, from all sources
v
produce ozone at the same efficiency when mixed with VOCs found in ambient air. TVA and its
SOS collaborators have also found that, in the Southeast during the "high ozone" season, natural
hydrocarbons and specifically isoprene are the major source of VOC reactivity in ozone formation
processes in non-urban areas. Urban area VOC emissions are clearly important for ozone mitigation
in populated areas, but different strategies may be required for urban and rural areas especially in
light of the new 8-hour NAAQS for ozone.
4-69
-------
Computational Studies of Oxidant Reactions of Volatile Organic Compounds Relevant to the
Formation of TropOSpheric Ozone: David A. Dixon. Thorn H. Dunning, Jr., Mloh«l Dupuls
Volatile organic compounds or VOCe play a key role in the global carbon cycle
- the direct formation of carbon monoxide CO from their oxidation by
radicals and ozone.
- primary seed compounds leading to the formation of aerosols which
provide reaction sites and act as carriers of condensed active species
Develop a fundamental molecular understanding of the oxidation of VOC's by
using advanced computational electronic structure methods on high performance
computing systems.
Calculate the thermodynamio and kinetic information needed to predict
degradation mechanisms. Provide both novel insights and chemically accurate
data. Provide spectrosooplo information for identifying key intermediates.
Develop a base capability in the modeling of rates and mechanisms of oxidation
processes of importance to the Atmospheric Chemistry Program. Such a
capability can be used to help guide further experiments and to extend limited
experimental data into new domains.
Reactions of oxidants such as OH. Cl, NOa, and Os with key VOCs
/,
HC C.
// \\
H,C CH,
CH38CH3
1 2 3
a-pinene {3-plnene ieoprene
TM&-9 Capabilities and Interest Relative to Atmospheric Chemistry
Reliable predictions of thermodynamic properties of atmospheric compounds
and intermediates including radical species, including predicting molecular
structures.
Reliable prediction of kinetics of important atmospheric chemistry reactions
(rate constants from ab initio molecular theory, variational transition state theory,
tunneling effects).
Models of reaction mechanisms and pathways.
Thermodynamics of cluster formation.
Global warming potentials (infrared intensity calculations).
Predtotions of exerted state chemistry.
Focus on aerosol formation from natural and anthropogenic sources.
Expertise In stratospheric chemistry, tropospheric degradation processes,
cluster formation, and global warming potential predictions.
Expertise In algorithm development end software implementation on massively
parallel computers for development of high performance, portable, and scaleable
software.
4-70
-------
EMSL Molecular Science Computing Facility
A DOE National Scientific User Facility
ZVNWChem
Experimental Computing Graphic* & Visualization
Laboratory _^^^ Laboratory
Applied Mathematics
Computer Science
Molecular Science
High Performaqce Computing
r'acility
Pacific Northwest National Laboratory
Multidisciplinary Teams for
HPC8I Software Development
Computer
Science
Applied
Mathematics
Application Discipline
- theories
- approaches
Applied Mathematics
- numerical analysis
- mathematical algorithms
Computer Science
- software methodolgy
- software tools
4-71
-------
Molecular Science Software Distributed Computing Model
Molecular Science Computing facility
///con/, Mdi/i'lui1^
-------
Environmental Molecular Sciences laboratory
Pacific Northwest National
Current NWChem Functionality
Quantum Mechanical Capabilities:
Hartree-Fock energies, gradients, and second derivatives.
Multiconfiguration self consistent field (MCSCF) energies and gradients.
Density functional theory at the local and nonlocal levels (with N3 and N4 formul scaling)
energies and gradients.
Many-body perturbation theory (MP2-MP4) energies plus MP2 gradients.
Coupled cluster [CCSD and CCSD(T)] energies.
Single and multireference configuration interaction energies.
Segmented and generally contracted basis sets including the correlation-consistent hasis
sets under development at EMSL.
Effective core potential energies, gradients, and second derivatives.
Classical Mechanical Capabilities:
Energy minimization
Molecular dynamics simulation
Free energy calculation
Supports variations such as: multiconfiguration thermodynamic integration or multiple-
step thermodynamic perturbation, first order or self consistent electronic polari/.ation,
simple reaction field or particle mesh Ewald, and quantum dynamics
i.it\ ironnit niul \}i>lt'ciiltir .S
High Performance Computational Chemistry
I'a i if 1C So rill* < vf .\\ilin/i ill I iil>i-tti!iir\
Current NWChem Functionality Targets
1F^Classical to Highly Correlated
(KT HF MP(2-4) MCSCF CCSD(T) MKSDCI ...
Staling
F
» 0.1 kcal
^ 10 atoms
1 000 bf 500 hf
100 atoms 5 000 000 CSFs
10 - 20 atoms
4-73
High f'crformance Computational Chemistry
-------
Environmental Molecular Sciences Laboratory
Pacific Northwest National laboratory
Measured Parallel Efficiency for NWChem - DFT on IBM-SP;
Wall Times to Solution for Full SCF Convergence
32 64 66 126 ISO 182
Number of Nodes
224
256
Zeolite
Fragment
SflH,
Wi
w»
Wi
Basis AQCD (Number of
Nodes
347/832 64
mm I2X
1199/2X18 256
I6S?H 256
Wall Tire to
Solution
238s
364s
1137s
2766s
High Performance Computational Chemistry
Theory, Modeling iV Sinnilnliiin ^^s=^~=ss=^^^=s=^=
Required Accuracies for Chemical Qystems:
Qeparafions and Catalysts
Absolute Rates (Speedup for Catalysts): Impact on Rates
- Factor of 10 @ 25°C is AEa = 1.4 kcal/mol
Relative Rates/Equilibrium Constants (Selectivity for Catalysts and Separations):
Impact on Selectivity
- Change from a 50:50 mixture to a 99:1 mixture @ 25°C
Keq =1 changes from Keq = 100
AO = 0 to AG = 2.8 kcal/mol
There is a dear requirement for accuracy in the computational results. How is
this accomplished?
i Environmental Molecular Sciences Laboratory/ _
4-74
-------
Expansion of the Many-Electron Wave Function :
Methods of Electron Correlation
MP/i
(MBPT)
Molecular Orbital
Wave Functions
SCF(RHI-VUHF)
^^H
CISD
C/SDTO
CCSO
CCSD(T)
CCSDT
MCSCF (CASSCF)
T
CASPT2
MRCI
How to best represent the molecular orbitals:
Correlation Consistent Basis Sets
How important are different types
of Gaussian functions?
100.0
_ 10.0
ui
i.o
0.1
Oxygen atom
\
(g) (0
1234
Number of Functions
4-75
Functions are added in correlation
consistent shells
cc-pVDZ
cc-pVTZ
cc-pVQZ
cc-pV5Z
-------
CH4
ID.ICHJ = 419.60 kcal/mol
ID.(CH4) = 420.71 kcal/mol including core correlation and spin orbit effects?
ZPE = 27.09 kcal/mol.
AHf (CHJ = -17.1 kcal/mol at 0 K
Experimental value of -16.0 _+ 0.1 kcal/mol
Use of larger A2PE = 27.71 kcal/mol (Grev et.al.): AHr(CH4) = -16.5 kcal/mol
LD.(CH4) = 419.2 kcal/mol by exponential extrapolation.
AH,(CH4) = -15.6 kcal/mol based on AZPE = 0.5IV,
Use of the Grev value for AZPE gives AH^CHJ = -15.0 kcal/mol.
Organic Thermochemistry & Kinetics
Estimates of computer time required for CCSD(T) calculations with cc-pVQZ basis
sets. Estimates are based on an algorithm that scales as Ne, rather than the N7
formal scaling.
Molecule Time (1 TFlop) Time (100 Tflop)
CeHe 1 min 0.6 sec
C8H18 40 min 25 sec
C,eH34 40 hr 25 min
C24H6o 600 hr 6 hr
Kinetics: Probably requires augmented cc basis sets: 10 x per point at the
aug-cc-pVQZ level
4-76
-------
Organic oxidation reactions: Thermodynamics
,MI?/DZ+P level and final wtfes calculated at the
in projected) level.
CH3OOH -> CH300' + H' AHca,c = 86.2 kcal/mol
AHexpt = 87.4 kcal/mol
CH3OH -> CH30* + H* AHcalc = 104.0 kcal/mol
AHexpt = 104.2 kcal/mol
CH4 -> CH3« + H* AHcalc = ,03.5 kcal/mol
AHext = 104.8 kcal/mol
Organic oxidation reactions: Kinetics
Calculate for the following prototypical reactions
- the barrier heights, AE*(0), and activation energies.
- Ea, from transition state theory with a Wigner tunneling correction
Geometries optimized at the MP2/DZ+P level and final energies calculated at the
PMP4/TZ2PF (spin projected) level.
CH4 + *OCH3 > CH3 OH -I- CH3* AE'(O) = 12.0 kcal/mol
Ea = 10.2 koal/mol
CH4 + CHs* > CHs* 4- CH4 AE*(0) = 19.3 kcal/mol
Ea = 17.9 kcal/mol
CH4 + *OOCH3 > CHs* + CHs OOH AE»(0) = 23.8 kcal/mol
Ea = 21.2 kcal/mol
4-77
-------
Isoprene Chemistry: 1
//
HC C
/
CH,
CH,
OH CH3
+ OH *~ HC C +
CH2 CH2
4a
HC' C
CH,
\
CH,
OH
4b
CH, /CH3
H ' , ,,-. /-*
C C OH + HC C-
CH,
4c
OH
4d
(6)
Isoprene Chemistry: 2
4a to4d+ O-,-
OH
HC
/
OOCH,
C
CH,
CH2
OO'
HC C
/
\
,CH3
CH,
CH,
H2C
5a
OH
5b
HC==C
/
OH
5c (c/t)
\
CH,
OO-
,CH,
HCC OH
CH2 CH2
OO-
5d
HC^~ COO'
S \
CH2 CH2
HO
Se
HC= C
oo.
\
C
OH
(7)
5f (c/t)
4-78
-------
Isoprene Chemisfry: 3
P" .CH, O. CHJ CH,
5ato5f+NO *- HC C + HC C + // \
/ % / ^ // ^>
OCH2 CH2 HOCH; CH2 HOCH2 H:CO-
6a 6b 6c (c/t)
CH3 CHj CH,
+ HC - COH + HC CO- + HC^=C
S \ -/ \ / \
CH. H2CO- H^C H:COH -OCH^ H:COH
Isoprene Chemistry: 4
OCH: CH2 H H
6a 7
O CH3
(K)
6d 6e 6f(c/t)
\ / II \ /
HC C P-sc'SSIQrV C + wr r (9a>
/ ^ / \ -
C C + >OOH (9h)
/ \ / ^
H CH CH:
OH
9
OH CH,
\ /
1.5-H Shift^ HC C (90
/ ^
HOCH2 CH;
10
4-79
-------
9(CH3)2 Chemistry
CH39(OH)'CH3 + 02 -> CH39(0)CH3 + H02<
-> CH390H + CH302-
-> (CH3)2902 + »OH
N03 (Nighttime) Chemistry
CH.i
V l°:l \
+ NO3 ^ «CH CH2 ^ CH CH:
ONO-. -O 0 ONO
CHj H
NO (20)
H V H
0
O
II
/\ + N02 ^
H H
CH2 ?i
| + II ^
°N°= H/XCH3
0
CH3
CH CH2
ONO2
C ON02 I [Oa]
H CH2
4-80
-------
Ozonolysis
XW { V /-\. O
\ / \ / \ y v
Criegee Intermediate Reactions
CH200' + R02* -» HCHO + RO- + 02
OCH20- -H 02 -> HC02- + H02-
OCH20- + HCHO -> HOCH2-0-CHO
Theory, Modeling & Simulation
Unimolecular Decomposition of CH3CH20
Method AE
MPO/TZ2PF 10.93 7.1 23.17 20.§
PMP2/" 11.16 7.3 16.32 13.7
MP49DQ/" 14.80 11.0 24.19 21.6
PMP49DQ/" 14.98 11.1 19.35 16.8
MP49DTQ/" 13.18 9.3 22.90 20.3
PMP49DTQ/" 13.35 9.5 18.21 15.6
QCI9D/" 15.78 11.9 22.04 19.5
QCI9D(T)/" 15.50 11.7 20.54 17.7
CC9D/" 16.13 12.3 22.40 19.8
CC9D(T)/" 16.68 11.7 20.85 18.3
CC9D(T)/aug-ooTZVP 16.61 11.6 19.53 17.0
BP/DZVP2 20.63 18.1 20.71 18.1
B31YP/DZP 17.87 14.6 19.96 16.8
B3LYP/TZ2PF 15.13 11.7 18.13 14.9
Expt 13.1(NA8A/JPl/94) 21.e(Batt)
Energies in kcal/mol, Geometry at MP2/TZ2P level except for NLDFT
! Environmental Molecular Sciences Laboratory ss
4-81
-------
Theory, Modeling & Simulation
Unimolecular Decomposition of
Rafes
RRKM. High Pressure limit, 298K, N2 collision pertner.s = 4.2 A, = 260 K
Method Ea(koel/mol) log A(«-l) k(n-l)
QC18D/TZ2PF 18.3 13.7 1.2
QCI9D(T)AZ2PF 20.3 13.7 0.041
CC8D(T)/TZ2PF 18.9 13.7 0.46
CCSD(T)/8ug-oo-VTZ 17.7 13.7 3.2
Expt 21.6 15.0 0.14
Tunneling estimate for an imaginary frequency of 642i with a Wigner correction
booed on the reverse reaction: 1.4 @ 298 K, 1.46 @ 277 K
BsH, M. J. Ch«n. Kinei. //, 977 (1979)
* Environmental Molecular Sciences Laboratory \
Benchmark Calculations
for Abstraction of H from CH4
CH4 + OH' > CH3* + H2O
- Model for hydroxyl radical decomposition of alkanes in troposphere
RH + OH* > R* + H2O
- Reactions have low activation energies leading to alkanes having
short atomospheric lifetimes.
CH4 + Cl* > CH3' + HC1
- Atmospheric sink for Cl atoms which participate in the destruction
of ozone.
CH4 + H* > CH3f + H:
- Simplest reaction of a radical with a hydrocarbon.
- Potential importance in the combustion mechanism of simple
hydrocarbons.
4-82
-------
Computational Model
Abstraction of H from CH4
[H3C-H-X]'*
CH + X
CH3* + HX
What does the TS structure look like?
What is the overall reaction enthalpy, A//298?
What is the barrier height, A£a*?
Kinetic Parameters From Transition State Theory
rigid rotor, harmonic oscillator
1 free internal rotor
Wigner tunneling correction
E. - 4.32 kcal/mol
A - 1.47 x 10-11 cc/moleoule - s
E. - 5.56 kcal/mol
A - 1.16 x 10-10 cnrvVmolecgte - s
Experimental values
E, - 3.6 kcal/mol
A - 2.9 x 10'12 crrvVmolecUe - s
200<;T<;420K
200 ^ T i 3000 K
4-83
-------
-20-
-24-
-28-
-32-
-36-
-40
O
D ln[k(calc)]
O ln[k(expt)]
8
0
D
CH4 + OH > CH3 +
0.0 1.0 2.0 3.0 4.0 5.0
1/T (x 100
HFC-23 and HFC-236fa Reactions with OH
HFC-23 = CF3H & HFC-236fa = CF3CH2CF3
Molecule
exot
calc
HFC-23
HFC-236fa
2.4x10-16
3.4x10-16
6.9x10"16
6.1
Experimental values from DeMore's work at JPL
Rate constants in cm3/molecule-sec
4-84
-------
ATMOSPHERIC CHEMISTRY OF ORGANIC COMPOUNDS
Roger Atkinson
Photochemical Reactivity Workshop
May 12-14, 1998
Tropospheric VOC Removal Processes
The tropospheric removal or transformation processes for
VOCs are:
Physical Removal Processes
Dry deposition
Wet deposition
Chemical Removal Processes
Photolysis
Reaction with ozone (O3)
Reaction with the hydroxyl (OH) radical
Reaction with the nitrate (NO3) radical
-------
VOC
ROOH
R
I0'
NO2
RO2 ^ *" ROON02
ROj
carbonyl
+
alcohol
NO
RONO2
RO-
products
Net photochemical formation of O3 versus net photochemical
loss of O3 in the troposphere depends on the rate of
HO2 + NO -» OH + NO2
versus
HO2 + HO2 -* H2O2 + O2
and
HO2 + 03 -» OH + 2O2
and also by the rate of
RO2 + NO - RO + NO2
versus
RO2 + HO2 -* ROOH + O2
-------
Organic Reactions (genera!)
Peroxy Radical Reactions
voc -»-
ROOM ^ R02'
carbonyl
+
alcoliol
T ROON02
RONO7
ROT
products
Reactions of organic peroxy radicals
Organic nitrate formation.
Reactions of alkoxy radicals.
Wet and dry deposition of VOCs and of their reaction
products.
RO, + NO
RONO2
RO + NO,
RO2 + NO2 ** ROONO2
RO, + HO, -» ROOH + O,
RCH(OO)R -f
RCH(O)R + RCH(6)R + O2
RCH(OH)R + RC(O)R -f O2
RO2 + NO3 -* RO + NO2 + O2 (or other products)
There is a need for kinetic and product data for the
reactions of a wide variety of organic peroxy radicals
with NO, HO2 radicals and NO3 radicals.
-------
R02 + NO
Organic Nitrate Formation
RONO2
RO + NO2
The nitrate yields increase with increasing pressure and
with decreasing temperature.
Data are available for 18 secondary alkyl radicals formed
from alkanes and for 4 other alkyl and /3-hydroxyalkyl
radicals formed from alkanes and alkenes (mainly at
room temperature and atmospheric pressure).
decomposition
x-"
|+ CH,CH2CH2
Jo,
NO I»- NO2
CH,CH2CH26
( CH,CH;CHo] .
CHjCH(O)CH2CH2CH,
isomen nation
HO2 + CH,C(O)CH2CH2CH,
H' CH-CH,
CH2
1
CH,CH(OII)CII2CH1CH2
I" '
NO -- * NO;
CH3CH(OH)CU2CH2CH2b
isomcrizanon
CH,C(OH)OI2CH2CH2OH
I0-
CH,C(O)CH2CH2CH2OH I + HO2
-------
OH and NO3 Radical Reactions with Alkanes and Alkencs
The initial reactions lead to the formation of alkyl or
substituted-alkyl (R) radicals, which then add O2 to form
RO2 radicals.
The present knowledge and uncertainties in the OH
radical-initiated reactions of alkanes and alkenes are:
RONO2 formation from RO2 + NO
RO2 + RO2, RO2 + HO2 and RO2 + NO, reactions
Reactions of alkoxy and hydroxyalkoxy radicals:
Reaction with O2 (if a-H atom present)
Decomposition
Isomerization through a 6-membered transition
state
isomenzation
CH3CHCH2CH(OH)CH2OH
CH3CH2CH2CH(OH)CH2O
\-
H02 +
decomposition
CH3CH2CH2CH(OH)CHO
CH3CH2CH2CHOH
CH3CH(OO)CH2CH(OH)CH2OH
NOj*- NO2
Cl I,CH(O)CH2CH(OH)CH2OH
isomcnzalion
OI3CH(OH)CH2Cll(OH)CHOH
02
CH3CH(OH)CH2CH(OH)CHO + HO2
M02
-------
OS
Isomerization reaction has been observed from alkane
and alkene reactions; quantification of the resulting
hydroxycarbonyl and dihydroxycarbonyl products is now
required.
The products and mechanisms of the NO3 radical
reactions are not well understood; in part because these
reactions occur in the essential absence of NO and hence
RO2 + HO2 and RO2 + RO2 reactions are important and,
especially in laboratory systems, ROONO2 are important
intermediate reservoir species.
R. B
o3 + )c=c(
T» /
R!-
f
R4
R,C(O)R2 + (RjR4COOJ*
followed by reactions of the biradicals
R3C(O)R4 » [R:R2CO6p
[R,CH2C(R2)OO]' + M
[R,CH2C(R2)00]'
R,CH2C(R2)OO + M
- [R,CH2C(0)OR2]'
decomposition
PRODUCTS
(including R]CH3 if R2 = H)
[R,CH2C(R2)00]'
[R,CH=C(OOH)R2]'
R|CHC(0)R2 f OH
-------
Areas of uncertainty:
Reactions of the thermalized biradicals.
Appear to be with water vapor under atmospheric
conditions.
The CH2OO biradical reacts with H2O to form
HOCHjOOH which (heterogeneously?) decomposes
to HC(0)OH + H2O.
Certain more complex biradicals appear to react
with water vapor to form the carbonyl (plus H2O2).
R,C(O6)R2 + H2O - R,C(0)R2 + H2O2
Reactions of the organic radical co-product to OH; e.g.,
CH3C(O)CH2 radical from the [(CH,)2CO6l* biradical.
AROMATIC HYDROCARBONS
For benzene and the alkyl-substiruted benzenes, the major
atmospheric reactions are with OH radicals (major) and
NO3 radicals (minor).
NO3 radical reactions proceed by overall H-atom
abstraction from the alkyl substituent groups.
OH radical reactions proceed by overall H-alOm
abstraction from the alkyl substituent groups (< 10%)
and by OH radical addition to the aromatic ring to form a
hydroxycyclohexadienyl radical (^90%).
-------
CH3
OH
oo
CHj
H20
Under tropospheric conditions, the
hydroxycyclohexadienyl radicals (OH-aromatic adducts)
react with O2; at elevated NO2 concentrations
encountered in some laboratory studies the OH-aromatic
adduct reactions with NO2 may be important.
The products and mechanisms of the reactions of the OH-
aromatic adducts with O2 and NO2 are not presently
understood in any detail, although product data
(sometimes contradictory) are available from a number of
laboratory product studies.
Formation of ring-opened unsaturated dicarbonyls
[-C(O)C=CC(O)-] and di-unsaturated dicarbonyls
[-C(O)C=CC=CC(O)-] have been observed and may be
very important.
-------
,OH
II
1
H
OH
+ HO2
,OH
OH
00
These radicals formed after O2 addition to the OH-benzene
adduct react further to (potentially) form:
HC(O)CHO + HC(O)CH = CHCHO
HC(O)CH =CHCHCHCHO
HC(O)CH=CHCH = CHCHO
Additionally benzene oxide/oxepin reacts to form
HC(O)CH = CHCH = CHCHO
-------
Tropospheric Chemistry of Oxygen-Containing Compounds
Aliphatic aldehydes, ketones and a-dicarbonyls
Aliphatic aldehydes, ketones and a-dicarbonyls.
Alcohols.
Ethers and glycol ethers.
a,|8-Unsaturated carbonyl compounds.
<^>
i
5 Unsaturated dicarbonyls.
Esters.
Hydroperoxides.
Other oxygenated compounds.
These react with OH radicals and (to a lesser extent) with
NO3 radicals, and also photoiyze.
Need absorption cross-sections and photolysis quantum
yields as a function of wavelength [apparently reliable
cross-section and quantum yield data arc available only
for HCHO, CH3CHO and (CHO)2].
The OH radical and NO3 radical reactions with >C2
aldehydes lead to peroxyacyl nitrate (PAN) formation.
-------
Esters
Reaction of RC(O)OCH(6)R radicals:
RC(O)OCH(6)R -» RC(O)OH + RCO
j_ Ethers and Glycol Ethers
Decomposition of > COC(6)RR radicals appears to be a
factor of ~ 103 faster than expected by analogy with tlic
alkoxy radicals formed from alkanes and alkenes.
NITROGEN-CONTAINING ORGANICS
Organic nitrates (RONO2) and peroxyacyl nitrates
(RC(O)OONO2) appear to be the most important N-
containing compounds.
Rate constants for the OH radical reactions are available
for alkyl nitrates; product data are needed.
For RC(0)OONO2 compounds (apart from PAN), data
are needed for photolysis and thermal decomposition.
COC(6)RR - > C(O)OR + R
-------
to
CONCLUSIONS
Much progress has been made over the past 2 decades:
Importance of NO3 radical reactions.
Kinetics of OH and NO3 radical and O3 reactions
with VOCs.
Studies of RO2 radical reactions.
Alkoxy radical reactions (isomerization)
Fate of hydroxycyclohexadienyl radicals, including
formation of ring-opened unsaturated dicarbonyls
from aromatic hydrocarbons
Product and mechanism studies of O3 + alkenes;
formation of OH radicals from these reactions.
Still many details to deal with!
Needed Research
Quantitative knowledge of the rate constants and
mechanisms of the reactions of organic peroxy (RO2)
radicals with NO, HO2 radicals, NO3 radicals and other
RO2 radicals (the latter mainly to allow accurate
modeling of irradiated NO, - VOC - air mixtures.
Additional data concerning the organic nitrates yields
from the reactions of organic peroxy radicals with NO,
preferably as a function of temperature and pressure.
The reaction rates of alkoxy radicals for decomposition,
isomerization and reaction with O2, especially of alkoxy
radicals other than those formed from alkanes and
alkenes (for example, from hydroxy-compounds, ethers,
glycol ethers and esters).
-------
Detailed mechanisms of the reactions of O3 with alkenes
and VOCs containing >C = C< bonds. This involves
understanding the reactions of the initially energy-rich
biradicals, and thermalized biradicals, formed in these
reactions.
Mechanisms and products of the reactions of
OH-aromatic adducts with O2 and NO2.
Tropospheric chemistry of many oxygenated VOCs
formed as first-generation products of VOC
photooxidations, including (but not limited to) carbonyls
(including unsaturated dicarbonyls, di-unsaturated
dicarbonyls, and unsaturated epoxy-carbonyls),
hydroperoxides, and esters.
-------
Atmospheric Chemistry of Oxygenated Organic
Compounds
Ray Wells
AFRL/MLQR
139 Barnes Drive
Tyndall AFB, FL 32403-5323
(850)283-6087
ray.wells@ccmail.aleq.tyndall.af.mil
Uncharacterized volatile organic compound (VOC) emissions from complex
formulations (coatings, coating strippers, cleaners) are involved in the production of
tropospheric ozone (63), a regulated pollutant. Since the detailed atmospheric chemistry
of several of these chemicals has never been investigated, experimental atmospheric
research coupled with incremental reactivity calculations is useful to more accurately
assess the atmospheric impact of coatings emissions. The atmospheric impact of the
coating systems was determined, using individual VOC incremental reactivity
calculations, coupled with a detailed description of coating system emissions. The
concentrations and identification of VOCs in the coating emissions were determined by
combining gas chromatography, mass spectroscopy and Fourier transform infrared
spectroscopy (GC/MS/FTIR) techniques. The OH rate constant for ethyl 3-
ethoxypropionate was determined using the relative rate technique. The products of the
OH + ethyl 3-ethoxypropionate reaction were determined and an atmospheric reaction
mechanism for ethyl 3-ethoxypropionate was proposed.
5-14
-------
AIR FORCE RESEARCH
LABORATORY
ATMOSPHERIC CHEMISTRY
OF OXYGENATED ORGANIC
COMPOUNDS
Ray Wells
AFRL/MLQR
Air Team
(850)283-6087
ray.wells@ccmail.aleq.tyndall.af.mil
AIR TEAM
Lt. Leon Perkowski
Dr. Ray Wells
Darrell Winner
Stewart Markgraf
Steve Baxley
Sheryl Wyatt
Bill Bradley
We determine impact on air quality.
5-15
-------
URGENCY
-1990 Clean Air Act-
200 new regulations
and guidance documents
They impact our missions!
DoD Releases by
Media, 1994
Land Water
1.31% 1.25%
Air
97.44%
From: 1994 Toxics Release Inventory for the DoD Publie Data Report, March 7, 1996
5-16
-------
Of That 97%...
Metals - 6%
Acids - 4%
Chlorinated
organics
and
chlorine
55%
Impacts
Stratospheric
ozone
Volatile
Organic
Compounds
35%
Impacts
Ground level
ozone
Compromise ozone formation and depletion
Sources of Pollution
Paints
Thinners
Solvents
Combustion
Exhausts
New replacements affect these sources
5-17
-------
Purpose
GOAL: Prevent pollution intelligently while maintaining performance
Achieved by addressing:
1. What is being emitted?
2. What happens to emitted chemicals?
Tech Need: 1940 - Replacement of chlorinated
cleaners for engines (High)
Experimental Apparatus
Reaction Chamber
Sample Loop
Analysis System
5-18
-------
Pertinent Radical Formation
Reactions
OH Radical
O + hv
O(1D) + H2O
RH + OH
0(1D) + O2
2 OH
R +H2O
NO3 Radical
NO2 + O3
NO2 + NO:
RH + NO3
NO3 + O2
N205
R + HNO3
O Atmospheric Transformation llSRi
v , Processes ^bss3^
DU -I- OU-
KM + \Jr\
+ o2
|)/^S . k 1 S-\
RO2 + NO
NO2 + hv
01 /^i i R It
+ O2 + M
RLJ + OH- + 9O
r\n ^ \j\\ ^w2
R. + u n
^ n2v/
» i~^k /*\
RO2
> M/^\ _L D/^
NU2 + KU
k hi /"^i . /^"X
NO + O
« /^\ i iv yi
Oq + M
o
PO + u n+ o
' r\w ~ n2\_/^ v^3
5-19
-------
Experimental Methods
I. Relative Rate Technique:
Compare unknown hydroxyl reaction rate to one that is
known:
1)
2)
Reference + OH-
Unknown + OH'
Products
Products
Dividing differential equations to remove OH concentration
and time and integrating yields:
In
[Unknown],,
[Unknown],
i\Unknown i
n
M i
K Reference
[Reference ]0
[Reference],
Hydroxyl Radical Rate Constant for EEP
CH3CH2OCH2CH2C(=0)OCH2CH3
1.2
0.00 0.01 0.02 0.03 0.04 0.05
{ln([Ref]0/[Ref]|)/kRe,}*10'12cm3molec"1 s'1
5-20
-------
Hydroxyl Radical Rate Constants
Rate constants and chemical structures are variable
Compound/Structure
kOH(10-12cm3molecule-1s-1)/lifetime(hr)
Ethyl 3-ethoxypropionate
CH3CH2OCH2CH2C(=0)OCH2CH3
Hexyl Acetate
CH3(CH2)5OC(=0)CH3
Isobutyl Acetate
(CH3)2CHCH2OC(=O)CH3
2-Butoxyethanol
CH3(CH2)3OCH2CH2(OH)
2-Butanol
CH3CH2CH(OH)CH3
23/12
9.3/30
6.5/43
22.5/12
8.1/34
5-21
-------
Experimental Methods
II. Product Identification and Yields:
Unknown + OH > Products
Must correct for transformation product/OH reaction to
determine yield:
[Unknown],
P _ k-...... ~ k....... [Unknown],
k°'"- f [Unknown] V [Unknown],
V [Unknown] ) [Unknown],
5-22
-------
CH,-CH9-O-CH
2-CH2-C(=O)O-CH2-CH3 + OH
CH3-CH-O-CH2-CH2-C(=O)O-CH2-CH3
- "NO,
* decompose
HC(=O)-O-CH2-CH2-C(=O)O-CH2-CH3
Ethyl (3-formvloxv) propionate (EFP)
CH3-CH2-O-CH-CH2-C(=O)O-CH2-CH3
O2. NO -
CH,-CHrO-CH2-CH-C(=O)O-CHrCH3
decompose
O2, NO
HC(=0)-CH2-C(=0)0-CH2-CH3
Ethyl (2-formyl) acetate (EFA)
CH3-C(=0)H
Acetaldehvde
decompose
I
CH(=0)-C(=O)O-CH2-CH3
Ethyl Glyoxatc (EG)
CH3-CH2-O-C(=O)H
Ethyl Formate (EF)
CH3-CH2-0-C(=O)-CH2-C(=O)0-CH2-CH3
Diethvl Malonate (DM)
2-Butanol
CH3CH2CH(OH)CH3 + OH
CH3CH2C(=0)CH3
? Where did oxygen come from in methyl ethyl ketone product?
?Is this major transformation pathway a source of ozone?
Experiment to reveal mechanism:
CH3CH2CH(18OH)CH3 + OH f - » - * CH3CH2C(=180)CH3
Experiment reveals that major transformation pathway is not
a source of ozone.
5-23
-------
Reaction Product Identification
V)
'£
.a
-2-
I
TO
o
JD
Experimental Spectrum
Pure 1,2 Ethanediol acetate formate
3000 2000
Wavenumbers (crrr1)
1000
Corrected Ethyl Formate Product Yield
0123
EEP Reacted (ppmv)
5-24
-------
Reactivity of Emissions
Reactivity(ozone forming potential) of individual chemicals requires
knowledge of atmospheric kinetics and mechanisms.
Reactivity of emissions is based on summation of reactivity of
individual chemicals.
Reactivity values for each formulation are used to assess impact
on air quality and minimize regulatory impact.
EMISSIONS - The Real Problem
Evaporation!
Emissions NOT content!
5-25
-------
8
ro
TIC for MIL-P-23377F
Methyl benzene
IBA
4-methyl-2-pentanone
MEK
1,3 DMe benzene
1,4 DMBZ
Ethyl benzene
Time (Minutes)->
10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00
Air Quality Impact
Coating A
= 0.27 g O3/g paint
Coating B
= 0.27 g O3/g paint
MIL-P-23377F = 0.75 g O3/g paint
jram O/gram paint = £ (gram emitted VOC/gram paint}* Factor,
voc
5-26
-------
Materials Benefits
Meet pollution prevention goals.
Formulation flexibility.
5-27
-------
Multicomponent Aerosol Generation System (MAGS) for the
Study of Fine Particulates on Photochemical Reactivity of Organu
nics
5 V. Kulkarni. KulTech Incorporated
Research Tnangle Park, North Carolina
M. B Ranade. Particle Technology Inc
Beltsville, Maryland
Considering the relative merits of using individual VOC species reactivities versus
VOC-group ratings (S Question/Issue 4 ) may address the effects of synergism within the
VOCs present in an organics emission source, but it does not consider the chemical
reactions that may affect photochemically active VOC species Mixed oxides and salts
such as SiO2/TiO2 are known photocatalysts in oxidation of ethylene In the study of
binary metal oxides as photocatalysts, it has been noted that there is a strong correlation
between surface acidity and reactivity Other reactions between gaseous species and
VOCs are also likely to affected The fine particles in industrial emissions may contain
metal oxides and sulfates and nitrates and may enhance or reduce the photochemical
reactivity depending upon the organics present in the emissions. We propose to study the
role of atmospheric fine particulates on the photochemistry of volatile organic
compoundsCVOCs).
We have developed a compact and portable multicomponent aerosol generation
system(MAGS) which produces a paniculate composition closely mimicking the ambient
and stack particulates The system has the ability to produce representative aerosols
containing inorganic oxides(and Ca, Al silicates), sulfates, ammonium nitrate, organics -
hygroscopic, solid and Hquid, volatile components in appropriate size ranges
MAGS, shown in the Figure below, can combine several types of particulates such
as mixed oxides, mixed salts and other species such as carbonaceous compounds each
type may be produced by nebulization and chemical conversion of precursor solution
droplets Particle size distribution of each component may be varied from submicrometer
to 10 micrometers
hotochemical Reactor
Aerosol Conversion
Reactors
Nebulizers
precursor Solutions
5-28
-------
Numerical Study of the Development of an Ozone Episode in
Germany: Relation of Anthropogenic and Biogenic Hydrocarbons
F. Fiedler, H. Vogel, B. Vogel
Institut fur Meteorologie und Klimaforschung
Forschungszentrum Karlsruhe / Universitat Karlsruhe
Postfach 3640, 76021 Karlsruhe, Germany
Abstract
A mesoscale numerical model is used to study the development of ozone concentrations
within the atmospheric boundary layer for different emission scenarios of anthropogenic and
biogenic emissions. The major results are:
In the reference case the maximum ozone concentrations are in the order of 100 ppb.
- When all anthropogenic emissions are switched of the maximum ozone concentrations are
in the order of 60 ppb.
When the anthropogenic emissions of the four most reactive hydrocarbons within the
RADM2 gas phase mechanism are switched off the ozone concentrations are reduced up
to 15 ppb. i
'. When all anthropogenic hydrocarbon emissions are switched of the ozone concentrations
are reduced up to 50 ppb.
1 The problem
The atmosphere is a huge deposit for gaseous waste from a large variety of anthropogenic and
biogenic sources. Due to the complex interaction of the chemical and physical processes in the
atmosphere a clear determination of the importance of individual substances for the develop-
ment e.g. of photooxidants like ozone is very difficult.
In this paper a comprehensive model system including the most relevant atmospheric and
chemical processes is used to estimate the influence of man made nitrogen oxides and hydro-
carbons on the development of maximum ozone concentrations during an ozone episode.
Especially four cases will be studied:
The reference case where both the anthropogenic and the biogenic emissions are included.
5-29
-------
The background case where all anthropogenic sources are switched off and only natural
emissions of nitrogen from soil and of hydrocarbons from biogenic sources are
considered.
The case where all anthropogenic emissions of hydrocarbons are switched off.
The case where the anthropogenic emissions of the four most reactive hydrocarbons
within the RADM2 chemical mechanism are switched off.
2 The model system and the data base
For this study the non-hydrostatic mesoscale model system KAMM (KAMM = Karlsruhe
Atmospheric Mesoscale Model) together with the gas phase mechanism RADM2 (Stockwell
et al., 1990) is used. The model system is driven by a basic state which is derived from the
larger scale observations. It is documented in more detail in Adrian and Fiedler (1991), Vogel
et al., (1995) and Fiedler (1993). As a result the model provides all important meteorological
variables and the concentrations of chemical species for episodes of several days.
Emission data of the most important anthropogenic emissions have been compiled in hourly
time steps and for an area of 177 km x 177 km with a horizontal resolution of 3 km.
Additional data like terrain height, land use and soil data have also been provided as close to
reality as possible. Those data are especially important for the parameterization of the
turbulent fluxes of momentum, energy, water vapour but also for the parameterization of the
emission and the dry deposition of chemical species at the surface. In addition with the
temperature and the photo synthetic active radiation calculated by KAMM the natural
emissions of nitrogen oxides from soil surfaces and of hydrocarbons from vegetation are
determined online by the model system.
3 Results
A situation for south-west Germany is selected, where high ozone concentrations have been
observed. It was accompanied by high air temperatures (-34 °C) and therefore enhanced
biogenic emissions. For the simulations with the model system a day at the beginning of
August 1990 has been chosen. On that day ozone concentrations have been observed up to
about 100 ppb in the early afternoon. Winds were rather weak and showed a dominant easterly
component in most of the of the area.
The dominant emission area for the anthropogenic emissions is within and in the surroundings
of the city of Stuttgart (Fig.l) which is located in the centre of the model domain. Fig. 2
shows the simulated ozone (O3) concentration for the reference case at about 18 m above
5-30
-------
ground. Areas of maximum ozone concentration appear in the west and south-west of the city
and approximately 30 to 40 km downstream. The highest concentrations reached are at about
100 ppb. They are comparable to the observations.
An extreme situation is a case where no anthropogenic emissions would be available.
Therefore only natural emission from soil and from biogenic source have been included. In
this case the maximum ozone concentration is around 60 ppb. This value gives the lowest
level which can be achieved by abatement strategies. The reductions in ozone at the same time
as for the reference case (Fig. 2) are given in Fig. 3. Maximum reduction areas are in the
range of 50 km downstream of the city complex.
The more realistic procedure would be the reduction of the most reactive species. In order to
estimate the level of reduction by eliminating the four most reactive groups of the
hydrocarbons, a scenario has been used, where emissions of propene, butene, toluene, and
xylene were switched off. Compared to the total anthropogenic emissions for the whole model
domain, about 20 to 25 percent of the hydrocarbons have been extracted by this procedure.
The ozone reduction in this case is shown in Fig. 4, where again the difference in ozone
concentration compared to the reference case (Fig. 2) is presented. Significant reductions in
, ozone concentration are confined to rather small areas in the west of the city of Stuttgart and
to the south-west of Heilbronn. Therefore Fig. 4 gives also those areas where the ozone
production is limited by the availability of hydrocarbons.
5 Conclusions '»
Numerical models are capable to study the effects of emission reductions but only according
to the state of the art of the understanding physical and chemical processes in the atmosphere.
However they provide the possibility to study quite realistic cases comparable to observations.
It is therefore possible to quantify the effects of different abatement strategies.
For the episode and the area of interest we focused on the ozone level is decreased by 30 to
50 % when all anthropogenic emissions are switched off.
When the emissions of the four most reactive groups of hydrocarbons are switched off the
ozone reduction reaches only about 15 % and the reduction is also confined to rather small
areas compared to the case with total reduction of anthropogenic emissions.
5 References
Adrian, G., F. Fiedler (1991): Simulation of unstationary wind and temperature fields over complex terrain and
comparison with observations - Contr. Phys. Atmos., 64, 27-48.
Fiedler, F. (1993): Development of meteorological computer models
Interdisciplinary Science Reviews, 18, 192-198.
Stockwell B.W., P. Middleton, J.S. Chang, X. Tang (1990): The second generation regional acid deposition
model chemical mechanism for regional air quality modeling, ]. Geophys. Res., 95, 16343-16368.
5-31
-------
Vosel. B., F Fiedler, H. Vogel (1995): Influence of topography and biogenic volatile organic compounds
emission in the state of Baden-Wuerttemberg on ozone concentrations during episodes of high air temperatures,
J. Geophys. Res., 100, 22907-22928.
75 100 125 150 175
> IOD.O
so.o - ioa.0
es.c - 6
£0.0 - 2G.O
1.5.0 - 20.0
10.0 - 15.0
7,5 - 10,0!
s.c - 7.5!
3,3 - 6,0
I I L.O - E,6
I I < l.o'
Figure 1: Horizontal distribution of the anthropogenic VOC emissions at 0800 CEST.
25
50
75 100 125 150
x in km
175
Figure 2: Horizontal distribution of ozone at 1400 CEST (18 m above ground).
5-32
-------
25
50 75 100
Y in km
125
150
Figure 3: Simulated ozone concentrations without anthropogenic emissions, minus simulated ozone
concentration with all emissions at 1400 CEST (18m above ground).
175
14:00 CIST
40, in ppb
HI > -I.E.
Hi -30- -1.6
|H -46 - -a.0
B-8 0 - -*.&
-T.5 - -5.0
-9.0 - -7.5
-10.6 - -B.O
-13.0 10.&
-13.6 - -1S.O
I I -16.0 13.&
I | <; -16,0
75 100
x in km
125
150
175
Figure 4: Simulated ozone concentrations without anthropogenic emissions of propene, butene, toluene, and
xylene, minus simulated ozone concentration with all emissions at 1400 CEST (18 m above eround).
5-33
-------
European Studies on the Photooxidation Mechanisms of Aromatic
Hydrocarbons and Oxygenates: Reactivity Implications
I. Barnes, K.H. Becker, B. Klotz and H. Geiger
Physikalische Chemie/FB 9, Bergische Unix ersitat - Gesamthochschulke Wuppertal.
GauBstraBe 20, D-42097 Wuppertal, Germany
Within the framework of the German Tropospheric Research Programme (TFS) in Germany
and the Chemical Mechanisms Development (CMD) subproject of EUROTRAC 2 research is
currently in progress in Europe to elucidate the photooxidation mechanisms of VOCs, in
particular aromatic hydrocarbons and oxygenates (e.g. dicarbonyls, ethers and acetals). The
aim of these efforts is the development of chemical mechanisms for inclusion in CT models to
better predict photoooxidation formation.
Since model calculations indicate that aromatic hydrocarbons mainly BTX (benzene, toluene
and the xylene isomers) can contribute up to as much as 40% to the formation of O3 and other
photooxidants in urban areas over Europe (Derwent el a!., 1996, 1998) substantial efforts
have been expended on investigating aromatic hydrocarbon oxidation mechanisms. However,
the accuracy of these predictions depends on the mechanism incorporated into the model and
thus model validation is required. The types of work performed embrace:
investigations on the atmospheric chemistry of benzene oxide and toluene oxide, possible
primary oxidation products for benzene and toluene, respectively (Klotz et al. 1997,1998).
detailed product and kinetic studies including verification/identification of carbonyl
products by GC/MS detection by their O-(2.3,4,5,6-pentafluroobenzyl)-hydroxylamine
(PFBHA) derivatives (Yu et al.. 1997; Kwok et al.. 1997).
measurement of photolysis frequencies of glyoxal. methylglyoxyl. biacetyl, butenedial and
3-he\ene-2.5-dione.
In the area of ethers and acetals (diethers) detailed chemical mechanisms are being developed.
Particular attention has been given to diethers and cyclic ethers, which in Europe are under
discussion for the use as fuel additives or alternative solvents. The developed chemical
mechanisms are tested against experimental data obtained in the outdoor EUPHORE
photoreactor in Valencia, Spain as well as in several indoor photoreactors. In collaboration
with the Fraunhofer Institute, Garmisch-Partenkirchen, FRG, the influence of these
oxygenates on tropospheric ozone formation is estimated by integration of the obtained
chemical degradation schemes into the ozone prognosis model RACM (Stockwell et al. 1997)
and application of the model to well defined scenarios.
The efforts in the area of aromatic hydrocarbon and oxygenates research within Europe will be
briefly summarised and the results discussed in terms of ranking these VOC classes with
regard to their reactivity.
References
Derwent, R.G., M.E. Jenkin, S.M. Saunders, Atmos. Em-iron. 30 (1996) 181-199.
Derwent, R.G., M.E. Jenkin, S.M. Saunders and M.J. Pilling Atmos. Environ. 32 (1998) in press.
Klotz, B., 1. Barnes, K.H. Becker, B.T. Golding, J. Chem. Soc. Faraday Trans,, 93 (1997) 1507-1516.
Klotz, B., 1. Barnes, K.H. Becker, accepted Chem. Phys., (1998).
Kwok, E.S.C., S.M. Aschmann, R. Atkinson and J. Arey, J. Chem. Soc., Faraday Trans. 93 (1997) 2847-
Stockwell, W.R., F. Kirchner, M. Kuhn and S. Seefeld, J. Geophys. Res. 102 (1997) 25847-25879.
Yu, J., H.E. Jeffries, K.G. Sexton, Atmos. Environ. 31 (1997) 2261-2280; Yu, J., H.E. Jeffries, Atmos. Environ.
31 (1997)2281-2287
5-34
-------
Experimental (symbols) and simulated (lines) concentration-time profile
in a toluene/NO photooxidation experiment
(730 ppb toluene /115 ppb NO).
O): toluene, left scale; (*): benzaldehyde; (): o-cresol; (T): m-cresol;
(A): p-cresol, right scale.
1.8 10
1.3 10°
12:00 13:00 14:00 15:00
time of day [hh:mm]
16:00
1200 1300 1400 1500 1600
time of day [hh-mm]
II 20 1200 12 40 13 20 14 00
limt of day [hh:mm)
I 00 II 20 II 40 12.00 12 20
lime of day (hhtmm)
10 JO 1045 II 00 II 15 11.30 II 45
limr of day [hhrmm)
000 S-
Bcrgische Universitat Wuppertal
Universitat Heidelberg
LT3 - Projekt D.I / D.2
Bergische Universitat Wuppertal
LT3 - Projekt D.I
-------
UJ
os
Untersuchung an ungesattigtenl,4-Dicarbonylen
im Europaischcn Photoreaktor
J(Dicarbonyl)
J(N02)
(40°N. Mittag. I Juli)
Lebensdauer T.
Hauplprodukte:
3H-2-Furanon
5-Methyl-3H-2-furanon
Maleinsa urea nhyd rid
Z-Butendial
l,62x 10' s"
9x 10 V
10 Minuten
30,1
3 1,0 %
I
k(OH)
(Bierbach et al. 1994)
Lebensdauer TO,,
[OH]= 1,6-10'cm'
[OH1 = 3,5 10'cm'
5,2 10" cm' s1
200 min
9 min
Z-4-Oxo-2-pentcnal
1.98 x 10's'
9x 10 V
8,4 Minulen
31,9 %
13.6 V
5.6-10" cm' s '
186 min
8 min
cis-butcncdial: product yields
2-3H-furanone
maleic anhydride
glyoxal
acrolein
carbon monoxide
(%perC)
sum of products
Photolysis
(Valencia)
30%
31 /.
6V.
3 /,
<3%
70-73 %
OH radicals
(CH3ONO/NO)
< 2 %
40-50 %
15-20%
observed
< 13 %
55-85 %
OH radicals
(H2O2)
not observed
6-12 %
20-45 %
not observed
(<27 %)
26-60 %
Bcrgische Uiiivcrsitat Wuppertal
LT3 - Projekt D.I
-------
cis-butenedial
U)
o-.. -"-
3H-2-funnon
o o
o o
O,/NO
Oi-U
Milfic (n
j\ *
The numben we product yields obiam«t in photolysis
c.penmenu made in EUPHORE in Aprile I997
O,/NO
C=0
Acroleln
5 %
o o
O,
NO
O,
NO
Abhangigkeit der Diacetyl-Photolysefrequenz von J(NO2)
V
T
C
,*
I
4.
c.
4,5
3.5
3
2.5
->
1.5
03
Q 1 -
0,5
0 -
C
X
x^
_/*
S
2 4 6 8 10 12 14
,T(N02)[10-Js-']
2»0 Bergische Universitat Wuppertal
"jfffff LT3 - Projekt D.I
GlTO.ml
C V,
-------
QUESTION/DISCUSSION SUMMARIES
-------
PRESENTATION SUMMARIES
Session I Summary of Questions and Discussions
Current EPA Regulatory Viewpoint on Reactivity
Bill Johnson, Ozone Policy and Strategies Group, EPA
Mr. Johnson was asked how to get a copy of Rule 66 and vapor pressure cutoffs. He
responded that inquiries on this rule, now referred to as Rule 442, can be made through Los
Angeles County (i.e., South Coast Air Quality Management District). Mr. Johnson was also
asked about obtaining a list of the 14 pending petitions. Mr. Johnson said that he will make this
list available.
William Carter (University of California) commented that this workshop group should
look at other scales in addition to the maximum incremental reactivity (MIR) scale. He added
that for compounds with low vapor pressure, the equilibrium vapor pressure should be
considered.
Robert Hamilton (Amway Corporation) asked about the mole versus mass based
reactivity and what the advantage would be if the standards are weight based. Mr. Johnson
commented that reactivity comparisons made on the mole basis would be more scientifically
sound. The regulation of emissions is still done on a weight-basis. A question was asked
concerning the basis for exempting compounds and what goal was trying to be accomplished.
Mr. Johnson responded that the goal is ozone reduction, but that other collateral effects are being
considered as well.
Brian Keen (Union Carbide) commented about the wide range of compound
concentrations and reactivities, and about the pit falls of using bright line cutoffs. Mr. Johnson
responded by stating that this is one of the issues he hopes will be discussed at this workshop. At
the present time, bright line cutoffs work well from a regulatory perspective. Dr. Keen also
commented on the problem of the uncertainty of the information.
Current EPA Research Viewpoint on Reactivity
Basil Dimitriades, National Exposure Research Laboratory, EPA
Alan Hansen (EPRI) asked about the difference between relative versus absolute
reactivity. He stated that the question of gram-based versus mole-based reactivity shouldn't be
important since the two are related by a proportionality constant. Dr. Dimitriades responded by
stating that the problem depends on how the data are used. The differences between cases
depends on the applications or how the material is used. For example, comparing paint solvents
with ethane on a per-gram basis is affected by the problem; but intercomparing solvents on a per-
gallon basis is not. John Festa (American Forest and Paper Association) asked if Dr. Dimitriades
had stated that there was no basis for excluding any VOC. Dr. Dimitriades responded that he
was referring to an exclusion from the inventory requirements. Another question was asked
about the distinction made between VOCs and non-VOCs. Dr. Dimitriades and William Carter
6-1
-------
(University of California) clarified this issue. The distinction is a legal or regulatory one; EPA
has defined compounds as being VOCs or non-VOCs and this information is provided by means
of a table.
Dr. Carter also commented that there are several issues that will drive the science. The
first issue is what type of policy is going to be used. If the policy continues to be the exemption
policy, then the question is whether ethane is the appropriate dividing line and, if not, what
substance should be used. One area of research is to determine what is the best dividing line or
bright line. The second issue, once the bright line has been determined, is to determine where
other compounds fall relative to the bright line. Dr. Carter commented that the most appropriate
comparison is on a mass basis, because this is how the VOCs are emitted. Dr. Dimitriades
responded by stating that the question of which basis to use is one that needs further discussion.
Dave Morgort (Eastman Kodak Company) asked if MOIR has been considered over MIR
as the basis for a reactivity scale. Dr. Carter commented that exemption decisions were based on
a number of reactivity scales (e.g. EBIR or MOIR) and not on just the MIR scale.
Dr. Dimitriades commented that the discussion on the reactivity scales is still an open one.
Jake Hales (ENVAIR) suggested that this workshop group try to develop a good definition for
the term reactivity.
California's Hydrocarbon Reactivity Program
Randy Pasek, California Air Resources Board
Can Roque (Naval Aviation Depot) asked if California was integrating their research
with those from other states. She continued by asking if there would be a shift from air pollution
to water pollution when changing over to material using water-based chemistry. Dr. Pasek
responded by stating that they are seeking opportunities to share studies. He also stated that
California doesn't plan to shift from air to water pollution. An addition, a comment was made
concerning the need to recognize instances where multimedia consideration must be made.
Bemie Zysman (Occidental Chemical Corporation) asked about other research facilities
doing work on MIR. William Carter (University of California) stated that he was not aware of
others doing reactivity research, except Harvey Jeffries's group (University of North Carolina).
Mr. Zysman add that he would like to see more research and development to clarify Dr. Carter's
work.
Bob Kozak (Atlantic Biomass Conversions, Inc.) asked if CARB was considering in-use
vehicle speciation testing for possible upgrading of smog check equipment. Dr. Pasek stated that
these types of measurements are not being done on a routine basis, but there is a need for this
type of information. Mr. Kozak also asked if tropospheric ozone production work might
interfere with any stratospheric ozone depletion work being done by CARB. Dr. Pasek stated
that they have taken this into account by considering the effects compounds might have on
stratospheric ozone and global warming, as well as its toxicity.
6-2
-------
VOC Reactivity - Beyond Ozone
Alan Hansen, EPRI
Dave Golden (SRI/Stanford) stated that no one knows the chemical mechanism of the
SO2 to sulfuric acid reaction. There is a lot of the chemistry in the models that may not be
correct, but care must be taken when jumping to the complexity of the model. Mr. Hansen
agreed with these comments.
Jake Hales (ENVAIR) asked about the counter-intuitive findings that result when you
increase NOX, the nitrate aerosol decreases. He asked if there was a mechanistic explanation for
this occurrence. Mr. Hansen responded that he didn't know the mechanistic reason why this
occurred. Ted Russell (Georgia Institute of Technology) commented that since you're decreasing
N02, you're increasing the OH radical pool. This allows for faster oxidation of VOCs, the ozone
concentration increases, and, with more ozone, there is more OH. The presence of more ozone
also causes faster nighttime conversion of NO2 to nitric acid.
Session II Summary of Questions and Discussions
Comparison ofPOCP and MIR Scales
Richard Derwent, Meteorological Office, United Kingdom
Eduardo Olaguer (The Dow Chemical Company) asked if by moving from the MIR scale
to a regional reactivity scale, there would be as much of an advantage in moving from highly
reactive to moderately reactive compounds. Dr. Derwent responded by stating that the best
benefit is from changing from a high reactive compound to a low reactive compound. He stated
that the MIR scale underestimates the impact of controlling the middle reactive compounds.
Dr. Olaguer also asked if there would be any significant impact on reactivity values if fast
vertical motions associated with convective activity, such as precedes thunderstorms, were
accounted for in models used to derive reactivity scales. Dr. Derwent stated that this has not
been considered, because ozone is not produced in thunderstorms.
Donna Carvalho (Pennzoil) asked if Dr. Derwent had made any assumptions about the
reactivity of C-13 or greater compounds. Dr. Derwent responded that he had not. These
compound were not included in his calculations, because they are not included in the emission
inventory.
Dr. Derwent was asked about why formaldehyde was not photolyzed. He responded that
it was, but whether or not to photolyze formaldehyde is not a user's choice, but is determined by
the hydroxyl radical concentration. In Dr. Carter's MIR scale, because of the high NO* levels
used, much of the formaldehyde is photolysed and so it appears to be highly reactive. In the
regional POCP scale, because NOX levels are lower, much of the formaldehyde reacts with
hydroxyl radicals and its reactivity appears lower.
6-3
-------
Session III Summary of Questions and Discussions
Dunn-Edwards Proposed NARSTO Research on Ozone Formation Potential ofVOC Emissions
from Architectural Coatings
Edward Edwards, Dunn-Edwards Corporation
Jeff Gaffney (Argonne National Laboratory) commented that by investigating lowNOx
chemistry, a great deal can be learned about regional scale reactivity. He also commented that on
a regional scale, consideration must be given to the formation of species other than ozone, such
as organic peroxides and the conversion of SO2 to sulfate. These species have impacts on other
environmental concerns such as regional haze.
Harvey Jeffries (University of North Carolina) commented on the importance of reactivity
in low NOX conditions. In low NOX environments, the VOCs determine the fate of the NOX; the
loss of NOX limits the formation of ozone. These effects can cause increases in ozone downwind
of urban areas. As illustrated by an SAI study, changes in solvent content from a more reactive
solvent based on toluene to a less reactive solvent based on paraffins can cause increases in
ozone concentrations in a downwind NOx-limited environment. Dr. Jeffries concluded by stating
that some of Dr. Edwards' concerns are legitimate and the issues of nitrate yields and nitrate
formation in these mechanisms are important pieces of information.
William Carter (University of California) commented about his chamber study that was
used to evaluate the isoprene mechanism. The changes made to the model were based on better
fundamental chemistry, but the predictive capability of the model was not evaluated. Dr. Carter
continued by stating that the current regional models are not designed to handle organic reactions
under low NOX conditions, with the exception of the RADM mechanism. He stated that the data
have not been adequate to evaluate these models and to gather the necessary data would require
the use of advanced analytical equipment to analyzes for other species.
Edward Edwards (Dunn-Edwards Corporation) added a comment about the need to do
full circle analysis: model, chamber, and ambient air.
CMA Research Initiatives
Jonathan Kurland, Union Carbide Corporation
Following Dr. Kurland's presentation, William Carter (University of California)
discussed the different factors that affect reactivity and the ways to evaluate whether a model can
adequately predict them all. In systems with VOC reacting in the presence of excess nitrous acid
and where the nitrous acid is forming all of the OH radicals, the amount of ozone or NO formed
is very sensitive to how fast the VOC reacts and the NO is converted to NO2. There is almost no
sensitivity to radical initiation or inhibition effects. This provides a way of testing that aspect of
the model independent of the other uncertainties. This method could potentially be used for very
low volatility compounds that cannot be done practically in environmental chambers. Dr. Carter
commented on his new program with the California Air Resources Board to develop more
generally applicable methods of measuring reactivity to replace the more expensive chamber
experiments.
6-4
-------
CSMA Position on the Importance of Relative Reactivity
Doug Fratz, Chemical Specialties Manufacturers Association
Harvey Jeffries (University of North Carolina) commented that the ozone formation in an
urban area cannot be determined by multiplying an inventory by the MIR; a whole airshed model
has to be used. He reiterated that nearly half of the ozone is produced by low reactivity
chemicals. Mr. Fratz responded by stating that they found very good agreement between the
MIR-weighted inventory approach and the urban airshed model approach. Both approaches
predicted very small amounts of ozone formed considering the VOC controls targeted. Both the
MIR-weighted approach and the modeling approach were being used to assess the effects of
marginal changes in VOC emissions on marginal ozone formation. William Carter (University
of California) commented that he agreed with the way Mr. Fratz had used the MIR to look at how
to prioritize controls and not to look at what caused ozone formation.
Reactivity Concerns
Phil Ostrowski, Occidental Chemical Corporation
John Festa (American Forest and Paper Association) commented that the branch of EPA
administering TRI requires the air program to declare a chemical as a negligibly reactive VOC
before it can be removed from TRI. Mr. Ostrowski responded that maybe there would be a
rethinking of the TRI exemption in the new policy.
Bob Avery (Eastman Chemical Company) commented that the cutoff for exemptions
should be raised in the short term, rather than lowered, in order to get more useful chemicals in
the 'tool box.' Mr. Ostrowski expressed his concern that this approach may end up impacting air
quality in a negative way. Mr. Avery continued by stating that there may be some local adverse
impacts, but overall, the substitution of low reactivity compounds would be directionally correct,
and, overall, an improvement.
Categorization of Low Reactivity Compounds
John Owens, 3M Company
William Carter (University of California) commented that there are procedures to
estimate upper limits of reactivity that could be used to establish the cutoff.
Harvey Jeffries (University of North Carolina) stated that, although compounds with a
low kOH must be looked at closely, he is not calling for compounds with lifetimes on the order
of months or years being studies in the smog chamber.
Richard Derwent (Meteorological Office, United Kingdom) stated that this is not so
simple for theoretical studies. When working on CFC replacements, they looked at long-lived
VOCs. It is important to have information about the degradation products and to consider the
by-products from the production of these low reactive VOCs. This will require more than just a
theoretical study, because there are a whole range of other problems. Mr. Owens responded by
stating that 3M does sponsor these studies of degradation products and does a life cycle analysis
of byproducts and products.
6-5
-------
Impact of a Molar Ethane Standard on the Number and Type of VOC-Exemptible Compounds:
Practical and Environmental Implications
Daniel Pourreau, ARCO Chemical Company
Harvey Jeffries (University of North Carolina) proposed a deferred control system of five
years with possible renewal based on further analysis. Dr. Pourreau responded by stating that
industry would be reluctant to use an alternative that might not be available in the future. He
proposed that the new reactivity-based policy revisit the exemptions, but he thinks the
exemptions granted now for the low-reactivity materials would still benefit under a reactivity
based policy. Dr. Jeffries continued by commenting that it would be necessary to inventory the
exempt products. He stated that it is necessary to know where and when and in what quantities
the VOCs are emitted to know the impact on the environment. Dr. Pourreau stated that, although
they are trying to replace highly reactive compounds to reduce ozone, there are practical
limitations on how they can do this.
Jim Berry (Berry Environmental) commented on a specific slide referring to 37% solids
from water-based solvents. He stated that for the comparison to be appropriate, the water must
be removed. Dr. Pourreau stated that the information was in pounds/ VOC and pounds/solids
and, therefore, water was not an issue. The weight percent solids was included as an illustration
to show that the non-water coating applies more solids per application.
Session IV Summary of Questions and Discussions
Computing Volatile Organic Compound Reactivities with a 3-D AQM
Zion Wang, University of North Carolina at Chapel Hill
Jake Hales (ENVAIR) commented that there are many choices: use kOH, use sensitivity
coefficient, use MIR or MOIR, etc or use 3-D models over EKMA. It needs to be remembered
that MIR isn't all that simple; even with the one-dimensional EKMA model, many parameters
need to be specified such as the different ways to titrate the NOX, how peak ozone is used in the
numerator of the equation, and what domain is used. Dr. Hales concluded by soliciting
comments from the 3-D modelers.
Ted Russell (Georgia Institute of Technology) commented that many of these are policy
questions. It needs to be asked: what is the metric for the policy makers. Once that decision is
made, then the modelers can react. There is a need to get away from boundary conditions and
initial conditions and to look at how these emissions are added. When looking at the impact of
change in fuel composition, the emissions change needs to linked to the fuel. For a solvent, the
solvent emission distribution used would be different. It is necessary to normalize the results to a
mixture of compounds and different people might have different ideas on this normalization
process. It will also be necessary to compare the 3-D models results to the box model results.
Gary Whitten (Systems Applications International, Inc.) agreed with Dr. Russell's
comments about policy. A policy decision is needed. Currently, xylene and ethane are treated
equally, but clearly they are not. He commented that there needs to be a methodology for trading
6-6
-------
solvents with lower reactivities. He proposed a 4-tier reactivity scheme in place of the 1-tier
scheme currently in place. Compounds in the highest tier (e.g. xylene) would be counted twice
as much; lower reactive compounds would be counted half as much. This would encourage
switching from solvents like xylene to paraffins and would be beneficial even if the alternative
was still somewhat reactive. In this way, industry can be a big credit for making the switch and
for making improvements in the reactivity in the atmosphere. Dr. Russell responded to this
proposal by stating that more bright lines don't help, especially there aren't huge gaps in the
reactivity spectrum where these different classes would divide. Dr. Russell believes that this is
defeating the purpose of reactivity and he would prefer the policy to be driven by the best
science.
William Carter (University of California) agreed that the policy will drive the approach.
He continued by stating that there are two main ways of looking at policy: (1) reactivity scale
which is generalized and (2) assessment or specific substitution scenario where replacement
should be handled in detail with a detailed model. There is also a need for a generalized scale to
be used for prioritizing, but the reactivity scale poses different modeling problems in trying to
answer the question: what is the optimum scale to use. The scale would need to be representative
of the criteria and to represent a distribution of conditions (set of scenarios). Lastly, Dr. Carter is
not convinced that an EKMA model is not adequate for the purposes of developing a scale.
Chris Walcek (State University of NY at Albany) commented that, because the scale can't
be made to be a simple one, this doesn't mean that it is impossible to have a scale. The scale
won't be simple and there will be a great deal of controversy about its development. Dr. Carter
added that it is important to have a standardized protocol for the scale; it is important not to vary
the metrics.
Barbara Francis (CMA) commented that CMA has not yet developed positions on the
issues raised in yesterday's or today's discussions and noted that the positions expressed here by
CMA member companies are company specific and not necessarily industry consensus positions.
She continued by saying that CMA is conducting research on specific chemicals and that CMA
believes that research programs should, wherever possible, be policy relevant.
Temperature Dependence of Ozone Chemilwninescent Reactions with Organics: Potential
Screening Method for VOC Reactivities
Jeffrey Gaffney, Argonne National Laboratory
Chris Walcek (State University of NY at Albany) asked about the cost of the
chemiluminescence system. Dr. Gaffney stated that the system costs between $6,000 and $7,000
which is cheaper than the NOX system and it could also be used to look at reduced sulfur gases.
VOC Receptor Modeling as an Aid to Evaluating the Effect of Reactivity Changes on Ozone
Formation
Donna Kenski, US Environmental Protection Agency
Alan Hansen (EPRI) asked about the reconstruction of the emissions mix at the
monitoring stations and the complicating factor of emissions in between the monitoring sites.
6-7
-------
Ms. Kenski responded by stating that the model a very simple, screening-level model and does
not include any reinforcement or deposition. Mr. Hansen continued by asking about the lack of
impact from refineries being an artifact of the location of intermediate sources. Ms. Kenski
answered that they tried to account for this issue with the trajectory, and by comparing the
upwind trajectories.
Computational Studies ofOxidant Reactions of Volatile Organic Compounds Relevant to the
Formation ofTropospheric Ozone
David Dixon, Pacific NW National Laboratory
Eduardo Olaguer (The Dow Chemical Company) asked all the presenters if neural
network techniques had been applied. Dr. Dixon responded that he was not sure if neural nets
are being used in the field of VOC reactivity. Neural nets are a data analysis tool and could
probably be used effectively to correlate a range of experimental measurements. Dr. Dixon was
not aware of any reasons why neural nets can not be used in this area and research on their use
and applicability would be appropriate.
It was asked if Dr. Dixon was going to look at the reaction between the hydroxyl radical
and NO2 and he answered affirmatively. Dave Golden (SRI) commented that the potential
energy surfaces for the hydroxyl radical and NQ reaction are inaccurate. There are two groups
that have estimated this, the IUPAC estimate is wrong and the JPL estimate is correct. It has
been measured to 5%, but could it be calculated better. Dr. Dixon responded by stating that
temperature and pressure corrections could be included in the calculations. A question was also
asked concerning the impact of water on this reaction. Dr. Dixon responded by stating that there
would be no effect from water, unless the reaction was taking place in a droplet, but because
these reactions are fundamental gas-phase processes, this would not be expected.
Session V Summary of Questions and Discussions
Oxygenates: Reactivity Implications
Ian Barnes, Bergische University at Wuppertal, Germany
Eduardo Olaguer (The Dow Chemical Company) commented that the ideal approach
would be to learn all that could be learned about reactivity options, then decide how to design the
reactivity strategy. It seems that currently, all the money goes into designing a policy index, and
the science is done to make that convenient. Dr. Olaguer would prefer to do the science first.
6-8
-------
FREE FORUM SUMMARY
Howard Feldman (Moderator) described this portion of the workshop as the opportunity
for any attendee to express their 'two cents worth.' There were many experts in the audience, not
all of whom made a presentation, and many others with opinions or points of view. The free
forum was their opportunity to make whatever presentation they wished to make.
S. Kent Hoekman (Chevron Products Company) began his comments by observing that
there had been talk about how desirable it was to separate policy and science issues. He
expressed his opinion that, in the case of reactivity, this is impossible, and that policy is actually
required to define reactivity. He continued by stating that science has been stymied by the lack
of policy or, at least, a broad policy outline. In California, there is a policy definition of
reactivity, but it's limited to the certification of new vehicles with respect to exhaust emissions.
It was not designed to result in an ozone benefit, but only to achieve an equivalent ozone impact.
Therefore, the absence of the automobile industry from this meeting is conspicuous, with the
exception of Honda. This is perhaps because the California system is complex and burdensome
and it does not achieve an air quality benefit. Therefore, automobile companies haven't taken
advantage of the opportunity to develop their own reactivity factors, but have relied on the
default values developed by CARB. Establishing the default values is a every expensive process;
CARB has spent hundreds of thousands of dollars to establish and maintain their ability to
determine reactivity adjustment factors.
When reactivity is applied to stationary sources and consumer products, there are some
simplifications and some additional complexities. The first simplification is the absence of
simultaneous emissions of VOC and NOX from stationary sources or consumer products. This is
very important when dealing with automobile exhaust. Another simplification is that, for
automobiles, the VOC's are changed during the combustion processes within the car. Therefore,
the fuel put into the car is not what is emitted in the exhaust. Thirdly, there is a great variability
between the vehicles, between operating conditions, and between technology classes whereas this
complication is not present for the stationary sources.
The first complication for stationary sources and consumer products over the mobile
sources is the great diversity of chemical structures and classes of solvents. In addition, the
atmospheric chemistry is not understood well. Secondly, it will be much more difficult to trade,
and trade fairly, for consumer products, because the materials being traded are very dissimilar.
For mobile sources, we are trading one gasoline emission mixture for another and these mixtures
are very much like one another. For consumer products and stationary sources, this will require
greater quantification and certainty in the reactivity of materials.
Dr. Hoekman continued by discussing two associated issues for which reactivity
arguments are being used, but that he believes must remain separate: (1) to reduce the ambient
ozone levels and (2) the exemption of specific VOCs. For the issue of exempting VOCs, the
focus is on establishing a bright line by looking at the least reactive VOCs. For the issue of
reducing the ambient ozone levels, the focus is on what materials should be controlled (VOCs,
which VOCs, or NOJ by looking at the most reactive VOCs. The attainment of the standard is a
broader, more urgent issue, but a VOC reactivity policy must be developed to encompass the full
spectrum of issues. It can also be asked whether or not there is any real justification for
exempting anything if it contributes to ozone.
6-9
-------
Dr. Hoekman summarized his remarks by stating that there are two main concerns for
industry. The first concern deals with all of the different decisions which must be made on
defining conditions, setting the scale, and measuring reactivity. There is a great deal of
arbitrariness in any policy, and although this is not good, instability is worse. It is very difficult
for industry to deal with changes in the rules. The second concern of industry is that reactivity
may become an attractive control measure, that appeals to different states (and districts) to
different extents. Dr. Hoekman encouraged EPA to be sure, when developing its policy, that it
can be applied generally and fairly to the whole country and not as a patchwork of different
requirements.
Anne Giesecke (American Bakers Association) discussed the concerns of the baking
industry. Yeast fermentation releases ethanol, which is released to the air. In addition, they
operate the third largest trucking fleet in the nation. Dr. Giesecke stated that this industry is
encouraged by these discussions. This industry has spent about $30,000,000 turning ethanol into
CO2 and increasing the NOX emissions through catalytic oxidation. The modeling discussions
were interesting and this work needs to move forward. Although access to the models is
important, Dr. Giesecke believes that not every state and industry needs to operate the model.
The resources of many states are strained as are many industries and she suggested that modeling
work could be out-sourced. Dr. Giesecke is encouraged by the work on relative or incremental
reactivity or the potential for ozone formation and the shifting of the emphasis away from the
'yes or no' system currently in place for evaluating VOC emissions. She expressed the need for
more sophisticated tools such as the baking industry's interest in the holistic or life cycle
approach. The baking industry also recommends looking at emissions trading and how to change
from the current system where all VOCs are considered equal to one where a more reactive VOC
can be replaced by a less reactive VOC. This industry would like to see more effort put into the
study of more reactive chemicals and those that have other complications. They would also like
to see more of their money going into fleet conversion instead of ethanol control. The baking
industry operates at a 2% profit margin and, therefore, they don't have a surplus of money to do
both.
Leslie Ritts (Hogan & Hartson), who represents a large number of stationary source
categories, commented that for 20 years billions of dollars have been spent on control strategies
and on decisions that have led to moving business offshore. She expressed her concern about
how the discussions held at this workshop will feed into regulations and whether or not there is a
time line for such regulations.
Barbara Francis (CMA) commented that the consensus among the solvents producers is
that the models are not really ready for the regulatory community to use. They believe that there
is enough information available now, especially on the incremental reactivity, to implement
reactivity-based regulations.
Neil Wheeler (MCNC-NC Supercomputing Center) began his comments by reminding
the audience that establishing a reactivity policy was not going to be easy and that the issue of
reactivity can not be simplified. It will take a great deal of work to apply the policy fairly and
effectively.
6-10
-------
Mr. Wheeler stressed the need for this to be a community effort and that people need to
participate and to share information. He commented that there is a wide range of possibilities of
how to use reactivity in the regulatory process: from setting exemptions with a bright-line (or
not) to reactivity-weighted emissions. He also commented that models do not make decisions.
There will be external information (scientific or policy) that will affect the decision making
process and consideration must be given to other environmental issues as well.
He expressed the need for multiple metrics to help the policy makers and the need for the
scientific community to clearly describe the meaning of these metrics. Consideration should be
given to the range of conditions from the current time out into the future. It is important to
monitor an exempt VOC both in the inventory and in the ambient environment into the future to
be certain that a poor decision has not been made. He agrees that stability is important in a
regulatory process but noted that we cannot have absolute stability, especially in light of possibly
making a bad decision. There needs to be a systematic process for dealing with necessary change
based on new information.
Mr. Wheeler discussed the range of metrics needed such as metrics to assess exposure,
various meteorological conditions, multiple locations, multiple pollutants, and total risk. He
emphasized the need to develop metrics using photochemical models with the best science
available. He feels that it may be possible to use simpler models, such as EKMA, for developing
a reactivity scale but they must first be evaluated against state-of-the-science models.
Roger Tanner (TVA) commented that the data shows clearly that ozone formation is NOX-
limited in some areas and is VOC-limited in others and that these conditions vary greatly by
location and by season, depending on the sources of VOCs and NOX. He believes that a metric is
needed that scales with the actual conditions within an airshed. In order to predict ozone
reduction in particular locations and at the times necessary get below the standard, a sliding
airshed-specific metric in needed. Dr. Tanner believes that this type of metric can be devised
with our current scientific knowledge.
Ken Schere (EPA/ORD) discussed the differences between developing assessment tools
and procedures and screening tools for implementation. Between these two ends of the
spectrum, there are many possibilities. The scientific tools are available to do a full scale
assessment to describe reactivity as a function of various environmental variables. Dr. Schere
commented that Dr. Russell showed that a sophisticated tool can be reduced down to a simple
screening tool. In order to decide where on the spectrum we need to be, it is necessary to know
what the policy makers need.
Randy Pasek (CARB) expressed his agreement with comments made previously about
California's reactivity regulation. The regulation is limited and complex. But California has
been moving ahead since the regulation was developed. From a regulatory perspective,
Dr. Pasek believes that the policy must be based on good science and that the complexity should
be understood. And it is very important that the policy be constant. He echoed Dr. Wheeler's
suggestion to build into the policy a process for change in order to make the policy more stable.
CARB has a policy based on the MIR scale, because it is a good complement to the NOX
controls. Lastly, Dr. Pasek expressed his agreement with the comment of Ms. Francis that the
science is adequate to develop regulations, as has been done in California.
6-11
-------
Dave Golden (SRI and Stanford University) believes that good models are needed to
understand any complex process. He continued by commenting that although the science has
advanced very far and the models are reasonable, there are a lot of things we don't know. He
warned that we shouldn't think that because we can model something that the models are
necessarily correct. He expressed his inability to believe in a model that predicts PM2 5. He
concluded by urging that the scientific work and the funding be continued
George Brown (National Aerosol Association) discussed the problems caused by the huge
variability in VOCs and the fact that the current regulations treat them as being equal. This
situation severely inhibits the use of certain substances and outlaws the use of others. As an
example, on January 1,1999, CARD will implement a rule that allows zero VOCs in one
product. Mr. Brown believed this is ludicrous. The National Aerosol Association (NAA) has
supported several reactivity studies over the last ten years and, based in these studies, they are
committed to the regulatory use of relative reactivity. The NAA realizes that some substances or
products will have difficulty operating under that system and they don't, as yet, have the
solutions to these problems. At the present time in California, the NAA believes that relative
reactivity ought to be used on an optional basis by sources having trouble meeting a mass-based
standard. Lastly, Mr. Brown believes we should stop thinking hi terms of cutting down the mass
of precursors of ozone and begin thinking in terms of ozone limitation itself.
Phil Ostrowski (Occidental Chemical Corporation) commented about the economic
benefits of reactivity regulations. Properly designed reactivity regulations should provide a tool
whereby small solvent users can avoid installing costly control equipment. This will provide
good environmental benefit at a low cost.
JeffGaffney (Argonne National Laboratory) echoed some of Dr. Golden's preceding
comments. He urged the group not to forget about VOC chemistry and not limit the discussion
to only the ozone-driven aspects. It should be remembered that VOCs form fine aerosols and
other products such as nitro-phenols that are very toxic and water soluble. Wet deposition, cloud
condensation, climate effect, and radiative properties (UV and IR absorption) of VOCs should
also be considered. He made a plea for the science and the sum-level support of the science. He
believes that by understanding the science better, the scientific community will be able to provide
a knowledge base to allow the policy markers and the modelers to do a better job.
Tun Lawrence (Georgia Pacific and American Paper and Forestry Association)
commented that, although the science has been fascinating, we must recognize the need for
pragmatic regulatory tool development in parallel with the existing scientific research. At this
time, EPA is moving forward with implementation plans for a national ambient air quality
standard for fine particulates, ozone, and regional haze; VOC and NOX are all listed for controls
under these programs. By November 2000,170 different industrial categories will receive
MACT standards. For organic HAPs, many of the MACT floors are being set based on existing
VOC controls. There is also implementation of the Kyoto Treaty that needs to be considered.
Mr. Lawrence feels that there need to be two parallel tracks: (1) one track moving forward very
quickly because of pending regulatory decision that have to be made and (2) the other track
continuing to understand the details of the science. Regulatory tools such as those described by
6-12
-------
Donna Kenski (EPA, Region 5) are important in the near term to assess the effectiveness of
proposed regulatory action.
Praveen Amar (NESCAUM) stressed the need to look at aerosol forming potential of
VOCs which might be exempted, in addition to the ozone forming potential. As an example,
between 33% and 40% of the fine particles in the Northeast are organics. Consideration should
also be given to wet deposition, dry deposition, and toxicity.
Doug Fratz (Chemical Specialties Manufacturers Association) made comments
comparing the use of regional or local reactivity scales versus the use of a linear national scale.
There are a number of control types that could be controlled locally. For other sources, such as
consumer products, it would not be practical to have a different product for every local region.
Therefore, to use reactivity, there must be a single national scale based on a single metric.
Mr. Fratz also commented on the funding of research as compared to the cost of controls. He
feels that research is still under-funded and that the ratio of money spent on controls to money
spent on research is far to high.
Jeffrey Holmstead (Latham & Watkins) stated that in the regulatory arena, we operate in a
legal framework that doesn't offer much flexibility. He hopes that in the long term, the
framework can be changed to allow flexibility. But in the near term, there are things that can be
done. Mr. Holmstead reminded that group that they already have a reactivity scale where
everything is either a 0 or a 1, and he believed that it would be hard for this group to do worse
than that. In the near term, he believes that the current knowledge can be used to help people
move from using highly reactive compounds to using lower reactive compounds. In the long
term, the scientific research needs to continue. Mr. Holmstead urged the group not to let the
perfect be the enemy of the good in the short term.
William Carter (University of California) wanted to echo what Mr. Holmstead said about
the near term problems. There are three approaches to using reactivity: the two currently used
reactivity scales (the binary national scale and the California MIR scale) and airshed/scenario
specific assessment. The scientific basis of the ethane exemption standard is not good. This is
why the EPA has frozen the exemption petitions, but these can't stay frozen for long. Dr. Carter
suggested that in the near term the EPA do a modeling assessment to recertify the ethane
standard or identify another standard. This would then allow the present method to continue
while alternative methods are being developed. In California, the regulations they are developing
require a scale which, if nothing else is put forward, is going to be the MIR scale. With the
demand for stability, if the MIR scale is implemented (later in the year), it will be difficult to
change. Therefore, this is the time to provide an alternative. Dr. Carter strongly emphasized the
urgency of this problem.
Bob Avery (Eastman Chemical Company) recommended two tracks be followed in the
future. Although the modelers are improving our understanding, the necessary results will not be
available for three to five years, optimistically. A more realistic time frame is between five and
ten years. The current system is bad and, therefore, a better interim system is needed for the next
two to five years. Mr. Avery suggested that a few dozen individuals should be able to sit down
in a room and improve on the current binary policy.
6-13
-------
Cyril Durrenburger (Texas Natural Resource Conservation Commission) discussed the
Control Measure Catalog developed for the last SIP submitted by Texas. It is a metric that ranks
the VOC controls based on the tons reduced, toxicity, and the reactivity using Dr. Carter's MIR
scale.
Robert Wendoll (Dunn- Edwards Corporation) discussed two approaches to the
policy/science issue first raised by Mr. Hokeman. The scaling approach is used to rank VOCs by
some reactivity metric for regulatory purposes. There are policy decisions that have to be made
and some of these will be arbitrary, by necessity. Other decisions would be arbitrary only
because of the lack of scientific information. The complementary approach is the systemic
approach which is the study of actual emissions into ambient air to determine their potential to
contribute to high ozone levels across the full range of environmental conditions. Mr. Wendoll
believes that the systemic approach is important because it is the area of research that will
develop the policy-relative science or the science that links actions with outcomes. This allows
the policy makers to choose action intelligently. Mr. Wendoll believes that both the scaling
approach and the systemic approach should be pursued simultaneously. The policy makers must
realize that both of these approaches together are embedded in the total ecological impacts.
There will be other factors that affect the regulatory decisions being made, such as the impact of
the regulations, including material resource and energy consumption, waste disposal, and water
quality. Mr. Wendoll echoed the comment made by Mr. Fratz about the ratio of research
spending to compliance spending. He believed that the amount of money that needs to be spent
on the research necessary to provide a better basis for sound regulations is minuscule compared
to the cost of compliance. Lastly, Mr. Wendoll reminded the group that control costs also have
indirect health impacts by reducing available income.
Dave Morgott (Eastman Kodak Company) commented that in order to assess the intrinsic
impact of VOCs on the environment, the more appropriate scale is the MOIR scale and not the
MIR scale. The MOIR scale provides information about the peak amount of ozone that can be
formed when a quantity of VOC is introduced into the environment.
6-14
-------
POLICY AND SCIENCE QUESTIONS SUMMARIES
Howard Feldman (Moderator) introduced this segment of the workshop by asking the
group to consider what needs to be done next. A review of these policy and science questions
will be a form of guidance. This is a genesis workshop and a group will be formed as a result of
this workshop and will continue these discussions. Mr. Feldman reviewed the policy questions
listed below and asked the group if these are the right policy questions. He also asked the group
if there are other questions that should be added to this list.
POLICY QUESTIONS
1. How should reactivity policy account for interaction with other air-pollution problems?
2. How should the reactivity concept be used to both maximize environmental benefit and
encourage environmentally superior product development? "Exemption" vs.
"substitution"?
3. What is the maximum uncertainty level that can be tolerated for reactivity-related
decision making?
4. Exemption policies:
° Bright line vs. bright band?
o Environmental cofactors?
o Where, ethane?
° Molar or mass basis?
5. Exemption protocols:
o Cookbook?
o Maintenance and tracking?
° Automatic testing criteria?
6. Procedures for modifying exemption criteria? Grandfathering? Grace periods?
7. Substitution protocols and guidelines?
8. How should vapor pressure be incorporated into the decision process?
Bob Avery (Eastman Chemical Company) raised the issue of national standards versus
regional standards, and he believes that this question needs to be added to the list.
Edward Edwards (Dunn-Edwards Corporation) was concerned about availability and
whether availability will be considered when making policy decisions. Availability is the issue
of the difference between what's in the can and what's in the air. He was also concerned about
looking at the time domain used to assess the ozone impact. For example, a paint may last from
between two to fifteen years, depending on how it is formulated. The question is will the ozone
impact be assessed only for the one-time application or for the full life cycle of the product.
These are both very critical policy decisions in analyzing how to- determine whether the VOC
life cycle is important for ozone forming potential.
Jonathan Kurland (Union Carbide) made comments on question 8 that deals with vapor
pressure. He stated that the issue of volatility is a subset of the general issue on how much of the
6-15
-------
content becomes emissions. There are other environmental non-evaporative fates (e.g. down the
drain). Dr. Kurland believes that this question should be expanded to include consideration of
whether the policy can properly determine the actual emissions released into the atmosphere as
opposed to other estimates such as gross sales.
John Durkee (Creative Enterprizes) commented that an additional policy question is
needed that deals with the issue of communication of the policy to the affected community.
Policy which can't be communicated may not be good policy. As the policy is developed,
consideration should be given to how the policy will be communicated, because, regardless of the
quality of the science, the end-user won't buy it if they don't understand it.
Howard Feldman (Moderator) reviewed the science questions listed below and asked the
group if these are the right science questions. He also asked the group if there are other questions
that should be added to this list or it any of these questions should be taken off the list.
SCIENCE QUESTIONS
1. Is a reactivity-based policy practical, feasible, and beneficial?
2. How do we best factor long-range transport into a reactivity-based strategy? How about
co-dependencies with PM?
3. What are the uncertainties of the various possible reactivity scales?
4. What are the advantages/disadvantages of using MODELS3 for estimating reactivity?
5. What are the merits of using speciated VOC reactivities, as opposed to lumped VOC
reactivities?
William Carter (University of California) commented that several of these questions are
actually policy questions. The questions on the practicality and feasibility of a reactivity-based
policy are policy issues. Although the question on the benefits of a reactivity policy is a
scientific issue, it would require a policy on how to measure the benefit. Also, the question on
using MODELS3 (question 4) has both a policy and a science component. Whether or not
photochemical grid models can be trusted is a science question, but the advantage or
disadvantage to using one is a policy question.
Anne Giesecke (American Bakers Association) commented that it is difficult to divide
science and policy issues. From an industry perspective, they are looking for regulatory
baselines: predictability, long-term planning, and basic functionality. Baking companies
currently work with a very high level of uncertainty on their emissions (+/- 30%). This is
acceptable, because it provides a regulatory baseline that tells them if they are in or out of an
EPA threshold for clear air requirements. For this reactivity-based policy, the same level of
science and policy would be acceptable, if it establishes a regulatory baseline.
Jason Ching (EPA) pointed out that a powerful system framework such as Models-3,
together with its Community Multiscale Air Quality (CMAQ) model provides a modeling
capability for estimating model reactivity. The Models-3/CMAQ system ability to provide a plug
and play capability allows the substitution or replacement of various science process options
including the means to modify (edit) existing chemical mechanisms or by exchanging and
6-16
-------
studying alternative chemical mechanisms. This capability and the other unique suite of
modeling capabilities and features in this system provide a powerful means to perform the
various model experiments and sensitivity tests needed to evaluate model response to changes in
reactivities. Further, because this is a publicly available system, it provides a means for the
entire community of interested parties to become involved and to contribute to the research,
development and assessment of reactivity models, and ultimately, to improved air quality
models.
Jake Hales (ENVAIR) commented that Question 4 was intended to be more generic than
just MODELS3. He suggested that MODELS3 be replace with 3-D or Eulerian models. Howard
Feldman (API) suggested that a Question 4a be added: how do you use one of these models to
determine a relative or incremental reactivity for substitution purposes. Other related questions
are: how many simulations are needed, how good do the databases need to be for a given urban
area to do these calculations on a national level, or, if a scale is used, was the modeling used to
develop the scale adequate.
Jim Vickery (EPA/ORD) would like to add a question about timing. He asked the
scientists what could they produce in the next two to three years that could help guide the policy.
This would allow the policy maker to decide if they should develop an interim policy or if they
should wait for a permanent answer.
William Carter (University of California) wanted to add to Question 4: What is the
distribution of conditions where ozone is a problem and where VOC control is relevant.
Howard Feldman (API) commented about the time period of a scale: 1-hour or 8-hour or
some other time period.
Ed Edwards (Dunn-Edwards Corporation) was concerned about the sensitivity of the
models and their ability to measure small changes in product formulations and how these small
changes are expanded into a airshed which contain only fractional percentage of VOCs.
Jim Berry (Berry Environmental) made several comments concerning the change in 1976
from Rule 66 to the beginning of the federal program. Rule 66 was based on reactivity
measurement for one solar day. Subsequent work based on multi-day exposures, recognized that
many of the compounds that Rule 66 exempted actually reacted over the longer term. In earlier
discussions, Dr. Jeffries stated that half of the ozone formation was the result of slower reacting
materials. Mr. Berry continued by discussing the impact that Rule 66 and, subsequently the
change to the federal program, had on various industries. He commented on the resources spent
on reformation to comply with Rule 66 and then again to comply with the federal program. He
urged that the lessons learned in 1976 not be forgotten, and that changes be made to the federal
program only after there is a longer term vision.
Paul Makar (Atmospheric Environment Service) asked if it would be possible to create a
hierarchy of methods for measuring reactivity that agree with one another. Many methods of
dealing with reactivity had been presented; his concern was that the use of more than one method
for regulatory purposes may necessitate intercomparison to ensure that the methods do not give
6-17
-------
conflicting results. For example, before using 3D models for regulatory purposes, their results
should be compared to the current ethane standard and/or MIR indicies. Any differences should
be fully understood in advance of implementation of regulations, to avoid undermining their
scientific credibility.
Jeffrey Gaffney (Argonne National Laboratory) suggested that a question be added about
the reactivity of secondary products of the primary emitted VOCs. The chemistry of these
secondary products needs to be understood, because they will play a role in the long-range
transport issue.
6-18
-------
PUBLIC/PRIVATE PARTNERSHIP DISCUSSION SUMMARY
Jake Hales (Moderator) began this portion of the workshop by describing NARSTO. Dr.
Hales then asked the group if it is appropriate to go into a public/private partnership to pursue the
reactivity issue, either within NARSTO or separate from NARSTO. Dr. Hales described his
views on the function of this committee as follows: (1) to establish the forum for communication
among interested participants, including the policy community and (2) to design a plan for the
future strategy for reactivity research and to provide a time table for the research. He asked the
group if there were any strong feeling against forming a public/private partnership. He took the
lack of response from the group as consent.
6-19
-------
AD HOC OPERATIONS TASK FORCE DISCUSSION SUMMARY
Jake Hales (Moderator) began by describing his thoughts about generally how this ad hoc
task force would operate. Anyone would be welcome to participate. At a minimum, there would
be several meetings per year. There would be communications on a regular basis. Travel
expenses would be the responsibility of the individuals participating.
A question was asked about whether the ad hoc task force would necessarily function
under NARSTO. Dr. Hales responded by stating that the task force could be either independent
or it could work as a functionary group under NARSTO. It could blend into all of the NARSTO
task activities: modeling, chemistry, observations, assessment, etc. If the group was to work
under NARSTO, it would have to adhere to the quality assurance and data management
guidelines. It would also have to adhere to the basic NARSTO principles which are to do good
research for the benefit of policy.
Dr. Hales continued by asking for volunteers to serve on the ad hoc task force. There was
a show of hands. Dr. Hales suggested that a sign-up list be circulated, and that the group caucus
immediately after lunch.
Dr. Hales began the discussion by asking the group if this ad hoc group should function
under the NARSTO umbrella. Robert Wendoll (Dunn-Edwards Corporation) asked if NARSTO
doesn't also have to agree to this association. Jake Hales described the standard procedures for
NARSTO to accept a field program under its umbrella. The Science and Resource Planning
group makes these decisions, but he expects no problem with this. William Carter (University of
California) suggested that the group encourage Europeans participation, who have made
important contributions to this work. Dr. Hales responded by saying that NARSTO has in the
past worked with the Europeans. EuroTrac is an affiliate member of NARSTO and other
European efforts could be in the future.
Bob Avery (Eastman Chemical Company) asked about alternatives to NARSTO. He
didn't feel that the workshop group had the information available to make the decision on
whether to associate with NARSTO. Mr. Avery was particularly interested in more information
about the costs associated with association with NARSTO. He asked if the EPA could provide
the leadership as an alternative. Jim Vickery (EPA/ORD) responded by stating that the EPA is
committed to conducting and organizing all of their ozone related research through NARSTO for
two primary reasons. This has helped to coordinate all of the different aspects of the research
and to effectively allocate the scarce resources of people and money. Dr. Vickery strongly
recommended that the ad hoc task force operate under NARSTO.
Jeffrey Holmstead (Latham & Watkins) asked how this effort would feed into the EPA's
policy decisions. He asked about the EPA's level of interest in pursuing these issues and
whether the EPA is in a position to accommodate the changes in the policy that this group would
recommend. Jake Hales expressed his observation that a simple liaison with a policy team is not
effective. There needs to be a strong presence of EPA policy people hi the partnership. Bill
Johnson (EPA/OAQPS) responded to Mr. Holmstead's question by stating that this question is
one that EPA's management would need to answer. Joe Paisie (EPA/OAQPS) added that as
6-20
-------
Sally Shaver (EPA/OAQPS) said in her introductory comments to this workshop, the EPA is here
to listen, and has been listening, but they are not prepared to say what the results will be yet.
It was asked if this activity could enjoy equivalent status with the recent FACA process.
That is, could there be direct EPA/OAQPS participation in the process.
Jeffrey Gaffney (Argonne National Laboratory) asked about extending invitations for
participation in the task force to people who had left or who didn't attend the workshop, such as
automobile makers. Ron Patterson (EPA/ and NARSTO) commented that several people who
are no longer here did complete the commitment forms made available at the beginning of the
workshop. He added that many of the automakers are members of NARSTO.
Tim Lawrence (Georgia Pacific and American Paper and Forestry Association)
commented, based on his experience with several FACA processes, that the level of interest,
participation, and commitment of resources, particularly by the regulating community, is directly
related to their sense of just how serious EPA is about moving the process forward to a useful
endpoint. He concluded that it will be very important at the beginning to see some real finite
indication of EPA's level of interest.
Howard Feldman (API) commented that the next step is to develop a plan with a
specified time horizon. It will be important to know what time horizon EPA would be receptive
to on the policy side. On the other hand, it is important to continue to work towards the good
science, because the policy makers will use it if it is there.
Jim Vickery (EPA/ORD) agrees with Mr. Feldman's comment about the importance of
good science and the fact that it will feed into the policy process. He described the difference
between the FACA processes and NARSTO. For the FACA process, there was a statutory driver
that required an output by a certain date. There is no such a driver for the reactivity policy.
Reactivity research is very much like the other ozone research organized under NARSTO. Under
NARSTO, the researchers and the policy makers are brought together to organize the research in
such as way as to use the resources efficiently and solve the most important questions for the
policy maker as quickly as possible. The EPA policy office is committed to using good science
as soon as it is produced.
Jake Hales (Moderator) summarized this discussion by stating that there is consensus to
form a public/private partnership on reactivity research and there are people interested in
participating in an ad hoc task force to develop the forum, the research plan and the time table.
Whether this task force operates under NARSTO is a question that will be deferred to the task
force itself. Everyone interested in serving on the task force was asked to meet after the
conclusion of the workshop.
6-21
-------
APPENDICES
-------
APPENDIX A
Photochemical Reactivity Workshop
May 12-14,1998
ATTENDEES LIST
Zac Adelman
Student
University of NC
Chapel Hill, NC 27599
919-966-1372
zac@ozone.sph.unc.edu
Robert Altenburg
Air Quality Program Specialist
PA Dept. Environmental Protection, Air Quality
400 Market St, 12th Floor
Harrisburg, PA 17105
717-787-9495 Fax 717-772-2303
altenburg.robert@al.dep.state.pa.us
Praveen Amar
Director, Science & Policy
NESCAUM
129 Portland St
Boston, MA 02114
617-367-8540 Fax 617-742-9162
pamar@nescaum.org
Viney Aneja
Research Professor
NCSU, Marine, Earth, & Atmospheric Sciences
Box 8208
Raleigh, NC 27695
919-515-7808 Fax 919-515-9802
viney_aneja@ncsu.edu
Anne Arnold
Environmental Engineer
US EPA Region I
JFK Federal Bldg
Boston, MA 02135
617-565-3166 Fax 617-565-4940
arnold.anne@epamail.epa.gov
Roger Atkinson
Air Pollution Research Center
University of California
Riverside, CA 92521
909-787-4191 Fax 909-787-5004
ratkins@mail.ucr.edu
Robert Avery
Sr. Associate, Product Issues Management
Eastman Chemical Company
PO Box 431,6-280
Kingsport, TN 37617
423-229-5409 Fax 423-224-0208
rjavery@eastman.com
Dennis Bahler
Professor
NC State University
Box 8206
Raleigh, NC 27695
919-515-3369 Fax 919-515-7896
bahler@ncsu.edu
Dan Baker
Staff Research Engineer
Shell Oil
3333Hwy6S
Houston, TX 77082
281-544-8437 Fax 281-544-8727
dcbaker@shellus.com
Ian Barnes
Bergische University Wuppertal
Gauss Str. 20
Wuppertal, D-42097 Germany
49-202-439-2510 Fax 49-202-439-2505
barnes@physchem.uni-wuppertal.de
Gary Beckstead
Environmental Protection Engineer
Illinois EPA
1340N9thSt
Springfield, IL 62702
217-524-4343 Fax 217-524-4710
epa2161 @epa.state.il.us
William Benjey
Physical Scientist
US EPA/NERL/AMD
MD-80
Research Triangle Park, NC 27711
919-541-0821 Fax 919-541-1379
benjey@hpcc.epa.gov
A-l
-------
Howard Berman
Senior Vice President
Kessler & Associates
510 llthSt, SE
Washington, DC 20003
202-547-6808 Fax 202-546-5425
Jim Berry
Berry Environmental
PO Box 20634
Raleigh, NC 27619
919-785-9631 Fax 919-785-9631
j imberryec@aol .com
Karen Borel
Environmental Engineer
US EPA Region IV
61 Forsyth St
Atlanta, GA 30303
404-562-9029
borel.karen@epamail.epa.gov
Kim Boudreaux
Environmental Engineering Advisor
Albemarle Corporation
451 Florida St
Baton Rouge, LA 70808
504-388-7776 Fax 504-388-7046
kim_boudreaux@albemarle.com
George Brown
Executive Director
National Aerosol Association
787 Windgate Dr
Annapolis, MD 21401
410-349-8614 Fax 410-349-8616
gwbjmb@annap.infi.net
Christine Brunner
Chemical Engineer
US EPA
2000 Traverwood
Ann Arbor, MI 48105
734-214-4287 Fax 734-214-4051
brunner.christine@epa.gov
Larry Bruss
Ozone & SIP Development Section Chief
WI Dept of Natural Resources
PO Box 7921
Madison, WI 53707-7921
608-267-7543 Fax 608-267-0560
brussl@mai!01 .dnr.state.wi.us
Daewon Byun
Physical Scientist
US EPA/AMD/NERL
MD-80
Research Triangle Park, NC 27711
919-541-0732 Fax 919-541-1379
bdx@hpcc.epa.gov
William Carter
Research Chemist
University of California
CE-CERT, University of California
Riverside, CA 92521
909-781-5797 Fax 909-781-5790
carter@cert.ucr.edu
Donna Carvalho
Pennzoil Products & Magie Brothers
PO Box 2967
Houston, TX 77252
713-546-8723 Fax 713-546-8930
Roy Carwile
Manager, Air Programs
ALCOA
1906 Alcoa Bldg
Pittsburgh, PA 15219
412-553-2680 Fax 412-553-4077
roy.carwile@alcoa.com
Janet Catanach
Environmental Planner
Exxon Chemical Company
13501 Katy Freeway
Houston, TX 77079
281-870-6959 Fax 281-588-4664
A-2
-------
William Chameides
Regents Professor
Georgia Institute of Technology
School of Earth & Atmospheric Sciences
221 Bobby Dodd Way
Atlanta, GA 30332-0340
404-894-1749 Fax 404-894-1106
wlc@blond.eas.gatech.edu
Grish Chandra
Scientist & Technical Manager
Dow Corning Corp.
PO Box 994
Midland, MI 48640
517-496-5990 Fax 517-496-5595
gchandra@dcrn.e-mail.com
Kirit Chaudhari
Director, Office of Air Data Anal) sis
VA Dept of Environmental Quality
PO Box 10009
Richmond, VA 23240
804-698-4414 Fax 804-698-4510
kochaudhar@dea.state.va.us
Mengdawn Cheng
Research Staff Member
Oak Ridge National Laboratory
Environmental Science Division, MS 6038
Oak Ridge, TN 37831
423-241-5918 Fax 423-576-8646
chngmd@ornl.gov
Qiao-Jung Chien
Student
University of NC - Chapel Hill
Chapel Hill, NC 27606
919-966-1372
chien@ozone.sph.unc.edu
Jason Ching
Chief, Atmospheric Model Development Branch
US EPA/AMD/NERL
MD-80
Research Triangle Park, NC 27711
919-541-4801 Fax 919-541-1379
ching.jason@epamail.epa.gov
Shao-Hang Chu
Environmental Scientist
US EPA
MD-15
Research Triangle Park, NC 27711
919-541-5382 Fax 919-541-7690
chu.shao-hang@epamail.epa.gov
Jeff Clark
Director, Policy Analyses & Communications
US EPA/OAQPS
MD-10
Research Triangle Park, NC 27711
919-541-5615 Fax 919-541-2464
John Clary
President
Bio Risk
PO Box 2326
Midland, MI 48641
517-839-8130 Fax 517-839-8130
bioriskl@aol.com
Andy Collantes
Technical Director - Chemicals
Sherwin-Williams, Diversified Brands
31500 Solon Rd
Solon, OH 44139
440-498-6092
Karla Colle
Senior Staff Chemist
Exxon Chemical Co
PO Box 4900
Baytown, TX 77520
281-834-5115 Fax 281-834-1904
Ted Creekmore
Environmental Engineer
US EPA
MD-15
Research Triangle Park, NC 27711
919-541-5699 Fax 919-541-0824
A-3
-------
Tammy Croote
Economist
USEPA/OAQPS/ISEG
MD-15
Research Triangle Park, NC 27711
919-541-1051 Fax 919-541-0839
croote.tammy@epa.gov
Larry Cupitt
Director, Human Exp. & Atmospheric Sci. Div.
US EPA/ORD/NERL
MD-77
Research Triangle Park, NC 27711
919-541-2454 Fax 919-541-0239
cupitt.larry@epamail.epa.gov
Tom Dann
Head Air Toxics
Environment Canada
ETC, 3439 River Rd
Ottawa, Ontario K1A OH3 Canada
613-991-9459 Fax 613-998-4032
dann.tom@etc.ec.gc.ca
Phil Davison
Issues & HSE Manager
BP Chemicals Ltd
Hull Works, Salt End
Hull, London HU12 80S UK
44-1482-2448 Fax 44-1482-2057
davisonp@bp.com
John Dege
Manager, Air Programs
DuPont - Environmental Excellence Center
1007 Market St
Wilmington, DE 19707
302-773-0900 Fax 302-774-1361
john.a.dege@usa.dupont.com
Robin Dennis
US EPA/NERL
MD-80
Research Triangle Park, NC 27711
919-541-2870 Fax 919-541-1379
rdennis@hpcc.epa.gov
Dick Derwent
Meteorological Office
London Road
Bracknell, Berkshire RG12 2SZ UK
44.1344854624 Fax 44-1344854493
rgderwent@meto.gov.uk
Folke Dettling
Dipl. Chem.
Umweltbundesamt
Postfach 330022
Berlin, Germany D14191
49-30-8903-3845 Fax 49-30-8903-3232
folke.dettling@uba.de
David Dewitt
Senior Engineer
Honda Research & Development
1900 Harpers Way
Torrance, CA 90501
310-781-5718 Fax 310-781-5655
ddewitt@hra.com
N. N. Dharmarajan
Senior Consultant
Central & South West Services, Inc.
1616 Woodall Rodgers Freeway
Dallas, TX 75202
214-777-1373 Fax 214-777-1320
ndharmarajan@csw.com
Basil Dimitriades
Senior Scientific Advisor
US EPA
MD-80
Research Triangle Park, NC 27711
919-541-2706 Fax 919-541 -1094
dimitriades.basil@epamail.epa.gov
David Dixon
Associate Director, Theory, Modeling &
Simulation Env. Molecular Sciences Lab
Pacific Northwest National Laboratory
906 Battelle Blvd, MSK1-83
Richland, WA 99352
509-372-4999 Fax 509-375-6631
da_dixon@pnl.gov
A-4
-------
Ian Dobson
External Affairs Manager, Solvents Business
BP Chemicals LTD
Pinners Hall, 105-108 Old Broad St
London, EC2N 1ER UK
44-171 -496-2786 Fax 44-171 -496-2706
dobsoni@bp.com
Paul Dugard
Executive Director
Halogenated Solvents Industry Alliance
200lLSt,NW, Suite 506A
Washington, DC 20036
202-775-0232 Fax 202-833-0381
srisotto@hsia.org
Robin Dunkins
Environmental Engineer
US EPA/OAQPS/AQSSD
MD-15
Research Triangle Park, NC 27711
919-541-5335 Fax 919-541-5489
dunkins.robin@epamail.epa.gov
John Durkee
Owner
Creative Enterprizes
105 Anyway, Suite 209
Lake Jackson, TX 77566
409-292-0244 Fax 409-292-0440
jdurkee@brazosport.cc.tx.us
Cyril Durrenberger
Senior Engineer
TX Natural Resource Conservation Commission
PO Box 13087, MC-164
Austin, TX 78711
512-239-1482 Fax 512-239-1500
cdurrenb@tnrcc.state.tx.us
Edward Edney
Research Physical Scientist
US EPA
MD-84
Research Triangle Park, NC 27711
919-541-3905 Fax 919-541-4787
edney.edward@epamail.epa.gov
Edward Edwards
Owner
Dunn-Edwards Corp.
4885 E 52nd Place
Los Angeles, CA 90040
213-771-3330 Fax 213-771-4440
Peter Ellis
Exxon Chemical Co
5200 Bayway Dr
Baytown, TX 77058
281-834-1681 Fax 281-834-2747
peter.s.ellis@exxon.sprint.com
Ron Evans
Group Leader
US EPA/OAQPS/AQSSD/ISEG
MD-15
Research Triangle Park, NC 27711
919-541-5488 Fax 919-541-0839
evans.ron@epamail.gov
Howard Feldman
Research Program Coordinator - Air
American Petroleum Institute
1220LSt,NW
Washington, DC 20005
202-682-8340 Fax 202-682-8270
feldman@api.org
John Festa
Senior Scientist
American Forest & Paper Association
1111 19thSt,NW
Washington, DC 20036
202-463-2587 Fax 202-463-2423
john_festa@afandpa.org
Franz Fiedler
Professor
University/Research Center Karlsruhe
Kaiserstr. 12
Karlsruhe, Germany
49-721-608-3355 Fax 49-721-608-6102
f.fiedler@phys.uni-karlsruhe.de
A-5
-------
Gary Foley
Director
US EPA/NERL
MD-75
Research Triangle Park, NC 27711
919-541-2106 Fax 919-541-0445
foley.gary@epamail.epa.gov
Yvonne Fong
Environmental Protection Specialist
EPA Region IX
75 Hawthorne St, AIR-4
San Francisco, CA 94105
415-744-1199 Fax 415-744-1076
fong.yvonnew@epamail.epa.gov
Barbara Francis
Director
Chemical Manufacturers Association
1300 Wilson Blvd
Arlington, VA 22209
703-741-5609 Fax 703-744-6091
barbara_francis@mail.cmahq.com
Douglas Fratz
Director of Scientific Affairs
Chemical Specialties Manufacturers Association
1913 Eye St, NW
Washington, DC 20006
202-872-8110 Fax 202-872-8114
dfratz@csma.org
Christopher Frey
Assistant Professor
Civil Engineering - NCSU
311 Mann Hall - CB 7908
Raleigh, NC 27695
919-515-1155 Fax 919-515-7908
frey@eos.ncsu.edu
Dawn Froning
Environmental Specialist
MO Dept of Natural Resources
PO Box 102
Jefferson City, MO 65101
573-751-4817 Fax 573-751-2706
dfroning@mo.state.us
Jeffrey Gaffney
Research Chemist
Argonne National Lab
9700 Cass Ave
Argonne, IL 60439
630-252-5178 Fax 630-252-8895
gaffney@anl.gov
Kenneth Galluppi
Senior Scientist
University of NC
107 Miller Hall, CB #1105
Chapel Hill, NC 27599
919-966-9926 Fax 919-966-9920
galluppi@unc.edu
Mark Garrison
Air Quality Meteorologist
ERM
855 Springdale Dr
Exton,PA 19341
610-524-3674 Fax 610-524-7798
mark_garrison@erm.com
Harald Geiger
Bergische University Wuppertal
Gauss Str. 20
Wuppertal, D-42097 Germany
49-202-439-3832 Fax 49-202-439-2505
geiger@physchem.uni-wuppertal.de
Nash Gerald
Environmental Engineer
US EPA/OAQPS/EMAD
MD-14
Research Triangle Park, NC 27711
919-541-5652 Fax 919-541-1903
gerald.nash@epa.gov
Sharon Gidumal
Technical Consultant
DuPont
PO Box 80711
Wilmington, DE 19808-0711
302-999-5325 Fax 302-999-2093
A-6
-------
Anne Giesecke
V.P., Environmental Activities
American Bakers Association
1350ISt,NW, Suite 1290
Washington, DC 20005
202-789-0300 Fax 202-898-1164
agiesecke@americanbakers.org
Gerald Gipson
Research Physical Scientist
US EPA/NERL
MD-80
Research Triangle Park, NC 27711
919-541-4181 Fax 919-541-1379
ggb@hpcc.epa.gov
David Golden
Senior Staff Scientist
SRI International
333 Ravenswood Ave
Menlo Park, CA 94025
650-859-3811 Fax 650-859-6196
golden@sri.com
Jack Goldman
Attorney at Law
Bryan Cave LLP
700 13th St,NW, Suite 600
Washington, DC 20005
202-508-6000 Fax 202-508-6200
j4g@bryancavellp.com
Joyce Graf
Director, Environmental Science
CTFA
1101 17th StNW
Washington, DC 20036
202-331-1770 Fax 202-331-1969
grafj@ctfa.org
David Graham
Business Manager
Occidental Chemical Corp
PO Box 809050
Dallas, TX 75380
972-404-4198 Fax 972-448-6676
dave_graham @oxy. com
Ross Gustafson
Technical Director
Florida Chemical Co
401 Somerset Dr
Golden, CO 80401
303-216-9420 Fax 303-216-9425
info@floridachemical.com
Jeremy Hales
Workshop Moderator
ENVAIR
60 Eagle Reach
Pasco, WA 99301
509-546-9542 Fax 509-546-9522
jake@odysseus.owt.com
Robert Hamilton
Research Associate
Amvvay Corp
7575 E Fulton Rd
Ada, MI 49355
616-787-7697 Fax 616-787-7941
bhamilton@amway.com
Adel Hanna
Manager, Environmental Research
MCNC - Environmental Programs
PO Box 12889
Research Triangle Park, NC 27709
919-248-9230 Fax 919-248-9245
adel@mcnc.org
Alan Hansen
Manager, Tropospheric Studies
EPRI
POBox 10412
Palo Alto, CA 94303
650-855-2738 Fax 650-855-2950
ahansen@epri.com
Madelyn Harding
Admin, Product Compliance & Registration
The Sherwin-Williams Co
101 W Prospect Ave
Cleveland, OH 44115-1075
216-556-2630 Fax 216-263-8635
mkharding@shenvin.com
A-7
-------
Kent Hoekman
Sr. Staff Scientist
Chevron Products Co
575 Market St
San Francisco, CA 94105
415-984-3060 Fax 415-894-2075
skho@chevron.com
Jeffrey Holm stead
Partner
Latham & Watkins
1001 Pennsylvania Ave,NW, Suite 1300
Washington, DC 20002
202-637-2287 Fax 202-637-2201
jeff.holmstead@lw.com
Harvey Jeffries
Professor
University of NC
CB #7400 Rm 120 Rosenan Hall
Chapel Hill, NC 27599
919-966-7312 Fax 919-933-2393
Harvey@unc.edu
William Johnson
Environmental Engineer
US EPA
MD-15
Research Triangle Park, NC 27711
919-541-5245 Fax 919-541-0824
johnson.williaml@epamail.epa.gov
Steve Jones
AQACS
SC Air Quality Management Division
21865 E Copley Dr
Diamond Bar, CA 91765
909-396-2094 Fax 909-396-3867
sjones@aqmd.gov
Norman Kaplan
Sr. Project Engineer
US EPA
MD-4
Research Triangle Park, NC 27711
919-541-2556 Fax 919-541-0579
nkaplan@engineer.aeerl.epa.gov
Richard Karp
US EPA, Region VI
1445 Ross Ave, 12th Fl, Suite 1200
Dallas, TX 75202-2733
karp.richard@epamail.epa.gov
Terry Keating
Harvard University - BCSIA
79 JFK St
Cambridge, MA 02138
617-495-1417 Fax 617-495-8963
terry_keating@harvard.edu
Brian Keen
Senior Technology Manager
Union Carbide Corporation
PO Box 8361
South Charleston, WV 25303
304-747-4897 Fax 304-747-4623
keenbt@ucarb.com
Gail Kelly
Project Manager
ARCO Chemical Co
3801 W Chester Pike
Newtown Square, PA 19073
610-359-6443 Fax 610-359-3155
cnsgbk@arcochem .com
Dale Kemmerick
Manager of Data & Modeling Unit
Georgia Environmental Protection Division
4244 International Pkwy, Suite 120
Atlanta, GA 30354
404-363-7092 Fax 404-363-7100
dale_kemmerick@mail.dnr.state.ga.us
Donna Kenski
Environmental Scientist
US EPA
77 W Jackson Blvd
Chicago, IL 60604
312-886-7894 Fax 312-886-5824
kenski.donna@epamail.epa.gov
A-8
-------
Marc Kessler
Post Doc
University of NC
CB# 7400
Chapel Hill, NC 27599
919-966-1372
marc_kessler@unc.edu
Tad Kleindienst
Principal Scientist
Mantech Environmental Technology, Inc.
PO Box 12313
Research Triangle Park, NC 27709
919-541-2308 Fax 919-549-4665
tkleindienst@man-env.com
Kenneth Knapp
Research Chemist
US EPA
MD-46
Research Triangle Park, NC 27711
919-541-1352 Fax 919-541-0960
knapp.ken@epamail.epa.gov
Robert Kozak
President
Atlantic Biomass Conversions, Inc.
1916-16thSt,NW
Washington, DC 20009
202-387-1838 Fax 202-483-6630
kzakr@aol.com
Michael Kravetz
Director Analytical Research
Cosmair Cosmetics Corp
159 Terminal Ave
Clark, NJ 07066
732-499-2934 Fax 732-499-2978
Shri Kulkarni
President & Principal Investigator
Kultech Incorporated
1323 Mellon Ct
Cary,NC 27511
919-467-0598 Fax 919-468-8805
kultecshri@aol.com
Naresh Kumar
Senior Air Quality Analyst
Sonoma Technology, Inc.
5510SkylaneBlvd, Suite 110
Santa Rosa, CA 95403
707-527-9372 Fax 707-527-9398
naresh@sonomatech.com
Jonathan Kurland
Research Scientist
Union Carbide Corp
PO Box 8361
S. Charleston, WV 25303-0361
304-747-3816 Fax 304-747-3752
kurlanjj@ucarb.com
William Kuykendal
Senior Environmental Engineer
US EPA/OAQPS
MD-14
Research Triangle Park, NC 27711
919-541-5372 Fax 919-541-0684
kuykendal.bill@epamail.epa.gov
Brian Lamb
Professor
Washington State University
Dept of Civil & Environmental Engineering
Pullman, WA 99164-2910
509-335-5702 Fax 509-335-7632
blamb@wsu.edu
John Langstaff
Senior Analyst
EC/R, Incorporated
1129 Weaver Dairy Rd
Chapel Hill, NC 27514
919-933-9501x239 Fax 919-933-6361
jlangstaff@mindspring.com
Sang-Mi Lee
Scientist
US EPA/AMD/NERL
MD-80
Research Triangle Park, NC 27711
919-541-2368 Fax 919-541-1379
smlee@hpcc.epa.gov
A-9
-------
Wen Li
Project Scientist
ManTech Environmental Technology, Inc.
PO Box 12313
Research Triangle Park, NC 27709
919-541-2596 Fax 919-549-4665
wli@man-env.com
Xiaoyn Lin
Student
University of NC
Chapel Hill, NC 27514
919-966-3932
xiaoyu@ozone.sph.unc.edu
Deborah Luecken
Physical Scientist
US EPA
MD-84
Research Triangle Park, NC 27711
919-541-0244 Fax 919-541-4787
luecken@geoid.rtpnc.epa.gov
James Magee
Environmental Chemical Specialist - Advanced
LA Dept of Environmental Quality
PO Box 82135
Baton Rouge, LA 70809
504-765-0146 Fax 504-765-0921
jamesm@deq. state, la.us
Paul Makar
Atmospheric Environment Service
4905 Dufferin St
Downsview, Ontario M3H 5T4 Canada
416-739-4692 Fax 416-739-4288
paul.makar@ec.gc.ca
Deborah Mangis
US EPA/NERL
MD-77B
Research Triangle Park, NC 27711
919-541-3086 Fax 919-541-7953
mangis.deborah@epamail.epa.gov
Mike Manning
Corporate Air Specialist
BASF Corporation
Sand Hill Rd
Enka, NC 28728
704-667-7481 Fax 704-667-7718
manninj@basf.com
Robert Matejka
Environmental Manager
AKZO Nobel Coatings, Inc.
1431 Progress Ave
High Point, NC 27261
336-801-0872 Fax 336-883-9525
Rohit Mathur
Research Scientist
MCNC - Environmental Programs
POBox 12889
Research Triangle Park, NC 27709
919-248-9246 Fax 919-248-9245
mathur@mcnc.org
Carolyn Matula
Manager, Solvents Regulatory Support
Shell Chemical Co
PO Box 4320
Houston, TX 77210
713-241-0579 Fax 713-241-3325
camatula@shellus.com
Gary McAlister
Chemist
US EPA/OAQPS
MD-19
Research Triangle Park, NC 27711
919-541-1062
Jim McCabe
Sr. Environmental Engineer
The Clorox Company
PO Box 493
Pleasanton, CA 94566
510-847-6674 Fax 510-847-2496
jim.mccabe@clorox.com
A-10
-------
Lesa McDonald
Environmental Manager
Gemini Coatings, Inc.
PO Box 699
El Reno, OK 73036
405-262-5710 Fax 405-262-9310
Mack McFarland
Principal Scientist, Environmental Programs
DuPont Fluoroproducts
PO Box 80702
Wilmington. DE 19880
302-999-2505 Fax 302-999-2816
mack.mcfarland@usa.dupont.com
Jan Meyer
Senior Environmental Engineer
US EPA/OAQPS
MD-13
Research Triangle Park, NC 27711
919-541-5254 Fax 919-541-5689
meyer.jan@epa.gov
Ned Meyer
US EPA/OAQPS
MD-14
Research Triangle Park, NC 27711
919-541-5594 Fax 919-541-0044
Meyer.Ned@epa.gov
Jana Milford
Associate Professor
University of Colorado
Dept. of Mechanical Engineering, CB# 427
Boulder, CO 80309
303-492-5542 Fax 303-492-2863
milford@spot.colorado.edu
Bruce Moore
Environmental Engineer
US EPA/ESD
MD-15
Research Triangle Park, NC 27711
919-541-5460 Fax 919-541-5689
moore.bruce@epa.gov
David Morgott
Eastman Kodak Co
1100 Ridgeway Ave
Rochester, NY 14652-6272
716-588-3704 Fax 716-722-7561
dmorgott@kodak.com
Jim Neece
Urban Airshed Modeler
TX Natural Resource Conservation Commission
POBox 13087, MC-164
Austin, TX 78711
512-239-1524 Fax 512-239-1500
jneece@tnrcc.state.tx.us
Robert Nelson
Director, Environmental Affairs
National Paint & Coatings Association
1500 Rhode Island Ave, NW
Washington, DC 20005
202-462-6272 Fax 202-462-8549
bnelson@paint.org
Monica Nichols
Principal Environmental Engineer, Flexible
Packaging Division
Reynolds Metals Company
2101 ReymetRd
Richmond, VA 23237
804-743-6154 Fax 804-285-5222
msnichol@lanmail.rmc.com
Becky Norton
Ecologist II
AR Dept of Pollution Control & Ecology
8001 National Dr
Little Rock, AR 72219
501-682-0060 Fax 501-682-0753
nortonb@adeq. state, ar.us
Joan Novak
Chief, Modeling Systems Analysis Branch
US EPA/AMD/NERL
MD-80
Research Triangle Park, NC 27711
919-541-4545 Fax 919-541-1379
novak.joan@epamail.epa.gov
A-ll
-------
Brenda Nuite
Senior Regulatory Advisor
The Dial Corporation
15101 N ScottsdaleRd
Scottsdale, AZ 85254
602-754-6151 Fax 602-754-6180
nuite@dialcorp.com
Anne O'Donnell
Manager, Analytical Development
Safety-Kleen Corp
PO Box 92050
Elk Grove Village, 1L 60009
773-825-7053 Fax 773-825-7850
aodonnell@safety-kleen.com
Jesse O'Neal
Environmental Science & Policy, Inc.
208 Beckley Court
Raleigh, NC 27615
919-676-1713 Fax 919-676-9056
jesoneal@pipeline.com
Eduardo Olaguer
Environmental Science Specialist
The Dow Chemical Co
1803 Bldg, The Dow Chemical Co
Midland, MI 48674
517-636-2927 Fax 517-638-9305
epolaguer@dow.com
Philip Ostrowski
Technical Service Manager
Occidental Chemical
PO Box 344
Niagara Falls, NY 14302
716-278-7346 Fax 716-278-7297
phil_ostrowski@oxy.com
Lawrence Otwell
Senior Environmental Engineer
Georgia-Pacific
POBox 105605
Atlanta, GA 30348
404-652-5081 Fax 404-654-4695
lpotwell@gapac.com
John Owens
Research Specialist
3M
3M Center, Bldg 236-3A-03
St. Paul, MN 55125
612-736-1309 Fax 612-733-4335
jgowens@mmm.com
Prasad Pai
Senior Scientist
AER
2682 Bishop Dr, Suite 120
San Ramon, CA 94583
5 ] 0-244-7123 Fax 510-244-7129
ppai@aer.com
Joe Paisie
US EPA/OAQPS
MD-15
Research Triangle Park, NC 27711
919-541-5556
Uay Palanski
Attorney
Wilmer, Cutler & Pickering
2445 M St, NW
Washington, DC 20037
202-663-6602 Fax 202-663-6363
ipalansky@wilmer.com
Randy Pasek
Manager, Atmospheric Processes Research Sect.
CA Air Resources Board
2020 L St
Sacramento, CA 95814
916-324-8496 Fax 916-322-4357
rpasek@arb.ca.gov
Ronald Patterson
Physical Science Administrator
US EPA/NERL
MD-77
Research Triangle Park, NC 27711
919-541-3779 Fax 919-541-0329
patterson.ronald@epamail.epa.gov
A-12
-------
John Patton
Environmental Protection Specialist 6
TN Air Pollution Control
L&C Annex, 401 Church St, 9th Floor
Nashville, TN 37243-1531
615-532-0604 Fax 615-532-0614
jpatton@mail.state.tn.us
Richard Paul
Manager, Environmental Health
American Automobile Manufacturers Assoc.
7430 2nd Ave, Suite 300
Detroit, MI 48202
313-871-5344 Fax 313-872-5400
Dib Paul
Project Manager
IT Corporation
1104CastaliaDr
Cary,NC 27513
919-233-7024 Fax 919-233-7027
dibp@mindspring.com
Rick Phelps
Technical Associate
Eastman Chemical Co
POBox431,Bldg280W
Kingsport, TN 37662
423-229-5164 Fax 423-229-4864
rcphelps@eastman.com
Daniel Pourreau
Coatings Development Manager
ARCO Chemical Co
3801 W Chester Pike
Newtown Square, PA 19073
610-359-6837 Fax 610-359-5753
cnsdbp@arcochem .com
Gene Praschan
Manager
American Automobile Manufacturers Assoc.
1000 Park Forty Plaza, Suite 300
Durham, NC 27713
919-547-7100 Fax 919-547-7102
praschea@ix.netcom .com
Harry Quarles
Research Scientist
Oak Ridge National Laboratory
PO Box 2008
Oak Ridge, TN 37831
423-241-2412 Fax 423-576-8543
hq3@ornl.gov
M. B. Ranade
President
Particle Technology Inc.
PO Box 925
Hanover, MD 21076
301-931-1037 Fax 301-931-1038
ranade@erols.com
S. T. Rao
Asst. Commissioner, Office of Science & Tech.
NYS Dept of Environmental Conservation
50 Wolf Rd, Room 198
Albany, NY 12233-3259
518-457-3200 Fax 518-485-8410
strao@dec.state.ny.us
Doug Raymond
Director, Regulatory Affairs
Diversified Brands
31500 Solon Rd
Solon, OH 44139
440-498-6049 Fax 440-519-663 8
djraymond@sherwin.com
Leslie Ritts
Attorney
Hogan & Hartson
555 Thirteenth St,NW
Washington, DC 20004
202-637-6573 Fax 202-637-5910
Isr@dc2.hhlaw.com
Can Roque
Materials Engineer
Naval Aviation Depot, Dept. of Navy
Code 4344
Jacksonville, FL 32212
904-542-4519x127 Fax 904-542-4523
roque.psd@navair.navy.mil
A-13
-------
Alexander Ross
Government Affairs Director
Rad Tech International, NA
400 N Cherry St
Falls Church, VA 22046
703-534-9313 Fax 703-533-1910
rossradtec@aol .com
Ted Russell
Professor
Georgia Institute of Technology
Env. Engineering, 200 Bobby Dodd Way
Atlanta, GA 30332-0512
404-894-3079 Fax 404-894-8266
trussell@pollution.ce.gatech.edu
Ron Ryan
Environmental Engineer
US EPA/OAQPS/EMAD
MD-14
Research Triangle Park, NC 27711
919-541-4330 Fax 919-541-0684
ryan.ron@epa.gov
Chris Salmi
Manager
NJ Dept of Environmental Protection
Bureau of Air Quality Planning
PO Box 418,401 E State St, 7th Fl
Trenton, NJ 08625
609-292-6722 Fax 609-633-6198
salmi@dep.state.nj.us
David Sanders
Environmental Engineer
US EPA/OAQPS/AQSSD
MD-15
Research Triangle Park, NC 27711
919-541-3356 Fax 919-541-0824
sanders.dave@epa.gov
Ken Schere
US EPA/ORD
MD-80
Research Triangle Park, NC 27711
919-541-3795 Fax 919-541-1379
skl@hpcc.epa.gov
Mark Schmidt
Statistician
US EPA
MD-14
Research Triangle Park, NC 27711
919-541-2416 Fax 919-541-1903,
schmidt.mark@epa.gov
John Schwind
Senior EHS Manager
Safety-Kleen
2110SYaleSt
Santa Ana, CA 92869
714-751 -0106 Fax 800-769-5841
Mohamed Serageldin
US EPA/OAQPS
MD-13
Research Triangle Park, NC 27711
919-541-2379 Fax 919-541-5689
serageldin.mohamed@epamail.epa.gov
Ken Sexton
Research Associate
University ofNC
CB#7400 School of Public Health
Chapel Hill, NC 27599
919-966-5451 Fax 919-966-7911
ken_sexton@unc.edu
Sally Shaver
Director
US EPA/AQSSD
MD-15
Research Triangle Park, NC 27711
919-541-5505 Fax 919-541-0804
shaver.sally@epamail.epa.gov
John Sherwell
Manager, Atmospheric Science
MD Dept of Natural Resources
Power Plant Assessment Division
Tawes Bldg B-3, 580 Taylor Ave
Annapolis, MD 21401
410-260-8667 Fax 410-260-8670
jsherwell@dnr.state.md.us
A-14
-------
David Smith
Principal Scientist
ManTech Environmental Technology, Inc.
POBox 12313
Research Triangle Park, NC 27709
919-406-2147
dsmith@man-env.com
Qingyuan Song
Postdoc
US EPA/UCAE/NOAA/AMD
MD-80
Research Triangle Park, NC 27711
919-541-1949 Fax 919-541-1379
qin@hpcc.epa.gov
Donald Spellman
Manager, Clean Air Alt Compliance
Louisville Gas & Electric
POBox 32010
Louisville, KY 40232
502-627-3425 Fax 502-627-2550
don.spellman@lgeenergy.com
Charlene Spells
Environmental Engineer
US EPA/0 AQPS/1PSG
MD-15
Research Triangle Park, NC 27711
919-541-5255 Fax 919-541-5489
spells.charlene@epamail.epa.gov
Bob Stallings
Environmental Engineer
US EPA
MD-15
Research Triangle Park, NC 27711
919-541 -7649 Fax 919-541 -0824
stallings.bob@epa.gov
William Stockwel!
Senior Scientist
Fraunhofer Institute D-82467
Kreuzeckbahnstrabe 19
Garmisch-Partenkirchen, Germany D-62467
011-49-8821-183262 Fax 011-49-8821-73573
stockwel@ifu.fhg.de
David Stonefield
Senior Environmental Engineer
US EPA/OAQPS/OPSG
MD-15
Research Triangle Park, NC 27711
919-541-5350 Fax 919-541-0824
stonefield.dave@epa.gov
Ron Stout
Technical Representative
Eastman Chemical Co
123 Lincoln St
Kingsport, TN 37662
423-229-3373 Fax 423-224-0414
ronstout@eastman.com
Dave Stringham
Manager, Reg & State Govt. Affairs
Safety-Kleen Corp
One Brinckman Way
Elgin, IL 60123
847-697-2221 Fax 847-468-8535
dstringham@safety-kleen.com
Gregory Suber
Ph.D. Student
Duke University
PO Box 2773
Durham, NC 27715
919-613-8054
gfs2@acpub.duke.edu
George Talbert
Assistant Director for Technology Transfer
Gulf Coast Hazardous Substance Research Cntr
POBox 10613
Beaumont, TX 77710
409-880-2183 Fax 409-880-2397
gotalbert@aol.com
Roger Tanner
Principal Scientist
TV A/Environmental Research & Services
POBox 1010, CEB2A
Muscle Shoals, AL 35662-1010
256-386-2958 Fax 256-386-2499
rltanner@tva.gov
A-15
-------
Philip Tham
Director, Regulatory Affairs
Estee Lauder Companies
125 Pinelawn Rd
Melville, NY 11747
516-531-1624 Fax 516-531-1565
ptham@estee.com
Stanley Tong
Environmental Protection Specialist
US EPA Region IX, Air Division
75 Hawthorne St, (AIR-4)
San Francisco, CA 94105
415-744-1191 Fax 415-744-1076
tong.stanley@epamail.epa.gov
Gail Tonnesen
Visiting Scientist
US EPA
MD-84
Research Triangle Park, NC 27711
919-541-4272 Fax 919-541-4272
tonnesen@olympus.epa.gov
James Vickery
Associate Lab Director for Air
US EPA/NERL
MD-75
Research Triangle Park, NC 27711
919-541-2184 Fax 919-541-3615
vickery.james@epamail.epa.gov
Darryl von Lehmden
Principal Environmental Engineer
Midwest Research Institute
5520 Dillard Rd
Cary,NC 27511
919-851-8181 x5167 Fax 919-851-3232
dvonlehmden@mriresearch.org
Fred Vukovich
Chief Scientist
Science Applications International Corp
615 OberlinRd, Suite 300
Raleigh, NC 27605
919-836-7563 Fax 919-832-7243
fVukovich@raleigh.saic.com
Kit Wagner
Principal Scientist
Atmospheric Information Systems
PO Box 721165
Norman, OK 73070
405-329-8707 Fax 405-329-8717
kit@ionet.net
Chris Walcek
Senior Research Associate
State University of NY at Albany
ASRC, 251 Fuller Rd
Albany, NY 12203
518-437-8720 Fax 518-437-8758
walcek@contrail.asrc.cestm.albany.edu
Zion Wang
University of NC
1152 College Ave
Palo Alto, CA 94306
650-424-8301 Fax 650-424-8301
zion_wang@unc.edu
Bob Wayland
Environmental Scientist
US EPA/OAQPS/IPSG
MD-15
Research Triangle Park, NC 27711
919-541-1045 Fax 919-541-5489
wayland.robertj@epamail.epa.gov
Ray Wells
Research Chemist
US Air Force (AFRL/MLQR)
139 Barnes Dr
Tyndall AFB, FL 32403
850-283-6087 Fax 850-283-6090
ray.wells@ccmail.aleq.tyndall.af.mil
Robert Wendoll
Director of Environmental Affairs
Dunn-Edwards Corp.
4885 E 52nd Place
Los Angeles, CA 90040
213-771-3330 Fax 213-771-4440
A-16
-------
Kurt Werner
Product Steward Specialist
3M
3M Center, Bldg 236-IB-10
St. Paul, MN 55144
612-733-8494 Fax 612-737-9909
kt\verner@mmm.com
Neil Wheeler
Chief of Environmental Applications
MCNC-NC Supercomputing Center
PO Box 12889
Research Triangle Park. NC 27709
919-248-1819 Fax 919-248-9245
Wheeler@ncsc.org
Gary Whitten
Chief Scientist
Systems Applications International, Inc.
101 Lucas Valley Rd
San Rafael, CA '94903
415-507-7152 Fax 415-507-7177
gzw@sai.icfkaiser.com
Carolyn Wills
Manager, Regulatory Affairs
Mary Kay, Inc.
1330 Regal Row
Dallas, TX 7524^
214-905-6360 Fax 214-905-6908
William Wilson
Physical Science Administrator
US EPA/NCEA
MD-52
Research Triangle Park, NC 27711
919-541-2551 Fax 919-541-5078
wilson.william@epa.gov
Ken Woodrow
Attorney
Baker & Hostetler LLP
1050 Connecticut Ave. Suite 1100
Washington, DC 20036
202-861-1739 Fax 202-861-1783
kwoodrow@baker-hostetler.com
Albert Yezrielev
Senior Staff Chemist
Exxon Chemicals
5200 Bayway Dr
Baytown. TX 77520
281 -834-2487 Fax 281 -834-2747
Rose Zaleski
Exxon Biomedical Sciences, Inc.
Mettlers Rd, CN 2350
East Millstone, NJ 08875-2350
732-873-6053 Fax 732-873-6009
rosemary.t.zaleski@erc.exxon.sprint.com
Guang Zeng
University of NC
Dept of ESE. CB# 7400, Prosenau Hall
Chapel Hill. NC 27599
919-966-3932
guang@ozone.sph.unc.edu
Elaine Zoeller
Technical Associate
Eastman Chemical Co
POBox431,B-280W
Kingsport, TN 37662
423^229-3983 Fax 423-229-4864
ezoeller@eastman.com
Bernard Z\sman
Technical Service Specialist
Occidental Chemical Corp
PO Box 344
Niagara Falls. NY 14302
716-278-7894 Fax 716-278-7297
bernie_z\ sman@oxy.com
A-17
-------
During the discussion period on the last day of the Photochemical Reactivity Workshop,
moderator Jake Hales explained the NARSTO organization and asked if there was interest in
forming a group under NARSTO to explore research needs concerning reactivity. There was
tentative interest among attendees. Interested persons were asked to remain in the room after
the Workshop itself was adjourned to discuss possible formation of such a group. At this time,
participants did not identify specific research tasks to sponsor. Here are the minutes of that
meeting.
Ad Hoc Task Force on VOC Reactivity
Minutes of Initial Meeting
May 14,1998
The meeting was called to order at around 1:00 pm by Robert Wendoll, who stated that
the first order of business was to establish the initial steps for Task-Force action.
John Dege noted that we need to establish EPA's position on this topic, and then we need
to formulate a concept paper to establish the basis for our downstream operations. It was
emphasized that we need to set forth a list of the relevant research needs. Cyril Durrenburger
suggested that we might want to develop one or more "issue papers" that define the primary
considerations at hand. He also suggested that we consider commissioning several "critical
review" papers, similar to those for the NARSTO Ozone Assessment, as a means of codifying
salient scientific aspects. Gary Foley stated that we need to start thinking about focusing our
scientific efforts, e.g., chamber studies, modeling efforts ..., in order to maximize relevance to
the policy community; we need to reach closure on how to produce the most definitive
information. He also recommended parallel, evolving, and communicative efforts between the
scientific and policy communities.
There was a general discussion of how the science/policy interface should be handled.
This arose at several points during the ensuing conversation. In particular it was asked whether
NARSTO had any direct chain-of-command linkage to EPA's Office of Air and Radiation. Jake
Hales said definitely no. Although OAQPS is a NARSTO signatory, NARSTO tries to be
scrupulous in observing the line between policy-making, and performing policy relevant
research. In observing this line, NARSTO - in Jake's estimation - has been less effective than
desirable to date in getting our research products conveyed to the policy community. Currently
NARSTO is designed to make this linkage through a Liaison Team for Policy, a standing box on
the NARSTO organization chart. This has been relatively ineffective, however, and because of
this, NARSTO's Executive Steering Committee feels that in the future such liaisons need to be
hard-wired into active, functioning groups. In view of this, Jake stressed the importance of
having OAQPS staff take a strong and active role directly in this Reactivity Task Force.
Bob Avery stated that we need to plan a meeting to formulate a mission statement
Robert Wendoll agreed, saying that we need to establish a list of meaningful scientific objectives
to go along with such a statement.
B-l
-------
Howard Feldman stated that it is desirable to develop an assessment of the current state of
the science associated with the reactivity issue. Is the scientific underpinning sufficiently "ripe"
for policy application? If not, when will it be?
It was asked if this Task Force is to be considered a science group, a funding group, or
both. If it's a funding group, what are the money sources? Basil Dimitriades responded that we
need some time after this meeting to think this over; then we should reconvene to write a
research plan.
Referring to the question of whether this effort should be incorporated as a part of
NARSTO, Ed Edwards recommend that everyone here review NARSTO's structure and
operational process. Jake Hales commented that the best way to do this is to visit the NARSTO
Web site on «htpp://narsto.owt.com/Narsto/>. Robert asked for a show of hands for those
favoring incorporation into NARSTO. The response was ambivalent, mainly because of
unfamiliarity of many of the attendees with NARSTO. There was a consensus that everyone
should visit the Web site in the near future.
There were some questions regarding alternatives to NARSTO, such as a possible FACA
[Federal Advisory Committee Act] committee or a dedicated EPA-coordinated arrangement. Jim
Vickery responded that EPA definitely prefers the NARSTO route to a dedicated EPA option, for
several reasons. First, NARSTO was established to promote public/private communications and
offers an established resource base for operations of this sort. Secondly, EPA desires to operate
on a multiorganizational, pooled resource basis with operations of this type, in order to
encourage all interested parties to enter in the discussion process. FACA arrangements, on the
other hand, are more suitable to short-term issues such as evaluating proposed standards and
similar concerns.
Robert Wendoll then asked for a list of action items for this initial meeting. These items
and their resolutions appear below:
1. Set date and place for our follow-on meeting. It was agreed that this meeting should
be during the first week in September at RTP, NC, at EPA's conference facilities. A pilot team
will convene by conference call in early June to draft a mission strategy and design this meeting.
This team will consist of Barbara Frances, Ed Edwards/Robert Wendoll, John Dege, John
Schwind, Cyril Durrenburger, Basil Dimitriades, and Ron Patterson (or Jake Hales).
2. Draft a Mission Strategy. This will be performed by the pilot team, as noted above.
3. Draft an agenda for the September meeting. Basil Dimitriades will do this.
4. Determine methods for communication. Jake Hales suggested that, for the time-being,
at least, the group use the NARSTO home page as a primary communication medium. He will
set up a reactivity sub-page there for that purpose. There was also some question of how we
B-2
-------
communicate activities of this group to individuals not present at this meeting. Jake suggested
that, as a first measure, we put an article describing our activities to-date in the NARSTO
newsletter, the 1998 summer/fall issue of which will go into press in early June. Robert Wendoll
agreed to write this article and send it to Jake for inclusion in this issue.
The meeting concluded at 1:40 pm.
B-3
------- |