United States
Environmental Protection
Agency
Office Of Air Quality
Planning And Standards
Research Triangle Park, NC 27711
EPA-454/R-00-030
April 2001
Air
HOT MIX ASPHALT PLANTS
EPA STAKEHOLDERS OPINIONS REPORT
-------
EPA-454/R-00-030
HOT MIX ASPHALT PLANTS
STAKEHOLDERS OPINIONS REPORT
This document was prepared by:
Emissions Monitoring and Analysis Division
Office of Air Quality Planning and Standards
United States Environmental Protection Agency
Research Triangle Park, NC
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
April 2001
-------
DISCLAIMER
This report has been reviewed by the Emissions Monitoring and Analysis Division of the
Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency, and has
been approved for publication. Mention of trade names or commercial products is not intended to
constitute endorsement or recommendation for use.
-------
PREFACE
This report was produced by the Source Measurement Technology Group of EPA's Emissions
Measurement Center located in Research Triangle Park, NC. It is one of a series of twelve reports
prepared to document an EPA program to characterize emissions to the air from hot mix asphalt plants.
These twelve reports and their associated EPA document numbers and publication dates are:
Document Title
Hot Mix Asphalt Plants
Emission Assessment Report
Hot Mix Asphalt Plants
Kiln Dryer Stack Instrumental Methods Testing
Asphalt Plant A, Gary, North Carolina
Hot Mix Asphalt Plants
Kiln Dryer Stack Manual Methods Testing
Asphalt Plant A, Gary, North Carolina
Volume 1 of 2
Volume 2 of 2
Hot Mix Asphalt Plants
Kiln Dryer Stack Instrumental Methods Testing
Asphalt Plant B, Clayton, North Carolina
Hot Mix Asphalt Plants
Kiln Dryer Stack Manual Methods Testing
Asphalt Plant B, Clayton, North Carolina
Volume 1 of 2
Volume 2 of 2
Hot Mix Asphalt Plants
Truck Loading and Silo Filling Instrumental Methods Testing
Asphalt Plant C, Los Angeles, California
Hot Mix Asphalt Plants
Truck Loading and Silo Filling Manual Methods Testing
Asphalt Plant C, Los Angeles, California
Volume Iof8
Volume 2 of 8
Volume 3 of 8
Volume 4 of 8
Volume 5 of 8
Volume 6 of 8
Volume 7 of 8
Volume 8 of 8
Hot Mix Asphalt Plants
Technical Systems Audit of Testing at Asphalt Plant C
Asphalt Plant C, Los Angeles, California
EPA Document
Number
EPA454/R-00-019
EPA 454/R-00-020
EPA454/R-00-021a
EPA454/R-00-021b
EPA 454/R-00-022
EPA 454/R-00-023a
EPA 454/R-00-023b
EPA 454/R-00-024
EPA 454/R-00-025a
EPA 454/R-00-025b
EPA 454/R-00-025c
EPA 454/R-00-025d
EPA 454/R-00-025e
EPA 454/R-00-025f
EPA 454/R-00-025g
EPA 454/R-00-025h
EPA 454/R-00-026
Publication Date
December 2000
April 2000
April 2000
April 2000
April 2000
April 2000
April 2000
May 2000
May 2000
May 2000
May 2000
May 2000
May 2000
May 2000
May 2000
May 2000
May 2000
-------
Document Title
Hot Mix Asphalt Plants
Truck Loading Instrumental Methods Testing
Asphalt Plant D, Barre, Massachusetts
Hot Mix Asphalt Plants
Truck Loading Manual Methods Testing
Asphalt Plant D, Barre, Massachusetts
Hot Mix Asphalt Plants
Response to Comments on Testing Program for Asphalt Plants
CandD
Hot Mix Asphalt Plants
Stakeholders Opinions Report
EPA Document
Number
EPA 454/R-00-027
EPA 454/R-00-028
EPA 454/R-00-029
EPA 454/R-00-030
Publication Date
May 2000
May 2000
May 2000
April 2001
These documents, including this Stakeholders Opinions Report document, are available for downloading,
on CD-ROM and in paper.
Downloads can be made from:
http//www.epa.gov/ttn/emc/asphalt.html
Copies of the CD ROM can be requested by mail at:
Emission Measurement Center, MD-19
US Environmental Protection Agency
Research Triangle Park, NC 27711
Paper copies of the reports can be obtained from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Phone orders 1-800-553-6847 or (703) 605-6000; FAX orders (703) 605-6900
http://www.ntis.gov/products/environment.htm
-------
ACKNOWLEDGMENTS
Many individuals contributed to the development of this report. Ron Myers of the Emission
Measurement Center's Source Measurement Technology Group of the Emissions Monitoring and Analysis
Division, Roy Smith of the Risk and Exposure Assessment Group of the Emissions Standard Division and
Gary Foureman of the National Center for Environmental Assessment of the Office of Research and
Development are the primary authors of the report. Bob McConnell of EPA's Region I office, David
Mobley, Acting Director of EPA's Emissions Monitoring and Analysis Division, Bill Lamason, Mike
Toney, Gary McAlister, and Candace Sorrell of EPA's Emission Measurement Center, Ron Ryan and
Dennis Beauregard of EPA's Emission Factor and Inventory Group, Laura Autry of EPA's Air Quality
Trends Analysis Group, Jim Crowder and Mary Johnston of the Minerals and Inorganic Chemicals Group
of the Emissions Standard Division participated in the review.
-------
Preceeding Page Blank
Contents
Chapter 1
1.0 Project Background 1-1
Chapter 2
2.0 EPA Response to Stakeholder Opinions 2-
2.1 Response to opinions of Wisconsin Department of Natural Resources 2-
2.2 Response to opinions of Connecticut Department of Environmental Protection 2-
2.3 Response to opinions of Coalition Against the Asphalt Plant et al 2-
2.3.1 Introduction (EPA interactions with citizens) 2-
2.3.2 Why are many citizen groups concerned about asphalt plants? 2-2
2.3.2.1 Asphalt health effects, epidemiological studies, and sound science 2-2
2.3.2.2 Differences in approach to assessing health impacts 2-2
2.3.2.3 Use of "best performance data" from sources 2-3
2.3.3 Unresolved issues 2-3
2.3.3.1 Volatility content and temperature issues 2-3
2.3.3.2 Issues relating to collection and analysis of emissions data 2-5
Background correction and enclosure problems 2-5
"EPA's flat denial of errors in the report" 2-6
"Unfounded leaps of faith 2-6
Problems with witnessing tests 2-6
2.3.4 Conclusions 2-8
23.4.1 Performance of epidemiological study by EPA 2-8
2.3.4.2 Include additional warnings in AP-42 2-8
2.3.4.3 Get tough with industries on excessive claims of CBI 2-8
2.4 Response to opinions of National Asphalt Paving Association 2-9
Chapter 3
Stakeholder Opinions on Hot Mix Asphalt Plants Emission Assessment Report EPA 454/R-00-019,
December 2000 3-1
Wisconsin Department of Natural Resources
Connecticut Department of Environmental Protection
Citizens Against the Asphalt Plant et al
National Asphalt Pavement Association
Chapter 4
Hot Mix Asphalt Plants Response to Comment on Draft Emissions Assessment Report, AP-42 Section
and AP-42 Background Report 4-1
Chapter 5
Stakeholder Comments on Draft Emissions Assessment Report, AP-42 Section and AP-42 Background
Report 5-1
Stephen F. Lynch
Coalition Against the Asphalt Plant
Massachusetts Department of Health
National Asphalt Pavement Association
Norman Ostroff, PhD, PE
Steven Zemba
vii
-------
Preceeding Page Blank
STAKEHOLDERS OPINIONS REPORT
Chapter 1
1.0 Project Background
This is the last in a series of EPA reports that document the development of methods to estimate
emissions from hot mix asphalt plants. The primary purpose of this report is to publish the opinions of the
stakeholders on the final "Hot Mix Asphalt Plants - Emissions Assessment Report." The stakeholders
involved in the processes leading up to the publication of the Emissions Assessment Report include national
and State legislators, State and local environmental and health agencies, national and regional industry
associations, citizen public interest groups and industry technical consultants. As an adjunct to the
opinions of the stakeholders on the final Emissions Assessment Report, this report includes EPA's
responses to stakeholder issues that had not been documented previously or that EPA believes require
additional clarification. Additionally, the stakeholder comments on the draft Emission Assessment Report
and EPA's responses to these comments are included to round out the stakeholder opinions and EPA's
assessment of the technical issues being raised.
The final "Hot Mix Asphalt Plants - Emissions Assessment Report" was published in December
2000. This report contains EPA's final emissions estimates from the two hypothetical facilities of average
production using the two predominant US production methods. This report also includes the revised AP-42
section, the revised AP-42 background report and the July 1996 Emission Inventory Improvement Program
(EIIP) report for hot mix asphalt plants. The AP-42 section contains the final emission estimation methods
and the background report provides detailed information on the development of these emission estimation
methods. Additionally, the background report contains summary information from all reference material
(including more than 350 emission tests) used in the development of the emission estimation methods. The
1996 EIIP report describes preferred and alternative emission estimation techniques and provides concise
example calculations to aid in the preparation of emission inventories.
The Emissions Assessment Report presents a comprehensive inventory with respect to the emission
sources characterized and the pollutants addressed. The emission sources characterized include the kiln
stack, hot mix asphalt (HMA) storage silo, truck load-out, hot asphalt storage tank, hot asphalt storage
tank heater, aggregate handling, RAP crushing, diesel exhaust, and road dust. For each of these sources,
the emissions data presented address the criteria pollutants (PM, VOC, CO, SO2 and NO x) and many of
the Clean Air Act hazardous air pollutants (19 PAH compounds, 27 volatile HAP compounds and 10 HAP
metals). While most of the emission estimates were based upon methodologies presented in the AP-42
section for hot mix asphalt plants, some emission estimates were based upon methodologies in other AP-42
sections (material handling, road dust and mobile sources) or journal articles (mobile source HAPs).
Emissions characterization data from three separate projects were utilized in drafting the
Emissions Assessment Report. Although each of the projects involved characterizing air emissions from
hot mix asphalt plants, each had a somewhat different goal. The goal of one project was to obtain and use
existing emissions test data from State and local air pollution control agencies to improve AP-42 emission
factors for hot mix asphalt plants. The second project was undertaken to determine whether there were any
previously unidentified kiln stack HAP emissions and, if such were present, to quantify these emissions.
1-1
-------
The third project was designed to characterize the fugitive emissions resulting from silo rilling and
transport truck loading operations.
1-2
-------
Chapter 2
2.0 EPA Response to Stakeholder Opinions
This chapter presents EPA's responses to the stakeholder opinions presented in Chapter 3 of this
report. For the purposes of brevity, a detailed response to all issues raised by stakeholders was not
prepared. However, each major issue cited by a stakeholder has been thoroughly evaluated and a brief
response is presented. For issues where there is previous documentation of the issue including EPA's
evaluation and response, the reader is referred to the appropriate section(s) in the prior report(s).
2.1 Response to opinions of Wisconsin Department of Natural Resources
EPA agrees with Wisconsin's point that the development of the additional technical information
that was presented in the Emissions Assessment Report was an ambitious project. Additionally, EPA
reviewed the guidance for emission inventories for hot mix asphalt submitted with Wisconsin's opinion and
will forward this document to the EIIP Point Source chair person for the committee's consideration.
2.2 Response to opinions of Connecticut Department of Environmental
Protection
EPA recognizes that representatives of the Connecticut Department of Environmental Protection
(DEP) were active participants in the development stages of the study, witnessed the emission test at Plant
D, and participated in the review process of the test reports and subsequent technical analyses. EPA
appreciates Connecticut DEP's recognition that the scope of the project was comprehensive, the collection
and analysis protocols were appropriate, the objectives were well defined and the EPA staff were very
capable.
2.3 Response to opinions of Coalition Against the Asphalt Plant et al
This section responds to the opinions of the citizen groups that were presented in the "Minority
Report on Fugitive Emissions from Asphalt Plants" (Minority Report). We have organized and presented
our responses based on the headings in the Minority Report.
2.3.1 Introduction (EPA interactions with citizens)
The citizens have provided a reasonable synopsis of the opportunities EPA afforded to all the
stakeholder groups to assist in the development of emissions estimation methodologies for fugitive emission
sources at hot mix asphalt plants. However, in the second half of the introduction and several other
locations in the Minority Report, the citizens disparage EPA's interactions with the citizens. We believe the
record shows that EPA provided an open and meaningful public participation process in the development
and performance of the emission testing program as well as for the review of data generated from the
testing. In doing so, EPA strived to balance its obligation to ensure responsible use of public monies with
the numerous requests made by the citizens groups, many of which added to the cost of the testing program.
A genuine effort was put forth, aided by facilitators brought in to assist the group discussions and to reach
outcomes agreeable to all parties. We believe we succeeded in that regard to the greatest extent possible.
2-1
-------
2.3.2 Why are many citizen groups concerned about asphalt plants?
2.3.2.1 Asphalt health effects, epidemiological studies, and sound science
In serving the public, the Agency has an obligation to use sound science in forming opinions on the
toxicity of chemical compounds. One aspect of applying sound science is to utilize and/or consider all of the
available credible scientific information on the toxicity of a given compound; credible scientific information
includes reports published in peer-reviewed scientific journals, conference proceedings, or other similar
peer-reviewed vehicles. Anecdotal information such as that referred to in the second paragraph of this
section of the Minority Report is, by definition, information that does not originate from a scientifically
peer-reviewed report and must be considered only with reservation. The recently published NIOSH hazard
review on the health effects of occupational exposure to asphalt (http://www.cdc.gov/niosh/01-110pd.html)
is a good example of an inclusive study of credible scientific information; in particular, it cites and evaluates
about 40 studies in an effort to present an overall view of all of the credible scientific information on the
acute and chronic health effects of asphalt emissions.1 This report concludes that current data are not
sufficient for quantifying the acute and chronic health risks of exposure to asphalt, asphalt-based paint, or
asphalt fumes and vapors. The NIOSH report also cites specific limitations concerning the Hansen cohort,
the one specifically cited in the Minority Report. The conclusions of the NIOSH report are of utility to EPA
especially as NIOSH is concerned with worker populations (e.g., pavers and roofers) that are more heavily
exposed to asphalt emissions than the general population, and in whom any adverse effects would be more
likely to occur. Any future formal review of the health effects of asphalt emissions by EPA would closely
approximate the sound science effort by NIOSH.
2.3.2.2 Differences in approach to assessing health impacts
The authors of the Minority Report express concern that the Agency is concerned solely with
long-term average exposures and not "... peak, short-term exposures for acute effects" We acknowledge
that the exposure guidelines issued by EPA for air pollutants concerning cancer and noncancer effects are
concerned with long-term exposures and, although a methodology has been under development for several
years, EPA has yet to institute methodologies for evaluating effects from short-term exposures.
Additionally, as stated clearly in the introduction to AP-42, the emission factors in AP-42 are average
values since its' primary purpose is to support the development of area wide emission inventories.
However, the emission factors for fugitive emissions at hot mix asphalt plants provide adequate flexibility
to estimate short term emission rates when combined with reasonable information on asphalt volatility and
temperature. Furthermore, reasonable estimates of short term stack emissions associated with normal plant
operations can be derived from the available supporting data. The authors of the Minority Report may
wish to consider the conclusions of NIOSH in their recently promulgated Hazard Review on asphalt
emissions. After considering the currently available information on workers under short-term exposure
conditions, their recommendation for worker exposure over a 15 minute period was not altered from their
original estimate determined in 1977, at the rather high level of 5 mg/m3 for total particulate.
1Hazard Review, Health Effects of Occupational Exposure to Asphalt; U.S. Department of Health
and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute
for Occupational Safety and Health; DHHS (NIOSH) Publication No. 2001—110, December 2000.
2-2
-------
2.3.2.3 Use of "best performance data" from sources
EPA does not agree with the citizens statement that "much of the data used for developing stack
emission factors was collected just after initial plant startup often during a preoperating test period where
the plant owner is demonstrating to the state regulatory authorities that the plant can operate for short
periods of time at the design operating rate while staying within the permitted emission limits." While some
of the data used were from tests performed within 60 days of the plants' initial start up (known as initial
compliance tests) and may be representative of the most favorable operating conditions, most of data were
from hot mix asphalt plant emission tests performed on facilities that had been in operation for many years.
The data utilized were drawn from 278 tests reports for emission tests performed between 1987 and 1994.
Given the slow production growth in the industry, it is unlikely that more than half of these tests were initial
compliance tests. Generally, emissions tests are only conducted during periods of representative
performance. EPA and most State and local agencies do not consider periods of startup, shutdown and
malfunction to be representative conditions. As a result, testing is usually not conducted during these
events. By definition, periods of non-representative operations are brief. Additionally, it is likely that a
facility being forced to test because of compliance concerns would correct any significant problem prior to
conducting the test. However, facilities that are compliance concerns are the exception. While it is
recognized that these worst case conditions are not represented in the supporting data, the available
supporting data is representative of the general range of facilities in operation. This is substantiated by the
wide range of the data for those pollutants where there are more than about ten data points supporting the
emission factor.
2.3.3 Unresolved issues
2.3.3.1 Volatility content and temperature issues
The citizens are concerned about the volatility content of asphalts and hot mix asphalt load-out
temperature and present five specific issues as the following five questions. We address these concerns
below specifically in regard to their impact on estimation of emissions using the emission estimation
methodologies presented in Appendix A of the "Hot Mix Asphalt Plants - Emission Assessment Report,"
EPA 454/R-OO-019, which is the text of Section 11.1 of AP-42, Hot Mix Asphalt Production.
a. Why did the industry change the specification for the maximum allowable volatile content of
asphalt from 0.5% to 1.0%? What are the implications of this change?
b. How is this volatile content measured now and how was it measured in the past?
c. How is the volatile content affected by seasonal variations in oil refining practice?
d. How is the volatile content affected by the various classes of additives that are sold to improve the
performance of asphalt?
e. Are the maximum operating temperatures recommended by industry groups actually followed by
plant operators? If not, what are the variations in temperature and what is the consequence of
these variations?
Response to questions a and b
As the citizens note, the industry has recently started using ASTM Method D2872-88, Standard
Test Method for Effect of Heat and Air on a Moving Film of Asphalt - Rolling Thin-Film Oven Test -
2-3
-------
RTFOT for assessing asphalt volatile content. In the past, the method commonly used was ASTM Method
D1754-97, Standard Test Method for Effect of Heat and Air on Asphaltic Materials - Thin-Film Oven Test
-TFOT). All information used to develop the emission estimation methodology for estimating load-out and
silo filling emissions, used to establish the default values for asphalt volatility which may be used in
estimating load-out and silo filling emissions when no volatility data are available, and specified for use in
the emission equation rely on the volatile content as determined by the RTFOT method. Therefore, the
industry's change in the specification for the maximum allowable volatile content of asphalt from 0.5% to
1.0% is not relevant to appropriate application of the emission estimation methodologies nor are the
differences in the loss on heating obtained using the two methods of concern.
Response to questions c and d
The five State DOT laboratory directors contacted for this project have stated that the independent
volatility assessments made by the states using the RTFOT method are conducted on samples of asphalt
that are to be used without further blending or addition of additives. State DOT's routinely perform a
variety of quality assurance tests including the RTFOT method on a representative sampling of asphalts to
insure that materials used to construct State and Federal roadways in their State meet required
specifications. There are potential legal ramifications to industry for changing the physical properties of
asphalts after the State quality assurance samples have been obtained. Since State agencies obtain and
independently determine the volatility of samples of asphalt which will be used without further
modification, we conclude that there is no potential impacts of seasonal refining practices, crude stocks,
and asphalt blending and additives in regard to the application of the emission estimation methodologies.
This information available from State DOT's is superior to the use of assumed values and should be more
than adequate in evaluating the variations in asphalt volatility over time.
Response to question e.
Since the emission estimation methodology for load-out and silo filling incorporates a variable for
actual measured load-out temperatures, the effect on emissions of exceeding the industry recommended
temperature guidelines as well as other load-out temperatures can be assessed (see, for example, Table 1 in
the Minority Report). The use of measured load-out temperature data is superior to the use of theoretical
extremes for load-out temperature. Facilities and/or their customers routinely monitor the hot mix asphalt
temperatures loaded into the trucks before leaving the plant property. If measured load-out temperatures
are not available, load-out temperatures for a given time can be approximated from other available
temperature records. Hot mix asphalt plants, for example, routinely monitor the mix temperature of their
heated aggregate and drum mix plants monitor the temperature of the product stored in silos. By
correlating historical measured mix temperature or silo temperatures with historical measured load-out
temperatures and combining this with plant records of mix temperature/silo temperature and RAP usage it
is possible to obtain improved estimates of load-out temperatures.
General
We believe it may also be helpful to clarify the criteria for selecting appropriate values for asphalt
volatility and load-out temperature. The values selected should not be arbitrary (e.g., based upon industry
desired targets or industry or State maximum allowable specifications) but instead should be based upon
legitimate verifiable data that is suitable for the health or environmental endpoint of interest. For example,
health or environmental endpoints that are the result of long term exposures should rely on emissions
estimations based upon a measure of central tendency (e.g., average, median, mode or other appropriate
2-4
-------
statistical measure) of the asphalt volatility and load-out temperature. Conversely, if the health or
environmental endpoint is the result of a short term exposure, a combination of reasonable upper bound
values for volatility and load-out temperature should be used to estimate emissions. As stated in the last
sentence of the first paragraph on page 105 of the "Response to Comment on Draft Emissions Assessment
Report, AP-42 Section and AP-42 Background Report," the selection of these upper bound values should
include an assessment of the probabilities of simultaneous events occurring. While the use of multiple
worst case assumptions (as depicted in the two columns labeled "Citizens" in Table 1 of the Minority
Report) may be a shortcut to evaluate maximum potential emissions, generally these emission estimates are
an unlikely result.
Previous responses or relevant information on asphalt volatility and temperature issues are
contained in Chapter 4 of this report (see Section 3.2 from page 54 through page 63, Section 3.5.34 on
pages 88 and 89, Section 3.6.4 on pages 104 and 105, Section 3.6.6 on page 105 and Section 3.6.10 on
page 106). In addition, the document EPA 454/R-00-029, "Hot Mix Asphalt Plants Response to
Comments on Testing Program for Asphalt Plants C and D" published in May 2000 also contains
responses and relevant information on these volatility and temperature issues (Item 5 on pages 9 and 10 and
Items 52 and 53 from page 61 through 68). Lastly, Appendix B of the "Hot Mix Asphalt Plants - Emission
Assessment Report, EPA 454/R-00-019 which is the "Emission Factor Documentation for AP-42 Section
11.1, Hot Mix Asphalt Production," clearly presents the methodology used as the basis for the final
emission factor equations recommended in the AP-42 section and the development of long term average
default values for volatility and temperature (Section 4.4.1, 4.4.2, 4.4.3 and 4.4.4 from page 4-114 through
4-123 and Tables 4-27 through 4-44 from page 4-286 through 4-307).
2.3.3.2 Issues relating to collection and analysis of emissions data
Background correction and enclosure problems
The citizens identify three issues related to the collection and analysis of the fugitive emissions
data. These issues include (1) the quality of the background emissions data obtained at Plant C, (2) the
methodology used to correct for background emissions due to truck operation (diesel exhaust and road dust
emissions), and (3) the effectiveness of the enclosure at Plant D to capture essentially all of the truck load-
out emissions. The citizen groups have raised each of these issues during meetings held in Boston, in their
written comments on the draft test reports for Plants C and D, and in their comments on the Draft
Emissions Assessment Report. We understand the concerns of the citizens, the supporting arguments of the
citizens, and their proposed methods to address these concerns. However, we remain unconvinced by the
arguments that the collection and analyses of the fugitive emissions are flawed.
Previous responses and relevant information on the issues expressed in the Minority Report on the
collection and analysis of emissions data are contained in Chapter 4 of this report (see Section 3.1 from
page 43 through page 54, Section 3.3 from page 63 through 70 and Section 3.5.38 on pages 90 and 91). In
addition, the EPA Report No. 454/R-00-029, "Hot Mix Asphalt Plants Response to Comments on Testing
Program for Asphalt Plants C and D" published in May 2000 also contains responses and relevant
information on these collection and analysis issues (see in particular Item 11 on pages 17 and 18, Item 13
on page 19, Item 15 from page 21 to 26, Item 41 from page 41 through 50, Item 51 from page 56 through
61, Item 55 on pages 68 and 69, Item 57 on page 70 and Item 60 from page 72 through page 75).
2-5
-------
"EPA's flat denial of errors in the report"
In one section of the Minority Report entitled "EPA's flat denial of errors in the report," the
citizens express dissatisfaction with EPA's earlier response to their prior comments regarding errors in the
draft reports distributed for review by the stakeholders. In particular, the citizens express dissatisfaction
with "incorrect references, missing and incorrectly referenced appendices and numerical and logical errors
in the analysis." We recognize that there were a few incorrect references and missing and incorrectly
referenced appendices which we have acknowledged in the past and since corrected: nevertheless, we do not
believe that these should have significantly affected the stakeholders reviews of the approximately 14 draft
reports distributed. We appreciate the citizens identification of the two more significant errors they cite
which we previously corrected. The first error was created when calculations to determine average
emission concentrations for one of three test runs on a silo filling operation included time periods where no
asphalt was being loaded into the storage silo; the second was an incorrect maximum concentration for the
intermittent load-out test reported in an appendix of the report for Plant C which fortunately had no effect
on data elsewhere in the report. We routinely distribute draft reports to all interested stakeholders for their
review to integrate the various perspectives and ensure that we have not made inappropriate assumptions,
overlooked relevant information, etc. Considering the large volume of data processed for this project and
the tight time constraints on its generation, we do not find these few errors unreasonable.
"Unfounded leaps of faith"
The citizens believe the selection of docosane and tricosane as model compounds for asphalt
emissions from storage tanks is an "unfounded leap of faith." They point out that docosane and tricosane
are not the major measured constituents of asphalt emissions and that EPA chose these compounds because
their Antoine's coefficients produce a working loss estimate of 32 pounds per million gallons of asphalt
throughput for their approximate molecular weight. The use of docosane and tricosane as the model
compounds in the development of Antoine's constants for asphalt for use in the TANKS program to
estimate working losses from storage tanks is based on sound scientific principles. EPA recognizes that the
specific compounds used are neither the major constituents of asphalt vapors nor major constituents of
liquid asphalt. As the citizens imply, our selection of these compounds to derive Antoine's constants for
asphalt is based in part on the fact that the Antoine's constants for the compounds produce working loss
emission estimates consistent with the actual test data from Plant C. In other words, when using the
Antoine's constants and the molecular weights developed, the TANKS program predicts the emissions
calculated from the maximum vapor concentrations for asphalt measured during silo emission testing at
Plant C. Each step of the development of the Antoine's constant is clearly described in Section 4.4.5
beginning on page 4-123 of Appendix B of the "Hot Mix Asphalt Plants - Emission Assessment Report,"
EPA 454/R-00-019 which is the "Emission Factor Documentation for AP-42 Section 11.1, Hot Mix
Asphalt Production." Previous responses or relevant information on emissions estimations methods for
asphalt storage tanks are contained in Chapter 4 of this report (Section 3.5.44 on page 93, Section 3.5.50
from page 96 through page 99, Section 3.7.2 on pages 115 and 116, Section 3.7.4 on page 116). While the
calculational details are preliminary, this basic methodology is described in the response to item 17 from
page 26 through 28 in the document EPA 454/R-00-029, "Hot Mix Asphalt Plants Response to Comments
on Testing Program for Asphalt Plants C and D" published in May 2000.
Problems with witnessing tests
The citizens raise a number of issues with regard to citizen observation of the emission tests at
Plants C and D: (1) it was difficult for the single witness allowed at Plant C to simultaneously observe
2-6
-------
operations at several places in the plant; (2) certain information was withheld from the citizen observer on-
site under a claim of confidentiality, (3) citizens were forbidden to take photographs to document their
observations, and (4) an EPA staff member commented during a dinner discussion that EPA had no
intention of regulating fugitive emissions from the asphalt industry.
The citizens were allowed unprecedented access to both Plant C and D. All process and emissions
sampling data were freely available to the observer for review during the test program. While EPA's
Confidential Business Information (CBI) procedures precluded providing a hard copy of this information to
the observer, hard copy was available later after it cleared CBI review. In addition, the citizen observer
was considered an integral part of the test team during on site decision meetings to discuss issues that could
potentially affect the test program. During both of the emission tests, the number of observers (citizen
plus agency) greatly exceeded the number of operators at the plants. To minimize competing requests for
the plant operators' attention, access to the plant operator was limited to the individual responsible for
obtaining the process operation data who was responsible for keeping the remainder of the test team
informed of plant operation changes. This limitation applied not only to the citizen observer but to the EPA
observers, the three EPA contractor organizations, and the industry observers as well.
While no photographs were taken prior to correcting the two ventilation system problems that
created the visible emissions condition observed by Dr. Nadkarni on the day before the first complete truck
load-out emission test run at Plant C, documentation (capture efficiency measurements and photographs)
have been supplied to the stakeholders that confirm that the capture of the emissions was not 100% and the
capture efficiency measurements have been used to correct the data for the emissions testing at this
emission point. Further responses to comments concerning the capture of emissions at both Plants C and D
are included in items 11, 13 and 14 of the report "Hot Mix Asphalt Plants Response to Comments on
Testing Program for Asphalt Plants C and D" (EPA 454/R-00-029) and sections 3.1.2 and 3.1.4 in
Chapter 4.
The citizens' assertion that EPA prejudged the final outcome of the project is incorrect. During the
many teleconferences and meetings with the stakeholders (citizens, industry, and State agencies) prior to the
emissions test, the EPA office responsible for regulatory decisions clearly identified the criteria established
by the Clean Air Act Amendments of 1990 (CAAA) for developing emissions regulations. In particular,
they explained that two major criteria that trigger development of a regulation are (1) a determination that
one or more facilities in a source category are a "major source" of hazardous air pollutants (HAPs) and (2)
that there is a demonstrated technology to control these HAP emissions. It was also explained to the
stakeholders, that the CAAA defines a "major source" as any contiguous facility with emissions of more
than 10 tons of any one HAP or 25 tons of any combination of HAPs. The objective of testing at both
Plants C and D was to quantify the fugitive HAP emissions from hot mix asphalt plants and thereby
determine whether any hot mix asphalt plants could be a "major source." Considering the substantial
emission testing resources being expended to quantify emissions from hot mix asphalt plants in order to
make the "major source" determination and that fact neither of the EPA staff present at the dinner is
responsible for regulatory decisions, it is unlikely that they would have stated that EPA had no intention of
regulating hot mix asphalt plant emissions.
2-7
-------
2.3.4 Conclusions
2.3.4.1 Performance of epidemiological study by EPA
The Minority Report recommends that the government conduct community-based epidemiological
studies to elucidate health effects from exposure to asphalt emissions. Justification for such a study would
be helped by further definitive studies in worker populations in whom any adverse effects would be more
likely to occur. Results from such a study among workers could then trigger and provide direction for
studies at the community level. The authors of the Minority Report may want to refer to the work of
Partanen, et al. (1995)2 and Burstyn, et al. (2000)3 which describe the feasibility of such studies among
asphalt workers in more depth.
2.3.4.2 Include additional warnings in AP-42
The citizens recommend that EPA include the limitations regarding individual emission factors in
each table in AP-42. As the citizens point out, EPA recognizes the great reliance on AP-42 emission
factors by industry and state agencies in permitting and other processes. The introduction to AP-42
currently provides a detailed discussion regarding limitations for using the emission estimation
methodologies contained in the numerous tables in the more than 200 sections of AP-42. We do not believe
it is reasonable that every table include a statement that the emission estimation methods represent an
average of available data and do not represent "worst case" impact scenarios. Such statements would
introduce additional and unnecessary complexity into the tables that are already complex and such
statements are not likely to eliminate or reduce any existing misuse of the available methodologies.
2.3.4.3 Get tough with industries on excessive claims of CBI
The citizens recommend that the government challenge industry regarding unsubstantiated demands
for confidentiality of business information. We would again point out that all available information was
provided to the stakeholders in sufficient time for evaluation and comment prior to the next decision point
of the process. Prior to their observation of testing at Plant C, the citizens were provided with a copy of the
OAQPS CAA Confidential Business Information (CBI) Security Manual. This manual includes the
procedures followed by OAQPS to insure that CBI that is inadvertently received is protected. We
recognize that much of the information collected during our plant visits and emission tests is unlikely to be
claimed as CBI. However, we must still recognize the potential to collect some limited amount of CBI.
The procedures used by OAQPS to protect and determine the CBI status of information collected during
inspections and testing is described on page 19 of the manual. These procedures incorporate the
requirements of Federal Rules established by Congress for insuring that information EPA needs from
businesses do not infringe on businesses ability to keep selected information confidential. As indicated in
third paragraph section "B. Original CBI," all information collected by EPA during data-gathering visits or
2Partanen, T.J., Boffetta, P., Heikkila, P.R, Frentzel-Beyme, R.R., Heederik, D., Hours, M.,
Jarvholm, E.G., Kauppinen, T.P., Kromhout, H. Langart, S. 1995. Cancer risk for European asphalt
workers. Scand. J. Work Environ. Health 21(4): 252-258.
3Burstyn, I., Kromhout, H., Cruise, P.J. and Brennan, P. 2000. Designing an international
industrial hygiene database of exposures among workers in the asphalt industry. Ann. Occup Hyg. 44(1):
57-66.
2-8
-------
tests must be forwarded to the responsible industry official for review to preclude the inadvertent release of
CBI. The sample letter on page 87 of the manual contains criteria for supporting a claim of confidentiality.
EPA staff have no authority to presuppose what material obtained during a visit or test does not meet these
criteria. Industries are usually provided 45 days to respond to EPA's requests to identify CBI. For all of
the information obtained during the hot mix asphalt fugitive emissions study, industry made no claims of
confidentiality. In addition, the information obtained by EPA was made available to all stakeholders less
than 30 days after it was submitted to the industry for clearance.
2.4 Response to opinions of National Asphalt Paving Association
EPA agrees that the methodologies published in the Emissions Assessment Report have resolved
the question of how best to estimate emissions from truck load-out and silo filling operations. Additionally,
EPA agrees that with the completion of this report there are now reliable emission estimation methodologies
for essentially all air emission sources at hot mix asphalt plants. EPA does not agree that the information
in the Emissions Assessment Report proves that there are no hot mix asphalt plants that are "major
sources." However, when combined with facility production information and the appropriate pollutant
emissions evaluation criteria, the available methodologies in the report will allow State and local agencies
to determine the major source status of a facility.
2-9
-------
Preceeding Page Blank
Chapter 3
Stakeholder Opinions on
Hot Mix Asphalt Plants
Emission Assessment Report
EPA454/R-00-019
December 2000
-------
Lloyd Eagan, Director
Bureau of Air Management
Wisconsin Department of Natural Resources
Madison, WI
-------
WISCONSIN
DEPT. OF NATURAL RESOURCES
State of Wisconsin \ DEPARTMENT OF NATURAL RESOURCES
101 S.Webster St.
Scott McCallum, Governor gox 7921
Darrell Bazzell, Secretary Madison, Wisconsin 53707-7921
Telephone 608-266-2621
FAX 608-267-3579
TTY 608-267-6897
February 28, 2001
Mr. Ron Myers
USEPA-OAQPS (MD-19)
Research Triangle Park, NC 27711
Subject: Asphalt Plant Emissions
Dear Mr. Myers:
We congratulate USEPA on its development of emission factors for hot mix asphalt plants. The Hot Mix
Asphalt Plants - Emission Assessment Report (EPA 454/R-00-019), December 2000, was a very
ambitious project that complements the Emission Inventory Improvement Program (EIIP) report
Preferred and Alternative Methods for Estimating Air Emissions from Hot Mix Asphalt Plants, July 1996.
Wisconsin was a major contributor to the EIIP report.
We would like to supply comments regarding the reporting of emissions from these asphalt plants that
you may want to consider in your final document. The Wisconsin asphalt companies approached DNR in
1996 with a concern regarding the consistency of reporting air emissions across the state. Cooperatively
we developed a document that met their concerns and we have included a copy of this document as an
attachment to this letter.
The attached guidance document specifies equipment common (noted as base in the guidance document)
to all asphalt plants. If a particular asphalt plant had not reported air emissions with that piece of
equipment, that piece of equipment was added as an emission source and asphalt plant was expected to
report air emissions for that piece of equipment. From this base, additional pieces of equipment (noted as
optional in the guidance document) were added to complete the assessment of the asphalt plant. If a
company did not have a base piece of equipment, it could notify WDNR and we would remove this piece
of equipment from our database.
We have found this approach to be very successful in ensuring consistent reporting from asphalt plants
across the state.
Thank you for allowing us to comment. If you should have any questions, please feel free to call Ralph
Patterson of my staff at 608-267-7546.
Sincerely,
Lloyd Eagan, Director
Bureau of Air Management
www.dnr.state.wi.us Quality Natural Resources Management
www.wisconsin.gov Through Excellent Customer Service
Primed on
Recycled
-------
Hot Mix Asphalt Plant
Air Emission Inventory Guidance
20-Sep-99
-------
DNR Disclaimer
This document is intended solely as guidance and does not contain any mandatory
requirements except where requirements found in statute or administrative rule are
referenced. This guidance does not establish or affect legal rights or obligations and is
not finally determinative of any of the issues addressed. This guidance does not create
any rights enforceable by any party in litigation with the State of Wisconsin or the
Department of Natural Resources. Any regulatory decisions made by the Department of
Natural Resources in any matter addressed by this guidance will be made by applying the
governing statutes and administrative rules to the relevant facts.
-------
Hot Mix Asphalt
Emission Inventory
Guidance
In the 1998 air emission inventory there were 133 asphalt plants reporting under SIC
2951 (Asphalt paving and mixtures) across all five DNR regions. It is air management's
emission inventory goal to have air emissions reported consistently for Hot Mix Asphalt
(HMAs) plants across the state. This document has been written so that air management
can achieve this consistency by defining base level equipment and emission sources
common to all HMAs.
This document is a revision of two previous documents. The HMA document was
written in 1996 after discussions with the industry. This document was later updated in
1998. This document uses much of the same information of the prior documents but also
includes summary tables for the HMA source and emission factor.
This guidance was developed by the Central Office and people that deal with hot mix
asphalt plants from the South Central and Southeast Regions. This guidance is only
applicable to the calculation of emissions for the air emission inventory. This guidance is
not intended for use on asphalt plant permits because the focus of permits tends to be on
potential to emit (PTE) and maximum total emissions (MTE) while the air emission
inventory focuses on the calculation of actual emissions. This guidance furnishes. ,
information on the emission factors and information for calculating emissions from each
hot mix asphalt plant. •
DATABASE NEEDS
The Air Emissions Management System (AEMS) is the air management computer system
used to calculate air emissions from all point sources. AEMS requires the identification
of the device (i.e. the piece of equipment), the process (i.e., the type of fuel burned or
material processed), and an eight digit source classification code (SCC) associated with
the process. The database assigns a default emission factor unless some other data is
placed into the database manually. The default emission factors are taken from four
sections of AP-42, 1/95 version. These four sections are 11.1 Hot Mix Asphalt Plants,
1.3 Fuel Oil Combustion, 1.4 Natural Gas Combustion and 1,5 Liquefied Petroleum Gas
Combustion.
USEPA HMA DESCRIPTIONS
[Note: This section taken from AP-42 Section 11.1 Hot Mix Asphalt Plants. Tables 1,
2, and 3 were developed from diagrams of the HMA plants in this section.]
There are different types of HMA set-ups. HMA paving materials can be manufactured
by: (1) batch mix plants, (2) continuous mix (mix outside drum plants), (3) parallel flow
drum plants and (4) counterflow drum mix plants. According to AP-42 Section 11.1 Hot
-------
Mix Asphalt Plants of the total amount of asphalt plants in the United States, 64% are
batch plants, 28% are parallel flow drum mix, and 8% are counterflow drum mix.
Continuous mix plants are rare and were considered insignificant (and thus not
considered) in Section 11.1. If a company has a continuous mix plant please notify the
Central Office and we will determine the typical emission sources after discussions with
the industry. Section 11.1 lists the emission sources for the three types of HMA plants.
1. Batch Mix HMA
Processing begins as the aggregate is hauled from the storage piles and is placed in the
appropriate hoppers of the cold feed unit. The material is metered from the hoppers
onto a conveyor belt and is transported into a rotary dryer (typically gas or oil fired).
Dryers are equipped with flights designed to shower the aggregate inside the drum to
promote drying efficiency.
As the hot aggregate leaves the dryer, it drops into a bucket elevator and is transferred to
a set of vibrating screens where it is classified into as many as four different grades
(sizes), and is dropped into individual "hot" bins according to size. To control aggregate
size distribution in the final batch mix, the operator opens various hot bins over a weigh
hopper until the desired mix and weight are obtained. Reclaimed asphalt pavement may
be added at this point, also. Concurrent with the aggregate being weighed, liquid asphalt
cement is pumped from a heated storage tank to an asphalt bucket, where it is weighed to
achieve the desired aggregate-to-asphalt cement ratio in the final mix.
The aggregate from the weigh hopper is dropped into the mixer (pug mill) and dry-mixed
for 6 to 10 seconds. The liquid asphalt is then dropped into the pug mill where it is
mixed for an additional period of time. Total mixing time is less than 60 seconds. Then
the hot mix is conveyed to a hot storage silo or is dropped directly into a truck and hauled
to the job site.
As with most facilities in the mineral products industry, batch mix HMA plants have two
major categories of emissions: ducted sources (those vented to the atmosphere through
some type of stack, vent, or pipe), and fugitive sources (those not confined to ducts or
vents but emitted directly into the ambient air). Ducted emissions are usually collected
and transported by an industrial ventilation system having one or more fans or air
movers, eventually to be emitted to the atmosphere through some type of stack. Fugitive
emissions result from process and open sources and consist of a combination of gaseous
pollutants and PM.
The most significant source of ducted emissions from batch mix HMA plants is the rotary
drum dryer. Emissions from the dryer consist of water as steam evaporated from the
aggregate, PM and small amounts of volatile organic compounds (VOC) of various
species (including hazardous air pollutants [HAP] derived from combustion exhaust
gases.
-------
Other potential process sources include the hot-side conveying, classifying, and mixing
equipment, which are vented to either the primary dust collector (along with the dryer
gas) or to a separate dust collection system. The vents and enclosures that collect
emissions from these sources are commonly called "fugitive air" or "scavenger" systems.
The scavenger system may or may not have its own separate air mover device, depending
on the particular facility. The emissions captured and transported by the scavenger
system are mostly aggregate dust, but they may also contain gaseous VOCs and a fine
aerosol of condensed liquid particles. This liquid aerosol is created by the condensation
of gas into particles during cooling of organic vapors volatilized from the asphalt cement
in the mixer (pug mill). The amount of liquid aerosol produced depends to a large extent
on the temperature of the asphalt cement and aggregate entering the pub mill. Organic
vapor and its associated aerosol are also emitted directly to the atmosphere as process
fugitives during truck loadout, from the bed of the truck itself during transport to the job
site, and from the asphalt storage tank. In addition to low molecular weight VOCs, these
organic emission streams may contain small amounts of polycyclic compounds. Both the
low molecular weight VOCs and the polycyclic organic compounds can include HAPS.
The ducted emissions from the heated asphalt storage tanks may include VOCs and
combustion products from the tank heater.
There are also a number of fugitive dust sources associated with batch mix HMA plants,
including vehicular traffic generating fugitive dust on paved and unpaved roads,
aggregate material handling and other aggregate processing operations. Fugitive dust
may range from 0.1 \j.m to more than 300 (jm in aerodynamic diameter. On average, 5
percent of cold aggregate feed is less than 74 jam (minus 200 mesh). Fugitive dust that
may escape collection before primary control generally consists of PM with 50 to 70
percent of the total mass less than 74 (im.
Table 1
Batch Mix HMA
Emission Sources
AP-42, 1/95 Figure 11.1.1
Emission Source
Hot screens
Hot bins
Mixer
Elevator
Rotary Dryer
Cold Aggregate Bins
Asphalt Cement Storage Heater
Loader
Fine Aggregate Storage Pile
Coarse Aggregate Storage Pile
Conveyor from Rotary Dryer
Conveyor to Cold Aggregate Bins
RAP Bin Conveyor
Primary Collector
Secondary Collector
Unpaved Haul road
Or
Paved Haul Road
Source Classification Code
3-05-002-02
3-05-002-02
3-05-002-02
3-05-002-02
3-05-002-01
3-05-002-04
3-05-002-06
3-05-002-07
3-05-002-08
3-05-002-04
3-05-002-03
3-05-002-03
3-05-020-06
3-05-020-06
3-05-020-06
3-05-020-33 (proposed)
3-05-020-34 (proposed)
-------
[Note: SCCs 3-05-020-33 and 3-05-020-34 have been proposed by DNR as additions to
the USEPA SCC list.]
2. Parallel Flow Drum Mix Plants
This process is a continuous mixing type process, using proportioning cold feed controls
for the process materials. The major difference between this process and the batch
process is that the dryer is used not only to dry the material but also to mix the heated and
dry aggregates with the liquid asphalt cement. Aggregate, which has been proportioned
by size gradations, is introduced to the drum at the burner end. As the drum rotates, the
aggregates, as well as the combustion products, move toward the other end of the drum in
parallel. Liquid asphalt cement flow is controlled by a variable flow pump electronically
linked to the new (virgin) aggregate and RAP weigh scales. The asphalt cement is
introduced in the mixing zone midway down the drum in a lower temperature zone, along
with any recycled asphalt (RAP) and particulate matter (PM) from collectors.
The mixture is discharged at the end of the drum and is conveyed to either a surge bin or
HMA storage silos. The exhaust gases also exit the end of the drum and pass on to the
collection system.
Parallel flow drum mixers have an advantage, in that mixing in the discharge end of the
drum captures a substantial portion of the aggregate dust, therefore lowering the load on
the downstream collection equipment. For this reason, most parallel flow drum mixers
are followed only by primary collection equipment (usually a baghouse or venturi
scrubber). However, because the mixing of aggregate and liquid asphalt cement occurs
in the hot combustion product flow, organic emissions (gaseous and liquid aerosol) may
be greater than in other processes.
The most significant ducted source of emissions is the rotary drum dryer. Emissions
from the drum consist of water as steam evaporated from the aggregate, PM, and small
amounts of VOCs of various species (including HAPs) derived from combustion exhaust
gases, liquid asphalt cement, and RAP, if utilized. The VOCs result from the incomplete
combustion and from the heating and mixing of liquid asphalt cement inside the drum.
The processing of RAP materials may increase VOC emissions because of an increase in
mixing zone temperature during processing.
Process fugitive emissions associated with batch plant hot screens, elevators, and the
mixer (pug mill) are not present in the drum mix process. However, there may be slight
fugitive VOC emissions from transport and handling of the hot mix from the drum mixer
to the storage silo and also from the load-out operations to the delivery trucks. Since the
drum process is continuous, these plants must have surge bins or storage silos. The
fugitive dust sources associated with drum mix plants are similar to those of batch mix
plants with regard to truck traffic and to aggregate material feed and handling operations.
-------
Table 2
Parallel Flow HMA
Emission Sources
AP-42, 1/95 Figure 11.1.2
Emission Source
Parallel Flow Drum Mixer
Cold Aggregate Bins
Asphalt Cement Storage Heater
Loader
Fine Aggregate Storage Pile
Coarse Aggregate Storage Pile
Conveyor from Parallel Flow Drum Mixer
to scalping screen
Conveyor from scalping screen to Cold
Aggregate Bins
Conveyor from RAP Bin to
Parallel Flow Drum Mixer
Primary Collector
Secondary Collector
Unpaved Haul road
Or
Paved Haul Road
Source Classification Code
3-05-002-05
3-05-002-04
3-05-002-06
3-05-002-07
3-05-002-08
3-05-002-04
3-05-002-03
3-05-002-03
3-05-020-06
3-05-020-06
3-05-020-06
3-05-020-33 (proposed)
3-05-020-34 (proposed)
[Note: SCCs 3-05-020-33 and 3-05-020-34 have been proposed by DNR as additions to
the USEPA SCC list.]
3.
Counterflow Drum Mix Plant
In this type of plant, the material flow in the drum is opposite or counterflow to the
direction of exhaust gases. In addition, the liquid asphalt cement mixing zone is located
behind the burner flame so as to remove the materials from direct contact with hot
exhaust gases.
Liquid asphalt cement flow is controlled by a variable flow pump which is electronically
limited to the virgin aggregate and RAP weigh scales. It is injected into the mixing zone
along with any RAP and particulate matter from primary and secondary collectors.
Because the liquid asphalt cement, virgin aggregate, and RAP are mixed in a zone
removed from the exhaust gas stream, counterflow drum mix plants are will likely have
organic emissions (gaseous and liquid aerosol) that are lower than parallel flow drum mix
plants. A counterflow drum mix plant can normally process RAP at ratios up to 50
percent with little or no observed effect upon emissions.
The most significant ducted source of emissions is the rotary drum dryer in a counterflow
drum mix plant. Emissions from the drum consist of water as steam evaporated from the
aggregate, PM, and small amounts of VOCs of various species (including HAPs) derived
from combustion exhaust gases, liquid asphalt cement, and RAP, if used.
Because liquid asphalt cement, aggregate, and sometimes RAP, are mixed in a zone not
in contact with the hot exhaust gas stream, counterflow drum mix plants will likely have
lower VOC emissions tan parallel flow drum mix plants. The organic compounds that
-------
are emitted from counterflow dram mix plants are likely to be products of a slight
inefficient combustion and can include HAP.
Process fugitive emissions associated with batch plant hot screens, elevators, and the
mixer (pug mill) are not present in the drum mix process. However, there may be slight
fugitive VOC emissions from transport and handling of the hot mix from the dram mixer
to the storage silo and also from the load-out operations to the delivery tracks. Since the
dram process is continuous, these plants must have surge bins or storage silos. The
fugitive dust sources associated with dram mix plants are similar to those of batch mix
plants with regard to track traffic and to aggregate material feed and handling operations.
TableS
Counter Flow HMA
Emission Sources
AP-42, 1/95 Figure 11.1.3
Emission Source
Counter Flow Drum Mixer
Cold Aggregate Bins
Asphalt Cement Storage Heater
Loader
Fine Aggregate Storage Pile
Coarse Aggregate Storage Pile
Conveyor from Counter Flow Drum
Mixer to scalping screen
Conveyor from scalping screen to Cold
Aggregate Bins
Conveyor from RAP Bin to
Counter Flow Drum Mixer
Primary Collector
Secondary Collector
Unpaved Haul road
Or
Paved Haul Road
Source Classification Code
3-05-002-05
3-05-002-04
3-05-002-06
3-05-002-07
3-05-002-08
3-05-002-04
3-05-002-03
3-05-002-03
3-05-020-06
3-05-020-06
3-05-020-06
3-05-020-33 (proposed)
3-05-020-34 (proposed)
[Note: SCCs 3-05-020-33 and 3-05-020-34 have been proposed by DNR as additions to
the USEPA SCC list.]
DNR INFORMATION ON HMAs
DNR has experience with asphalt plants specific to their operations in Wisconsin. .The
following is a list of DNR observations regarding HMAs.
1. Emissions from processes
Processes are those sources ducted to a vent or to a piece of control equipment.
For the most part emissions from these processes are assumed to be gaseous.
Particulate matter emissions are assumed to be emitted from fugitive sources.
-------
A. Particulate Matter Sources
The PM and PM10 emissions associated with ducted or vented processes
at HMA plants are assumed to be negligible.
B. Sulfur dioxide (SOa), nitrous oxides (NO*), reactive organic gasses
(ROG), lead (Pb), carbon monoxide (CO)
The discharge of theses criteria pollutants should be linked to the fuel
burned at either the batch mix or drum mix asphalt plants. Information
from AP-42 Sections 1.3 and1.4 are used in calculating these emissions
because these emission factors are supported through higher quality data
(AP-42 gives them an A rating) as compared to the emissions for these
pollutants in AP-42 Section 11.1 which have a D or E rating. [Note:
US EPA assesses a quality rating of A (the best) to E (the worst) based on
data compiled to support these emission factors.] These criteria emissions
vary based on fuel burned.
1. Natural Gas
Information from AP-42 Section 1.4 using SCC code 1-02-006-02
for small industrial boilers for burners with a heat input rating of
less thanlOO MMBtu/hr are used in the calculation of criteria
pollutant from natural gas burning. SCC code 1-02-006-01 is used
for burners with a heat input rate greater than 100 MMBtu/hr.
2. Distillate Fuel Oil
Information from AP-42 Section 1.3 using SCC code 1-03-005-02
for boilers using distillate fuel oil The criteria pollutant from
distillate fuel oil burning should be calculated. Please note that
distillate fuel oils are #1, #2 and #4 (if made up solely of heavy
distillate fuel oil). #1 and #2 distillate fuel oils have a maximum
sulfur content of 0.5% by weight.
3 Waste Fuel Oil
Waste fuel oil is any oil refined from crude oil or synthetic oil that
becomes contaminated during use. It includes engine oil, gear oil,
lubricating oil, hydraulic oil, cutting oil, transformer fluids and
tempering or quenching oils. Waste fuel oil is divided into three
categories: specification, off-specification and hazardous waste.
Waste fuel oil can then be mixed with either distillate or residual
fuel oils to obtain a required viscosity.
-------
10
As a default, DNR assumes that #3 waste oil emission factor(s) are
the same as the distillate fuel oil emission factors, provided the #3
oil is made up of either distillate fuel oils or meets the
specifications of #1 and #2 distillate fuel oils. Because of this
assumption, we use the same SCC code of 1-02-005-02 for the #3
waste fuel oil meeting the above criteria.
4. Liquified Petroleum Gas Combustion
The criteria pollutant emissions from Liquified Petroleum Gas
Combustion should be calculated using information from AP-42
Section 1.5 using SCC code 1-02-010-02 for industrial boilers.
You will need to know the amount of Liquified Petroleum Gas
burned at the HMA in 1000 gallons in order to calculate these
emissions.
2. Fugitive Emissions
A. Asphalt Heaters
Asphalt heaters are used to heat the liquid asphalt cement. These units are
typically fired with fuel oil. Again the combustion emissions factors from
AP-42 Section 1.4 (natural gas fired) and 1.3 (fuel oil fired) could be
applied to the heaters, as appropriate to fuel burned.
B. Fuel Oil Tank Heaters
Fuel oil tank heaters are used to heat the waste oil/residual fuel oil. These
units are typically fired with fuel oil. Again the combustion emissions
factors from AP-42 Section 1.4 (natural gas fired) and 1.3 (fuel oil fired)
could be applied to the heaters, as appropriate to fuel burned. Generally,
the emissions from fuel oil tank heaters are small or insignificant.
C. Roadways
Asphalt plants may have either paved or unpaved roadways, depending on
their location and whether they are permanent or portable units. The same
emissions factors and fugitive dust control efficiencies applied to
roadways in quarries and sand and gravel operations under the current
nonmetallic mining guidance document should be used to characterize
these emissions.
D. Stockpiles
Asphalt plants will have several stockpiles of different types of stone, sand
and ground recycled asphalt (RAP). The same emissions factors and
-------
11
fugitive dust control efficiencies applied to roadways in quarries and sand
and gravel operations under the current nonmetallic mining guidance
document should be used to characterize these emissions.
E. Diesel/Gas Generators
Portable asphalt plants generally use diesel/gas generators in remote areas
to supply power for the asphalt plant equipment. The same emissions
factors applied to diesel/gas generators in crusher plants. The combustion
emissions factors from AP-42 for Reciprocating engines (diesel fuel) 2-
02-001-02 and (gas) 2-02-003-01 could be applied to the diesel/gas
generators, as appropriate to fuel burned.
F. Hauling/Loading to Bins
Asphalt plants generally will have several bins that a loader operator is
required to keep full by traveling between the storage piles and dumping
into the bins. The same emissions factor and fugitive dust control
efficiency applied to portable crusher operations can be applied to asphalt
plants. The hauling emission factor (developed for crushers) is listed in
the Nonmetallic Air Emissions Guidance document applicable to the
calendar year the emission was reported. The SCC code applicable to this
fugitive emission is 3-05-020-11.
G. Screening
Asphalt plants generally will have screen(s) that separate out over sized
materials before it enters the mix. The same emissions factor and fugitive
dust control efficiency applied to portable crusher operations can be
applied to asphalt plants. The SCC Code applicable to screening is 3-05-
020-04.
H. Soil Remediation
Asphalt plants that remediate soil discharge ROG and benzene emissions.
These emissions are calculated and recorded according to their permit
requirements and can be entered into AEMS under a separate process
number. The SCC code associated with asphalt plants should be entered.
The emission factor for ROG should reflect the units of reporting.
Typical HMA Plant
There are similarities and differences between (1) batch mix plants, (2) parallel flow
drum and (3) counterflow drum mix plants. Each HMA plant in AEMS should begin
with similar devices and proceses and have the specific HMA tailored using the devices
and processes specific to the type of HMA. Table 4 lists the HMA sources identified by
-------
12
DNR and USEPA. The table then lists whether these HMA sources should be part of the
base emission configuration or can be added to make the HMA specific to the type of
HMA.
-------
Table 4
HMA Emission Sources
Base or Optional
Emission Source
Rotary Dryer
or
Parallel Flow, Counter Flow
Drum Mixer
Hot screens
Hot bins
Mixer
Elevator
Cold Aggregate Bins
Asphalt Cement Storage
Heater
Loader
Fine aggregate storage pile
Coarse aggregate storage pile
Conveyor
Primary collector
Secondary collector
Unpaved Haul Road
or
Paved Haul Road
Screen
Combustion tank heaters
Diesel/Gas Generators
(Gensets)
Soil Remediation
sec
3-05-002-01
or
3-05-002-05
3-05-002-02
3-05-002-02
3-05-002-02
3-05-002-02
3-05-002-04
3-05-002-06
3-05-002-07
3-05-002-08
3-05-020-11 (unpaved)
or
3-05-020-37
(paved)
3-05-002-03
3-05-002-03
3-05-020-06
3-05-020-33 (proposed)
3-05-020-34 (proposed)
3-05-020-04
1-03-005-02 oil
1-02-006-03 natural gas
2-02-001 -02-diesel
2-02-003-01 -gas
3-06-22-001 Underground
Storage Remediation and
Other Remediation
3-06-22-002
Underground Storage and
Other Remediation: Soil:
Residual Oil
30622003 Underground
Storage and Other
Remediation: Soil: Natural
Gas
30622004 Underground
Storage and Other
Remediation: Soil: Distillate
Oil
30622005
Underground Storage and
Other Remediation: Soil:
LPG
30622006 Underground
Storage and Other
Remediation: Soil: Waste
Oil
Base or optional
Base
Optional
Optional
Optional
Optional
Optional
Base
Base
Optional
Optional
Base
Base
Optional
Base
Base
Optional
Base
Optional
Comment
Emission factors for SO2,
NOx, ROG, and CO should be
removed for asphalt mix
process with SCC 3-05-002-05
SO2, NOx, ROG, and CO
should be removed for asphalt
mix process with SCC 3-05-
002-04
Assign 50%, 75%, or greater
than 90% control efficiency
Assign 50%, 75%, or greater
than 90% control efficiency
Assuine baghouse of 95%
control
Assign 50%, 75%, or greater
than 90% control efficiency
Emission factors for PM and
PM10 for natural gas should
be switched off
-------
14
Summary
Air Management wants to consistently characterize HMA emissions. USEPA and DNR
information were reviewed to determine what sources should be part of a base emission
characterization and which sources could be optional.
-------
Newt Rowe, Analytical Scientist
State of Connecticut
Department of Environmental Protection
Hartford, CT
-------
STATE OF CONNECTICUT
DEPARTMENT OF ENVIRONMENTAL PROTECTION
March 12,2001
Mr. Ronald Meyers
USEPA
Mail Drop 19
Research Triangle Park, NC 27711
RE: Hot Mix Asphalt Plant Emissions Study
Dear Mr Meyers:
As a representative of the Connecticut DEP, I tracked the progress of the asphalt plant emissions study
from the outset. As you know, I played an active role in the developmental stages of the study by
attending meetings, participating in conference calls, providing written comments, and witnessing
emissions testing. During the developmental stages of this project, I became very aware that the USEPA
staff was conducting the project in an exemplary fashion. It was quite evident that the scope of the
project was comprehensive, the collection and analysis protocols appropriate, the objectives well-defined,
and the USEPA staff very capable.
Given this, it was decided that my time could be more efficiently and effectively spent on Title V and
NSR permitting as well as other areas where technical support was needed within the DEP. As such, I
started to track the progress of the asphalt study. It was important to track the progress of the study since
a moratorium on the issuance of permits to asphalt batch and drum mix plants had been put into effect in
Connecticut. This moratorium was established to assure that the results of the USEPA asphalt study
could be considered in the NSR permit application review process. I believe that there are six
applications in-house to which the results of the hot mix asphalt study will be applied.
To this end, the DEP has developed two sets of spreadsheets one for batch mix the other for drum mix
plants. Each set contains eight individual tables, only natural gas and #2 fuel oil combustion are
considered since the use of #4 or #6 fuel oil for a new or modified batch or drum mix plant in Connecticut
would probably not be considered BACT. The eight tables are described briefly below:
1) Dryer/Hot Screens/Mixers (as applicable) - Criteria Pollutants
2) Dryer/Hot Screens/Mixers (as applicable) - HAPs (federal & CT)
3) Load & Silo Filling Operations - Criteria Pollutants
4) Load & Silo Filling Operations - Semivolatile HAPs (federal & CT)
5) Load & Silo Filling Operations - Volatile Organic HAPs (federal & CT)
6) Silo Filling & Storage Operations - Volatile Organic HAPs (federal & CT)
7) Asphalt Hot Oil Heaters - VOC, PAH & PCDD/PCDF
8) Premise-Wide Summary (Criteria Pollutant & federal HAP)
If you have any questions, please contact me at (860)-424-3465 or by E-mail at
Newt.Rowe@po.state.ct.us.
Sincerely,
Newt Rowe
Analytical Scientist
cc: Robert McConnell, USEPA Region 1
( Printed on Recycled Paper )
79 Elrn Street • Hartford, CT 06106-5127
http://dep.state.ct.us
An Equal Opportunity Employer
elebrating Connecticut Coastal Resource Management: 1980 - 2000
-------
Lloyd Pillion
Citizens Against the Asphalt Plant
Boston, MA
Ravi Nadkarni
Wrentham Environmental Coalition
Wrentham, MA
Henry W. Nowick
Nowick Environmental Associates
Springfield MA
Hyla Napadensky
Napadensky Energetics
Lutsen MN
Janet &Lou Zeller
Blue Ridge Environmental Defense League
Glendale Springs, NC
Jeff Gearhart
Ecology Center
Ann Arbor MI
Stephen F. Lynch
Massachusetts State Senator
First Suffolk District
Boston, MA
-------
Minority Report on Fugitive Emissions from
Asphalt Plants
Lloyd Fillion, Citizens Against the Asphalt Plant, Boston MA
Ravi Nadkarni, Wrentham Environmental Coalition, Wrentham MA
Henry W. Nowick, Nowick Environmental Associates, Springfield MA
Hyla Napadensky, Napadensky Energetics, Lutsen MN
Lou Zeller, Blue Ridge Environmental Defense League, Glendale Springs, NC
Janet Zeller, Blue Ridge Environmental Defense League, Glendale Springs, NC
Jeff Gearhart, Ecology Center, Ann Arbor MI
Stephen F. Lynch, Massachusetts State Senator, First Suffolk District, Boston MA
-------
Minority Report on Fugitive Emissions from Asphalt Plants
Introduction
In the mid-1990s, several groups of citizens and environmental organizations began petitioning the
US EPA to quantify so-called "fugitive" emissions from bituminous concrete, or asphalt hot mix,
plants. These requests were based on observations and sub sequent theoretical analysis suggesting that
such emissions could be significantly greater in their volume and concomitant impact on the local
environment than the word "fugitive" implied. In late 1996, the EPA decided to move forward with
its own testing after the asphalt industry failed to conduct their own testing in an open manner as they
had promised the EPA, producing a test report1 which was heavily criticized by academic and
research scientists, public health officers, citizens, environmental activists and even from within the
EPA itself. The citizen or public representatives who participated in this program ultimately included
staff from public health departments in Chelsea, Massachusetts and Boston, Massachusetts,
engineering/science consultants from several diverse disciplines including chemists, environmental,
chemical and mechanical engineers, environmental organizations from Massachusetts, Connecticut,
North Carolina and Michigan, and citizen activists from those states and from Minnesota, New York
and Virginia. Additionally the Massachusetts Department of Public Health and one Massachusetts
State Senator with experience in construction provided comments.
The EPA committed to providing an opportunity for citizens, along with industry and other
stakeholders, i.e., other government agencies, to help frame the test protocols. This opportunity was
provided through a series of conference calls and day long meetings and written communications. The
interaction was hosted and coordinated by the EPA's New England Regional office, and continued
through the testing, compilation of test data, analysis and production of reports. The commitment of
the EPA was laudable; however, the several year long interactions were adversarial at best, with
hostility or distrust between the multi-state citizen group with its technical consultants and the
technical staff from EP A's Research Triangle Park (EP A-RTP). The barely contained rancor resulted
in most of the conference calls and meetings being conducted with mediators from the EPA's New
England Regional Alternative Dispute Resolution Program.
At the point of testing, the EPA provided opportunity (travel and lodging) for only one citizen
technical advisor from the East Coast cluster of public commenters to witness the primary test which
took place in California at a drum style facility, the same facility that had been previously selected by
industry1, code labeled by the EPA as "Plant C." At a secondary test in Massachusetts, at a batch
style facility labeled "Plant D," three citizens were on site for the three days of testing. The primary
test observer was severely handicapped because of an inability to be at more than one place
simultaneously, a prohibition on citizen photography coupled with the EPA's reluctance and even
refusal to take requested photographs as had been promised the citizens, and the exclusion of the
citizen from certain meetings which occurred between plant personnel and EPA officers with their
testing consultants who were responsible for the test. It is against this less-than-open venture in
cooperation that the test and subsequent data are to be viewed.
1. "Emissions Testing: July 9-11, 1996" AIRx Testing, Ventura Ca - Job # 1030.
-------
Why are many citizen groups concerned about asphalt plants?
Hot Mix Asphalt plants (HMA plants) and asphalt terminals have numerous characteristics that result
in a large number of the general public being exposed to their hydrocarbon and parti culate emissions.
This is in contrast to the general experience where workers in industrial plants are frequently exposed
to higher concentrations of pollutants for longer durations than the people who live near the plant.
In the case of asphalt plants, this generalization is not correct because such plants, of which there are
about 3,600 within the US, are often located in urban areas close to homes, schools and playgrounds
with minimal setbacks. Further, the plants operate around the clock when fulfilling certain types of
contracts so that while the workers might be exposed only eight hours per day, the neighbors breathe
in the fumes day and night. While the operation of an asphalt plant might be seasonal, particularly in
northern latitudes, the intensity of exposure during the peak production months is quite acute.
Anecdotal information collected by Boston Health Commission's Office of Environmental Health and
by others shows that some individuals exposed to asphalt plant emissions have the same or similar
symptoms: asthma, nausea, headaches, skin rash, etc. This information is not sufficient to indicate
whether these physiological reactions were a result of hypersensitivity and allergies or whether this
was a reaction of non-allergic people to levels of hydrocarbon and fine particulates high enough to
elicit a response. Available epidemiological studies have shown statistically significant links between
exposure to hydrocarbons and/or metal fume and childhood leukemia2 and between exposure to
asphalt fume and a variety of cancers. Dr. Eva Hansen3 measured excess cancers in asphalt workers
in Scandinavia. These included cancers of the mouth, esophagus, lung, and rectum. Dr. Hansen also
found increases in non-pulmonary cancer, liver cirrhosis, bronchitis, emphysema and asthma in asphalt
workers.
As a part of the interaction with the EPA, the multi-state citizen coalition held a conference call with
EPA-RTP to discuss health effects of asphalt fume and to see if they could regulate emissions from
HMA plants so that the exposure of the general population would be reduced. These discussions
were unsatisfactory. EPA-RTP has acknowledged that their limited resources permit the
consideration of only about 50 chemicals annually, out of a field of 50,000 chemicals not yet tested.
Asphalt fume, a complex emission, is comprised of over 2,000 individual chemicals. Since EPA's
current approach is based on considering each chemical by itself, knowledge about the health effects
of each individual chemical will not be available for many decades. Further, even after this data has
been compiled, the synergistic interactions between these chemicals in a complex mixture will not be
available and would require further study. EPA-RTP suggested that the citizens undertake an
epidemiological study to establish a statistical link between exposure to asphalt emissions and health
2. E.G. Knox & E.A. Oilman, "Hazard proximities of childhood cancers in Great
Britain from 1953 -80", Journal of Epidemiology and Community Health, 1997, 51, 151-159.
3. Hansen ES, "Cancer incidence in an occupational cohort exposed to bitumen
fumes". Scand J Work Environ Health 1989;15:101-5. Hansen ES. "Mortality of mastic
asphalt workers". Scand J Work Environ Health 1991;17:20-4; Hansen ES, "Author's reply",
Scand. J. Work Environ. Health. 18: 135-41, (1992)
-------
effects. EPA would not act to curb these emissions in the absence of such a study. For environmental
pollutants like asphalt fume, neither exposure nor toxicity are acute enough to cause immediate death.
The physiological response might take time (for example, cancer often takes several decades to
develop even with exposure to known carcinogens like tobacco smoke) or the irritants will produce
a variety of negative health responses in a general population which, while significant in terms of loss
of productivity and "life, liberty and pursuit of happiness", don't cause immediate death but still have
an effect on mortality and morbidity. To track asphalt plant neighbors on a national basis over a long
period of time is a massive job that is totally outside the ability and resources of any citizen group4.
The current work is the culmination of an almost 6-year effort to quantify fugitive emissions from
HMA production/handling. On the one hand, EPA agreed to measure these emissions by collecting
and analyzing data from two plants, labeled Plant C, a large drum plant and Plant D, a batch plant.
For this we are grateful. On the other hand, the work has certain shortcomings and this Minority
Report presents a discussion of these issues which have not been resolved to our satisfaction. Given
the contentious and contradictory nature of the latest document "RESPONSE TO COMMENT ON
DRAFT EMISSIONS ASSESSMENT REPORT, AP-42 SECTION AND AP-42 BACKGROUND
REPORT"5, a point by point rebuttal has not been attempted; instead a few select examples are
presented to illustrate that these differences continue due to lack of valid response from EPA. The
discussion focuses on the maj or questions that this work was expected to answer which unfortunately
have not been answered.
It should be noted that there is a difference between the EPA and the citizens in the way they
respectively approach the issue of asphalt fume emissions. The citizens approach the issue from a
public health perspective. Thus their concern is with both average exposures for long term effects
4. There are probably political reasons for this reluctance. After all, the asphalt industry is at the
confluence of several large, powerful and politically active industries with a long reach into Washington DC.
These include the oil industry who produces the asphalt during the refining of petroleum, the sand and gravel
business which produces aggregate, the other component of hot mix, the highway lobby and the State Departments
of Transportation or Highways who have a symbiotic relationship with the hot mix asphalt industry as the major
customer for the product. The article by Sarah Gibson ("How OSHA Dealt with Asphalt Fume", New Solutions, 5-
1, 24 - 47, Fall 1994.) is essential reading for any one interested in learning how the industry dictated the final
OSHA standards for occupational exposure to asphalt fume. The industry also conducted a vehement, coordinated
attack on Dr. Hansen, the epidemiologist who has shown a significant relationship between exposure to asphalt
fume and various cancers, at an ACGIH Satellite Meeting on "Developing Occupational Exposure Values from
Toxicology and Epidemiology Studies" on March 6, 1998 in Seattle. Her use of statistics was vehemently
challenged by a statistician as a part of an industry presentation by an employee of Ashland Oil. The industry also
paid to bring other attackers to the meeting including an employee of a Scandinavian industry trade group and
another Scandinavian researcher whose results are somewhat different from Dr. Hansen's.
5. "RESPONSE TO COMMENT ON DRAFT EMISSIONS ASSESSMENT REPORT,
AP-42 SECTION AND AP-42 BACKGROUND REPORT" released December 20, 2000. This
report and related information can be downloaded from: http://www.gov/ttn/emc/asphalt.html.
Copies of CD ROM can be requested by mail from Emissions Measurement Center, MD-19, US
EPA, Research Triangle Park, NC, 27711. Paper copies can be obtained from National Technical
Information Service, 5285 Port Royal Road, Springfield VA 22161. 1-800-553-6847.
-------
and also with peak, short-term exposures for acute effects: particularly since much of the anecdotal
information deals with the latter. The EPA's approach, as stated in the latest response5 is to collect
data to primarily enable the estimation of emission inventories and for air quality planning, although
EPA acknowledges that the information could be used in other contexts, for example, in applications
to State authorities to build new plants. EPA's published data on emission factors is not just an
average number for all plants when data from individual plants varies by orders of magnitude, as much
of this data is collected under ideal operating conditions. For example, much of the data used for
developing stack emission factors was collected just after initial plant startup often during a
preoperating test period where the plant owner is demonstrating to the state regulatory authorities
that the plant can operate for short periods of time at the design operating rate while staying within
the permitted emission limits. In these situations, it is not unusual to have engineers from the
equipment suppliers standing by to adjust and tune the equipment (as was the case at Plant C for
EPA's test) to make sure that the equipment is operating optimally. Once the test is successfully
completed, these engineers depart and the on-going emissions are dependent on the skills of the plant
operators and their internal maintenance staff.
Of equal importance, and most critically, citizens and public health officers are inclined to consider
worst case scenarios while the EPA almost consistently opted to publish data that reflects the best
case scenario. Additionally, these numbers don't show measures of variability despite our request.
Such measures are buried in the appendices.
Recently, the EPA has been criticized for the lack of science to back up their decision-making6.
Specifically, National Research Council mentioned "weak scientific performance" and "weak scientific
credibility". The problems uncovered by citizen reviewers in this case indicate that the current proj ect
also suffers from similar shortcomings.
Unresolved Issues
Rather than repeat the arguments, pro and con, under each topic, the highlights of the positions of
the two sides will be presented in this discussion of unresolved issues with emphasis on issues that
concern the citizens. Additional details, particularly of the EPA position, are in the most recent
document5. Specific references are provided to the Section number in this "Response to Comments"
report.
1. The issue of volatile content and operating temperature
If asphalt did not contain a volatile fraction, there would not be any emissions of hazardous
organic chemicals from hot asphalt. Unfortunately, all asphalts contain a volatile fraction
which evaporates at the operating temperature of between 275° and 375° F. Further, the
emissions increase exponentially as temperature increases. The issues here are as follows:
6. "Strengthening Science at the U.S. Environmental Protection Agency: Research
Management and Peer Review Practices (2000)", National Research Council, National
Academy Press, http://www.nap.edu/openbook/030971275/html/
-------
a. Why did the industry change the specification for the maximum allowable volatile
content of asphalt from 0.5% to 1.0%? What are the implications of this change?
b. How is this volatile content measured now and how was it measured in the past?
c. How is the volatile content affected by seasonal variations in oil refining practice?
d. How is the volatile content affected by the various classes of additives that are sold
to improve the performance of asphalt?
e. Are the maximum operating temperatures recommended by industry groups actually
followed by plant operators? If not, what are the variations in temperature and what
is the consequence of these variations?
The above issues are repeated below as captions for the discussion to follow which delineates
the differing positions of the citizens and EPA in bold letters:
a_ Why did the industry change the specification for the maximum allowable volatile
content of asphalt from 0.5% to 1.0%? What are the implications of this change?
Traditional grades of asphaltuse "proxy" properties such as viscosity at 140° F for specifying
and differentiating between grades of asphalt. Under such specifications, the volatile content
is a "not to exceed" value, typically 0.5% for the most common grade of asphalt AC-20.
Certain users might allow a relaxation of these specifications. For example, in Massachusetts,
the Mass Highway Department will allow their engineer in the field to relax the requirement
for Loss on Heating from 0.5% to 1.0%. One of the problems with this type of specification
is that the volatiles content of different batches of asphalt meeting the same viscosity
specification will vary dramatically. For example, an article published by industry researchers7
shows that the volatiles content of two different batches of AC-20 varied from 0.053% to
0.5%, a variation of one order of magnitude. This means that the emissions of asphalt fume
would have also varied by an order of magnitude. Second, the "not to exceed" value for
volatiles contentis "based on original asphalt", i.e., before any blending. In other words, these
specifications apply to the material that is shipped from a refinery, not necessarily to material
that is shipped from a terminal to a hot mix plant. At a terminal, the terminal operator will
add various diluents for altering viscosity. The potential effect of these diluents is discussed
later in section d.
The old system of viscosity specification is being replaced by a new specification called
"Superpave" which is based on measuring the properties of hot mix under simulated service
conditions. This standard was j ointly developed in concurrence with many parties including
the government. However, the newer Superpave specification increases the permissible
volatiles content to "not to exceed 1.0%".
Our concern is that the specification was changed by industry in order to make it easy and
feasible for the industry to meet it under widely varying production conditions; otherwise, it
7. V.P. Puzinauskas & L.W. Corbett, of the Asphalt Institute, "Report on Emissions
from Asphalt Hot Mixes", Paper presented at ACS Meeting, Chicago IL, August 1975.
-------
would have remained at 0.5%. This means that the emissions from certain plants at certain
times of the year, will be quite a bit higher than what is now shown in these reports. This
increase is quantified in Table 1 later in this report. EPA's position is that this change is not
relevant, and that anyone can use the volatility value they want in the equations presented.
EPA also believes that the average volatility content of asphalts is less than 0.5% based on
a set of samples analyzed in 1993 as a part of the SHRP program. (3.2.2) We strongly
disagree that the range of volatility values measured in the SHRP program in 1993 is the
relevant range for asphalts in use today after blending. Further, EPA arrives at conclusions
about volatility of asphalts used around the country based on the SHRP data. This
extrapolation would be correct only if equal quantities of each of the various asphalts in the
database were sold each year in the US. There is absolutely no information to support that
key assumption.
h. How is this volatile content measured now and how was it measured in the past?
The volatiles content of asphalt has traditionally been measured by the Thin Film Oven Test
(TFOT), but the industry is moving to a slightly different test, the Rolling Thin Film Oven
Test (RTFOT). We understand that a major reason for this change is because the results of
the Rolling Thin Film Oven Test are more predictable, i.e., the method produces results with
less systematic and random error.
There are two problems with either of these tests. First, both tests measure loss in weight on
heating, which is not the same as measuring the volatiles content, particularly since some
asphalt samples will gain weight as a result a pickup of oxygen and nitrogen. Nevertheless,
because these tests are performed routinely, it is the only available data source for estimating
the volatile content of original asphalt. Second, the loss of weight number obtained from the
Thin Film Oven Test is lower than the loss of weight measured by the Rolling Thin Film Oven
Test. However, the limited data, published by PES in the Plant C test report Appendix B.8,
indicates that this difference is 0.16% at 325° F. (For example, if the TFOT measured loss
of weight as 0.3% and RTFOT measured the loss of weight as 0.46%, the difference would
be 0.16%). When this difference was measured at two other temperatures, it was found to
be less than 0.1%.
Thus, EPA's other explanation (3.2.2) that the volatiles specification was increased from
0.5% to 1% because of a change in the testing procedure from the Thin Film Oven Test to
the Rolling Thin Film Oven Test is not supported by data. The citizens don't agree that a
change of less than 0.16% should cause a relaxation of the specification by 0.5%, a value
three times the measured change.
a. How is the volatile content affected by seasonal variations in oil refining practice?
Refinery operations change seasonally. For example, summer runs are designed to maximize
gasoline production while winter runs decrease gasoline production and maximize fuel oil
production. Seasonal variations in refining operations should affect the volatiles content of
asphalt.
-------
7
The citizens are concerned that no data is available on this subject. Similar concerns were
raised by the Mass Department of Public Health. Such data might provide further insight why
the specification was changed and might provide a better understanding of exposure suffered
by neighbors of such plants.
d_ How is the volatile content affected by the various classes of additives that are sold
to improve the performance of asphalt?
In recent years, the industry has produced dozens of proprietary diluents and modifiers for
the purpose of blending and for improving asphalt properties. These additives fall into
different modifier families including the following: fibers, fillers, plastic/rubber, rejuvenating
oils, antistripping agents, extenders and antioxidants. A partial list includes hydrolene (Sun
Chemicals), Kraton polymers (Shell Chemicals), Elvaloy (Du Pont) and so on. The fibers and
fillers are not expected to change the volatility but all others, which interact chemically with
asphalt constituents, should alter the volatility of the asphalt, based on engineering principles.
These substances are added in significant amounts: one or several percent by weight of the
asphalt. In other words, the recommended amount of additive exceeds the percentage of the
volatile component. Not only are these additives light compounds, but they can also alter the
vapor pressure of the asphalt by increasing the volatility of some of the lighter components
which otherwise might not have volatilized in the temperature range of 275 to 375° F.
The citizens are concerned that the measurement of loss of heating of the original asphalt can
bear little relation to the emissions from the asphalt that is actually used to prepare a hot mix.
EPA's position is that this is not a significant issue. The two tests at Plant C and Plant D did
not use any additives so there is no available data.
e.. Are the maximum operating temperatures recommended by industry groups actually
followed by plant operators? If not, what are the variations in temperature and what
is the consequence of these variations?
The second factor that increases asphalt fume emissions is operating temperature. As such,
emissions increase exponentially with temperature. State Highway Departments will specify
a minimum and a maximum temperature for HMA when delivered to the job site where the
pavement is being installed. In Massachusetts, this range is 275° to 325° F. Note that the
actual temperature of the asphalt leaving the plant and being loaded into a truck is higher.
The temperature has to be higher in order to compensate for cooling during transportation
to the job site and this issue becomes critical when delivering to a distant job site. Also, note
that there is no such restriction when delivering HMA for non-state supervised jobs. Also,
according to an engineer who was the general manager of an asphalt plant and a plant
designer internationally, who attended many of the planning meetings, many small private
contractors who drive small trucks will demand a higher temperature at loadout since a
smaller mass of HMA will cool faster.
There are also other factors that will increase emissions. For example, when a plant switches
from one HMA mix formula to another, emissions will increase. The best example of this is
-------
when a plant making hot mix with Recycled Asphalt Pavement (RAP) switches to a formula
without RAP. In the first instance, the plant would have operated at a higher temperature to
provide the extra heat to evaporate the water associated with RAP since the RAP is stored
in the open and not dried. When the change occurs to a formulation without RAP, the plant
loadout will emit a higher level of organics because of overheating. This is a frequent
problem at batch plants though this type of problem is not restricted to only batch plants.
(Examples of this type of excursion exist in the data collected at Plant C, a drum mix plant
where emissions increased by a factor of two to three over a 40 minute period8.) This again
means that the episodes of high emissions caused by variations such as high temperatures are
missed by the total reliance on averages of data collected under ideal conditions, even though
the high emissions on the day of the test did affect the average.
The citizens, based on personal observations, believe that such periods of high emissions are
frequent, especially at batch plants. Further, the public health impacts of these emissions are
more significant since batch plants are more numerous than drum plants, located closer to
homes and service the small independents who are more likely to demand hotter asphalt. The
Table below, Table 1, is based on equations published by EPA relating volatiles content and
production temperature to various emissions. (Please refer to the Emissions Assessment
Report9, Table 1 on page 6 and Table 11.1-14 on page 11.1-31 in appendix A.) Table 1 on
page 9 of this report shows the differences between EPA's numbers and ours on the basis of
assuming 1% volatile content and an operating/loadout temperature of 375°F.
It can be seen that the emissions calculated by using EPA-derived equations, particularly
emissions of noxious organic compounds, increase by more than 600% under conditions of
higher operating temperature and volatility content. Both the EPA and the Citizen numbers
would be increased by another 20 to 40% to compensate for the low bias introduced by the
"background correction" and "Method 204", discussed later in this report. Finally, it should
be noted that although the numbers in Table 1 are shown on an annual basis to help compare
them to Table 1 in the Executive Summary of the Emission Assessment Report, the citizens
are aware that actual annual emissions will be lower since a plant will not always operate with
an asphalt with a high volatiles content at high temperatures. On the other hand, the table
clearly shows the type of variation in emissions that is likely to occur under such conditions
with its effects on nearby residents.
2. Issues relating to how the data was collected and analyzed
This process of planning, data collection and analysis has involved many discussions,
agreements and disagreements.
8. Ravi Nadkarni, Comments on EPA Reports on Asphalt Fume Measurements.
September 15, 1999.
9. "Hot Mix Asphalt Plants - Emissions Assessment Report" with Appendices A, B & C.
EPA-454/R-00-019, December 2000.
-------
The citizens are concerned that with time, EPA's responses have become more rigid,
inconsistent and not based on data (or the lack thereof) in the reports. Only a few of the key
issues are mentioned below. Others are found in the previous comments by the citizens to
various draft reports by EPA8'10.
Table 1: Effect of Different Volatile Contents and Operating Temperatures on emissions
EPAa Citizens" EPAC Citizens"
Loadout emissions6
- Total Paniculate Matter 52 257 104 515
- Organic Paniculate Matter 34 239 68 478
- Total Organic Compounds (Method 25A) 416 2,918 832 5,836
- Carbon Monoxide 135 947 270 1,893
Silo filling emissions'
- Total Paniculate Matter 59 211 117 423
- Organic Paniculate Matter 25 178 51 356
- Total Organic Compounds (Method 25A) 1,219 8,550 2,437 17,100
- Carbon Monoxide 118 828 236 1,656
a. EPA estimates for batch plant in lb/100,000 tons of HMA. Volatility of 0.5%, 325° F.
b. Citizen estimates for batch plant in lb/100,000 tons of HMA. Volatility of 1.0%, 375° F.
c. EPA estimates for drum plant in lb/200,000 tons of HMA. Volatility of 0.5%, 325° F.
d. Citizen estimates for drum plant in lb/200,000 tons of HMA. Volatility of 1.0%, 375° F.
e. Loadout emissions for both batch and drum plants - See Table 11.1-14. AP-42.
f. Loadout emissions forplants with silo storage- mainly, but not exclusively, drum plants. See Table 11.1-14.
AP-42.
a. The issue of the "Background" correction
Many of the methodological problems which occurred throughout this program are
crystallized under EPA's topic of the "background correction". Background correction at
Plant C refers to EPA's attempts to separately measure loadout emissions mixed with truck
exhaust and just truck exhaust in order to subtract the truck exhaust values from the former
to obtain a "pure" value for loadout emissions. But, the equipment for collecting the
emissions in the tunnel at Plant C was not 100% efficient. Under such conditions of
inefficient collection, the standard engineering practice is to use a tracer gas to measure the
collection efficiency and then correct the raw emission data for this inefficiency of capture
to arrive at a more appropriate estimate of the actual emissions. (Whether a single average
10. Ravi Nadkarni & Lloyd Pillion, Comments on EPA reports relating to AP-42 Section
11.1. August 31.2000
-------
10
collection efficiency factor should be used for adjusting all three runs or whether more time-
specific collection efficiency factors should be used has been discussed extensively in previous
comments8 by the citizens. Also, the math works out such that if the truck exhaust value is
high, the net value for the "pure" loadout emissions will be low.)
EPA conducted such a "background" run to measure just truck emissions. However, the
data collected during the run had so many problems that EPA abandoned standard data
reduction procedures to obtain an answer that they liked. There are three problems: First, the
raw data shows a doubling in the background emissions from the first half to the second half
of the run. What caused this doubling? Second, even if one uses the background data from
the low first half of the run, the numbers are still too high. What is the cause of this result?
Third, why did EPA abandon the standard engineering procedure for correcting measured
emissions for capture efficiency? These various problems with this background run are
discussed below:
i. Inconsistent raw data: The raw data, i.e., data as recorded, and not corrected
for capture efficiency, shows two very distinct regimes in the run, with a break
in between. The first regime is flat with little fluctuation and shows an
average reading of 0.8 ppm of total hydrocarbons (THC). The second regime,
which is also relatively flat, shows a value about twice this level. The first
issue is what caused this doubling of measured emissions. EPA was unable
to provide a credible explanation. The citizens believed that this increase was
achieved by parking a second truck at the entrance to the tunnel so that the
second regime was based on measuring emissions from two trucks. In the
final RESPONSE TO COMMENTS report5, EPA used only the first portion
of the background run, hoping to avoid this inference that the background
numbers were inflated. However, the second truck was at the entrance even
in the first portion of the run, though for shorter time periods. Thus, as shown
below, it is not clear that the lower values measured in the first regime
represent exhaust from just one truck.
ii. Is the background reading from the first half of the run low enough?: By
selecting only the data from the first half of the run and by not averaging the
two regimes, it would appear that EPA has avoided problems with data
reduction. Unfortunately, this is not the case.
iii. Problems with data corrected for capture efficiency: Standard protocol
requires that when collection is not 100% efficient, the raw data has to be
corrected for capture efficiency. Capture efficiency is measured with a tracer
gas. Unfortunately, when correction is applied to the background data from
the first half of the background run and subtracted from the combined loadout
plus truck emissions to get "pure" loadout emissions, one gets negative values
for several hazardous air pollutants. Since this is an impossible result, EPA
decided to use the raw uncorrected background number rather than the
capture-efficiency-corrected background number. Even after adopting this
unusual procedure, some "pure" loadout emissions were still negative and
they were assumed to be zero. There is a major problem with this procedure.
-------
11
There is no technical/ scientific justification for ignoring capture efficiency.
The only justification is that it produces numbers acceptable to EPA. The
citizens suggested that the background "correction" be eliminated and the
loadout emissions data reported as "truck plus loadout emissions". This
suggestion was summarily rejected. We reproduce the first paragraph of
EPA's summary rejection (3.2.2):
"The background adjustment was appropriate. There was no improper
manipulation of the data from the background test at Plant C and EPA did
not manipulate the placement of the trucks to obtain higher uncorrected
emissions for the background run. Further, we do not agree that the
background run demonstrates that data was manipulated to produce biased
results and do not believe that concurrently measured truck exhaust and
road dust emissions should be included in the emission factor for load-out
emissions."
EPA states further that:
"The only additional instruction provided to the truck drivers during the
background test was to reduce the time of their travel from the exit of the
tunnel to the arrival at the tunnel entrance At about the time trucks began
driving faster to reduce gaps between trucks, the wind speed increased. This
may have caused an increase in the diesel exhaust that entered the tunnel
entrance or increased the capture of the diesel exhaust of the truck that was
inside the tunnel. " (Emphasis added.)
The problem with these explanations is that the raw data clearly shows that the reading
doubled in the second half of the run. This doubling needs to be explained. Above, EPA is
admitting that there may have been an increase in diesel exhaust entering the tunnel entrance,
i.e., exhaust from two trucks was being counted, but this possible explanation is two
paragraphs after the original denial. Also, note the statement, "Dr. Nadkarni [the citizen
observer at the Plant C tests] observed the operation of the trucks during the background
run. The issue of manipulating the placement of the trucks was not raised by him during the
test to either of the EPA personnel present. " This statement is incorrect. On the day of the
run, he left early to catch a plane and was not present when the second half of the data was
collected. Further, problems of this type are visible only after the data has been collected and
viewed. Such insights are impossible in the field. Dr. Nadkarni had obj ected originally to the
background "correction" because it seemed to be an unnecessary manipulation of data and
the results. In retrospect, his concerns were well founded.
The explanations of why the background reading was not corrected for capture efficiency, in
3.3.4, are mutually contradictory. The tracers used to measure capture efficiency showed that
capture efficiency was high early in the morning and decreased as the day wore on during all
the runs at Plant C. The most reasonable explanation was that as the land around the plant
heated up, there was an on-shore breeze which blew through the tunnel and decreased the
-------
12
capture efficiency. The capture efficiency data for the background run is consistent with this
general statement. This also means that truck emissions captured in the second half of the
background run, when corrected for capture efficiency, were three times those in the first half.
This further confuses the issue. To the citizens, these problems raise serious questions about
both the first and second half of the run. The general scientific procedure when data can't be
explained is to reject it.
Finally, it should also be noted that the auditors from RTI were apparently not involved in a
detailed analysis of this procedure of collecting "background" data.
This background correction introduces a low-bias in the published results for load out
emissions. EPA's estimate of this bias introduced by not correcting for capture is about 20%.
Both EPA and citizens' numbers in Table 1 would have to be increased by this amount to
compensate for this unexplainable "background" factor. The citizens believe that the entire
procedure is faulty and should be rejected.
b._ Problems with enclosures
At Plant C, because the loadout enclosure did not meet the requirements of Method 204, an
EPA protocol for constructing total enclosures for measuring pollutant emissions, atracergas
was used to measure the capture efficiency. At Plant D, because the specially constructed
loadout enclosure met the requirements of Method 204, it was assumed that all the emissions
would be "captured". Method 204 gives precise design requirements which put limits on the
size of Natural Draft Openings compared to the surface area of the walls, floor and ceiling of
the enclosure with the desired ratio being less than 5%, the velocity at Natural Draft Openings
to be at least 200 ft per minute and other limitations on the distance between the opening and
the emission source. In essence, under Method 204, an induced draft fan collects the fumes
from inside the enclosure and delivers them to the instrumentation at the sampling point. By
controlling the openings (Natural Draft Openings) to a specific size, the intent is to force
outside air into the enclosure, avoiding any loss of the material being sampled. As the data
shows8'10, complying with these requirements still does not result in proper sampling of fumes
inside this enclosure. This point is illustrated by the following situation.
i. The loadout enclosure is empty. The instrumentation measuring hydrocarbons
is showing zero hydrocarbons since the fan is pulling in just air from the empty
enclosure. (See paragraph vi below.)
ii. A truck stops under the loadout point. Hot mix asphalt is dumped into the
truck from a silo (a single dump) or from a batch mixer (several dumps with
a waiting period in between). During this period, the air and asphalt fume in
the enclosure are pulled past the hydrocarbon measuring instrumentation by
a fan.
iii. The truck loadout is complete, but the instrumentation continues to indicate
and record hydrocarbon emissions from the hot mix sitting in the truck
iv. The truck leaves the enclosure. One should expect the fan to evacuate all of
the fume from inside the enclosure and for the hydrocarbon reading to go to
-------
13
zero quickly. But this does not happen. As a matter of fact, the readings
persist at a non-zero value for a long period of time showing that Method 204
does not do a good job of delivering hydrocarbons from inside the enclosure
to the measuring point. In other words, since the emissions lingering in the
enclosure are not delivered to the instrumentation for measurement, they are
not "captured".
v. The truck will continue to emit hydrocarbons in the yard of the manufacturing
plant. These are not captured by the sampling system attached to the
enclosure. In the report, these have been estimated as the so-called "yard
emissions".
vi. The next truck enters the enclosure. If the enclosure was doing the proper
job, it would have shown an initial reading of zero, as mentioned above under
i. However, this is rarely the case since these emissions linger in the
enclosure.
Criticism of Method 204 was provided to the EPA before the Plant D tests but was ignored.
One of the problems with Method 204 was illustrated during the first day of testing at Plant
D. Method 204 requires that the inward velocity at any opening be maintained over 200 feet
per minute (fpm) so that air is flowing into the enclosure and not out of the enclosure. Note
that 200 fpm is less than 2.3 miles per hour, which is not much of a breeze. At Plant D, the
citizens observed that some of the fumes were escaping from the top or the bottom openings
in the downwind door of the tunnel because of an ambient breeze. The contractors corrected
this by decreasing the size of the openings by about half at the end of the first day of testing.
Once this was done, this upset condition did not occur again. EPA's recollection of this event
is quite different. (3.1.2).
In spite of these deficiencies, EPA position was:
"As the analysis is an extrapolation of only two data runs and the enclosure was designed
to Method 204 criteria, EPA believed that the actual uncaptured emissions are most likely
smaller than estimated by our analysis. Therefore, no further adjustments were made to the
loadout emissions from Plant D. " (3.1.2)
This is a confusing argument. The citizens position is that EPA again biased the results by
this action. EPA's estimate of this bias is of the order of 10% and this adjustment would have
to be used to adjust the EPA and citizen numbers in Table 1, but this adjustment could easily
be double this value, i.e. 20%, in our opinion.
a. EPA 's flat denial of errors in the report
In the Introductory Section of our last response10, reproduced in the current report as 1.1.1,
we commented that we were dissatisfied with the report review process because EPA released
draft reports that were not sufficiently finalized. We specifically referred to the following
problems; incorrect references, missing and incorrectly referenced appendices and numerical
and logical errors in the analysis. We were referring to the entire review process, not just the
-------
14
current cycle. Also, our point was that EPA should produce draft reports that are free from
error and the review process should not become a search for a needle in a haystack if a large
document is referred to without providing a page number. Further, once the responsibility for
finding each mistake is entirely passed to the reviewers who are volunteering their time for
the review, such mistakes will persist in the final report as such volunteer reviewers miss
them. In their current response to comments5, EPA responded on page 1:
"The commentwith respect to numerical and calculational errors is unsubstantiated
in that no instances of any numerical or calculational errors were provided by the
commentors."
This comment is not accurate. In the previous round of comments8, such numerical and
calculational errors were shown. These included averaging data over a period when there
were no emissions because trucks were absent, and showing a wrong number as a maximum
value. These mistakes were not caught by EPA'S auditor Research Triangle Institute (RTI).
Although chances are that RTFs contract focused on data collection procedures and did not
include an auditing of data reduction and data analysis, their role is being exaggerated in the
final report. At industry's urging, EPA is including the following statement in the Final
Report. "For example, two scientific auditors from the Research Triangle Institute were
employed to independently audit the test and reporting process. " (1.4.38). If RTI is to be
given credit for the audit process, they should also be assigned blame for its failures.
EPA did admit to problems with references and appendices but stated that the missing
appendices were provided on October 6 (1.1.2) or October 3 (1.4.54). They fail to mention
that this was over a month after the end of the comment period.
At Plant D, two citizen observers noted what appeared to be errors in protocol for sampling.
Specifically, when the sampling train was rinsed to recover condensed organic matter, the
observers saw that the entire tube was not being rinsed. When this issue was mentioned,
EPA's denial was immediate and strong, even before the citizens had explained their
observations.
Numerous logical errors and examples of bias were pointed in comment letters8'10, To this,
EPA's response is, "The information presented in the reports as drafted are (sic) unbiased."
(1.2.6). An example of an "unbiased" statement, we suppose, is the following from 2.4.6:
"The lack of data to substantiate the statement is not a reason to eliminate the statement".
d. Unfounded Leaps of Faith
In 3.7.2, while criticizing the use of model compounds by others at a time when no data on
the composition of asphalt fume was available11, EPA defends its own use of model
11. R. Nadkarni, written communication to Chief, Emission Factor and Methodologies
Section, USEPA-RTP, November 7, 1994; R. Nadkarni, written communication to Mr. Ron
-------
15
compounds when such data on composition of asphalt fume is available. Furthermore, this
data clearly shows that the two substances selected by EPA as model compounds, docosane
and tricosane, are not the major measured constituents of asphalt fume. Standard scientific
practice would be to use the best information available for this purpose. When data is
available on the maj or constituents of asphalt fume, EPA chose these two compounds purely
for expediency when data shows otherwise. EPA states, "The Antoine's coefficients for
aliphatic hydrocarbons that come the closest to producing a working loss emission estimate
of 32 pounds per million gallons of asphalt throughput for this approximate molecular
weight are docosane and tricosane. " In other words, these compounds are chosen because
they give the desired result.
A similar contradiction exists in 3.7.6. Here EPA admitted that reference 360 in Appendix
B, a report of certain tests by Division of Air Quality, North Carolina Department of
Environment and Natural Resources, 1998, was not evaluated critically but then EPA ignored
the comments offered by the citizens. These comments had pointed out that the authors of
the reference had selectively ignored high readings of ambient benzene concentrations and
used only the low readings. Further, they had used questionable methodology to convert an
ambient concentration to an emission rate. Nevertheless, EPA decided to retain this defective
information in the final report stating: "The reference was not read critically the
results that were developed for truck load-out fin this reference,) do provide limited support
for the load-out emission estimates for benzene. " In other words, EPA will continue to use
this faulty information because it, purely by accident, appears to provide limited support to
an EPA-derived number.
e._ Problems with witnessing tests
The citizens were permitted to witness the actual tests at Plant C and Plant D. One individual,
Dr. Nadkarni, was allowed to be present at Plant C and up to 3 individuals, were allowed at
Plant D. While there can be justifiable concerns about visitor safety and owner liability when
visitors are allowed to wander unfettered around an operating industrial plant, the usefulness
and limitations of allowing a single witness need to be discussed.
The first issue is what a single witness can see when data is being gathered using complex
instruments or complex sampling trains in various places simultaneously. Second, when
access is also prevented under the claim of confidentiality, much information is shielded from
citizen observers and discrepancies are seen by the citizen observers only after the data has
been published in a draft report. (See references 8 and 10 for details on the questions raised
about the data.) In an industry where the manufacturing equipment is available from several
competing firms who will provide all the necessary details in order to sell this equipment, and
the product is produced to published customer specifications, our view is that there IS little
proprietary content. If plant operating procedures are proprietary, they would be of interest
Ryan of EPA of October 26, 1995 and November 16, 1995; and R. Nadkarni, written
communication to Mr. David Mobley of July 5, 1996.
-------
16
only to another HMA manufacturer and access to these details would be obtained, as in other
industries, by hiring away a plant operator. Yet, access to information was blocked on many
occasions under the cloak of confidentiality. Following are specific incidents where there were
differences between the observations by the citizen observer and the EPA.
i. Emissions from the downwind end of the tunnel: In the discussions leading up
to the actual testing at Plant C, EPA asserted several times that they had not
observed any fugitive emissions exiting the tunnel at Plant C. However, they
agreed to use tracers to measure capture efficiency. When Dr. Nadkarni
arrived at Plant C on the day before the first test, he was surprised to see
visible emissions exiting from the tunnel. (Such emissions were seen during
each day of the test, typically later in the day when the ambient breeze became
stronger.) Several times when this phenomenon was pointed out to EPA
staff, they ascribed these emissions to a plant malfunction. Since citizens
could not take photographs but EPA could, he requested that this fume be
photographed and several times his request to document this photographically
was turned down because the fumes might not be visible in a photo. When
the draft report was published, the capture efficiency numbers corroborated
this observation that fume was escaping from the downwind side of the
tunnel, contrary to EPA assertions during planning, but there is little
photographic evidence. Yet, at Plant C, the EPA Project Director spent
almost an entire day with a videographer hired by the industry association
consulting him on what to videotape.
In case of Plant D, no tracers were employed because the specially
constructed enclosure was presumed to meet Method 204 requirements. The
shortcomings of method 204 were obvious to the citizen observers on the first
day of testing. Two citizen observers noted that the streamers at the Natural
Draft Openings on the downwind side (above and below the door) were
pointing towards the outside showing that a breeze was counteracting the
draft induced by the fan. Again, there is no photographic evidence. While
this problem was corrected for days 2 and 3 by installing plywood to reduce
the size of the openings by half, EPA claims that this observation was not
communicated to them until much later. (See 3.1.2) Similarly, the citizens
observed that the fume lingered for a long time inside the enclosure at Plant
D and was not transferred to the measurement point, as was discussed earlier
in Section l.b. Again, the citizens could not document this because of the
restriction against photography.
ii. Lack of information sharing and access to instrumentation: At Plant C, the
control room was cleared of all observers several times. Therefore it was not
possible to take actual readings to double check the readings taken by EPA
contractors. (This is not to cast any aspersions on the contractors who took
these readings. Our point is that it was not possible to double check these
readings). EPA had information on the volatile content of asphalt prior to and
-------
17
during testing at both plants. This information was not shared until much
later.
iii. Prejudging of final outcome: On the day before the start of testing at Plant C,
the industry hosted a dinner attended by the senior EPA staff member and the
citizen observer. The citizen observer was quite surprised when the industry
was told by the EPA staff member that EPA had no intention of regulating
fugitive emissions from their industry.
Conclusion
Citizens, industry and EPA-RTP have undergone a lengthy process during the last six years to
examine fugitive emissions from asphalt plants. This process suggests that there are lessons to be
learned regarding a number of aspects of this exercise.
1. An epidemiological study should be undertaken by the government to better understand the
quantity and intensity of health problems caused by exposure to asphalt fume. In our very
limited casual interaction with individuals across the country, we have had too many
anecdotal histories come to our attention to ignore this public health problem. These histories
include the deaths of otherwise healthy farm animals barned within several hundred yards of
an asphalt plant in the northeast section of the country, rashes and difficulty breathing by
humans all across the country, each living in close proximity to one of these facilities, and
cancer clusters in proximity to asphalt storage tanks at terminals. It is hard to believe that
each of these instances is due solely to uncaring plant operators lacking concern for the
impact of their business on the surrounding neighborhood. Even if the cause is determined to
be hypersensitivity by individuals to certain chemicals, this is an issue that should not be faced
by each individual citizen as his or her "problem" but rather is a responsibility for society as
a whole, as well as for the industry.
2. A number of the EPA's regional offices have extensive experience in working through
environmental concerns in concert with citizens rather than in opposition to the public. Some
of the citizens who are working with this problem of asphalt fume have heretofore had
productive relations with their regional offices of EPA on other industrial concerns. It may
be that EPA-RTP has not had reason to develop protocols for interaction with the public,
possibly because more of its business is conducted with industrial representatives.
Nevertheless, it may be appropriate that EPA-RTP review its approach to citizens to find
ways of encouraging that dialogue with an emphasis on cooperation and mutual respect. It
is important that the public understands that it IS no less a partner with government at every
level than is industry.
3. Industry and state environmental departments place great reliance on AP-42 emissions factors
during the process of local permitting. However, AP-42 contains averaged data collected
under the best of circumstances and is a poor source to determine "worst case" environmental
impact scenarios. Many of the factors representing these worst cases are not as clearly
represented in that document as it could be. To avoid misuse, simple changes in the layout
-------
18
could be accomplished such that where appropriate, the limitations of factors are made more
evident, without recourse to the introductory language of this lengthy government
publication. Even a standard "warning" or notice attached to each table could begin to
highlight this complexity.
4. Industrial proprietary information varies from industry to industry. While some industries'
competition may involve truly cutting edge technology, methodology and technique, other
industries' competition is based on non-technical non-proprietary factors such as location and
market power. Restrictions on oversight by the public during EPA financed tests as were
conducted may be appropriate in one instance, and in other instances may more reflect an
attempt by industry to hinder open and frank discourse. We believe that Government should
aggressively challenge, where appropriate, unsubstantiated demands for secrecy by industry
which ultimately reflect negatively on both that industry and on the government.
5. Henry Nowick, Nowick Environmental Associates, Springfield MA
-------
Gary Fore, Vice President
Environment and Safety
National Asphalt Pavement Association
Lanham, MD
-------
NATIONAL ASPHALT PAVEMENT ASSOCIATION
NAPA Building . 5100 Forbes Boulevard . Lanham, Maryland, USA 20706-4413
Toll Free: I-888-468-6499 * Tel: 301-731-4748 . Fax: 301-731-4621
Mike Acott, President
February 1,200 1
Mr. Bob McConnell
EPA Region' 1
1 Congress Street/CAQ "
Suite 1000
Boston, MA 02114
Reference: Draft Hot Mix Asphalt Plants - Emission Assessment Report; AP-42
Section 11.1; Emission Factor Documentation for AP-42 Section 11.1
Dear Mr. McConnell;
In 1996, the National Asphalt Pavement Association began working with government and
citizen groups to determine the amount of emissions that are emitted during silo and truck
load-out operations. In order to develop emission factors for the HMA industry, more
than 390 emission test reports and other types of documentation were compiled and
reviewed. The question of truck load-out and silo emissions has been answered. As a
result, the Hot Mix Asphalt industry is now one of the most thoroughly studied industries
in the United States. During testing, quality assurance scientists from Research Triangle
Park, North Carolina were deployed on-site to independently audit the testing program
and procedures. We have every reason to believe that the test data are representative of
average emissions from HMA facilities.
This test data has proven conclusively that emissions from asphalt plants are low, well
controlled and not "major sources" of emissions. The evidence is quite extensive, the
industry and states need the data, and it is now time to bring finality to the process of
testing.
We appreciate the diligence put forth in successfully completing a project of this
magnitude.
Sincerely, ,
R. Gary Fc
Vice Pr
Environment, ; Health" & Safety
£«0100% RECVCLABI
email: napa® hotmix.org £46100% RECYCLABLE www.hotmix.org
-------
Chapter 4
Hot Mix Asphalt Plants
Response to Comment on
Draft Emissions Assessment Report,
AP-42 Section and AP-42 Background Report
Released December 2000
-------
RESPONSE TO COMMENT ON DRAFT EMISSIONS ASSESSMENT
REPORT, AP-42 SECTION AND AP-42 BACKGROUND REPORT
(Released December 20, 2000 and to be published in the Stakeholder Opinions Report)
This response to comments was prepared by:
Source Monitoring and Technology Group
Emissions Monitoring and Analysis Division
Office of Air Quality Planning and Standards
United States Environmental Protection Agency
Research Triangle Park, NC
December 2000
-------
Contents
1.0 Process. Report Organization. Presentation and Bias 1
1.1 Process 1
1.2 Organization 3
1.3 Perspective 6
1.4 Presentation 7
1.5 Reference Citations 25
2.0 Industry Characterization and Structure 28
2.1 Capacity and Production Levels 28
2.2 Type of Fuels Used by Typical Production Facility 36
2.3 Process Descriptions 39
2.4 Plant Characteristics 40
3.0 Development and Evaluation of Available and Derived Information 43
3.1 Deficiencies in EPA Method 204 43
3.2 Compensation for Asphalt Temperature and Volatility 54
3.3 Adjustment for Background Truck Operation Emissions 63
3.4 Assignment of Quality Ratings 71
3.5 Emission Factor Development Methods 76
3.6 Characterization of Emissions 103
3.7 Evaluation of Unused Information 114
-------
1.0 Process. Report Organization. Presentation and Bias
1.1 Process
1.1.1 CAAP Process Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Commentors Introduction. Section A
Before discussing the results of our review, we want to comment on the review process
and our general dissatisfaction with it. When complex, technical documents are sent to a
wide audience for review, one assumes that both the EPA and their contractors did their
best to produce as final a draft as possible. One would not only expect that most of the
typos and numerical errors have been found and eliminated but that the authors wrote the
report in a way that makes it easy for the reviewer to go back to the primary references
and original sources (assuming that they are available) to check a particular point. On
both these counts, the reports, just like the previous set of reports reviewed last year, fail
resoundingly. Many key comments are not referenced at all; in some cases, the references
are incorrect; in other cases, the reference is made to a large document without giving a
page number, making it difficult and time-consuming to check. Appendix B refers to its
Appendix B and Appendix C on page 4-137. But these appendices, which presumably
contain the statistical results and the data sets, are not provided, making it impossible to
even spot check the analysis. This, combined with the severe time constraints, makes it
impossible to review the reports in any but a cursory fashion, using random spot checks,
to make sure that the numbers are correct. Unfortunately, the reports also fail on that
count since numerical or logical errors in the analysis have been found.
1.1.2 Response to CAAP Process Comment Number 1 - The commentors have identified a
few areas where the report could benefit by more specificity in the wording of the
statements. However, EPA believes that these reports present this large amount of highly
technical information in a clear and organized fashion. Most respondents did not
comment on the reports structure which appears to indicate most reviewers could
understand the reports as drafted. The comment with respect to numerical or
calculational errors is unsubstantiated in that no instances of any numerical or
calculational errors were provided by the commentors. The primary references for this
report are the almost 350 emission test reports used to develop the emission factors for
the source category. All of the key comments in the report refer to the information
contained within these test reports or were derived from compiled information contained
in these test reports. While some ancillary information contained in the report was not
referenced in the draft report, additional reference citations supporting this information
were added to the final report. While we do not believe it is necessary to cite individual
references in the main report, each of the test reports are cited in Appendix A and B. To
assist the reader, in the main report we have included the location in Appendix A, B or C
where more detailed reference citations are available. These test reports (which were
summarized in Chapter 4 of Appendix B) are generally over 50 pages in length and some
may be over one thousand pages (like the Plant C test report). EPA obtained most
non-EPA test reports from State or local air pollution control agency files. Typically,
copies of these reports are not routinely provided to reviewers. However, specific source
1
-------
test reports can be provided upon request. While the detailed results of the statistical
analysis were inadvertently excluded from the report, the data were presented in the tables
in chapter 4 and the methodology and conclusion was presented in that chapter as well.
The detailed results of the statistical analysis were provided to all of the commentors on
October 6, 2000 and are included in the final report.
1.1.3 CAAP Process Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Commentors Introduction, Section A
We also need to air another complaint regarding the process. For about 6 weeks before
the report was finally issued, we received phone calls informing us that the release of the
reports was imminent. While we appreciate receiving this advance warning, it also turns
out that the actual release of the reports was much later than announced. One of us (Ravi
Nadkarni) had cancelled attendance at the Annual Meeting of IPMI in mid-June in
response to these calls. Because of weddings in the family, Ravi was going to be
unavailable after July 17 and wanted to start the review process as soon as possible, hence
the cancellation. Unfortunately, the reports were received only on June 28. This means
that he missed an important Annual Meeting where he is on the Board of Directors. We
appreciate getting the extra time until the end of August to complete this review.
Appendix C was prepared in 1996. Why was this 69-page document not provided earlier
so that it could have been reviewed before this "crunch" period? Also, are comments on
such a document, which has undergone a "final" review, relevant at this stage? Finally,
why was this document not released to us when the citizen groups and the EPA were
discussing sampling issues, when that document might have provided an additional
viewpoint? For all these reasons, this third document has not been reviewed.
1.1.4 Response to CAAP Process Comment Number 2 - EPA reasonably accommodated
reviewers by providing additional time and flexibility for the review of the Emissions
Assessment Report. While the report was release later than originally anticipated, EPA
provided notice of this delay and extended the review period for the report to provide
flexibility to the reviewers schedules. Additionally, most of the information in the draft
report had been available for over two years on the Emissions Factors and Inventory web
site and both of the commentors had previously commented on at least two items in this
earlier draft report. EPA postponed completing this 1998 draft AP-42 Section and
background report pending completion of the load-out emissions testing. Almost all of
the stack test data and analyses in the current draft were also contained in the previous
draft. The information added since 1998 concerned the development of factors for the
silo filling, load-out and yard emissions based largely on tests conducted at the request of
the commentors and other stakeholders with similar interests. The data and EPA analyses
for this information were clearly available to all of the stakeholders and were identified in
the EPA report Response to Comments on Testing for Asphalt Plants C and D. There
was a limited amount of information that was not available to the stakeholders until we
circulated the draft. This information included the approximately thirty emission test
reports submitted by the State of Wisconsin in response to the circulation of the first draft
of the AP-42 Section and background report for Hot Mix Asphalt Plants. In addition, the
-------
36 page main body of the Emissions Assessment Report was also new to reviewers. The
review time provided was more than adequate to review information not previously made
available for review.
With regard to Appendix C, the purpose of this appendix was to provide background
information to reviewers on how the revised emission factors could be used with
available techniques for preparing an emission inventory for hot mix asphalt plants. This
type of information is not presented in AP-42 sections nor in the background reports for
the AP-42 sections. As stated in the introduction to the EIIP document (Appendix C),
"The purposes of the preferred methods guidelines are to describe emission estimation
techniques for stationary point sources in a clear and unambiguous manner and to provide
concise example calculations to aid in the preparation of emission inventories. While
emissions estimates are not provided, this information may be used to select an emission
estimation technique best suited to a particular application." It should be noted that the
primary audience for the EIIP reports are engineers and environmental scientists that are
assembling local or regional emissions inventories. Information on the availability of the
draft of this EIIP document was provided in a letter from Ronald B. Ryan of EPA to Dr.
Ravinda Nadkarni (with copy to Lloyd Pillion) dated September 5, 1995. The letter
clearly states that the purpose of the EIIP document is to explain how to inventory
emissions from hot mix plants. The letter further stated that additional information on
fugitive emissions from truck loading was not included in the document. The draft
document was available for comments on the EPA web from October 1995 to about
March of 1996. The commentors did not request a paper copy of the draft EIIP document
and did not provide any comment on this document at that time. As with most emission
factor and inventory documents, the documents are revised from time to time. All of the
commentors were advised that Appendix C was a final report in EPA's June 2000 cover
letter transmitting the Emissions Assessment Report. Additionally, this cover letter stated
"Comments on Appendix C of the EA Report will be forwarded to the EIIP Point Source
Committee for their consideration in future revisions."
1.2 Organization
1.2.1 CAAP Organizational Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Commentors General Comments, Section B
AP-42 is essentially a "cookbook" for engineers/contractors who need to produce
emissions estimates for local regulatory authorities to obtain permits to construct and
operate plants. In view of this, the overall organization of the three reports does not make
sense. In the long run, Appendix A will be used most often for estimating emissions.
When necessary, the background information in Appendix B will be reviewed. The
"main" report will remain either as a political document used by lobbyists or fade away.
As it now stands, the "main" volume repeats in a vague and biased way, the information
in first two appendices. The same information is presented in a marginally better format
and with slightly less "spin" in Appendix A. If the main document is supposed to be an
Executive Summary to Appendix B, it is a poor summary and reads more like a political
tract. This is particularly true since only emissions data on plants using natural gas is
-------
given in the main report; i.e. there is a deliberate attempt to show only the lowest
emissions. Most of the caveats associated with the summary information in the main
report are well concealed in Appendix B. If EPA is serious about making the main
document useful, it should use the main document to calculate the total emissions to the
environment from a batch and a drum asphalt plant. This type of rewrite, where other
emissions (for example from materials handling and storage piles, storage tanks, from
traffic and truck exhaust, and so on) are also calculated and shown for typical plants, will
result in a guidebook for industry contractors and help them by demonstrating how such
calculations should be made. (In Massachusetts, specific plants will employ a wide range
of pollution control technology and then argue that it is "state of the art" and is BACT
(Best Available Control Technology)). Other emissions are currently only listed and
cross-referenced in Section 2.3 on page 16. These include emissions from storage tanks
(AP-42 Chapter 7), materials handling and storage piles (AP-42 Sectionsl 1.19.2
and!3.2.4), vehicular traffic (AP-42 Sections 13.2.1 and!3.2.2) and emissions from
trucks bringing in the raw materials and taking away the hot mix asphalt (AP-42 Volume
II).
1.2.2 Response to CAAP Organizational Comment Number 1 - The organization of the
three reports was designed to meet the needs of a diverse audience. The "main" report
presents a summary of the detailed technical information presented in Appendices and
provides emission estimates for a typical facility. Because of limiting the volume of
technical information, information representing HMA plants using fuel oil (an estimated
10 to 30% of the fuel used by the industry) were omitted from the "main" report. The
presentation of technically factual information representing the two different production
methods of average size, using fuel used by between 70 and 90% of the industry, is most
representative of nationwide emissions from hot mix asphalt plants. While it is
recognized that the use of fuel oil results in higher emissions of some pollutants, fuel oil
is seldom used when natural gas is available and is primarily used as a backup fuel.
However, it is recognized that the "main" report should include information that fuel oil
is used to produce approximately 10 to 30% of the HMA, that the use of fuel oil creates
higher emissions. The main body of the report will include a revised summary of
emissions for both oil fired and gas fired facilities.
As stated above, the purpose of the "main" report is to provide a summary of more
technical information. It is agreed that the "main" report may benefit by the addition of
fugitive emissions calculated from emission factors contained in other AP-42 sections.
These emissions include diesel exhaust, dust from truck traffic and dust from raw
aggregate handling. The main body of the report is not a stand alone guidebook for
contractors. Industry contractors and State agency permit and inventory specialists will
use AP-42, the EIIP document (included as Appendix C) and other supplementary
information to make the required calculations.
-------
1.2.3 CAAP Organizational Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Commentors General Comments, Section B
Finally, it is unclear whether this "main" report has a number or whether
EPA-454/R-00-OXX is the final designation. The reason for asking this is because
reference 1 of Appendix A has the same name and a designation of EPA-454/R-00-019,
but this reference is dated May 2000. What is going on?
Commentors Summary Comments, Section C
Confusing Report Designations: The designation of the reports is confused. The main
report is called EPA-454/R-00-OXX. But Reference 1 of Appendix A has the same name
and a designation of EPA-454/R-00-019, and this reference is dated May 2000 while the
main report is dated June 2000.
1.2.4 Response to CAAP Organizational Comment Number 2 - The EPA report number
was requested for this report in advance so that all of the reports associated with the Hot
Mix Asphalt projects would be numbered consecutively. The EPA report number on the
title page of the report was changed to XX to avoid the draft report being considered a
published EPA report. The actual date of publication of the report was used for the final
report. The use of this report as the citation for the first reference in Appendix A is
circuitous. Appendix A and Appendix B are being made available as separate
independent reports through the various mechanisms used to distribute the AP-42
sections and their associated background reports. As a result, a more proper citation
reflecting the Appendix B material was used.
1.2.5 CAAP Organizational Comment Number 3 (Ravi Nadkarni & Lloyd Pillion) -
Commentors Summary Comments. Section C
Sloppy Report Preparation: The reports are full of minor errors, are poorly organized,
repeat the same paragraphs in different documents, and have incorrect references. Many
key comments are not referenced at all; in some cases, the references are incorrect; in
other cases, the reference is made to a large document without giving a page number,
making it difficult and time-consuming to check. The data sets used for statistical
analysis are not provided, making it impossible to even spot check the analysis. All
indications are that there was no complete internal review of the report within EPA before
releasing it.
Commentors Summary Comments. Section C
Poor Organization: The overall organization of the three reports does not make sense. In
the long run, Appendix A will be used most often for estimating emissions. When
necessary, the background information in Appendix B will be reviewed. The "main"
report will remain either as a political document used by lobbyists or fade away. The
main report shows the industry's emissions in the best possible light with most of the
caveats associated with information being well concealed in Appendix B. If EPA is
serious about making the main document useful, it should use the main document to
calculate the total emissions to the environment from a batch and a drum asphalt plant,
rather than show the information from AP-42 with the decimal point moved to the right.
-------
1.2.6 Response to CAAP Organizational Comment Number 3 - These reports present the
large amount of highly technical information in a clear and organized fashion, and
therefore it is reasonable to encounter some minor editorial errors in the final report
review. While the document benefitted by some editorial corrections, the reports
contained few errors and these errors were not substantive.
The organization of the three reports was designed to meet the needs of a diverse
audience. In order to meet these needs, some information was repeated in all four parts of
the report. As explained elsewhere, all three Appendices are available separately.
Therefore, some background information is repeated so users do not have to obtain all
three Appendices to understand any one Appendix. While future uses of the "main"
report will be modest compared to Appendix A and B, it is unlikely to be used to support
any air emissions permit for any new or existing facility. The information presented in
the reports as drafted are unbiased.
1.2.7 CAAP Organizational Comment Number 4 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
1. Page 1-1. final paragraph: The organization of the entire report is confusing. The last
line refers to Section 5, which is the same as Appendix A. Which is the main report
and which are the subsidiary appendices?
1.2.8 Response to CAAP Organizational Comment Number 4 - As stated previously, the
report is organized such that each part of the report can stand by itself. The main body of
the report provides summary information from the three Appendices. Appendix A is the
AP-42 section for Hot Mix asphalt which is generally written for use by air pollution
professionals. Appendix B is the background information for the AP-42 section and is
typically used by researchers and the few air pollution professionals looking for more
detailed information than is presented in the AP-42 section. Appendix C is an existing
document describing emission inventory methodologies applicable to Hot Mix Asphalt
Plants and was published for air pollution professionals specializing in emission
inventories and air quality management.
As each of the Appendices is intended to be stand alone reports, the last line was revised
to "The final AP-42 Section 11.1, Hot Mix Asphalt Production is presented elsewhere."
1.3 Perspective
1.3.1 CAAP Comment on Report Perspective (Ravi Nadkarni & Lloyd Pillion) -
Commentors General Comments. Section B
The main report is a political document for the following reasons. The EPA's
pro-industry bias shows on almost each page. Every attempt is made to minimize
emissions, for example, by calculating emissions for plants burning natural gas only in
the "main" report. The report also provides qualitative comments to indicate that
emissions could be even lower in many cases; for example, "counterflow drum mix plants
-------
will likely have lower organic compound emissions than parallel flow drum plants", page
11.1-9. Such comments are not substantiated in the detailed information provided in
Appendix B. Yet, caveats regarding conditions under which the emissions might be
higher are few and far between. This tilting is not surprising given the proindustry bias
expressed many times by the Project Officer during the course of testing.
Commentors Summary Comments. Section C
Pro-industry Bias: The EPA's pro-industry bias shows on almost each page. Every
attempt is made to minimize emissions shown. The report also provides qualitative
comments to indicate that emissions could be even lower in many cases. We don't see
why EPA makes such statements in the absence of data. Yet, caveats regarding conditions
under which the emissions might be higher are few and far between.
1.3.2 Response to CAAP Comment on Report Perspective - The EPA has conducted a fair
and balanced process to consider stakeholder's opinions during the development of
emissions information for hot mix asphalt industry. Consistently, EPA-has
accommodated CAAP concerns at additional cost and delay in completing EPA's
evaluation of hot mix asphalt plant emissions. In response to these latest comments, EPA
has made additional changes to the Emissions Assessment Report. As stated elsewhere,
the report was revised to suggest that the natural gas is used to manufacture HMA at
approximately 70 and 90% of the HMA facilities in the US. Where appropriate, the
report includes separate emission factors for oil-fired and gas-fired HMA plants.
The commentor quotes a statement in a part of the report that describes general process
information. There are no statements in the "main" report or Appendix A which would
provide support for using a lower emission factor in the report. It is not clear how the
reporting of technically accurate information based on known engineering principles to
generally characterize "typical" facilities of each, major production process, fired with the
most commonly used fuel to fire most of the HMA plants in the U.S., shows a bias.
1.4 Presentation
1.4.1 Industry Presentation Comment Number 1 (Steve Zemba, Cambridge
Environmental) -
General: A list of acronyms at the front of the document might e helpful to readers.
1.4.2 Response to Industry Presentation Comment Number 1 - A list of acronyms was
added to the front of the document.
1.4.3 Industry Presentation Comment Number 2 -(Steve Zemba, Cambridge
Environmental) -
Hot Mix Asphalt Plants Emissions Assessment Report
The last sentence third pgh. of pg 1 contains a typo: the 100,000 tons of production
corresponds to a batch plant, not a drum plant.
-------
1.4.4 Response to Industry Presentation Comment Number 2 - The word drum was
changed to batch in this sentence.
1.4.5 Industry Presentation Comment Number 3 - (Steve Zemba, Cambridge
Environmental) -
Hot Mix Asphalt Plants Emissions Assessment Report
1st pgh. of p. 10: Does the emission assessment report provide stack emission rates for a
typical counterflow or parallel flow drum plant? If the estimates are for a parallel flow
plant, it might make sense to mention in the text the approximate difference expected for
emissions from a counterflow plant (e.g.. this could be mentioned in a footnote to
Table 2).
1.4.6 Response to Industry Presentation Comment Number 3 - The emissions estimates
provided in Table 2 are appropriate for both parallel flow and counter flow drum plants.
While the statements in the AP-42 section and background reports concerning potential
differences in organic emissions between these two dryer designs are based on valid
engineering criteria, sufficient information is not available to quantify these differences.
As a result, data from both types of facilities were combined to determine the volatile
organic and organic particulate emission factors.
1.4.7 Industry Presentation Comment Number 4 - (Steve Zemba, Cambridge
Environmental) -
Hot Mix Asphalt Plants Emissions Assessment Report
p. 3: EPA might want to provide the regulatory definition or a better explanation of
criteria pollutants and HAPs.
1.4.8 Response to Industry Presentation Comment Number 4 - A brief regulatory
definition of criteria pollutants and HAPs was included in the Glossary which be added to
the document.
1.4.9. Industry Presentation Comment Number 5 - (Steve Zemba, Cambridge
Environmental) -
Hot Mix Asphalt Plants Emissions Assessment Report
p. 2, 3rd pgh: Might want to mention the number of tests (> 300) that survived the
screening process (i.e., were usable for the development of emission factors).
1.4.10 Response to Industry Presentation Comment Number 5 - We agree that more
detailed information on the number of tests that were used to develop the resulting
emissions factors is needed in this paragraph. In addition to stating that over 300 tests
cleared the initial screening process, general information on the number of tests that
support broad classes of pollutants was provided. Also, the reader was referred to
Appendix A and B for more specific information on the number of supporting tests and
quality ratings of individual pollutants. The paragraph was revised as follows: "To
develop emission factors for the HMA industry, data from about 370 emission test reports
and other documents on the industry were compiled and reviewed. Through a careful
-------
screening process, about fifty of the reports were determined to be unusable for emission
factor development and were excluded from further evaluation. About 310 reports
remained and were compiled by plant type, emission source, pollutant, and emission
control. For each emission test, emission factors were calculated by dividing the
measured emission rates by the HMA production rate measured at the time of the
emission test. These emission factors were then grouped by source, pollutant, and control
device, and an average emission factor was calculated for each group. While particulate,
CO2, CO and TOC emission factors are based upon over one hundred tests, most of the
remaining criteria pollutant emission factors are based on between five and ten tests. A
few HAP emission factors are based upon more than five tests, although the majority are
based upon between two and five tests. Information on the supporting data for specific
emission factors and the quality rating assigned to the emission factor is included in the
emission factor tables in Appendix A and B. Generally, the number of supporting data
are typical of many AP-42 sections. However, the number of data supporting particulate,
CO2, CO and TOC emission factors is greater than most AP-42 sections."
1.4.11 Industry Presentation Comment Number 6 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 4: A. Add in (source classification code in parentheses) under Figure 1 Typical
Page 5: A. Add in (source classification code in parentheses) under Figure 2
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-3: A. Add in (source classification code in parentheses) under Figure 11.1-1
General...
Page 11.1-4: A. Add to Figure 11.1-2. General process flow diagram for parallel-flow
drum mix asphalt plants.
B. Add in (source classification code in parentheses) under Figure 11.1-2 General...
Page 11.1-6: A. Add in (source classification code in parentheses) under Figure 11.1-3
General...
1.4.12 Response to Industry Presentation Comment Number 6 - The titles of the figures
were revised to include that the source classification codes are in parentheses within the
figures and that Figure 11.1-2 is a parallel-flow drum mix plant..
1.4.13 Industry Presentation Comment 7 - (Gary Fore, Vice President National Asphalt
Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 2: A. Paragraph 1, second line -please add the following: ...and PM-2.5 and very
small amounts of hazardous air pollutant. ..
-------
C. Second sentence - please add the following: ...VOC and very small amounts of
volatile HAP organic compounds.
Page 10: B. 4th paragraph - Please correct as follows: ... CO; and very small amounts of
organic compounds ...
Page 11: A. 1st paragraph - Please add the following: ... organic aerosol may contain
very small amounts ...
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-7: A. In the fourth line, second paragraph under 11.1.2.1 - Batch Mix plants,
please add the word very before the word small. The sentence will then read "and very
small amounts of organic compounds"
Page 11.1-8: B. In the second line from the top of page - please add the word very before
the word small. The sentence will then read "may contain very small amounts of HAP"
Page 11.1-9: A. I n the first paragraph under section 11.1.2.3 - please add the word very
before the word small. The sentence will then read "and very small amounts of organic
compounds of various species"
1.4.14 Response to Industry Presentation Comment Number 7 - The additional modifier
"very" would not add any substantial information to the sentences and may be
misinterpreted to be a relative indication that the emissions of these compounds or the
relative proportion of these compounds from HMA plants are less than other similar type
sources.
1.4.15 Industry Presentation Comment Number 8 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-10: A. 2nd paragraph -Suggest a more typical loss-on-heating of 0.25 not 0.41
a. To estimate total PM emissions from drum mix plant load-out operations using an
asphalt loss-on-heating of 0.25 percent and temperature of 290°F, HMA mix, the
following calculation is made:
EF = 0.000181 + 0.00067 (_v)e«0-0251H290 + 46°)
0.000181 + 0.00067 (.(0.25))e((ao251)(290
0.000181 + 0.00067(0.25)e(-L605)
0.000181 + 0.00067(0.25)(0.2009)
0.000181 +0.0000485
0.000215 Ib/ton of asphalt loaded
10
-------
NOTE: The correlation factor in the equation should be 0.00067 and not 0.00059.
This loss-on-heating value is a value that is seen on a daily basis within the Industry.
1.4.16 Response to Industry Presentation Comment Number 8 - As indicated in the
sentence preceding the equations, this is an example to show the user how to calculate the
emissions with the various equations presented in Table 11.1-14. A temperature and a
loss-on-heating value was selected for this equation that would not be confused with the
recommended default values in the footnote to Table 11.1-14. We agree that the
constant for the MCEM portion of the equation should have been 0.00067. However,
based upon other comments, the batch and drum mix equations were combined into one
equation. As a result, the revised constant value of 0.00141 was used in the example
equation.
1.4.17 Industry Presentation Comment Number 9 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-17 : A. In the title for Table 11.1-6, please delete the word formaldehyde from
title
1.4.18 Response to Industry Presentation Comment Number 9 - The word formaldehyde
was deleted since it is not included in the table.
1.4.19 Industry Presentation Comment Number 10 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
General comment - Tables tend to be confusing. For example, Place HAPs contiguously
with HAPS and follow with Non-HAPs together. Also, suggest adding a set of batch table
and drum tables with contiguous HAPs listing, non- HAPs listing and metals HAPs vs.
non-metal HAPs listing.
General comment - Suggest following the same table order as found in the Emissions
Assessment Report. Much less confusing.
Page 11.1-20 :
A. Sub title: Hazardous air pollutants under Pollutant Heading - For easier reader
understanding, please change subtitle from Hazardous air pollutants to non PAH/HAPs
B. Please add a line totaling the non-PAHs hazardous air pollutants section. Currently
very confusing.
C. Sub title PAHs - Please change from PAHs to PAH/HAPs
D. Please add row and sum the NO.6 fuel "organic HAPs"
11
-------
E. Please change non-HAP subtitle to Non Organic HAPs
F. Sub title: Hazardous air pollutants -Please change to PAH/HAPs
G. Left side column under Dryer with fabric filter - Please add natural gas, propane,
fuel oil, or waste oil fired. Confusing as is and fails to distinguish between these fuels
and No.6 fuel oil.
G. Footnote b -Please change as follows: Natural gas, propane, fuel oil, or waste oil fired
dryer. I1 or pollutcints tlicit cire nicUKed witn 3, , scp3.i3.tc emission icictors cire presented
for No. 6 fuel oil fired dryers. Substitute these factors for above factors when burning
No.6 oil.
Page 11.1-21,22,23: A. Use the same format and comments as suggested for above Table
11.1-9
Page 11.1-22 and 23: A. Please add Dioxins and Furans to above referenced Organics
and HAPs
Page 11.1-24:B. Footnote b - Please change as follows: Natural gas, propane, fuel oil, or
Wciste oil tired dryer, l1 or pollutcints tlicit cire 1113.1 Ked witn ci , sep3.i3.te emission icictors
are presented for No. 6 fuel oil fired dryers. Substitute these factors for above factors
when burning No.6 oil.
Page 11.1-27:
A. Overall comment - Should create subtitles for Metal HAPs category and a Metal Non-
HAPs categories. Also suggest moving to 11.1-9.
B. Adjust footnotes accordingly.
Page 11.1-28:
B. Use the same format and comments as suggested for above Table 11.1-9. Should be
the same for both drum and batch.
C. Please delete Footnote b. unnecessary.
D. To make the table easier to read, suggest switch last two sections of the table. The
section pertaining to lead and mercury should come last.
Page 11.1-30:
A. Place PAH/Organic PM (%) up above PAHs line in the left column. It should be a
title, not a part of the data.
B. Change Subtitle from PAHs to PAH/HAPs
12
-------
C. Total PAHs, subtitle -Change to total PAH/HAPs
Page 11.1-31:
A. Move subtitle -Compound TOC (%) up to be included in the titles for the columns.
B. Create Sub title - non-VOC/non-HAP's
C. Please a Percent sign (%) to every EF as in table 11.1-15. It is currently very confusing
because the other tables in this document do not contain percents.
D. Total- add line total volatile organic HAPs
E. Please add to the example on page 11.1-10. It would be helpful to illustrate the use of
Table 11.1-16.
1.4.20 Response to Industry Presentation Comment Number 10 - The reorganization of
the emission factors tables for organic and metal HAPs address most of the issues
identified by the commentor. Although the revised presentation causes repetition of most
of the compound or element specific emission factors, this presentation makes this
information less confusing. Within source category types, the compounds were
segregated into broad classifications (non-HAP organic compounds, PAH HAP,
non-PAH HAP, Dioxin and Furan) and the sums of these classifications were provided.
These three changes should make it easier for the reader to use this information. The
following suggested revisions of specific details in the tables were also accommodated.
Specific comment for page 11.1-27 (A & B) - Rather than creating HAP and non-HAP
categories of metals in this table, a separate footnote was created. To more clearly alert
the reader of the HAP classification of the element, the footnote letter now follows the
metal name in the pollutant column. Specific comment for page 11.1-30 (A, B & C) -
The title was moved to the top of the column as a single heading for the load-out and yard
column and the silo filling column. Also, the subtitles in the Pollutant column for PAH
and Total PAH's were changed to PAH HAPs and Total PAH HAPs. Specific comment
for page 11.1-31 (A) - The subtitle was moved to the top of the column as a single
heading for the load-out and yard column and the silo filling column. Specific comment
for page 11.1-31 (B) - The speciation profile for VOC was moved to the top of the
column and a subtitle non-VOC/non-HAP was added between VOC and the other
compounds. Specific comment for page 11.1-31 © - Percent signs were added to the
values for each speciation profile percentage. Specific comment for page 11.1-31 (D) - A
row was added to the bottom of the table to present Total Volatile Organic HAPs.
Specific comment for page 11.1-31 (E) - An example to illustrate the use of speciation
information in Tables 11.1.15 and 11.1.16 was added to page 11.1-10 between the
paragraph describing the speciation profiles and the paragraph describing methods to
estimate emissions from asphalt storage tanks.
13
-------
1.4.21 Industry Presentation Comment Number 11 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-29: A. Questioning] why [the table] include[s] total PM when it includes
organic PM? This is very confusing.
1.4.22 Response to Industry Presentation Comment Number 11 - Total PM does include
organic PM. Organic PM is included as one of the known components of fine particulate
matter as this information may be needed in air quality management analysis for PM fine.
The speciation of the total PM (also PM-2.5) into the organic and inorganic components
provides some information that will be needed to help prepare regional control strategies.
EPA has established a network of air samplers that determine the constituents of fine
parti culate matter. The important major constituents include sulfates, nitrates,
ammonium, organic carbon (a.k.a. organic PM), elemental carbon and a number of
elements. Similar information is presented for the stack emissions although the
nomenclature is slightly different. This difference in nomenclature recognizes that the
organic constituents in stack gas are vapors and will condense after cooling to ambient
temperatures. The users of AP-42 and FIRE should be aware that, for stack sources, total
PM is the summation of Filterable PM, condensible inorganic PM and condensible
organic PM.
1.4.23 Industry Presentation Comment Number 12 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-29: B. Footnote a. -Please change the word asphalt to HMA mix to prevent
confusion as to mix temperature or liquid temperature.
1.4.24 Response to Industry Presentation Comment Number 12 - The text in footnote a
which defines the term T in the equation was changed to "T = FDVIA mix temperature
in °F".
1.4.25 Industry Presentation Comment Number 13 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-32: A. Please correct Reference No.2. Katherine is spelled Kathryn
1.4.26 Response to Industry Presentation Comment Number 13 - The correct spelling
was used.
1.4.27 Industry Presentation Comment Number 14- (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 2: B. Second sentence - delete the "e" off of the end of with
14
-------
1.4.28 Response to Industry Presentation Comment Number 14 - The spelling of with
was corrected.
1.4.29 Industry Presentation Comment Number 15 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
D. paragraph 2 - please add the following: ...emissions for specific facilities where
source specific emissions is not available or where source testing is cost prohibitive.
1.4.30 Response to Industry Presentation Comment Number 15 - The phrase "where
source specific emissions are not available or where source testing is cost prohibitive"
was added to the end of the last sentence in the second paragraph.
1.4.31 Industry Presentation Comment Number 16 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 10: A. 2nd paragraph - Please add the following: ... counterflow drum mix plants
will likely have organic stack emissions ...
1.4.32 Response to Industry Presentation Comment Number 16 - The clarifying word
"stack" was added to the sentence.
1.4.33 Industry Presentation Comment Number 17 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 11: B. Last paragraph - Please add the following: An initial screening the of these
documents...
1.4.34 Response to Industry Presentation Comment Number 17 - The sentence was
corrected to say "An initial screening of these documents...."
1.4.35 Industry Presentation Comment Number 18 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 12: A. 1st paragraph - Please add the following: ... truck prior to these departure for
the job site")". Missing end parentheses.
1.4.36 Response to Industry Presentation Comment 18 - The closing parenthesis was
added to the end of the phrase.
1.4.37 Industry Presentation Comment Number 19 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
PagelS: B. 4th paragraph - Prior to the last sentence please add the following sentence. ...
developed from the data. For example, two scientific auditors from the Research
15
-------
Triangle Institute were employed to independently audit the test and reporting
process. These additional steps ...
1.4.38 Response to Industry Presentation Comment Number 19 - The additional sentence
was added to the paragraph prior to the last sentence.
1.4.39 Industry Presentation Comment Number 20 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 20: A. Sub title PAH - Please change as follows: PAH (semi volatile HAPs).
1.4.40 Response to Industry Presentation Comment Number 20 - The parenthetical
phrase was added to the sub title.
1.4.41 Industry Presentation Comment Number 21 - (Gary Fore, Vice President National
Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 20: B. Please delete line between methylene chloride and MTBE
1.4.42 Response Industry Presentation Comment Number 21 - The line was deleted.
1.4.43 CAAP Presentation Comment Number 1 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
12. Page 4-2 to page 4-117: On these pages, each reference is reviewed and commented
upon. There are several problems with this section. Initially, section 4.2.1.1
discusses reference 1, section 4.2.1.2 discusses reference 2 and so on. By the time
one reaches the end of this chapter, the numbering is no longer synchronized. Section
4.2.1.327 refers to reference 354. In between, some references are not discussed (42,
43, 115, 116, 120, 127, 131 and so on). In other cases, missing references are
discussed, for example reference 258. In addition, the same reference numbers refer
to different documents. For example, on page 4-6, during the discussion of reference
23, references 26 and 27 are discussed. One has to read the citations carefully to
realize that these are references from the document called reference 23. But the
exact citations for 26 and 27 are not given. A careful reading also shows many
inconsistencies. In some cases, the data is downrated because some information is
missing; in other cases, it is not. See for example reference 181 where the data is
still rated A although information about RAP processing was not provided. On page
4-106, it appears that the cross references are incorrect.
1.4.44 Response to CAAP Presentation Comment Number 1 - Historical consistency of
reference numbers with previous versions of this section was considered a more
important criterion than synchronizing the paragraph numbering with the reference
numbering. This consistency facilitates the understanding of some of the differences in
national, regional and site specific emission inventories over time. In addition, this
16
-------
consistency assists individuals reviewing or revising existing permits that include
emission estimates made with emission factors from previous versions of AP-42.
The lack of synchronization beginning at paragraph 4.2.1.42 is the result of
approximately 40 source tests excluded from emission factor derivation due to critical
deficiencies. Rather than discussing the nature of the critical deficiencies for each test in
the text, Table 4-1 provides the reasons the references cited are not considered further.
Some of the critical deficiencies include "Flow rates provided; cannot calculate emission
rates," "No production data provided," "Insufficient process description and production
data," and "Test methods not comparable to EPA reference methods."
Reference 258 (described in paragraph 4.2-233) was not missing and was summarized on
page 4-84 and listed in the references on page 4-365 of the draft Appendix B report.
Reference 284 (described in paragraph 4.2.1.258) was also not missing and was
summarized on page 4-91 and 4-92 of the draft Appendix B report and listed in the
references on page 4-336.
Reference 23 is a document that provides information concerning the particulate size
distribution of kiln stack emissions. As indicated in the discussion, the information is
considered a secondary reference since the report provides limited detail on the contents
of unavailable test reports. Also the discussion states that primary reference material is
available for some information cited in the report. The discussion explains that because
of a lack of the primary reference information, some emission factors were developed
based on the secondary reference when no primary reference information was available.
It is recognized that a reader may misinterpret the reference numbering within the
discussion of Reference 23 and that citations for references 26 and 27 of Reference 23
need to be provided. We have revised the way this information is presented and cited so
that the reader can better discern that these citations are internal to reference 23.
Data are down rated because of the lack of critical information for quantifying the mass
emissions or for quantifying the production rate in units of the emission factor developed.
Data were not down rated due to the lack of information on normal operating parameters
although this information might result in a slightly better understanding of the effects of
these operating parameters on emissions. While it would have been helpful in evaluating
the relationship between RAP usage and emissions, the lack of information on RAP usage
does not affect the ability to assessing the quality of the measured emissions.
1.4.45 CAAP Presentation Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
12. Page 4-2 to page 4-117: The section ends with the review of reference 354 on page
4-117. This review resumes on page 4-167. This is poor formatting for a report.
17
-------
1.4.46 Response CAAP Presentation Comment Number 2 - To facilitate the issuance of
this report, including the addition of approximately 30 references, we have adopted the
format summarizing all of the references in one section. As with the draft background
report, we considered keeping the information related to the separate processes together
to improve the users comprehension of the emission factor development for each
emission point within a facility. However, this would have required renumbering the
more recent references for load-out and silo filling to present unrelated processes
separately. We believe that maintaining the numerical order of the references in the final
report with the numbering in the draft is more important.
1.4.47 CAAP Presentation Comment Number 3 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
13. Pages 4-118 through 4-147: There are two sections in this report: 4.2.4 Results of
Data Analysis and 4.3 Statistical Approach which more or less cover the same
ground. They should be combined so that the exposition is complete and consistent.
In many cases, common sense should tell us that the emissions to the atmosphere are
either a result of the performance of a control device or the result of the presence of
a pollutant in the feedstock. A good example of the former is the PM emission
factor from dryers controlled with a fabric filter. Here, the emission factor for PM is
not dependent on fuel type. An example of the latter, is the sulfur dioxide emission
factor for dryers. Here, the emissions are dependent on fuel type since the control
devices do not remove any substantial amount of this pollutant. Yet, these issues are
never discussed.
1.4.48 Response to CAAP Presentation Comment Number 3 - We did not integrate the
two analyses because we believed it would make it more difficult for users familiar with
other AP-42 background reports to understand the basis upon which the emission factors
for HMA were developed. Essentially all of the statistical analyses performed could not
discern effects that many knowledgeable individuals, with good engineering judgement,
would support. The inability of the statistical analyses to discern statistically significant
effects are discussed elsewhere (see response 3.5.2 and 3.5.16). The HMA source
category is one of few AP-42 sections that had sufficient data for us to consider a
statistical analysis.
Generally, decisions on combining or separating data from different production methods
or different control types are justified with good engineering judgement (e.g. common
sense). Integrating the inconclusive statistical analyses, which provide no useable
information, and the practical information typical users need, will obfuscate their
understanding of the decisions we made to develop the emission factors. Additionally,
we believe that some users would misinterpret the statements in the statistical analyses
and perceive the quality of the factor is lower than the data indicate.
18
-------
We have identified similar commentor misinterpretations concerning the criteria used for
establishing the emission factor quality ratings. Specifically, in the comments presented
in section 3.4.1, 3.4.3, 3.4.5 and 3.4.9, the commentor considered variation between
sources (either range of data or standard deviation) as a primary criterion for assigning a
low quality rating. Variations between sources is normally used as one of several
secondary criteria to make minor adjustments to a rating. We believe that integrating the
statistical analyses would further exacerbate misinterpretations of the established
emission factor rating criteria. This is especially true when most statistical analyses were
inconclusive even when there were hundreds of supporting data. Presenting the
traditional emission factor derivation methods separate from the voluminous inconclusive
statistical analyses will enable customary users of the background reports to better
understand the derivations.
1.4.49 CAAP Presentation Comment Number 4 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
19. Page 4-137. paragraph 1: The description of the box plots is extremely short and
incomplete. Obviously, the authors do not intend to make it easy for readers to
understand these plots. Technically, what is shown is a "Box-and-Whisker" Plot,
which is used to show characteristics of univariate data. The data is ordered and
ranked and then divided into four equal parts. For example, a data set of 20 points
would be divided into four portions, each containing five data points. The box
encloses the middle 50% of the data. The text states that the whiskers "represent
adjacent values". Our understanding is that the whiskers either enclose the entire
data range, i.e. the upper whisker shows the range of values for the higher 25% of the
data and the lower whisker encloses the range of values for the lower 25% of the
data. Alternately, in modified Box plots, the upper whisker is drawn to the highest
point within 1.5 times the interquartile range and the lower whisker is drawn to the
lowest point within 1.5 times the interquartile range. The interquartile range is the
difference between the 75th and 25th percentile values. Values outside the range
bounded by the whiskers are plotted individually. None of this has been bought out
in the terse description.
1.4.50 Response to CAAP Presentation Comment Number 4 - There are a number of
common variants of box-plots or box-and-whisker plots. Although short, the description
on page 4-137 is sufficient to explain the information presented in the plots. We agree
that the statement "The horizontal lines above and below each box represent adjacent
values," could easily be misunderstood by many readers and requires clarification. We
have added the following definition "The upper adjacent value is defined as the largest
data point less than or equal to the 75th percentile plus 1.5 times the interquartile range;
the lower adjacent value is defined as the smallest data point greater than or equal to the
25th percentile minus 1.5 times the interquartile range."
19
-------
1.4.51 CAAP Presentation Comment Number 5 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
20. Page 4-137. paragraph 3: The paragraph lists the statistical analyses performed.
However, the box plots shown are for only a few of the analyses listed. Where are
the remaining plots?
1.4.52 Response to CAAP Presentation Comment Number 5 - Box plots and scatter
diagrams were not developed for small data sets because these statistical visualization
tools provide little assistance in understanding the potential relationships that may exist
between and among the individual data points. All of the box plots and scatter diagrams
that were developed to facilitate understanding the distribution and relationships of the
data are presented.
1.4.53 CAAP Presentation Comment Number 6 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
21. Page 4-137. paragraph 5: This paragraph refers to Appendices B & C. Where are
they?
1.4.54 Response to CAAP Presentation Comment Number 6 - There were only two
Appendices (A and B) that were developed for the 1997 draft of this background report.
These Appendices were inadvertently left out of the revised draft provided to the
commentors. The indication that there are three Appendices is a typographical error.
Appendix A is the statistical analysis of batch mix driers and Appendix B is the statistical
analysis of the drum mix driers. A copy of these two appendices were e-mailed to all of
the commentors on October 3 and was included in the final report.
1.4.55 CAAP Presentation Comment Number 7 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
22. Page 4-138. paragraph 2: This paragraph refers to Appendix A. Where is it?
1.4.56 Response to CAAP Presentation Comment Number 7 - As stated above Appendix
A was inadvertently left out of the revised draft.
1.4.57 CAAP Presentation Comment Number 8 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
23. Page 4-139. paragraph 1: This paragraph refers to Figure 4.2. Note that Figure 4.7
has an identical caption. The two figures should be labeled properly. One has to
read the text very carefully to realize that Figure 4.2 refers to batch plants and 4.7
refers to drum plants.
20
-------
1.4.58 Response to CAAP Presentation Comment Number 8 - The captions for Figures
4-1 through 4-6 were revised to include the term "batch plants" and the captions for
Figures 4-7 through 4-11 were revised to include the term "drum mix plants." This
should clarify which processes are presented in each set of figures.
1.4.59 CAAP Presentation Comment Number 9 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
26. Page 4-144. paragraph 2: The first sentence is incomplete.
1.4.60 Response to CAAP Presentation Comment Number 8 - We revised the first
sentence to read , "The linear model analysis indicated that neither of the continuous
variables modeled (RAP content, production rate) had a significant effect on filterable
PM emissions. Furthermore, analysis of the scrubber data indicated that the effect of
scrubber pressure drop on filterable PM emissions also was negligible."
1.4.61 CAAP Presentation Comment Number 10 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
29. Page 4-147. paragraph 4: This is the beginning of the review of the more recent data
on load-out emissions. As noted earlier, there is a change in format of the report at
this point in that the discussion of data sets is split into three separate parts.
1.4.62 Response to CAAP Presentation Comment Number 10 - Separate discussions of
stack and fugitive emission sources were intended to keep the reference numbers the
same as previous AP-42 sections and present unrelated emissions processes separately.
However, approximately thirty additional emission test reports on HMA plants in
Wisconsin were submitted for evaluation. These reports include tests supporting industry
comments on the emission factors for SO2 and benzene for stack emissions from drum
mix plants. As a result, the summaries and evaluations of test reports for the stack
emissions and fugitive production emissions were aggregated and presented in the
approximate order that the references were reviewed and summarized for emission factor
development.
1.4.63 CAAP Presentation Comment Number 11 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
34. Page 4-151. paragraph 4: The last line is ambiguous. Silo emissions were measured
only at Plant C. They can't be "combined" with any other measurements.
1.4.64 Response to CAAP Presentation Comment Number 11 - The last sentence was
revised to read "The following sections will describe the basis for performing these
adjustments to arrive at load-out and silo filling emissions at a standardized temperature
and asphalt volatility. This will allow the two load-out data sets to be compared and,
where appropriate, combined."
21
-------
1.4.65 Industry Presentation Comment Number 22 - (Norman Ostroff) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
On page 4-153 you present the following equation:
EFcor = ((Cprod) - (Cback))*(EFprod/Cprod)
If this equation were written in "algebra" rather than "computerese," it might look like:
\\
ack )
This small change would make the equation more "readable" to me and perhaps others. I
understand the physics of the equation, but recommend that the definition of the
parameter Cprod be repeated here; it is not specifically stated. I assume that this equation
is dimensionally correct, but recommend you include the units for the respective terms.
In the numerical example that follows (in the text), I recommend that you indicate to the
reader the exact source of the input data. I could not find the reference to Run 1 MCEM,
but could have missed it while reading the report. Also, demonstrating the calculation of
Cprod from raw data would have been helpful to me.
1.4.66 Response to Industry Presentation Comment Number 22 - The equation was
placed in an equation format to make it more readable. In addition, the text that follows
was revised to include units for the various terms and the table containing the source of
the input data will included. The text will read as follows:
M
Ep = \(r- \_ (n \\ „ EFProd
Where:
EFcor = Background corrected emission factor (Ib/ton).
= Capture efficiency corrected production concentration (gr/dscf).
= Measured background concentration (gr/dscf).
EFprod = Capture efficiency corrected emission factor (Ib/ton).
The following values were obtained from Table 4-22. REPORTED
PARTICULATE-BASED LOAD-OUT EMISSIONS - PLANT C.
Cprod = 1 -68x10 "3 gr/dscf (from MCEM row, second column).
cback = 3.78x10 '4 gr/dscf (from MCEM row, eighth column).
EFprod = 3.12x10 '4 Ib/ton (from MCEM row, third column).
EFcor = ((1.68x10 '3 ) - (3.78x10 '4 )) * (3.12x10 '4 / 1.68x10 '3 )
= 1.30x10-3* 1.86x10-1
= 2.42x10-4 "
22
-------
1.4.67 Industry Presentation Comment Number 23 - (Norman Ostroff) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
On the bottom of page 4-154 and the top of page 4-155, you present a short discussion of
vapor pressure. I agree that the Clausius-Clapyron (CC) equation generates a linear
relationship between (the logarithm of the) vapor pressure and (the reciprocal of the
absolute) temperature. The Antoine equation is a refinement of the CC equation in which
a third constant has been added. The CC equation is a special case of the Antoine
equation in which the third constant is 273.16. Please double check your references, the
(CRC) Handbook of Chemistry and Physics provides tabulated numerical vapor pressure
(verses temperature) data for a large number of compounds and Lange's handbook
provides Antoine equation constants.
1.4.68 Response to Industry Presentation Comment Number 23 - The commentor is
correct that the Clausius-Clapyron equation uses two constant terms to predict vapor
pressure and that Antoine's equation adds a third constant for improved accuracy under
some conditions. The TANKS program can accept vapor pressure information in three
ways. Two of the ways use vapor pressure curve equations. One of these ways uses the
three terms as described by the commentor, this is described as "Constants for Antoine's
equation (using C)". The other way uses two terms and is described as "Constants for
Antoine's equation (using K)". In the TANKS documentation, Antoine's Equation (using
K) has the form
Log P = (-0.05223 A) / T) + B
where:
log (P) is the logarithm (base 10) of the vapor pressure (P).
P = vapor pressure in mmHg.
T = temperature for vapor pressure determination in °K.
The TANKS documentation describes the two parameters A and B as the constants in
Antoine's equation. The constants A and B for the aliphatic hydrocarbons used to
empirically develop the "Antoine's (using K) constants" were obtained from the
Handbook of Chemistry and Physics; 45th Edition; CRC Press; June 1973 from the table
titled Vapor Pressures, Critical Temperatures and Critical Pressures of Organic
Compounds. The two term form of Antoine's equation as stated in the TANKS software
program was used in the draft report since interpolation between constants for two
compounds is simpler than when using the three constants version of Antoine's equation.
Based on the commentors statements we considered deriving the more proper Antoine's
constants using three terms. However, the highest molecular weight aliphatic
hydrocarbon in Lange's handbook is eicosane. Since eicosane is more volatile than the
compounds best representing asphalt, the more proper Antoine's constants are not
available to predict vapor pressure. In addition, the improved accuracy is not required.
While the proper terminology would refer to the Clausius-Clapyron equation, the
23
-------
terminology used in the text is consistent with the terminology in the TANKS program.
Therefore, to avoid confusing some readers, the terminology will not be revised.
1.4.69 Industry Presentation Comment Number 24 - (Norman Ostroff) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
In the second line of the next paragraph, I recommend that you remove the words "The
industry indicates that higher temperatures are avoided. . ."
1.4.70 Response to Industry Presentation Comment Number 24 - The text was removed.
1.4.71 Industry Presentation Comment Number 25 - (Norman Ostroff) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
On the bottom of page 4-155, you cite equation numbers 4-1 1 and 4-12. Where are
equations 4-1 through 4-10? Again, these equations, as presented, are somewhat difficult
to follow. The equations can be simplified and perhaps the following sentence added:
"Equation 4-11 can be rearrange to give:
z2l0-0231(325-T)
My Equation (4.13) is mathematically identical to your Equation (4. 1 1). In addition, I
have found (in my own writing) that the use of the dotted line, connecting the equation
and the equation number, simplifies reading. Am I correct in my understanding that EFstd
is the emission factor to be used for prediction of future emissions? If so, then I believe
this should be clearly stated.
1.4.72 Response to Industry Presentation Comment Number 25 - Equation 4-1 is on page
4-139 and Equation 4-10 is on page 4-145. It is recognized that the equations are
somewhat difficult to follow since the standard condition values remain in the equation.
These equations were not simplified to demonstrate the methodology EPA used to adjust
the measured data to the standard condition. In order to use the report, the user is not
required to use either of these equations. The equations that are to be used to predict
future emissions are developed later in the background report and are presented on pages
4-160 through 4-162. The equations to be used for prediction of future emissions are
reduced to the simple form suggested by the commentor.
1.4.73 Industry Presentation Comment Number 26 - (Norman Ostroff) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
Fifteen figures are referred to at various points. In their current location, reader must stop
reading the document and look for the figures then return to the text. This is distracting,
interrupts the smooth flow of information, and can be avoided if the figures were placed
directly into the text (perhaps reduced in size).
24
-------
1.4.74 Response to Industry Presentation Comment Number 26 - We agree that moving
the figures so that they are closer to the text where they are referenced will help the reader
locate them easier. However, we also believe that a significant size reduction will make
some of details too difficult to see. Most of the figures are associated with the statistical
analysis that was performed. We explored moving the figures for batch mix dryers to
follow Section 4.3.1 and the figures for drum mix dryers to follow Section 4.3.2. This
arrangement would also be distracting to some and interupt the smooth flow of
information in the text. As a result, the location of the figures was not changed.
1.4.75 CAAP Presentation Comment Number 12 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
54. Table 4-15: When will this table be completed?
1.4.76 Response to CAAP Presentation Comment Number 12 - There is no table to be
inserted at this location. The final document will have no blank tables and the table
numbering was revised so that all tables are numbered consecutively.
1.4.77 CAAP Presentation Comment Number 13 - (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
55. Table 4-19: Footnote a is incorrect. R2 is the correlation coefficient, not the "squared
correlation coefficient." We believe the terminology calls R the coefficient of
determination, so that R2 could be called the "squared coefficient of determination".
The more common usage is to stick with "correlation coefficient" or R2.
1.4.78 Response to CAAP Presentation Comment Number 13 - The footnote is correct.
The term R is the correlation coefficient. R2 is called the coefficient of determination or
the correlation index.
1.5 Reference Citations
1.5.1 CAAP Reference Citations Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
3. Page 2. paragraph 3: This paragraph needs a cross reference, even if it is to the
relevant sections in Appendix B. In general, we need page number references to
information in Appendix B which is over 400 pages long.
1.5.2 Response to CAAP Reference Citations Comment Number 1 - To help locate the
primary references for the key information presented in the main report, the reader is
referred to specific tables or page numbers in the Appendices. However, to produce a
report that is uncluttered with voluminous citations that would hinder the readability of
the report, all of the primary references will not be used in the main body of the
Emissions Assessment Report. Instead, a sentence was added to inform the reader that a
comprehensive listing of references is contained in the Appendices. In addition, a table
25
-------
has been added to the main body of the Emissions Assessment Report that provides a
cross reference directing the reader to the specific section or table in the appendices
where the more important information is located.
1.5.3 CAAP Reference Citations Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
16. Page 15. paragraph 3: The first line refers to "emission tests described in previous
paragraphs". Since the previous paragraphs refer to more than one plant test, it is
preferable to have a proper reference; for example, the text can refer to the test at
Plant C or Plant D.
1.5.4 Response to CAAP Reference Citations Comment Number 2 - Since there are a
limited number of emission tests which were used to establish emission factors for hot oil
heaters, load out, silo filling, asphalt storage tanks and yard emissions, the specific
reference numbers used in Appendix B were cited to describe the test(s) used to establish
these emission factors.
1.5.5 CAAP Reference Citations Comment Number 3 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
16. Page 11. paragraph 2: What exactly is reference 1? Is it the main report with a
different date and number? Also, the derivation of the factor is properly described,
not in Reference 1 but in Appendix B.
1.5.6 Response to CAAP Reference Citations Comment Number 3 - Reference 1 for this
part of the report should be Appendix B or the Emission Factor Documentation for AP-
42 Section 11.1, Hot Mix Asphalt Production. This reference citation was revised.
1.5.7 CAAP Reference Citations Comment Number 4 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
17. Table 1: There are many problems here. Footnotes k through p refer to Reference 1.
That is incorrect. If you want to refer to Appendix B, provide specific page number
and citation.
18. Table 3: Same problems as Table 1; i.e. use of Referencel which does not really
contain the primary data and uncontrolled emissions equal controlled emissions.
19. Table 5: Same problems as Table 1; i.e. use of Referencel which does not really
contain the primary data.
23. Table 14: Same problems as Table 1; i.e. use of Referencel which does not really
contain the primary data.
1.5.8 Response to CAAP Reference Citations Comment Number 4 - Reference 1 for this
part of the report should be Appendix B or the Emission Factor Documentation for
AP-42 Section 11.1, Hot Mix Asphalt Production. This reference citation was revised.
26
-------
We decided that citing each of the over 150 references used to derive these emission
factors would add text that would confuse the typical user. Additionally, citing over 150
references would potentially hide information that adds value. As a result, we decided
that we would reference only the background report. We agree that presenting a specific
location in the footnote will assist a user find the information needed to better understand
the basis for the emission factor. To provide this assistance, the table number in the
background report was specified.
1.5.9 CAAP Reference Citations Comment Number 5 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
20. Table 6: Unlike previous tables, descriptive statistics about the data sets are not
presented.
22. Table 8: Unlike previous tables, descriptive statistics about the data sets are not
presented.
1.5.10 Response to CAAP Reference Citations Comment Number 5 - Generally, we
present descriptive statistics only when more than 20 supporting data are available.
1.5.11 CAAP Reference Citations Comment Number 6 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
23. Table 14: This and many other tables, especially in appendix B, have footnote "a"
obscured because it is superimposed over the last letter of the table heading
1.5.12 Response to CAAP Reference Citations Comment Number 6 - The print over of
footnote "a" is not evident in the electronic file or page printed from the file. It is
possible that the Adobe Acrobat software used by the commentor is not the version
provided on the CD-ROM that the commentor is using. There also may be a problem
with the printer or printer driver used by the commentor. However, all of the footnotes in
the tables were evaluated for the situation the commentor describes.
1.5.13 CAAP Reference Citations Comment Number 7 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
2. Page 2-1, Section title: Reference 1 is to an earlier version of AP-42. Why is an
obsolete version of the same document being used as the reference? Reference 3 is a
personal communication from the head of an industry lobbying group. Is this
information in the public domain? Can we get a copy?
1.5.14 Response to CAAP Reference Citations Comment Number 7 - Reference 1 was
revised to indicate the original source of the information characterizing the industry. All
of the information cited in this report is in the public domain. A copy of Reference 3 was
provided to the commentor.
27
-------
2.0 Industry Characterization and Structure
2.1 Capacity and Production Levels
2.1.1 CAAP Production Level Comment Number 1 - (Ravi Nadkarni & Lloyd Pillion) -
Commentors' General Comments. Section B
The report and its appendices contain little or no information about industry structure.
For example, Appendix B asserts that drum plants are 90% larger than batch plants in
terms of annual throughput. This really falls out of an assumption made by the authors.
There is little or no backup for this conclusion. Similarly, there is no information on what
proportion of plants use natural gas and what proportion use other fuels.
Commentors' Summary Section C
Undocumented Information on Industry Structure: The report and its appendices contain
little or no documented information about industry structure. Yet such information is used
to derive average plant sizes and share of annual production between batch and drum
plants.
Commentors' Summary Section C
Industry Comments on Capacity Are Given More Weight than Data Collected by EPA or
State Agencies: The report states that the hourly capacity data was averaged from the
capacities of precisely 98 batch plants and 162 drum plants. Since most of the data was
gathered during performance tests at full capacity and witnessed by State Agencies, it is
unlikely that hourly production rate data is incorrect. It is interesting that "conversations
with industry" are all that is needed to supercede this data. In the report, it is clear that
the authors believe that batch plants generally operate fewer hours per year than drum
plants. Reasons for this (not necessarily explained in the report) include the lack of silos
for HMA storage, small customers using smaller trucks, and so on. The report keeps
talking about hourly production rates when the discussion should focus on annual
production, i.e. hourly rate times hours operated per year.
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
1. Page 1. paragraph 3: In this paragraph, 1996 production statistics are presented along
with an estimate of the number of plants in the country. There are no references to
the source of these data. In Appendix B, page 2-1, the same paragraph is repeated.
On page 2-2 and 2-3 of Appendix B, calculations are presented which derive average
annual production rates for batch and drum plants. We will comment on the
problems with these calculations in our page-by-page discussion of Appendix B.
However, the entire discussion is aimed at showing that drum plants produce more
HMA compared to batch plants without actual data.
5. Page 8. paragraph 4: More industry statistics are presented without a source or a
measure of precision or accuracy.
28
-------
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
1. Page 1. paragraph 3: This paragraph is repeated in the main volume, and in the first
two appendices. It would be more useful to discuss the source of this data and
comment on its reliability, rather than repeat it in three places.
Comments on Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot
Mix Asphalt Production, June 2000 Draft
3. Page 2-1. paragraph 3: The key information in this paragraph about the number of
plants in the country and the type of plants being built is not referenced so the reader
can't judge the reliability of the information.
4. Page 2-2. paragraph 1: The key information in this paragraph about 1996 production
is not referenced so the reader can't judge the reliability of the information. Further,
since road building tends to be a cyclical activity (at least locally), production or
output over several years might be a more reliable way to establish these levels than
the output in a single year. Dept of Commerce or USGS/Mineral Industry Surveys
probably collect and publish this information.
5. Page 2-2. paragraph 2 and 3: If the hourly capacity data was averaged from the
capacities of precisely 98 batch plants and 162 drum plants, the numbers are
probably correct, unless these plants consistently misrepresented the capacity. Such
misrepresentation is not likely to go unnoticed since most of the data was gathered
during performance tests at full capacity in the presence of State agencies to make
sure that emissions limits were being met. It is interesting that "conversations with
industry" are all that is needed to supercede this data. Why are we not surprised!
Overall, this paragraph is attempting to relate hourly capacity to annual production.
That goal is not achieved. The discussion in paragraph 3 is totally confused and
needs to be rewritten. Hourly production rates are easy to measure. If the data set
referred to in the previous paragraph is for obtaining operating permits from the
relevant State Air Pollution Control Agencies, as is the case in Massachusetts, the
plant is supposed to be running at or close to the design hourly capacity when the
emissions are sampled. Therefore, We believe that the hourly capacity
measurements are correct because these numbers are measured during State
compliance testing and should not be superceded by "industry comments".
The discussion should be talking about the fact that batch plants, especially batch
plants without silos for HMA storage, don't operate for as many hours a year as
drum plants because they service customers using small trucks/small volume
customers and because of the difficulties in matching hot mix production with truck
arrivals. (Note that because batch plants store dried, hot stone, there is storage
available, but the stored quantity is generally much smaller than that available with
HMA storage silos.) In contrast, because of their large production contracts, drum
plants operate for more hours each year and the ability to store HMA in silos also
facilitates long production runs. The entire paragraph keeps talking about hourly
29
-------
production rates when the discussion is really trying to focus on annual production,
i.e. hourly rate times hours operated per year.
Also note that the key assumption is that drum plants produce 90% more HMA than
batch plants ON AN ANNUAL BASIS. This assumption is not backed up with
anything; no footnotes or anything. If this is EPA's assumption, say so.
Interestingly, in the main report, this assumption transforms into a finding that drum
plants produce 100% more than batch plants on an annual basis.
If one assumes that the empirical 80/20 rule applies, the average annual output from
batch plants would be 108,000 tons per year and that from drum plants would be
194,000 tons per year. This means that the purported annual industry output of 500
million tons is shared almost equally between batch and drum plants.
It should also be noted that there is no information provided on distribution between
small jobs and large jobs, which would also provide information on how the total
production is split between batch and drum plants. Is it possible that this is because
industry information was provided by a lobbying group supported mainly by the
large producers? This means that the production information has a bias favoring
large producers and making them appear more important.
2.1.2 Response to CAAP Production Level Comment Number 1 - Information presented
in the Emissions Assessment Report (and the AP-42 Section and Background Report)
includes references to the original source of information when it is relevant to the purpose
of the document. The primary purpose of the document is not intended to provide a
detailed information on the industry structure. The information considered most relevant
to the purpose of the Emissions Assessment Report is the development of methodologies
to estimate air emissions from available production and air pollution control device
information. National production statistics, the number and type of production facilities
and the average annual production at the different types of facilities are provided for the
user to have a basic understanding of the source category when conducting an inventory
of air emissions on a regional or national scale. Significant differences in the production
level, the total number of plants and the number of plants by production method may
exist between States and even within regions of States. Local emission inventories or
individual facility permits should use locally developed information on the number, type,
capacity and demonstrated or expected actual production levels to derive improved
estimates of emissions.
Although it is believed that the national trade association for the HMA industry has a
good assessment of the national production levels, EPA performed an independent
assessment of HMA production. As suggested by the commentor, reviewed information
in the USGS Mineral Yearbook1 for production of aggregate by identified end uses. It
1 Available by internet at URL http://minerals.usgs.gov/minerals/pubs/commodity/myb/
30
-------
was determined that 1996 has the greatest reported use of bituminous aggregate.
Information useful for estimating HMA production is divided between reports on
Crushed Stone and Construction Sand and Gravel. Within these two categories the
information is further divided.
For Crushed Stone, the following information is presented in Table 13:
Coarse aggregate, graded:
Bituminous aggregate, coarse 88,900,000 Mt
Bituminous surface-treatment aggregate 22,900,000 Mt
Fine aggregate (-3/8 inch):
Stone sand, bituminous mix or seal 25,500,000 Mt
Special:
Asphalt fillers or extenders 1,280,000 Mt
Unspecified: 4/
Actual 370,000,000 Mt
Estimated 182,000,000 Mt
Total 1,330,000,000 Mt
4/ Includes production reported without a breakdown by end use and estimates for
nonrespondents.
Figure 1. Excerpt from Table 13 of the 1996 USGS Minerals Yearbook.
In addition, Table 20 presents a Total of 1,350,000 Mt of Recycled asphalt for 1996.
For Construction Sand and Gravel, the following information is presented in Table 6:
Asphaltic concrete aggregates and
other bituminous mixtures 70,800,000 Mt
Unspecified: II
Actual 174,000,000 Mt
Estimated 203,000,000 Mt
Total 914,000,000 Mt
II Includes production reported without a breakdown by end use and estimates for
nonrespondents.
Figure 2. Excerpt from Table 6 of the USGS Minerals Yearbook.
In addition, Table 14 presents a Total of 3,740,000 Mt of Recycled asphalt for 1996.
Both chapters in the Minerals Yearbook recommend that the unspecified uses' categories
be distributed as the specified uses categories. Adjusting the Total Crushed Stone
production of 1,330,000,000 Mt to the ratio of reported total specified use totals for HMA
usage verses the total specified uses of Crushed Stone yields 236,904,000 Mt
(1,330,000,000 X (138,580,000 - 778,000,000). Adjusting the Total Sand and Gravel
31
-------
production of 914,000,000 Mt to the ratio of reported total specified use total for HMA
usage verses the total specified uses for Sand and Gravel yields 120,505,000 Mt
(914,000,000 X (70,800,000 + 537,000,000). Asphalt is added to this newly quarried
aggregate total of 357,409,000 Mt to produce HMA that is 6% asphalt (or 94% aggregate)
to yield 380,222,000 Mt of HMA. In addition to newly quarried aggregate, the USGS
reported that a total of 5,090,000 MT of recycled asphalt was used in HMA. Based on
these reported values from USGS, the production of HMA for 1996 is estimated to be
385,312,000 Mt or 424,614,000 tons. However, given the emphasis on recycling in State
paving contracts, a recycled asphalt to new HMA ratio of only 1.0% appears very low and
is probably significantly under estimated by USGS.
In a 1993 joint Federal Highway Administration and EPA report (A Study of the Use of
Recycled Paving Material - Report to Congress; FWHA-RD-93-147; EPA 600/R-93-095;
June 1993), it was estimated that 73 million metric tons (80.4 million tons) of RAP were
recycled annually. This report documents several methods of reprocessing RAP for reuse
as pavement or other materials. However, the report does not provide estimates of
reprocessing by each method. Based on this report, EPA concluded that the majority of
RAP reprocessing is in HMA plants. Assuming all of the RAP is reprocessed in HMA
plants, an early 1990's upper bound estimate of 16.1% recycled asphalt produced can be
calculated [73,000,000 - (380,222,000 Mt + 73,000,000 Mt) xlOO% = 16.1%].
Recongizing this estimate includes reprocessing not in HMA plants, EPA examined an
alternative method of estimating national RAP usage in HMA plants.
At Plants C and D between 80 and 90% of the asphalt produced included RAP. When
RAP was used, Plants C and D included 30 percent and 10 percent RAP in their
respective final asphalt mixes during EPA emission testing. Extrapolating this
production information to an annual estimate, Plants C and D use between 8 and 27%
RAP with a midpoint of 17.5 percent. While this number is larger than the FHA derived
upper bound estimate of 16.1%, it may be more representative of RAP usage rates in the
late 1990's. Using the midpoint (17.5%) of this range yields a revised RAP usage of
80,653,000 Mg (88,905,000 tons). Using the revised RAP usage yields a total estimated
HMA production for 1996 of 469,102,000 Mg or 516,096,000 tons.
Since the industry has an economic interest in understanding the annual production of
their industry, it is reasonable that this independently developed value is close to the
value provided by the industry and used in the draft report. EPA's review of the USGS,
FHA and EPA studies indicates the national RAP usage rate is likely between 15-20%.
Accordingly, EPA used the above information and the mid-point of the estimated RAP
usage range (17.5%) to estimate annual HMA production of 516,096,000 tons for the
United States.
The commentor is correct that the paragraph including the explanation and comparison of
estimated average drum mix plant annual production and average batch mix plant is
confusing and is revised. However, based on statements in the fifth comment on
Appendix B, it appears that the commentor understands the derivation and rationale in the
32
-------
comparison of drum mix and batch mix asphalt plants; the development of annual
production levels for typical batch mix and drum mix plants. In addition to presenting the
development of the estimated annual production of HMA, we have revised Section 2.1 of
the background report to read as follows:
"Information provided by the HMA industry indicates that HMA is produced by
approximately 2,300 batch plants and 1,300 drum-mix plants. Using a national asphalt
production estimate of 517 million tons of HMA, an estimate of the national annual
production capacity at drum and batch mix plants was determined as follows:
Based on available production capacity data from emission compliance tests of 98 batch
mix plants and 162 drum-mix plants, the average maximum production rates are:
Batch-214 tons/hr
Drum - 272 tons/hr
Extrapolating these averages to the entire HMA industry yields an estimated, theoretical
national production capacity of 7,409 million tons of HMA if all plants could operate
8760 hours in a year. The 2,300 batch mix plants would produce 4,311.7 million tons
and the 1,300 drum mix plants would produce 3,097.5 million tons.
Based upon the above estimates of HMA production and available plant capacity, the
estimated utilization rate of the industry is only about 7 percent (517 million + 7,409
million). This significant under utilization is caused by limitations on when pavements
can be laid that are created by weather conditions, contract specifications on times of the
day when pavement construction can be performed, the local demands for paving
construction and repair, the distance that HMA can be trucked to a paving site, the desire
to be able to meet short term peak production demands and a variety of other factors. A
number of differences between drum and batch mix asphalt plants suggest the estimated 7
percent utilization ratio is not likely to be equal distributed among batch mix and drum
mix plants. These differences include:
1) production methods,
2) capability to make and store product ahead of the time,
3) ability for loading to occur significantly quicker at facilities with storage silos,
4) the general lack of storage capability at batch mix plants, and
5) the economic desire to shift higher production demands to the higher capacity and
more cost efficient drum mix plants.
All drum mix plant require HMA storage silos to store product ahead of demand. It is
estimated that for a maximum production day, a typical drum mix plants will begin
production three hours prior to the first truck load-out in order to stay ahead of demand.
For days with less than maximum production, a typical drum mix plants will maintain
this relative production advantage, although it is not necessary to begin production three
hours ahead of the first truck load-out. Since a typical batch mix plant does not have
storage for a significant amount of aggregate, it cannot produce significantly ahead of
demand. For an eight-hour load-out schedule and equal production capacity, the drum
33
-------
mix plant would be able to produce 38% more product (11-^8= 1.38). It is also
estimated that for about 30 minutes over the eight hour day, a typical batch mix plant will
need to stop production because there are no transport trucks to load. As a result, the
batch mix plant will only be able to produce 94 percent of its hourly target production
(7.5 + 8 = 0.938). As a result the eight-hour load-out capability for drum mix plants
would be 147% of the eight-hour production capacity for batch mix plants (1.38 H- 0.938
= 1.47). Since the average production capacity of drum mix plants is 27% greater than
batch mix plants (272 H- 214 = 1.27), the overall eight-hour load-out advantage of the
average drum mix plant (as compared to the average batch mix plant) would be 187
percent of the eight-hour production capacity (1.47 x 1.27 = 1.87). Using this estimate, if
517 million tons of HMA are produced per year by the 2,300 batch and 1,300 drum-mix
plants then:
517 x 106 = 2,300 x B + 1,300 x D
and
D = 1.87B
where:
B = average production of a batch mix plant (tons/yr)
and
D = average production of a drum-mix plant (tons/yr)
Solving the equations for B:
517xl06 = 2,300xB+ 1,300(1.87x8)
517xl06 = 4,731xB
B = 109,000 tons/yr
D = 204,000 tons/yr
Using these average production rates, the total 1996 HMA production from batch and
drum-mix plants is estimated at about 251 million tons and 265 million tons,
respectively."
At the estimated annual production levels, the average batch plant would have a 5.5%
utilization rate compared to an 8.1% utilization rate for drum mix plants. It is believed
that using these estimates of annual production provides a reasonable differentiation in
the utilization rates of batch mix plants and drum mix plants. Given the conditions stated
above that would tend to influence the use of newer, higher production and more efficient
plants, a 50% difference in utilization appears reasonable. In addition, the information
that was obtained on the capacity and annual production of Plant C and Plant D
(References 355 and 356 in the AP-42 section and background reports) supports these
utilization rates. In the test report for Plant D, the hourly capacity is listed at 255 tons per
hour and the typical annual production is about 100,000 tons per year. In addition to
being very close to the average capacity and estimated annual production cited above, the
34
-------
utilization rate is 4.5%. While this is slightly less than the estimated national average
utilization rate, it is very close to the value estimated. In the test report for Plant C, the
hourly capacity is listed as 650 tons per hour and the annual production is about
1,000,000 tons per year. While this facility has a much higher capacity than the estimate
for the average plant, it is one of the newest plants in the nation and is likely one of the
highest capacity plants in the nation. Based upon plant C production and capacity
information, the utilization rate of this plant is 17.5%. This is significantly higher than
the average utilization estimated above. However, the plant is located in southern
California where the temperature is suitable for paving all year and where there are on
average only 30 rain days a year to limit the laying of pavement. In addition, the facility
was built to provide HMA for the construction of California Highway 241, which is a
new 30 mile toll road west of Irvine, CA. It is unlikely that there are many other
locations where the combination of weather and demand for major road construction
would support the high utilization rate of Plant C. Weather conditions in the majority of
the United States restricts paving operations to about nine months due to low
temperatures. Further, for the eastern half of the United States, the average 100 days of
precipitation further limits the number of paving days. Considering these situations, it is
not unreasonable for plant C to have twice the utilization of the average HMA facility in
the United States.
As indicated in the more detailed response provided later, information on the derivation
of an estimated proportion of gas and fuel oil usage by the industry was presented in the
revised industry description contained in Appendix B. In addition, statements were added
to the "main" report and Appendix A that between 70% and 90% of the energy estimated
to be used at HMA plants in 1998 was from natural gas. Further detail on the
characteristics of the HMA industry such as the distribution between small jobs and large
jobs are outside the scope of this report. This type of information would be highly
specific to the location and type of the facility. It is not known whether this type of
information is available outside the individual companies that own the 3600 plants and
obtaining this information would require significant effort and time.
2.1.3 CAAP Production Level Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot
Mix Asphalt Production, June 2000 Draft
10. Page 3-2. paragraph 1: The sentence states that emission data must be from a primary
reference. In chapter 2, however, secondary references and anecdotal comments are
used to supercede primary reference data on hourly capacity.
2.1.4 Response to CAAP Production Level Comment Number 2 - The information
presented in chapter 2 which is derived from secondary references is not emission data
and does not supercede primary reference data. As indicated in previous response, the
discussion on the derivation of annual production from reported hourly capacity data was
revised to more clearly indicate the basis for the derivation.
35
-------
2.2 Type of Fuels Used by Typical Production Facility
2.2.1 State Fuel Usage Comment Number 1 (Massachusetts State Senator Stephen F.
Lynch)-
In reading this report and its appendices, along with comments sent to my office by Ravi
Nadkarni, it is my understanding that this report only shows emissions results from
asphalt plants burning natural gas as a source of fuel.
The report also fails to provide information about what percentage of asphalt plants cited
in the report used natural gas and what percentage of those plants used alternative fuels.
Although I am not an environmental engineer, it is obvious to see that emissions result
from asphalt plants burning only natural gas, as opposed to other fuels, will show
considerably lower emissions result. This would clearly indicate to me that the results in
this report are skewed. I would like to see more thorough testing of various data included
in the final report, especially if this report is to be considered an emissions standard for all
asphalt plants.
2.2.2 CAAP Fuel Usage Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
18. Page 16. paragraph 3: The second line refers to Tables 4 through 11 which present
more detailed information about "typical" plants. Unfortunately, this is the emission
factor data of Appendix A multiplied by 100,000 for batch plants and by 200,000 for
drum plants. The implication is that the reader can't move the decimal point to the
right or multiply by 2. Further, in keeping with their bias for showing only low
numbers, these tables are for natural gas only and not for oil-fired heaters. Why?
Finally, what is the basis for assuming (third bullet on page 17) that PM emissions
from load-out and silo filling are entirely PM-10? We would have thought that a
better assumption was that these were PM-2.5.
2.2.3 Response to State and CAAP Fuel Usage Comments Number 1 - Presentation of
emissions representative of an average gas fired HMA facility data in the draft Emissions
Assessment report was reasonable. Over 95% of the fuels (by energy content) used in
industrial sources is gas and fuel usage by HMA plants is likely similar. However, in
response to commenters concerns we conducted a limited survey to obtain additional
information on fuel useage and added emission factor tables for fuel oil fired HMA
plants.
A very limited survey of available, but limited, fuel usage information from five State
agencies suggests that approximate 70 - 90% of HMA plants are gas fired. Limitations
on the available data from these five State agencies could not precisely discern fuel usage
at HMA plants. However, we have revised the Emissions Assessment Report by adding
text on page 3 characterizing this range of fuel usage in the industry.
We also recognize that there are local situations where a HMA plant may primarily use
fuel oil and, as a result, may generate more emissions than the typical gas fired HMA
36
-------
facility. We have revised the Emissions Assessment Report to indicate that many HMA
plants have the capability to use fuel oil as an alternative to gas. The report was also
revised to note that fuel oil fired HMA plants would have higher emissions of SO2, NOX
and HAP compounds. In addition, Emissions Assessment Report Tables 1 and 2 were
revised to include a column for fuel oil. The footnotes to these tables were revised to
indicate that natural gas is used to produce between 70 and 90% of the HMA and fuel oil
to produce between 10 and 30% of the HMA. Also, Tables 4 and 7 include columns
presenting emissions for fuel oil. The footnotes to Tables 4 and 7 now indicate that
between 70 and 90% of the HMA produced is with natural gas fuels and between 10 and
30% of the HMA is produced with fuel oil.
We have also documented the basis of the statements added to the main report in
Appendix B. The following paragraph was added to Section 2.1 following the derivation
of the estimates for the average annual production of batch and drum mix plants.
"The Department of Energy indicates that annual distillate fuel sales to industrial
customers in the US for 1998 was 2,462,355,000 gallons2 compared to natural gas sales
of 8,686,147,000,000 cubic feet3. At a typical energy content of 140,000 Btu/gal for
distillate oil and 1050 btu/ft3 for natural gas, 96.3% of the energy used by industries was
natural gas. We expect that many of the factors that promote the preferential use of gas
fuels are common within many industries. Therefore, we expect fuel usage at Hot Mix
Asphalt plants is very similar to other industrial sources. Some of these factors are fuel
cost, delivery system requirements, and equipment maintenance requirements. The
Energy Information Agency reported in the Manufacturing Consumption of Energy 1994
(Combined Consumption and Fuel Switching) report4 that the national average cost
industry paid for a million Btu of energy was $2.15 for natural gas, $4.84 for distillate oil
and $4.71 for LPG. This report also separates these energy costs by four regions of the
United States. Of the four regions, the differences in the costs of the various fuels are
smallest in the northeast region. In this region, the average fuel costs were $3.39 for
natural gas, $4.89 for distillate oil and $5.69 for LPG. In addition, the delivery of fuel oil
and LPG must be scheduled and stored near the production unit. The storage tanks and
supporting mechanical equipment require monitoring and maintenance that is not
required when the plant is fueled with natural gas. Also, burners for firing fuel oil require
a higher level of maintenance than natural gas burners. Lastly, it is recognized that the
combustion of fuel oil produces more air emissions than natural gas combustion.
Therefore, it is believed that, where it is available, natural gas is and will remain the
predominant fuel used at HMA plants. However, many plants will maintain the
2 Available by internet at URL
http://www.eia.doe.gov/pub/oil_gas/petroleum/data_publications/fuel_oil_and_kerosene_sales/current/pdf/tablel.pdf
3 Available by internet at URL
http://www.eia.doe.gov/pub/oil_gas/natural_gas/data_publications/natural_gas_annual/current/pdf/table_014.pdf
4 Available by internet at URL http://www.eia.doe.gov/emeu/mecs/mecs94/consumption/mecs4a.html
37
-------
capability to use fuel oil as an alternate or supplementary fuel. A few plants will use only
fuel oil due to the unavailability or high local cost of natural gas. As a spot check of the
DOE information on industrial fuel usasge, an informal telephone survey of five States
was conducted. The survey confirmed that HMA plants use natural gas when it is
available. The fuel usage information in the States' emission inventories shows a range
of about 50 percent to 99 percent gas usage. However, the fuel usage information
reported by industry was incomplete. Also, for some States, energy usage per ton of
product was over 50 percent higher than emission tests where fuel usage and production
information was available. As a result, it is estimated that between 70 and 90 percent of
HMA is produced with gas fuels."
While the main body of the Emissions Assessment report only presented estimated
emissions for gas fired HMA plants, the appendices present emission factors for not only
natural gas fired facilities but also coal, butane, propane, fuel oil and waste oil fired
facilities. Most industrial facilities can fire both natural gas and fuel oil. Additionally,
emissions performance tests with few exceptions are conducted under worst case
conditions. This includes conducting emission tests while firing fuel oil in plants that
almost always use natural gas. As a result, the use of test data presented in this report to
estimate the percentage of facilities firing different fuels would provide a skewed
estimate of the actual fuel usage by the industry.
The Emissions Assessment Report does not establish a national emission standard. The
introduction to AP-42, EPA advises users that emission factors are not recommended as
source-specific permit limits because emission factors are averages. Since emission
standards are a "not to exceed" value, using the average emission factor without any
engineering analysis would likely cause half the facilities to violate their permit limit.
The background report compiles a significant amount of data that may not otherwise be
available in one document. While not establishing emission standards or rules, this data
can be part of the information used in rule or State Implementation Plan development.
EPA's primary purpose for AP-42 is for air quality planning purposes. This includes
providing State and local agencies with reliable information to assist them to develop
comprehensive emission inventories. The information presented in the report is based
upon emission tests of many facilities. Although all facilities were not tested for all
possible pollutants, the information available to characterize emissions from HMA plants
is more than adequate to characterize emissions from these facilities for air quality
planning purposes.
As the commentor states, the information presented in the tables is the emission factor
data of Appendix A multiplied by 100,000 or 200,000. This information is presented in a
simpler and more concise format to help non-technical readers understand the results.
More technical information is included Append!cies A and B. The multipliers used in the
main body of the Emissions Assessment Report are the approximate, average production
levels for the typical batch mix plant and drum mix plant respectively. We have
38
-------
explained how we determined the average production levels in Section 2.1.2. Response
to C AAP Production Level Comment Number 1.
While it is reasonable to assume that organic emissions from load-out and silo filling are
predominantly PM-2.5, the inorganic emissions are likely to be rock dust. The organic
particulate emissions result from a condensation phenomenon which is expected to
predominantly form particulate that is less than one micrometer in diameter. Therefore,
all emissions from load-out and silo filling are PM-10 as well as PM-2.5. Footnote b in
Table 11.1-14 in Appendix A of the draft report states "Total PM is assumed to be
predominantly PM-2.5 since emissions consist of condensed vapors." While PM-2.5 has
recently been designated by EPA as an additional particulate pollutant of concern for
ambient air quality, the use of PM-10 was selected because it is the particulate pollutant
with which most people are familiar.
2.3 Process Descriptions
2.3.1 CAAP Process Description Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
2. Page 1. paragraph 4: The statement in line 1 is incorrect. The primary emission
source is the dryer, not the mixer, as stated. This is the case because combustion of
fuel for generating heat for drying generates large quantities of criteria pollutants.
This error of confusing dryers and mixers is repeated several times, (for example in
Table 1), and we have not bothered to find all instances of such misuse. The best
way to clarify this issue is to substitute "dryer/mixer" for drum plants where the two
units are connected. In the case of batch plants, the dryer and the mixer (pug mill)
are typically uncoupled through hot storage bins. Emissions from both however are
often captured by the same piece of air pollution control equipment.
2.3.2 Response to CAAP Process Description Comment Number 1 - Whether the dryer or
the mixer is the primary emission source for batch plants depends on the pollutant that is
being considered. As stated by the commentor, emissions from both sources are often
captured by the same air pollution control device. Compared to other combustion
sources, the majority of emissions of particulate, VOC and possibly CO are not due to
fuel combustion. However, it is likely that most of the particulate emissions are
generated in the dryer drum at batch mix plants. At batch mix plants, substantial
particulate emissions may also be generated by the hot screens and the pug mill when the
hot aggregate is transferred to these processes. However, to clarify the issue, the text was
revised to clarify that the source of the batch mix emissions is the dryer, hot bins and
mixer.
2.3.3 CAAP Process Description Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
10. Page 9. paragraph 1: The sentence on line 6 is confusing. Do you mean RAP is
added to the "hot" bins or is it added to the pug mill?
39
-------
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
3. Page 2. paragraph 3: This paragraph repeats the ambiguity in the main report on page
9 paragraph 2. Is RAP added to the weigh hopper or to the hot bins?
Comments on Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot
Mix Asphalt Production. June 2000 Draft
6. Page 2-3. last paragraph: The writing should make clear whether the RAP is added
to the hot storage bins or directly to the mixer/muller.
2.3.4 Response to CAAP Process Description Comment Number 2 - Although a few
batch mix plants may mix RAP with the hot aggregate prior to the combined aggregate
storage in the "hot" bins, it is more typical for the cold RAP to be combined with the hot
aggregate in the pug mill. Some newer batch mix plants have a RAP silo with the "hot"
bins to dry and partially heat the RAP before combining the RAP and the hot aggregate in
the pug mill. The wording in the EA-Report, AP-42 Section and Background Report was
revised to more clearly describe the more prevalent process.
2.3.5 Industry Process Description Comment Number 1 (Gary Fore, Vice President
National Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 8: A. 3rd paragraph - ...(2) continuous mix (mix outside dryer drum) plants, ...
Please add the word dryer when using the "mix outside dryer drum" phrase. This will
allow for both past and present types of plants.
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-1: A. Hot mix asphalt paving materials can be manufactured by: ... (2)
continuous mix (mix outside dryer drum) plants, ... Please add the word dryer when
using the "mix outside dryer drum" phrase. This will allow for both past and present
types of plants.
2.3.6 Response to Industry Process Description Comment Number 1 - The word dryer
was added to the sentence to allow for both past and present types of plants.
2.4 Plant Characteristics
2.4.1 CAAP Plant Characteristics Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
2. Page 2. paragraph 1: The statement that raw aggregate is stockpiled "near" the plant
is ambiguous. Such aggregate is normally stockpiled within the plant site and is
moved to the cold feed bins with a front-end loader.
2.4.2 Response to CAAP Plant Characteristics Comment Number 1 - The term plant in
this statement is intended to designate the kiln, hot screens and pug mill and not the
property site in general. The statement was revised to state that the raw aggregate is
stockpiled near the production unit.
40
-------
2.4.3 CAAP Plant Characteristics Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
4. Page 5, paragraph 2: In counterflow plants, the baghouse is preceded by a cyclone
to reduce the load on the baghouse. This is a trade off issue where it is cheaper to
use a cyclone than to increase bag house size and bag cleaning frequency to achieve
the same result. Since the industry prefers counterflow plants to parallel flow
plants(page 1 paragraph 3), the use of an additional particulate separation device in
the former is obviously not a disadvantage, on an economic or a technical basis.
2.4.4 Response to CAAP Plant Characteristics Comment Number 2 - The decision on
using a cyclone before the baghouse is more of a consideration of the baghouse
manufacturer than the manufacturer of the HMA plant. The size of the baghouse depends
on the gas flow, the bag cleaning method, and the type of fabric used for the bags and not
a function of the inlet dust concentration. The statement only identifies one air pollution
control advantage of a parallel flow plant. Other production tradeoffs may be considered
by the plant manufacturer and operator in the selection of the final design of the plant.
The statement as drafted provides information that is useful to individuals and the report
will not be changed.
2.4.5 CAAP Plant Characteristics Comment Number 3 (Ravi Nadkarni & Lloyd Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
4. Page 5, paragraph 2: If you have data to support the statement that organic emissions
may be greater from this process, present it; otherwise eliminate the statement.
2.4.6 Response to CAAP Plant Characteristics Comment Number 3 - The lack of data to
substantiate the statement is not a reason to eliminate the statement. As with many other
emissions that were statistically evaluated, there is insufficient information about the
relevant process variables, an insufficient number of data and relatively high variation in
the emissions to properly evaluate the potential effect. However, the statement was
revised to indicate that the available data is insufficient to discern any differences in
emissions because the process.
2.4.7 CAAP Plant Characteristics Comment Number 4 (Ravi Nadkarni & Lloyd Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
5. Page 5, paragraph 5: We believe the real issue is that with counterflow plants, the
air from the mixing drum acts as secondary and tertiary air for the combustion
process which eliminates many of the organic compounds through combustion. This
point needs to be brought out. Further, is there data to support the statement that a
counterflow plant can process RAP at ratios up to 50% with little or no observed
effect on emissions? If not, the statement belongs in industry sales literature and not
in an EPA document.
2.4.8 Response to CAAP Plant Characteristics Comment Number 4 - It is unlikely that a
significant portion of the air from the mixing drum is used as combustion air for the
41
-------
burners. Typically, the exit CO2 concentration from HMA stacks is around 4% indicating
that the combustion gases are only about 30% of the stack gas. The mixing of a
significant portion of the air from the mixing drum in the combustion zone of the burner
would quench the combustion and result in increased emissions of CO and uncombusted
hydrocarbons (THC). As a result it is also unlikely that a significant reduction in the
organic emissions results from the combustion process. With respect to the processing of
high percentages of RAP, several emissions tests of facilities processing RAP at ratios of
50% are used in the development of the emission factors. The statistical analysis
performed to evaluate the effect RAP percentage on emissions indicated that only the
condensible organic particulate was affected by the RAP percentage. However, the
squared correlation coefficient was only 0.11 (with a correlation of 1.0 being perfect)
indicating that the RAP percentage explained very little of the variability in the data.
2.4.9 CAAP Plant Characteristics Comment Number 5 (Ravi Nadkarni & Lloyd Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
5. Page 5, paragraph 5: Similarly, the comment about "improved thermal efficiencies"
begs the question of what is being compared to what and what is the degree of
improvement. The statement belongs in industry sales literature and not in an EPA
document.
2.4.10 Response to CAAP Plant Characteristics Comment Number 5 - While it is clear
that the report text on thermal efficiency compares currently designed plants to those of
earlier designs, it does not add information needed to characterize the industry with
respect to air emissions and was removed.
2.4.11 CAAP Plant Characteristics Comment Number 6 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
6. Page 7. paragraph 1: The statement that RAP with new aggregate and fresh asphalt
is used to "produce a high quality grade of PDVIA" is another salesmanlike statement.
The point is that the HMA industry produces products to meet a specification. If the
specification requires the use of RAP, it is used; if not, RAP is not used. The point
you want to make is that the use of RAP does not impair the ability to meet the
specifications for certain grades of HMA since the specs might actually require that
RAP is used. Or is EPA implying that the industry used RAP when specifications
call for the use of only virgin materials?
2.4.12 Response to CAAP Plant Characteristics Comment Number 6 - The statement
was revised to state "new asphalt cement is added to produce HMA that meets the
required quality specifications".
42
-------
3.0 Development and Evaluation of Available and Derived Information
3.1 Deficiencies in EPA Method 204
3.1.1 CAAP Capture Efficiency Comment Number 1 - (Ravi Nadkarni & Lloyd Pillion) -
Commentors Summary Section C
Deficiencies in Method 204: Tracers were not used at Plant D because it met the
requirements of Method 204. Unfortunately, Method 204 is inadequate to achieve the
results it tries to achieve. Further, the contention that the estimate of unmeasured
emissions is an "upper bound" is incorrect. At Plant D, a mild ambient breeze was
sufficient to cause some of the fumes to escape from the top or the bottom openings in the
downwind door of the tunnel. Therefore, the size of the openings was decreased further
at the end of the first day of testing. This means that the Method 204 criterion of
maintaining over 200 fpm at a natural draft opening is not adequate. Note that 200 fpm is
less than 2.3 miles per hour, which is not much of a breeze. Therefore it is not surprising
that these criteria are inadequate to assure total containment within a TTE as
contemplated by method 204.
Further, Method 204 has a more serious and fundamental problem. Although the size of
natural draft openings is specified, the method does not contain any criteria to ensure that
the emissions are pulled/sucked past the sampling point in a reasonable period of time. As
is well-known, residence time calculations, based on enclosure volume and fan capacity,
generally provide incorrect information by underestimating the time to evacuate the
enclosure because of channeling. At Plant D, the tunnel was not evacuated in the
approximately 15 second time gap between the final HMA drop and the opening of the
doors. This can be clearly seen in the fact that THC readings did not drop to zero before
the doors were opened. This fact was also observed visually inside the TTE enclosure
since SVOCs are visible. Thus, the report contains compelling evidence that the
collected sample had a low bias, not only because some emissions escaped through the
natural draft openings but also because these emissions did not travel to the sampling
point but were lost when the doors were opened.
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
11. Page 13. paragraph 4: This paragraph is wrong in several ways. The capture
efficiency was measured and calculated using tracers; it was not "estimated" as
stated. To me, the word "estimation" denotes a weaker approach to quantification
than calculations based on actual measurements. Equally, important, the EPA
requirements for total enclosure were shown to be totally inadequate during the
testing at Plant D in Massachusetts. A mild breeze was sufficient to cause visible
emissions from the total enclosure even when a hand-held velometer showed that
Method 204 criteria were being met. Why is this factor, which suggests that Method
204 criteria were inadequate and that the measured emissions have a low bias, not
discussed in this paragraph?
43
-------
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
45. Page 4-163. paragraph 3-5: We further find that the contention that the estimate of
unmeasured emissions being an "upper bound" to be unpersuasive and incorrect.
Our reasons for saying so result from the inadequacies of Method 204. These were
outlined in Ravi Nadkarni's letter of September 15, 1999. They are:
a. At Plant C, two precautions were taken to ensure that we could measure or
estimate the amount of organic fugitive emissions that did not reach the primary
sampling ports. The first was the use of sulfur hexafluoride tracer gas, and the
second was the use of impingement/deposition plates in the ducts which collected
organic particulates which would be deposited upstream of the sample port.
Although all the emissions did not reach the sampling ports (some was deposited
inside the tunnel and some was pumped out of the tunnel by ambient breeze or by
truck movement), the tracer gas enabled an estimation the VOC portion of this
loss. (As was pointed out during the planning, the tracer gas was a good proxy
only for non-condensible emissions but not for condensible emissions.)
At Plant D, a tracer gas was not used although requested by the citizens because
the TTE was constructed to meet Method 204 criteria, yet visual observations and
THC readings showed the following:
b. An ambient breeze was sufficient to cause some of the fume to escape from the
top or the bottom openings in the downwind door of the tunnel. Therefore the
size of the openings was decreased further. This means that the Method 204
criterion of maintaining over 200 fpm at a natural draft opening is not adequate.
Note that 200 fpm is less than 2.3 miles per hour, which is not much of a breeze.
Therefore it is not surprising that these criteria are inadequate to assure total
containment within the TTE. We observed visible emissions of fume from the
entrance, further emphasizing that the Method 204 criteria are inadequate.
c. Further, Method 204 has a more serious and fundamental problem. Although the
size of natural draft openings is specified, the method does not contain any criteria
to ensure that the emissions are pulled/sucked past the sampling point in a
reasonable period of time. As is well-known, residence time calculations, based
on enclosure volume and fan capacity, generally provide incorrect information by
underestimating the time to evacuate the enclosure because of channeling. Note
that Method 204 does not require a specific location for the natural draft openings,
a problem that was pointed out during the planning sessions. At Plant D, even
after the size of the opening was decreased, and the little pieces of colored tape at
the Natural Draft Openings were indicating airflow towards the inside of the TTE,
the tunnel was not evacuated in the approximately 15 second time gap between
the final drop and the opening of the doors. This can be clearly seen in the fact
that THC readings did not drop to zero even when the sample averaging time is
less than 1 minute. This fact was also observed visually inside the TTE enclosure
44
-------
since SVOCs are visible. Thus, in the absence of a tracer gas, we have
compelling evidence that the collected sample had a low bias. For example, the
extended period test results in MRI-D-Table 4-4, show that final concentration did
not drop to zero but stayed between 1 ppm and 2.1 ppm . Yet, this low bias is not
mentioned anywhere in the report nor in Appendix B of the MRI Report which
contains the original data. These emissions did not reach the instrumentation but
were emitted to the environment causing a low bias.
3.1.2 Response to CAAP Capture Efficiency Comment Number 1 - There was no
indication during the testing at Plant D that EPA Method 204 was inadequate for
determining the ability of the enclosure to completely capture the emissions. The
individuals that opened and closed the doors of the enclosure also periodically observed
the tail tale streamers attached above the openings above and below the doors. At all
times when both doors were closed they noticed that the tail tales showed that air was
being drawn into the enclosure. Additionally, while it is stated that the citizen observers
present at the test observed visible emissions leaving the natural draft openings to the
enclosure, EPA was not advised of this until a meeting to discuss EPA's draft responses
to comments on the test reports. Three EPA representatives were present at the test to
address and resolve these type of issues. If EPA were advised of the presumed
observations by the citizens, the cause could be investigated during the emission test. If it
were determined that the complete capture of emissions was being compromised,
corrective action would have been instituted.
The adequacy of EPA Method 204 criteria to assess whether the enclosure attained
essentially complete capture was addressed in the draft of the "Response to Comments on
Testing Program for Asphalt Plants C and D Report" submitted to commentors
previously. Response number 15 "THC Emissions During Extended Period Tests" in the
draft document was divided into two areas. The first area addressed the issue of the
capture efficiency at Plant D. The second area addressed the issue of additional
emissions following the load-out operation. In a meeting held in Boston on January 20,
2000, one point of discussion was EPA's interpretation of Ravi Nadkarni's comments.
Based upon this discussion, EPA understood that the intention of Ravi Nadkarni's
comments was not to question the capture effectiveness of the enclosure at Plant D since
it met EPA Method 204 criteria. As a result, EPA proposed to delete the first area of the
response. In the capture efficiency portion of the draft response, several evaluations were
presented that provided various estimates of the capture efficiency obtained by the
enclosure at Plant D. The text on the following six pages is reproduced from page 21
through 26 of the draft "Response to Comments on Testing Program for Asphalt Plants C
and D Report" submitted to commentors.
45
-------
Capture Efficiency and Post Loadout Emissions
15. THC Emissions During Extended Period Tests
COMMENT: Several commentors (Nadkarni, Pillion, Toxics Action Center, Yatzyshyn,
Lynch) expressed concern over the determination of static emission rates (e.g., trucks sitting in the
yard) based on the Plant D extended period tests. As a related issue, the same commentors were
also concerned that holding the trucks for 15 sec following the final drop was not long enough,
and that some fraction of the emission was missed. Additional clarification on both of these issues
was requested.
RESPONSE: The commentors raise a valid concern that some portion of the emissions
may have been missed due to opening and closing of the TTE doors. This represents a situation
different from that in California, where tracer gases were used to determine capture. As discussed
earlier in Response 11, MRI insured that the TTE design did meet all necessary criteria to assume
100% capture. However, the commentors are correct in that the TTE was intended to allow trucks
to remain stationary until emission levels returned to baseline, which required 3 to 7 min as shown
by the extended period tests. Since this is clearly longer than the 15 sec allowed most of the
trucks, some accounting for the less than 100% capture efficiency was considered. Emissions due
to trucks sitting in the yard are also related to these calculations and are discussed below.
Capture Efficiency
Commentor Nadkarni performs a mathematical analysis of the data set to arrive at a factor
of 1.74, meaning that the measured emissions should be multiplied by 1.74 to arrive at the total
(Total = sum of measured + unmeasured). The basic approach presented by the commentor is one
of several of methods that seem reasonable. However, since the commentors calculations are
21
Figure 3. Excerpt from page 21 of the Draft "Response to Comments on Testing Program for
Asphalt Plants C and D Report."
46
-------
not clearly presented, calculations based upon the commentors approach have been performed as part
of this response, and a different value is presented.
Figures 2 and 3 show load-out data for two truck loadings during the Plant D testing. One of
the loadings was during Test 2 and logged 10-sec data points prior to the extended period test from
1333 to 1338. The second loading was during Run 3 and logged 1-sec data points for an entire
load-out cycle. These two data sets provide the only detailed data clearly showing both the load-out
cycle and the extended tail-off to the baseline. Thus, these two data sets can be used to estimate the
capture efficiency.
Measured emissions are represented by the area under the curve (integrated value) from the
beginning of load-out until the doors are opened (i.e., 15 sec after the final drop). Figures 2 and 3 are
marked with vertical lines at these times to provide clarity. It is believed that the commentor
determined what he called "unmeasured" emissions by integrating from the 15-sec mark until the
emissions return to baseline. The Extended Period column in Table 1 shows the results of this
integration.
Table 1. Estimate of Capture Efficiency Adjustment
Test 2 Data Set
Measured
Unmeasured
Total
Capture Efficiency Adjustment Factor
Test 3 Data Set
Measured
Unmeasured
Total
Capture Efficiency Adjustment Factor
Extended Period
(>5 min)
18.14 ppm-min
7.66 ppm-min
25.8 ppm-min
1.42
29.35 ppm-min
9.75 ppm-min
39.10 ppm-min
1.33
Typical Period
(30 sec)
18.14 ppm-min
1.80 ppm-min
19.94 ppm-min
1.10
29.35 ppm-min
1.67 ppm-min
31.02 ppm-min
1.06
Thus, using the Extended Period (as used by the commentor), the above calculations show that
the estimated Capture Efficiency Adjustment Factor might be 1.33 to 1.42 (or a capture efficiency of
between 70% and 75%), rather than the 1.74 (capture efficiency of 57%) calculated by the commentor.
Furthermore, these calculations are overly conservative (i.e., high) for three reasons:
a. Integration of a greater than 5-min tail-off period is not representative of actual plant
operations during the test. Under normal operating conditions, a truck left the load-out area
no more than 30 sec after the final drop. In fact, as the commentors noted, a high volume of
truck traffic forced MRI to shorten the waiting time to 15 sec during the actual
22
Figure 4. Page 22 from Draft "Response to Comments on Testing Program for Asphalt Plants C
and D Report."
47
-------
c
o
O)
w T~
c -o
, .2
*
ET3
a>
ui -S
T3 £
£ x
3 W
-------
re
a> ^
If
"in ts
§ O
"i o
HI
•D
l>
_o
y
-C
mddj '»ue>3
O/j
a ^
§ .S
1 88
i-2
w «
g*>
83
S H
a ^
0/j
24
Figure 6. Page 24 from Draft "Response to Comments on Testing Program for Asphalt Plants C
and D Report."
49
-------
tests. Thus, the greater than 5 min period integrated above clearly over estimates load out
emissions. The greatest portion of what the commentors refer to as "unmeasured" emissions
would be better termed "yard emissions," which are dealt with later in this section.
b. Almost all of the time periods between truck loadings were 2 min or less, and most were less
than 1 min. Thus, not all of the emission went unmeasured when the doors were opened and
closed around a new truck. In other words, not all of the lingering fumes were dissipated
when a new truck entered the TTE in quick succession. Some of the "unmeasured"
emission calculated above was actually measured as part of the next truck load-out and
therefore was double counted.
c. As discussed in Response 11, both Plant C and Plant D enclosures were very similar except
for the operating doors. The minimum measured capture efficiency for Plant C was 52%
during the background run when only two trucks were available and 61% during production
operations. One would expect that the addition of operable doors that significantly
improved capture during about 80% of the time would improve capture efficiency to 90 to
95% rather than the 70 to 75% as estimated by the commentor.
Evidence of better capture than estimated by the commentor can be seen in the presentations
of 1-min THC readings in Appendix B of the test report. Many times the drop in THC concentrations
between loadings is less than 1 ppm from a value of about 3 to 5 ppm. Sometimes there is no
perceptible drop in the THC concentrations between successive loading operations. The subsequent
loadings begin from one to two minutes after the end of the previous loading. Based upon this
information, one would expect the capture efficiency to be better than 90% {((6min*3ppm) +
(2min*2ppm))/(8min*3ppm) = 0.92}. For loading times longer than 6 min, higher loading
concentrations than 3 ppm, and shorter times between loadings than 2 min, the capture efficiency may
be as high as 98% {((8min*5ppm) + (lmin*4ppm))/(9min*5ppm) = 0.98}.
Additional evidence of better capture efficiency can be made based upon the assumption that
the emissions from one complete drop out of seven is missed. This provides an estimated capture
efficiency of 85%.
Note also that the two integrated data sets were from load-outs of 6 drops (Run 2) and 8 drops
(Run 3). The lower factor (1.33) is from the 8-drop load-out. This makes sense since the measured
emission is greater for 8 drops. Truck-loading operations for Plant D used both drop rates frequently,
so an actual factor would fall somewhere between the two factors calculated.
A better method of estimating capture efficiency would be to select an interval more
representative of actual plant operations than the greater than 5-min period used above. Under normal
operations, a truck would not sit for more than 15 to 20 sec after the final drop before leaving the
load-out area. Additionally, the doors were open an equally short period to allow an empty truck to
replace the truck that was just loaded. Thus, an interval of 30 sec after the final drop can be used as a
representative time period and will give a biased low estimate of the capture efficiency. Data
presented in the table show that this method results in a biased high estimate for unmeasured emission
of 6% to 10%, and that the scaling factor would be from 1.06
25
Figure 7. Page 25 from Draft "Response to Comments on Testing Program for Asphalt Plants C
and D Report."
50
-------
to 1.10 (capture efficiency between 91% and 94%). As discussed in items b and c above,
these estimates are also biased low due to multiple counting of emissions and comparison to
the capture efficiency testing at Plant C.
Although biased low, Capture Efficiency Adjustment Factors in the range of 1.06 to
1.10 appear to be the more representative of actual plant operations and test conditions
encountered than the commentors calculations. However, the analysis that follows for Yard
emissions indicates that other factors appear to overestimate emissions, therefore we did not
apply this adjustment to the data. Inclusion of the long tail-off period (emissions returning to
baseline) is not appropriate since these emissions are more realistically included as yard
emissions and will be described below.
26
Figure 8. Excerpt from Page 25 from Draft "Response to Comments on Testing Program for
Asphalt Plants C and D Report."
51
-------
This earlier, draft response indicates that, if commenters assertions are correct that
opening doors 15 seconds after loadout did release some load out emissions from the
enclosure before capture and measurement, the uncaptured and unmeasured release were
small (less than 10%). As the analysis is an extrapolation of only two data runs and the
enclosure was designed to Method 204 criteria, EPA believed that the actual uncaptured
emissions are most likely smaller than estimated by our analysis. Therefore, no further
adjustments were made to the loadout emissions from Plant D.
Although the capture efficiency for Plat C was based upon measurements and calculations
using a tracer, these measurements still provide an estimate of the capture efficiency of
the ventilation system during the emission testing. First, EPA Method 204 recommends
that sources to use lower 90% confidence level of the measured data as a safety factor for
demonstrating compliance with regulatory requirements. While the calculated lower 90%
confidence level capture efficiency for Plant C averaged 61.3%, the actual measured
capture efficiency averaged 67.9%. Second, as explained in item 60 on page 74 of the
Response to Comments on Testing Program for Asphalt Plants C and D Report, many
factors could contribute to variations in the correlation between what is measured from
the release of process emissions and the tracer gas emissions. Quoting from this report,
the first four items state:
"1. The tracer gas was released outside the edge of the tallest truck bed, whereas
actual emissions occurred generally from the entire bed area but primarily from the
center. Therefore, the tracer gas was not thoroughly mixed with the gas containing
the emissions.
2. The hot asphalt generated a buoyant stream that quickly rose to the ventilation
system. This buoyant stream contained a concentrated mass of the emissions,
whereas the tracer gas was injected toward the edge of this buoyant stream.
Therefore, the gas containing the emissions would be captured better than the air
containing the tracer gas.
3. The asphalt was loaded intermittently with emissions (both visible and measured)
that peaked during the loading, decreased rapidly while the truck waited to be
released, and then tailed off slowly after the truck left. This is significantly different
than the constant continuous release of the tracer gas. Most of the visible emissions
were released during a very short period when the truck was stopped in the tunnel.
The truck being loaded (and the next truck waiting to enter the tunnel) blocked the
movement of air through the tunnel, allowing for more effective capture of emissions
and tracer gas. If a truck were not in the tunnel, wind effects would lower capture
efficiency below the capture efficiency during actual load out. The tracer gas was
also being injected during the time that trucks were moving in the tunnel and when
there was no buoyant effect to help transport the tracer gas to the hood. The piston
effect of truck movement would create air movement across the whole cross-section
of the tunnel to draw the tracer gas away from the capture hood, thereby lowering the
measured capture efficiency. Higher load-out rates (trucks per period) would
52
-------
exacerbate this situation and increase the differences in the capture of the emissions
and tracer gas.
4. Not only were the nozzles that injected the tracer gas at the edge of the gas stream
of the emissions, but they were splayed to inject the tracer gas from 45° below
horizontal to 45 ° above horizontal. For the majority of trucks, about 33% of the
tracer gas was injected down from a point even with, and about two feet from, the
upper edge of the truck bed. As a result, the capture of the tracer gas would be
diminished."
While the design of the system to inject the tracer could not replicate exactly the release
conditions of the emissions, it was as close as could be achieved given the dynamic
nature of the load out process. The inability to replicate the release conditions results in
situations that predominantly underestimate the actual capture efficiency. The magnitude
of this underestimation cannot be determined and thus capture efficiency is considered an
estimate. To avoid confusion created by using the non-qualified term "estimated," the
term "quantitatively estimated " was used in the Emissions Assessment Report.
3.1.3 CAAP Capture Efficiency Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot
Mix Asphalt Production. June 2000 Draft
32. Page 4-150. paragraph 3: There are several fundamental issues here that need to be
explained. First, EPA argued for a long time that the tunnel at Plant C was adequate
to capture all the emissions, based on visual observations by EPA - which were
contradicted later by observations by the citizens. EPA reluctantly agreed to use
tracers only because the tunnel did not meet Method 204 criteria. (In discussion of
Plant D, We will discuss why Method 204 criteria are inadequate). Next, EPA
averaged the efficiency over several runs and used a single capture efficiency
correction. It was only after the citizens showed that capture efficiency varied during
the day, typically declining as the on-shore wind picked up as the land around the
plant heated up, that EPA agreed to correct data for the individual capture
efficiencies. Finally, the so-called correction for truck emissions, discussed later,
continues to be a sham. But that sham is not really discussed here when the quality
of data is being considered. See our comments with reference to page 4-152 below.
Also, note that there is a typographical error in EMD GD-035 " Guidelines for
Determining Capture Efficiency". On page 9, in the formula for p in equation 7, x
value should be 100.8 not 110.8. The calculated value of p is correct, however.
3.1.4 Response to CAAP Capture Efficiency Comment Number 2 - It is not clear what
information contained in the paragraph is incorrect and the commentor wants EPA to
revise. The commentors statements concerning EPA's decision to use a tracer to
characterize capture efficiency are inaccurate. We had previously concluded that, with
doors open, Plant C enclosure did not meet EPA Method 204 criteria. We estimated that
the capture efficiency was between 70% and 90%. EPA's decision to revise the
calculation of capture efficiency was based on our established criteria in the guidance
53
-------
document "Capture Efficiency Testing Guidance" (EMC GD-035). This guideline
specifies that the lower 90% confidence level capture efficiency is to be used rather than
the average measured capture efficiency. The variations in the capture efficiency were
not germane in our decision to revise the calculation of the capture efficiency. The
addition of this information to the current information describing capture efficiency
adjustment is not needed. We agree that the intermediate value presented in equation 7 of
the example calculation on page 10 of the guideline document is a typographical error.
As the commentor notices, the value for p that results from the calculation is correct. In
addition, the text preceding the equation presents the correct calculated value for term
used in the equation.
3.2 Compensation for Asphalt Temperature and Volatility
3.2.1 CAAP Temperature and Volatility Comment Number 1 (Ravi Nadkarni & Lloyd
Pillion) -
Commentors Summary Section C
Incorrect Approach in Compensating for Volatiles Content of Asphalt: The methods used
in these reports to correct for volatiles content of asphalt are contrary to the agreement
that was reached between the EPA and various citizen groups. The citizens had argued
that the asphalt specifications allow the use of asphalt with a much higher weight loss
compared to the asphalt used in Plant C and Plant D tests. The older AC specifications
allow for a 0.5% loss on heating, which can be waived by an engineer on site who can
then allow the use of an asphalt with up to 1% weight loss. The new Superpave
specifications allow for a 1% loss on heating and therefore represent a relaxation of the
previous AC grade standards. In view of this and the industry's use of various additives
and diluents (which would increase the vapor pressure), the citizens argued and the EPA
agreed that maximum number of 0.5% should be used. Note that this was a compromise
since the newer specification allows the use of asphalt with twice that amount of
volatiles. This issue is particularly important since the measurements by State agencies
quoted here appear to be for asphalt prior to blending with additives, in spite of
suggestions to the contrary, in some cases. The EPA, by providing an equation where a
hypothetical RTFOT value can be plugged in, is encouraging misrepresentation by a
proponent of any new asphalt plant, who, once the permit is received, can then switch to
an asphalt with a different RTFOT value with impunity.
As long as the specifications allow for up to 1% volatiles in asphalt (even before
additives are used), it is misleading for EPA to state that their approach in Appendix B -
Table 4-27 "encourages the use of site specific data". The primary use of AP-42 is to
prepare calculations to get permits to build and operate hot mix plants. Therefore, there
is no site specific data that can be used other than fictional data. Unfortunately, the
industry has a long history of providing low-ball estimates, designed to fool the public
and get permits from agencies that are not particularly anxious to ask difficult questions.
Incorrect Substitution of Delivery Temperature for Loadout Temperature: The statement
that "325°F, which is the maximum load-out temperature recommended by industry" is
54
-------
imprecise. Most specifications require this to be the maximum temperature at a job site
where HMA is to be spread and compacted. (The minimum temperature is 275°F.) The
maximum load out temperature is much more variable and can be higher if the plant is
producing asphalt for a distant job or for a small job, where there can be considerable
cooling of the HMA mass between loadout and delivery.
Hot Mix Asphalt Plants Emissions Assessment Report
13. Page 14. paragraph 2: The statement that "325°F, which is the maximum load-out
temperature recommended by industry" is imprecise. Most specifications require
this to be the maximum temperature at a job site where FDVIA is to be spread and
compacted with rollers. (The minimum temperature is 275°F.) The maximum load
out temperature is much more variable and can be higher if the plant is producing
asphalt for a distant job or for a small job, where there can be considerable cooling of
the FDVIA mass between loadout and delivery. This same error is repeated in the last
sentence of the next paragraph and in many portions of Appendix B.
This paragraph also states that -0.5% was selected as a default value for weight loss
in a rolling thin film oven test. The methods used to apply the correction for
volatility are contrary to the agreement that was reached between the EPA and
various citizen groups. The citizens had argued that the asphalt specifications allow
the use of asphalt with a much higher weight loss. The older AC specifications
allow for a 0.5% loss on heating, which can be waived by an engineer on site who
can then allow the use of an asphalt with up to 1% weight loss. The new Superpave
specifications allow for a 1% loss on heating and therefore represent a relaxation of
the previous AC grade standards. In view of this and the industry's use of various
additives and diluents (which would increase the vapor pressure) which are often
added after the asphalt has been sampled for the measurement of the weight loss, the
citizens argued and the EPA agreed that maximum number of 0.5% should be used
at all times. Note that this was a compromise since the newer specification allows
the use of asphalt with twice that amount of volatiles. This issue is particularly
important since the measurements by State agencies quoted here appear to be for
asphalt prior to blending with additives. The EPA, by providing an equation where a
hypothetical RTFOT value can be plugged in, is encouraging misrepresentation by a
proponent of any new asphalt plant, who, once the permit is received, can then
switch to an asphalt with a different RTFOT value with impunity.
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
14. Page 10. paragraph 2. 3: The comments here are a major problem. The methods
used to apply the correction for volatility are contrary to the agreement that was
reached between the EPA and various citizen groups. The citizens had argued that
the asphalt specifications allow the use of asphalt with a much higher weight loss
than that suggested by limited sampling. The older AC specifications allow for a
0.5% loss on heating, which can be waived by an engineer on site who can then
allow the use of an asphalt with up to 1% weight loss. The new Superpave
specifications allow for a 1% loss on heating. In view of this and the industry's use
55
-------
of various additives and diluents which would increase the vapor pressure after
sampling for the measurement of the weight loss, the citizens argued and the EPA
agreed that maximum number of 0.5% should be used at all times. Note that this
was a compromise since the newer specification allows the use of asphalt with twice
that amount of volatiles. This issue is particularly important since the measurements
by State agencies quoted here appear to be for asphalt prior to blending with
additives. The EPA, by providing an equation where a hypothetical RTFOT value
can be plugged in, in encouraging misrepresentation by a proponent of any new
asphalt plant, who, once the permit is received on the basis of a low RTFOT value,
can then switch to an asphalt with a different RTFOT value. This entire paragraph
has to be rewritten with -0.5% weight loss and 325° F or higher temperature.
Comments on Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot
Mix Asphalt Production. June 2000 Draft
39. Page 4-155. paragraph 2: The statement is that "The industry has stated that good
paving practices dictate that load-out temperatures in excess of 325° F should be
avoided." Who has stated this and in what context? The specification for highways
is that hot mix temperature before the pavement is laid down should be between 325
and 275° F. Hot mix outside this temperature range can be rejected. The
temperature at load out depends on the distance between the plant and the location
where the new pavement is being laid and the size of the truck since smaller loads
will cool faster than larger loads.
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
43. Page 4-159 - 162: As discussed earlier, the entire loss on heating discussion should
be rewritten so that a RTFOT value of -0.5% is used in all cases. This way, the EPA
will not be providing yet another opportunity for builders of new plants to
misrepresent what might happen once the plant starts operating. The same applies
for the formulae which promote the use of variable temperatures.
3.2.2 Response to CAAP Temperature and Volatility Comment Number 1 - EPA does
not agree that we should present a single emission factor based upon the maximum
allowable asphalt volatility and maximum load-out temperature. Both the volatility (as
determined by the Rolling Thin Film Oven Test) and the FDVIA load-out temperature vary
with a number of parameters. EPA has obtained data showing a regional variation in the
volatility of asphalts used in road construction. EPA also has obtained information
showing that the storage and mixing temperatures for different asphalt binder grades are
different. Additionally, EPA recognizes that there are minimum delivery temperature
requirements for paving operations. Therefore, manufacturers will vary the load-out
temperature to accommodate transport times and ambient temperatures. EPA's equations
provide a reasonable estimate of emissions when used with asphalt volatility and
temperature information determined by valid measurements.
56
-------
We recognize that the various citizens' groups requested EPA to require the tested
facilities to use an asphalt with 1% loss on heating. Also, as we explained in the many
meetings preceding the tests, it should be recognized that EPA could not require HMA
plants to use 1% loss on heating asphalt during the emission tests. We also recognize that
the various citizens' groups requested EPA to adjust the data to a 1% loss on heating as
an alternative to using an asphalt with 1% loss on heating. EPA's analysis before the
emissions tests revealed that the adjustment to a 1% loss on heating is clearly not
supported by data collected by the Strategic Highway Research Program (SHRP)/National
Academy of Sciences (NAS). Additional data collected by EPA from selected State
Highway Departments, also does not support adjustment to a 1% loss on heating value.
The limited data obtained from the State Highway Departments also supported a much
lower loss on heating value. The State Highway Department data also indicates that the
use of a single loss on heating value for all States or regions is not supported.
EPA's commitment is clearly stated in item 4 on page 3 and in the third paragraph on
page 63 of the Response to Comments on Testing Program for Asphalt Plants C and D
Report. This commitment is that while emissions presented in the EA-Report would be
adjusted to 0.5% loss on heating for consistency, additional adjustments would be
recommended to accommodate local and State specific conditions. In March 2000, EPA
provided to all of the stakeholders 1) copies of all changed pages of the Plant C and D test
reports and 2) copies of all changed pages of the Response to Comments on Testing
Program for Asphalt Plants C and D report. EPA committed to change these pages as a
result of written and verbal comments from stakeholders. Specifically, page 3 of the
Response to Comments on Testing Program for Asphalt Plants C and D was included.
EPA provided the stakeholders 30 days to review the changed pages to verify that EPA's
commitments were properly reflected in the changed pages. There was no response from
any stakeholder.
As stated in Item 52 on page 62 through 65 of the Response to Comments on Testing
Program for Asphalt Plants C and D Report, less than 3% of the asphalts analyzed by the
SHRP/NAS nationally have a loss on heating greater than 1%. As shown in the test
reports for plants C and D, the earlier analysis method (used by SHRP/NAS) results in
loss on heating values about half the presently recommended RTFOT analysis method for
the same asphalt. The revised maximum loss on heating specifications is a recognition of
the differences between the previous analysis method and the presently recommended
RTFOT analysis method. The revised specification is not a relaxation of the standards for
asphalt.
RTFOT data provided by State Departments of Transportation and presented in Table 10
on page 64 of the Response to Comments on Testing Program for Asphalt Plants C and D
Report also show that adjustment to a 1% loss on heating using the presently
recommended test method is not supported. For four of the five States less than 0.0004%
of the asphalts analyzed would exceed the 1% loss on heating value. Only about 3% of
the asphalts obtained and analyzed by one state (Minnesota) would exceed the 1% loss on
heating value. As stated on page 63 of the Response to Comments on Testing Program
57
-------
for Asphalt Plants C and D Report the value of 0.5% reflects the 99.99% upper
confidence level for the average loss on heating value obtained from the State of
Minnesota.
While EPA agreed to present data adjusted to 0.5% loss on heating for consistency, this
would not be appropriate for the remainder of the States shown in Table 10. For the State
of Massachusetts, only 1.5% of the 44 individual asphalts tested had a loss on heating
greater than 0.5%. For most pollutants and most situations, the average loss on heating
value is the most appropriate value to be used. However, some State Agencies may wish
to use an alternate value for estimating emissions to address specific conditions of interest
to them. This alternate approach could be based upon a statistical analysis of loss on
heating data representative of the specific locality rather than a maximum value allowed
by a product specification.
The commentors have made statements previously that diluents and additive are added at
the distribution terminal after the State Transportation Department has obtained samples
for analysis. However, as presented on page 64 of the Response to Comments on Testing
Program for Asphalt Plants C and D Report each of the State Department of
Transportation employees that provided the data on loss on heating indicated that the
samples obtained were to be used without further blending. We recognize that asphalt
distributors blend asphalts and include other additives so that the asphalt supplied to the
HMA plant meets the required specifications. However, there is no evidence that
indicates diluents are added following the collection of samples by the State Department
of Transportation. Furthermore, the loss on heating analyses of samples collected during
the pre-test survey and during the emission test of plant D compare very favorably with
the loss on heating analyses performed by the State of Massachusetts Department of
Transportation.
The equation presented in the report is not to be used with a hypothetical loss on heating
value as implied by the commentor. The report specifically states that without regional or
site-specific data for asphalt volatility, a default loss on heating value of-0.5% should be
used. In addition, if State permitting authorities are concerned that a source may switch
to asphalts that are significantly more volatile than what was stated in their application for
the permit, the State could include enforceable limitations in the permit to preclude this
switch. This enforceable limitation could be based upon a statistical analysis of loss on
heating data representative of the specific locality rather than a maximum value allowed
by a product specification.
We do not know of any specifications for maximum delivery temperature. However, as
shown on Figures 1 and 2, the Asphalt Pavement Environmental Council's published
"Best Practices" brochure published on 4/00 provides guidance for controlling fumes,
emissions and odors from HMA plants and paving operations. The second side of the
brochure (Figure 2) includes recommendations for the range and midpoint temperatures
for both the storage of asphalt and the mixing of the HMA product. These temperatures
vary by the asphalt binder grades. The numbers in the binder grade are indications of the
58
-------
Asphalt Pavement Environmental Council Best Practices
Controlling Fumes. Emissions and Odors
from HMA Plant and Paving Operations
AT THE PLANT
• Select plant mixing temperature by:
— Contacting your asphalt supplier.
— Using the chart on the back.
• Do not use laboratory mixing tempera-
ture as plant mixing temperature.
• Make sure RAP and aggregates are dry.
• Do not use RAP containing coal tar.
• Do not expose RAP to flame.
• Do not over-heat RAP.
• Look for other sources of fumes
such as:
— Slag aggregate
— Shingles
— Crumb rubber mixtures
— Other products from construction
and demolition waste.
• Read the Material Safety Data Sheet
(MSDS) for all materials.
• Regularly calibrate thermocouples
and other sensors.
• Tune up the burner.
• Contact the manufacturer and find
out the limits on CO and 02.
• When the stack is tested, compare
the plant's thermocouple reading to
the tester's thermocouple.
• Gather data on aggregate moisture
content and fuel usage. If fuel usage
goes up for the same or less moisture,
find the reason.
• Have stack gases tested to see if
they are in limits. If not, contact
manufacturer to make adjustments.
• Compare mix temperatures with
plant temperatures. Look for
changes with time.
• Measure and record the pressure
drop in the baghouse. Look for
changes over time.
• Keep a record of fuel usage over
time. Find the reason for any
big changes.
• Keep track of this information and
discuss it with co-workers and the
manufacturer.
• Do not use diesel fuel and kerosene
as release agents.
AT THE PAVING SITE
• Try increasing the mat lift thickness
before calling for a higher plant
temperature.
• Do not use diesel fuel and kerosene
as release agents.
• Maintain engineering controls on
paving equipment.
ASPHALT PAVEMENT ENVIRONMENTAL COUNCIL
APEC is comprised of the following organizations: National Asphalt Pavement Association, Asphalt Institute, State Asphalt Pavement Associations
Figure 9. Asphalt Pavement Environmental Council, Best Practices Brochure, Side 1.
Reprinted with the permission of the National Asphalt Pavement Association.
59
-------
^^^^1 Asphalt Pavement Environmental Council Best Practices H^HI
^I^^^H Tvdii
cal Asphalt Binder Tern
nppfltimpfi ^^^^^^1
Binder Grade
PG 46 -28
PG 46 -34
PG 46 -40
PG 52 -28
PG 52 -34
PG 52 -40
PG 52 -46
PG 58 -22
PG 58 -28
PG 58 -34
PG 64 -22
PG 64 -28
PG 64 -34
PG 67 -22
PG 70 -22
PG 70 -28
PG 76 -22
PG 76 -28
PG 82 -22
HMA Plant Asphalt Tank
Storage Temperature (°F)
Range Midpoint
260 - 290 275
260 - 290 275
260 - 290 275
260 - 295 278
260 - 295 278
260 - 295 278
260 - 295 278
280 - 305 292
280 - 305 292
280 - 305 292
285-315 300
285-315 300
285-315 300
295 - 320 308
300-325 312
295 - 320 308
315-330 322
310-325 318
315-335 325
HMA Plant Mixing
Temperature (°F)
Range Midpoint
240 - 295 264
240 - 295 264
240 - 295 264
240 - 300 270
240 - 300 270
240 - 300 270
240 - 300 270
260-310 285
260-310 285
260-310 285
265 - 320 292
265 - 320 292
265 - 320 292
275 - 325 300
280 - 330 305
275 - 325 300
285-335 310
280 - 330 305
290-340 315
Use mid-point temperature for test strip construction.
I^^H ASPHALT PAVEMENT ENVIRONMENTAL COUNCIL ^^H
APEC is comprised of the following organizations: National Asphalt Pavement Association, Asphalt Institute, State Asphalt Pavement Associations
Figure 10. Asphalt Pavement Environmental Council, Best Practices Brochure, Side 2.
Reprinted with the permission of the National Asphalt Pavement Association.
60
-------
project-specific temperature extremes (in degrees centigrade) for which the asphalt
mixture is being designed. As such, a PG82-22 grade asphalt is intended for use when
average 7-day maximum pavement design temperature is 82° C (179° F) and the
minimum pavement design temperature is 22° C (-8° F). The midpoint FDVIA Plant
Mixing temperatures range from 264° F to 315° F. As shown on in Figure 2, the highest
HMA mixing temperature is associated with a binder used for the most severe
temperature conditions.
In addition, in a letter dated September 14, 1999 commenting on the Hot Mix Asphalt
Emissions Tests and Report for "Plants C & D" sent to Robert McConnell of EPA
Region I by Massachusetts State Senator Stephen F. Lynch, a copy of the Massachusetts
Highway Department 1995 Standard Specifications for Bituminous Materials was
attached. The standard includes three separate maximum temperatures measured at the
production facility. The maximum temperature standard for drum mix plants is 163 ° C
(325 ° F) measured at the point where FDVIA is loaded onto the hot mix conveyor (see
page III.25 Paragraph I). For batch mix plants there are two maximum temperature
standards, one for HMA containing RAP and one for HMA without RAP. The maximum
temperature for HMA containing RAP is 150° C (302° F). HMA that does not include
RAP is allowed to have a maximum temperature of 190° C (374° F).
The 1995 standard specifications provide no explanation for why there are three different
maximum production temperatures. However, process differences in batch mix and drum
mix plants do provide a reasonable rationale for why there would be differences in
achievable temperatures based upon the target temperature and RAP content of the
predominant product. Drum mix plants make and store multiple loads ahead of their
customers need for a given product. With relatively long production runs, they can
change the new aggregate temperature to accommodate the temperature and percentage of
RAP for that product. As a result, the production temperature can be the same despite the
type of product or the ratio of RAP required. On the other hand, batch plants store
sufficient preheated and screened aggregate for two to four truck loads of HMA.
Therefore, any changes in aggregate temperature must be anticipated at least two
transport trucks in advance and there is little capability to change aggregate temperatures
for every product specification. As a result, it would be difficult for a batch plant to
concurrently produce HMA with and without RAP and meet the same temperature
requirements. It is believed that most batch mix facilities would heat the new aggregate
to the temperature requirements of the predominant product but within a range that is also
acceptable for production of the other products. As a result, for the more typical case
where the predominant product includes RAP, products which do not contain RAP will
be at a higher temperature. For example, a batch mix plant with a predominant product
which contains 10% RAP at the maximum load-out temperature of 150° C (302° F)
would heat the new aggregate to approximately 165° C (330° F). As a result, when a
customer requests HMA without RAP, the load-out temperature will be 165° C (330° F).
The statements with respect to the maximum recommended load out temperature are
sufficiently accurate. Therefore, the use of a default temperature of 325 ° F is a
reasonably high default value to accommodate all but the most extreme conditions.
61
-------
3.2.3 CAAP Temperature and Volatility Comment Number 2 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
23. Table 14: This table has to be redone so that an actual emission factor is given; not a
formula for calculating low emissions based on fictitious values of asphalt volatility.
3.2.4 Response to CAAP Temperature and Volatility Comment Number 2 - Many other
Sections of AP-42 employ equations to improve the precision of emission factors when
there are documented characteristics which are readily available for use to develop a
regionally or locally specific emission factor. Some examples are paved and unpaved
roads, surface coal mines, wind erosion, and several emission factors for fuel combustion.
The table presents formulas that allow emissions to be calculated based upon measured
(not fictitious) parameters representative of the regional or site specific conditions. In the
December 17, 1999 draft and the May 2000 final version of the Response to Comments
on Testing Program for Asphalt Plants C and D Report, the adjustments for the volatility
and load out temperature were specified. The basis of these adjustments is clearly
documented in the text and tables in response numbers 52 and 53. While the equations
are different in appearance from that presented in response numbers 52 and 53, the
equations implement the tabular values for the adjustments presented.
3.2.5 CAAP Temperature and Volatility Comment Number 3 (Ravi Nadkarni & Lloyd
Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
14. Page 14. paragraph 3: EPA again makes the same recommendation which is to
assume that the load out temperature can never exceed 325° P. This is incorrect. For
example, Mr. Webster, the industry expert who has attended many of the
EPA/citizen group/industry meetings has looked at the photographs that Ravi
Nadkarni submitted to the EPA in 1994 and are in the EPA files. His comment was
that the photos were typical of a plant loading out material at a high temperature
when supplying small contractors. If this practice is common enough to be readily
identified at a glance, EPA's comments about 325° F being the maximum
temperature under all conditions are wrong and need to be corrected.
3.2.6 Response to CAAP Temperature and Volatility Comment Number 3 - The
statement is correct. At the time the citizens, EPA and industry agreed to request the
facilities tested by EPA to target a load-out temperature of 325° F, the industry stated that
this temperature was their best practices' recommendation. While a facility may target a
specific load-out temperature, variations in storage time and some characteristics of the
aggregate and RAP may cause small variations in the actual load out temperature. It is
expected that temperature variations at batch mix plants would be greater than at drum
mix plants. This is because batch mix plants generally do not store HMA product.
However, they store aggregate heated to the desired temperature for their projected
principal product. Using the previous example, if the principal product was HMA
containing 10% RAP with a desired load-out temperature of 302° F, the new aggregate
would be stored at about 330° F. If a single load of HMA with no RAP was required, the
62
-------
load-out temperature for that load would be about 330° F. Based upon these two
temperatures, for load-outs of equal amounts of these two materials, the HMA with no
RAP would have emissions that are about twice the emissions of the HMA with RAP.
Because of the difference in these emissions, the visual appearance of these two load-outs
would be markedly different. It is believed that this is the type of situation described to
Ravi Nadkarni by Mr. Webster. This type of situation occurred during emissions testing
at Plant D about 10% of the time. It is expected that the situation described by the
commentor would occur at about the same percentage of time. However, batch mix
facilities that routinely produce significantly more than about 10% of their product
without RAP should accommodate the emissions differences by adjusting the average
load-out temperature used in the emission estimation equation.
3.3 Adjustment for Background Truck Operation Emissions
3.3.1 CAAP Background Emissions Comment Number 1 (Ravi Nadkarni & Lloyd
Pillion) -
Commentors Summary Section C
Improper Manipulation of Data from the "Background" Test Run 4. at Plant C: The
report states correctly: "The most reliable method to adjust for emissions measured
during background operations would be to separately adjust each run for the measured
capture efficiency and then subtract these adjusted background emissions from the
adjusted emissions measured during production operations." However, because this
"procedure produces negative values for both the PM and MCEM and many other HAP
compounds", a new unsound procedure is adopted because it gives results that please
EPA and industry by showing low but positive emissions. This new procedure is to use
background emissions uncorrected for capture efficiency. If the actual emissions have to
be corrected for capture efficiency because the enclosure does not capture all the
emissions, there is no justification for ignoring the capture efficiency for the background
run. Having used this incorrect analysis, the authors then have the temerity to advise us
that this result might even have a "high bias" relative to the correct method mentioned
above.
The report justifies this faulty procedure because this situation "cannot be accommodated
retroactively." This statement is not correct. The best way to eliminate improper data
manipulation is to discard the phony background adjustment and report the data as that
from loadout plus truck exhaust. Since silos always dump hot mix into a truck, these
combined emissions are present at each silo loadout point. Further, in the unlikely event
that a local permitting authority requires inclusion of truck exhaust emissions (unlikely
because we have not seen such calculations associated with any local permitting
requirements), note that the truck spends about /^ minute under the silo but many more
minutes, typically 3 to 8, on site. As a result, the truck emissions that are included under
loadout correspond to the /^ minute portion. Therefore, the double counting of truck
exhaust emissions would be minimal and can be adjusted for by subtracting the 1A minute
from residence time of the truck on site.
63
-------
Overall, the background run demonstrates how data collection was manipulated to
produce biased results. The EPA Project Officer was interested in recording a high
background and the original data showed that he was able to manipulate truck placement
near the entrance of the tunnel in such a way that uncorrected emissions in the second half
of the run were about twice the emissions in the first half. If these emissions are
corrected for the lower capture efficiency in the second half of the run, the corrected
emissions from the second half are almost three times the corrected emissions from the
first half. In spite of using the emissions from the first portion of the background run, it
appears that they were still too high when corrected for capture efficiency and resulted in
negative or zero emissions to the environment.
Commentors Summary Section C
Unexplained Differences Between MCEM and TOC: The comparison of Plant C and
Plant D results shows a wide variation between emission rates for MCEMs and TOC.
MCEM PMs and TOC are generated the same way and from the same source. Therefore,
if MCEM generation is a function of loadout time, TOC generation should also be a
function of loadout time and TOC emissions should be higher for a batch plant in the
same ratio. The report does not explain this discrepancy. Of course, the comparison is
further complicated and made meaningless by the bogus "background" correction for
drum plants discussed earlier. To eliminate the confusion, separate emission factors
should be prepared for batch and drum plants in all cases.
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
12. Page 13. paragraph 5: This paragraph discusses the subtraction of "background
emissions", which reduced the calculated loadout emissions significantly. Our
detailed comments are in our page-by-page discussion of Appendix B, which show
that the method used is totally without foundation or theoretical support, and the only
excuse for using it is that it gives non-negative results in several cases.
3.3.2 Response to CAAP Background Emissions Comment Number 1 - The background
adjustment was appropriate. There was no improper manipulation of the data from the
background test at Plant C and EPA did not manipulate the placement of the trucks to
obtain higher uncorrected emissions for the background run. Further, we do not agree
that the background run demonstrates that data was manipulated to produce biased results
and do not believe that concurrently measured truck exhaust and road dust emissions
should be included in the emission factor for load-out emissions.
As stated in Appendix B (AP-42 Background Report) and in item 51 on page 56 through
61 of the Response to Comments on Testing Program for Asphalt Plants C and D Report,
the background adjustment was performed to accommodate the emissions from the
operation of the trucks within and near the enclosure. The information presented on
pages 56 through 61 of the Response to Comments on Testing Program for Asphalt Plants
C and D Report shows that the emissions measured during the background run agree
reasonably well with published data for diesel engine emissions. This is also explained in
Section 3.3.6 of this response. As explained in Section 3.3.6, the truck drivers were
64
-------
instructed to operate their trucks (including the placement prior to entering the tunnel and
placement to simulate loading) as they did during normal operation. The only additional
instruction provided to the truck drivers during the background test was to reduce the
time of their travel from the exit of the tunnel to the arrival at the tunnel entrance.
During the emission test at Plant C, Dr. Nadkarni observed the operation of the trucks
during the background run. The issue of manipulating the placement of the trucks was
not raised by him during the test to either of the EPA personnel present.
The differences in THC concentrations recorded for the first and second half of the
background run do not indicate that either truck placement or data were manipulated. At
about the time trucks began driving faster to reduce gaps between trucks, the wind speed
increased. This may have caused an increase in the diesel exhaust that entered the tunnel
entrance or increased the capture of the diesel exhaust of the truck that was inside the
tunnel. Given that the THC concentrations during the background run were low, the
differences in concentrations appear more significant.
Commentors raised concerns with the calculated average background THC concentrations
at a stakeholders meeting to discuss EPA's draft document "Response to Comments on
Testing Program for Asphalt Plants C and D." At the meeting, commentors raised
concerns with the differences between the first and second half of the background test run
suggesting the second half run measured emissions from two trucks; one in the tunnel and
one waiting at the tunnel entrance. While EPA disagreed with the commentors
assertions, we agreed to use only the first half run data (0.83 ppm) in lieu of the average
of the entire run (1.2 ppm) to determine the background concentration to be used in
adjusting for background concentrations that were attributed to truck engine emissions.
This change of background concentration lowered the THC background adjustment from
9.7% to 6.7% of the capture efficiency adjusted THC emissions. If the background
concentration of 0.83 ppm were adjusted for the 45% capture efficiency for Run 4, the
background adjustment to the THC emission factor calculation would be 1.84 ppm.
Adjusting for capture efficiency would have resulted in a 14.5% reduction of THC
emissions compared to the 6.7% reduction used by EPA.
Publishing load-out emission factors, which includes diesel exhaust and road dust
emissions as the commentor suggests does not address diesel exhaust or road dust
emissions. These emissions were excluded since more comprehensive methods to
calculate these emissions are available in other parts of AP-42. Specifically, mobile
source factors are available for different engine type by milage and year of production and
fugitive dust emission factors are available that accommodate the type and dustiness of
the road surface. Excluding emissions not related to load-out operations isolates the
emissions of concern, precludes future double counting emissions and precludes other
improper manipulation of emissions data.
65
-------
3.3.3 CAAP Background Emissions Comment Number 2 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot
Mix Asphalt Production, June 2000 Draft
35. Page 4-152. paragraph 3: This paragraph presents a major problem in this report.
The second line states correctly: "The most reliable method to adjust for emissions
measured during background operations would be to separately adjust each run for
the measured capture efficiency and then subtract these adjusted background
emissions from the adjusted emissions measured during production operations."
However, because this "procedure produces negative values for both the PM and
MCEM and many other HAP compounds", a new unsound procedure is adopted
because it gives results that please EPA and industry by by showing low but positive
emissions. This new procedure is to use background emissions uncorrected for
capture efficiency. If the actual emissions have to be corrected for capture efficiency
because the enclosure does not capture all the emissions, there is no justification for
ignoring the capture efficiency for background.
Having used this incorrect analysis, the authors then have the temerity to advise us
that this result might even have a "high bias" relative to the most correct method
mentioned above. In a very narrow sense, this comment might be correct, but in the
current context, given the consistent efforts of the Project Officer to bias the results
to favor industry, this comment is totally off base.
Line 5 states that this situation "cannot be accommodated retroactively." This
statement is not correct. The best way to eliminate improper data manipulation is to
discard the phony background adjustment and report the data as that from loadout
plus truck exhaust. Since silos always dump hot mix into a truck, these combined
emissions are present at each silo loadout point. Further, if a local permit application
requires the inclusion of an estimate for truck emissions on site (and we have yet to
see a local permit that requires this), note that the truck spends about !/2 minute under
the silo but many more minutes, typically 3 to 8, on site. As a result, the truck
emissions that are included under loadout correspond to the !/2 minute portion.
Therefore, the double counting of truck exhaust emissions would be minimal and can
be adjusted for by subtracting the !/2 minute from residence time of the truck on site.
We strongly recommend that Tables 4-25 and 4-26 be revised to reflect this
approach.
3.3.4 Response to CAAP Background Emissions Comment Number 2 - The goal of the
test program and the document being commented on was to develop emission factors for
publication in AP-42 for the various criteria and hazardous air pollutants resulting from
truck loading and silo filling. Emission factors for specific process operations reduce the
possibility of inappropriate and inconsistent manipulation of emissions to satisfy the
differing State and local requirements for permit and inventories. Different methods are
used by some states to evaluate and control fugitive, mobile source and point source
emissions which occur within a facility.
66
-------
Since emission factors for diesel truck tailpipe emissions and for paved and unpaved road
emissions are available elsewhere, emissions sources may satisfy some States request to
segregate these emissions. A known example of this occurs with unpaved and paved road
emission factors. Both emission factors include the vehicle tailpipe emissions since they
were not accommodated in the emission tests that provide the supporting data for the
emission factors. EPA, and many State and local agencies subtract the estimated tailpipe
emissions as calculated with the emission factors for the type of vehicle generating the
fugitive emissions. This is reasonable for situations where the driving conditions of the
vehicles are similar to the Federal Test Procedure for motor vehicles. However, in some
situations (such as low load and mild acceleration conditions) this adjustment would over
correct the emissions estimate. This latter situation is a distinct possibility for the
combined load-out, diesel exhaust and fugitive dust emissions described in this comment.
Emissions for many pollutants measured during the background run agreed reasonably
well with information published by EPA's Office of Transportation and Air Quality and
in two journal articles for diesel emissions. It should be noted that additional information
is being developed by EPA and others to better characterize diesel engine emissions for a
wider variety of engines. Many AP-42 emission factors are revised from time to time to
reflect improved information and changes that may occur due to improvements in
emissions control. For diesel engine emissions, this is highly probable. Many diesel
engine manufacturers are developing revised engine designs and exhaust controls. These
revisions could potentially result in a significant reduction of diesel engine emissions.
The availability of separate emission factors for individual, distinctly different processes
allow State and local agencies to more effectively meet the individual requirements of
their programs. Additionally, separate emission factors eliminate the possibility for
inconsistent and erroneous adjustments to the composite emission factor for truck
load-out.
There are several justifications for not adjusting the background emissions for the
measured capture efficiency as part of the process for isolating the load-out emissions
from the composite emissions measured. First, it is evident that most of the visible
particulate emissions were from the load-out operations and relatively little came from
road dust and truck exhaust. Correcting the background emissions for the measured
capture efficiency resulted in background particulate that exceeded the capture efficiency
corrected emissions measured during the production tests. In addition, the capture
efficiency corrected background organic particulate emissions (MCEM) were 99.5% of
the capture efficiency corrected emissions measured during the production tests. This is
clearly contrary to the observed situation. Since only two trucks were available during
the background run, the entrance to the tunnel was obstructed less than half the time by a
truck waiting to enter the tunnel. During the production tests, the tunnel entrance was
obstructed over 90% of the time. The lack of obstruction allowed the prevailing winds to
scavenge the tracer gas away from the capture hood and result in poorer capture of the
tracer than would occur otherwise. Without the thermal lift created by the hot asphalt, the
tracer was not drawn into the capture hood like during the production runs. And lastly,
67
-------
excluding the capture efficiency adjustment would tend to understate the significance of
the road dust and truck exhaust and provide a moderate and acceptable safety factor to the
truck load-out emissions.
The adjustments to the data were reasonable. As stated above, the modification of the
method used to adjust the composite emissions measured during the production periods
provide at most a moderate safety factor to the truck load-out emissions while resulting in
adjusted emissions that agree to the situations observed during the emissions tests.
Further, published emission factors for diesel exhaust use the number of vehicle miles
traveled and not on the time of operation. Adjusting road dust and diesel exhaust
emissions for the time period or distance trucks' travel introduces additional complexities
into calculating emissions for a facility and for an area. The availability of separate
emission factors representing only truck load-out emissions will allow State and local
agencies to meet multiple individual requirements with a minimum of inconsistent and
potentially erroneous adjustments to the composite emission factor for truck load-out.
3.3.5 CAAP Background Emissions Comment Number 3 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot
Mix Asphalt Production, June 2000 Draft
36. Page 4-152. paragraph 4: This paragraph further demonstrates the problems with
EPA's background run. The EPA project officer was interested in recording a high
background and the original data showed that he was able to manipulate truck
placement in such a way that uncorrected emissions in the second half of the run
were about twice the emissions in the first half. If these emissions are corrected for
the lower capture efficiency in the second half of the run, the corrected emissions
from the second half are almost three times the corrected emissions from the first
half. This further points out the problems with EPA's manipulation of and
placement of trucks to maximize exhaust emissions inside the tunnel.
We have to apologize for not taking this particular analysis far enough a year ago
during a review of the Plant C draft. When the data for Run 4, the background run,
was reviewed, it was clear that there was data manipulation going on since the raw
data showed that emissions in the second half of the run were double those in the
first half. When these emissions were adjusted for capture efficiency, the emissions
from the second half almost tripled. We felt that the data from the first half of the
run was more appropriate than that from the second half. We did not realize that this
data, after correcting for capture efficiency, still overwhelmed and negated the load
out emissions and was therefore also tainted.
3.3.6 Response to CAAP Background Emissions Comment Number 3 - Differences
between the THC measured during the first and second half of the background run cannot
be fully explained without some conjecture. However, there was neither manipulation of
the truck placement nor manipulation of the data. As explained in response number 41 of
the "Response to Comments on Testing Programs for Asphalt Plants C and D" document,
68
-------
we instructed the two truck drivers used during the background run to replicate normal
operations as much as possible. This included, the location where they stopped before
entering the tunnel and the time and location within the enclosure to simulate loading.
During normal production, there was almost always a truck parked about two to four feet
from the tunnel entrance waiting for a signal to enter the tunnel for loading. Since only
two trucks were available for the background run, the tunnel entrance was blocked
between 20 and 70 percent of the time. The remainder of the time was required for a
truck to exit the tunnel, circle the plant and return to the tunnel entrance. At about the
midpoint of the background run, the truck drivers were asked to drive a little faster to
return to the tunnel entrance quicker. As a result, the last half of the background run
approximated normal operations more closely than the first half. In addition to increasing
the time the tunnel entrance was blocked by a truck, the wind appeared to increase in
velocity during the second half of the background run. During the tests during
production, generally the emission testing was complete by the time the wind velocity
increased.
While the commentors compare the differences in THC emissions between the first and
second half of the background run, they fail to notice that the first and second half
background emissions are only 10 and 20 percent of the emissions during production. In
addition, by adjusting only the production runs for capture efficiency, the correction for
background emissions for all pollutants is minimized. Contrary to the statement by the
commentors, few background corrections overwhelmed or negated the emissions
measured during the production runs. The above corrections of 10 to 20 percent for THC
are comparable to the background corrections for most other pollutants. There are
explanations for the few instances where there was a significant background correction
for a pollutant. For example, the non-MCEM particulate matter is probably fugitive dust
generated from the roadways approaching and within the tunnel. Additionally, emissions
of chlorinated hydrocarbons (Methylene Chloride, Tetrachloroethene, 1,1,1 -
Trichloroethane, Trichloroehtene etc.) would not be expected from asphalt.
In addition, the commentors ignore the comparison of mass emissions measured during
the background run with the published emission rates from diesel trucks made in response
number 51 of the "Response to Comments on Testing Programs for Asphalt Plants C and
D" document and specifically in Tables 7 and 8 in this document. The last paragraph on
page 57 and the first full paragraph on page 59 of this document summarizes the
comparisons. The last sentences of the paragraph on page 57 states, "The uncorrected
emissions measured during the background run was 0.66 g/min for THC, 1.2 g/min for
CO, and 0.54 g/min for PM. Except for PM, these agree very well with the estimated
emissions from AP-42. It should be noted that the fugitive PM emissions are not
accounted for in the above comparison." And the last sentences of the paragraph on page
59 states, "For all but four compounds, the emission rate during the background run was
lower than presented in the articles. For these four compounds, the emission rate was
within a factor of two of the rates presented in the articles. More background run
compounds do not compare more favorably to the emission rates presented in the articles
69
-------
because the Federal Test Procedure begins with a cold start and the acceleration
requirements are greater than the truck accelerations used during the test."
3.3.7 CAAP Background Emissions Comment Number 4 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
56. Table 4-23. 4-25: In view of the comments earlier about the background run, these
tables need to be revised.
3.3.8 Response to CAAP Background Emissions Comment Number 4 - Based on EPA's
response to comments in Section 3.3.6, the tables were not changed.
3.3.9 CAAP Background Emissions Comment Number 5 (Ravi Nadkarni & Lloyd
Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
17. Page 16. paragraph 1: How do the truck exhaust emission factors published by EPA
compare with the background measurement from Run 4 at Plant C? Why is such a
comparison missing from the entire report?
3.3.10 Response to CAAP Background Emissions Comment Number 5 - A comparison
of the background measurement from Run 4 at Plant C was compared to the truck exhaust
emission factors published by EPA and others were presented on pages 57 to 59 and in
Table 8 in item 51 of the Response to Comments on Testing Program for Asphalt Plants
C and D Report. While the acceleration conditions for the trucks during Run 4 are less
severe than during conditions use to develop EPA published truck exhaust emission
factors, the emissions agreed reasonably well. A comparison is not made in this report
since the validity of the test results is independent of other measures of diesel truck
emissions by others.
3.3.11 CAAP Background Emissions Comment Number 6 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot
Mix Asphalt Production, June 2000 Draft
33. Page 4-151. paragraph 2: Note that Cambridge Environmental has calculated the
cooling of a mass of PDVIA. Their results are not supported by any measurements.
3.3.12 Response to CAAP Background Emissions Comment Number 6 - It is not clear
how the calculations made by Cambridge Environmental relate to this paragraph. In
addition, the emission factor developed for the period following load-out account for only
the first eight minutes following load-out and do not account for any cooling of the hot
mix asphalt in the transport truck.
70
-------
3.4 Assignment of Quality Ratings
3.4.1 CAAP Quality Ratings Comment Number 1 - (Ravi Nadkarni & Lloyd Pillion) -
Commentors Summary Section C
Inconsistent Assignment of Ratings: In one portion of the report, the emission factor for
sulfur dioxide was assigned an E rating because the data ranged over an order of
magnitude. This might have been simply the result of the variation in sulfur content of
the fuel. In the case of filterable PM, the range is over two orders of magnitude but the
emission factor got an A rating
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
1. Page 12. paragraph 2: Is the rating system applied consistently? The discussion in
Appendix B indicates that it was not.
3.4.2 Response to CAAP Quality Ratings Comment Number 1 - The emission factor
rating system was applied consistently. The discussion in Appendix B Section 3 includes
a synopsis of the emission factor rating system. More detail is included in the reference
cited at the end of Section 3. The commentor may be confusing the between source
variations of the supporting data as the basis for the emission factor rating. While
characteristics such as the variability of the supporting data may be used to adjust the
emission factor rating, the primary criteria for assigning the emission factor rating are the
quality of the supporting emission test data, the representativeness of the facilities tested
to the source category and the number of facilities tested.
3.4.3 CAAP Quality Ratings Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
9. Page 13. paragraph 1: The quality rating needs to be quantified so that one can get a
numerical feel for the differences in ratings. If, as stated, the quality ratings are a
"function primarily of the number of data points", these various quality ratings can
be related to the standard error of estimate and using t-values, a range can be shown
about the mean.
3.4.4 Response to CAAP Quality Ratings Comment Number 2 - As stated, the emission
factor quality ratings are related to the number of data points used to develop the emission
factor. As such, the rating is primarily an indicator of the robustness of the emission
factor. For example, an emission factor that is likely to change significantly with the
addition or deletion of data would be rated low. Whereas, an emission factor that is not
likely to change significantly with the addition or deletion of data would be rated high.
Information on the variability of the data supporting the emission factors are presented in
an appropriate location when sufficient supporting data are available to provide a
reasonable estimate of the variability.
71
-------
3.4.5 CAAP Quality Ratings Comment Number 3 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
11. Page 3-3. final paragraph: The comments here refer to variations or fluctuations in
measured results which could not be explained readily. While we agree that such
results might be given a lower rating, in many cases, variations are a normal part of
the production process. For example, if a batch plant is producing a mix with a high
RAP content, the crushed stone is heated to a high temperature to provide the heat
for evaporation of the moisture associated with RAP stored in the open. If another
customer comes in for a RAP-free mix at this stage, the hot mix will be loaded out at
a much higher temperature and will emit more than the usual amount of fugitive
organics. This is a normal transient.
3.4.6 Response to CAAP Quality Ratings Comment Number 3 - The point of the
comment is not clear. The paragraph on page 3-3 is addressing the criteria used to
evaluate source test reports for sound methodology and adequate detail. As stated in the
paragraph if the test results have a high variation without process information that may
explain the high variation, the source test is downrated. While one could assume that
undocumented process variations were the cause of the variation in emissions, facilities
generally monitor their process closely during emissions tests. While the information
monitored may not be documented in the source test report, a source test run is not begun
until the process is operating within the normal variation and source test runs are stopped
when the process begins operating out on normal. It is recognized that variations are a
normal part of the production process. It is also recognized that there will be a normal
variation in the emissions of a facility. However, when the emission variation for one
source test is significant compared to data from multiple facilities, it is likely that the
variations are an indicator of the quality of the source test contractor. The example the
commentor gives would result in short term variations in emissions. However, the
emission tests cited in the background report for the batch mix facility were conducted
over six or more hours. Although most of the HMA produced during the test included
10% RAP, approximately 15% of the HMA produced had no RAP. It is recognized that
the short term variations that may occur with this situation are normal for this process and
the source test was not downgraded because of the variations.
3.4.7 CAAP Quality Ratings Comment Number 4 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
11. Page 3-3. final paragraph: There are many factors which contribute to the emissions
which were not measured. For example, SO2 emissions from combustion are
dependent on sulfur in the fuel. The emission factors for different fuel types are a
result of the maximum sulfur content allowed under the fuel specifications. No such
analysis is presented. The same comments hold true for organic fugitive emissions
being dependent on volatile content of the asphalt.
72
-------
3.4.8 Response to CAAP Quality Ratings Comment Number 4 - It is agreed that there
may be many factors which might contribute to differences in emissions which were not
measured. However, if the information is not presented in the test report, an analysis
cannot be performed. The vast majority of the emission tests used to develop the
emission factors were obtained from the source test files of State and local air pollution
control agencies. As a minimum, these source tests contain the information the State
agency requires for determining compliance with applicable emission standards for the
facility. As indicated in the summaries of the over 350 test reports reviewed for this
section, additional information that would be helpful to evaluate a number of process
variables were not available in most of the source test reports. Specifically for SO2, there
were very few source test reports for oil fired HMA plants. In addition, the fuel usage
and fuel analysis for the fuel was not available for all of the oil fired tests. Therefore, a
detailed analysis was not possible. However, the industry has supplied 14 source test
reports for SO2 which include fuel usage and sulfur analysis. These tests were evaluated
to determine whether there was any relationship between the sulfur in the fuel and SO2
emissions. Because of a variable reduction of the SO2 emissions, no relationship between
SO2 emissions and the more typical parameters (fuel sulfur content, fuel usage, control
device temperature) that affect SO2 emissions could be found. With respect to the
availability of the volatile content of the asphalt used during emissions testing, if this
information were included in any of the over 350 kiln stack test reports it was not noticed
during the review of the source test reports and generally State Agencies do not require
this information to be collected during compliance tests.
3.4.9 CAAP Quality Ratings Comment Number 5 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
15. Page 4-122. paragraph 4: This paragraph states that the emission factor for sulfur
dioxide was assigned an E rating because the data ranged over an order of
magnitude. This might have been simply the result of the variation in sulfur content
of the fuel. In the case of filterable PM (page 4-118), the range is over two orders of
magnitude but the emission factor got an A rating. A little consistency would help.
3.4.10 Response to CAAP Quality Ratings Comment Number 5 - The assignment of the
emission factor ratings is consistent. The paragraph does not state that the sole reason for
assigning an E rating to the emission factor for sulfur dioxide for #2 fuel oil fired drum
mix dryers was the variability of the data. It should be noted that all of the sulfur dioxide
emission factors for drum mix dryers are rated D or E. The primary reason for the low
rating was the availability of only three to six source tests. An additional adjustment
from a D rating to an E rating was made because of the low number of tests and the
higher variability of the data. The rating of A was assigned to the emission factor for
filterable PM for fabric filter-controlled drum-mix dryers because there were 145 tests.
This is well in excess of the number of tests that would justify an A rating. While the
data ranged over two orders of magnitude, the large number of data contributed to the
magnitude or the range of the data. In addition, the relative standard deviation of this
data is near one and is typical of many emission factors for PM.
73
-------
3.4.11 CAAP Quality Ratings Comment Number 6 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
16. Page 4-125. paragraph 4: This paragraph states that the emission factor for trace
metals was assigned an E rating because the data ranged over two orders of
magnitude. This might have been simply the result of the variation in trace metal
content of the feed materials and fuel.
3.4.12 Response to CAAP Quality Ratings Comment Number 6 - The paragraph states
"An emission factor rating of E was assigned to data sets with only one or two data
points. An emission factor rating of D generally was assigned to data sets with three or
more data points, and an emission factor rating of C generally was assigned to data sets
with seven or more data points." The paragraph further states that a rating of E was
assigned for silver because there were only three data points and the data ranged over two
orders of magnitude. While it is likely that the variation of all of the metals emission
factor are the result of variations in trace metal content of the feed materials, the metals
content of the fuel and the effectiveness of the control device, the factor rating system
would not accommodate these variations unless there was data to correlate the emissions
with the process information.
3.4.13 CAAP Quality Ratings Comment Number 7 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
30. Page 4-148. paragraph 1: Reference 6 (fourth line) is incorrect. What is the proper
reference? We are also amused that this report has been given an A rating for data
because of the "few problems in the Technical systems Audit Report" etc.
Unfortunately, this audit report missed all of the errors found by the citizens.
Further, the citizens caught the errors through spot checking and there are no
guarantees that they caught everything. If such errors can exist in a report that was
guaranteed a high degree of scrutiny, how sure is the EPA that the other data is error
free? We am particularly concerned about this point because the data, which is
supposedly in appendices A, B and C, is not attached to the material sent to us for
review. Also, it raises the question of whether EPA staff do any of their own
reviews and spot checks before accepting a report from a contractor. It does not
appear to be the case.
3.4.14 Response to CAAP Quality Ratings Comment Number 7 - We agree that citing
Reference 6 is a typographical error. In addition, the "Response to Comments on Testing
Programs for Asphalt Plants C and D" document cited is not included in the references at
the end of the Chapter. This document was added to the list of references at the end of
the chapter and the proper citation replaced the citation in this paragraph. The audit
report was an independent evaluation of the sampling and analyses performed for the
testing at Plant C. The citizens did not identify any additional errors that were not
identified in the audit report. The citizens did comment on one statement in the audit
report which was not accurate and is identified in response number 43 on page 50 and 51
74
-------
of the Response to Comments on Testing Program for Asphalt Plants C and D. While the
draft audit report stated that the "VOST and Modified Method 5 methods have not been
validated for all chemical compounds of interest in asphalt plant emissions," response
number 43 corrected this statement by providing information on why the test methods
used provided a reasonable assessment of the emissions. Page 2 of the Response to
Comments on Testing Programs for Asphalt Plants C and D reports provides summaries
of changes made to the test reports. There were two changes that might be considered
significant errors in the original test reports. The other seven identified changes in the
test reports either result in a very minor difference in the emissions presented in the draft,
added emissions of some pollutants that an instrument was capable of measuring but was
not presented in the draft, make the presentation of some information more prominent, or
provide some editorial changes in the text. The final reports are free of significant errors
and have been reviewed multiple times by EPA staff and EPA contractors to produce a
document which can withstand the most critical unbiased evaluation of the quality of the
data presented. Because of the thorough documentation of the high quality emission
testing activities, and the adherence to an established and approved quality assurance test
plan, the source test exceeds the criteria for receiving an A rating.
3.4.15 Local Health Agency Quality Ratings Comment Number 1 (Elaine T. Krueger,
Environmental Toxicology Program (BEHA) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
As we have previously noted, it is important to understand the limitations of the emission
testing at the Barre and Irvine plants, some of which were discussed in EPA: response to
comments (i.e., "Emission Tests of Hot Mix Asphalt Plants, Response to Comments for
Asphalt Plants C and D"). EPA noted that resources were not available to perform
emission testing at different times of year or at different locations to determine whether
there is a relationship between emissions and asphalt cooling or variability of emissions
of specific chemical compounds. This information could be important in determining
emissions of specific chemical compounds. This information could be important in
determining emission from plants in different areas of the country. While we appreciate
that limited resources may have prevented more comprehensive testing, unfortunately a
number of uncertainties remain regarding opportunities for exposure and health concerns.
In addition to those noted above, uncertainties also exist due to the variations in modes of
operation, materials, and plant designs. Therefore emission data from previous asphalt
studies may not be applicable to the Barre plant. We would recommend that these
important limitations be outlined in the final document, rather than just in the response to
comments.
3.4.15 Response to Local Health Agency Quality Ratings Comment Number 1 - It would
not be reasonable to assess the limitations of the emissions information without knowing
the intended final use and without comparing this information to other source categories.
The most significant uncertainties and limitations of the emission factors for HMA plants
is very similar to the uncertainties and limitations of emission factors for most source
categories. The quality and quantity of chemical specific emissions information available
for HMA plants is at least comparable to many source categories in AP-42 where this
75
-------
information is available. In fact, many people familiar with chemically specific emissions
information would consider the information for HMA plants to be superior to the
information for most other source categories. First, chemically specific emissions
information is not available for many source categories that may potentially create the
greatest exposure to the individuals. Second, compound specific stack emission factors
based on more than three to five emission tests is atypical of most source categories in
AP-42. Third, the use of additional laboratory analyses was very successful and as a
result allows site specific conditions with respect to asphalt volatility and load-out
temperature to be used as parameters to more reliably estimate the load-out and silo
filling emissions. Many of the uncertainties and limitations of emission factors are
presented in the Introduction to AP-42, pages 3 through 5
(http://www.epa.gov/ttn/chief/ap42/cOOsOO.pdf). It is highly recommended that AP-42
emission factors users read this introduction to understand the development, use and
limitations of these emission factors. When developing AP-42 emission factors, EPA
assembles the best and most complete information available. As a result, the information
that is presented in the AP-42 section and background report for HMA plants is the best
data available and additional comments with respect to general uncertainties and
limitations was not included in the final report.
3.5 Emission Factor Development Methods
3.5.1 CAAP Factor Development Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
8. Page 12. paragraph 4: This paragraph presumably refers to the statistical analysis
presented in Appendix B, but there is no direct link or reference. Our detailed
comments regarding the statistical analysis are in our page-by-page discussion of
Appendix B. However, the entire paragraph is wrong or inconsistent with
information in other parts of the report. It states that the statistical analysis showed
"no strong correlation" between parameters such as fuel type and emissions factors.
When such correlation is absent, the approach has been to use the same emission
factor in both cases. Yet, Table 11.1-5 in Appendix A shows different emission
factors for NOX and SO2 for different fuel types. For NOX the emission factor is
different for natural gas and oil fired dryers. For SO2 there are three different
emissions factors, one for gas, one for No. 2 oil and one for No. 6 oil. The backup
analysis is not shown in Appendix B. But obviously, the analysis showed that fuel
type did affect emissions, otherwise a single factor would have been used. Overall,
the two sets of comments need to be made more explicit and explained more
carefully and the inconsistencies removed.
3.5.2 Response to CAAP Factor Development Comment Number 1 - There is no error in
what was stated and there is no inconsistency between what is stated in this paragraph and
the remainder of the report. The information presented in this paragraph is based upon
the detailed statistical analysis which attempted to discern a relationship between several
operating characteristics and emissions. As stated in Appendix B, where the quantity of
data should be sufficient to discern a relationship, no relationship could be documented.
76
-------
As a result, the data were combined to arrive at a single emission factor for the differing
characteristics. However, for some pollutants such as NOX and SO2, there were few
supporting data and the reports lacked the necessary information that would help to
discern the relationship. For example the NOX and SO2 emission factors for batch plants
were based on less than six test reports. While there were less than 12 test reports
supporting the NOX and SO2 factors for drum plants, this was still not sufficient to discern
a statistical difference. In addition, few test reports include information on fuel usage and
detailed fuel analysis that would provide the additional information needed to establish a
correlation. Although there was insufficient data to establish a statistically based
correlation, separate emission factors were developed for gas and oil fired HMA plants.
An additional 30 Wisconsin test reports that were submitted which contain fuel usage and
fuel analyses. However, a statistical evaluation of these data was unable to discern any
strong statistical correlations between the available operating characteristics (fuel usage,
fuel sulfur content, process type and control device temperature) and SO2 emissions.
3.5.3 CAAP Factor Development Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
10. Page 13. paragraph 2: This paragraph is confusing. It is stated that only a single test
report was available for an oil-fired hot oil heater. Does this mean that only one hot
oil heater was sampled out of 364 plants and test reports? Finally, why does EPA go
out of their way Further, if most of such heaters are fired with natural gas, why was
an oil-fired heater sampled? to here to point out that "emission factors developed
from this data would not be representative of gas-fired heaters". While this comment
is correct, since oil fired burners emit more pollutants than gas-fired burners, there is
a consistent bias in such comments. The reports point out the possibility of a
high-bias each time such possibility exists, inviting the user of AP-42 to discount the
published number. But, when the reported emission factors have a low bias, there is
no such comment. For example, see the issue of the phony correction for
background emissions to truck loadout discussed later on the same page.
3.5.4 Response to CAAP Factor Development Comment Number 2 - As stated in the
paragraph, only one emission test of a hot oil heater was available. As indicated on page
4-17 of Appendix B (Emission Factor Documentation for AP-42 Section 11.1 Hot Mix
Asphalt Production), the hot oil heater test was conducted as part of a more
comprehensive test program to quantify trace metals and trace organic pollutants from the
combustion sources as required by California Air Resources Board (CARB) rules. The
title of the test report (see reference 35 on page 4-350 of Appendix B) indicates that the
CARB rule that the test was conducted to satisfy was the Air Toxics "Hot Spots"
Information and Assessment Act of 1987 (AB 2588). The test report did not provide
specific information on the rationale for testing the hot oil heater nor on the use of oil as
the fuel for the unit when propane was the fuel for the dryer. However, two goals of the
"Hot Spots" program were to collect emissions data and to identify facilities with
localized impacts (see http://www.arb.ca.gov/ab2588/ab2588.htm). As this test was
conducted for the California "Hot Spots" program, it likely that the oil fired tests were
performed to obtain a reasonable "worst case" scenario for possible "hot spot" evaluation.
77
-------
The comment with respect to the applicability of this emissions information to gas fired
facilities was made to place the information into the proper perspective. The primary
reason that there are no indications of low bias is that when information is not available
to make a reasonable unbiased evaluation, an assumption is made that provides a bias that
is protective of the public health. The objective is not to invite the user to discount the
published number but to use the information as a screening tool and to collect additional
data if the biased information results in concerns with respect to the potential
environmental impacts.
The correction for the measured background conditions created by the emissions from
diesel engines during the emissions test at plant C does not provide a positive bias to the
resulting data. As explained in item 51 beginning on page 56 of the Response to
Comments on Testing Program for Asphalt Plants C and D Report, the emissions
measured during the background run are consistent with other measurements of diesel
exhaust and a methodology was used that could be considered to provide a high bias for
load-out emissions.
3.5.5 CAAP Factor Development Comment Number 3 (Ravi Nadkarni & Lloyd Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
15. Page 11. paragraph 1: The industry's own MSDS sheets provide a molecular weight
of asphalt as "over 2000" ( a number we personally don't believe). It is refreshing to
see EPA use a lower number of 345.69
3.5.6 Response to CAAP Factor Development Comment Number 3 - The molecular
weight provided is not the molecular weight of the asphalt as specified in the MSDS's.
Data presented in the document SHRP Materials Reference Library: Asphalt Cements: A
Concise Data Compilation (SHRP-A-645; Strategic Highway Research Program;
National Research Council; Washington, DC; May 1993) indicates that the molecular
weight of asphalts from single crude oil sources ranges from 700 to 1300. As indicated in
Appendix C, the molecular weight of 345.69 is the average molecular weight of
tetracosane and pentacosane. These two compounds were single compound surrogates
that resulted in approximately the same vapor concentration at saturation as the complex
mixture emitted from the storage silos. The molecular weight of 345.69 is also not the
estimated average molecular weight of the emissions. The FTIR analysis of the silo vent
indicated that most of the hydrocarbons emitted were similar in spectra (and therefore
molecular structure) to hexane. Since the aliphatic hydrocarbons from about pentane to
nonane have similar spectra, the molecular weight of the vaporous emissions are probably
between 86 and 128. Therefore, to prevent additional misconceptions by users that the
specified molecular weight is either the average molecular weight of the liquid asphalt or
the head space vapors, the information presented on page 11 was revised to show the
estimated molecular weights of the liquid and vapor. This required the TANKS program
to be run to determine the appropriate surrogate compound to use to estimate the
appropriate Antoine's coefficients. The Antoine's coefficients for aliphatic hydrocarbons
that come the closest to producing a working loss emission estimate of 32 pounds per
million gallons of asphalt throughput for this range of molecular weights were docosane
78
-------
and tricosane. As a result, the background report was changed to describe this revised
methodology. The revised Antoine's coefficients are "A = 75350.06" and "B = 9.00346"
with a vapor molecular weight of 105 g/g-mole and are included in the background report
and AP-42 section.
3.5.7 CAAP Factor Development Comment Number 4 (Ravi Nadkarni & Lloyd Pillion) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
21. Table 7: Same problems as Table 1; i.e. use of Referencel which does not really
contain the primary data. Furthermore, there is no discussion why a batch plant dryer
using the same range of fuels e.g. natural gas, No. 2 fuel oil and other fuel oil and a
drum dryer using the same fuel oil should have different emission factors for NOX
and SO2.
3.5.8 Response to CAAP Factor Development Comment Number 4 - Reference 1 for this
part of the report should be Appendix B or the Emission Factor Documentation for
AP-42 Section 11.1, Hot Mix Asphalt Production. This reference citation was revised.
However, it was decided that since citing each of the over 150 references used to derive
these emission factors would add text that would confuse the typical user and potentially
hide information that adds value, only the background report would be referenced. We
agree that presenting a specific location in the footnote will assist a user find the
information needed to better understand the basis for the emission factor. To provide this
assistance, the table number in the background report was specified. A discussion on the
reasons why two different technologies would have different emission factors is not
germane for the development of emission factors. However, one of the primary reasons
for industries migration from batch plants to parallel flow drum mix plants to counterflow
drum mix plants is improved energy efficiency of the basic process. In addition, the
higher production rates for drum mix plants makes fuel efficiency more critical. As a
result of energy efficiency concerns, it seems reasonable that the CO2, and SO2 emission
factors for batch mix plants are slightly higher than for drum mix plants. It is also
realistic that the NOX emissions may be slightly higher from attempts to improve
combustion efficiency.
3.5.9 CAAP Factor Development Comment Number 5 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
14. Page 4-119. paragraph 1: Was there any attempt to correlate capture efficiency in
venturi scrubbers with the pressure drop?
3.5.10 Response to CAAP Factor Development Comment Number 5 - Information on the
scrubber pressure drop was not available for many of the facilities where venturi
scrubbers were used to control the particulate emissions. As a result, a statistical analysis
was performed only for drum mix facilities. This analysis attempted to correlate the
emission factor with the scrubber pressure drop. As stated on page 4-143 of Appendix B,
only the type of control device significantly affected filterable PM emissions. On this
79
-------
page it is explained that the lack of statistical power associated with small data sets and
the large variability in the data is one reason for not finding a statistical association.
3.5.11 CAAP Factor Development Comment Number 6 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
17. Page 4-126. paragraph 2:. This paragraph points out the inconsistencies in the
approach taken in several places in the report. Because there was no analysis of the
feed materials or of the fuel used, it is not clear whether the measured trace metals
are present on the site or are a result of sample contamination in the lab or elsewhere
or whether the detection of trace metals at the exhaust and non-detection at the
entrance to the control device represents random error in the measurements. The
approach of discarding inconvenient data is not acceptable.
3.5.12 Response to CAAP Factor Development Comment Number 6 - It is not clear how
this paragraph points out any inconsistencies in sampling or analysis. The cited reference
should be 340 rather than 356. Data were not discarded because they were inconvenient,
the data were discarded because they provided no useful information. Sufficient
information is provided to explain the reasons for discarding the data. Where the metal
was not detected, other more useful data are available. In the one situation where the
silver was higher at the outlet than was measured at the inlet of the control device, the
outlet data was judged to be superior since the volume of flue gas sampled was greater.
One would expect the control effectiveness to be greater than 90% as with other metals
measured at both the inlet and outlet, and the potential for matrix effects is less for the
inlet samples. An analysis of the feed material and the fuel used would not have altered
this result or provided information to explain the inability to measure some metals.
However, the differences in sample volumes and sample mass collected do provide an
explanation of the differences in the mass of metal that were detectable. Because of the
high particulate, loading at the inlet of the control device only about 50 cubic foot of flue
gas was sampled. At the outlet of the control device where the parti culate loading was
significantly lower, approximately 200 cubic foot of flue gas was sampled. This
difference in sample volumes is reasonable to detect metals at both the inlet and outlet. It
is unlikely that the measured trace metals are the result of sample contamination. As
provided in Chapter 6 (Quality Assurance/Quality Control Procedures and Results) in the
test report (which was provided to the commentors), of the sixteen metals analyzed, the
only metal that showed possible contamination was nickel.
3.5.13 CAAP Factor Development Comment Number 7 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
18. Page 4-135. last line: We are told that the statistical analysis is restricted to
references up to number 338. Why? Several of the subsequent references, up to
80
-------
reference 355, contain the same type of data as that being analyzed and should have
been incorporated.
3.5.14 Response to CAAP Factor Development Comment Number 7 - The statistical
analysis presented was performed in 1997 when the previous draft of the AP-42 section
and background report were circulated for review. Only fifteen additional emissions tests
of HMA kiln stacks were submitted in response to the draft section. Considering the over
300 source tests used in the statistical analysis, the lack of correlation that existed within
these data, the resources required to reanalyze the data and the high probability that the
results would be the same as for the earlier analysis, it was decided that the statistical
analysis would not be repeated.
3.5.15 CAAP Factor Development Comment Number 8 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
24. Page 4-139. paragraph 2: There is a discussion of attempts to correlate wet scrubber
performance with other parameters. Any engineering handbook will tell you that
scrubber performance will correlate with pressure drop, yet this measurement was
either not made or not used. As it now stands, this paragraph finds that although the
emission factor for venturi scrubbers is less than half that for other, unspecified
scrubbers, this result is not statistically significant.
3.5.16 Response to CAAP Factor Development Comment Number 8 - As stated
elsewhere, the vast majority of emission tests were obtained from the files of State and
local air pollution control agencies. It is recognized that with other parameters being
equal, scrubber performance will correlate with pressure drop. However, there are many
other design and operating parameters that can also affect the performance of wet
scrubbers. Some more important parameters are scrubber geometry, scrubber water flow,
mixing of scrubber water and flue gas, scrubber water solids content, flue gas cooling,
mist eliminator performance, inlet particulate concentration and inlet particle size
distribution. As explained in the paragraph, "the lack of statistical power associated with
the small data sets is the likely explanation for this result." The lack of statistical power
is caused primarily by the large variations in the data that may be attributable to
parameters other than differences in the pressure drops. The inability to statistically
differentiate the performance of these control devices can be seen on the box plots
(figures 4-2 and 4-7) where the middle 50% of the data overlap. For emission factor
development, statistical analyses provide additional support for engineering evaluations.
Although no statistical support was found for segregating scrubber and fabric filter
controls, the magnitude of the differences in the emission factors provides sufficient
support to provide separate emission factors for these two control types.
81
-------
3.5.17 CAAP Factor Development Comment Number 9 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
25. Page 4-141. paragraph 1: This entire paragraph is confusing. Figure 4.3 shows that
there is a relationship between RAP content and condensible inorganic PM. The
relationship appears to be non linear. Alternately, looking at just the 0% RAP data
points, it is clear that there is great variability in the condensible inorganic PM
emissions even without RAP. Without any analysis of feedstocks and fuels, the
source of these condensible inorganic PM.s can't be isolated. To add further
confusion, the last sentence jumps to a discussion of carbon dioxide emissions. The
reference to Figure 4.3 in context of CO2 emissions is incorrect.
3.5.18 Response to CAAP Factor Development Comment Number 9 - Although initially
one may observe a relationship between RAP content and condensible inorganic PM, as
the commentor points out that there is great variability in the condensible inorganic PM
emissions without RAP. A similar variability would be expected if more than a single
emission test were available for the higher RAP contents. As a result, when the data are
evaluated with this additional information, it is evident that any correlation developed
from this data would be highly speculative. In addition, when one considers that the
primary components of condensible inorganic particulate would be formed from sulfur (in
the fuel or in the asphalt) and nitrogen (from the fuel, air or asphalt), the possible
relationship may be related more to fuel usage than RAP content. While there may be a
correlation between RAP usage and fuel usage, other factors (such as moisture content of
the new aggregate, final mixing temperature and fuel efficiency) may have a more
dominant role in the fuel usage. The reference made to CO2 is a typographical error and
was changed from CO2 to RAP.
3.5.19 CAAP Factor Development Comment Number 10 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
27. Page 4-145. paragraph 1: The invocation of engineering principles in the fifth line
is interesting because they have been ignored in many instances up to this point; e.g.
discussion of scrubber efficiencies without consideration of the pressure drops.
VOCs, by definition, don't condense even when cooled to ambient temperatures.
Therefore, if a control device has an effect on VOC emissions, this attenuation must
be a result of a mechanism other than condensation, such as adsorption on solids.
The data quoted show lower emission factors (by a factor of 4) for fabric filters
compared to wet scrubbers, but the authors find this result surprising. Although they
don't tell us which engineering principle they are referring to, one has to assume that
they are referring to the ability of water to capture gases through dissolution.
Unfortunately, the components of VOCs are only sparingly soluble in water and the
pressure drops/retention times in such scrubbers are quite modest. Therefore, this is
a bad place to invoke such engineering principles when they contradict the data.
82
-------
3.5.20 Response to CAAP Factor Development Comment Number 10 - As stated
elsewhere, in the development of emission factors are generally supported by a known
engineering principle. Statistical analyses are used to characterize the significance of an
observed effect. Although the engineering principles are not specifically cited in all of
the summaries of the individual statistical analyses performed, they were the basis for
many evaluations. While water-based scrubbers are not generally installed at HMA
facilities to control VOC's, they may provide some incidental control. Air Pollution
Technology Fact Sheets and Technical Bulletins5 are available on the Clean Air
Technology Center (CATC) web site that provides information on the ability of various
scrubbers to control VOC's. Information on the ability of water-based scrubbers to
control VOC's are specifically mentioned in the fact sheets for Impingement-Plate/
Tray-Tower Scrubbers6, Packed-Bed/ Packed-Tower Wet Scrubbers7, Spray-Chamber/
Spray-Tower Wet Scrubbers8 and Venturi Scrubbers9. As indicated in the paragraph, one
would generally expect that the fabric filter to provide little or no control of VOC. It is
also stated in the paragraph that the data sets were small which would make any
difference suspect. Therefore, the paragraph states that data were not segregated to
develop the emission factors.
3.5.21 CAAP Factor Development Comment Number 11 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
28. Page 4-147. paragraph 2: The comment in the last line is that emissions are a
function of RAP content and production rate. Yet, earlier on page 4-139, these same
conclusions were rejected because of low correlation coefficients. Which conclusion
is correct?
3.5.22 Response to CAAP Factor Development Comment Number 11 - Both of the
statements cited by the commentor are correct. Statements on both pages indicate that
filterable PM emissions show an association with production rate and RAP content.
However, in the discussion on batch mix dryers on page 139, there is an additional
qualifying statement. The qualifying statement is, "However, the squared correlation
coefficient (R2) value for the model is 0.22, which indicates that the model explains only
a small percentage of the variability in the data."
5 Available by internet at URL http://www.epa.gov/ttn/catc/products.html#aptecfacts
6 Available by internet at URL http://www.epa.gov/ttn/catc/dirl/fimpinge.pdf
7 Available by internet at URL http://www.epa.gov/ttn/catc/dirl/fpack.pdf
8 Available by internet at URL http://www.epa.gov/ttn/catc/dirl/fsprytwr.pdf
9 Available by internet at URL http://www.epa.gov/ttn/catc/dirl/fventuri.pdf
83
-------
3.5.23 CAAP Factor Development Comment Number 12 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
31. Page 4-148. paragraph 2: The discussion should point out that Plant C is atypical of
drum plants in the industry. Figure 4.6 shows CO2 emissions factors for drum plants
as a function of production. Although there is considerable scatter, it is interesting to
note that all high emission points are associated with lower production and none are
associated with high production. Does this behavior extend to the emission of other
pollutants?
3.5.24 Response to CAAP Factor Development Comment Number 12 - As explained in
many meetings during the planning phase for testing of Plant C, and in the "Response to
Comments on Testing Programs for Asphalt Plants C and D" document, Plant C is typical
of drum mix plants with respect to all characteristics affecting emissions generation from
silo filling and truck loading operations. The method of loading hot mix asphalt into the
storage silos is identical to other hot mix asphalt plants. Additionally, the method of
loading hot mix asphalt into the transport trucks is identical to other hot mix plants.
While there are characteristics that differentiate this plant from drum mix plants in other
parts of the US, these characteristics relate more to the collection of emissions generated
during silo filling and truck loading, the maximum production capacity of the plant and
the ability of the plant to operate for a much larger percentage of the time than other
facilities. It should be noted that to test the batch mix plant (Plant D) EPA constructed an
enclosure and emissions collection system almost identical to Plant C.
With respect to the association between CO2 emission factor and production rate noted by
the commentors, the higher emission factors appear to be centered on about 180 tons per
hour rather than the lower production rates. However, this observation may be
misleading due to visual aspects of the data. The one or two high data points at 100 tons
per hour production and the four high data points at about 200 tons per hour may be
visually misleading since the remaining data do not show a strong relationship. The
summary of the statistical analysis (page 4-142) states "The linear model analysis
indicated that CO2 emissions can be estimated as a function of RAP content (p = 0.052),
production rate (p = 0.0002), and the RAP content-production rate cross-product (p =
0.043). However, the squared correlation coefficient ®2) value for the model is 0.23,
which indicates that the model explains only a small percentage of the variability in the
data." For most combustion sources, CO2 emissions are primarily a function of the fuel
usage and the carbon content of the fuel used. Differences in fuel efficiency, moisture
content of the aggregate and mixing temperature would also affect the CO2 emissions.
Other than the weak associations identified in the statistical analysis, no association
between production rate and emission factor was evident.
84
-------
3.5.25 Industry Factor Development Comment Number 1 (Gary Fore, Vice President
National Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-18 : B. Footnote D - There is a question that the range to 96 Ib/ton is too high.
Where is source for this number?
3.5.26 Response to Industry Factor Development Comment Number 1 - The results of
individual source test analyses that were used to develop the CO2 emission factor are
presented in Table 4-7 of the background report (Appendix B of the Emissions
Assessment Report). In this table near the bottom of page 4-272 is summary information
from Reference 347 which is a drain oil fueled facility that had measured CO2 emissions
of961b/ton.
3.5.27 Industry Factor Development Comment Number 2 (Gary Fore, Vice President
National Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-16: A. Footnote C -Suggest striking SO2 or add 50% attenuation of SO2.
Please review attached stoechiometry and SO2 attenuation data.
Page 11.1-18 : A. Footnote c -Suggest striking SO2 or add 50% attenuation of SO2.
Please review attached stoechiometry and SO2 attenuation data.
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
The Hot Mix Asphalt Industry is providing additional data specifically relevant to SO2
attenuation and benzene to EPA for possible inclusion into the Emissions Factors
Document (AP-42). Attached are three binders containing stack test reports. All of the
tests were consist of three 1-hour runs and were conducted for compliance purposes.
NOTE: NAPA is sending one copy of this material and one copy of the comments to Ron
Myers. Only the comments document is being sent to EPA Region I.
SO2
The Draft Emissions Assessment (EA) offers an EF for SO2 of 0.078 Ibs/ton hot mix
from drum mix plants firing waste oil (Draft EA,, Table 11.1-7). The Draft EA also
suggests via footnote (c.) of Table 11.1-7 that AP-42, Chapter 1 also be used to estimate
SO2. Chapter 1 Emission Factors (EF) incorporate fuel sulfur content to determine SO2
emissions.
It is believed that a more accurate approach should be used to determine SO2 emissions
from plants firing oil, especially waste oil.
Several stack tests for SO2 were conducted. Samples of each plant's burner fuel supply
during each test were taken and analyzed for sulfur content. The amount of SO2
85
-------
produced during combustion of the burner fuel was calculated. There was an assumption
that all of the sulfur in the fuel would be converted to SO2. Then, using a mass balance
approach, the actual stack emissions of SO2 was compared with the calculated SO2 as a
product of combustion, and determined a control (i.e., emission reduction) factor for SO2,
from a hot mix asphalt plant, firing oil.
The proposed means of control would be the adsorption of SO2 by dust particles
suspended in the ductwork and captured on the bags in the baghouse. For a plant
producing 250 tons/hour hot mix and firing 500 gallons/hour waste oil having a sulfur
content of 0.046% by weight and a specific gravity of 0.89, the amount of SO2 generated
by combustion would be 500 gal/hr x 8.34 Ibs/gal x 0.89 x 0.46/100 x 2 (Ib SO2/ Ib S) =
34.14 Ib SO2/hr. At the same time, the uncontrolled particulate emissions from the plant
would be 28 Ibs PM/ton (Draft Table 11.1-3) x 250 tons/hour = 7000 Ibs PM/hour. The
available reaction surface area of the dust particulate provides significant acid gas capture
opportunity by any alkaline solid material suspended in the exhaust stream or
accumulated on the bag house bags.
The results from the 15 tests demonstrate SO2 control ranging from 47% to 97%, with an
average control of 62%. Test result averages show little variation in control among
different plant types, aggregate types or recycle asphalt pavement (RAP) content in the
mix. Please refer to the attached table: "Asphalt Plant SO2 Emissions Summary"
The Wisconsin Department of Natural Resources (WDNR) has reviewed these stack tests
and will adopt a conservative control factor of 50% for future permitting and emission
inventory purposes. Please refer to the attached letter from Lloyd Eagan, Director of
Bureau of Air Management from the State of Wisconsin - Department of Natural
Resource.
3.5.28 Response to Industry Factor Development Comment Number 2 - The emission
tests that were submitted were evaluated and added to the test information in the draft
background report (Appendix B). The measured emissions were used to develop a
revised SO2 emission factor of 0.058 Ib/ton for drain oil/waste oil/No. 6 fuel oil-fired
dryers. The additional emission test data also included fuel usage and fuel sulfur content.
This data was analyzed to discern a relationship between measured emissions, the results
of sulfur mass balances (as SO2) and several available operating parameters. Like the
commentors, EPA could not find a correlation with the stated operating parameters.
While EPA found no reasonable correlation, we did verify that control levels for SO2
ranged from 47% to 92% and averaged 62%. Additionally, the available data indicate
that a maximum of 0.1 Ib/ton of SO2 is retained in the product (which includes recycled
baghouse dust). This analysis was described in more detail in the final background
report. Based upon this analysis, we have added the following sentence to footnote c of
Table 11.1-7: "50 percent of the fuel-bound sulfur, up to a maximum (as SO2) of 0.05
kg/Mg (0.1 Ib/ton) of product, is expected to be retained in the product, with the
remainder emitted as SO2."
86
-------
3.5.29 Industry Factor Development Comment Number 3 (Gary Fore, Vice President
National Asphalt Pavement Association) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
The Hot Mix Asphalt Industry is providing additional data specifically relevant to SO2
attenuation and benzene to EPA for possible inclusion into the Emissions Factors
Document (AP-42). Attached are three binders containing stack test reports. All of the
tests were consist of three 1-hour runs and were conducted for compliance purposes.
NOTE: NAPA is sending one copy of this material and one copy of the comments to Ron
Myers. Only the comments document is being sent to EPA Region I.
Benzene
Stack tests were also conducted for benzene from drum mix plants, firing either waste oil
or natural gas. The reports from 17 tests are enclosed. These reports indicate average
benzene emissions of approximately 0.0003 Ibs benzene/ton hot mix asphalt, with no
significant variation for fuel type. The draft EA lists a benzene EF of 0.00051 Ibs/ton
(Table 11.1-10).
3.5.30 Response to Industry Factor Development Comment Number 3 - The emission
tests that were submitted were evaluated and added to the background report (Appendix
B). As with the benzene data in the draft report, the fuel type did not appear to affect
emissions. Therefore all 19 sets of data were averaged to produce a single emission
factor for all fuel types. The background document section and table describing the
development of benzene emission factors for drum mix plants were revised. The
resulting benzene emission factor of 0.00040 Ib/ton replaced the draft benzene emission
factor in Table 11.1-10.
3.5.31 CAAP Factor Development Comment Number 13 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
37. Page 4-153. last paragraph: In the 10th line, there is an unnecessary caveat about the
relationship being uncertain. The simplest assumption is one of proportionality.
Unless lab scale data are collected that indicate otherwise, this is the best way to deal
with the issue. Are you forgetting Occam's Razor?
3.5.32 Response to CAAP Factor Development Comment Number 13 - The simplest
assumption that emissions are proportional to the results obtained from the RTFOT is
being made. However, future laboratory and field data may reveal that the loss in mass
obtained by the RTFOT analysis may not explain all the emissions from the asphalt.
Specifically, there are some asphalts which lose very little weight or gain weight during
this analysis. It is unlikely that these asphalts do not have any emissions or remove
pollutants from the air. While the commentors may believe that this caveat is
87
-------
unnecessary, it is provided because there have been only two emission tests of fairly
similar asphalt (with respect to the RTFOT analysis) to verify the relationship.
3.5.33 CAAP Factor Development Comment Number 14 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
38. Page 4-154. paragraphs 1-2: These paragraphs repeat industry propaganda (along
with that from State highway department laboratories who share the same
road-building mind set). The specifications allow for measurements of unblended
asphalt and the information represents the results of tests on unblended asphalt. In
the Boston area, the terminal in Chelsea has a tank holding such additives for
blending with the asphalt. After a meeting at EPA-Region 1, where the industry
representatives had denied that any blending took place, Mr. Frank Singleton, the
Board of Health Agent for Chelsea showed me this tank which is clearly labeled as
containing an asphalt additive. The citizens submitted a list of many such additives
from Roads and Bridges magazine to the EPA/citizen group which was ignored by
the EPA Project Manager, who instead chooses to believe the asphalt industry's
lobbyists who insist that such additives are not used. Also, we have examined the
application from one of the terminals in Chelsea which said that they proposed to
add these additives directly to the tanker truck whereby they would mix in the truck
while being transported to the user. This would mean that a sample taken at the
terminal would be a sample of unmixed asphalt.
The third paragraph needs to be deleted. As long as the specifications allow for up to
1% volatiles in asphalt (even before additives are used), it is misleading for EPA to
state that their approach in Table 4-27 "encourages the use of site specific data". The
primary use of AP-42 is to prepare calculations to get permits to build and to operate
hot mix plants. Therefore, there is no site specific data that can be used other than
fictional data. Unfortunately, the industry has a long history of providing low-ball
estimates, designed to fool the public and get permits from agencies that are not
particularly anxious to ask difficult questions.
3.5.34 Response to CAAP Factor Development Comment Number 14 - The information
in these paragraphs is not from industry but are results of preliminary laboratory analyses
performed as part of a broader quality control program for State and Federal road
construction. The State Departments of Transportation laboratories perform a quality
control function to verify that the pavement being purchased (directly or indirectly
through a contractor) meets the specifications for the road being constructed. The
RTFOT analysis is only a preparatory step for the rheological analyses that determine the
predicted performance of the asphalt. It is recognized that there are a variety of additives
which are designed to be added to blended asphalt to improve the physical characteristics.
It is also recognized that to meet some physical characteristics of some more recent
performance graded asphalts, the use of additives is necessary. However, the contacts at
the State Department of Transportation laboratories stated that they analyze samples of
88
-------
asphalts that are to be used without further blending. If an asphalt distributor or a hot mix
asphalt production plant adds a diluent or other additive that changes the physical
characteristics, it is possible that they could be considered to be violating the State
contract specifications.
In addition, no test information has been provided which would support the contention
that asphalt distributors modify the asphalt provided to hot mix asphalt plants following
the collection of samples for the State Department of Transportation analysis. In fact, the
samples obtained during the pre-test survey and the emission test at Plant D in
Massachusetts were within 2% of the average volatility for asphalts analyzed by the
Massachusetts Department of Transportation laboratory. Also as stated previously, State
and local departments of environment have the option of including enforceable
requirements in the operating permits which limit the volatility of the asphalts used by the
hot mix asphalt plants. If there are concerns that asphalt distribution terminals or hot mix
asphalt plants are altering the asphalt, the collection and analysis of the final product by
any authorized State or local transportation, environmental or health inspector seems in
order.
The paragraph discussing the use of site specific information does not need to be deleted.
There is ample information that the volatility of asphalts can vary both between and
within States. The available information does not support the use of 1% as the only
volatility value to use. There is also ample information that independent information on
the volatility of asphalts used within States are readily available. It is not misleading to
allow State and local agencies to use this readily available and independently collected
data as a locally specific parameter for calculating emissions. The primary use of AP-42
is not to prepare calculations to get permits to build and operate plants. The primary use
of AP-42 is to calculate unbiased actual emissions estimates for use in understanding and
managing the combined emissions from all sources in a geographic area of interest.
While AP-42 is used by industries and State agency reviewers to calculate emissions for
various permits, these emissions are generally maximum limitations that are not to be
exceeded by the facility. Although some facilities may provide low ball estimates in their
permits for a variety of reasons, these values in permits are not estimates but are
enforceable limitations. When the permit is issued, the facility must operate within this
limitation and the company official signing the permit is held accountable.
3.5.35 CAAP Factor Development Comment Number 15 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
39. Page 4-155. paragraph 2: On the 13th line, the statement is that "this hypothesis has
not been validated by emissions testing but provides an adjustment that is
directionally correct." This comment is totally unnecessary and needs to be deleted.
Is EPA saying that the Clausius-Clapeyron equation should be thrown out?
Alternately, is EPA saying that the degree of precision and accuracy associated with
emissions testing methodology is good enough to provide this type of data?
89
-------
3.5.36 Response to CAAP Factor Development Comment Number 15 - The EPA is
neither saying that the Clausius-Clapeyron equation nor that the precision and accuracy
associated with emissions testing are in question. However, the relationship between
temperature and emissions has been derived from a standard laboratory test that measures
the loss of weight of the asphalt during heating and not on a measurement of emissions
during this test. It is expected that there will be some differences in the mass emissions
and loss of weight due to the uptake of oxygen or nitrogen by the asphalt and result in
some small weight gain. It is believed that this phenomenon is at least one explanation
for the few asphalts which have demonstrated a weight gain during the test. EPA
obtained the temperature relationship for asphalts by performing three laboratory tests.
While one of these tests was the standard laboratory test, the other two were at
temperatures 25 ° F above and below the standard temperature. Asphalts that do not lose
weight or that gain weight during the standard laboratory test, would have weight losses
that are incompatible for developing the proper constants for the Clausius-Clapeyron
equation.
3.5.37 CAAP Factor Development Comment Number 16 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
40. Page 4-157. paragraph 2: In this paragraph, the groundwork is being laid to explain
why inorganic PMs were much higher at Plant D than Plant C. This is further
elaborated on page 158. We strongly disagree with the conclusion that the additional
dust was a result of truck movements. While Plant D was unpaved, water sprays
were used to keep the dust down as noted in the report and as observed. At Plant C,
because of the practice of cleaning the silos at the end of the day with crushed stone,
there was considerable dust present in the tunnel. Furthermore, one ofthe silos had
dumped hot mix on the floor and cleaning it up generated a lot of dust. Our
recollection is that if anything, the tunnel at Plant C was dustier than the one at Plant
D. We believe the explanation for the higher PM load at Plant D is the relatively
short residence time in the mixer at any batch plant. Because of this, during mixing,
there are small pockets of stone dust that are not coated with asphalt and these are
the source ofthe high PM readings. In contrast, the vapors inside a silo have time to
permeate through the mass of hot mix in the silo and coat any uncoated dust
particles. Once coated, the fine dust coagulates and is no longer airborne particulate
matter. This hypothesis is supported by the visual observation that dumps from the
mixer at Plant D were dustier than those at Plant C when observed from inside the
tunnel.
3.5.38 Response to CAAP Factor Development Comment Number 16 - The justification
presented in paragraph 2 and supported by the presentation on page 158 are valid and the
conclusion is reasonable. It is recognized that the unpaved road at Plant D was watered to
minimize the dust created by truck movement. In first full paragraph on page 4-158 of
the draft background report, an assumed surface moisture content of 15 percent was used
to calculate the road dust emissions. This moisture content was used because it is
90
-------
representative of a highly watered unpaved gravel road. At Plant C, emission testing was
completed by the end of the day and the tests did not include the dust created by the silo
cleaning with crushed stone. Additionally, emission testing was stopped temporarily
during the clean up of the hot mix accidently dropped on the floor of the enclosure.
Although the suspended paniculate within the Plant C enclosure appeared greater than
within the Plant D enclosure, this increased suspended particulate was not due to
suspended road dust as purported by the commentors. At Plant C, the load-out emission
rate was much greater than at Plant D making the collection and removal by the
ventilation system more difficult. First, at Plant C, the production rate averaged four
times the production rate of Plant D with some periods of high production being six to
seven times greater. Second, the volatility of the asphalt used at Plant C was about 40%
greater than at Plant D. Third, the ventilation system at Plant C captured an estimated 60
to 70 percent of the emissions compared to essentially 100 percent capture at Plant D. In
addition, the suspended particulate at Plant C was more visible since the enclosure was
open on both ends thereby presenting a background more conducive to seeing the
suspended particulate.
The hypothesis which the commentors contended explain the high inorganic PM
measured at Plant D is remote. While there may be a small probability that small pockets
of uncoated stone dust may periodically occur during batch mix operations, the one
minute of vigorous agitation in the pug mill makes this highly unlikely. Additionally, the
coating process for both batch and drum mix plants depends on the significance of this
agitation and mixing not on additional coating in the silo due to the availability of vapors.
Since the aggregate is also heated to approximately the same temperature as the asphalt,
vapors inside the silo would not condense on any uncoated dust particles. Moreover, dust
emissions even from highly watered unpaved roads have been measured and can account
for a significant portion of the inorganic PM measured at Plant D.
3.5.39 CAAP Factor Development Comment Number 17 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
41. Page 4-158. paragraph 2. 3: See the comments above. The adjustment is indeed
"speculative" and should be eliminated. Separate emissions factors should be used
for inorganic PM emissions from batch and drum plants.
3.5.40 Response to CAAP Factor Development Comment Number 17 - The equation
used to adjust for background dust emissions from the unpaved road is not speculative
but is based upon emissions measured at other sources. Since the basis of the unpaved
roads emission factor equation is over 180 emissions tests we believe that it is a
reasonably reliable indicator of unpaved road emissions.
91
-------
3.5.41 CAAP Factor Development Comment Number 18 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
42. Page 4-159. paragraph 2. 3: There is a fundamental problem with the two paragraphs
which has not been explained. MCEM PMs and TOC are generated the same way
and from the same source. Therefore, if MCEM generation is a function of loadout
time, TOC generation should also be a function of loadout time. Of course, the
comparison is further complicated and made meaningless by the bogus "background"
correction discussed earlier. To eliminate the confusion, separate emission factors
should be prepared for batch and drum plants in all cases.
3.5.42 Response to CAAP Factor Development Comment Number 18 - The commentor
is correct in that both MCEM PM and TOC emissions are generated as a result of organic
material vaporizing from the asphalt binder. While the more volatile organic material
remains a vapor and is measured as TOC, the less volatile material condenses to become
a particulate and is measured as MCEM PM. The sum of the TOC and MCEM PM is the
total organic emissions from the load-out operation and should be compared to determine
whether drum mix and batch plants should have separate emission factors. Table 4-32
presents the emissions adjusted to a consistent asphalt volatility and temperature. The
sum of the MCEM PM and the TOC load-out emissions presented in this table are
0.00421 (4.21 e-3) Ib/ton for Plant C and 0.00477 (4.77 e-3) Ib/ton for Plant D. The Plant
D emissions are only 13% greater than Plant C. First, this agreement is better than most
of the data supporting the kiln stack emission factors. Second, this level of agreement is
remarkable given that the data have been adjusted to account for differences in measured
load-out temperature and asphalt volatility. Given that the load-out times at Plant D were
from six to eight times longer than at Plant C one would expect that differences in the
total organic emissions would be close to this ratio. However, the emissions are very
nearly the same and well within the uncertainties of the adjustments performed and the
precision of the measurement methods. Therefore, there is insufficient data to conclude
that emissions are a function of the load-out time. By presenting a single set of load-out
equations for batch plants and drum mix plants, there should be no confusion on the
applicability of the equation. As a result, the two paragraphs were revised as follows:
"The next most significant difference in emissions between Plant C and D is the MCEM
PM. The MCEM PM from Plant D is slightly over four times the emissions from Plant
C. This difference could be explained by the longer time required to complete the
load-out operations at batch plants compared to drum-mix plants and other test-specific
factors. However, the asphalt dependent mechanism that generates emissions of MCEM
PM and TOC is the same for both pollutants. This volatilization should cause similar
MCEM PM and TOC load-out emissions after adjustments for asphalt volatility and
temperature. Both emissions are the result of vaporization of organic material from the
asphalt binder. The more volatile organic material remains a vapor and is measured by
Method 25A and is generally referred to as TOC. The less volatile organic material
condenses into an aerosol and is measured by Method 315 and is referred to as MCEM
92
-------
PM. When summed, the TOC and MCEM PM emissions from Plant D are only 13
percent higher than the TOC and MCEM PM emissions from Plant C. Given the
variations in the run-by-run data, the low number of runs and the uncertainty in adjusting
emissions to a consistent temperature and volatility, the difference is not significant.
Therefore, for the purposes of developing emission factors for load-out operations, both
the MCEM PM and TOC data from Plant C and Plant D were averaged and an equation
that represents the averaged data was developed."
3.5.43 CAAP Factor Development Comment Number 19 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Hot Mix Asphalt Plants Emissions Assessment Report
15. Page 15. paragraph 2: It is ironic that EPA, after rejecting our repeated pleas for
collecting some fundamental data on asphalts such as its vapor pressure, states
"Although vapor pressure information on paving asphalt is not available ". A
more appropriate way to state what has happened would be to state that vapor
pressure measurements are not available because of EPA's refusal to measure the
vapor pressure.
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
44. Page 4-163. paragraph 1: The first line states that "vapor pressure information .... is
not available". The report fails to state that this was only because of EPA's refusal
to collect this information.
3.5.44 Response to CAAP Factor Development Comment Number 19 - Vapor pressure
information is not available for paving asphalt because there is no established
methodology to measure vapor pressures for this type of material. First, the collection of
vapor pressure information would have required the development of a suitable test
method to measure the vapor pressure of this material. While within the last year, the
Lawrence Berkeley Laboratory has submitted a procedure for determining vapor pressures
of heavy crude oils, this method was not available at the time the HMA project was
started. To develop and validate their method required over a year of effort. This method
requires quantification of almost all of the individual organic species in the vapor space.
This method has not been attempted for hydrocarbons that are more volatile than the
heavy California crude oils their method addresses. Although no acceptable standard
methodology to measure the vapor pressure of asphalt is available, determining the
maximum head space concentration of organic compounds above an organic substance is
suitable for use in the TANKS equations. The use of this concentration provides an
alternative to developing a laboratory method to determine vapor pressure for the
complex mixtures of the compounds and mixtures requiring elevated storage
temperatures.
93
-------
3.5.45 CAAP Factor Development Comment Number 20 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
45. Page 4-163. paragraph 3-5: Please explain what you mean by "pegged TOC
readings"? As it stands we can't understand this at all. Furthermore, Figures 4-9
through 4-13 are missing from our copy. Finally, it is not clear whether you are
referring to Plant C or Plant D in this discussion, though the context suggests
Plant C.
3.5.46 Response to CAAP Factor Development Comment Number 20 - The phrase
"pegged" TOC readings means that the emissions being measured exceeded the
maximum concentration of 1,000 ppm that the instrument was capable of measuring. The
paragraphs explain how an estimate of the maximum TOC concentration was
extrapolated from the measured data converging on 1,000 ppm. The paragraph and the
figure cited describe how the lines of increasing and decreasing concentration on both
sides of the plateau of 1,000 ppm were extended to determine the concentration where
these lines crossed. In addition, the paragraphs explain by comparison to other similar
situations which were within the instruments measurement capability why the actual
maximum TOC concentration is likely to be less than was estimated by this extrapolation.
The paragraph was edited to make the discussion more clear and to reinforce the source
of the emission information.
All of the figures the commentors state were missing from their copy are included in the
version available from the hot mix asphalt web page (www.epa.gov/ttn/emc/asphalt.html)
and on the master use to produce the approximately 50 CD-ROMs that were mailed to
individuals that requested the reports in this format (including the CD-ROM mailed to
Ravi Nadkarni on July 24, 2000). If the commentors would have informed EPA of
missing pages from the few paper versions mailed to stakeholders, we could have
supplied the missing pages or informed the commentors on how to print them from the
CD-ROM version.
3.5.47 CAAP Factor Development Comment Number 21 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
45. Page 4-163. paragraph 3-5: We further find that the contention that the estimate of
unmeasured emissions being an "upper bound" to be unpersuasive and incorrect.
Our reasons for saying so result from the inadequacies of Method 204. These were
outlined in Ravi Nadkarni's letter of September 15, 1999. They are:
a. At Plant C, two precautions were taken to ensure that we could measure or
estimate the amount of organic fugitive emissions that did not reach the primary
sampling ports. The first was the use of sulfur hexafluoride tracer gas, and the
second was the use of impingement/deposition plates in the ducts which collected
94
-------
organic particulates which would be deposited upstream of the sample port.
Although all the emissions did not reach the sampling ports (some was deposited
inside the tunnel and some was pumped out of the tunnel by ambient breeze or by
truck movement), the tracer gas enabled an estimation the VOC portion of this
loss. (As was pointed out during the planning, the tracer gas was a good proxy
only for non-condensible emissions but not for condensible emissions.)
At Plant D, a tracer gas was not used although requested by the citizens because
the TIE was constructed to meet Method 204 criteria, yet visual observations and
THC readings showed the following:
b. An ambient breeze was sufficient to cause some of the fume to escape from the
top or the bottom openings in the downwind door of the tunnel. Therefore the
size of the openings was decreased further. This means that the Method 204
criterion of maintaining over 200 fpm at a natural draft opening is not adequate.
Note that 200 fpm is less than 2.3 miles per hour, which is not much of a breeze.
Therefore it is not surprising that these criteria are inadequate to assure total
containment within the TTE. We observed visible emissions of fume from the
entrance, further emphasizing that the Method 204 criteria are inadequate.
c. Further, Method 204 has a more serious and fundamental problem. Although the
size of natural draft openings is specified, the method does not contain any criteria
to ensure that the emissions are pulled/sucked past the sampling point in a
reasonable period of time. As is well-known, residence time calculations, based
on enclosure volume and fan capacity, generally provide incorrect information by
underestimating the time to evacuate the enclosure because of channeling. Note
that Method 204 does not require a specific location for the natural draft openings,
a problem that was pointed out during the planning sessions. At Plant D, even
after the size of the opening was decreased, and the little pieces of colored tape at
the Natural Draft Openings were indicating airflow towards the inside of the TTE,
the tunnel was not evacuated in the approximately 15 second time gap between
the final drop and the opening of the doors. This can be clearly seen in the fact
that THC readings did not drop to zero even when the sample averaging time is
less than 1 minute. This fact was also observed visually inside the TTE enclosure
since SVOCs are visible. Thus, in the absence of a tracer gas, we have
compelling evidence that the collected sample had a low bias. For example, the
extended period test results in MRI-D-Table 4-4, show that final concentration did
not drop to zero but stayed between 1 ppm and 2.1 ppm . Yet, this low bias is not
mentioned anywhere in the report nor in Appendix B of the MRI Report which
contains the original data. These emissions did not reach the instrumentation but
were emitted to the environment causing a low bias.
3.5.48 Response to CAAP Factor Development Comment Number 21 - The paragraph
three to five on page 4-163 pertains only to the methodology used to estimate a maximum
TOC concentration from data collected during silo filling operations at Plant C and the
95
-------
rationale why this methodology would lead to an "upper bound estimate of this
concentration. It does not discuss any measurements at Plant D. However, since the
commentor makes specific statements related to the ability of the enclosure constructed at
Plant D to capture essentially all of the emissions, a response to the remainder of these
statements is provided in Section 3.1.2.
3.5.49 CAAP Factor Development Comment Number 22 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
46. Page 4-164. paragraph 2: There is reason to believe that the numbers derived are low.
We base this on several facts: Material Safety Data Sheets contain the following
warning: "Studies have shown that low flash point substances, such as hydrogen
sulfide and low-boiling hydrocarbons, may accumulate in the vapor space of hot
asphalt tanks and bulk transport compartments. Such vapors may exhibit
flammability characteristics of a significantly lower flash product than would be
indicated by the open cup test." A review of light hydrocarbon compound properties
in the Handbook of Chemistry and Physics indicates that the lower explosive limit is
reached at vapor concentrations around 1%. Therefore, this warning makes sense
only if the TOC concentration is around 10,000 ppm, not the 2000 ppm level used.
Second, storage tank temperatures are generally higher than 325° F and closer to 350°
F. Finally, the assumed vapor pressure is low in comparison to data from Nelson's
Petroleum Refinery Engineering, which would indicate a vapor pressure over 10 mm
Hg.
3.5.50 Response to CAAP Factor Development Comment Number 22 - It is assumed that
the commentor is directing the statements to not only the paragraph stated but to all of
section "4.4.2.5 Storage Tank Emissions." The warning cited by the commentor from an
asphalt suppliers Material Safety Data Sheet (MSDS) could not be found in other
MSDS's from other asphalt suppliers. However, neither this warning nor the
concentration correlating with the lower explosive limit support using a vapor
concentration of around 10,000 ppm. First, the open cup test method yields flash points
that are from 10 to 25% higher than the closed cup test methods. For asphalts, this flash
point difference could be more than 100 °F. While the flash point supplied in the MSDS
should use the results obtained from the Pensky-Martens Closed Tester, the statement
provides additional insurance that a user would not rely on other flash point information.
For a closed liquid storage tank, the head space vapor would be at the lower explosive
limit when the tank temperature is at the closed cup flash point. Measured flash points
are available in a report that was previously provided to the commentors. The range of
flash points for 86 asphalts analyzed by the Asphalt Institute (Properties of Asphalt
Cements, V. P. Puzinauskas, Asphalt Institute Research Report No. 80-2, Asphalt
Institute, November 1980) was from 450 °F to 685 °F.
The derivation of the information presented in this paragraph was explained on page
4-162 through 4-163 of the draft report. As explained on these pages the emissions for
96
-------
storage tanks are derived from estimates of the maximum THC concentrations measured
during the silo filling operations at Plant C. The validity of the derivation of these
emissions depends mostly on the assumption that when operated at the same temperature,
the estimated maximum THC concentration observed from the silo filling is the
maximum THC concentration in the head space of asphalt storage tanks. At the meeting
held in Boston on January 20, 2000, Dr. Nadkarni and others appeared to agree that it was
reasonable to assume that the vapors in the exhaust from the silo were saturated. As
stated by the commentor, we estimated that this maximum concentration was 2,000 ppm.
At this concentration, emissions from an asphalt storage tank maintained at an average
temperature of 325 °F would be 32 Ib per million gallons of asphalt throughput.
If the average storage tank temperature was raised to 450 °F, the emissions would be
increased to 570 Ib per million gallons of asphalt throughput. Since these emissions are
the result of the venting of the vapors due to the periodic displacement by liquid asphalt
(working losses), the head space vapor concentration for the higher temperature tank can
be calculated. The head space vapor concentration of the storage tank maintained at 450
°F would be 35,600 ppm. Given that the FTIR analysis of the silo exhaust indicated that
most of the vapors were likely to be aliphatic hydrocarbons between pentane and nonane,
the lower explosive limit concentration of 1% estimated appears reasonable. For the head
space vapor concentration to be 10,000 ppm, the average temperature of the asphalt
storage tank would need to be only 390 °F. As a result, the statement in the MSDS may
also be a warning about the relative flammability of the head space vapors compared to
the flash point. Therefore, the information derived in this section supports the statements
in the MSDS.
Many safety and health related requirements are instituted at a percentage of the lower
explosive limit rather than at the lower explosive limit. For example EPA safety
requirements preclude entry into an environment that is above 10% of the lower explosive
limit. As a result, the commentors inference that the vapor concentration in the asphalt
storage tank is at or above the lower explosive limit is inappropriate. Therefore, the
information presented in the paragraph does not contradict the information in the Material
Safety Data Sheets cited by the commentors.
The information presented in the paragraph also does not contradict the information in
Nelson's Petroleum Refinery Engineering. The commentors statement concerning the
vapor pressure of 10 mm Hg neither specifies the boiling point for the hydrocarbon nor
the temperature corresponding to the vapor pressure. To have a vapor pressure of 10mm
Hg at a temperature of 325 °F, the boiling point of the hydrocarbon would have to be
about 600 °F (See Point "A" on Figure R-l which is adapted from Petroleum Refinery
Engineering, Fourth Edition, Wilbur L. Nelson, McGraw-Hill, 1958). In a review of the
physical properties of asphalts (CONCAWE: Bitumens and Bitumen Derivatives
(92/104), CONCAWE, Brussels (December 1992)}, the range of boiling points for
asphalts is from greater than 400 °C (752 °F) to greater than 550 °C (1,022 °F). Based
upon the boiling point data presented in the CONCAWE review and the vapor pressure
relationships presented in Petroleum Refinery Engineering, the vapor pressure for an
97
-------
asphalt at 325 °F with a boiling point of 900 °F (about the midpoint of the range) would
be less than 0.01 mm Hg. The range of vapor pressures at 325 °F corresponding to the
extremes of the boiling points would be from about 0.0002 mm Hg to about 0.2 mm Hg.
These vapor pressure extremes are shown as Points "B" and "C" on Figure R-l. The
Asphalt Institute (http://www.asphaltinstitute.org/faq/acbfaqs.htm) states "It is estimated
that at a typical inventory temperature of 325° C, the vapor pressure of petroleum asphalt
is less than 0.01 psia (1.5e-3 kPa)." A vapor pressure of 0.01 psia is equal to 0.52 mm
Hg. In addition, EPA confirmed with the Asphalt Institute that the stated temperature is a
typographical error and should be 325 °F. EPA does not know what type of asphalt the
Asphalt Institutes estimate of vapor pressure is based upon nor what methodology was
used to estimate the vapor pressure. However, the vapor pressure of 0.95 mm Hg
estimated by the silo filling data is reasonably consistent with these other estimates of
vapor pressure for asphalt.
The purpose of the commenters statement that the storage tank temperatures are generally
greater than 325 °F and closer to 350 °F is unclear. There is nothing in the paragraph
that indicates the typical storage tank temperature. The storage tank temperature used in
the paragraph was the recorded temperature for the material being loaded into the silo on
July 27, 1998 at 7:36 when the peak concentration was observed. The same temperature
is used to provide consistency in the concentration of the vapors emitted for both
emission sources to estimate the most appropriate Antoine's constants representing the
asphalt. The use of a storage tank temperature of 350 °F at this point would generate
Antoine's constants that represent a significantly less volatile substance than is indicated
by the data. In addition, the commentors statement is in contrast to the industry
recommendations for asphalt storage temperature (Asphalt Pavement Environmental Best
Practices, Asphalt Pavement Environmental Council, April 2000) as previously shown in
Figure 2. The industry guidance provide different recommended storage temperatures for
different performance graded asphalts. The highest midpoint storage temperature
recommended is 325 °F for PG82-22 asphalt. The numbers in the grade are indications
of the project-specific temperature extremes (in degrees centigrade) for which the asphalt
mixture is being designed. As such, a PG82-22 grade asphalt is intended for use when
average 7-day maximum pavement design temperature is 82 °C (179 °F) and the
minimum pavement design temperature is 22 °C (-8 °F). Of the asphalts included in the
recommendations, this asphalt grade is for use in a severe application where the
maximum design temperature is the highest.
98
-------
206 PETROLEUM REFINERY ENGINEERING
50 IOO ISO 200 ,300
600
400
300
200
100
60
40
§ 20
o
c
u
* .0
fe 8
R PRESSURE, mm
3
D — ro 01 * 0)
6 0.6
>
0.4
03
0.2
0.1
0.08
0.06
0.04
0.03
0.02
0.01
0.008
0.006
0.004
0.003
O.OO2
0.001
0.0008
0.0006
0.0004
0.0003
0.0002
0.0001
s
Fio. 5-
Co.)
VAPOR PRESSURE OF
; HYDROCARBONS
/
/
'
This chart applies to stocks having a
characterization factor "K"of 12.0
ESS
,
•/—
/
/
/
0 R
J
/
7
-^
/
/
ESI
/
f
/
/
—7
/
/
/
:AR
,
/
/
f
/
f
'
CH
§
-/
^
'
^
/""
'
/
AN
f
/
/
J
/
/
D
f
f
/
/
'
/
f
/
EN
/
/
-4
/
/
/
/
/
31 N
/
/
'
^
$
/
/
-f-
f
'
/
EE
7
/
/
}
/
'
-j
/
/
:R
y
/
/
i_
7
f
>
f
t
\
NC
uc
X
/
_/
^
'
f
/
'
f
c
r
•/•
,
/
u
*
/
/
/
o^
f
f
/
'
M
«<
j.
f
>
f
/
1PAN
/
/;
/-7
/ ^
/
y-
/ /
/ u
fl
^-/
/
7
7
Y 19
/
/ r
f/
t
' /
/ f
>/
•/—
t
/ 1
' 1
ft
/]
55
•f-
/
f
f
'/
•f-
/
' /
t
P
/-
1
' f
7
/
/
* /
/ /
>
&
/
A
f
•/•
-/•
/
/ f
i
i
/
/;/
/ /
/ j
y
f / 1
'A
-f=ff
t\
1
f /
f
, 1
(•\
"I
S*Z
/ (
//
/I
/
/
/
/
y
f
/
/
/
/
^
I
4-
t
1
1
'I,
-j-
J\
t
/ (
/
/,
'/
/
-f
/
/ ,
/f
/;
/
/ /
/
/
/
-/-
t
n
ft
/
i
.Q>
* $1
'///
40O
~f
i
f
/
^
/
/
/
j
tj
/
(
f
1
k
f
1
f
1
f
f
/
f
1
/
/
i
'
/
/
/
f
' i
1
/
-L
I
1
i
\
1
/
t
I
I
i
/jj
—
f
t
' t
*
(
/
/
/
/
/
/
',
-L
/
1
1
1
1
I
1
1,
\
1
1
/
/
.
/
/
/
/
j
/
/
/
1
J-
1
1
/
A
/
/
/
/
1
/
t
'
i
/
/
/
i
1
/
I
1
i
/
/
/
"l
/
/
/
_/
1
1
1
1
500
7
i
' t
/
/
<-i
',
'
ff
/
*•
/,
f
1
t
i
i
i
i
n
1
i
i
j
/
-f
/
/
/
/
f
/
/
/
/
f
1 \c\
0 100 ISO 200 300 400 500
TEMPERATURE.-F
25. Vapor pressure of hydrocarbon oils (low range). (Esso Research and Eng.
Figure 11. Vapor Pressure of Hydrocarbons. Adapted from
Petroleum Refinery Engineering, W. L. Nelson, 1958.
99
-------
3.5.51 CAAP Factor Development Comment Number 23 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
47. Page 4-166. paragraph 4: This needs to be rewritten. It is obvious that this paragraph
is recycled from some other writeup. There is no relevant "response 53" in the
present writeup. Further, emissions are not only dependent on temperature, they are
also dependent on convective effects, i.e. evaporation rate increases in the presence
of strong convection.
3.5.52 Response to CAAP Factor Development Comment Number 23 - We agree that
there is no relevant "response 53". In addition, the document to which "Response 53"
refers is missing from the reference list at the end of the chapter. The statement refers to
Response number 53 in the EPA report "Response to Comments on Testing Program for
Asphalt Plants C and D." The reference number of this report was added to the text. It is
not clear that convective effects would have a significant impact on the emissions from
asphalt in the bed of transport trucks. First, emissions of low molecular weight
hydrocarbons are more likely a function of the rate at which volatile compounds can
reach the asphalt surface, the rate that the high molecular weight asphalt molecules crack
thereby creating the more volatile compounds within the asphalt and the rate at which
these volatile compound can migrate from air spaces inside the asphalt pile to the surface
of the pile. Second, near the surface of the asphalt pile in the transport truck without
ambient air movement, the concentration of these volatile compounds was measured at
about 10 ppm. This is much lower than the saturation concentration of these compounds.
In order for convection effects to significantly influence the evaporation rate, the surface
concentrations would need to approach saturation without air movement. Third, in many
States, the asphalt in the transport truck bed is required to be covered. This cover limits
the convective effects the asphalt is exposed to during transport.
3.5.53 CAAP Factor Development Comment Number 25 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
48. Page 4-166. final paragraph: This paragraph presents a single emissions factor for
yard emissions. This is exactly what needs to be done with the RTFOT equations.
They should be eliminated and only a single factor, based on RTFOT of 0.5%
presented in AP-42.
3.5.54 Response to CAAP Factor Development Comment Number 25 - EPA believes that
a single RTFOT value would overestimate emissions in the majority of locations in the
US and also would underestimate the emissions in a few locations. The adjustment of
measured emission to accommodate differences in emissions due to the effects of
temperature and asphalt volatility was encouraged by the commentors. EPA believes that
this adjustment has accommodated two parameters that would have otherwise been
viewed as imprecision in the measurement of these emissions. Additionally, the use of
100
-------
these two parameters improves the reliability of the final emissions estimates and the
ability to adapt for different regional, regulatory and health impact situations and goals.
3.5.55 CAAP Factor Development Comment Number 26 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42, Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
53. Tables 4-4 to 4-14: There are numerous deletions of data shown in these tables
where emission factors for each plant are presented. The reasons for the deletion are
not explained in the text.
3.5.56 Response to CAAP Factor Development Comment Number 26 - There were 89
pieces of data presented in Tables 4-4 to 4-14 that were not used in developing emission
factors. For 56 pieces of data the quality rating of the data is D. In Chapter Three of this
appendix, the criteria for assigning the emission factor ratings are explained. As
indicated, generally only lowest quality rated emission factors include D rated data and D
rated data are generally not used when data of a higher quality rating are available. For
12 pieces of data, an explanation is provided in the text summarizing the specific
reference or in the text describing the development of the emission factor. For the
remaining 21 pieces of data, explanations were added to either the text summarizing the
specific reference or the text describing the development of the emission factor. The
following table provides information on the data that were not used and the 21 instances
where explanations for the reason why the data was not used were added.
Emission Test Data Referenced and Not Used to Develop an Emission Factor
Table No.
4-6
4-6
4-6
4-7
4-7
4-7
4-8
4-8
4-8
4-8
4-8
Page No.
4-263
4-266
4-266
4-267
4-274
4-275
4-276
4-277
4-278
4-278
4-279
Reference No
22
56
11
209
350
236, 22
142
142
142
340
340
Comment
D rated data.
Explanation in Table.
Report rating was revised from C to D.
D rated data.
Explanation was added.
D rated data.
D rated data (three instances).
D rated data (two instances).
D rated data.
D rated data.
D rated data & explanation in text.
101
-------
Emission Test Data Referenced and Not Used to Develop an Emission Factor
Table No.
4-9
4-9
4-9
4-9
4-9
4-9
4-9
4-9
4-9
4-9
4-9
4-9
4-9
4-11
4-11
4-11
4-11
4-11
4-11
4-11
4-12
4-12
4-12
4-13
4-14
Page No.
4-281
4-281
4-281
4-282
4-282
4-283
4-283
4-284
4-285
4-285
4-285
4-286
4-287
4-293
4-293
4-296
4-298
4-298
4-298
4-299
4-300
4-305
4-305
4-306
4-308
Reference No
40
50
48
141
40
40
146
48
50
48
340
340
340
240, 46
331
1
1
76
15,40
40
296
90
161
317,318,319
47,49
Comment
Explained in text.
Explanation was added.
Explanation was added.
D rated data.
D rated data (six instances).
D rated data (eight instances).
D rated data.
Explanation was added.
Explained in text.
Explanation was added.
D rated data (eight instances).
D rated data (eight instances).
D rated data.
Explanation was added.
Data was used for factor development.
Explained in text (three instances).
Explained in text (two instances).
Explanation was added.
Explanations were added (five instances).
Explanations were added (two instances).
D rated data.
D rated data.
Explained in text.
Explanation was added.
Explained in text.
102
-------
Emission Test Data Referenced and Not Used to Develop an Emission Factor
Table No.
4-14
4-14
4-14
Page No.
4-310
4-310
4-311
Reference No
40
143
49
Comment
D rated data (nine instances).
D rated data.
Explained in text.
3.6 Characterization of Emissions
3.6.1 State Emissions Characterization Comment Number 1 - (Massachusetts State
Senator Stephen F. Lynch) -
Additionally, as I mentioned in my last letter dated September 14, 199, it is my
understanding that this report, again, neglects to specify the composition (ratio of
bituminous cement to aggregate) of the asphalt used in the study. Due to various
engineering standards/requirements associated with asphalt, it is very important that this
information be provided so that test results can be adjusted accordingly.
3.6.2 Response to State Emissions Characterization Comment Number 1 - Information
on the ratio of bituminous cement to aggregate was included in the draft and final reports
of both load-out emission tests. As indicated in Section 2.1 Process Description of the
EPA test report Hot Mix Asphalt Plants, Truck Loading and Silo Filling Instrumental
Methods Testing, Asphalt Plant C, Los Angeles, California, Final Report
(EPA/R-00-024, May 2000), "The percent by weight of liquid asphalt in the mix varies
from 4.8% to 6.0% depending on the size of the aggregate (the smaller the aggregate, the
higher the liquid asphalt content)." Additionally, as shown on the last four pages of
Appendix B of the EPA test report Final Report, Hot Mix Asphalt Plants, Truck Loading
Manual Methods Testing, Asphalt Plant D, Bare, MA, (EPA/R-00-028, May 2000), the
percent weight of asphalt in the mix ranged from 4.4% to 7.2%. During the emissions
test for plant C, the average asphalt used was 5.0% and for the five days of the test ranged
from 4.9% to 5.2%. During the emission test for plant D, the average asphalt used was
5.2% and for the three days of the test ranged from 4.5% to 5.7%. Although the above
information was added to the text reviewing the specific data sets in the AP-42
Background Report (Emissions Assessment Report Appendix B) EPA does not support
adjusting the data further to accommodate the variations of the asphalt content of the
HMA produced. While the range of asphalt content in any specific HMA formulation
may vary significantly, as indicated by the two EPA tests, the range of asphalt content
averaged over a days operation of a typical plant would be significantly less.
Furthermore, the range asphalt content averaged over longer periods (months or entire
year) would show very little variation. It is believed that the average asphalt content of
5.1% for the HMA produced during the EPA emissions tests are very close to the typical
average asphalt content of most HMA plants. Additionally, differences between this
value and measurement made at other plants are within the precision and accuracy of the
103
-------
weigh scales used by the plants for measuring the materials. As a result, it would be
unclear whether differences in the adjusted emissions are due to actual differences in the
asphalt content in the HMA or differences due to measurement errors of the weigh scales.
3.6.3 Local Health Agency Emissions Characterization Comment Number 1 (Elaine T.
Krueger, Environmental Toxicology Program (BEHA) -
Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
A final comment regards EPA reporting averages of data collected without presenting the
range of numbers. This concern has also been raised by others. The use of average
values may be appropriate for evaluating risks of long-term health effects e.g., cancer
risk, but may underestimate the potential for health effects associated with other issues,
e.g., short-term odors, that have been reported to be associated with these types of
facilities. Clearly, some of the chemicals, e.g., naphthalene, contained in the asphalt have
very low odor thresholds. For that reason, we recommend that EPA present the full range
of predicted emissions data including upper-bound or maximum estimates of emissions as
they relate to the potential for shorter term effects, such as odors. This is particularly
important given that odors have been important issues at other facilities like this
elsewhere in the U.S. (E.g., Michigan).
3.6.4 Response to Local Health Agency Emissions Characterization Comment
Number 1 - There are a variety of reasons why AP-42 does not include additional
emissions factors that would predict upper-bound or maximum values. First, AP-42 is
primarily used for making source-specific emission estimates for area wide inventories.
These inventories have many purposes including ambient dispersion modeling and
analysis, control strategy development, and in screening sources for compliance
investigations. Generally, long averaging times reduce the concern over short term
variabilities and the over and under estimations (due to disparities between estimated
emissions and actual emissions) tend to cancel each other. Second, not all emission
factors are based upon the same amount of data nor the same quality of data. Emission
factors based upon large data sets (like the HMA section) will have a greater difference
between the average value and the maximum value. Conversely, emission factors based
upon few data are likely to have a relatively small difference between the average value
and the maximum value. Third, emission factors for source categories where emissions
can be highly controlled, also have some supporting data from a few facilities with
relatively high emissions. Their high emissions may be due to their location in a State
without stringent emission limits or because they are an older facility that is grand
fathered. Fourth, this type of information is rarely required and presenting this type of
information in the AP-42 section would further complicate information that is already
technically complex. Fifth, the range of possible time frames different individuals would
want is great. For example, where one individual might want a 15 minute value while
another would want a one hour value. And lastly, local situations may necessitate the use
of either a value based on unique conditions or may require the collection of additional
locally specific information that cannot be reflected in a "one size fits all" value.
Additionally, different compounds may have health and nuisance thresholds that may
have differing averaging times.
104
-------
The HMA section has a background report which contains a significant amount of
information which individuals can use to accommodate individual concerns related to a
variety of uncertainties including estimating upper bound and short-term emissions. If
required, copies of a few test reports could be obtained if more detailed information is
required. In using this information, the user should develop an understanding for the
conditions under which the data were collected. For example most tests are three run
averages where each run was collected over one to six hour time frames. Estimating
shorter term values from this data may involve significant uncertainties. In addition, the
user should assess the probabilities (even if they can only be assumed) of simultaneous
unlikely scenarios occurring since using multiple worst case type assumptions generally
infers a highly improbable or almost impossible result.
3.6.5 Industry Emissions Characterization Comment Number 1 (Steve Zemba,
Cambridge Environmental) -
Hot Mix Asphalt Plants Emissions Assessment Report
General: Tables 1 and 2 claim to present emissions for a "typical" plant. This is probably
true for mixer emissions, but is not the case for the fugitive sources, as indicated by
statements on p. 14. Specifically, the fugitive emission estimates assume a high-end
volatility of asphalt cement, and also a maximum recommended load-out temperature. I
would thus characterize the fugitive emission estimates as "high-end" for a
typically-sized plant. EPA might want to clarify this matter somewhat, since the actual
ratio of fugitive-to-stack emissions for a typical plant running at typical conditions will be
lower than that implied in Tables 1 and 2.
3.6.6 Response to Industry Emissions Characterization Comment Number 1 - It is
recognized that the loss on heating of 0.5% is likely an upper bound estimate of volatility
of asphalt for a number of States based upon information that has been collected. It is
also recognized that an average load-out temperature of 325° F is near the maximum
target value for HMA plants. As a result of both of these conditions, it is estimated that
the resulting emission estimate may be higher but as much as a factor of three over the
operating conditions at the typical HMA plant. However, as currently drafted, the HAP
emissions for load-out, storage, the eight minutes following load-out and HMA storage
silo (for drum mix plants) are only 1.5% of the total of the HAP emissions for batch
plants and only 5% of the total for drum mix plants. Since this is a relatively small
percentage, it is not necessary to add text regarding the relative significance of these
emission as this is readily apparent from the estimates derived.
3.6.7 Industry Emissions Characterization Comment Number 2 (Steve Zemba,
Cambridge Environmental) -
Hot Mix Asphalt Plants Emissions Assessment Report
p. 13: Are emissions from hot oil heaters included in any of the categories in Tables 1 & 2
(e.g., the "mixer" category)? If not, does it make sense to mention the likely magnitude
of these emissions relative to other sources (qualified, of course, by the large degree of
uncertainty resulting from the dearth of test data)? Might it also be possible to use
105
-------
emission factors from similar burners (used in other industries) to evaluate hot oil heater
emissions?
3.6.8 Response to Industry Emissions Characterization Comment Number 2 - As stated
elsewhere, emissions from other sources at HMA plants were added to Tables 1 and 2. In
addition to adding an emissions estimate for hot oil heaters, road dust emissions, material
handling emissions, and emissions from diesel truck exhaust were added.
3.6.9 CAAP Emissions Characterization Comment Number 1 (Ravi Nadkarni & Lloyd
Pillion) -
Hot Mix Asphalt Plants Emissions Assessment Report
4. Page 2. paragraph 4: In this and in other paragraphs and tables to follow, emissions
and/or emission factors are given as averages. If this data was derived by averaging
emissions from many plants, as indicated in Appendix B, measures of variability
such as standard deviation should also be given. At present, this information is
buried in the tables in the two appendices.
3.6.10 Response to CAAP Emissions Characterization Comment Number 1 - The
presentation of information on the variability of the supporting data in the body of the
Emissions Assessment Report would only complicate the already complex set of values
presented. The two appendices provide increasingly technical discussions which support
the information presented in the body of the report. As the commentor has stated in other
comments, the appendices do include the standard deviation for some emission factors.
Most emission factors do not include information on the variability of the supporting
data. However, appendix B includes the individual data supporting all of the emissions
factors that are provided. With the individual data, the user is not limited to one measure
of variability and can develop statistical information that is tailored to the specific
situation that they may be assessing.
3.6.11 CAAP Emissions Characterization Comment Number 2 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
1. Page 1', paragraph 5: The fourth line from bottom uses the word "gas" when " vapor
of organic compounds" would be more correct.
3.6.12 Response to CAAP Emissions Characterization Comment Number 2 - The report
has been revise to use the more proper term vapor in place of gas.
3.6.13 CAAP Emissions Characterization Comment Number 3 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
8. Page 8. paragraph 1: The second line states that both the low and high molecular
weight compounds "may" contain small amounts of HAP. We were under the
impression that such asphalt-derived compounds always contain some HAPs. Is
106
-------
EPA suggesting that certain asphalts don't contain any HAPs? Is there any data to
support this statement?
3.6.14 Response to CAAP Emissions Characterization Comment Number 3 - The word
"may" was deleted from the report.
3.6.15 CAAP Emissions Characterization Comment Number 4 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
9. Page 8. paragraph 3: In the fourth line, a particle size analysis of cold feed is
presented. We don't see the relevance of this sentence since the previous sentence
and the subsequent sentence both talk about fugitive dust and cold feed is not a
major contributor to fugitive dust.
3.6.16 Response to CAAP Emissions Characterization Comment Number 4 - The
percentage of silt (less than 74 //m or minus 200 mesh) is used as a variable in calculating
some fugitive paniculate emissions. Although a silt content that is specific for the facility
improves the predictive accuracy of the emission estimate, the presentation of a average
estimate is useful when a facility specific value is unavailable.
3.6.17 CAAP Emissions Characterization Comment Number 5 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
10. Page 8. paragraph 4: If as you say in the last sentence that data are not available to
support the hypothesis, EPA should not be repeating the hypothesis. Again, such a
statement belongs in industry sales literature and not in an EPA document.
3.6.18 Response to CAAP Emissions Characterization Comment Number 5 - While the
available data to support the hypothesis are very weak and other process parameters are
probably responsible for the majority of the variations, the hypothesis is worthy of
mention. The statement that data are not available to support the hypothesis was revised
to "temperature during processing, data supporting this hypothesis is very weak.
Specifically, although the data show a relationship only between RAP content and
condensible organic particulate emissions, 89% of the variations in the data were the
result of other unknown process variables."
3.6.19 CAAP Emissions Characterization Comment Number 6 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
11. Page 9. paragraph 1: Most of this paragraph is again a replay of industry assertions.
The only relevant issue is whether or not these process modifications work to reduce
"blue smoke". We notice that the industry favorite, the wet electrostatic precipitator
or smog hog is conspicuous by its absence from this list.
107
-------
3.6.20 Response to CAAP Emissions Characterization Comment Number 6 - The
technologies mentioned have been demonstrated and are mentioned to provide
individuals that encounter facilities with these problems to understand that solutions are
available. It is unknown whether any wet electrostatic precipitator has been used to
control PM and/or blue haze from the kiln exhaust. As a result, this control technology
was not mentioned.
3.6.21 CAAP Emissions Characterization Comment Number 7 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
12. Page 9, paragraph 2: A repeat of the comment on page 5 about how such plants "will
likely have lower emissions". If you have the data, show it; if not, let the
industry do its own PR.
3.6.22 Response to CAAP Emissions Characterization Comment Number 7 - The
statements concerning the organic compound emissions address the location that organic
material (from either RAP or asphalt cement) is added to the new aggregate and whether
those emissions would be ducted to the kiln exhaust stack. Information, of this type is
useful to State and local agencies that review permits. However, to clarify the statement,
it was revised to "not in contact with the hot exhaust gas stream, counterflow drum mix
plants will likely have lower organic compound emissions from the kiln stack than
parallel flow drum mix plants. However, variations in the emissions due to other
unknown process variables are more significant. As a result, emission factors for parallel
flow and counter flow drum mix plants are the same."
3.6.23 Industry Emissions Characterization Comment Number 3 (Gary Fore, Vice
President National Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 15: A. Last paragraph - Please add the following sentence to the end of the
paragraph. ...equation developed from the emission data. It should be noted that truck
emissions would decrease rapidly with time as a result of surface cooling and crusting and
industry practice of covering loads with tarps.
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-8: A. In the first partial sentence on the page, please add prior to tarping after
truck. The sentence will then read: Organic vapor ... bed of the truck prior to tarping,
3.6.24 Response to Industry Emissions Characterization Comment Number 3 -
Although the surface material of tarped or un-tarped HMA may cool rapidly during
transport, it is believed that only the organic particulate emissions may be reduced due to
condensation on the cooler surface material. The emission factor discussed in the
paragraph is for the compounds that remain vapors after cooling to ambient conditions. It
is believed that the emissions of these more volatile compounds would not be reduced as
rapidly as the particulate or visible emissions. It is believed that the more volatile
108
-------
emissions are reduced more slowly over time as a result of the more general cooling of
the hot mix asphalt as a whole. This cooling is much slower than the cooling of the
surface material. The addition of the phrase "prior to tarping" infers that the tarping of
the HMA somehow inhibits the emissions of organic vapor from the bed of the truck.
While tarping may limit to some extent the visible emissions from the truck bed, it is not
believed that tarping inhibits the emissions of the volatile organic compounds that were
measured during load-out. Even if there were information that supported a reduction in
organic vapor emissions due to tarping, it is likely that these emissions would only be
delayed until the tarp was removed. In addition, the presence of some covers may inhibit
the loss of heat from the asphalt in the bed of the truck. The higher temperature retained
by the HMA may in turn result in slightly higher emissions for a given load-out
temperature. As a result the changes suggested by the commentor were not made.
3.6.25 Industry Emissions Characterization Comment Number 4 (Gary Fore, Vice
President National Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-17 :B. Note there is an error under VOC for natural gas-fired dryer and the
No.2 fuel oil-fired dryer. They should be 0.0076 for both not 0.0082
3.6.26 Response to Industry Emissions Characterization Comment Number 4 - While
there appears to be a mathematical error in the table, the differences in the VOC emission
factor are due to the rounding of the formaldehyde, methane and THC emission factors.
The arithmetic was reviewed. The emission factor is correct. A clarifying footnote was
added to indicate that the values presented are different due to rounding.
3.6.27 Industry Emissions Characterization Comment Number 5 (Gary Fore, Vice
President National Asphalt Pavement Association) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
Page 11.1-22 and 23: B. Delete uncontrolled from this table because to be consistent
with other HAPs, we should use data collected after filtration in the stack. Please make a
footnote reference that the uncontrolled numbers are in the Emission Factor
documentation for AP-42 Section 11.1 Hot Mix Asphalt Production.
Page 11.1-24: A. Delete uncontrolled from this table because to be consistent with other
HAPs, we should use data collected after filtration in the stack. Please make a footnote
reference that the uncontrolled numbers are in the Emission Factor documentation for
AP-42 Section 11.1 Hot Mix Asphalt Production.
Page 11.1-28: A. Delete uncontrolled from this table because to be consistent with other
HAPs, we should use data collected after filtration in the stack. Please make a footnote
reference that the uncontrolled numbers are in the Emission Factor documentation for
AP-42 Section 11.1 Hot Mix Asphalt Production.
3.6.28 Response to Industry Emissions Characterization Comment Number 5 - When
available, emission factors prior to air pollution control devices are provided in AP-42.
109
-------
Information on uncontrolled emissions is used by some individuals to assess the
regulatory status of some facilities, assess the potential impact due to air pollution control
device by-pass during start up or shut down or to assess the potential impact of a
temporary malfunction of the air pollution control device.
3.6.29 Industry Emissions Characterization Comment Number 6 (Gary Fore, Vice
President National Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
Page 3: A. At the end of the last paragraph -Please add an editors note: "HAPs
emissions are typical of emissions that result from fossil fuel combustion."
3.6.30 Response to Industry Emissions Characterization Comment Number 6 - A
limited comparison of the AP-42 emission factors for a drum mix HMA plant with a fuel
oil fired boiler firing the same quantity of fuel oil (1.6 gal per ton of HMA) indicates that
HMA plants generally have higher emissions than an industrial or commercial boiler.
Therefore, the statement will not be added.
3.6.31 Industry Emissions Characterization Comment Number 7 (Gary Fore, Vice
President National Asphalt Pavement Association) -
Hot Mix Asphalt Plants Emissions Assessment Report
PagelS: A. 2"d paragraph -Please add to the end of the paragraph. ..."representative of
gas-fired heaters As a practical matter theses units represent a very small
combustion source when compared to the dryer.
3.6.32 Response to Industry Emissions Characterization Comment Number 7 - At this
location in the report, this statement is not believed to be appropriate. However,
estimated emissions from the hot oil heater were added to Tables 1, 2, 6 and 10. In
addition, the basis for the emissions estimates are included in Section 2.5. The following
information was included in this basis. The typical HMA plant has two asphalt storage
tanks that are 50 ft long by 8 ft diameter. It is estimated that these storage tanks require a
total heating capacity of about 200,000 Btu/hr based on a heat loss of 60 Btu/ft2 of tank
surface area. The asphalt storage tanks are kept hot continuously for the five months the
HMA plant operates. As a result 720 million Btu's are used to maintain the temperature
of the asphalt in the storage tank. For a gas fired hot oil heater, 720,000 cubic feet of gas
would be combusted. For an oil fired hot oil heater, 5,100 gallons of fuel oil would be
combusted. It is estimated that this fuel usage is about 3% of the fuel used in a typical
batch mix plant (100,000 tons HMA/year) and 1.6% of the fuel used in a typical drum
mix plant (200,000 tons HMA/year).
3.6.33 CAAP Emissions Characterization Comment Number 8 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
1. Page 2-4. last paragraph: This is a repeat comment in the last line. If you don't have
the data about organic emissions, don't speculate about "potential".
110
-------
3.6.34 Response to CAAP Emissions Characterization Comment Number 8 - The
statements concerning the organic compound emissions address the location that organic
material (from either RAP or asphalt cement) is added to the new aggregate and whether
those emissions would be ducted to the kiln exhaust stack. Information, of this type is
useful to State and local agencies that review permits. However, to clarify the statement
it was revised to "asphalt cement occurs in the hot combustion product flow, there is a
potential for kiln stack organic emissions (gaseous and liquid aerosols) to be greater than
in counterflow plants and some batch plants. However, this increased emissions is not
evident in the data due to variations in the emissions due to other unknown process
variables are more significant."
3.6.35 CAAP Emissions Characterization Comment Number 9 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
8. Page 2-7. paragraph 3: Same comment as above under item ? regarding dataless
conclusions. The last sentence belongs in an industry sales brochure, not in an EPA
document.
3.6.36 Response to CAAP Emissions Characterization Comment Number 9 - The lack of
data to substantiate the statement is not a reason to eliminate the statement. As with
many other emissions that were statistically evaluated, there is insufficient information
about the relevant process variables, an insufficient number of data and relatively high
variation in the emissions to properly evaluate the potential effect. However, the
statement was revised to indicate that the available data is insufficient to discern any
differences in emissions because of the differences in the process. With respect to the
processing of high percentages of RAP, a number of emissions tests of facilities
processing RAP at ratios of 50% are used in the development of the emission factors.
The statistical analysis performed to evaluate the effect RAP percentage on emissions
indicated that only the condensible organic particulate was affected by the RAP
percentage. However, the squared correlation coefficient was only 0.11 (with a
correlation of 1.0 being perfect) indicating that the RAP percentage explained very little
of the variability in the data.
3.6.37 CAAP Emissions Characterization Comment Number 10 (Ravi Nadkarni & Lloyd
Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
9. Page 2-10. paragraph 4: The last sentence belongs in an industry sales brochure, not
in an EPA document.
3.6.38 Response to CAAP Emissions Characterization Comment Number 10 - The
technologies mentioned have been demonstrated and are mentioned to provide
individuals that encounter facilities with these problems to understand that solutions are
available.
Ill
-------
3.6.33 CAAP Emissions Characterization Comment Number 11 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
13. Page 9, paragraph 4: The last three lines appear to be a plug for one particular firm
"Compliance Monitoring Service". Is it EPA's role to do this?
3.6.34 Response to CAAP Emissions Characterization Comment Number 11 - The only
source test information available indicating the potential benefits of combustion
monitoring was this report supplied by the State of New Jersey. While the report was
produced by Compliance Monitoring Service, it is no more an endorsement of their
services than any other test contractor referenced as performing any of the tests of Hot
Mix Asphalt Plants. In addition, we would not think that this statement would provide a
strong recommendation for this company since there was insufficient information in this
report for EPA to quantify the emissions benefits of this monitoring and the contractor
used a variety of assumptions rather than collecting the needed process information to
allow the use of this test data.
3.6.35 Industry Emissions Characterization Comment Number 8 (Steve Zemba,
Cambridge Environmental) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
General: Many HMA plants are now installing condensers on their asphalt storage tank
vents and silo vents; some are enclosing or partially enclosing their loadout areas. The
fugitive emissions discussed in this document are for uncontrolled sources. In fact, the
estimates in Tables 1 and 2 implicitly assume controls on the mixer (dryer) stack but no
controls for fugitive emissions. This assumption should be made explicit (perhaps in a
footnote to the tables). EPA might also want to offer some opinion/conjecture on the
efficacy of engineering controls for the various fugitive sources.
3.6.36 Response to Industry Emissions Characterization Comment Number 8 - The
potential for controlling the emissions from asphalt storage tanks and HMA storage silos
can be evaluated based upon the composition of these emissions. The composition of the
organic emissions can be determined from the measured emissions from the silo vent
during the emissions test at Plant C. The emissions measured were about 95% volatile
organic compounds and about 5% particulate organic compounds. In addition, the FTIR
analysis inferred that the majority of the volatile organic compounds were aliphatic
hydrocarbons between about pentane and nonane. While it is likely that atmospheric
condenser would provide some control of these organic compounds, the control efficiency
would only be a conjecture. Normally, the control effectiveness of condensers is
dependent on the temperature of the condenser and the concentration of the components
in the vent stream. While atmospheric condensers should be highly effective at reducing
the particulate components of the emissions and therefore the visible emissions, these are
a small portion of the overall emissions from these sources. Atmospheric condensers
would be significantly less effective in controlling relatively low concentrations (2,000
ppm) of low molecular weight compounds. The effectiveness of enclosing the load-out
areas provides no control of emissions unless combined with an emissions control device.
112
-------
In addition to having about the same situation as the asphalt storage tank and HMA
storage tank emissions, these emissions would be more dilute because of the high air flow
rates required to capture these emissions from load-out operations. Although the wet ESP
at Plant C may not have been operating properly during the industry sponsored emissions
test by AIRx Testing, this test showed essentially no control of either the particulate or
vapor phase emissions. While there may be a cost effective control strategy that would be
effective on these sources, it is premature to include such a statement in the AP-42
section since there is no basis at this time.
3.6.35 CAAP Emissions Characterization Comment Number 12 (Ravi Nadkarni & Lloyd
Pillion) -
Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
17. Table 1: We are totally confused by the data on condensible inorganic and organic
PM. These are defined in footnote b as being method-driven. It would have been
helpful if additional details had been presented. In particular, we are curious about
the chemical components of the condensible inorganic PM fraction. The table shows
that uncontrolled emissions are numerically the same as controlled emissions.
Footnote j states that "uncontrolled "values are assumed to be equal to the maximum
controlled value measured". Since the controlled values of 0.013 and 0.0042 are
averages (arithmetic means), and are the same as the uncontrolled values, is the table
saying that the average is the same as the maximum value in the data set? Please
explain. We recall that this problem has been in table 1 in previous versions of
AP-42.
3.6.36 Response to CAAP Emissions Characterization Comment Number 12 - The
commentor is correct that the components of particulate matter are method defined. The
addition of details to explain the test method is beyond the scope of AP-42 and the
background report for this section. The presentation of PM emission factors in this
manner is common throughout AP-42 where information on the condensible fractions of
particulate matter are available. More detail on pollutant terminology and conventions
are contained in the Introduction to AP-42. Beyond the Introduction to AP-42, the
introductions to the various EPA reference test methods may include additional detail on
the constituents that are measured by the test methods. This information is presented for
future assistance in reconciling ambient speciation information on fine particulate with
the speciated information from the emission sources. While EPA Method 202 does not
include determining the components in the inorganic or organic fractions, the inorganic
fractions are generally sulfates and nitrates with lesser amounts of other inorganic
substances. These inorganic substances are vapors at the filtration temperature used to
avoid water condensation on the filter of EPA Method 5, EPA Method 20 la or EPA
Method 17. The values for the controlled and uncontrolled condensible (inorganic and
organic portions) particulate are the same because fabric filter control devices (the most
prevalent in use by HMA plants) are not expected to control vapors and there is no
uncontrolled data for these components of particulate. While wet scrubbers may be
expected to control these components of particulate, no test data was available to indicate
113
-------
a suitable controlled emission factor with this type of control. The values are not
maximums but are averages, the word maximum was removed from the footnote.
3.7 Evaluation of Unused Information
3.7.1 CAAP Unused Information Comment Number 1 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
49. Page 4-167. paragraph 2-4: This is a critique of Ravi Nadkarni's original
submission in 1994. There are several problems here. If you are going to critique
this work, you should refer to all of submissions to the EPA, not just the first one
from 6 years ago. The complete list would include his response to Mr. Ryan of EPA
of October 26, 1995 and November 16, 1995 and to Mr. Mobley of July 5, 1996. We
find it particularly interesting that the estimate of vapor pressure of asphalt from
1994, based on several model compounds has been represented as one based on a
single model compound. Second, "people who live in glass houses should not throw
stones". In using the TANKS program, (see page 4-165) the authors used another set
of model compounds whose presence at the required concentrations is also not
supported by the same data that is used to attack this work. For example, if
naphthalene and 2-methyl naphthalene are predominant species of PAH in emissions
from asphalt DRYERS, why were tetracosane and pentacosane used as model
compounds for TANKS? Can you please recalculate the TANK emissions using
naphthalene and 2-methyl naphthalene?
Next, why is the Cambridge Environmental letter not listed in the references? It
should have been reference 358. Cambridge Environmental used another set of
model compounds to come up with a different figure. That does not make their
calculation any more valid since both estimates were done years ago in the absence
of data on components of asphalt vapor. (This also begs the question of why
DRYER emissions are used here when better data is now available from silo
emissions.) Given Cambridge Environmental's track record as an extreme advocate
for their industrial clients in the public health area, we would not characterize their
work as "independent". (In one case, they compared emissions from a single plant
versus emissions from all industrial plants of all types in a county and the state.
Even the most polluting plant can appear to be a winner under that criterion.) There
is more to the Cambridge Environmental story. In a public meeting in Uxbridge, Dr.
Zemba, the principal author of the letter, reversed his position and stated that he now
believed that the calculations are in the correct range, or only modestly overestimate
the actual emissions. Specifically, he said, "When I originally looked at the
calculations, I thought that these were grossly overestimated. I can see your point
though and they might not be overestimated by as much as I originally thought." He
also admitted that he had not studied Ravi Nadkarni's two 1995 letters to the EPA. A
specific critique of the Cambridge Environmental work is in the 1996 letter to David
Mobley.
114
-------
We agree that calculations done in the absence of actual measurements should be
discarded when actual measurements are available. But any critique should fairly
represent the work and its shortcomings.
3.7.2 Response to CAAP Unused Information Comment Number 1 - We believe that the
summary and the shortcomings identified are an accurate representation of the reference.
The original submission is cited since it was submitted to EPA for use as the basis for
emission factor for this source and has received wide distribution by the author. The
reference cited is intended to include not only the submission by Dr. Nadkarni but also
the various follow up submissions by Dr. Nadkarni and others. Data collected during the
EPA emissions tests of Plant C and Plant D and analyzed by agreed upon procedures
support the shortcomings that were identified in the analysis. The purpose of the
discussion is to provide a brief summary of the information and describe its usefulness in
developing the final emission factor. It is not intended to provide detailed critiques of all
aspects of the methodology described nor the critiques by others.
With respect to EPA's selection of compounds to represent asphalt in the TANKS
equations, this selection was based upon measured emissions and not assumptions about
the minor constituents of the mixture and the physical processes that are occurring. It
would not be appropriate to use molecular weight information from other compounds
unless the Antoine's coefficients are changed to produce emission concentrations at
325 °F that are the same as the saturated vapor concentration measured at the HMA
storage silos at Plant C. The TANKS software uses both the vapor pressure calculated
with the Antoine's coefficients and the molecular weight to calculate mass emissions.
While the molecular weights of naphthalene and 2-methyl naphthalene could be
substituted, the Antoine's coefficients for these compounds would not be appropriate.
Instead, a revised set of Antoine's coefficients would have to be selected so that the mass
emissions are 32 pounds per million gallons of asphalt throughput. In addition, these two
compounds comprise only 0.004% and 0.01% of the organic emissions (PM plus TOC)
from silo filling and all PAH's are only 0.024% of the organic emissions. At these
percentages, these compounds are not a significant portion of the vapor pressure of head
space organic compounds. The test data strongly support the conclusions that the
emissions are mostly aliphatic hydrocarbon vapors, a small component of aromatic
hydrocarbon vapors, a small component of organic particulate and trace amounts of
specific identified organic HAP compounds. The FTIR data also support the assumption
that the molecular weight of the vapor is between pentane (72 g/g-mole) and nonane (128
g/g-mole). Since the presentation of the liquid and vapor molecular weight as well as the
Antoine's coefficients for a model compound between tetracosane and pentacosane has
been misinterpreted by the commentors and may be misinterpreted by others, the median
molecular weight of 105 g/g-mole was used for tanks emissions. The Antoine's
coefficients for aliphatic hydrocarbons that come the closest to producing a working loss
emission estimate of 32 pounds per million gallons of asphalt throughput for this
approximate molecular weight are docosane and tricosane. As a result, the background
report was changed to describe this revised methodology. The revised Antoine's
115
-------
coefficients "A = 75,350.06" and "B = 9.00346", vapor molecular weight of 105 g/g-mole
and liquid molecular weight of 1,000 g/g-mole were included in the AP-42 section.
3.7.3 CAAP Unused Information Comment Number 2 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
50. Page 4-168. paragraph 2-4: The evaluation of reference 359 indicates that the
reference was not read critically. Tables 4-42 and 4-43 are presented without any
critical evaluation of the contents. The table shows about the same concentrations at
an upwind site and a downwind site. How is this possible, unless the measurements
represent background readings? There is no comment as to why higher
concentrations were detected at a residential site and how the residential site
compared to the upwind and downwind sites. If you are not going to evaluate the
information, don't present the tables.
3.7.4 Response to CAAP Unused Information Comment Number 2 - The reference was
not read critically since it was impossible to derive an emission factor from the test. A
more critical review would have been attempted should the samples from the railcar hatch
and storage tank vent been within the calibration range of the GC/MS. However, it does
offer limited support for the estimated concentration of THC for asphalt storage tank
emissions that was used to develop Antoine's constants for use with the TANKS
equations. Assuming that the residential site was near the source, an insignificant
difference between upwind and downwind concentrations provides information that
indicates that emissions are low relative to the background concentrations. Finding
significant increases in downwind concentrations requires at least four conditions to
occur. The source strength (mass per unit of time) be significant relative to the other
conditions. The downwind measurement must be close enough to the source so that the
pollutant is not significantly diluted due to dispersion in the air. The test method should
be capable of detecting the difference in concentrations. Last, the variations in the
upwind and downwind concentrations should be insignificant when compared to the
differences in upwind and downwind concentrations.
3.7.5 CAAP Unused Information Comment Number 3 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1, Hot Mix Asphalt
Production. June 2000 Draft
51. Page 4-169. paragraph 2-6: Again, the comments indicate that reference 360 was
not read critically. For example, the high outliers were rejected. Having read the
original report, we know that the authors don't plead that the instrument
malfunctioned. Therefore, the high measurements must be explained. Or are they
supposed to be background? It is likely that the wide variation in measured
concentrations was a function of measuring technique? Low readings were obtained
upwind of the emission source and high readings were obtained near the source and
downwind from it. Under these conditions, when the instantaneous wind direction
and sampling location with respect to the source were not recorded, there is no
justification for rejecting the high data points. Note that the determination of the
116
-------
high data points as outliers was based on a purely statistical argument (the "outliers"
were many standard deviations away from the rest of the data; therefore they don't
belong with the rest of the dataset); not one based on measurement errors or other
relevant factors.
The final paragraph again uncritically quotes the process used by the authors to
convert the measurement into a mass rate in g/sec. The concentration of benzene
was measured in concentration units i.e. g/1 which can be converted to g/m3. This
was then converted to an "emission factor" (actually a mass rate) by multiplying the
concentration by wind velocity in m/sec and the vehicle area in m2. This is data
manipulation of the worst sort. If one is trying to measure the flux of benzene
escaping from the truck, the flux would be expressed in units of mass flow per
second per unit area or g/sec-m2. The relevant area is that perpendicular to the
direction of flow of the flux that you are trying to characterize. In this reference,
concentration, wind velocity and truck area are combined in a way that might be
dimensionally correct but make no physical sense. In particular, it is difficult to see
how a measured wind velocity relates to the area of the truck. EPA does disservice
to the reader when material of this type is presented in detail over 4 tables without
critical evaluation.
3.7.6 Response to CAAP Unused Information Comment Number 3 - The reference was
not read critically since the derivation of an emission factor from the test would require a
number of additional unsubstantiated assumptions. For example, the amount of asphalt
loaded during the one minute period, the temperature of the asphalt and the volatility of
the asphalt. In addition, based upon the description of the sample collection, the air
velocities and concentrations are probably biased high since the funnel was located within
the bed of the truck near the falling HMA. While the authors excluded the high values
due to statistical criteria, for emission factor development outlier evaluations are
typically used to identify data requiring further evaluation. However, the test does
provide a reasonable semi-quantitative estimate of emissions for use in a dispersion
model used to determine the need for more sophisticated emissions testing. Within this
context, the results that were developed for truck load-out do provide limited support for
the load-out emission estimates for benzene.
3.7.7 CAAP Unused Information Comment Number 4 (Ravi Nadkarni & Lloyd Pillion) -
Appendix B: Emission Factor Documentation for AP-42. Section 11.1. Hot Mix Asphalt
Production, June 2000 Draft
52. Page 4-172. reference 364: This reference is the AIRx report of measurements at
Plant C. We want to see exactly what procedure was used to estimate capture
efficiency at 70 to 90%. Otherwise, this comment should be eliminated.
3.7.8 Response to CAAP Unused Information Comment Number 4 - The statement
concerning the estimate of the capture efficiency of the enclosure should have indicated
that the assessment was made during the evaluation of the test report done prior to
capture efficiency testing. While the capture efficiency testing at Plant C supports the
117
-------
estimates of capture efficiency that were made, this test report is not being used for
developing the final emission factor. Therefore, no additional evaluations are necessary.
The paragraph was revised to:
"Capture efficiency of the control system was not measured during the test. However, in
an evaluation of the test report by an EPA contractor, an assessment of the capture
efficiency was made. The assessment was made based upon the available information on
the load-out area in this test report. The assessment indicates that an average face
velocity of building air of about 42 feet per minute was estimated from data in the report.
It was recognized that this velocity is outside the recommended range of 50 and 100 feet
per minute specified in the Industrial Ventilation Manual. It also was recognized that the
enclosure did not meet all of the criteria for a permanent total enclosure (PTE) specified
in the EPA document The Measurement Solution: Using a Temporary Total Enclosure for
Capture Efficiency Testing. Finally the assessment provides an estimated range of
capture efficiency of the system at between 70 and 90 percent. While this capture
efficiency assessment could be re evaluated using the measured capture efficiency and
emissions documented in reference 355, this is unnecessary since this test was not used in
developing the final emission factor."
118
-------
Chapter 5
Stakeholder Comments on
Draft Emissions Assessment Report,
AP-42 Section and AP-42 Background Report
-------
Stephen F. Lynch
Massachusetts State Senator
First Suffolk District
-------
COMMONWe/VUTH OF MASSACHUSETTS
MASSACHUSETTS SENATE
STATE MOUSE. BOSTON O2I33-1OS3
SENATOR STEPHEN F. LYNCH
FIRST SUFFOLK DISTRICT (CHAIRMAN)
STATE HOUSE. ROOM 424
TEL. (017)722-1150 WAYS AND MEANS
FAX r« (7)722-2191 EDUCATION
PUBLIC SERVICC
COUNTIES
August 31,2000 «™»
Mr. Robert McConnell
Environmental Engineer
Air Quality Planning Unit
U.S. Environmental Protection Agency
John F. Kennedy Federal Building
Boston, MA 02203-0001
RE: EPA-454/R-00-OXX Hot Mix Asphalt Plants Emission Assessment
Report.
Dear Mr. McConnell:
This letter is in regard to the EPA-454/R-00-OXXHot Mix Asphalt Plants
Emission Assessment Report. As you may know, there have been various
proposals to locate an asphalt plant in my district in the South Bay area of
Boston. As the State Senator for the Dorchester, South Boston, South End
and Roxbury neighborhoods which would be most impacted by this
proposed plant, I appreciate the opportunity to comment on this report.
In reading this report and its appendices, along with comments sent to my
office by Ravi Nadkarni, it is my understanding that this report only shows
emissions results from asphalt plants burning natural gas as a source of fuel.
The report also fails to provide information about what percentage of asphalt
plants cited in the report used natural gas and what percentage of those
plants, used alternative fuels. Although I am not an environmental engineer,
it is obvious to see that emission results from asphalt plants burning only
-------
PAGE TWO
August 31,2000
Mr. Robert F. McConnell, Environmental Engineer
natural gas, as opposed to other fuels, will show a considerably lower
emissions result. This would clearly indicate to me that the results in this
report are skewed. I wollld like to see more thorough testing of various data
included in the final report, especially if this report is to be considered an
emissions standard for all asphalt plants.
Additionally, as I mentioned in my last letter dated September 14, 1999, it is
my understanding that this report, again, neglects to specify the composition
(ratio of bituminous cement to aggregate) of the asphalt used in the study.
Due to the various engineering standards/requirements associated with
asphalt, it is very important that this information be provided so that the test
results can be adjusted accordingly.
As the elected Senator for this district, I would appreciate a response to these
issues.
Thank you for the opportunity to comment on this important matter. I look
forward to your response. If I can be of any further assistance, please do not
hesitate to contact my office.
Sincere
'STEPHEN F/LYNCH
State Senator
-------
Lloyd Pillion - Boston, MA
Ravi Nadkarni - Wrentham, MA
Coalition Against the Asphalt Plant
-------
Comments on EPA reports relating to AP-42 Section 11.1
August 31. 2000
Table of Contents
A. Introduction 1
B. General Comments about the Package 2
C. Summary of Major Deficiencies 3
D. Page-by page Comments on Hot Mix Asphalt Plants Emissions Assessment Report - EPA-454/R-
00-OXX, June 2000 7
E. Page-by page Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft . 11
F. Page-by page Comments on Appendix B: Emission Factor Documentation for AP-42, Section
11.1, Hot Mix Asphalt Production, June 2000 Draft 14
A. Introduction
This letter is a review of the Draft of "Hot Mix Asphalt Plants Emissions Assessment Report" dated
June 2000. The report is designated EPA-454/R-00-OXX. It comes with three appendices:
Appendix A is Section 11.1 of AP-42 of June 2000
Appendix B is "Emissions Factor Documentation" for the above
• Appendix C is on Preferred and Alternative Methods for Estimating Air Emissions from Hot
Mix Asphalt Plants dated July 1996.
The overall package is 3" thick and contains over 500 pages.
Before discussing the results of our review, we want to comment on the review process and our
general dissatisfaction with it. When complex, technical documents are sent to a wide audience for
review, one assumes that both the EPA and their contractors did their best to produce as final a draft
as possible. One would not only expect that most of the typos and numerical errors have been found
and eliminated but that the authors wrote the report in a way that makes it easy for the reviewer to
go back to the primary references and original sources (assuming that they are available) to check a
particular point. On both these counts, the reports, just like the previous set of reports reviewed last
year, fail resoundingly. Many key comments are not referenced at all; in some cases, the references
are incorrect; in other cases, the reference is made to a large document without giving a page number,
making it difficult and time-consuming to check. Appendix B refers to its Appendix B and Appendix
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
C on page 4-137. But these appendices, which presumably contain the statistical results and the data
sets, are not provided, making it impossible to even spot check the analysis. This, combined with the
severe time constraints, makes it impossible to review the reports in any but a cursory fashion, using
random spot checks, to make sure that the numbers are correct. Unfortunately, the reports also fail
on that count since numerical or logical errors in the analysis have been found.
We also need to air another complaint regarding the process. For about 6 weeks before the report
was finally issued, we received phone calls informing us that the release of the reports was imminent.
While we appreciate receiving this advance warning, it also turns out that the actual release of the
reports was much later than announced. One of us (Ravi Nadkarni) had cancelled attendance at the
Annual Meeting of IPMI in mid-June in response to these calls. Because of weddings in the family,
Ravi was going to be unavailable after July 17 and wanted to start the review process as soon as
possible, hence the cancellation. Unfortunately, the reports were received only on June 28. This
means that he missed an important Annual Meeting where he is on the Board of Directors. We
appreciate getting the extra time until the end of August to complete this review.
Appendix C was prepared in 1996. Why was this 69-page document not provided earlier so that it
could have been reviewed before this "crunch" period? Also, are comments on such a document,
which has undergone a "final" review, relevant at this stage? Finally, why was this document not
released to us when the citizen groups and the EPA were discussing sampling issues, when that
document might have provided an additional viewpoint? For all these reasons, this third document
has not been reviewed.
In our review, we present general comments about the package and then a summary of major
deficiencies. Finally, a page-by-page sets of comments are presented. The "major deficiencies" will
repeat discussions in the page-by-page comments and readers might argue with our selection of
certain areas to highlight as major deficiencies. However, because of the numerous problems with
the reports, the page-by-page comments became very voluminous. Hence the decision was made to
highlight some of the deficiencies in a separate section, even at the cost of some repetition, which
mirrors the repetitions in various segments of the reports.
B. General Comments about the Package
AP-42 is essentially a "cookbook" for engineers/contractors who need to produce emissions estimates
for local regulatory authorities to obtain permits to construct and operate plants. In view of this, the
overall organization of the three reports does not make sense. In the long run, Appendix A will be
used most often for estimating emissions. When necessary, the background information in Appendix
B will be reviewed. The "main" report will remain either as a political document used by lobbyists
or fade away. As it now stands, the "main" volume repeats in a vague and biased way, the
information in first two appendices. The same information is presented in a marginally better format
and with slightly less "spin" in Appendix A. If the main document is supposed to be an Executive
Summary to Appendix B, it is a poor summary and reads more like a political tract. This is
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
particularly true since only emissions data on plants using natural gas is given in the main report; i.e.
there is a deliberate attempt to show only the lowest emissions. Most of the caveats associated with
the summary information in the main report are well concealed in Appendix B. If EPA is serious
about making the main document useful, it should use the main document to calculate the total
emissions to the environment from a batch and a drum asphalt plant. This type of rewrite, where
other emissions (for example from materials handling and storage piles, storage tanks, from traffic
and truck exhaust, and so on) are also calculated and shown for typical plants, will result in a
guidebook for industry contractors and help them by demonstrating how such calculations should be
made. (In Massachusetts, specific plants will employ a wide range of pollution control technology
and then argue that it is "state of the art" and is BACT (Best Available Control Technology)). Other
emissions are currently only listed and cross-referenced in Section 2.3 on page 16. These include
emissions from storage tanks (AP-42 Chapter 7), materials handling and storage piles (AP-42
Sectionsl 1.19.2 and!3.2.4), vehicular traffic (AP-42 Sections 13.2.1 and!3.2.2) and emissions from
trucks bringing in the raw materials and taking away the hot mix asphalt (AP-42 Volume II).
The main report is a political document for the following reasons. The EPA's pro-industry bias
shows on almost each page. Every attempt is made to minimize emissions, for example, by
calculating emissions for plants burning natural gas only in the "main" report. The report also
provides qualitative comments to indicate that emissions could be even lower in many cases; for
example, "counterflow drum mix plants will likely have lower organic compound emissions than
parallel flow drum plants", page 11.1-9. Such comments are not substantiated in the detailed
information provided in Appendix B. Yet, caveats regarding conditions under which the emissions
might be higher are few and far between. This tilting is not surprising given the proindustry bias
expressed many times by the Project Officer during the course of testing.
The report and its appendices contain little or no information about industry structure. For example,
Appendix B asserts that drum plants are 90% larger than batch plants in terms of annual throughput.
This really falls out of an assumption made by the authors. There is little or no backup for this
conclusion. Similarly, there is no information on what proportion of plants use natural gas and what
proportion use other fuels.
Finally, it is unclear whether this "main" report has a number or whether EPA-454/R-00-OXX is the
final designation. The reason for asking this is because reference 1 of Appendix A has the same name
and a designation of EPA-454/R-00-019, but this reference is dated May 2000. What is going on?
C. Summary of Major Deficiencies
This section will highlight our major concerns some of which have already been covered in the
introduction or will be discussed in the page-by-page review in greater detail.
• Sloppy Report Preparation: The reports are full of minor errors, are poorly organized, repeat
the same paragraphs in different documents, and have incorrect references. Many key
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
comments are not referenced at all; in some cases, the references are incorrect; in other cases,
the reference is made to a large document without giving a page number, making it difficult
and time-consuming to check. The data sets used for statistical analysis are not provided,
making it impossible to even spot check the analysis. All indications are that there was no
complete internal review of the report within EPA before releasing it.
Poor Organization: The overall organization of the three reports does not make sense. In
the long run, Appendix A will be used most often for estimating emissions. When necessary,
the background information in Appendix B will be reviewed. The "main" report will remain
either as a political document used by lobbyists or fade away. The main report shows the
industry's emissions in the best possible light with most of the caveats associated with
information being well concealed in Appendix B. If EPA is serious about making the main
document useful, it should use the main document to calculate the total emissions to the
environment from a batch and a drum asphalt plant, rather than show the information from
AP-42 with the decimal point moved to the right.
Pro-industry Bias: The EPA's pro-industry bias shows on almost each page. Every attempt
is made to minimize emissions shown. The report also provides qualitative comments to
indicate that emissions could be even lower in many cases. We don't see why EPA makes
such statements in the absence of data. Yet, caveats regarding conditions under which the
emissions might be higher are few and far between.
Undocumented Information on Industry Structure: The report and its appendices contain
little or no documented information about industry structure. Yet such information is used to
derive average plant sizes and share of annual production between batch and drum plants.
Confusing Report Designations: The designation of the reports is confused. The main report
is called EPA-454/R-00-OXX. But Reference 1 of Appendix A has the same name and a
designation of EPA-454/R-00-019, and this reference is dated May 2000 while the main
report is dated June 2000.
Deficiencies in Method 204: Tracers were not used at Plant D because it met the requirements
of Method 204. Unfortunately, Method 204 is inadequate to achieve the results it tries to
achieve. Further, the contention that the estimate of unmeasured emissions is an "upper
bound" is incorrect. At Plant D, a mild ambient breeze was sufficient to cause some of the
fumes to escape from the top or the bottom openings in the downwind door of the tunnel.
Therefore, the size of the openings was decreased further at the end of the first day of testing.
This means that the Method 204 criterion of maintaining over 200 fpm at a natural draft
opening is not adequate. Note that 200 fpm is less than 2.3 miles per hour, which is not much
of a breeze. Therefore it is not surprising that these criteria are inadequate to assure total
containment within a TTE as contemplated by method 204.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
Further, Method 204 has a more serious and fundamental problem. Although the size of
natural draft openings is specified, the method does not contain any criteria to ensure that the
emissions are pulled/sucked past the sampling point in a reasonable period of time. As is well-
known, residence time calculations, based on enclosure volume and fan capacity, generally
provide incorrect information by underestimating the time to evacuate the enclosure because
of channeling. At Plant D, the tunnel was not evacuated in the approximately 15 second time
gap between the final HMA drop and the opening of the doors. This can be clearly seen in
the fact that THC readings did not drop to zero before the doors were opened. This fact was
also observed visually inside the TTE enclosure since SVOCs are visible. Thus, the report
contains compelling evidence that the collected sample had a low bias, not only because some
emissions escaped through the natural draft openings but also because these emissions did not
travel to the sampling point but were lost when the doors were opened.
• Incorrect Approach in Compensating for Volatiles Content of Asphalt: The methods used
in these reports to correct for volatiles content of asphalt are contrary to the agreement that
was reached between the EPA and various citizen groups. The citizens had argued that the
asphalt specifications allow the use of asphalt with a much higher weight loss compared to
the asphalt used in Plant C and Plant D tests. The older AC specifications allow for a 0.5%
loss on heating, which can be waived by an engineer on site who can then allow the use of an
asphalt with up to 1% weight loss. The new Superpave specifications allow for a 1% loss on
heating and therefore represent a relaxation of the previous AC grade standards. In view of
this and the industry's use of various additives and diluents (which would increase the vapor
pressure), the citizens argued and the EPA agreed that maximum number of 0.5% should be
used. Note that this was a compromise since the newer specification allows the use of asphalt
with twice that amount of volatiles. This issue is particularly important since the
measurements by State agencies quoted here appear to be for asphalt prior to blending with
additives, in spite of suggestions to the contrary, in some cases. The EPA, by providing an
equation where a hypothetical RTFOT value can be plugged in, is encouraging
misrepresentation by a proponent of any new asphalt plant, who, once the permit is received,
can then switch to an asphalt with a different RTFOT value with impunity.
As long as the specifications allow for up to 1% volatiles in asphalt (even before additives are
used), it is misleading for EPA to state that their approach in Appendix B - Table 4-27
"encourages the use of site specific data". The primary use of AP-42 is to prepare
calculations to get permits to build and operate hot mix plants. Therefore, there is no site
specific data that can be used other than fictional data. Unfortunately, the industry has a long
history of providing low-ball estimates, designed to fool the public and get permits from
agencies that are not particularly anxious to ask difficult questions.
• Incorrect Substitution of Delivery Temperature for Loadout Temperature: The statementthat
"325°F, which is the maximum load-out temperature recommended by industry" is imprecise.
Most specifications require this to be the maximum temperature at a job site where HMA
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
is to be spread and compacted. (The minimum temperature is 275°F.) The maximum load out
temperature is much more variable and can be higher if the plant is producing asphalt for a
distant job or for a small job, where there can be considerable cooling of the HMA mass
between loadout and delivery.
Improper Manipulation of Data from the "Background" Test Run 4. at Plant C: The report
states correctly: "The most reliable method to adjust for emissions measured during
background operations would be to separately adjust each run for the measured capture
efficiency and then subtract these adjusted background emissions from the adjusted emissions
measured during production operations." However, because this "procedure produces
negative values for both the PM and MCEM and many other HAP compounds", a new
unsound procedure is adopted because it gives results that please EPA and industry by
showing low but positive emissions. This new procedure is to use background emissions
uncorrected for capture efficiency. If the actual emissions have to be corrected for capture
efficiency because the enclosure does not capture all the emissions, there is no justification
for ignoring the capture efficiency for the background run. Having used this incorrect analysis,
the authors then have the temerity to advise us that this result might even have a "high bias"
relative to the correct method mentioned above.
The report justifies this faulty procedure because this situation "cannot be accommodated
retroactively." This statement is not correct. The best way to eliminate improper data
manipulation is to discard the phony background adjustment and report the data as that from
loadout plus truck exhaust. Since silos always dump hot mix into a truck, these combined
emissions are present at each silo loadout point. Further, in the unlikely event that a local
permitting authority requires inclusion of truck exhaust emissions (unlikely because we have
not seen such calculations associated with any local permitting requirements), note that the
truck spends about /^ minute under the silo but many more minutes, typically 3 to 8, on site.
As a result, the truck emissions that are included under loadout correspond to the 1A minute
portion. Therefore, the double counting of truck exhaust emissions would be minimal and can
be adjusted for by subtracting the 1A minute from residence time of the truck on site.
Overall, the background run demonstrates how data collection was manipulated to produce
biased results. The EPA Project Officer was interested in recording a high background and
the original data showed that he was able to manipulate truck placement near the entrance of
the tunnel in such a way that uncorrected emissions in the second half of the run were about
twice the emissions in the first half. If these emissions are corrected for the lower capture
efficiency in the second half of the run, the corrected emissions from the second half are
almost three times the corrected emissions from the first half. In spite of using the emissions
from the first portion of the background run, it appears that they were still too high when
corrected for capture efficiency and resulted in negative or zero emissions to the environment.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
7
Industry Comments on Capacity Are Given More Weight than Data Collected by EPA or
State Agencies: The report states that the hourly capacity data was averaged from the
capacities of precisely 98 batch plants and 162 drum plants. Since most of the data was
gathered during performance tests at full capacity and witnessed by State Agencies, it is
unlikely that hourly production rate data is incorrect. It is interesting that "conversations
with industry" are all that is needed to supercede this data. In the report, it is clear that the
authors believe that batch plants generally operate fewer hours per year than drum plants.
Reasons for this (not necessarily explained in the report) include the lack of silos for HMA
storage, small customers using smaller trucks, and so on. The report keeps talking about
hourly production rates when the discussion should focus on annual production, i.e. hourly
rate times hours operated per year.
• Unexplained Differences Between MCEM and TOC: The comparison of Plant C and Plant
D results shows a wide variation between emission rates for MCEMs and TOC. MCEM
PMs and TOC are generated the same way and from the same source. Therefore, if MCEM
generation is a function of loadout time, TOC generation should also be a function of loadout
time and TOC emissions should be higher for a batch plant in the same ratio. The report does
not explain this discrepancy. Of course, the comparison is further complicated and made
meaningless by the bogus "background" correction for drum plants discussed earlier. To
eliminate the confusion, separate emission factors should be prepared for batch and drum
plants in all cases.
• Inconsistent Assignment of Ratings: In one portion of the report, the emission factor for
sulfur dioxide was assigned an E rating because the data ranged over an order of magnitude.
This might have been simply the result of the variation in sulfur content of the fuel. In the
case of filterable PM, the range is over two orders of magnitude but the emission factor got
an A rating.
D. Page-by page Comments on Hot Mix Asphalt Plants Emissions Assessment Report - EPA-
454/R-OO-OXX, June 2000
1. Page 1. paragraph 3: In this paragraph, 1996 production statistics are presented along with
an estimate of the number of plants in the country. There are no references to the source of
these data. In Appendix B, page 2-1, the same paragraph is repeated. On page 2-2 and 2-3
of Appendix B, calculations are presented which derive average annual production rates for
batch and drum plants. We will comment on the problems with these calculations in our
page-by-page discussion of Appendix B. However, the entire discussion is aimed at showing
that drum plants produce more HMA compared to batch plants without actual data.
2. Page 1. paragraph 4: The statement in line 1 is incorrect. The primary emission source is the
dryer, not the mixer, as stated. This is the case because combustion of fuel for generating
heat for drying generates large quantities of criteria pollutants. This error of confusing dryers
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
and mixers is repeated several times, (for example in Table 1), and we have not bothered to
find all instances of such misuse. The best way to clarify this issue is to substitute
"dryer/mixer" for drum plants where the two units are connected. In the case of batch plants,
the dryer and the mixer (pug mill) are typically uncoupled through hot storage bins. Emissions
from both however are often captured by the same piece of air pollution control equipment.
3. Page 2. paragraph 3: This paragraph needs a cross reference, even if it is to the relevant
sections in Appendix B. In general, we need page number references to information in
Appendix B which is over 400 pages long.
4. Page 2. paragraph 4: In this and in other paragraphs and tables to follow, emissions and/or
emission factors are given as averages. If this data was derived by averaging emissions from
many plants, as indicated in Appendix B, measures of variability such as standard deviation
should also be given. At present, this information is buried in the tables in the two
appendices.
5. Page 8. paragraph 4: More industry statistics are presented without a source or a measure of
precision or accuracy.
6. Page 9, paragraph 1: The sentence on line 6 is confusing. Do you mean RAP is added to the
"hot" bins or is it added to the pug mill?
7. Page 12. paragraph 2: Is the rating system applied consistently? The discussion in Appendix
B indicates that it was not.
8. Page 12. paragraph 4: This paragraph presumably refers to the statistical analysis presented
in Appendix B, but there is no direct link or reference. Our detailed comments regarding the
statistical analysis are in our page-by-page discussion of Appendix B. However, the entire
paragraph is wrong or inconsistent with information in other parts of the report. It states that
the statistical analysis showed "no strong correlation" between parameters such as fuel type
and emissions factors. When such correlation is absent, the approach has been to use the
same emission factor in both cases. Yet, Table 11.1-5 in Appendix A shows different
emission factors for NOX and SO2 for different fuel types. For NOX the emission factor is
different for natural gas and oil fired dryers. For SO2 there are three different emissions
factors, one for gas, one for No. 2 oil and one for No. 6 oil. The backup analysis is not shown
in Appendix B. But obviously, the analysis showed that fuel type did affect emissions,
otherwise a single factor would have been used. Overall, the two sets of comments need to
be made more explicit and explained more carefully and the inconsistencies removed.
9. Page 13. paragraph 1: The quality rating needs to be quantified so that one can get a
numerical feel for the differences in ratings. If, as stated, the quality ratings are a "function
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
primarily of the number of data points", these various quality ratings can be related to the
standard error of estimate and using t-values, a range can be shown about the mean.
10. Page 13. paragraph 2: This paragraph is confusing. It is stated that only a single test report
was available for an oil-fired hot oil heater. Does this mean that only one hot oil heater was
sampled out of 364 plants and test reports? Further, if most of such heaters are fired with
natural gas, why was an oil-fired heater sampled? Finally, why does EPA go out of their way
to here to point out that "emission factors developed from this data would not be
representative of gas-fired heaters". While this comment is correct, since oil fired burners
emit more pollutants than gas-fired burners, there is a consistent bias in such comments. The
reports point out the possibility of a high-bias each time such possibility exists, inviting the
user of AP-42 to discount the published number. But, when the reported emission factors
have a low bias, there is no such comment. For example, see the issue of the phony
correction for background emissions to truck loadout discussed later on the same page.
11. Page 13. paragraph 4: This paragraph is wrong in several ways. The capture efficiency was
measured and calculated using tracers; it was not "estimated" as stated. To me, the word
"estimation" denotes a weaker approach to quantification than calculations based on actual
measurements. Equally, important, the EPA requirements for total enclosure were shown
to be totally inadequate during the testing at Plant D in Massachusetts. A mild breeze was
sufficient to cause visible emissions from the total enclosure even when a hand-held velometer
showed that Method 204 criteria were being met. Why is this factor, which suggests that
Method 204 criteria were inadequate and that the measured emissions have a low bias, not
discussed in this paragraph?
12. Page 13. paragraph 5: This paragraph discusses the subtraction of "background emissions",
which reduced the calculated loadout emissions significantly. Our detailed comments are in
our page-by-page discussion of Appendix B, which show that the method used is totally
without foundation or theoretical support, and the only excuse for using it is that it gives non-
negative results in several cases.
13. Page 14. paragraph 2: The statement that "325° F, which is the maximum load-out
temperature recommended by industry" is imprecise. Most specifications require this to be
the maximum temperature at a job site where HMA is to be spread and compacted with
rollers. (The minimum temperature is 275°F.) The maximum load out temperature is much
more variable and can be higher if the plant is producing asphalt for a distant job or for a small
j ob, where there can be considerable cooling of the HMA mass between loadout and delivery.
This same error is repeated in the last sentence of the next paragraph and in many portions
of Appendix B.
This paragraph also states that -0.5% was selected as a default value for weight loss in a
rolling thin film oven test. The methods used to apply the correction for volatility are
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
10
contrary to the agreement that was reached between the EPA and various citizen groups. The
citizens had argued that the asphalt specifications allow the use of asphalt with a much higher
weight loss. The older AC specifications allow for a 0.5% loss on heating, which can be
waived by an engineer on site who can then allow the use of an asphalt with up to 1% weight
loss. The new Superpave specifications allow for a 1% loss on heating and therefore
represent a relaxation of the previous AC grade standards. In view of this and the industry's
use of various additives and diluents (which would increase the vapor pressure) which are
often added after the asphalt has been sampled for the measurement of the weight loss, the
citizens argued and the EPA agreed that maximum number of 0.5% should be used at all
times. Note that this was a compromise since the newer specification allows the use of
asphalt with twice that amount of volatiles. This issue is particularly important since the
measurements by State agencies quoted here appear to be for asphalt prior to blending with
additives. The EPA, by providing an equation where a hypothetical RTFOT value can be
plugged in, is encouraging misrepresentation by a proponent of any new asphalt plant, who,
once the permit is received, can then switch to an asphalt with a different RTFOT value with
impunity.
14. Page 14. paragraph 3: EPA again makes the same recommendation which is to assume that
the load out temperature can never exceed 325° F. This is incorrect. For example, Mr.
Webster, the industry expert who has attended many of the EPA/citizen group/industry
meetings has looked at the photographs that Ravi Nadkarni submitted to the EPA in 1994 and
are in the EPA files. His comment was that the photos were typical of a plant loading out
material at a high temperature when supplying small contractors. If this practice is common
enough to be readily identified at a glance, EPA's comments about 325° F being the maximum
temperature under all conditions are wrong and need to be corrected.
15. Page 15. paragraph 2: It is ironic that EPA, after rejecting our repeated pleas for collecting
some fundamental data on asphalts such as its vapor pressure, states "Although vapor
pressure information on paving asphaltis not available ". Amore appropriate way to state
what has happened would be to state that vapor pressure measurements are not available
because of EPA's refusal to measure the vapor pressure.
16. Page 15. paragraph 3: The first line refers to "emission tests described in previous
paragraphs". Since the previous paragraphs refer to more than one plant test, it is preferable
to have a proper reference; for example, the text can refer to the test at Plant C or Plant D.
17. Page 16. paragraph 1: How do the truck exhaust emission factors published by EPA compare
with the background measurement from Run 4 at Plant C? Why is such a comparison missing
from the entire report?
18. Page 16. paragraph 3: The second line refers to Tables 4 through 11 which present more
detailed information about "typical" plants. Unfortunately, this is the emission factor data of
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
11
Appendix A multiplied by 100,000 for batch plants and by 200,000 for drum plants. The
implication is that the reader can't move the decimal point to the right or multiply by 2.
Further, in keeping with their bias for showing only low numbers, these tables are for natural
gas only and not for oil-fired heaters. Why? Finally, what is the basis for assuming (third
bullet on page 17) that PM emissions from load-out and silo filling are entirely PM-10? We
would have thought that a better assumption was that these were PM-2.5.
E. Page-by page Comments on Appendix A: AP-42 Hot Mix Asphalt Plants June 2000 Draft
1. Page 1. paragraph 3: This paragraph is repeated in the main volume, and in the first two
appendices. It would be more useful to discuss the source of this data and comment on its
reliability, rather than repeat it in three places.
2. Page 2. paragraph 1: The statement that raw aggregate is stockpiled "near" the plant is
ambiguous. Such aggregate is normally stockpiled within the plant site and is moved to the
cold feed bins with a front-end loader.
3. Page 2. paragraph 3: This paragraph repeats the ambiguity in the main report on page 9
paragraph 2. Is RAP added to the weigh hopper or to the hot bins?
4. Page 5. paragraph 2: In counterflow plants, the baghouse is preceded by a cyclone to reduce
the load on the baghouse. This is a trade off issue where it is cheaper to use a cyclone than
to increase bag house size and bag cleaning frequency to achieve the same result. Since the
industry prefers counterflow plants to parallel flow plants(page 1 paragraph 3), the use of an
additional paniculate separation device in the former is obviously not a disadvantage, on an
economic or a technical basis.
If you have data to support the statement that organic emissions may be greater from this
process, present it; otherwise eliminate the statement.
5. Page 5. paragraphs: We believe the real issue is that with counterflow plants, the air from
the mixing drum acts as secondary and tertiary air for the combustion process which
eliminates many of the organic compounds through combustion. This point needs to be
brought out. Further, is there data to support the statement that a counterflow plant can
process RAP at ratios up to 50% with little or no observed effect on emissions? If not, the
statement belongs in industry sales literature and not in an EPA document.
Similarly, the comment about "improved thermal efficiencies" begs the question of what is
being compared to what and what is the degree of improvement. The statement belongs in
industry sales literature and not in an EPA document.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
12
6. Page 7. paragraph 1: The statement that RAP with new aggregate and fresh asphalt is used
to "produce a high quality grade of HMA" is another salesmanlike statement. The point is
that the HMA industry produces products to meet a specification. If the specification requires
the use of RAP, it is used; if not, RAP is not used. The point you want to make is that the
use of RAP does not impair the ability to meet the specifications for certain grades of HMA
since the specs might actually require that RAP is used. Or is EPA implying that the industry
used RAP when specifications call for the use of only virgin materials?
7. Page 7. paragraph 5: The fourth line from bottom uses the word "gas" when " vapor of
organic compounds" would be more correct.
8. Page 8. paragraph 1: The second line states that both the low and high molecular weight
compounds "may" contain small amounts of HAP. We were under the impression that such
asphalt-derived compounds always contain some HAPs. Is EPA suggesting that certain
asphalts don't contain any HAPs? Is there any data to support this statement?
9. Page 8. paragraph 3: In the fourth line, a particle size analysis of cold feed is presented. We
don't see the relevance of this sentence since the previous sentence and the subsequent
sentence both talk about fugitive dust and cold feed is not a maj or contributor to fugitive dust.
10. Page 8. paragraph 4: If as you say in the last sentence that data are not available to support
the hypothesis, EPA should not be repeating the hypothesis. Again, such a statement belongs
in industry sales literature and not in an EPA document.
11. Page 9. paragraph 1: Most of this paragraph is again a replay of industry assertions. The only
relevant issue is whether or not these process modifications work to reduce "blue smoke".
We notice that the industry favorite, the wet electrostatic precipitator or smog hog is
conspicuous by its absence from this list.
12. Page 9, paragraph 2: A repeat of the comment on page 5 about how such plants "will likely
have lower emissions". If you have the data, show it; if not, let the industry do its own
PR.
13. Page 9, paragraph 4: The last three lines appear to be a plug for one particular firm
"Compliance Monitoring Service". Is it EPA's role to do this?
14. Page 10. paragraph 2. 3: The comments here are a major problem. The methods used to
apply the correction for volatility are contrary to the agreement that was reached between the
EPA and various citizen groups. The citizens had argued that the asphalt specifications allow
the use of asphalt with a much higher weight loss than that suggested by limited sampling.
The older AC specifications allow for a 0.5% loss on heating, which can be waived by an
engineer on site who can then allow the use of an asphalt with up to 1% weight loss. The
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
13
new Superpave specifications allow for a 1% loss on heating. In view of this and the
industry' s use of various additives and diluents which would increase the vapor pressure after
sampling for the measurement of the weight loss, the citizens argued and the EPA agreed that
maximum number of 0.5% should be used at all times. Note that this was a compromise since
the newer specification allows the use of asphalt with twice that amount of volatiles. This
issue is particularly important since the measurements by State agencies quoted here appear
to be for asphalt prior to blending with additives. The EPA, by providing an equation where
a hypothetical RTFOT value can be plugged in, in encouraging misrepresentation by a
proponent of any new asphalt plant, who, once the permit is received on the basis of a low
RTFOT value, can then switch to an asphalt with a different RTFOT value. This entire
paragraph has to be rewritten with -0.5% weight loss and 325° F or higher temperature.
15. Page 11. paragraph 1: The industry's own MSDS sheets provide a molecular weight of
asphalt as "over 2000" ( a number we personally don't believe). It is refreshing to see EPA
use a lower number of 345.69
16. Page 11. paragraph 2: What exactly is reference 1 ? Is it the main report with a different date
and number? Also, the derivation of the factor is properly described, not in Reference 1 but
in Appendix B.
17. Table 1: There are many problems here. Footnotes k through p refer to Reference 1. That
is incorrect. If you want to refer to Appendix B, provide specific page number and citation.
We are totally confused by the data on condensable inorganic and organic PM. These are
defined in footnote b as being method-driven. It would have been helpful if additional details
had been presented. In particular, we are curious about the chemical components of the
condensable inorganic PM fraction. The table shows that uncontrolled emissions are
numerically the same as controlled emissions. Footnote] states that "uncontrolled "values are
assumed to be equal to the maximum controlled value measured". Since the controlled values
of 0.013 and 0.0042 are averages (arithmetic means), and are the same as the uncontrolled
values, is the table saying that the average is the same as the maximum value in the data set?
Please explain. We recall that this problem has been in table 1 in previous versions of AP-42.
18. Table 3: Same problems as Table 1; i.e. use of Reference 1 which does not really contain the
primary data and uncontrolled emissions equal controlled emissions.
19. Table 5: Same problems as Table 1; i.e. use of Reference 1 which does not really contain the
primary data.
20. Table 6: Unlike previous tables, descriptive statistics about the data sets are not presented.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
14
21. Table 7: Same problems as Table 1; i.e. use of Referencel which does not really contain the
primary data. Furthermore, there is no discussion why a batch plant dryer using the same
range of fuels e.g. natural gas, No. 2 fuel oil and other fuel oil and a drum dryer using the
same fuel oil should have different emission factors for NOX and SO2.
22. Table 8: Unlike previous tables, descriptive statistics about the data sets are not presented.
23. Table 14: This and many other tables, especially in appendix B, have footnote "a" obscured
because it is superimposed over the last letter of the table heading. This table has to be
redone so that an actual emission factor is given; not a formula for calculating low emissions
based on fictitious values of asphalt volatility.
Same problems as Table 1; i.e. use of Referencel which does not really contain the primary
data.
F. Page-by page Comments on Appendix B: Emission Factor Documentation for AP-42,
Section 11.1, Hot Mix Asphalt Production, June 2000 Draft
1. Page 1-1. final paragraph: The organization of the entire report is confusing. The last line
refers to Section 5, which is the same as Appendix A. Which is the main report and which
are the subsidiary appendices?
2. Page 2-1. Section title: Reference 1 is to an earlier version of AP-42. Why is an obsolete
version of the same document being used as the reference? Reference 3 is a personal
communication from the head of an industry lobbying group. Is this information in the public
domain? Can we get a copy?
3. Page 2-1. paragraph 3: The key information in this paragraph about the number of plants in
the country and the type of plants being built is not referenced so the reader can't judge the
reliability of the information.
4. Page 2-2. paragraph 1: The key information in this paragraph about 1996 production is not
referenced so the reader can't judge the reliability of the information. Further, since road
building tends to be a cyclical activity (at least locally), production or output over several
years might be a more reliable way to establish these levels than the output in a single year.
Dept of Commerce or USGS/Mineral Industry Surveys probably collect and publish this
information.
5. Page 2-2. paragraph 2 and 3: If the hourly capacity data was averaged from the capacities of
precisely 98 batch plants and 162 drum plants, the numbers are probably correct, unless these
plants consistently misrepresented the capacity. Such misrepresentation is not likely to go
unnoticed since most of the data was gathered during performance tests at full capacity in the
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
15
presence of State agencies to make sure that emissions limits were being met. It is interesting
that "conversations with industry" are all that is needed to supercede this data. Why are we
not surprised!
Overall, this paragraph is attempting to relate hourly capacity to annual production. That goal
is not achieved. The discussion in paragraph 3 is totally confused and needs to be rewritten.
Hourly production rates are easy to measure. If the data set referred to in the previous
paragraph is for obtaining operating permits from the relevant State Air Pollution Control
Agencies, as is the case in Massachusetts, the plant is supposed to be running at or close to
the design hourly capacity when the emissions are sampled. Therefore, We believe that the
hourly capacity measurements are correct because these numbers are measured during State
compliance testing and should not be superceded by "industry comments".
The discussion should be talking about the fact that batch plants, especially batch plants
without silos for HMA storage, don't operate for as many hours a year as drum plants
because they service customers using small trucks/small volume customers and because of the
difficulties in matching hot mix production with truck arrivals. (Note that because batch plants
store dried, hot stone, there is storage available, but the stored quantity is generally much
smaller than that available with HMA storage silos.) In contrast, because of their large
production contracts, drum plants operate for more hours each year and the ability to store
HMA in silos also facilitates long production runs. The entire paragraph keeps talking about
hourly production rates when the discussion is really trying to focus on annual production,
i.e. hourly rate times hours operated per year.
Also note that the key assumption is that drum plants produce 90% more HMA than batch
plants ON AN ANNUAL BASIS. This assumption is not backed up with anything; no
footnotes or anything. If this is EPA's assumption, say so. Interestingly, in the main report,
this assumption transforms into a finding that drum plants produce 100% more than batch
plants on an annual basis.
If one assumes that the empirical 80/20 rule applies, the average annual output from batch
plants would be 108,000 tons per year and that from drum plants would be 194,000 tons per
year. This means that the purported annual industry output of 500 million tons is shared
almost equally between batch and drum plants.
It should also be noted that there is no information provided on distribution between small
jobs and large jobs, which would also provide information on how the total production is split
between batch and drum plants. Is it possible that this is because industry information was
provided by a lobbying group supported mainly by the large producers? This means that the
production information has a bias favoring large producers and making them appear more
important.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
16
6. Page 2-3. last paragraph: The writing should make clear whether the RAP is added to the hot
storage bins or directly to the mixer/muller.
7. Page 2-4. last paragraph: This is a repeat comment in the last line. If you don't have the data
about organic emissions, don't speculate about "potential".
8. Page 2-7. paragraph 3: Same comment as above under item Tregarding dataless conclusions.
The last sentence belongs in an industry sales brochure, not in an EPA document.
9. Page 2-10. paragraph 4: The last sentence belongs in an industry sales brochure, not in an
EPA document.
10. Page 3-2. paragraph 1: The sentence states that emission data must be from a primary
reference. In chapter 2, however, secondary references and anecdotal comments are used to
supercede primary reference data on hourly capacity.
11. Page 3-3. final paragraph: The comments here refer to variations or fluctuations in measured
results which could not be explained readily. While we agree that such results might be given
a lower rating, in many cases, variations are a normal part of the production process. For
example, if a batch plant is producing a mix with a high RAP content, the crushed stone is
heated to a high temperature to provide the heat for evaporation of the moisture associated
with RAP stored in the open. If another customer comes in for a RAP-free mix at this stage,
the hot mix will be loaded out at a much higher temperature and will emit more than the usual
amount of fugitive organics. This is a normal transient.
There are many factors which contribute to the emissions which were not measured. For
example, SO2 emissions from combustion are dependent on sulfur in the fuel. The emission
factors for different fuel types are a result of the maximum sulfur content allowed under the
fuel specifications. No such analysis is presented. The same comments hold true for organic
fugitive emissions being dependent on volatile content of the asphalt.
12. Page 4-2 to page 4-117: On these pages, each reference is reviewed and commented upon.
There are several problems with this section. Initially, section 4.2.1.1 discusses reference 1,
section 4.2.1.2 discusses reference 2 and so on. By the time one reaches the end of this
chapter, the numbering is no longer synchronized. Section 4.2.1.327 refers to reference 354.
In between, some references are not discussed (42, 43, 115, 116, 120, 127, 131 and so on).
In other cases, missing references are discussed, for example reference 258. In addition, the
same reference numbers refer to different documents. For example, on page 4-6, during the
discussion of reference 23, references 26 and 27 are discussed. One has to read the citations
carefully to realize that these are references from the document called reference 23. But the
exact citations for 26 and 27 are not given. A careful reading also shows many
inconsistencies. In some cases, the data is downrated because some information is missing;
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
17
in other cases, it is not. See for example reference 181 where the data is still rated A although
information about RAP processing was not provided. On page 4-106, it appears that the
cross references are incorrect.
The section ends with the review of reference 354 on page 4-117. This review resumes on
page 4-167. This is poor formatting for a report.
13. Pages 4-118 through 4-147: There are two sections in this report: 4.2.4 Results of Data
Analysis and 4.3 Statistical Approach which more or less cover the same ground. They
should be combined so that the exposition is complete and consistent.
In many cases, common sense should tell us that the emissions to the atmosphere are either
a result of the performance of a control device or the result of the presence of a pollutant in
the feedstock. A good example of the former is the PM emission factor from dryers
controlled with a fabric filter. Here, the emission factor for PM is not dependent on fuel type.
An example of the latter, is the sulfur dioxide emission factor for dryers. Here, the emissions
are dependent on fuel type since the control devices do not remove any substantial amount
of this pollutant. Yet, these issues are never discussed.
14. Page 4-119. paragraph 1: Was there any attempt to correlate capture efficiency in venturi
scrubbers with the pressure drop?
15. Page 4-122. paragraph 4: This paragraph states that the emission factor for sulfur dioxide was
assigned an E rating because the data ranged over an order of magnitude. This might have
been simply the result of the variation in sulfur content of the fuel. In the case of filterable
PM (page 4-118), the range is over two orders of magnitude but the emission factor got an
A rating. A little consistency would help.
16. Page 4-125. paragraph 4: This paragraph states that the emission factor for trace metals was
assigned an E rating because the data ranged over two orders of magnitude. This might have
been simply the result of the variation in trace metal content of the feed materials and fuel.
17. Page 4-126. paragraph 2:. This paragraph points out the inconsistencies in the approach
taken in several places in the report. Because there was no analysis of the feed materials or
of the fuel used, it is not clear whether the measured trace metals are present on the site or
are a result of sample contamination in the lab or elsewhere or whether the detection of trace
metals at the exhaust and non-detection at the entrance to the control device represents
random error in the measurements. The approach of discarding inconvenient data is not
acceptable.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
18
18. Page 4-135. last line: We are told that the statistical analysis is restricted to references up
to number 338. Why? Several of the subsequent references, up to reference 355, contain the
same type of data as that being analyzed and should have been incorporated.
19. Page 4-13 7. paragraph 1: The description of the box plots is extremely short and incomplete.
Obviously, the authors do not intend to make it easy for readers to understand these plots.
Technically, what is shown is a "Box-and-Whisker" Plot, which is used to show
characteristics of univariate data. The data is ordered and ranked and then divided into four
equal parts. For example, a data set of 20 points would be divided into four portions, each
containing five data points. The box encloses the middle 50% of the data. The text states
that the whiskers "represent adjacent values". Our understanding is that the whiskers either
enclose the entire data range, i.e. the upper whisker shows the range of values for the higher
25% of the data and the lower whisker encloses the range of values for the lower 25% of the
data. Alternately, in modified Box plots, the upper whisker is drawn to the highest point
within 1.5 times the interquartile range and the lower whisker is drawn to the lowest point
within 1.5 times the interquartile range. The interquartile range is the difference between the
75th and 25th percentile values. Values outside the range bounded by the whiskers are plotted
individually. None of this has been bought out in the terse description.
20. Page 4-137. paragraph 3: The paragraph lists the statistical analyses performed. However,
the box plots shown are for only a few of the analyses listed. Where are the remaining plots?
21. Page 4-137. paragraph 5: This paragraph refers to Appendices B & C. Where are they?
22. Page 4-138. paragraph 2: This paragraph refers to Appendix A. Where is it?
23. Page 4-139. paragraph 1: This paragraph refers to Figure 4.2. Note that Figure 4.7 has an
identical caption. The two figures should be labeled properly. One has to read the text very
carefully to realize that Figure 4.2 refers to batch plants and 4.7 refers to drum plants.
24. Page 4-139. paragraph 2: There is a discussion of attempts to correlate wet scrubber
performance with other parameters. Any engineering handbook will tell you that scrubber
performance will correlate with pressure drop, yet this measurement was either not made or
not used. As it now stands, this paragraph finds that although the emission factor for venturi
scrubbers is less than half that for other, unspecified scrubbers, this result is not statistically
significant.
25. Page 4-141. paragraph 1: This entire paragraph is confusing. Figure 4.3 shows that there
is a relationship between RAP content and condensable inorganic PM. The relationship
appears to be non linear. Alternately, looking at just the 0% RAP data points, it is clear that
there is great variability in the condensable inorganic PM emissions even without RAP.
Without any analysis of feedstocks and fuels, the source of these condensable inorganic PM.s
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
19
can't be isolated. To add further confusion, the last sentence jumps to a discussion of carbon
dioxide emissions. The reference to Figure 4.3 in context of CO2 emissions is incorrect.
26. Page 4-144. paragraph 2: The first sentence is incomplete.
27. Page 4-145. paragraph 1: The invocation of engineering principles in the fifth line is
interesting because they have been ignored in many instances up to this point; e.g. discussion
of scrubber efficiencies without consideration of the pressure drops. VOCs, by definition,
don't condense even when cooled to ambient temperatures. Therefore, if a control device has
an effect on VOC emissions, this attenuation must be a result of a mechanism other than
condensation, such as adsorption on solids. The data quoted show lower emission factors (by
a factor of 4) for fabric filters compared to wet scrubbers, but the authors find this result
surprising. Although they don't tell us which engineering principle they are referring to, one
has to assume that they are referring to the ability of water to capture gases through
dissolution. Unfortunately, the components of VOCs are only sparingly soluble in water and
the pressure drops/retention times in such scrubbers are quite modest. Therefore, this is a bad
place to invoke such engineering principles when they contradict the data.
28. Page 4-147. paragraph 2: The comment in the last line is that emissions are a function of
RAP content and production rate. Yet, earlier on page 4-139, these same conclusions were
rejected because of low correlation coefficients. Which conclusion is correct?
29. Page 4-147. paragraph 4: This is the beginning of the review of the more recent data on
load-out emissions. As noted earlier, there is a change in format of the report at this point
in that the discussion of data sets is split into three separate parts.
30. Page 4-148. paragraph 1: Reference 6 (fourth line) is incorrect. What is the proper
reference? We are also amused that this report has been given an A rating for data because
of the "few problems in the Technical systems Audit Report" etc. Unfortunately, this audit
report missed all of the errors found by the citizens. Further, the citizens caught the errors
through spot checking and there are no guarantees that they caught everything. If such errors
can exist in a report that was guaranteed a high degree of scrutiny, how sure is the EPA that
the other data is error free? We am particularly concerned about this point because the data,
which is supposedly in appendices A, B and C, is not attached to the material sent to us for
review. Also, it raises the question of whether EPA staff do any of their own reviews and
spot checks before accepting a report from a contractor. It does not appear to be the case.
31. Page 4-148. paragraph 2: The discussion should point out that Plant C is atypical of drum
plants in the industry. Figure 4.6 shows CO2 emissions factors for drum plants as a function
of production. Although there is considerable scatter, it is interesting to note that all high
emission points are associated with lower production and none are associated with high
production. Does this behavior extend to the emission of other pollutants?
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
20
32. Page 4-150. paragraphs: There are several fundamental issues here that need to be explained.
First, EPA argued for a long time that the tunnel at Plant C was adequate to capture all the
emissions, based on visual observations by EPA - which were contradicted later by
observations by the citizens. EPA reluctantly agreed to use tracers only because the tunnel
did not meet Method 204 criteria. (In discussion of Plant D, We will discuss why Method
204 criteria are inadequate). Next, EPA averaged the efficiency over several runs and used
a single capture efficiency correction. It was only after the citizens showed that capture
efficiency varied during the day, typically declining as the on-shore wind picked up as the land
around the plant heated up, that EPA agreed to correct data for the individual capture
efficiencies. Finally, the so-called correction for truck emissions, discussed later, continues
to be a sham. But that sham is not really discussed here when the quality of data is being
considered. See our comments with reference to page 4-152 below. Also, note that there
is a typographical error in EMD GD-035 " Guidelines for Determining Capture Efficiency".
On page 9, in the formula for p in equation 7, xavg value should be 100.8 not 110.8. The
calculated value of p is correct, however.
33. Page 4-151. paragraph 2: Note that Cambridge Environmental has calculated the cooling
of a mass of HMA. Their results are not supported by any measurements.
34. Page 4-151. paragraph 4: The last line is ambiguous. Silo emissions were measured only at
Plant C. They can't be "combined" with any other measurements.
35. Page 4-152. paragraph 3: This paragraph presents a major problem in this report. The
second line states correctly: "The most reliable method to adjust for emissions measured
during background operations would be to separately adjust each run for the measured
capture efficiency and then subtract these adjusted background emissions from the adjusted
emissions measured during production operations." However, because this "procedure
produces negative values for both the PM and MCEM and many other HAP compounds", a
new unsound procedure is adopted because it gives results that please EPA and industry by
by showing low but positive emissions. This new procedure is to use background emissions
uncorrected for capture efficiency. If the actual emissions have to be corrected for capture
efficiency because the enclosure does not capture all the emissions, there is no justification
for ignoring the capture efficiency for background.
Having used this incorrect analysis, the authors then have the temerity to advise us that this
result might even have a "high bias" relative to the most correct method mentioned above.
In a very narrow sense, this comment might be correct, but in the current context, given the
consistent efforts of the Project Officer to bias the results to favor industry, this comment is
totally off base.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
21
Line 5 states that this situation "cannot be accommodated retroactively." This statement is
not correct. The best way to eliminate improper data manipulation is to discard the phony
background adjustment and report the data as that from loadout plus truck exhaust. Since
silos always dump hot mix into a truck, these combined emissions are present at each silo
loadout point. Further, if a local permit application requires the inclusion of an estimate for
truck emissions on site (and we have yet to see a local permit that requires this), note that the
truck spends about /^ minute under the silo but many more minutes, typically 3 to 8, on site.
As a result, the truck emissions that are included under loadout correspond to the 1A minute
portion. Therefore, the double counting of truck exhaust emissions would be minimal and can
be adjusted for by subtracting the 1A minute from residence time of the truck on site. We
strongly recommend that Tables 4-25 and 4-26 be revised to reflect this approach.
36. Page 4-152. paragraph 4: This paragraph further demonstrates the problems with EPA's
background run. The EPA project officer was interested in recording a high background and
the original data showed that he was able to manipulate truck placement in such a way that
uncorrected emissions in the second half of the run were about twice the emissions in the first
half. If these emissions are corrected for the lower capture efficiency in the second half of the
run, the corrected emissions from the second half are almost three times the corrected
emissions from the first half. This further points out the problems with EPA's manipulation
of and placement of trucks to maximize exhaust emissions inside the tunnel.
We have to apologize for not taking this particular analysis far enough a year ago during a
review of the Plant C draft. When the data for Run 4, the background run, was reviewed, it
was clear that there was data manipulation going on since the raw data showed that emissions
in the second half of the run were double those in the first half. When these emissions were
adjusted for capture efficiency, the emissions from the second half almost tripled. We felt that
the data from the first half of the run was more appropriate than that from the second half.
We did not realize that this data, after correcting for capture efficiency, still overwhelmed and
negated the load out emissions and was therefore also tainted.
37. Page 4-153. last paragraph: In the 10th line, there is an unnecessary caveat about the
relationship being uncertain. The simplest assumption is one of proportionality. Unless lab
scale data are collected that indicate otherwise, this is the best way to deal with the issue. Are
you forgetting Occam's Razor?
38. Page 4-154. paragraphs 1-2: These paragraphs repeat industry propaganda (along with that
from State highway department laboratories who share the same road-building mind set). The
specifications allow for measurements of unblended asphalt and the information represents
the results of tests on unblended asphalt. In the Boston area, the terminal in Chelsea has a
tank holding such additives for blending with the asphalt. After a meeting at EPA-Region 1,
where the industry representatives had denied that any blending took place, Mr. Frank
Singleton, the Board of Health Agent for Chelsea showed me this tank which is clearly labeled
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
22
as containing an asphalt additive. The citizens submitted a list of many such additives from
Roads and Bridges magazine to the EPA/citizen group which was ignored by the EPA Proj ect
Manager, who instead chooses to believe the asphalt industry's lobbyists who insist that such
additives are not used. Also, we have examined the application from one of the terminals in
Chelsea which said that they proposed to add these additives directly to the tanker truck
whereby they would mix in the truck while being transported to the user. This would mean
that a sample taken at the terminal would be a sample of unmixed asphalt.
The third paragraph needs to be deleted. As long as the specifications allow for up to 1%
volatiles in asphalt (even before additives are used), it is misleading for EPA to state that their
approach in Table 4-27 "encourages the use of site specific data". The primary use of AP-42
is to prepare calculations to get permits to build and to operate hot mix plants. Therefore,
there is no site specific data that can be used other than fictional data. Unfortunately, the
industry has a long history of providing low-ball estimates, designed to fool the public and get
permits from agencies that are not particularly anxious to ask difficult questions.
39. Page 4-155. paragraph 2: The statement is that "The industry has stated that good paving
practices dictate that load-out temperatures in excess of 325° F should be avoided." Who has
stated this and in what context? The specification for highways is that hot mix temperature
before the pavement is laid down should be between 325 and 275° F. Hot mix outside this
temperature range can be rejected. The temperature at load out depends on the distance
between the plant and the location where the new pavement is being laid and the size of the
truck since smaller loads will cool faster than larger loads.
On the 13th line, the statement is that "this hypothesis has not been validated by emissions
testing but provides an adjustment that is directionally correct." This comment is totally
unnecessary and needs to be deleted. Is EPA saying that the Clausius-Clayperon equation
should be thrown out? Alternately, is EPA saying that the degree of precision and accuracy
associated with emissions testing methodology is good enough to provide this type of data?
40. Page 4-157. paragraph 2: In this paragraph, the groundwork is being laid to explain why
inorganic PMs were much higher at Plant D than Plant C. This is further elaborated on page
158. We strongly disagree with the conclusion that the additional dust was a result of truck
movements. While Plant D was unpaved, water sprays were used to keep the dust down as
noted in the report and as observed. At Plant C, because of the practice of cleaning the silos
at the end of the day with crushed stone, there was considerable dust present in the tunnel.
Furthermore, one ofthe silos had dumped hot mix on the floor and cleaning it up generated
a lot of dust. Our recollection is that if anything, the tunnel at Plant C was dustier than the
one at Plant D. We believe the explanation for the higher PM load at Plant D is the relatively
short residence time in the mixer at any batch plant. Because of this, during mixing, there are
small pockets of stone dust that are not coated with asphalt and these are the source of the
high PM readings. In contrast, the vapors inside a silo have time to permeate through the
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
23
mass of hot mix in the silo and coat any uncoated dust particles. Once coated, the fine dust
coagulates and is no longer airborne particulate matter. This hypothesis is supported by the
visual observation that dumps from the mixer at Plant D were dustier than those at Plant C
when observed from inside the tunnel.
41. Page 4-15 8. paragraph 2.3: See the comments above. The adjustment is indeed "speculative"
and should be eliminated. Separate emissions factors should be used for inorganic PM
emissions from batch and drum plants.
42. Page 4-159. paragraph 2. 3: There is a fundamental problem with the two paragraphs which
has not been explained. MCEM PMs and TOC are generated the same way and from the
same source. Therefore, if MCEM generation is a function of loadout time, TOC generation
should also be a function of loadout time. Of course, the comparison is further complicated
and made meaningless by the bogus "background" correction discussed earlier. To eliminate
the confusion, separate emission factors should be prepared for batch and drum plants in all
cases.
43. Page 4-159 - 162: As discussed earlier, the entire loss on heating discussion should be
rewritten so that a RTFOT value of -0.5% is used in all cases. This way, the EPA will not
be providing yet another opportunity for builders of new plants to misrepresent what might
happen once the plant starts operating. The same applies for the formulae which promote the
use of variable temperatures.
44. Page 4-163. paragraph 1: The first line states that "vapor pressure information .... is not
available". The report fails to state that this was only because of EPA's refusal to collect this
information.
45. Page 4-163. paragraph 3-5: Please explain what you mean by "pegged TOC readings"?
As it stands we can't understand this at all. Furthermore, Figures 4-9 through 4-13 are
missing from our copy. Finally, it is not clear whether you are referring to Plant C or Plant
D in this discussion, though the context suggests Plant C.
We further find that the contention that the estimate of unmeasured emissions being an "upper
bound" to be unpersuasive and incorrect. Our reasons for saying so result from the
inadequacies of Method 204. These were outlined in Ravi Nadkarni' s letter of September 15,
1999. They are:
a. At Plant C, two precautions were taken to ensure that we could measure or estimate
the amount of organic fugitive emissions that did not reach the primary sampling
ports. The first was the use of sulfur hexafluoride tracer gas, and the second was the
use of impingement/deposition plates in the ducts which collected organic parti culates
which would be deposited upstream of the sample port. Although all the emissions did
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
24
not reach the sampling ports (some was deposited inside the tunnel and some was
pumped out of the tunnel by ambient breeze or by truck movement), the tracer gas
enabled an estimation the VOC portion of this loss. (As was pointed out during the
planning, the tracer gas was a good proxy only for non-condensible emissions but not
for condensible emissions.)
At Plant D, a tracer gas was not used although requested by the citizens because the
TIE was constructed to meet Method 204 criteria, yet visual observations and THC
readings showed the following:
b. An ambient breeze was sufficient to cause some of the fume to escape from the top
or the bottom openings in the downwind door of the tunnel. Therefore the size of the
openings was decreased further. This means that the Method 204 criterion of
maintaining over 200 fpm at a natural draft opening is not adequate. Note that 200
fpm is less than 2.3 miles per hour, which is not much of a breeze. Therefore it is not
surprising that these criteria are inadequate to assure total containment within the
TTE. We observed visible emissions of fume from the entrance, further emphasizing
that the Method 204 criteria are inadequate.
c. Further, Method 204 has a more serious and fundamental problem. Although the size
of natural draft openings is specified, the method does not contain any criteria to
ensure that the emissions are pulled/sucked past the sampling point in a reasonable
period of time. As is well-known, residence time calculations, based on enclosure
volume and fan capacity, generally provide incorrect information by underestimating
the time to evacuate the enclosure because of channeling. Note that Method 204 does
not require a specific location for the natural draft openings, a problem that was
pointed out during the planning sessions. At Plant D, even after the size of the
opening was decreased, and the little pieces of colored tape at the Natural Draft
Openings were indicating airflow towards the inside of the TTE, the tunnel was not
evacuated in the approximately 15 second time gap between the final drop and the
opening of the doors. This can be clearly seen in the fact that THC readings did not
drop to zero even when the sample averaging time is less than 1 minute. This fact
was also observed visually inside the TTE enclosure since SVOCs are visible. Thus,
in the absence of a tracer gas, we have compelling evidence that the collected sample
had a low bias. For example, the extended period test results in MRI-D-Table 4-4,
show that final concentration did not drop to zero but stayed between 1 ppm and 2.1
ppm . Yet, this low bias is not mentioned anywhere in the report nor in Appendix B
of the MRI Report which contains the original data. These emissions did not reach the
instrumentation but were emitted to the environment causing a low bias.
46. Page 4-164. paragraph 2: There is reason to believe that the numbers derived are low. We
base this on several facts: Material Safety Data Sheets contain the following warning:
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
25
" Studies have shown that low flash point substances, such as hydrogen sulfide and low-boiling
hydrocarbons, may accumulate in the vapor space of hot asphalt tanks and bulk transport
compartments. Such vapors may exhibit flammability characteristics of a significantly lower
flash product than would be indicated by the open cup test." A review of light hydrocarbon
compound properties in the Handbook of Chemistry and Physics indicates that the lower
explosive limit is reached at vapor concentrations around 1%. Therefore, this warning makes
sense only if the TOC concentration is around 10,000 ppm, not the 2000 ppm level used.
Second, storage tank temperatures are generally higher than 325° F and closer to 350° F.
Finally, the assumed vapor pressure is low in comparison to data from Nelson's Petroleum
Refinery Engineering, which would indicate a vapor pressure over 10 mm Hg.
47. Page 4-166. paragraph 4: This needs to be rewritten. It is obvious that this paragraph is
recycled from some other writeup. There is no relevant "response 53" in the present writeup.
Further, emissions are not only dependent on temperature, they are also dependent on
convective effects, i.e. evaporation rate increases in the presence of strong convection.
48. Page 4-166. final paragraph: This paragraph presents a single emissions factor for yard
emissions. This is exactly what needs to be done with the RTFOT equations. They should
be eliminated and only a single factor, based on RTFOT of 0.5% presented in AP-42.
49. Page 4-167. paragraph 2 - 4: This is a critique of Ravi Nadkarni' s original submission in 1994.
There are several problems here. If you are going to critique this work, you should refer to
all of submissions to the EPA, not just the first one from 6 years ago. The complete list
would include his response to Mr. Ryan of EPA of October 26, 1995 and November 16,
1995 and to Mr. Mobley of July 5, 1996. We find it particularly interesting that the estimate
of vapor pressure of asphalt from 1994, based on several model compounds has been
represented as one based on a single model compound. Second, "people who live in glass
houses should not throw stones". In using the TANKS program, (see page 4-165) the
authors used another set of model compounds whose presence at the required concentrations
is also not supported by the same data that is used to attack this work. For example, if
naphthalene and 2-methylnaphthalene are predominant species of PAH in emissions from
asphalt DRYERS, why were tetracosane and pentacosane used as model compounds for
TANKS? Can you please recalculate the TANK emissions using naphthalene and 2-
methylnaphthalene?
Next, why is the Cambridge Environmental letter not listed in the references? It should have
been reference 358. Cambridge Environmental used another set of model compounds to
come up with a different figure. That does not make their calculation any more valid since
both estimates were done years ago in the absence of data on components of asphalt vapor.
(This also begs the question of why DRYER emissions are used here when better data is now
available from silo emissions.) Given Cambridge Environmental's track record as an extreme
advocate for their industrial clients in the public health area, we would not characterize their
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
26
work as "independent". (In one case, they compared emissions from a single plant versus
emissions from all industrial plants of all types in a county and the state. Even the most
polluting plant can appear to be a winner under that criterion.) There is more to the
Cambridge Environmental story. In a public meeting in Uxbridge, Dr. Zemba, the principal
author of the letter, reversed his position and stated that he now believed that the calculations
are in the correct range, or only modestly overestimate the actual emissions. Specifically, he
said, "When I originally looked at the calculations, I thought that these were grossly
overestimated. I can see your point though and they might not be overestimated by as much
as I originally thought." He also admitted that he had not studied Ravi Nadkarni's two 1995
letters to the EPA. A specific critique of the Cambridge Environmental work is in the 1996
letter to David Mobley.
We agree that calculations done in the absence of actual measurements should be discarded
when actual measurements are available. But any critique should fairly represent the work
and its shortcomings.
50. Page 4-168. paragraph 2-4: The evaluation of reference 3 59 indicates that the reference was
not read critically. Tables 4-42 and 4-43 are presented without any critical evaluation of the
contents. The table shows about the same concentrations at an upwind site and a downwind
site. How is this possible, unless the measurements represent background readings? There
is no comment as to why higher concentrations were detected at a residential site and how
the residential site compared to the upwind and downwind sites. If you are not going to
evaluate the information, don't present the tables.
51. Page 4-169. paragraph 2-6: Again, the comments indicate that reference 360 was not read
critically. For example, the high outliers were rejected. Having read the original report, we
know that the authors don't plead that the instrument malfunctioned. Therefore, the high
measurements must be explained. Or are they supposed to be background? It is likely that
the wide variation in measured concentrations was a function of measuring technique? Low
readings were obtained upwind of the emission source and high readings were obtained near
the source and downwind from it. Under these conditions, when the instantaneous wind
direction and sampling location with respect to the source were not recorded, there is no
justification for rejecting the high data points. Note that the determination of the high data
points as outliers was based on a purely statistical argument (the "outliers" were many
standard deviations away from the rest of the data; therefore they don't belong with the rest
of the dataset); not one based on measurement errors or other relevant factors.
The final paragraph again uncritically quotes the process used by the authors to convert the
measurement into a mass rate in g/sec. The concentration of benzene was measured in
concentration units i.e. g/1 which can be converted to g/m3. This was then converted to an
"emission factor" (actually a mass rate) by multiplying the concentration by wind velocity in
m/sec and the vehicle area in m2. This is data manipulation of the worst sort. If one is trying
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
27
to measure the flux of benzene escaping from the truck, the flux would be expressed in units
of mass flow per second per unit area or g/sec-m2. The relevant area is that perpendicular to
the direction of flow of the flux that you are trying to characterize. In this reference,
concentration, wind velocity and truck area are combined in a way that might be
dimensionally correct but make no physical sense. In particular, it is difficult to see how a
measured wind velocity relates to the area of the truck. EPA does disservice to the reader
when material of this type is presented in detail over 4 tables without critical evaluation.
52. Page 4-172. reference 364: This reference is the AIRx report of measurements at Plant C.
We want to see exactly what procedure was used to estimate capture efficiency at 70 to 90%.
Otherwise, this comment should be eliminated.
53. Tables 4-4 to 4-14: There are numerous deletions of data shown in these tables where
emission factors for each plant are presented. The reasons for the deletion are not explained
in the text.
54. Table 4-15: When will this table be completed?
55. Table 4-19: Footnote a is incorrect. R2 is the correlation coefficient, not the "squared
correlation coefficient." We believe the terminology calls R the coefficient of determination,
so that R2 could be called the "squared coefficient of determination". The more common
usage is to stick with "correlation coefficient" or R2.
56. Table 4-23. 4-25: In view of the comments earlier about the background run, these tables
need to be revised.
Comments by Ravi Nadkarni & Lloyd Pillion August 31, 2000
-------
Elane Kruger, Chief
Environmental Toxicology Unit
Bureau of Environmental Health Assessment
Massachusetts Department of Health
Boston, MA
-------
The Commonwealth of Massachusetts
Executive Office of Health and Human Services
Department of Public Health
250 Washington Street, Boston, MA 02108-4619
ARGEO PAUL CELLUCCI
SOVERNOR
JANE SWIFT
LIEUTENANT GOVERNOR
WILLIAM D. O'LEARY
SECRETARY
HOWARD K. KOH. MO, MPH
COMMISSIONER
September 7, 2000
Robert F. McConnell
Environmental Engineer
Air Quality Planning Group
United States Environmental Protection Agency
Region I
John F. Kennedy Building
Boston, MA 02203-0001
Dear Mr. McConnell:
Thank you for extending the deadline for submitting our comments. The Massachusetts Department of
Public Health (MDPH) Bureau of Environmental Health Assessment (BEHA) is "Writing this letter in
response to the U.S. Environmental Protection Agency's (EPA) request for comments on the draft
emissions assessment report and draft AP-42 section regarding emissions from hot-mix asphalt plants
(i.e., "Hot Mix Asphalt Plants Emission Assessment Report" and "Hot Mix Asphalt Plants AP-42
Section 11.1"). These documents are the culmination of work done by EPA to characterize uncontrolled
emissions during loading operations at hot mix asphalt plants. As well as reviewing asphalt plant test
reports and related documents, EPA performed its own emission testing at asphalt plants in Irvine,
California, and Barre, Massachusetts. Using these data, EP A produced emission factors which can be
used to estimate area-wide emissions from many facilities as welt as emissions from specific facilities.
Previously, MDPH had submitted comments concerning the two-part protocol for emission testing at the
hot mix asphalt plant in Barre, Massachusetts, and the 8-volume draft report on results for emission
testing at the Barre and Irvine plants. MDPH comments focused primarily on whether speciated
emissions data, which were gathered solely from the California facility, were applicable to the Barre
facility due to differences in modes of operation and material between the two plants and whether dust
particles collected at the elbow junctions of the Barre plant exhaust system were included in the study.
In response, EPA explained that limited resources precluded conducting analyses at the Barre plant to
better speciate emission data. However, EPA also stated that applicable specific emissions data for
asphalt could be gathered from previous asphalt studies. In addition, EPA stated that those dust particles
that were deposited at the elbow junctions were recovered and analyzed.
-------
As we have previously noted, it is important to understand the limitations of the emission testing at the
Barre and Irvine plants, some of which were discussed in EPA: response to comments (i.e., "Emission
Tests of Hot Mix Asphalt Plants, Response to Comments for Asphalt Plants C and D"). EPA noted that
resources were not available to perform emission testing at different times of year or at different
locations to determine whether there is a relationship between emissions and asphalt cooling or
variability of emissions of specific chemical compounds. This information could be important in
determining emissions from plants in different areas of the country. While we appreciate that limited
resources may have prevented more comprehensive testing, unfortunately a number of uncertainties
remain regarding opportunities for exposure and health concerns. In addition to those noted above,
uncertainties also exist due to the variations in modes of operation, materials, and plant designs.
Therefore emission data from previous asphalt studies may not be applicable to the Barre plant. We
would recommend that these important limitations; be outlined in the final document, rather than just in
the response to comments.
A final comment regards EPA reporting averages of data collected without presenting the range of
numbers. This concern has also been raised by others. The use of average values may be appropriate
for evaluating risks of long-term health effects e.g., cancer risk, but may underestimate the potential for
health effects associated with other issues, e.g., short-term odors, that have been reported to be
associated with these types of facilities. Clearly, some of the chemicals, e.g., naphthalene, contained in
the asphalt have very low odor thresholds. For that reason, we recommend that EPA present the full
range of predicted emissions data including upper-bound or maximum estimates of emissions as they
relate to the potential for shorter term effects, such as odors. This is particularly important given that
odors have been important issues at other facilities like this elsewhere in the U.S. (e.g.) Michigan).
We appreciate the substantial amount of technical work that was undertaken by EPA and others in
regard to this project and applaud EPA's efforts to freely discuss the results of this with a variety of
concerned individuals and organizations. Thank you for your attention and consideration. If you have
any questions, please do not hesitate to call me or Michael Celona at 617-624-5757.
Sincerely
Elaine T. Krueger, Chief
Environmental Toxicology Program (BEHA)
cc: Suzanne K. Condon, Assistant Commissioner
Martha J Steele, Deputy Director, BEHA
Michael Celona, Environmental Analyst, BEHA
-------
Gary Fore, Vice President
Environment and Safety
National Asphalt Pavement Association
Lanham, MD
-------
NATIONAL ASPHALT PAVEMENT ASSOCIATION
NAPA Building P 5100 Forbes Boulevard P Lanham, Maryland 20706-4413
Toll Free: 1-888-468-6499 P Tel: 301-731-4748 P Fax: 301-731-4621
Mike Acott, President
August 17, 2000
Bob McConnell
EPA Region I
Q Congress Street/CAQ
Suite 1000
Boston, MA 02114
Reference: Draft Hot Mix Asphalt Plants -Emission Assessment Report;
Draft AP-42 Chapter 11.1; Background Report
Dear Mr. McConnell:
We have reviewed the above referenced draft reports and offer comments and
suggestions for your consideration.
This letter will be broken into two sections. Section I will discuss the suggested
corrections and comments per above referenced document. Section II will
discuss additional stack test report data. This data offers significant additional
information relative to Benzene and Sulfur Dioxide (SO2).
Section I
Chapter 11.1 Hot Mix Asphalt Plants (Draft 2000)
General comment - Tables tend to be confusing. For example, Place HAPs
contiguously with HAPS and follow with Non-HAPs together. Also, suggest
adding a set of batch table and drum tables with contiguous HAPs listing, non-
HAPs listing and metals HAPs vs. non-metal HAPs listing.
General comment - Suggest following the same table order as found in the
Emissions Assessment Report. Much less confusing.
>- Page 11.1-1
A. Hot mix asphalt paving materials can be manufactured by: ... (2) continuous
mix (mix outside dryer drum) plants, ...
Please add the word dryer when using the "mix outside dryer drum" phrase.
This will allow for both past and present types of plants.
email: napa@hotmix.org OL£D 100% RECYCLABLE www.hotmix.org
-------
August 17, 2000
EPA -OAR
Page 2
Page 11.1-3
A. Add in (source classification code in parentheses) under Figure 11.1-1
General...
Page 11.1-4
A. Add to Figure 11.1-2. General process 1 low diagram for parallel-flow drum
mix asphalt plants.
B. Add in (source classification code in parentheses) under Figure 11.1-2
General...
Page 11.1-6
A. Add in (source classification code in parentheses) under Figure 11.1-3
General...
Page 11.1-7
A. In the fourth line, second paragraph under 11.1.2.1 - Batch Mix plants,
please add the word very before the word small. The sentence will then read
"and very small amounts of organic compounds"
Page 11.1-8
A. In the first partial sentence on the page, please add prior to tarping after
truck. The sentence will then read: Organic vapor... bed of the truck prior to
tarping,
B. In the second line from the top of page - please add the word very before
the word small. The sentence will then read "may contain very small
amounts of HAP"
Page 11.1-9
A. In the first paragraph under section 11.1.2.3 - please add the word very
before the word small. The sentence will then read "and very small amounts
of organic compounds of various species"
-------
August 17, 2000
EPA -OAR
Page 3
Page 11.1-10
A. 2nd paragraph -Suggest a more typical loss-on-heating of 0.25 not 0.41
a. To estimate total PM emissions from drum mix plant load-out operations
using an asphalt loss-on-heating of 0.25 percent and temperature of
290T, HMA mix, the following calculation is made:
EF = 0.000181 + 0.00067 (-v)e ((a°251) (29°+ 460) '2a43)
0.000181 + 0.00067 (-(0.25))e «a°251>(290 + 460> -2a43>
0.000181 + 0.00067(0.25)6 (-1605)
0.000181 + 0.00067(0.25)(0.2009)
0.000181+0.0000485
= 0.000215 Ib/ton of asphalt loaded
NOTE: The correlation factor in the equation should be 0.00067 and not
0.00059.
This loss-on-heating value is a value that is seen on a daily basis within
the Industry.
Page 11.1-16
A. Footnote C -Suggest striking SO2 or add 50% attenuation of SO2. Please
review attached stoechiometry and SO2 attenuation data.
Page 11.1-17
A. In the title for Table 11.1-6, please delete the word formaldehyde from title
B. Note there is an error under VOC for natural gas-fired dryer and the No.2
fuel oil-fired dryer. They should be 0.0076 for both not 0.0002
Page 11.1-18
A. Footnote C - Suggest striking SO2 or add 50% attenuation of SO2. Please
review attached stoechiometry and SO2 attenuation data.
B. Footnote D - There is a question that the range to 96 Ib/ton is too high.
Where is source for this number?
-------
August 17, 2000
EPA -OAR
Page 4
Page 11.1-20
A. Sub title: Hazardous air pollutants under Pollutant Heading - For easier
reader understanding, please change subtitle from Hazardous air pollutants to
non PAH/HAPs
B. Please add a line totaling the non-PAHs hazardous air pollutants section.
Currently very confusing.
C. Sub title PAHs - Please change from PAHs to PAH/HAPs
D. Please add row and sum the NO. 6 fuel "organic HAPs"
E. Please change non-HAP subtitle to Non Organic HAPs
F. Sub title: Hazardous air pollutants - Please change to PAH/HAPs
G. Left side column under Dryer with fabric filter - Please add natural gas,
propane, fuel oil, or waste oil fired. Confusing as is and fails to distinguish
between these fuels and No.6 fuel oil.
G. Footnote b - Please change as follows: Natural gas, propane, fuel oil, or
waste oil fired dryer, for pollutants that are marked with a "*", separate
emission factors are presented for No. 6 fuel oil fired dryers. Substitute these
factors for above factors when burning No.6 oil.
Page 11.1-21,22,23
A. Use the same format and comments as suggested for above Table 11.1-9
Page 11.1-22 and 23
A. Please add Dioxins and Furans to above referenced Organics and HAPs
B. Delete uncontrolled from this table because to be consistent with other
HAPs, we should use data collected after filtration in the stack. Please make
a footnote reference that the uncontrolled numbers are in the Emission Factor
documentation for AP-42 Section 11.1 Hot Mix Asphalt Production.
-------
August 17, 2000
EPA -OAR
Page 5
Page 11.1-24
A. Delete uncontrolled from this table because to be consistent with other
HAPs, we should use data collected after filtration in the stack. Please make
a footnote reference that the uncontrolled numbers are in the Emission Factor
documentation for AP-42 Section 11.1 Hot Mix Asphalt Production.
B. Footnote b - Please change as follows: Natural gas, propane, fuel oil, or
waste oil fired dryer, for pollutants that are marked with a "*", separate
emission factors are presented for No. 6 fuel oil fired dryers. Substitute these
factors for above factors when burning No.6 oil.
Page 11.1-27
A. Overall comment - Should create subtitles for Metal HAPs category and a
Metal Non-HAPs categories. Also suggest moving to 11.1-9.
B. Adjust footnotes accordingly.
Page 11.1-28
A. Delete uncontrolled from this table because to be consistent with other
HAPs, we should use data collected after filtration in the stack. Please make
a footnote reference that the uncontrolled numbers are in the Emission Factor
documentation for AP-42 Section 11.1 Hot Mix Asphalt Production.
B. Use the same format and comments as suggested for above Table 11.1-9.
Should be the same for both drum and batch.
C. Please delete Footnote b. unnecessary.
D. To make the table easier to read, suggest switch last two sections of the
table. The section pertaining to lead and mercury should come last.
Page 11.1-29
A. Question why include total PM when it includes organic PM? This is very
confusing.
B. Footnote a. - Please change the word asphalt to HMA mix to prevent
confusion as to mix temperature or liquid temperature.
-------
August 17, 2000
EPA -OAR
Page 6
»- Page 11.1-30
A. Place PAH/Organic PM (%) up above PAHs line in the left column. It
should be a title, not a part of the data.
B. Change Subtitle from PAHs to PAH/HAPs
C. Total PAHs, subtitle - Change to total PAH/HAPs
»- Page 11.1-31
A. Move subtitle - Compound/TOC(%) up to be included in the titles for the
columns.
B. Create Sub title - non-VOC/non-HAPs
C. Please a Percent sign (%) to every EF as in table 11.1-15. It is currently
very confusing because the other tables in this document do not contain
percents.
D. Total - add line total volatile organic HAPs
E. Please add to the example on page 11.1-10. It would be helpful to
illustrate the use of Table 11.1-16.
»- Page 11.1-32
A. Please correct Reference No.2. Katherine is spelled Kathryn.
Emission Assessment Report (Draft 2000)
»- Page 2
A. Paragraph 1, second line -please add the following: ...and PM-2.5 and
very small amounts of hazardous air pollutant...
B. Second sentence - delete the "e" off of the end of with.
C. Second sentence - please add the following: ...VOC and very small
amounts of volatile HAP organic compounds.
-------
August 17, 2000
EPA -OAR
Page 7
D. paragraph 2 - please add the following: ...emissions for specific facilities
where source specific emissions is not available or where source
testing is cost prohibitive.
Page 3
A. At the end of the last paragraph -Please add an editors note: "HAPs
emissions are typical of emissions that result from fossil fuel
combustion."
Page 4
A. Add in (source classification code in parentheses) under Figure 1 Typical
Page 5
A. Add in (source classification code in parentheses) under Figure 2
Page 8
A. 3rd paragraph - ...(2) continuous mix (mix outside dryer drum) plants, ...
Please add the word dryer when using the "mix outside dryer drum" phrase.
This will allow for both past and present types of plants.
Page 10
A. 2nd pai
will likely have organic stack emissions ...
A. 2nd paragraph - Please add the following: ... counterflow drum mix plants
B. 4th paragraph - Please correct as follows: ... CO; and very small amounts
of organic compounds ...
Page 11
A. 1st paragraph - Please add the following: ... organic aerosol may contain
very small amounts ...
B. Last paragraph - Please add the following: An initial screening the of these
documents...
. Page 12
A. 1st paragraph - Please add the following: ... truck prior to these departure
for the job site")". Missing end parentheses.
-------
August 17, 2000
EPA -OAR
PageS
»- Page13
A. 2nd paragraph - Please add to the end of the paragraph. ..."representative
of gas-fired heatlers. As a practical matter theses units represent a very
small combustion source when compared to the dryer.
B. 4th paragraph - Prior to the last sentence please add the following
sentence. ... developed from the data. For example, two scientific
auditors from the Research Triangle Institute were employed to
independently audit the test and reporting process. These additional
steps ...
»- Page 15
A. Last paragraph - Please add the following sentence to the end of the
paragraph. ...equation developed from the emission data. It should be
noted that truck emissions would decrease rapidly with time as a result
of surface cooling and crusting and industry practice of covering loads
with tarps.
»- Page 20
A. Sub title PAH - Please change as follows: PAH (semi volatile HAPs).
B. Please delete line between methylene chloride and MTBE
Section II
The Hot Mix Asphalt Industry is providing additional data specifically relevant to
S02 attenuation and benzene to EPA for possible inclusion into the Emissions
Factors Document (AP-42). Attached are three binders containing stack test
reports. All of the tests were consist of three 1-hour runs and were conducted for
compliance purposes.
NOTE: NAPA is sending one copy of this material and one copy of the
comments to Ron Myers. Only the comments document is being sent to EPA
Region I.
-------
August 17, 2000
EPA -OAR
Page 9
S02
The Draft Emissions Assessment (EA) offers an EF for SO2 of 0.078 Ibs/ton hot
mix from drum mix plants firing waste oil (Draft EA,, Table 11.1-7). The Draft EA
also suggests via footnote (c.) of Table 11.1-7 that AP-42, Chapter 1 also be
used to estimate SO2. Chapter 1 Emission Factors (EF) incorporate fuel sulfur
content to determine SO2 emissions.
It is believed that a more accurate approach should be used to determine SO2
emissions from plants firing oil, especially waste oil.
Several stack tests for SO2 were conducted. Samples of each plant's burner fuel
supply during each test were taken and analyzed for sulfur content. The amount
of SO2 produced during combustion of the burner fuel was calculated. There
was an assumption that all of the sulfur in the fuel would be converted to SO2.
Then, using a mass balance approach, the actual stack emissions of SO2 was
compared with the calculated SO2 as a product of combustion, and determined a
control (i.e., emission reduction) factor for SO2. from a hot mix asphalt plant,
firing oil.
The proposed means of control would be the adsorption of SO2 by dust particles
suspended in the ductwork and captured on the bags in the baghouse. For a
plant producing 250 tons/hour hot mix and firing 500 gallons/hour waste oil
having a sulfur content of 0.046% by weight and a specific gravity of 0.89, the
amount of SO2 generated by combustion would be 500 gal/hrx 8.34 Ibs/gal x
0.89 x 0.46/100 x 2 (Ib S02/ Ib S) = 34.14 Ib SO2/hr. At the same time, the
uncontrolled particulate emissions from the plant would be 28 Ibs PM/ton (Draft
Table 11.1 -3) x 250 tons/hour = 7000 Ibs PM/hour. The available reaction
surface area of the dust particulate provides significant acid gas capture
opportunity by any alkaline solid material suspended in the exhaust stream or
accumulated on the bag house bags.
The results from the 15 tests demonstrate SO2 control ranging from 47% to 97%,
with an average control of 62%. Test result averages show little variation in
control among different plant types, aggregate types or recycle asphalt pavement
(RAP) content in the mix. Please refer to the attached table: "Asphalt Plant SO2
Emissions Summary"
-------
August 17, 2000
EPA -OAR
Page 10
The Wisconsin Department of Natural Resources (WDNR) has reviewed these
stack tests and wil adopt a conservative control factor of 50% for future
permitting and emission inventory purposes. Please refer to the attached letter
from Lloyd Eagan, Director of Bureau of Air Management from the State of
Wisconsin - Department of Natural Resource.
Benzene
Stack tests were also conducted for benzene from drum mix plants, firing either
waste oil or natural gas. The reports from 17 tests are enclosed. These reports
indicate average benzene emissions of approximately 0.0003 Ibs benzene/ton
hot mix asphalt, with no significant variation for fuel type. The draft EA lists a
benzene EF of 0.00051 Ibs/ton (Table 11.1-10).
Thank you for your attention to these matters. If you have any questions or
comments, please feel free to call either Una Connolly or myself at 301-731-
4748.
Sincerely
Gary
Vice President -
Environment & Safety
attachments
-------
ASPHALT PLANT SO2 EMISSIONS SUMMARY
AH Plants Fired Using Waste 01
Plant
UL
Ten
Locations
Plant
Type*
%SuKur
ioJEual
HMA.
IEH
Aggregate
Type,
Ki Limeatojif
%S02
KSNPit
CD
0.46
362
24
Gravel
73
50.5
IS
25
26
27
28
29
34
Cedar Lake Pit
SaukvUla Pit
Michel's Pit
SE of Markesan
Quarry Near
fish Creek
Quarry SE of
Horicon
Pit NW of
Wautoma
Vulcan Pit
Near Dousman
Denmark
53 Newberry, Ml
68 Cold Spring
10 NW Wisconsin
25 NWWisconsin
41 NWWisconsin
CO
CD
PO
PD
PO
PO
PO
CD
PD
PD
PO
PD
46 NW Wisconsin PO
0.37
0.63
0.44
0.53
0.39
0.43
0.47
0.43
0.64
0.54
0.36
0.5
0.47
0.36
287
358
250
303
300
270
282
273
227
245
1 83
1 27
280
272
25
25
1 0
20
20
15
10
Gravel
68
Gravel
79
Quarry
100
Quarry
100
Quarry
100
Gravel
61
Gravel
81
Gravel
74
Gravel
62
Quarry
100
Gravel
30
Gravel
0
Gravel
0
Gravel
10
32.2
64.3
77.8
55.0
52.1
78.8
62.3
56.3
61.5
47.6
97.1
48.1
47.2
50.9
Standard
UMiflQ
IS.88
EC
57.9
CD
63.3
97.1
Statistical Analyses Based on Operational Variables
All values represent Average % SO2 Controlled
%LRAP
0%-14%
BAE
62.6
10%-2S
RAP
61.28
Aggregate
Ixflfi
0%-5Q%
Limestone
60.8
S.l.%-.100%
Limestone
62.5
PO means parallel-flow drum
CD means counter-flow drum
6 means batch plant
C0mp\PJtU«3tSumhAPASQ2 Sit
-------
MATHV
ASPHALT PLANT STACK TEST SUMMARY
n*m
LB.
2
4
s
6
7
8
15
24
25
26
27
28
28
29
31
34
52
53
SS
SS
56
59
63
85
es
63
10
25
41
46
la*
B
CO
CO
CO
CO
CO
CO
PO
TO
TO
PD
PO
PD
M
8
TO
PO
TO
B
CO
PO
PO
CO
PD
CO
PO
•
PO
PO
n
T«*
am
SMS
10/97
was
7»S
7»S
807
W7
to«s
Ocl-M
5/97
10/97
9,-as
Oct-M
t'97
sra*
Oct-SS
e/ss
9 -98
SOI
Sw-93
fc9J
7/9$
S/JS
7»7
JUB-99
7«7
SM-99
Aug-93
J.J-59
Qet-W
fuX
SMK
NG
WO
NQ
NG
WG
wo
wo
wo
wo
wo
wo
wo
wo
wo
HO
wo
wo
wo
NO
NG
NC
WO
WO
NQ
WO
WO
WO
WO
WO
KM A.
350
3«2
300
395
29!
28 7
35B
290
2iO
303
100
244
270
262
175
S75
1SS
227
260
450
292
240
420
319
450
24S
183
127
280
272
»*E
15
24
20
20
20
2S
25
24
0
10
20
24
0
20
0
IS
0
0
IS
20
15
0
20
21
25
10
0
0
0
0
UM/Tofl LbiTim tM/Ta* IbtfTen
3SE neon £0 jiajs
9.04
0.0099 0.0104
0.0012 0.0048 0.1053 0.017
0,0023 0.0021 0.08S 0.0228
o.ccsa o.occ
-------
AUG-16-2000 14'-27 PflYNE DOLfiN P.02/82
State of Wisconsin \ DEPARTMENT OF NATURAL RESOURCES
,, - - 101S. Webster St
Tommy G. Thompson, Governor
~*s*^^s^***^j^s*^>
WI8CONSIN~^
DEFT. OF NATURAL RESOURCES
George E. Meyer. Secretary Madison, Wisconsin 53707-7921
Telephone 008-266-2621
FAX 608-267-3579
TDD 608*267-6897
August 8, 2000
Peter Tolsma
Construction Resources Management, Inc.
P.O. Box 1632
Waukesha,W153187
Subject: Determination of a Sulfur Dioxide emissions factor for hot mix asphalt plants
Dear Mr. Tolsma:
Members of my staff have met with the Wisconsin Asphalt Pavement Association Environmental
Technology Transfer Team (WAPA ETTT). The meetings covered several topics including the one
addressed in the subject heading of this letter.
The Wisconsin Department of Natural Resources (DNR) decision making process regarding
environmental protection is always based on good engineering practices and sound engineering
judgement. In the absence of commonly accepted scientific fact we will, on rare occasion, consider
extrapolation if it is based on sound engineering principles.
It is my understanding, based on discussion with my staff; the Wisconsin hot mix asphalt (HMA) industry
has provided DNR with a considerable number of SO2 emissions test results. In addition, data from the
Michigan Department of Natural Resources has been made available to the DNR. The Michigan data and
the DNR data are consistent in showing that 50% of the sulfur in the fuel oil does not result in SO?
emissions.
The consistency in data among these sources, in addition to other scientific information supports our
conclusion that an emission factor of 0.5 pounds per pound is reasonable and appropriate. We also note
that the data we have evaluated suggests a range that could possibly indicate 0.5 pounds per pound to
result in a conservative estimate of SO2 emissions from HMA sources.
The Wisconsin Department of Natural Resources reserves the nght to alter its current position without
notice if a review of existing or new relevant data dictates.
If you have questions or concerns regarding our position or this issue, feel free to contact either Mr.
Patrick Kirsop or Mr. Keith Pierce of my staff,
Sincerely,
Lloyd L. Eagan, Director
Bureau of Air Management.
cc: P. Kirsop - AM/7, K. Pierce - AM/7
Quality Natural Resources Management
Through Excellent Customer S&rvice «•«-«
" KKitHO
TQTftL P.02
-------
Norman Ostroff, PhD, PE
Stamford, CT
-------
NORMAN OSTROFF, PhD, P.E,
87 FISHING TRAIL
STAMFORD, CONNECTICUT 06903
(203) 322-5559
2000.190
August 25, 2000
Mr. Robert McConnell
Environmental Engineer
Air Quality Planning Group
USEPA - Region 1
1 Congress Street - Suite 1100
Boston. MA 02 11 4-2023
Dear Bob:
I am writing to discuss several points in the recently issued draft report titled "Hot Mix
Asphalt Plants, Emission Assessment Report" EPA-454/R-00-OXX. Please extend my
congratulations to all those who worked on its preparation for a job very well done. Please
also accept the comments I make in this letter as purely constructive.
In general, I my comments will concern the readability of equations, rather than their
derivation, which I took to be correct. My primary area of concern at this time involves VOC
emissions during truck load-out operations; my comments will be mainly concerned with this
section of the draft and will cite examples based upon Appendix B, pages 4-147 et seq.
I recommend you consider adding equation numbers wherever needed. When reviewing the
document (especially Appendix B) I found it somewhat difficult to follow because of the lack
of equation numbers.
On page 4.153 (Appendix B) you present the following equation:
If this equation were written in "algebra" rather than "computerese," it might look like:
FF
EFcor = (C prod ) - (C back ) x 7;
*• •* V- #
This small change would make the equation more "readable" to me and perhaps to others. I
understand the physics of the equation, but recommend that the definition of the parameter
Cprod be repeated here; it is not specifically stated. I assume that this equation is dimensionally
correct, but recommend you include the units for the respective terms. In the numerical
-------
Mr. Robert McConnell
August 25, 2000
2000.190 Page 2
example that follows (in the text), I recommend that you indicate to the reader the exact source
of the input data. I could not find the reference to Run 1 MCEM, but could have missed it
while reading the report. Also, demonstrating the calculation of Cprod from raw data would
have been helpful to me.
On the bottom of page 4-154 and the top of page 4-155, you present a short discussion of
vapor pressure. I agree that the Clausius-Clapyron (CC) equation generates a linear
relationship between (the logarithm of the) vapor pressure and (the reciprocal of the absolute)
temperature. The Antoine equation is a refinement of the CC equation in which a third
constant has been added. The CC equation is a special case of the Antoine equation in which
the third constant is 273.16. Please double check your references, the (CRC) Handbook of
Chemistry and Physics provides tabulated numerical vapor pressure (versus temperature) data
for a large number of compounds and Lange's handbook provides Antoine equation constants.
I suggest that the form of the two equations at the top on page 4-155 are somewhat difficult to
read. For example:
J3SS -
would be easier to read if it were in the form:
Loss = - EXP [0.0231 (t + 460) -19.28
I don't think the minus sign is necessary in these equations, but I understand it being placed
there to re-enforce the fact that the asphalt is losing mass during heating.
Return to the CC equation for a moment. The linearized form of the equation is:
ln(VP) = A + -
v ; OT
Where A and B are compound-specific constants. If one would make the same transformation
on the empirical equations (ignoring the minus sign) the form of the equations would be:
In(Loss) = AT + B
Please note the difference in form between the two equations.
In the second line of the next paragraph, I recommend that you remove the words "The
industry indicates that higher temperatures are avoided ."
On the bottom of page 4-155, you cite equation numbers 4-11 and 4-12. Where are equations
4-1 through 4-10? Again, these equations, as presented, are somewhat difficult to follow,
The equations can be simplified and perhaps the following sentence added:
-------
Mr. Robert McConnell
August 25, 2000
2000. 190 Page 3
"Equation 4-1 1 can be rearranged to give:
EF EF e- C413V'
crStcl = nrCorr i y J ........................... \ '
My Equation (4.13) is mathematically identical to your Equation (4.11). In addition, I have
found (in my own writing) that the use of the dotted line, connecting the equation and the
equation number, simplifies reading. Am I correct in my understanding that EFSt(I is the
emission factor to be used for prediction of future emissions? If so, then I believe this should
be clearly stated.
Fifteen figures are referred to at various points. In their current location, reader must stop
reading the document and look for the figures then return to the text. This is distracting,
interrupts the smooth flow of information, and can be avoided if the figures were placed
directly into the text (perhaps reduced in size).
Finally, I have been interested in the emissions from this phase of hot mix asphalt manufacture
for some time and have developed a purely theoretical model to describe VOC emissions from
freshly prepared asphalt starting the instant that it is discharged from a storage hopper or a pug
mill. My model predicted temperature changes during discharge and during transfer to the
paving site, and the VOC emissions accompanying the the transfer. My (theoretical) analysis
of the period in the truck is extremely close to your measurements. Your measured emission
factor was of 0.001 1 pounds of VOC per ton of asphalt in 8 minutes, my value was 0.00046
pounds of VOC per ton of asphalt in 3.3 minutes. I am enclosing a copy of my report, you
may find it interesting.
Please accept my congratulations for a job very well done. Thank you for the printed review-
copy you provided, and for the opportunity to comment on it.
Very truly yours,
Norman Ostroff, PhD, P.E.
Consulting Engineer
enclosure
cc: Mr. Ron Myers, US EPA, RTPNC
-------
DEVELOPMENT OF EMISSION FACTORS
FOR
FUGITIVE EMISSIONS
FROM
HOT MIX ASPHALT PLANTS
A THEORETICAL EVALUATION
by
Norman Ostroff, Ph.D, P.E.
87 Fishing Trail
Stamford, CT 06903
(203) 322 5559
April 1998
(Revised January 2000)
-------
DEVELOPMENT OF
EMISSION FACTORS FOR FUGITIVE EMISSIONS
FROM HOT MIX ASPHALT PLANTS
A THEORETICAL EVALUATION
by
Norman Ostroff, PhD, P.E.
Consulting Engineer
1. INTRODUCTION AND THEORY
1.1 Introduction
There are no published emission factors to quantify the fugitive emissions at hot mix asphalt
plants. The fugitive emissions in this study will be limited to VOCs. Fugitive particulate
emissions from blended product are expected to be negligible because of the binding action of
the asphalt.
The emission factors will be studied by following the temperature history of the product and
relating it to the rate of evaporation of volatile compounds.
There are two points in the process during which fugitive emissions can occur. These are:
1. The period during which the blended material is being dropped from the pug
mill into the transport truck, and
2. The period when the material is in the truck and is being transported to the
paving site.
In the case of blended hot mix asphalt, there are three mechanisms by which heat can be lost
from the mass. These are:
1. Radiation of heat from the hot particles to the atmosphere,
2. Convection from the surface of the mix to the atmosphere, and
3. Evaporative cooling accompanying evaporation of volatile components from the
mass.
-------
1.2 Radiation
The equation for heat transfers by radiation between a solid wall and a gas [1] is:
Q_
A
0.173s
wo)
oo
(1-1)
where e is the emissivity of the gas which will be assumed to be 0.1 based on published data
for water vapor and CO, [2]. Take 325°F (785 °R) as the surface temperature of the asphalt
mix and 60°F (520°R) as the air temperature. Substitute these quantities in Equation (1) to
get:
— = 0 . 173x0. 1 x f7.854 - 5.2041 = 50.3
A L Krft2
BTU
hrfr
(1-2)
1.3 Convection
The equation for heat loss from the mass to the atmosphere by convection is given by:
(1-3)
Where Ts is the surface temperature and !„ is the air temperature. The heat transfer coeffi-
cient for heat loss from a flat plate to still air [3a] is:
h
0.38(T
\0±5
(1-4)
Substitute in Equation (1-4) to get:
h = 0.38 x (325-60p = 1.53
BTU
ft2hr°F
(1-5)
and in Equation (1-3) to get:
- = 1 .53 x (325 - 60) = 406.3
A ^ ' ft2hr
(1-6)
The heat loss per unit area (assuming an average convective heat transfer coefficient approxi-
mately of 1.5 BTU/hrft2°F) is approximately 400 BTU/hr ft2.
-2-
-------
1.4 Evaporative Cooling
The heat loss accompanying evaporative cooling of asphalt compounds from the surface of the
mass is approximately 100 BTU/lb of asnhalt evaporated. Since asphalt only comprises
approximately 5 percent of the mass, and only a small amount of the asphalt will be evapo-
rated, this quantity of heat will not be included in the heat balances that follow.
1.5 Particle Heat Loss Summary
Convective heat transfer (406 BTU/hrft2) from the surface of the mixed mass is the primary
mechanism for heat loss from the surface and radiative heat transfer (53 BTU/hrft2) accounts
for approximately ten percent of the convective loss. To simplify the calculations that follow,
a combined convective-radiative heat transfer coefficient whose value is (1.1x1.5) 1.65
BTU/hr ft2 °F will be used in this study.
1.6 Conduction
The mechanism of heat transfer within the mass is conduction. The equation describing the
temperature profiles in a solid is the Laplace Equation [5a] of conduction which, in this case
is:
7
In this case, the asphalt mass will be modeled as a semi infinite solid, Equation (1-7) becomes:
In Equation (1-8), T is temperature, t is time, and a is a parameter called "thermal diffusivity"
defined by:
(1-9)
in which k is the thermal conductivity, p is the density of the solid, and Cp is its heat capacity.
Equation (1-8) is made dimensionless by introducing the following parameter:
e = -P (i-io)
-3-
-------
Substitute Equation (1-10) in Equation (1-8) to get:
a20
at
The initial and boundary conditions used to solve Equation (1-11) are:
X
(i-ii)
© = 0
® = 1
0 = 0
when t = 0
when t ) 0
when t } 0
all
X
X -> 00 (
0
1
1
The solution to Equation (1-1 1) is [5d]:
r0 -T x
= 1 - erf —==
Lv4at_
(1-13)
In which erf(Z) is defined by:
(1-14)
erf (0) = 0 (1-15)
erf(m)=l (1-16)
The heat flux is given by:
A
(1-17)
x- o
Differentiate Equation (1-13) with respect to X to get:
l-erf
ax
4at
(1-18)
Evaluate Equation (1-18) using the identity:
[ erf(Z)]
ax
(1-19)
-4
-------
with:
4at
(1-20)
and:
az
ex
4at
(1-21)
Substitute Equations (1-19) and (1-20) in Equation (1-18), and rearrange slightly to get
X
j= I exp I - ......... --
.v/4at
(1-22)
The temperature gradient at the solid surface (X = 0) is:
x=o
Ts-To
,/7t at
(1-23)
and the conductive heat flux is:
Q IcTl
A ~ UXJ
= k
' T,
4 a t
.. (1-24)
Heat transfer by convection (corrected for radiation) is the primary mechanism for removing
heat from the surface of the mass of the asphalt; it will be assumed that it is the only mecha-
nism, and the conductive heat flow in the mass will be equated with the convective heat flux.
It will be necessary to adjust the surface temperature (T,) as time progresses.
-5-
-------
2.
PHYSICAL PROPERTIES
Asphalt is not a single compound, but a mixture of hundreds of individual compounds that are
found in the residuum of petroleum refining. Asphalt is the material that remains after the
more volatile components have been distilled away, often at high vacuum. The hydrocarbon
components in asphalt are generally C25 and higher. Since it was impossible to find all
required physical properties of asphalt in available literature, the compound hexacosane
(C26H54) was used as a physical property model. Physical properties that were not available
were estimated using standard methods. The following data were available [4]:
366.72
0.7783
262°C at 15 mmHg
Molecular Weight
Specific Gravity
Boiling Point
Vapor Pressure:
mmHg
T°C
T°F
The vapor pressure data were correlated by:
P° (mm Hg) = 9.695- 109 [erp
API gravity 50.3"
UOP K factor 13.7
1
204.0
399.2
10
257.4
495.3
40
295.2
563.4
100
323.2
614.1
760
399.8
751.6
-197SOV
T°R J
(2-1)
The following properties were estimated using accepted methods:
Property
Critical Temperature
Critical Pressure
Critical Volume
Heat of Vaporization
Liquid Viscosity
1.16x10
-3
Method
Nokay [6a]
Lyderson [6a]
Lyderson [6a]
Othmer [3b]
Orick and Erbar [6b]:
2846V
T°KJ
Value
543.2°C
9.412 atm
1470cm3/mol
103.7 BTU/lb at 325°F
(2-2)
Liquid Heat Capacity Luria and Benson [6c]:
CP(molar) = - 16.724 + 1.69(T) - 5.063 x 1(T3 (l2) + 5.79 x 1(T6(T3) (2-3)
Thermal Conductivity
Robbins and Kingrea [6d]:
k = 3.314 xlO"4
T°R
Note: CP in molar units in this equation (2-4)
-6-
-------
3. TEMPERATURE PROFILES
3.1 Introduction
The system will be modeled in two stages. The first is the period during which a batch of
asphalt product is dropped from the pug mill mixer into the truck. The second is the period
when the batch is in the truck. The second period will be assumed to begin at the conclusion
of the first.
3.2 First Period
During this period, the asphalt will be assumed to be in the form of 1/4 inch (2.08xlO~3 ft)
spheres at a uniform temperature of 325°F. Since the spheres are moving (they are falling into
the truck), it is assumed that the heat transfer coefficient, between the spheres and the air, is
(10x1.5) 15 BTU/ft2hr°F.
The Biot number [7] in the sphere is:
_. Lk 2,08 x 10~2 x 0.132 ,
B.= — = -UfxIO-* •• -.(3-1)
Since the Biot number is very low, one can safely assume that the convective heat transfer
(from the surface) is controlling and the internal temperature is relatively constant. Assume
that the heat transfer coefficient will be approximately ten times as high as the value calculated
using Equation 1-5, the rate of heat loss from the sphere is:
Q / \ s BTl
* = h Ts - T* ) = 1 5x(325- 60 = 3980----
^ ' hrrt
BTU
.. (3-2)
In which Tm is the temperature of the air and Ts is the surface temperature of the particle. The
area of the sphere is:
A = TtD2 = W2.081(T2)2J36x 10"3 ft2 (3-3)
Assume that the particle must fall 1 0 feet from the mill outlet to the truck. The elapsed time
is:
= 0.788 sec .... (3-4)
Which is equivalent to (0.78813600) 2. 19xlO"4 hours. The amount of heat lost from the
particle is:
-7-
-------
Q pride = 3980 x 1.36 x 10^ x 2.19 x 1 0'4 = 1.18 xl(T3 B T U ( 3 - 5 )
The mass of the particle is:
_n3 n x (.08 xl(T3)x (0.7885x62.4)
Mass = — p = — ^ - 1 - = 2.32xlO-4lbs ( 3 - 6 )
6 6
The temperature change in the particle (using the heat capacity of asphalt) is:
AT= = 7 8 o F (3.
2.32xl(T4x0.656 v
This estimate can be refined with several additional iterations, but for the purpose of this
study, it is sufficiently accurate to take the temperature change as 7°F and the initial tempera-
ture of the asphalt in the truck to be (325 - 7) 318°F.
3.3 Temperature Changes in the Truck
This period will be evaluated using the equations developed in Sections 1.3 and 1.5. It will be
assumed that the mass in the truck is homogeneous, and its physical properties are equal to
those of asphalt (or hexacosane). In this case, the Biot number is very large and the internal
temperature profiles must be considered. The development of the temperature profiles will be
developed first using numerical methods; the emission rate from the mass will be computed
using the results of the temperature calculation.
The conductive heat flux through the asphalt has been given by Equation (1-24):
O { dr\\ f T - To)
^-= -kkv = k ° s (1-24)
A VaXyx=o V V4at J
and the convective heat flux from the surface has been given by Equation (1-3):
(1-3)
Since the heat flux is the same in the solid and gas phases, equate Equations (1-23) and (1-5)
to get:
Q_ = ,, f T0 - Ts
A
Now solve Equation (3-8) for the surface temperature Ts to get:
-8-
(3-8)
-------
s = " ', -- - , " ....... (3-9)
+ k
Equation (3-9) is a relationship between the surface temperature of the mass, and the elapsed
time since it was deposited in the truck.
The thermal boundary layer is that portion of the mass in which all of the temperature changes
occur. Outside of this boundary layer, the temperature is constant at the temperature of the
initial deposit (318°F). In order to evaluate the boundary layer thickness, a trial and error
approach was used. The temperature profile, given in Equation (1-12) was solved for T:
T(x,t)=T0-(T0- T,)
X
. (3-10)
The time was assumed and values of X were found so that the temperature was within
tolerance of the initial temperature. The boundary layer thickness was found to be approxi-
mately 1 inch at 3.3 minutes.
-9-
-------
4. EMISSION RATES
The emission rates were determined based upon published AP-42 emission factors. These are:
TABLE 4-1
AP-42 Emission Factors for VOC
From Hot Mix Asphalt Plants
Entries in Ib/ton
Batch Mode Drum Mode
Natural Gas 0.017 0.051
Fuel Oil 0.046 0.069
Of these, the factor for VOC emission using fuel oil in the drum mode was chosen for this
study because it provided the most conservative (high) estimate.
The data in Table 4-1 represents the amount of VOC that will be emitted during the production
of a ton of hot mix asphalt. The amount of fuel required to produce a ton of hot mix asphalt
has been reported to be 0.5 gallons of Number 2 fuel oil. This is equivalent to (05x140,000)
70,000 BTU or (70,000/1000) 70 standard cubic feet of natural gas.
The AP-42 factor for VOC emission during Number 2 fuel oil combustion is 0.20 lb/mgal;the
AP-42 factor for VOC emission during natural gas combustion is 2.78 Ib/mmSCF. The VOC
emission associated with fuel combustion during hot mix asphalt production are (0,5xlO"3x0.2)
l.OxlO-4 Ib/ton for fuel oil and (70xlO'6x2.78)
1.95xlO'4 Ib/ton. These quantities are small and will not be considered here.
For the purposes of this investigation, the emission factors given in Table 4-1 were interpreted
as Ib/ton per hour. That is, if one ton of hot mix asphalt could be maintained under manufac-
turing process conditions, it would emit the AP-42 factor.
When dealing with fugitive emissions, the process conditions are not maintained. If fact the
mass cools quickly, and a thermal boundary layer is created. The emissions of VOC are
dependent upon three parameters:
1. The rate at which liquid can flow to the surface,
2. The ease with which the liquid can be vaporized (volatility), and
3. The rate with which the volatilized asphalt components can be removed from
the vicinity of the cooling mass (diffusion).
In order to develop a liquid flow factor, it will be assumed that the liquid flow to the surface
of the mass occurs by capillary flow, which is a form of laminar flow. The equation for
pressure drop in a fluid system is:
- 10-
-------
(4.1)
The (D'Arcy -Moody) friction factor in laminar flow is:
64 64 ,
~
substitute Equation (4.2) in Equation (4.1) and solve for v to get:
HfD2pg
If it is assumed that the capillary pressure is constant, and the diameter of the passages arc
constant, Equation (4-3) demonstrates that the liquid flow rate is inversely proportional to the
viscosity of the liquid.
The volatility is simply expressed as the vapor pressure (of hexacosane) at the surface
temperature of the mass.
The diffusion coefficient is more difficult to estimate. There are no published data describing
diffusion of compounds of this type in air. The diffusion coefficient for asphalt vapor in air
(at 25°C) was estimated by noting the a strong correlation between diffusion coefficient and
geometric mean molecular weight shown in Figure 4-1. The data were correlated using the
method of least squares; the resulting empirical equation is:
D^B = 20.78 (MW}~1392 (4-4)
The index of correlation (r2) is 0.91, indicating a good fit to the data. Equation (4-4) was
evaluated at the geometric mean molecular weight of air and hexacosane [(29+ 366)° 5] 103 to
obtain an estimate of 0.0327 cm2/sec at 25°C. Admittedly, this is a small extrapolation but
represents the best available information.
The Chapman-Enskog [5b] kinetic theory was used to relate the diffusion coefficient to
temperature. The Chapman-Enskog equation is rather complex, but for this discussion has
been simplified as:
in which ( is a constant related to the system and QD is a function of temperature that is
usually presented in tabular form [5c]. For use in this investigation, QD was plotted against
the temperature function (which is the product of two system-specific constants and the
absolute temperature). The graph is presented Figure 4-2.
-11-
-------
FIGURE 4.1
Diffusion Coefficients of Various
Compounds in Air at 25°C [8]
0
(oas/aiDbs)
uoisn,y.jQ
O
, o
O)
'CD
O
_0>
o
c
CU
0)
OJ
E
O
CD
-+s
To
sO
- 12-
-------
FIGURE 4-2
The QD Function [5c]
o
T§
o
o
CD
g
"o
CD
OJ
CL
f~
0)
o
-13-
-------
A power law equation (similar to Equation 4-3) was fit to the published data again using the
method of least squares; the resulting empirical equation is:
nD = 1.583 (KT / s)"034 (4-6)
The index of correlation (r2) is 0.93, indicating another good fit. Equation (4-5) is a simple
and relatively accurate relationship which is satisfactory for use here. Since both K and e are
compound-specific constants, Equation (4-6) indicates that the QD function is proportional to
the (absolute) temperature raised to the (-0.34) power. When this is substituted in Equation
(4-5) it becomes:
' I'15 \ x
1 ' '-'"^ (4.7)
Equation (4-7) will be used to estimate the variation of diffusion coefficient with temperature,
In order to predict emission rates as the mass cools, it was assumed that the full value of the
AP-42 factor was applicable when the mass was being prepared and was maintained at the
mixer temperature of 325°F.
The AP-42 factor is adjusted using three "parameter correction factors" defined:
_ pi (TsoQ
4 • 8
in which P is the parameter (vapor pressure, viscosity, diffusivity) and F; is the respective
correction factor defined as the ratio between the value of the parameter at the instantaneous
surface temperature of the asphalt, divided by the value of the same parameter at 325°F. The
overall correction factor is:
Visc
The emission rate is computed as the product of the overall correction factor and the AP-42
factor. Recall that the oil/drum factor will be used here because it provides the most conserva-
tive (high) estimate:
Emission Rate 1 — -- - = F0 x(AP42EmissionFactor) (4-10)
\ ton /
-14-
-------
5. CALCULATIONS
The calculations were in the following order:
5.1 A time was assumed.
5.2 The surface temperature was calculated using Equation (3-9).
5.3 The physical and transport properties of the mass were evaluated at the surface using
the surface temperature computed in Step 5.2.
5.4 The thickness of the thermal boundary layer was computed using Equation (3-10) and a
trial and error solution. The assumed thickness (X) was varied until the computed
temperature [T(X,t)] was equal to the initial temperature (T,).
5.5 The vapor pressure, liquid viscosity, and diffusion coefficient, at the surface tempera-
ture were computed using Equations (2-1), (2-2), and (4-6), respectively.
5.6 The parameter correction factors for vapor pressure, liquid viscosity, and diffusion
coefficient were computed using Equation (4-7), and the overall correction factor was
computed using Equation (4-8).
5.7 The emission rate was computed using Equation (4-9). Recall that this rate is only
valid for the period defined in Step 5.1.
5.8 The amount of VOC emitted during the period is equal to the product of the emission
rate calculated in Step 5.7 and the duration of the time period. The emission is
expressed in the units of Ib/ton.
5.9 The total emission is calculated by adding the amount emitted (Step 5.7) to that
computed in previous iterations.
5.10 Assume a new temperature and repeat Steps 5.2 through 5.9.
These calculations were performed by computer, the output is appended.
-15-
-------
6. RESULTS
The results of this study are summarized in Figures 6-1, 6-2, and 6-3 which relate surface
temperature, thermal boundary layer thickness, total emissions, and instantaneous emission
rate; respectively; to elapsed time. The study was conducted over a time period of 3.3
minutes from the time that the pug mill is empties into a truck. The emission rate has decayed
considerably and only adds a small increment to the total.
The "3.3 minute emission factor" is 4.21xlO~4 pounds per ton, which will be increased by 10
percent to (1. Ix4.21xl0~4) 4.63xlO"4 pounds per ton to account for any emissions that might
occur after the 3.3 minute period has elapsed.
This number is approximately two orders of magnitude smaller than the AP-42 factors (Table
4-1) because the asphalt is cooling and thereby severely restricting its ability to emit,
The stack emissions from a batch mix hot mix asphalt facility, operating in the batch mix
mode, using natural gas fuel, and producing 300 tph of product are:
VOC emissions = 0.017 x 300 = 5.10 Ib/hr (6-1)
The fugitive emission from the truck would be:
Fugitives = (4.63xlO-4) x 300 = 0.14 Ib/hr (6-2)
Which is about 2.7 percent of the stack emissions.
The emission rate, defined as the rate at which VOC evaporates from the surface of the
asphalt, is initially equal to the evaporation rate from the asphalt during the mixing, falls to
approximately ten percent of the initial rate after approximately 3 minutes. This is shown in
Figure 6.4.
-16-
-------
FIGURE 6.1
Surface Temperature of Cooling Asphalt Mass
i-CO
_ ; .
i ©
I
o
ro
CO
o
CM
CO
(-j
CNJ
CD
e
-17-
c:\tcwin2\asptemp.prn
-------
FIGURE 6.2
Development of the Thermal Boundary Layer
in Cooling Asphalt Mass
1
^
\
X
\
\
\
«
i
I
\
\
h-
c
i
A
\
\
>
•
i
j
\
o
.
Vx>
co
0 0
-co
-C\J
-------
FIGURE 6.3
Fugitive VOC Emissions from Cooling Asphalt Mass
Development of the Emission Factor
0
»;
\
-> x
\
\
a
*- if
i
\
\
\
i
•) r
j
\
\
V :
\
\
\
\
i
) LO Cs
•"d" co c\j
\
'\,
I
J IT
j
; i
- -- •••- ' I-CO
" ' ' " \™$
\ "3
i C
! E
0
; 1 E
; i~
1 • , :
••• j ! O
) T- m o
-r- O
OOO'OUdieqdsv UOJ/QQA Ql) suoisstiug
- 19
c:\tcwin2\asp-cram.prn
-------
FIGURE 6.4
Fugitive Emission Rate from
Cooling Asphalt
*
I
f
*
I
f
0
(-CM
I
0)
I
0)
p
;_, o
CD
O
O
O
UD
O
o
p
o
xt-
o
o
o
CO
o
o
p
o
CM
o
o
p
o
[jnoij/(}|eijdsy UOJ/OQA
o
o
p
o
uoissiuug
o
- 20
-------
Appendix 1: Computer Output Sheets
-21-
-------
Appendix 2 Literature Cited
[1] 1989 Ashrae Handbook, Fundamentals, Page 3.11
[2] RH Perry, D Green "Perry's Chemical Engineers' Handbook, 6/ed" McGraw
Hill (1984) pp 10.62, 10.63
[3] JH Perry, ed. "Chemical Engineers' Handbook, 3/ed" McGraw Hill (1950) p
474
[3b] Ibid, Page 300
141 RC Weast, ed. "Handbook of Chemistry and Physics, 49/ed" (1968) assorted
pages
[5a] RB Bird, WE Stewart, EN Lightfoot "Transport Phenomena" Wiley (1960),
Chapter 10
[5b] Ibid, Page 510
[5c] Ibid, Page 746
[5d] Ibid, Section 11.1
[6a] RC Reid, JM Prausnitz, TH Sherwood "The Properties of Gases and Liquids,
3/ed" McGraw Hill (1977), Page 12
[6b] Ibid, Page 439, 440
[6c] Ibid, Page 155, 156
[6d] Ibid, Page 519
[7] DQ Kern "Process Heat Transfer" McGraw Hill (1950) Page 37
[8] CO Bennett, JE Myers "Momentum, Heat, and Mass Transfer, 2/ed" McGraw
Hill (1974) Page 787
-22
-------
Steven Zemba
Cambridge Environmental Inc.
Cambridge, MA
-------
zemba@cambridgeenviro To: Ron Myers/RTP/USEPA/US
nmental.com cc:
Subject: Belated Comments on the Hot Mix Asphalt Emissions Assessment Report
09/22/00 10:34 AM
Hi Ron:
I apologize for not geeting these comments to you within the official
comment period, but since I did review the emission assessment report at
a cursory level, I thought I would pass along my notes belatedly in the
hope that they still might be useful to you. My notes/comments are
contained in the attached file. Feel free to call/write with any
questions on them. I hope that this represents the end of the process
for you!
Take care,
Steve
Stephen Zemba (mailto:Zemba@CambridgeEnvironmental.com)
Cambridge Environmental Inc.
58 Charles St.
Cambridge, MA 02141
617-225-0810 (Voice)
617-225-0813 (FAX)
http://www.CambridgeEnvironmental.com
- asphalt_comments.PDF
-------
Comments and notes on "Hot Mix Asphalt Plants, Emission Assessment Report"
General: A list of acronyms at the front of the document might e helpful to readers.
General: Tables 1 and 2 claim to present emissions for a "typical" plant. This is probably true for mixer
emissions, but is not the case for the fugitive sources, as indicated by statements on p. 14. Specifically,
the fugitive emission estimates assume a high-end volatility of asphalt cement, and also a maximum
recommended load-out temperature. I would thus characterize the fugitive emission estimates as "high-
end" for a typically-sized plant. EPA might want to clarify this matter somewhat, since the actual ratio
of fugitive-to-stack emissions for a typical plant running at typical conditions will be lower than that
implied in Tables 1 and 2.
General: Many HMA plants are now installing condensers on their asphalt storage tank vents and silo
vents; some are enclosing or partially enclosing their loadout areas. The fugitive emissions discussed in
this document are for uncontrolled sources. In fact, the estimates in Tables 1 and 2 implicitly assume
controls on the mixer (dryer) stack but no controls for fugitive emissions. This assumption should be
made explicit (perhaps in a footnote to the tables). EPA might also want to offer some
opinion/conjecture on the efficacy of engineering controls for the various fugitive sources.
Specific comments
The last sentence third pgh. of pg 1 contains a typo: the 100,000 tons of production corresponds to a
batch plant, not a drum plant.
1st pgh. of p. 10: Does the emission assessment report provide stack emission rates for a typical
counterflow or parallel flow drum plant? If the estimates are for a parallel flow plant, it might make
sense to mention in the text the approximate difference expected for emissions from a counterflow plant
(e.g.. this could be mentioned in a footnote to Table 2).
p. 3: EPA might want to provide the regulatory definition or a better explanation of criteria pollutants
and HAPs.
p. 2, 3rd pgh: Might want to mention the number of tests (> 300) that survived the screening process
(i.e., were usable for the development of emission factors).
p. 13: Are emissions from hot oil heaters included in any of the categories in Tables 1 & 2 (e.g., the
"mixer" category)? If not, does it make sense to mention the likely magnitude of these emissions
relative to other sources (qualified, of course, by the large degree of uncertainty resulting from the
dearth of test data)? Might it also be possible to use emission factors from similar burners (used in
other industries) to evaluate hot oil heater emissions?
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA-454/R-00-030
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Hot Mix Asphalt Plants
Stakeholders Opinions Report
5. REPORT DATE
April 2001
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Ron Myers (EPA)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The United States Environmental Protection Agency (EPA) Source Measurement Technology Group (SMTG) and
Emission Factors and Inventory Group (EFIG) are investigating the Hot Mix Asphalt industry to identify and quantify
criteria and hazardous air pollutants (HAP's) emitted from kiln stacks, transport truck loading and silo filling at hot mix
asphalt plants. The EPA report titled "Hot Mix Asphalt Plants - Emissions Assessment Report" (EPA 454/R-00-019,
December 2000) presented the findings from these investigations. As part of the investigation, EFIG obtained over 300
emission tests from kiln stacks that characterize emissions of criteria pollutants and hazardous air pollutants' emissions.
Additionally, SMTG sponsored two kiln stack emission tests and two emission tests of the transport truck and silo filling
operations. The primary objective of the testing program was to characterize uncontrolled emissions of the criteria
pollutants particulate matter (PM) and total hydrocarbons (THC) and emissions of volatile and semi-volatile organic
HAP's. The results of the four test reports and responses to comments on these test reports are covered in separate EPA
reports (EPA 454/R-00-020, EPA 454/R-00-021, EPA 454/R-00-022, EPA 454/R-00-023, EPA 454/R-00-024,
EPA 454/R-00-025 (a through h), EPA 454/R-00-026, EPA 454/R-00-027, EPA 454/R-00-028 and
EPA 454/R-00-029). This document includes stakeholder comments on the draft version of the Emissions Assessment
Report, EPA's response to these comments, stakeholder opinions on the final version of the Emissions Assessment
Report and EPA responses on new issues raised by stakeholders.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATT Field/Group
Hot Mix Asphalt
Silo Filling
Truck Loading
Particulate Matter
Volatile Organic Compounds
Total Hydrocarbons
Hazardous Air Pollutants
Air Pollution control
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
288
20. SECURITY CLASS (Page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
-------
This Page is intentionally left blank.
------- |