EPA-600/2-76-120
April 1976
Environmental Protection Technology Series
CONTROL TECHNOLOGY FOR
ASPHALT ROOFING INDUSTRY
PRCT
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development. U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2 Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research oerformed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality'standards
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield Virginia 22161
-------
EPA-6QO/2-76-120
April 1976
CONTROL TECHNOLOGY
FOR
ASPHALT ROOFING INDUSTRY
by
Paul G. Gorman
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
Contract No. 68-02-1324. Task 35
ROAPNo. 21AFA-106
Program Element No. 1AB015
EPA Task Officer: E. J. Wooldridge
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
This report was prepared for Industrial Environmental Research Laboratory-
RTF, Environmental Protection Agency, to present results of the work carried
out by MRI under Contract No. 68-02-1324 (Task 35). This work was performed
in the Physical Sciences Division of Midwest Research Institute by Mr. Paul
G. Gorman, Dr. K. P. Ananth, Dr. F. Honea, Dr. L. Rust, and Dr. A. K. Rao.
Approved for:
MIDWEST RESEARCH INSTITUTE
L. J.( shannon, Assistant Director
Physical Sciences Division
April 30, 1976
111
-------
CONTENTS
Page
List of Figures vii
List of Tables . ix
Summary. ... ...... 1
Conclusions and Recommendations 3
Introduction • 4
Discussion 6
Survey of the Industry ........ 6
Technical and Economic Evaluation of Candidate Control
Techniques 17
Recommendations. 32
Priority I 32
Priority II 34
Priority III 34
Priority IV (Optional) 35
Planning ...... 36
Priority I - Evaluation of Results from Current EPA Test Program . 36
Priority II - Characterization of Emissions 36
Priority III - Pilot Scale Testing of HEAP and Wet ESP on Air
Blowing ..... 38
Priority IV - Optional Research 39
References 40
v
-------
CONTENTS (Concluded)
Appendix A - List of Asphalt Roofing Plants and Emission Control
Devices 41
Appendix B - Evaluation of Afterburners for the Asphalt Roofing
Industry 45
Appendix C - Evaluation of the Performance of HEAF (High Energy
Air Filtration) System for Controlling Emissions
from Asphalt Saturators/Blowers 65
Appendix D - Theoretical Analysis of the Applicability of Venturi
Scrubbers for Control of Asphalt Saturator
Emissions 82
Appendix E - Evaluation of Electrostatic Precipitators for Con-
trolling Emissions from Asphalt Saturators/
Blowers 91
vi
-------
LIST OF FIGURES
No. Page
1 Schematic Diagram of Vertical Asphalt Blowing Still 7
2 Schematic Diagram of Asphalt Saturating Line 8
3 Particle Size Distribution in Uncontrolled Saturator Exhaust. 14
B-l Consideration for Successful Incineration of Dilute Fumes . . 47
B-2 Afterburner Operating Temperature Versus Removal Efficiency . 48
B-3 Typical Temperature-Performance Curves for Various Molecular
Species Being Oxidized Over Pt/A^O-j Catalysts 51
B-4 Thermal Afterburner Heat and Fuel Requirements Versus Tem-
perature and Volume Flow Rate 53
B-5 Afterburner With Recovery Boiler 55
B-6 Estimate Capital Costs for Installed Afterburner Versus Flow
Rate and Temperature Rise 57
B-7 Estimated Annual Fuel Costs for Thermal Afterburners Without
Heat Recovery 58
B-8 Estimated Annual Costs for Thermal Afterburners Without Heat
Recovery ..... 59
C-l Operation of a HEAP Unit 67
E-l General Classification of Electrostatic Precipitators .... 93
E-2 Hypothetical ESP Collection Plate Configuration 98
VII
-------
LIST OF FIGURES (Concluded)
No. Page
E-3 Particle Size Distribution in Uncontrolled Saturator
Exhaust 99
E-4 Schematic of "Smog-Hog"™ 102
viii
-------
LIST OF TABLES
No.
1 Summary of Test Data Reported for Blowing Emissions 11
2 Summary of Test Data Reported for Saturator Emissions .... 12
3 Selected Emission Criteria 19
4 Evaluation of Control Equipment for Asphalt Saturator
Emissions 20
5 Evaluation of Control Equipment for Air Blowing Emissions . . 22
6a Afterburner Advantages and Disadvantages 24
6b HEAF Advantages and Disadvantages 25
6c Electrostatic Precipitator Advantages and Disadvantages ... 26
7 Summary of Recommendations 33
8 Summary of Estimated Costs and Time Requirements for Recom-
mended Research and Development Plans 37
B-l Estimate of Thermal Afterburner Costs for Saturators .... 61
B-2 Estimate of Thermal Afterburner Costs for Air Blowing .... 62
C-l Impaction Efficiency as a Function of Particle Size and
Particle Velocity 71
C-2 Efficiency of Interception as a Function of Particle Size . . 73
C-3 Combined Target Efficiency as a Function of Particle Size
and Particle Velocity . 74
ix
-------
LIST OF TABLES (Concluded)
No.
C-4 Asphalt Saturator Emission Tests 76
C-5 Particulate Emission Data for Asphalt Saturator Controlled
With HEAF 77
C-6 Economics of HEAF System for Controlling Asphalt Emissions . . 79
C-7 Economics of HEAF Systems at One Roofing Plant 80
D-l Tabulation of Particle Size Distribution Data for Asphalt
Saturator Emissions 86
D-2 Overall Efficiency of Removal for Asphalt Saturator Particu-
late Versus Pressure Drop 88
E-l Number of Elementary Charges Acquired by a 0.1 urn Particle
as a Function of Time by Diffusion and Field Charging ... 96
E-2 Theoretical Overall and Fractional Efficiencies for Saturator
Operations 100
E-3 Tests at Celotex Corporation, Lockland, Ohio 104
E-4 Tests at Celotex Corporation, Fairfield, Alabama (March 4 and
5, 1975) 105
E-5 Test Data Obtained for a Small "Hydro-Precipitrol"™ Unit for
a Saturator Application 107
E-6 Cost Estimate for "Smog-Hog"™ Units Controlling Asphalt
Saturator Emissions 110
E-7 Cost Estimates for Environmental Control Unit Controlling
Asphalt Saturator Emissions Ill
E-8 Cost Estimate for "Elektrofil"™ Units Controlling Asphalt
Saturator Emissions 113
-------
SUMMARY
Asphalt roofing plants represent a source of hydrocarbon emissions for
which control technology has not been well characterized. The technical and
economic feasibility of candidate control methods which may be capable of
99% removal of total hydrocarbons emitted from asphalt-saturating and air-
blowing operations in asphalt roofing plants was evaluated in this study.
Information from the literature, theoretical analysis of control systems,
and contacts with equipment manufacturers and plant operators were utilized
in making the evaluations.
Results of an industry survey showed that thermal incinerators or after-
burners are currently the only technique being employed for control of air-
blowing emissions. Control techniques for saturator emissions consist of
afterburners, wet scrubbers, filters (HEAP) and electrostatic precipitators.
Analysis of test data for these devices did not reveal any that had
demonstrated efficiency as high as 99%. For this reason, it was necessary
to employ theoretical analysis to evaluate the capability of various de-
vices for achieving 99% removal of total hydrocarbons. Further, the avail-
able test data indicated that emissions from both saturating and air-blowing
operations contain considerable amounts of gaseous hydrocarbons, which may
range from 2 to 48% of the total hydrocarbons. It would be necessary to ef-
fect condensation of these hydrocarbons in order to make it possible for
particulate control devices to remove 99% of all hydrocarbons. However, the
available data were not sufficient for determining the temperature to which
the effluents would have to be cooled for more than 99% of the hydrocarbons
to be in particulate form. A current EPA test program should provide more
information in this area, but in the interim the assumption was made that
precooling below 52 C (125 F) will be required. If the pending test data
show that precooling could not achieve the requisite condensation, then
afterburners are the only devices that may be capable of achieving 99% re-
moval of total hydrocarbons.
Another concern of this study was removal of POMs, which may be only
a small fraction of the total hydrocarbons. However, investigation of pre-
vious work in this area indicated that the control devices would be ex-
pected to reduce the POM emissions in direct proportion to the particulate
removal.
-------
Theoretical analysis of candidate control systems indicated that
thermal afterburners, high energy air filters (HEAP) and electrostatic
precipitators (ESP) could be capable of 99% removal of particulates, but
it is doubtful that wet scrubbers could achieve 99% removal. In addition
to the theoretical analysis, further technical and economic evaluations
of the devices were performed in order to identify those candidate de-
vices that should be recommended for more research and development.
These evaluations showed that afterburners are already well developed
and should certainly be capable of 99% removal efficiency, but they
have much higher costs than HEAP and ESP units and fuel availability
can be a constraint to widespread usage.
HEAP and dry ESP units have already been applied to the control of
saturator emissions. Further development work in this area should not
be necessary unless the results of the current EPA test program casts
serious doubt on their capability for providing 99% removal of particu-
late matter.
Neither the HEAP nor the ESP units have been used for control of
air-blowing emission. It was recognized that the higher grain loadings
in air-blowing effluents would increase the filter mat usage rate in a
HEAP and could seriously compound the buildup and fouling problems in
an ESP. Additional development work is warranted for this application.
Based on the technical and economic evaluations made in this study,
several recommendations were made for further research and development
efforts and these are presented in the next section. The primary con-
trol device development recommendation was that pilot scale HEAP and wet
ESP devices be tested on an air-blowing source.
If the HEAP or wet ESP performed successfully on a pilot scale and
were proven on a full-scale demonstration project, alternatives to after-
burners for high efficiency removal of hydrocarbon emissions from air-
blowing operations would be available at lower cost and energy consump-
tion.
-------
CONCLUSIONS AND RECOMMENDATIONS
Inadequate data exist to (a) completely characterize the nature of
emissions from saturator and air-blowing operations at asphalt roofing
plants, and (b) define the actual performance of candidate control sys-
tems. With regard to the former, data are not available to determine the
percentage of the hydrocarbon emissions that exist in gaseous form at
typical control device operating conditions. Lacking reliable data on
the nature of the hydrocarbon emissions, it is difficult to define actual
performance of particulate control devices. Based on existing data for
other sources, afterburners may be the only method capable of achieving
99% removal of total hydrocarbons.
Theoretical analysis indicated that HEAP and ESP units should be
capable of providing 9970 removal of particulate hydrocarbons. Cooling
of the gases to achieve hydrocarbon condensation prior to entering the
HEAP and ESP units may make it possible to achieve 99% removal of total
hydrocarbons. Analysis of the performance of wet scrubbers indicated
that 99% removal of total hydrocarbons was doubtful.
The principal recommendations for additional work are:
1. Evaluate results of ongoing EPA sampling work as soon as
it is completed.
2. Conduct any additional field sampling as may be necessary to
characterize emission streams for design of control equipment.
3. Undertake development work to define the performance of HEAP
and wet ESP units on air-blowing emissions.
-------
INTRODUCTION
The Clean Air Act directs the Environmental Protection Agency to de-
fine and develop technology for control of air pollutant emissions. Much
of that effort has been directed to specific industrial source categories,
One of these source categories is the asphalt roofing industry.
Preliminary surveys and investigations have indicated that the emis-
sions from sources within this industry are primarily particulate and
gaseous hydrocarbons contained in the effluents from the asphalt blowing
stills and asphalt saturator operations. The investigations also showed
that there were not adequate data available for characterizing these emis-
sions. Application of control devices was very limited on air blowing,
but several different types have been applied for controlling saturator
emissions, with varying degrees of success.
EPA contracted with MRI to perform a study for evaluation of the
technical and economic feasibility of control process candidates for
abating air emissions from the asphalt roofing industry.
This study was composed of three parts:
1. Survey the asphalt roofing industry to (a) identify existing
control technology, and determine the device usage for controlling air
emissions from saturating lines and blowing stills, and (b) identify
feasible control process candidates which may be applicable to those
sources.
2. For each control process or device identified, determine whether
the technology is technically and economically feasible for reducing air
emissions by 99%.
3. Recommend the most feasible control process or processes, based
on the above evaluation of technical, economic and energy considerations,
and determine the cost, time and approach necessary for developing or
demonstrating this control technology.
During initial discussions with the project officer (Mr. E. J.
Wooldridge) it was mutually agreed that the stipulation of 99% removal
-------
efficiency would be defined on the basis of 99% removal of total hydro-
carbons. This means all hydrocarbons, whether in the form of solid, liq-
uids, condensible vapor, or gases.
It was logical to assume that the significant emissions from air
blowing and saturators would all be hydrocarbons in one form or another,
and it was felt that the above definition would be most meaningful. It
was also decided that although 99% removal of total hydrocarbons was the
primary goal, the devices should also be evaluated on the basis of a lower
removal efficiency (e.g., 95%) in order to compare cost differences and
other factors.
Although the above definition of removal efficiency (i.e., total hy-
drocarbons) is probably the most meaningful one for evaluating emission
control techniques for the two asphalt industry sources, its use did lead
to some difficulties, primarily because of the lack of adequate informa-
tion on characteristics of emissions from these sources and the associ-
ated fact that emission test procedures often do not include determination
of gaseous hydrocarbons. Because of the lack of such detailed information
it was necessary to make certain assumptions regarding the form and fate
of the hydrocarbons. These assumptions are discussed in another section
of this report, but basically they consisted of the following:
1. Available test data indicated that for both sources the effluent
at stack conditions probably contained a significant portion of gaseous
hydrocarbons; probably greater than 5% of the total hydrocarbons. It was
therefore assumed that particulate control devices could not provide the
required hydrocarbon removal efficiencies unless they were preceded by
precoolers (water sprays) to reduce the temperature and maximize condensa-
tion of gaseous hydrocarbon constituents.
2. It was further assumed that if the effluent streams are precooled
(e.g., water sprays) below 52 C, less than 1% of the hydrocarbons would
remain in gaseous form so that 99% removal of total hydrocarbons would
be possible via particulate control devices if this were within their
capability and they were so designed.
This latter assumption is rather tenuous. Another EPA project that
is currently in progress involves testing of various control devices cur-
rently used in asphalt roofing plants and the test method being used should
provide more information on the percent gaseous hydrocarbons at the sam-
pling temperature of 38 C. If data obtained in that project do not vali-
date the above assumption, some of the results in this report may need
to be reconsidered. That is, it would not be appropriate to pursue further
testing and development of control devices that remove only the particulate
matter.
Details of the various activities on the program are presented in the
following sections of this report.
-------
DISCUSSION
SURVEY OF THE INDUSTRY
Initial work on this project consisted of a survey of the industry to
(a) identify the extent of usage and capability of current control devices
used to control emissions from asphalt-blowing operations and saturator
lines, and (b) identify other possible control techniques. Activity also
included collection of existing data to characterize the emissions and con-
trol techniques for the two emission sources. The major outputs from the
survey are presented in the following subsections.
Process Descriptions
Figures 1 and 2 illustrate the asphalt-blowing operations and the
saturator lines. Air blowing is a batch operation in which compressed
air is forced to bubble up through the liquid, at temperatures of 220
to 270 C, for a period of about 2 to 4 hr or longer in horizontal stills.
The purpose of this operation is dehydrogenation, causing some polymeriza-
tion of the asphalt to increase its melting point.—' The longer the blow-
ing time the higher the melting point, and the length of time is dependent
on whether the asphalt is to be used as saturant or coating in the satu-
rating line. The blowing operation uses 0.006 to 0.06 m^/min of air per
gallon of asphalt and the quantity of asphalt to be blown is on the order
of 20 to 40,000 liters. Water may also be added at the top of the vessel
to form steam to blanket the liquid surface and maintain proper tempera-
tures of the asphalt. Effluent flow rates may therefore be approximately
140 to 560 nrVmin at a temperature of about 170 to 200°C.
The effluent gas from the blowing stills is composed primarily of un-
reacted air and water vapor but also contains carbon oxides and some sul-
fur compounds plus gaseous hydrocarbons and entrained asphalt droplets.
This effluent is usually passed through a cyclone separator to collect
the larger droplets, presumably those that are greater than 10 \im in size.
This recovered material may be drained back into the blowing vessel, mixed
with raw asphalt or burned in one of the asphalt heaters. The exhaust from
the cyclone is then ducted to an afterburner or a process heater furnace.
-------
Air
45 Nm3/min
700 g/cm2
^
1
«-.— ^.-.-j-t
^ ^ -n
J
Spar
Blower
232 °C
Charge Asphalt |X] '
from Heated
Storage
to Atmosphere
50-150 Nm3/min
232-270°C
Recovered
Asphalt
H2O
(Steam)
40,000 Liter
Capacity
Pump
Fan
Blowing Time
2-4 hrs.
Blown Asphalt
to Storage
Figure 1. Schematic Diagram of Vertical Asphalt Blowing Still
-------
00
Return
Asphalt
Storage
Saturant Asphalt
Dry
Felt
Dry
Looper
OJ
1~~"tWWW\
Coating
Asphalt
Sand
MixerX /
V 280-560 Nm3/min
e-
Drying-ln
Section T°P
Coating
>- Continue to:
Top Surfacing
Backcoating
Cooling
Drying
Cutting & Packing
Saturating Tank
Figure 2. Schematic Diagram of Asphalt Saturating Line
to Atmosphere
Fan
-------
Saturator lines consist of dip tanks or sprays, or both, where the
saturant is applied to the felt. The saturated felt passes through a drying-
in section followed by application of coating asphalt to one side of the
saturated felt. Although mechanical problems or breakage of the felt may
cause intermittent shutdowns of the saturating line, it is essentially a
continuous operation.
These saturating and coating operations may be partially or totally
enclosed with air and fumes being exhausted from the enclosure to a con-
trol device (or directly to atmosphere). The volume of effluent exhausted
is on the order of 560 Nm^/min and consists primarily of air, water vapor,
asphalt liquid droplets (fume) and gaseous hydrocarbons.
Variations in process and operating parameter do affect resultant
emissions. Air-blowing emissions are generally lower from vertical stills,
but the emissions are also a function of the asphalt feedstock, blowing
rates and temperatures, and the meltpoint of the desired product. Saturating
line emissions are affected by indraft air and hooding arrangements, but
are also influenced by the characteristics of the asphalt and the felt,
as well as variations in the spraying/dipping process and line speeds,
etc.
Control Practices
One of the first steps in this study was to identify the types of
control equipment presently being used to control emissions from air-
blowing and saturator operations and determine the extent that they have
been applied to these two sources. A survey of the industry disclosed
that all blowing stills are controlled to some degree, either by ducting
the fume to a direct fired process heater or ducting them to an afterburner.
However, other information indicated that only about 40% of the plants do
their own blowing. With regard to the extent of control of asphalt satu-
rators, the MRI survey of 76 saturators showed the following:
Number of saturato_rs Control devices
28 Afterburners
18 HEAP
10 Electrostatic precipitator
9 Wet scrubber
11 Uncontrolled
76
It was found that all of the air-blowing operations are controlled
by some means of fume incineration, whether by ducting of the fume to
process heaters or installation of afterburners. However, it is possible
that the process heater control method might not provide the same destruc-
tion of hydrocarbons as an afterburner (i.e., proper operating temperatures
-------
and residence time). It was also noted that in some cases an afterburner
may be used to control emissions from both the air blowing and saturators,
but to our knowledge fume incineration techniques are the only method that
has been used, or tried, for control of air blowing operations. On the
other hand, several types cf devices have been installed for control of
asphalt saturator emissions. A list of all asphalt roofing plants is given
in Appendix A, with partial information on control equipment.
In order to evaluate the control techniques that are being used, a
search of the literature and other sources was made in an attempt to char-
acterize the emissions from air blowing and saturators, and to compile re-
sults of the test data on the control devices, as discussed in the next
section.
Characterization of Emissions
A search of the literature was carried out in order to compile avail-
able data on air blowing and saturator effluents and compile test data
for determining efficiency of control devices. One of the reasons for com-
piling this type of data was to form a basis for selecting representative
effluent characteristics for a plant model that could be used as a common
basis for evaluating the technical and economic feasibility of all types
of candidate control techniques.
Results of the above data compilation work are presented in Tables
1 and 2. For air blowing operations, these data showed that uncontrolled
particulate concentrations (which included condensible particulate) ranged
from 3.0 to 25.6 g/Nm3. Gaseous hydrocarbon content was determined separately
in some of these tests but because, of the sampling methods used, it is dif-
ficult to determine how much of these gaseous hydrocarbons may have been
collected as condensible particulate. Even so, the magnitude of the gaseous
hydrocarbon quantities indicate that a significant portion (> 1%) of the
emissions are probably in gaseous form.
These same data also showed that the removal efficiency of the after-
burners (process heaters) ranged from 84.6 to 95.8%, which is lower than
might have been expected. The reason for this cannot be definitely deter-
mined, but it is suspected that these process heaters, which served as
afterburners, probably did not provide proper mixing, operating tempera-
ture, and residence time for efficient destruction of the asphalt hydro-
carbons. In such heaters, firebox temperature is usually controlled by
the exit temperature of the asphalt, so control efficiency may be reduced
when firebox temperature is reduced..!'
Our examination of the literature did not reveal any data on particle
size distribution of the air-blowing emissions. It should be noted that
the air-blowing emissions are at higher temperatures (94 to 153 C), and
10
-------
Table 1. SUMMARY OF TEST DATA REPORTED FOR BLOWING EMISSIONS
Data
source
11
11
21
Gas flow
(Ntn3/min)
62.3
62.3
238.0
Temp.
CO
153
94-99
99
Inlet
Cone. (g/Nm3)
F. F. part.
part. I/ and cond.-'
25.6
3.39-8.15
2.97
Outlet
Cone . (g/Nm3)
Emissions^/
(kg/hr)
96W
ll-32.sk/
42. 5£/
Control
device F. part .I/
ABl/
AB*/
None
F. part.
and cond.6/
0.23
0.053-0.275
Emissions.!/ Control eff.
(kg/hr) (%)
4.0b-/ 95.8
0.86-5.0£/ 84.6-94.5
£/ Control device was a fume incinerator used to preheat asphalt.
b/ Separate tests for gaseous HC's showed average values of 15.9 mg/hr at the inlets and 8.5 kg/hr at the outlet. However, it is
not possible to determine what portion of these "gaseous" hydrocarbons, if any, may have been collected and reported as con-
densible particulates.
£/ Samples collected in plastic bags and analyzed by FID showed 2,500 ppm gaseous HC at room temperature. This would be equivalent
to about 24 kg/hr. However, the special particulate sampling method, intended for POM, cools the sample stream to -18°C so may
include some of the "gaseous HC" as condensible vapor.
d/ Filterable particulate.
£/ Filterable particulate and condensibles.
t] Total emission rate, in kg/hr, comprised of both filterable particulate and condensibles whenever such data were available.
I/ Gerstle, R. W., "Atmospheric Emissions from Asphalt Roofing Processes," EPA Report No. EPA-650/2-74-101, pp. 51 and 56, October
1974.
2_t Von Lehmden, D. J., R. P. Hongebrauck, and J. E. Meeker, "Polynuclear Hydrocarbon Emissions from Selected Industrial Processes,"
J. Air Pol. Cont. Assn., 15:7, July 1965.
-------
Table 2. SUMMARY OF TEST DATA REPORTED FOR SATURATOR EMISSIONS
Data
source
I/
T/
I/ •
I/
I/. 2/
-2'
I/. I/
i/
4/
^
£/
I/
I/
I/
Gas flow
(Nm3/min) Temp. (°
354 54
286 127
773 53-57
340 59
1| ...
Inlet (uncontrolled)
Cone. (s/Nm3)
F. part, and
C) F, part.-' cond.V
1.35
1.81
0.18-0.23
1.224
Emissions^/
(ke/hr)
28.6
30.9
8.6-10.9
25
567 60
595 52-61
708 53-58
575 33
391 (Unknown
1,195
844
940 58-74
297 60-92
173 99
0.952
0.034 0.066
0.080 0.101
0.114 0.162
1.227
ets) - 0.602
0.863
0.897
0.11-0.73
0.69-1.08
1.146
32.3
2.4
4.3
5.56
27.4
11.2
14.1
12.0
6.8-37.3^
12.3-15.4-
11.9
Control device
(with orecooline)
Scrubber (H20)
Scrubber (KMn04)
Scrubber and ESP
ESP (no)
ESP (no)
ESP (unknown)
HEAF (no)
HEAF (no )
HEAF (no)
HEAF (no)
HEAF (no)
ABi/
AB£'
Gas flow
(Nm3/mtn)
345
652
570
425
538
595
(See
inlets)
.
-
249
Outlet
Cone.
Temo. (°C) F. Dart.<*/
28
32 0.016
28
49 .0.0041
52 0.0059
32 0.0014
48
64
69
56
54-80
593-677
393
(B/Nm3)
F. part, and
cond-J^/
0.169
-
0.133
0.025
0.043
0.0076
0.1293
0.0494
0.0208
0.0188
0.037-0.183
0.183-0.275
0.076
EraisslonsS/
(kg/hr)
3.5
0.62
4.5
0.60
1.40
0.27
3.3
0 91"
uooi/
0.73
1.86-9.68^
4.23-8.05-
1.14
Control
eff.
(7.)
86.0
-
86
74.4
67.7
95.2
88
92
93
94
67.2-92.8
34.4-72.6
90.4
.a/ Filterable particulate.
£/ Filterable particulate and
.c/ Total
emission rate, in kg/
condensible particulate.
hr, comprised of both filterable
and condensible
particulate whenever such data
were available.
.d/ Two tests included sampling for gaseous hydrocarbons, showing these to be 0.27 and 1.95 kg/hr, but it could not be determined what portion of these may have been collected
and reported as condensible particulates.
j;/ Separate tests for gaseous hydrocarbons (HC) showed average values of 1.0 and 1.9 kg/hr at the inlets with 1.14 and 2.50 kg/hr at the outlets, respectively. However, it was
not possible to determine what portion of these gaseous hydrocarbons, if any, may have been collected and reported at condensible particulates.
_f/ Control device was a process heater furnace.
\l Cerstle, R. W., "Atmospheric Emissions from Asphalt Roofing Processes," EPA Report No. EPA-650/2-74-101, pp. 37 and 38, October 1974.
I/ Weiss, S. M., Air Pollution Engineering Manual. Public Health Service Publication 999-AP-40 (1967).
j/ White, H. R., Source Emission Test Report by Alar Engineering Corporation (Chicago, Illinois) for Fry Roofing Company of Brookville, Indiana, dated July 23, 1974.
1>I DeWees, W., and R. Gcrstle, Emission Test Report by PEDCo-Environmental (Cincinnati, Ohio) for the Celotex Corporation of Lockland, Ohio, dated April 1974.
,57 DeWees, W., and R. Cerstle, Emission Test Report by PEDCo-Environmental (Cincinnati, Ohio) for the Celotex Corporation of Fairfield, Alabama, dated March 1975.
,6_/ Netzley, A. B., "Control of Asphalt Saturators by Filtration," paper presented at the 68th annual meeting of the Air Pollution Control Association, Boston,
Massachusetts, June 15 to 20, 1975.
_7/ Nance, J., and W. L. Oaks, Emission Test Report No. C-2095 by Los Angeles Air Pollution Control District at Bird and Son, Inc., of Wilmington, California, dated
January 10, 1974.
-------
have higher grain loadings than saturator emissions. Opinions have also
been expressed that the asphalt emissions from air blowing are of a more
tar-like nature than those from saturator operations.
Data for saturator operations (Table 2) showed a wide variation in
uncontrolled grain loading of 0.066 to 1.81 g/Nm^. From those tests that
included determinations of both filterable and condensible particulate,
it is indicated that the percentage of hydrocarbons that exist in vapor
form in the stack is significant; being no less than 2% and as high as
48%. Some information is available on the particle size distribution for
saturator emissions (Figure 3) indicating a median particle diameter of
about 0.8 M>m.
Reported efficiencies of particulate (and condensible hydrocarbons)
removal for the various types of control devices were as follows:
Range of
particulate removal
Control device efficiency^
Afterburners 34.4-90.4%
ESP 74.4-95.2%
HEAP 62.7-94%
Wet scrubber 86%
Sif Removal efficiency for filterable particulate and conden-
sible hydrocarbons.
Most of the above efficiency data did not include analysis for gas-
eous hydrocarbons so these values do not necessarily reflect total hydro-
carbon removal efficiency.
In some cases, the control system employed on saturator emissions
includes some type of preceding, usually water sprays, which are believed
to enhance control efficiency. This is consistent with the expectation
that precooling would promote condensation and particle size growth for
the hydrocarbon emissions, thereby improving collection efficiency. How-
ever, it is difficult to quantify such effects because sufficient test
data were not available (i.e., efficiency tests with and without precool-
ing).
Nature of Gaseous Hydrocarbon Emissions
Some of the emission tests on air blowing and saturators did include
sampling and analysis of gaseous hydrocarbons. This was usually done in
conjunction with the tests for filterable and condensible particulates.
13
-------
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
I
O
2.0
N
t/i
"j i.o
u 0.9
f= 0.8
< 0.7
°" 0.6
0.5
0.4
0.3
0.2
0.1
I I
I
I
I
I
I I
I
I
I
5 10 20 30 40 50 60 70 80 90
PERCENT BY WEIGHT SMALLER THAN INDICATED SIZE
98 99
Figure 3. Particle Size Distribution—' in Uncontrolled
Saturator Exhaust
14
-------
These gaseous hydrocarbon tests were based on samples collected in plastic
bags that were later analyzed by flame ionization detectors. Such samples
probably also included some of the liquid droplets so it is difficult to
interpret the results in terms of how much gaseous hydrocarbons would re-
main in the normal sampling train after all filterable and condensible
hydrocarbon had been removed. Even so, the test data in Tables 1 and 2
do shed some light on the problem.
? /
Data for air blowing in tests conducted by PEDCo=' showed that the
controlled HC emissions exceeded the total of filterable and condensible
particulate. However, the control devices were process heaters, so the
gaseous hydrocarbon data could be very misleading.
The same set of data for uncontrolled emissions showed gaseous HC
values of 14.7 and 16.7 kg/hr while the filterable and condensible partic-
ulate for all tests averaged 32.1 kg/hr. As mentioned earlier, these data
cannot necessarily be interpreted to mean that gaseous hydrocarbons amount
to about 50% of the total filterable and condensible material. On the
other hand, the data do support the assumption that the uncontrolled ef-
fluent contains a significant portion of gaseous hydrocarbons, especially
considering the process and temperatures involved (232 to 260 C).
Considerably more test data were available for saturator effluent
and associated control devices. Three of these tests, when averaged, showed
filterable particulate concentrations of 0.075 g/Nm^ while respective total
particulate (filterable and condensible) concentrations averaged 0.110
g/NnP. These data would indicate that the condensible vapors were 31% of
the total. However, since the total concentrations for these tests were
much lower than several of the other test data G« 0.7 to 1.1 g/NnH) it
would be possible to argue that the percentage of condensible vapors might
be much lower (i.e., 2%) but this is still a significant amount relative
to the objective of 99% removal of all hydrocarbons. Therefore, it certainly
is reasonable to assume that the effluent from both air blowing and satu-
rators contain gaseous hydrocarbons in sufficient proportions (> 1%) such
that particulate removal devices could not achieve 99% removal effici-
encies unless some means could be provided to reduce the gaseous hydrocar-
bons to less than 1% of the total hydrocarbons.
It would, of course, be expected that precooling of effluent gases
would cause condensation of gaseous hydrocarbon. Very little data were
available on the two emission sources of interest that could be utilized
to determine how low a temperature would be necessary. The only data that
were available are that from Refs» 4 and 5 shown in Table 2.
15
-------
Data from Ref. 5 is the most interesting because of the low operat-
ing temperatures (33°C). Inlet concentrations showed 0.048 g/Nm^ of con-
densible particulate at 33°C while outlet concentrations showed 0.0062
g/Nm^ at 32°C. This would indicate that the control device (an ESP) was
removing most of the gaseous hydrocarbons, which is highly unlikely. It
is more probable that these data reflect the sampling problems associated
with this source. That is, liquid collected on the filter seeps through
the filter, is carried over into the impingers, and is reported as conden-
sible particulate.
Considering such problems as those discussed above and previous dis-
cussion of difficulties in interpreting the gaseous hydrocarbon sampling
data, it must be concluded that sufficient information is not available
for determining the temperature to which the effluent streams would have
to be cooled in order that less than 1% of total hydrocarbons would re-
main in gaseous form. Further, within the normal range of temperatures,
it may be that this is not possible at all. However, a current EPA sam-
pling program utilizes sampling and analyses methods that should provide
considerably more information in this area. This method requires that
the effluent sample be cooled to 38°C with collection of the filterable
and condensible particulate on fiberglass. It is then passed through
a final filter and remaining hydrocarbon gases determined by FID (flame
ionization detector).—' Results from this sampling method should provide
a considerable amount of data on the percent of total hydrocarbons that
remain in gaseous form at 38°C. In the interim, it has been assumed that
less than 1% of the total hydrocarbons will remain in gaseous form if
the effluent streams are cooled below 52°C.
POM Emissions
Control devices considered in this study were to be evaluated on the
basis of total hydrocarbon removal efficiency, as stipulated earlier. It
must be realized, however, that this does not directly address removal of
POM, which may be of concern. These compounds comprise a very small por-
tion (<< 1%) of the total hydrocarbons.—' Therefore, even if a device pro-
vided 99% removal of total hydrocarbons it might provide negligible control
of POMs.
It might be suspected that the air-blowing operations would be a more
significant source of POM than saturators. Tests reported by Hangebraucfc^'
on POM emissions from air blowing led him to conclude that the process
does not appear to emit significant amounts of BaP or other POMs of equal
or greater molecular weight, as compared to the quantities emitted by other
sources and effects on ambient air levels. He noted, however, that it does
emit large quantities of unidentified alkyl polynuclear hydrocarbons which
might be carcinogenic.
16
-------
With regard to removal of POM via control devices, test reported by
PEDCo=.' showed that in most cases the control methods reduced the POM in
direct proportion to the particulate removal. This led them to conclude
that it does not appear that roofing plants with typical controls are a
major contributor of these compounds (POM) to the ambient air.
Removal of POMs by "typical controls" as reported by PEDCo is consis-
tent with that found by Hangebrauck in studies of other industrial processes.
It was found that combustion of the emissions (afterburning) or wet scrub-
bing considerably reduced these emissions (> 90% reductions). Similar work
by Cuffed/ mentions that fly ash. collectors in coal-fired power plants
(e.g., electrostatic precipitators) showed significant collection of POM,
which was attributed to the expectation that POMs are, or are associated
with, particulate matter that is removed by the collectors.
Based on the foregoing information, we concluded that high efficiency
control devices (99%) applied to asphalt blowing and saturator emissions
would also provide high efficiency removal of POMs.
It is important to mention that the PEDCo report—' showed that in
tests on a process heater furnace used to control saturator emissions,
there was no decrease in POMs. Again, it is suspected that this oil-fired
process heater did not provide proper operating conditions for efficient
destruction of hydrocarbons, including POM. As mentioned earlier, properly
designed afterburners should provide efficient removal of these compounds.
The current EPA test program does include testing for POMs and should pro-
vide additional information in this area.
TECHNICAL AND ECONOMIC EVALUATION OF CANDIDATE CONTROL TECHNIQUES
The industry survey information discussed previously was used to
identify candidate control techniques and available data on these control
devices. This information was combined with engineering analysis and cost
estimates in order to evaluate the technical and economic feasibility for
95 and 99% removal of total hydrocarbons.
The industry survey revealed that the only control technique that
is being used or has been tried on air-blowing emissions is thermal in-
cineration (afterburners). For saturator emissions control, several tech-
niques are being used, and include afterburners, electrostatic precipita-
tors, filters (HEAF) and wet scrubbers. These devices were included in
the technical and economic evaluations, but other possible control methods
were also given at least preliminary consideration. Possible methods con-
sidered were wet electrostatic precipitators and catalytic afterburners.
The technical and economic evaluations for all control methods basically
consisted of three areas:
17
-------
1. Theoretical and engineering analysis of their capability for 95
and 997» removal of total hydrocarbons.
2. Estimates of capital, operating and annualized costs.
3. Tabulation of energy requirements, waste disposal aspects, operat-
ing problems, and other advantages and disadvantages including their appli-
cability for controlling both emissions sources (air blowing and saturating
lines).
Control efficiency for all devices except afterburners was based on
the assumption that use of precoolers would provide for more than 99% of
the hydrocarbons being in particulate form.
In order to provide a common basis for evaluating the control devices,
certain criteria were selected from the industry survey data to represent
a plant model. These criteria are shown in Table 3. Specific cost estima-
tion criteria were also set out as follows:
Depreciation period 15 years (straight line)
Interest and taxes 9%
Utilities
Natural gas $1.222/109 joules
Fuel oil $2.63/109 joules
Electrical $0.016/kw-hr
Operating time 5,480 hr/year for saturator controls
2,080 hr/year for blowing controls
Most of the actual test data on the control devices were not suffi-
cient for determining whether a specific type of device was or was not
capable of 95 or 99% removal efficiency. It was, therefore, necessary to
employ theoretical analysis techniques for this purpose. These techniques
usually necessitated use of particle size distribution data, some of which
were available for saturator emissions. No such data were available for
air-blowing emissions, but it seems reasonable to assume that the satu-
rator emissions represent a control problem no less difficult to control
than air-blowing emissions because for saturators the grain loading is
low and the mean particle diameter is quite small (0.8 P-m).
Details of the technical and economic analysis, calculations that
were carried out for specific types of control devices, descriptions of
the devices and their characteristics are presented in Appendices B, C,
D, and E.
Results of the technical and economic analyses, along with other
information considered important for comparison purposes, have been tabu-
lated and are presented in Tables 4 and 5. A listing of advantages and
disadvantages is given in Table 6.
18
-------
Table 3. SELECTED EMISSION CRITERIA
Range of avilable data Selected value
Saturator emissions
Gas flow 173-1,195 Nm3/min 567 Nm3/min
Temperature 33-127°C 68°C
Grain loading 0.066-1.81 g/Nm3 0.80 g/Nm3
Moisture content 1.5-3.8% 3%
Blowing emissions
Gas flow 62-567 Nm3/min 567 Nm3/min£'
Temperature 94-153°C 121°C
Grain loading 3.0-25.6 g/Nm3 8.4 g/Nm3
Moisture content 5.1-19.57. 16%
ji/ Gas flow of 567 Nm3/min was selected to represent control devices de-
signed for control of more than one air blowing vessel and to
facilitate comparison with saturator controls.
19
-------
Table 4. EVALUATION OF CONTROL EQUIPMENT FOR ASPHALT SATURATOR EMISSIONS
Afterburners
Thermal
95%.
A. Collection efficiency for No
99%
No
Catalytic
9-5% 99%
No No
HEAP
957.
Yes
99%
Yes
Electrostatic precipitators Wet
Wet
95%
Yes
99%
Yes
Dry scrubbers
95%
Yes
99% 957. 99%
Yes No No
total hydrocarbons.
(Refers to collection of
particulate hydrocarbon
equivalent to precooling
to 52°C.) Indicate need
to precool (Yes or No).
B. Can specified efficiency
be reasonably expected
without being preceded
by other control devices,
except cyclones and/or
precobling (Yes or No).
C. Estimated costs'^
Installed cost, $
Operating cost, $/year
Annual!zed cost, $/year
D. Does system generate
other effluent problems
(Yes or No - and what
type).
E. Has similar system been
operated on this source
(Yes or No). If yes, at
what range of particulate
collection efficiency.
F. Are plugging, poisoning,
or other problems con-
sidered likely (Yes or
No). Specify if yes.
Yes
Yes
312,000^'
61,200f
110,
327,000r.
63,900; .
116, 000^'
Yes
(34 to over 907.)
No
Yes
Unknown
Removal costs comparable
with thermal afterburners
(higher capital cost but
lower fuel costs)
-100,C
-19,000*'—
-34,700^'-
(Disposal of used
mats and precooling
effluents)
(88 to 92%)
5,500^
32,000^
Yes
5,500
113,000^'
6,000- .
23,7004il/
Some (periodic washing)
Yes
(74 to 95%)
—Unknown—
(Potential plugging and
poisoning of catalyst bed
with asphalt liquid par-
ticles)
No
No
No further
information
listed, be-
cause analysis
indicates that
required effi-
ciency cannot
be reasonably
achieved.
-------
Table 4. (concluded)
N5
G.
H.
I.
J.
K.
It
~£./
Thermal
957. 997.
controlling both types
of sources (airblowing
and saturator) (Yes
or No).
Energy Reauired (for
567 Nm3/min') 109 joules/hr
- Electrical Power None None
- Natural Gas or Oil 38.2 39.9
- Other (max. without
heat recovery
Process Availability NP NP
(Proprietary or
NonProprie tary )
Are any design or No No
operating parameters
tical limits of the
required efficiency.
(Yes or No).
Are there any unique No No
operating or main-
tenance requirements
(Yes or No) . Identify
if yes.
Afterburners Electrostatic precipitators
Catalytic HEAF Wet Dry Wet scrubbers
957. 997. 957. 997. 957. 997. 957. 997. 95?. 99%
(Unless preceded with information
filter to remove listed, be-
asphalt liquid cause analysis
particles) indicates that
required effi-
ciency cannot
be reasonably
— JO. 0 i 0
) covery)
P P PP NPNPNPNP
catalyst bed volume are
near maximums at 957. re-
moval
of electrodes due
to asphalt liquid
particles
Installed cost based on 807. heat recovery, because this represents lowest annual ized cost (see Appendix B).
Operating cost with natural gas fuel.
For more information on HEAF cost estimates see Appendix C.
For more information on ESP cost estimates see Appendix E.
Costs do not include any additional operating labor.
Includes costs for precooling the effluent to effect condensation.
-------
Table 5. EVALUATION OF CONTROL EQUIPMENT FOR AIR BLOWING EMISSIONS
Afterburners
Electrostatic precipitators
Thermal Catalytic HEAF Wet Dry
95%
A. Collection efficiency for No
99% 95% 99% 95%
No No No Yes
99% 95% 99% 95%
Yes Yes Yes Yes
997.
Yes
Wet
scrubbers
95%
No
997.
No
total hydrocarbons.
(Refers to collection of
particulate hydrocarbon
equivalent to precooling
to 52°C.) Indicate need
to precool (Yes or No).
to
B. Can specified efficiency
be reasonably expected
without being preceded by
other control devices, ex-
cept cyclones and/or pre-
cooling (Yes or No).
C. Estimated costs"*"
Installed cost, $
Operating cost, $/year
Annuallied cost, J/year
D. Does system generate
other effluent problems
(Yes or No - and what
type).
E. Has similar system been
operated on this source
(Yes or No). If yes, at
what range of particulate
collection efficiency.
F. Are plugging, poisoning,
or other problems con-
sidered likely (Yes or
No). Specify if yes.
Yes
Yes
304,000?'
290,000r.
24-500Ib/ -' --.b/
69,900-^' 73,000"U-2/
Yes
(85 to over 95%)
No No
(HEAF unit recommended
in front)
Costs comparable with
thermal afterburner
(higher capital cost
but lower fuel cost)
No
-No-
-Unknown—
-Unknown-
No
potential plugging
and poisoning of
catalyst bed with
asphalt liquid
particles)
100, O
20,700; .
36,400s-"*'
(Disposal of used
mats and precool-
ing effluents)
Po s s ibly
(Plugging of mat)
32
^*
,f
,000^'
169,000^*
—Unknown—
(Fouling of electrodes
is expected to be a
serious problem)
No further
information
listed, be-
cause analysis
indicates that
required effi-
ciency cannot
be reasonably
achieved
-------
Table 5. (concluded)
Afterburners Electrostatic precipitators
Thermal Catalytic HEAP Wet Drv Wet scrubbers
G.
H.
I.
;.
K.
Is device amenable
controlling both types
of sources (air blowing
and saturator) (Yes
or No)
Energy Required (for
567 NnrVmin) 109 joules/hr
- Electrical Power
- Natural Gas or Oil
- Other
Process Availability
(Proprietary or
Non Proprietary)
Are any design or
operating parameters
near or outside prac-
tical limits of the
device, for the
required efficiency.
(Yes or No) .
Are there any unique
operating or main-
tenance requirements
957. 997. 957. 997. 957. 997. 95% 99% 95% WZ 957.
(unless preceded info
with filter to list
remove asphalt caus
liquid particles) indi
requ
cien
be r
heat recovery) recovery)
NP NP P P P P NP NP NP NP
catalyst bed volume
are near maximums at
957. removal.
(Fouling of
electrodes)
997.
urther
rmatlon
ed be-
e analysis
cates that
ired effi-
cy cannot
easonably
eved.
(Yes or No). Identify
if yes.
j/ Installed cost based on 807. heat recovery because this represents lowest annualized cost (see Appendix B).
b/ Operating cost with natural gas fuel.
jc/ For more information on HEAP cost estimates set Appendix C.
ji/ For more information on ESP cost estimates see Appendix E.
.e/ Costs may be about the same as for saturators but dry ESP's are not thought to be
applicable to air blowing.
_f/ Costs do not include any additional operating labor.
£/ Includes costs for precooling the effluent to effect condensation.
-------
Table 6a. AFTERBURNER ADVANTAGES AND DISADVANTAGES
Technical
considerations
Thermal
Advantages
Presently used in the
asphalt roofing industry.
Removal efficiency from
89 to over 95%.
997= removal efficiency is
possible with little in-
crease in fuel consumption.
Capable of controlling
emissions from blowing
and saturators.
Disadvantages
Fuel availability is a
potential problem.
Catalytic
Advantages
Lower operating temper-
ature than thermal
afterburner saves fuel.
Can achieve over 957.
removal efficiency.
Disadvantages
Potential plugging and
poisoning of catalyst bed.
HEAF unit recommended to
precede catalytic after-
burner to prevent plugging.
(Requires high pressure
drop.)
Design and development tests
are required.
Fuel availability.
N>
-P-
Low maintenance require-
ments.
Long operating life.
Available designs and
hardware.
No secondary wastes are
generated.
Economic
considerations
Low maintenance costs.
Operating cost may be
quite small if recovered
energy (steam) is re-
quired by plant.
Fuel cost is high especially
if no heat recovery system
is used.
Capital cost is high.
Lower fuel cost than for
thermal afterburner
(custom design and
catalyst costs).
Potential low mainte-
nance costs.
Potential recovery of
spent catalyst.
HEAF unit may also be re-
quired at added cost.
Fuel costs are high with no
heat recovery system.
Increased electric fan power
costs over thermal after-
burner .
Potential long-lead delivery
time for development.
Higher capital cost.
-------
Table 6b. HEAF ADVANTAGES AND DISADVANTAGES
Technical
considerations
N5
Economic
considerations
Advantages
• Presently used in roofing industry
(saturators)
• Capable of obtaining high collection
efficiencies (87 to 94% on particulates)
• Available designs and hardware.
• Ease of maintenance and operation.
• Operation of unit dependent only on
electrical power.
• Lower cost than afterburner.
• Mat costs are low.
Disadvantages
Not capable of controlling gaseous
hydrocarbons.
Requires cooling of inlet gases to
maximize collection of condensible
particulates. This adds to mainte-
nance and operating requirements.
Requires frequent replacement of
fiber mats.
Disposal of used fiber mats is a
potential problem.
With water cooling, maintenance/
operation costs probably will in-
crease and resultant liquid may
need to be treated.
-------
Table 6c. ELECTROSTATIC PRECIPITATOR ADVANTAGES AND DISADVANTAGES
Technical
considerations
ro
Economic
considerations
Advantages
Have been used successfully in asphalt
roofing industry on saturators.
Modular design allows achievement of high
removal efficiency (up to 997=,).
Long operating life.
Available designs and hardware.
Wet-wall units may be applicable to
blowing operations.
Lowest operating cost.
Low maintenance cost.
Disadvantages
Precooling is probably necessary to
condense gaseous hydrocarbons.
Dry unit not applicable to blowing
operations because it is expected
that higher grain loading would re-
sult in excessive fouling of
electrodes.
Secondary water treatment may be
required for wet-wall units.
Further testing required on wet-wall
units for saturator/blowing opera-
tions.
Requires frequent cleaning.
Lower capital cost than afterburners,
-------
A summary of results and interpretation of this and other information
are presented in the following paragraphs for each type of device.
Afterburners
Theoretical analysis of thermal afterburners indicated that properly
designed units (0.3 sec residence time at 816 C) should be capable of
providing 99% removal of total hydrocarbons. However, in comparison with
other types of devices, their capital cost is high. Also, due to the fuel
requirements their operating costs are also high. Of course, these operat-
ing costs are a function of the heat recovery equipment that may be used,
but such equipment increases the capital cost. Operating experience indi-
cates that when afterburners are used to control air blowing, the emission
may provide 20 to 257» of the heat requirement but the fuel costs are still
quite high.
Under certain circumstances it is possible that the operating cost
for a thermal afterburner system would in reality be much lower if the
fuel would still be consumed for plant energy requirements. For instance,
if the asphalt plant is one that manufactures its own felt, a boiler would
probably be used to provide the steam necessary for this process. However,
the fuel that would be used to fire the boiler could instead be diverted
to an afterburner equipped with a waste heat boiler that would supply part
or all of the steam required. It is obvious that under these circumstances
the fuel used in the afterburner would not represent an increase in plant
operating costs. Such situations could make afterburners more attractive
but this is dependent on the type of plant and other individual plant
factors.
Aside from capital and operating cost considerations, thermal after-
burners have additional advantages in that they can and are being used
for controlling emissions from both air blowing and saturating lines. Con-
trol methods which might offer lower capital cost and/or operating cost
exist but these have not been used for control of air blowing. The reason
that some of these other devices have not been used, or tried, was not
clear.
Catalytic afterburners were also considered in this study as a pos-
sible candidate control device. This type of afterburner employs a catalyst
to increase rate of reaction. The required operating temperature is lower
(427 to 538°C) so the fuel requirements, and therefore the operating costs,
are lower. However, the capital costs of such units are higher than those
for thermal afterburners, and our analysis of the total annualized costs
for both types of afterburners indicated that they would be about equal.
Further evaluation of the catalytic afterburners indicated that the
quantity of catalyst required to achieve the desired removal efficiencies
27
-------
(95 to 99%) was approaching the maximum for which such units have been
designed. More important considerations were that the asphalt particulates
contained in the emissions could plug or foul the catalyst, and that the
effluent stream would contain sulfur or other compounds that could poison
the catalyst. Our discussions with manufacturers of catalytic afterburners
led us to conclude that the particulate matter could very well be a serious
problem and that a particulate removal device would be necessary upstream
of a catalytic afterburner. Assuming that to be true, it was concluded that
the catalytic afterburners probably do not have any advantage over thermal
afterburners for controlling emissions from asphalt blowing or saturator
lines.
HEAP
The HEAP unit has been used at several plants for control of satura-
tor line emissions. It is basically a particulate filter consisting of a
moving fiberglass mat through which the effluent passes. The system ex-
hibits a moderately high pressure drop of about 64 cm H~0. Because of the
mechanisms involved in filtration and the properties of the emissions, it
was difficult to project the control efficiency capability of this device,
either by theoretical means or analysis of existing test data.
Our interpretations indicated that the HEAP might be capable of up to
99% removal efficiency when preceded by a precooling section of water
sprays. Use of such sprays does add some cost, and may require treatment
of liquid effluent from the precooling section. Also, the HEAP device in-
volves the secondary problem of disposal of the used mats.
Estimates of the cost of a HEAP system indicated that the costs were
higher than for ESP devices, but both were considerably lower than after-
burners.
The HEAP has been used only for control of saturator emissions. It
would appear that when equipped with precoolers they might also be appli-
cable to control of air blowing. Air-blowing effluent does have a higher
particulate loading which would necessitate a faster usage rate for the
filter mat, but the cost involved is small in comparison with other operat-
ing cost factors for HEAP units. Other problems might occur due to the
nature of the air-blowing emissions, including the possibility of rapid
plugging of the mat, but none could be identified with enough certainty
to reject the HEAP as a candidate device for control of air-blowing emis-
sions.
28
-------
Wet Scrubbers
Wet scrubbers have been used for control of saturating-line emissions,
primarily in those plants operated by Fry Roofing Company. Very little test
data were available for the wet scrubbers, but the highest reported effici-
ency was 86%.
Because of the lack of data, we proceeded with a theoretical analysis
of the removal efficiency capability of these devices (see Appendix D). The
analysis was aimed at determining the pressure drop that would be required
for efficient removal (95 to 997.) of a source exhibiting the particle size
distribution shown previously in Figure 3. Results of the analysis indi-
cated that such devices are not technically feasible for the desired re-
moval efficiencies because of the excessively high pressure drops that would
be required (> 1,000 cm K^O).
Some additional information and evaluation of the applicability of
wet scrubbers for control of asphalt emissions by other authors was lo-
cated during this study. Goldfield^' reported that in small-scale tests
of different control devices, a packed tower wet scrubber operating at
15 cm t^O. AP was not promising because it achieved only 58% removal ef-
ficiency. He also mentions that the system that included exhaust from
blowing stills showed signs of excessive corrosion and may necessitate
use of special materials.
Another article, published in the Oil and Gas Journal"" stated that
aerosols (from air blowing) are highly resistant to control by wet scrub-
bing. Similarly, the Los Angeles Engineering Manual— indicates that spray
scrubbers have met with limited success on control of saturators, and even
though their efficiency may be as high as 90%, the effluent may still be
50 to 1007» opaque because of their low collection efficiency for small
(< 1 M-m) particles.
MRI has recently visited two plants that involve scrubbers on satu-
rator emissions. One was using a scrubber preceding an ESP and the other
was using a scrubber alone. From the combination unit there were no visi-
ble emissions, which was not the case for the scrubber alone.
Based on the above information, it appeared very doubtful that wet
scrubbers would be capable of achieving the desired removal efficiency
(95 to 99%). It was not considered worthwhile to proceed with cost esti-
mates for these devices nor evaluate other factors, such as possible
water treatment problems.
The above conclusion seems justifiable for wet scrubbers in general,
but it must be pointed out that the scrubber systems employed by Fry Roof-
ing involve use of a permanganate scrubbing solution. This solution is a
29
-------
strong oxidizing agent generally used to remove organic vapor contaminants
from gas streams. Use of such solutions may or may not improve particulate
collection efficiency, but it is more likely that they could provide for
removal of certain organic vapors, such as odors and possibly POMs. It is
possible that weak permanganate solutions could also be used as the pre-
cooling medium upstream of a particulate removal device (HEAF or ESP).
(Further testing of such permanganate scrubber units may be of interest.)
Electrostatic Precipitators
ESPs have been used to control saturator emissions. One such device
that has been utilized at several plants is a two-stage dry ESP termed
the "Smog Hog."™ A description of this device and ESPs in general is con-
tained in Appendix E.
uTM
"Smog Hogs" are equipped with a mechanical prefilter to remove
larger particles (> 10 Urn) and minimize carry-over of water droplets
into the ESP. They also are equipped with a similar afterfilter (pre-
sumably to collect particles that may be exhausted due to reentrainment).
The "Smog Hog" manufacturer indicates that for proper operation in-
let air must be lower than 60 C. A premist or water spray systems is nor-
mally included for this purpose as well as to extend the required mainte-
nance cycle. Tar-like buildup on collector plates requires cleanout or
detergent washing at periodic intervals. The manufacturer did not believe
the dry ESP units were suitable for control of air blowing because of the
higher grain loading (i.e., > 2.3 g/Nra3).
ESPs can be very efficient for removal of fine particles (< 2
and the theoretical analysis carried out by MRI verified that they should
be capable of 95 to 99% removal of particulates. For high efficiency re-
moval of total hydrocarbon emissions it was assumed that this could be
accomplished if the ESP were preceded by water spray precoolers. Some
plants already employ precoolers upstream of an ESP or HEAF.
Economic analysis of the ESPs indicated that their costs would be
somewhat lower than HEAF units. The ESPs also have the advantage of lower
pressure drop and consequently, lower energy requirements. However, the
dry ESPs are subject to some operating problems due to fouling of the
electrodes and collection surfaces with asphalt. It is commonly required
that they be shut down and washed out with water or detergent on a peri-
odic basis.
30
-------
It has been indicated to MRI that the problem of fouling in ESPs
probably accounts for the fact that they have not been used for control
of air-blowing emissions where the grain loading is higher and the emis-
sions may be of a more tar-like nature. An alternate approach could be a
wet ESP for control of such emission sources. Wet ESPs have not been tried
on either source to our knowledge, but in other industries they have been
used succesfully for control of sources having similar characteristics.
The cost of wet ESPs is somewhat higher than the dry type, but their pos-
sible advantages appear to justify further investigation of their appli-
cability.
31
-------
RECOMMENDATIONS
The technical and economic evaluation of candidate control devices
was used as the basis for specific recommendations for further research
and development efforts. The objective of these efforts would be to pro-
vide the capability to reduce total hydrocarbon emissions by 99% from air
blowing and saturating lines. Some of these recommendations are founded
on unproven assumptions, as discussed earlier. It is important that the
following recommendations should be reexamined using results of the on-
going EPA test program when they become available. The recommendations
are discussed and presented below. These recommendations and their sug-
gested priority are also summarized in Table 7.
PRIORITY I
It was found in this study that there were little actual data avail-
able from which one could ascertain that various control devices are cap-
able of 95 or 997» particulate removal efficiency for emissions from air
blowing and saturating lines. Likewise, there were even less data rela-
tive to total HC removal efficiency (i.e., percent of the hydrocarbons
that are in gaseous form).
Because a current EPA program is directed to testing of the subject
control devices, and does include some determination of gaseous hydrocar-
bons, it is recommended that these results be compiled and evaluated as
soon as they are available. Specifically, they should be analyzed in an
attempt to ascertain the .total HG removal efficiency of afterburners on
air blowing and saturator emissions, and total HC removal efficiency of
HEAP and ESP units on saturators with notation of inlet temperature and
whether the units were or were not equipped with precoolers.
If this analysis shows that the air-blowing emissions still contain
a significant amount of gaseous hydrocarbons at the sampling (filter)
temperature of 38°C (e.g., > 1%) it would have to be concluded that no
jpart leu late control device could achieve 99% removal of total HC on air-
blowing emissions. A similar situation would be true if the saturator
32
-------
Table 7. SUMMARY OF RECOMMENDATIONS
Priority I
A. Evaluate results of current EPA test program, as soon as they be-
come available, to assess percent of hydrocarbons that exist as
gases at 38 C.
Use this assessment to determine if Priority II recommendations
should be undertaken.
Priority II
A. Conduct additional tests, as may be necessary, to characterize
both emission sources for specification and design of control
devices. This testing may include the following:
1. Determine particle size distributions (especially air-
blowing emissions).
2. Measure in situ resistivity of particulate matter.
3. Use EPA test train, with temperature control, to deter-
mine percent gaseous HC as a function of temperature.
Priority III
A. Obtain and install a pilot scale (=" 30 m3/min) HEAP and wet ESP
(with precooling section) on a sidestream drawoff from an air-
blowing operation. Test both devices to determine total HC
removal efficiency and assess possible operating problems.
If results of the above tests are favorable, proceed with full
scale tests of these control devices. Such tests would pre-
ferably be carried out at a location where air blowing and
saturator emissions could be diverted to the test units either
singularly or in combination.
Priority IV (Optional)
A. Perform experiments to determine efficiency of gaseous HC re-
moval via permanganate scrubbing. Tests could be conducted at
an operating KMnO/ scrubber facility or by using KMnO^ solu-
tion as the precooling medium in Priority III activity.
33
-------
emissions arc more than 1% gaseous HC at 38°C. In this case, the Priorities
II and III recommendations should be reevaluated.
Conversely, if the EPA tests show that the emissions are less than
17. gaseous HG at 38° C the Priorities II and III recommendations should be
pursued.
It is important to note that the technical evaluation of afterburners
has led us to conclude that properly designed afterburners should certainly
be capable of 99% removal of total hydrocarbon emissions either from air
blowing or saturators. Because of this, and considering the fact that after-
burners are presently being used on both sources, we believe their appli-
cability has been sufficiently demonstrated. No further research and de-
velopment has been recommended for afterburners. If evaluation of the EPA
test program should cast doubt on their capability for 99% removal, some
further R&D work may be needed.
PRIORITY II
Priority II recommendations center on characterization of emissions,
primarily to permit optimum design of control equipment especially as it
may be needed for the pilot-scale devices recommended under Priority III.
It is recognized that some of the necessary design information should be
available from the current EPA test program. However, our familiarity with
that program indicates that some additional testing will probably be re-
quired to sufficiently characterize the emissions for specification and
design of control devices. Therefore, the following work has been recom-
mended pending evaluation of the data from the current EPA test program:
1. Determine particle size distributions of emissions, especially
those from air-blowing operations.
2. Measure in situ resistivity of particulate matter in the ef-
fluent streams.
3. For both air-blowing and saturator emissions, conduct tests to
determine percent of the total hydrocarbon emissions that exist in gaseous
form as a functon of temperature.
PRIORITY III
Technical evaluation of candidate control techniques indicated that
HEAP and wet ESPs (equipped with precoolers) may be feasible for control
of air-blowing emissions, at a significant cost reduction over afterburn-
ers. Likewise, the same should be true for saturator emissions since HEAP
and dry ESPs are already used in this application and are currently being
34
-------
tested. It is recommended that a program be undertaken to install and test
a pilot-scale HEAP and wet ESP on air-blowing emissions. The objective of
this program would be to perform tests to determine removal efficiency and
identify operating problems. These preliminary pilot-scale tests would be
evaluated to decide if full-scale tests of either or both units are war-
ranted.
PRIORITY IV (OPTIONAL)
Evaluations carried out in this program indicated that wet scrubbers
could not reasonably achieve 99% reduction in particulate emissions. How-
ever some operating wet scrubbers utilize KMnO^ scrubbing solution which
might be advantageous for reducing gaseous hydrocarbon emissions. It is
therefore optionally recommended that this be investigated, either by tests
on an operating KMnO/^ scrubber at one of the asphalt production facilities
that is so equipped or as an addition to the pilot-scale tests recommended
under Priority III using KMnO^ solution as the precooling medium.
The recommendations that resulted from this study vary in their com-
plexity and requirements. To further delineate the work and procedures that
would be involved in carrying out these recommendations, estimates have been
prepared showing the cost, time and approach for each. The next section pre-
sents these estimates.
35
-------
PLANNING
Final work on this study was directed, to the estimation of the cost,
time and approach necessary for carrying out the recommendations that have
been made. These are discussed below and results have been summarized in
Table 8.
PRIORITY I - EVALUATION OF RESULTS FROM CURRENT EPA TEST PROGRAM
This recommendation will require reduction and analysis of data to
determine total HC removal efficiency of devices tested and determine the
percent of the emissions that remain in gaseous form at 38 G. When all
the data from this test program are available, it should be possible to
conduct the data analysis in 1 man-month at an estimated cost of about
$4,000, which includes preparation of a summary report.
PRIORITY II - CHARACTERIZATION OF EMISSIONS
This effort will be directed to characterization of emissions, to
the extent it is needed for design of control devices and is not already
available from other studies. This testing is expected to include, but
not necessarily be limited to, those test plans presented below.
Particle Size Tests on Air-Blowing Emissions
These tests should require only about two men for 1 week at a cost
estimated to be $3,100. The total time involved should be about 1 month
but these tests could be carried out in conjunction with the other tests
described below.
Measurement of In Situ Resistivity
Resistivity of the fume from asphalt saturators and air-blowing
operations should be determined at least twice at two different facili-
ties, preferably including one that utilizes a precooling system. It
is estimated that this would cost about $2,000 and require one calendar
month. Again, these tests could be carried out in conjunction with other
tests, probably at less cost.
36
-------
Table 8. SUMMARY OF ESTIMATED COSTS AND TIME REQUIREMENTS FOR
RECOMMENDED RESEARCH AND DEVELOPMENT PLANS
I.
Recommendation
Evaluation of results from current
EPA test program.
II. Characterization of emissions
a. Particle size testing of
air-blowing emissions
b Resistivity measurements
c. Tests to determine percent
gaseous KG as a function
of temperature
III. Pilot-scale testing of HEAF
and wet ESP on air blowing
IV. Optional research
a. Gaseous HC removal by
scrubbing
Time required Estimated
(months) cost ($)
4,000
3,100
2,000
15,400
62,400
16,000
37
-------
Tests to Determine Percent of HC Emissions that are in Gaseous Form as a
Function of Temperature
The simplest method to obtain these results would be to use the sam-
pling train being employed in the present series of EPA tests. However,
this train would have to be modified to include capability for varying
the temperature at the filter from 177°C down to 38°C. It would again
necessitate use of the FID for determination of gaseous HCs, or perhaps
special collection media in impingers after the filter. It is estimated
that the sampling train could be suitably modified at a cost of less than
$1,000 during the 1-month period required for site arrangements and sam-
pling preparations.
Actual field tests at each site (one air blowing and one saturator)
would require four men for 1 week at a cost of $6,200, for a total of
$12,400. Compilation and evaluation of the data in report form would re-
quire another 2 man-weeks at a cost of about $2,000 but the calendar time
required would probably be 1-1/2 months. Therefore, the total cost is esti-
mated to be $15,400 and would require about three calendar months.
PRIORITY III - PILOT SCALE TESTING OF HEAP AND WET ESP ON AIR BLOWING
This is the major recommendation of the study, and requires the most
effort. It will require procurement of the pilot scale test devices, ar-
rangement for their installation at an operating plant with modifications
necessary and utility hookups, testing of the devices, and reduction and
evaluation of data obtained.
Specification and procurement of the devices can be carried out dur-
ing the same period that arrangements are being made for their installation.
We estimate that this will require 2 man-months over a 3-month period, at
a cost of $7,600.
Pilot-scale tests should be conducted with units designed for flow
rates of about 30 nrYmin (or more). Conversations with HEAF and wet ESP
manufacturers verified that such units are available. The HEAF unit can
be purchased for close to $5,000 but apparently this size unit cannot be
rented or leased. Wet ESP units of this size are available for lease, but
at a cost of $5,000/month. Based on these figures we estimate that total
cost including auxiliaries and installation, would probably be a minimum
of $10,000 each. If water treatment facilities are necessary for the short-
term use of these test units, the cost will be higher. Assuming that this
will not be required, total cost for both units will be $20,000.
38
-------
After the units are operational, they should be operated at least
8 hr/day over a 2-week period for shakedown and adjustment and for identi-
fication of possible operating problems. This can probably be accomplished
with two men for 2 weeks (1 man-month, $5,800). Following that, each unit
should be tested over a 1-week period (a minimum of five tests) which will
require a test crew of eight men at a cost of approximately $12,500 or
$25,000 for testing both units.
Once the pilot-scale tests have been completed it will probably re-
quire at least 1 month to obtain all of the laboratory test results. Eval-
uation of these data and report preparation can probably be carried out by
the project leader in 1 man-month at a cost of about $4,000.
In summary, the pilot-scale testing of a HEAP and wet ESP on air-
blowing emissions can probably be done in a 6-month period at a cost
estimated to be $62,400.
PRIORITY IV - OPTIONAL RESEARCH
Optional testing of KMnO, scrubbing to determine efficiency of gaseous
hydrocarbon removal might be carried out at one of the asphalt plants where
this system is being used or it could be investigated as an addition to
the pilot-scale tests recommended under Priority III. For the purpose of
this report, it has been assumed that tests would be carried out at one
of the existing installations.
Actual tests of one KMnO/ scrubber installation would require an
eight man test crew for 1 week at a cost of about $12,500. However, it
will also be necessary to allow 1 man-month for making test arrangements,
observing field tests, and evaluating data including report preparation.
Therefore, the total cost is estimated to be $16,000 and will require about
three calendar months, allowing for lag in obtaining laboratory data.
39
-------
REFERENCES
1. Laster, L. L., "Atmospheric Emissions from the Asphalt Industry,"
EPA Report No. EPA 650/2073-046 (December 1973).
2. Gerstle, R. Wo, "Atmospheric Emissions from Asphalt Roofing Processes,"
EPA Report No. EPA 650/2-74-101, prepared by PEDCo-Environmental
of Cincinnati, Ohio (October 1974).
3. Von Lehmden, D. J., R0 P. Hangebrauck, and J. E. Meeker, "Polynuclear
Hydrocarbon Emissions from Selected Industrial Processes," J. Air
Pol. Con. Assn., 15(7) (July 1965).
4. Cuffe, S. T., "Emissions from Coal-Fired Power Plants: A Comprehen-
sive Summary," U.S. Department of Health, Education, and Welfare,
Publication No. 999-AP-35, Cincinnati, Ohio (1967).
5. Goldfield, J., and R. G. McAnlis, "Low-Voltage Electrostatic Precipi-
tators to Collect Oil Mists from Roofing Felt Asphalt Saturators
and Stills," published in the American Industrial Hygiene Journal,
24(4) (July-August 1963).
6. "Americans Fume Burner Cleans Air," Article Published in The Oil and
Gas Journal (March 20, 1972).
7. Air Pollution Engineering Manual, Public Health Service Publication
999-AP-40 (1967).
8. "Emission Tests in Support of Development of New Source Performance
Standards for the Asphalt Roofing Industry," Emission Standards
and Engineering Division, OAQPS, Environmental Protection Agency,
Research Triangle Park, North Carolina (1975).
40
-------
APPENDIX A
LIST OF ASPHALT ROOFING PLANTS AND
EMISSION CONTROL DEVICES
41
-------
LIST OF PLANTS^/ ASPHALT ROOFING MANUFACTURERS
AND EMISSION CONTROL DEVICES UTILIZED
Name and Address
Allied Material Corporation
Stroud, Oklahoma
Artie Roofings, Inc.
Edge Moor
Wilmington, Delaware 19809
Atlas Roofing Manufacturing
Company, Inc.
P.O. Box 1606
Meridian, Mississippi
Bear Brand Roofing, Inc.
P.O. Box 206
Bearden, Arkansas
Big Chief Roofing Company
P.O. Box 980
Ardmore, Oklahoma
Bird and Sons, Inc.
East Walpole, Massachusetts
Carey, Philip Manufacturing
Company
Lockland
Cincinnati, Ohio 45215
(now part of Celotex)
The Celotex Corporation
1500 North Dale Mabry
Tampa, Florida 33607
Certain-teed Products
Corporation
Valley Forge, Pennsylvania 19481
Plant Location
Stroud, Oklahoma
Albuquerque, New Mexico
Edge Moor, Delaware
Meridian, Mississippi
Bearden, Arkansas
» Ardmore, Oklahoma
Charleston, South Carolina
Martinez, California
Norwood, Massachusetts
Perth Amboy, New Jersey
Portland, Oregon
Shreveport, Louisiana
Wilmington, California
Houston, Texas
Lockland, Cincinnati, Ohio
Memphis, Tennessee
Perth Amboy, New Jersey
Wilmington, Illinois
Birmingham, Alabama
Camden, Arkansas
Chester, West Virginia
Chicago, Illinois
Los Angeles, California
Philadelphia, Pennsylvania
San Antonio, Texas
Goldsboro, North Carolina
Avery, Ohio
Chicago Heights, Illinois
Kansas City, Missouri
Miller (Dallas), Texas
Port Wentworth, Georgia
Richmond, California
Tacoma, Washington
York, Pennsylvania
Minneapolis, Minnesota
Saturator-/
Control
(type)
AB
AB
HEAP
AB
AB
ESP
HEAF
AB
AB
SH
SH
AB
HEAF
SH
NONE
HEAF
HEAF
HEAF
HEAF
AB
AB
AB
AB
AB
AB
AB
AB
AB
Blower^/
Control
(type)
AB
AB
N/A
N/A
AB
N/A
AB
N/A
AB
AB
N/A
N/A
42
-------
Name and Address
Congoleum-Nairn, Inc.
Cedarhurst, Maryland
Consolidated Fiberglass
Products, Inc.
Delta Roofing Mills, Inc.
P.O. Box 546
Slldell, Louisiana
Daingerfield Manufacturing
Company
Elk Roofing Company
Stephens, Arkansas
The Flintkote Company
400 Westchester Avenue
White Plains, New York
Lloyd A. Fry Roofing Company
5818 Archer Road
Summit, Illinois 60501
GAP Corporation
140 West 51 Street
New York, New York 10020
Plant Location
Cedarhurst, Maryland
Bakersficld, California
Slidell, Louisiana
Daingerfield, Texas
Stephens, Arkansas
Chicago Heights, Illinois
Ennis, Texas
Los Angeles, California
Portland, Oregon
Atlanta, Georgia
Brookville, Indiana
Compton, California
Denver, Colorado
Detroit, Michigan
Ft. Lauderdale, Florida
Hazelwood, Missouri
Houston, Texas
Irving, Texas
Jacksonville, Florida
Kearney, New Jersey
Lubbock, Texas
Medina, Ohio
Memphis, Tennessee
Minneapolis, Minnesota
Moorehead City, North California
North Kansas City, Missouri
Oklahoma City, Oklahoma
Portland, Oregon
San Leandro, California
Summit, Illinois
Waltham, Massachusetts
Woods Cross, Utah
Jessup, Maryland
Baltimore, Maryland
Dallas, Texas
Denver, Colorado
Erie, Pennsylvania
Joliet, Illinois
Kansas City, Missouri
Millis, Massachusetts
Minneapolis, Minnesota
Mobile, Alabama
Mount Vernon, Indiana
Savannah, Georgia
South Bound Brook, New Jersey
Tampa, Florida
Saturator^
Control
(type)
HEAP
HEAP
NONE
SH
HEAP
ESP
NONE
WS
ESP
NONE
WS
WS
WS
AB
NONE
WS
NONE
NONE
NONE
WS
WS
NONE
WS
NONE
AB
AB
WS
AB
NONE
AB
HEAP
AB
Blower^/
Control
(type)
N/A
N/A
N/A
AB
N/A
N/A
AB
AB
43
-------
Name and Address
Name and Address
Johns-Manville Corporation
at Greenwood Plaza
Denver, Colorado 80217
The Logan-Long Company
6600 South Central Avenue
Chicago, Illinois 60638
Lunday-Thagard Oil Company
9301 Garfield Avenue
South Gate, California
Malarkey, Herbert Roofing
Company
3131 North Columbia Boulevard
Portland 17, Oregon
Owens-Corning Fiberglass
Corporation
Fiberglass Tower
Toledo, Ohio 43659
Royal Brand Roofing, Inc.
A Subsidiary of
Tamko Asphalt Products, Inc.
Box 385
Phillipsburg, Kansas
Southern Asphalt Roofing
Corporation
2500 E. Roosevelt Road
Little Rock, Arkansas
(presently owned by Masonite)
Tamko Asphalt Products, Inc.
601 North High Street
Joplin, Missouri 64801
Tilo Company, Inc.
Stratford, Connecticut
United States Gypsum Company
101 South Wacker Drive
Chicago, Illinois 60606
Plant Location
Fort Worth, Texas
Los Angeles, California
Manville, New Jersey
Marrero, Louisiana
Pittsburg, California
Savannah, Georgia
Waukegan, Illinois
Chicago, Illinois
Franklin, Ohio
Tuscaloosa, Alabama
South Gate, California
Portland, Oregon
Santa Clara, California
Kansas City, Missouri
Phillipsburg, Kansas
Little Rock, Arkansas
Joplin, Missouri
Stratford, Connecticut
Jersey City, New Jersey
St. Paul, Minnesota
South Gate, California
Saturatort/ Blowerk/
Control Control
(type) (type)
HEAP
HEAP
HEAF
HEAF
HEAF
HEAF
HEAF
AB
AB
AB
Cyclone
(ESP)
AB
AB
ESP
ESP
N/A
N/A
N/A
N/A
N/A
N/A
N/A
AB
N/A
AB
AB
AB
N/A
AB
N/A
b/
This list is based on information obtained from the Asphalt Roofing Manufacturers
Association as well as private communication with companies.
N/A - Not Applicable HEAF - High Energy Air Filter
AB - Afterburner SH - Smog Hog (modular ESP)
ESP - Electrostatic Precipitator WS - Wet Scrubber
Blanks indicate information is not available.
44
-------
APPENDIX B
EVALUATION OF AFTERBURNERS FOR THE
ASPHALT ROOFING INDUSTRY
45
-------
OPERATING PARAMETERS AND PERFORMANCE
Thermal Afterburners
In many industrial processes, combustible gaseous and vapor fumes and par-
ticulates have been successfully controlled by thermal afterburners. Es-
sentially, the thermal afterburner incinerates the combustible gas, vapor
or particulate fumes to form the principal combustion products of carbon
dioxide and water. Figure B-l indicates the key considerations involved
in successful incineration of these fumes. The three T'S (temperature,
turbulence, and time) are the primary considerations, as in many incin-
eration processes. In the thermal afterburner, temperature is achieved by
combustion of an auxiliary fuel to heat the incoming fume. Turbulence is
achieved by injection nozzles and flow-diverting baffles, and time is ob-
tained by providing adequate space for retention of the fumes for the de-
sired time. Detailed discussions of these afterburner designs and operat-
ing considerations are included in other reports.*>2/
Operating Conditions - Following years of practical experience, a reten-
tion time of at least 0.3 to 0.5 sec is desired in afterburner designs.A'
Temperature requirements to achieve desired levels of control (or removal)
efficiency have also been determined for gases, vapors, and odors .ill' The
temperatures and ranges versus control efficiency for (a) odors, (b) hydro-
carbons (HC) and carbon monoxide (CO), and (c) methane are presented in
Figure B-2. Based on these temperature ranges, the minimum design tempera-
tures to assure control of HC and CO in the asphalt roofing industry should
be 750°C for 90% control, 782°C for 95%, and 816°C for 99%. Note that these
recommended minimum temperatures are still far below the temperatures re-
quired for incineration of dilute fumes of methane. The additional reten-
tion time required to burn a 100-um hydrocarbon droplet has been shown to
be about 0.02 sec at 760°C.^ Since drops larger than 50 to 100 urn are
easily removed in simple cyclones or knockout vessels, no special problems
46
-------
Supplemental
Fuel
(Gas or Oil)
TEMPERATURE
TURBULENCE
TIME
Fuel Combustion
Dilute Fumes
from Asphalt
Blowing or Saturation
Mixing of Fume
and Hot
Combustion Gases
Retention of Fumes
at High Temperature
for Sufficient Time
Clean
Effluent
Figure B-l. Consideration for Successful Incineration of Dilute Fumes
-------
90
600
Recommended
Design
Temperatures
at 0.3-0.5
Seconds
Retention
Time
LEGEND:
O Odors
•f Hydrocarbons + CO
A Methane
800 1000
Afterburner Mixed Gas Peak Temperature
Figure B-2. Afterburner Operating Temperature Versus Removal Efficiency
48
-------
are expected relative to rates of pollutant destruction. Knockout vessels
are routinely used in the asphalt roofing industry in flue gas lines from
asphalt blowing, storage vessels, and saturators. Certain practical prob-
lems such as fouling of burners and buildup in duct work may be encount-
ered with asphalt droplets.
User Survey - A survey of users of thermal afterburners was conducted by
Shell.2.1 The survey results for 214 afterburners indicated that the per-
formance of 89% was satisfactory, with seven not yet run. Main operating
problems involved safety controls, refractory linings and heat exchanger
fouling or mechanical failure. However, most units were available 95 to
99% of the time so that maintenance downtime was a minor problem.
In the same survey,^.' 39% of the afterburners included heat exchangers
for energy conservation. Only 66 gave an indication of hydrocarbon reduc-
tion efficiency with 62% of these respondents reporting over 95% reduc-
tion efficiency and 94% reporting achievement of over 90% reduction ef-
ficiency.
Catalytic Afterburners
Catalytic afterburners are similar to thermal afterburners in requiring a
supplemental fuel source for heating the flue gas, but the catalytic after-
burner also includes a catalyst bed or matrix structure to accelerate the
oxidation reaction. As a result of the use of the catalyst, the catalytic
afterburner can be operated at lower temperature than a thermal afterburner
to achieve the same hydrocarbon reduction efficiency. This lower tempera-
ture will result in savings in fuel requirements. However, to achieve high
levels of hydrocarbon reduction efficiency, above 95%, either the amount
of catalyst used must be increased at a high economic cost and high pres-
sure drop or the temperature must be increased to approach the thermal af-
terburner temperature. If fuel costs continue to increase at a faster rate
than catalysts, the costs of catalytic afterburners with reduction effic-
iencies above 95% may be practicable.
Aside from the operation parameters of temperature, turbulence, and reten-
tion time, catalytic afterburners have another set of variable parameters
associated with the catalyst. These catalyst parameters include volume of
catalyst, volume/flow of flue gas, surface area of catalyst, porosity of
catalyst, oxidation reactivity, size of catalyst pellets or interstices
and type of catalyst. Some contaminants in the flue gas may also react with
the catalyst in irreversible reactions (called "poisoning"). Other contami-
nants can clog the catalyst pores. As a result, the reduction efficiency
of catalytic afterburners is expected to decrease with age but with proper
design, the catalyst bed in the catalytic afterburner can be sized for 10
to 15 years of operation without replacement.
49
-------
Operating Conditions - For most hydrocarbons and catalysts, the minimum
temperatures required for 90% conversion with solvent concentrations of
10% LEL have been reported as 249 to 349°C.^/ These are minimum tempera-
tures for a new catalyst with no aging or poisoning. Because of aging and
poisoning, the operating temperature should probably be at 370 to 427°C
to assure 907o conversion. Figure B-3 shows typical performance versus tem-
perature curves. For this example, the temperatures required are 482°C to
achieve 90% conversion of hydrocarbon and about 650°C to achieve 95% con-
version. At lower temperatures than the 750°C for 90% and 782°C for 95%
conversions required for thermal afterburners, a savings of 42 and 23%,
respectively, in heat energy and supplemental fuel requirements is possi-
ble. However, 99%, conversion might not be achieved by this catalytic af-
terburner even at the 816°C thermal afterburner temperature. The 99% con-
version for the example in Figure B-3 would require additional catalyst
volume resulting in increased cost and increased pressure drop through
the bed or matrix.
User Survey - In a 1972 survey of users of catalytic afterburners only a
small response with 24 units being reported was obtained.2/ Of these 24
units, 25% included heat exchangers. Only 30% of the units were reported
to be satisfactory. The major maintenance problem was catalyst poisoning
reported in 71%. Reduction efficiency was only reported for one unit at
90%.
AFTERBURNERS IN USE IN THE ASPHALT ROOFING INDUSTRY
A total of 76 asphalt roofing plants have been identified from information
obtained from the Asphalt Roofing Manufacturers Association and private
communication with companies.^-' Up to 82% of these plants purchase blown
asphalt from refineries and do not have asphalt blowing operations. Of the
13 plants identified which have blowing operations, all 13 use thermal af-
terburners. For the saturator operations, 28 of 76 plants or 37% use thermal
afterburners and 9 of 76 have no emission control devices at all. No cata-
lytic afterburners were reported in use. The opinion was expressed that the
asphalt droplets in these roofing plant emissions would cause rapid foul-
ing of the catalyst.
A detailed survey of 33 roofing plants showed only 13 with asphalt blowing
operations.—' The results of this survey also indicated the following in-
formation on afterburners in use in the industry:
Flow Rates
Range: 85 to 1,416 Nm-Vmin
Mean: about 283 Nm3/min
50
-------
100
80
60
c
o
C
O
U
40
Ln
20
0
H2
Reference: R. W. Rolke et al.—
2/
0
100
200
500
600
300 400
Temperature, °C
Figure B-3. Typical Temperature-Performance Curves for Various Molecular Species Being Oxidized
Over Pt/Al203 Catalysts
-------
Saturator Emissions
Average grain loading: 0.80 g/Nm^ (0.64 to 1.08)
Average temperature: ~ 68°C (60 to 74)
Average moisture content: ~ 3% (1.5 to 3.8)
Blower Emissions
Average grain loading: 8.4 g/Nm3 (0.7 to 25.6)
Average temperature: ~ 121°C (94 to 153)
Average moisture content: ~ 1670 (5.1 to 19.5)
An analysis of the operation schedules of the 13 plants weighted with re-
spect to plant production in tons per year indicated that the weighted
average annual operating time for saturators is about 5,480 hr/year and
for asphalt blowing is about 2,080 hr/year (only two samples for blowing).
Note that some plants combine blower and saturator gas streams for treat-
ment in a common afterburner which can then have an average operating time
in excess of the 5,480 hr/year.
Afterburners in use in the asphalt roofing industry are operated at temper-
atures from 538 to 1090°C.—' Data from another survey indicated that after-
burners are operated at 638°C up to 816°C._t' Test results in a Los Angeles
plant indicate a reduction efficiency of 90.47o of organics by Rule 52 and
of 88% of total organics for an afterburner operating at 250 Nm-Vmin (no
indication of afterburner operating temperature is given). In another af-
terburner test in Illinois at about 708 Nm3/min with natural gas and with
No. 6 fuel oil as supplemental fuels, the afterburner met all state require-
ments for average particulate (0.038 g/Nm^ measured for gas fuel, 0.088
g/Nm3 for oil fuel) for sulfur dioxide (7.91 ppm for gas, 14.64 ppm for
oil) and total hydrocarbons (121 to 112 ppm). Nitrogen oxide emissions were
also measured at 16.3 ppm. This afterburner was reported to be designed to
operate at 704 to 760°C, but no data on input emissions were measured to
evaluate the reduction efficiency since the existing exhaust emission stan-
dard was met.
Afterburner Fuel/Energy Requirements
Fuel/Energy Requirements - Thermal and catalytic afterburners both require
supplemental fuel to heat the flue gas to the desired combustion or oxida-
tion temperature. The amount of energy required is directly related to the
mass flow rate of the flue gas, the specific heat of the flue gas and the
temperature difference from afterburner inlet to the desired combustion
temperature. For flue gas with 16% moisture, the heat energy requirement
as a function of flue gas volume flow rate and peak afterburner mixed gas
temperature is presented in Figure B-4. For use in asphalt blowing, the
inlet temperature has been selected at 121°C. The same figure can also be
used for an approximation of the requirements for saturator flue gas if
35°C is added to the peak mixed gas temperature (to correct for an inlet
52
-------
Fuel Oil
No. 2 Gas Heat
( Liters/Hr) ( Nm3/Min) ( 109 Joules/Hr)
500 -
400 J
300 '
200
100 -
0 J
100-
80 -
60-
40 -
20
0J
200-
Inlet Gas Temperature: 121 °C
Combustion & Mixing Efficiency: 80%
Moisture Content: 16%
Volume
Flow
Rate
(Nm3/M>n)
1,416
1,133
850
800 900 1000 1100 1200
Peak Afterburner Mixed Gas Temperature °C
Figure B-4. Thermal Afterburner Heat and Fuel Requirements
Versus Temperature and Volume Flow Rate
53
-------
of 68°C). The requirements for natural gas at 35.3 x 10^ joules/Nm^ and
for No. 2 fuel oil at 38.2 x 10& joules/liter are also presented in Figure
B-4. For example, an afterburner operated at 816°C (~ 99% 1IC reduction ef-
ficiency) and 283 Nm3/min would require about 22 x 1.0^ joules/hr of energy,
which represents about 708 m-Vhr of natural gas or about 75 liters/hr of
No. 2 fuel oil. The combustion and mixing efficiency for the heated gases
is assumed to be about 8070 to make allowance for the added fuel require-
ments .
The pressure drop in an afterburner is usually around 2.5 cm of water.
Since this pressure drop is small and the fan is required anyway for as-
phalt blowing and saturator venting, the fan power requirements for the
afterburner is negligible. However, if a heat exchanger is used for en-
ergy recovery and conservation, the pressure drop may increase to 5 to 10
cm of water and some added fan power will be required, but still much less
than is required for a HEAF filter at 66 cm of
Heat /Recovery/Energy Conservation - In the last 2 years the costs of fuel
have more than doubled, and adequate supplies of natural gas and fuel oil
have become hard to find. As a result, some form of heat recovery is al-
most mandatory for afterburners. Heat energy has been recovered in three
ways in the asphalt roofing industry: (a) by use of a preheater heat ex-
changer using the exhaust hot gas to heat the incoming flue gas, (b) by
use of a heat exchanger using the hot exhaust gas to heat thermal liquids
or asphalt flux, or (c) by use of the hot exhaust gases in an auxiliary
steam boiler. Afterburner heat exchangers currently on the market can
achieve 67.5 to 80% heat recovery, or more if additional paths or modules
are used ..3' For a known heat recovery percent, the energy and fuel require-
ments presented in Figure B-4 will decrease proportionately. For example,
a 50% recovery heat exchanger reduces the energy and fuel requirements by
507».
The catalytic afterburner, by operating at a lower temperature than the
thermal afterburner will require less energy and fuel with savings of about
427o for 9070 hydrocarbon conversion efficiency and about 237o for 9570 hydro-
carbon conversion efficiency. Many catalytic afterburners also include heat
exchangers for heat recovery.
Problems with certain types of heat exchangers and with catalytic after-
burners can be anticipated in applications for the asphalt roofing indus-
try because of potential fouling of surfaces with coatings from the asphalt
droplets. This potential fouling problem might be avoided by using the hot
exhaust gases in a gas-liquid heat exchanger to heat a thermal liquid or
the asphalt flux or in a steam boiler to generate steam. A steam boiler
installation similar to that 'used at GAF in Kansas City is indicated in
Figure B-5. Here, the flue gases from the asphalt blowing, heated storage
tanks and saturators flow through two knockout vessels and two fans at 425
Nm3/min each into the thermal afterburner. The exhaust gases from the
54
-------
©-
4>
Fuel [XI—*.
Asphalt
Fume
Stream
Afterburner
Knockout
Vesse I
Blower
Water
Preheater
Notes:
A
C
B, C
Automatic Diversion to Bypass
1. When Stream Pressure/temp. Excessive and Boiler
Fuel Cut to Minimum (Using Dual-Range Controller
on Boiler Fuel).
Upon Loss of Boiler Feed Water.
Upon Loss of Boiler Flame.
(Closure of Both) Upon Shutdown of Afterburner.
Blind When Boiler Not in Service
Blinds When Afterburner Not in Service.
2.
3.
4.
Figure B-5. Afterburner with Recovery Boiler
-------
afterburner are then vented through an auxiliary steam boiler to the stack.
The boiler can also be bypassed for operations when the afterburner is not
in use. Supplemental fuel and air is available to the boiler to maintain
the required steam output rate.
Thermal Afterburner Economics
A large amount of economic information on the capital and operating costs
of thermal afterburners is available from surveys in two major environ-
9 R /
mental studies.£i2' The capital cost data from both studies has been up-
graded to 1975 estimated costs by using the ratio of construction cost
indexes as a corrective multiplier. The updated capital costs for thermal
afterburners for both studies have been combined to obtain the estimated
1975 capital costs as presented in Figure B-6. The capital costs in Figure
B-6 are presented as a function of gas volume flow rate and temperature
rise. The capital costs include an allowance of 10070 for installation
costs based on current estimates of relative costs of installation.£.'
Information on the operating costs of afterburners and estimates of the
annualized costs of capital, depreciation and operation have also been pre-
sented in the previous studies. >-*/ jn both studies, the cost of fuel is
the major operating and annualized cost for the afterburner. The cost of
fuel versus peak afterburner mixed gas temperature is presented for as-
phalt blowing and saturation in Figure B-7. The fuel costs are presented
for No. 2 fuel oil at 10c/liter and for natural gas at $1.22/109 joules.
These fuel costs are for February 1975 from the U.S. Department of Labor,—'
and are more than double the 47c/l()9 and 57£/l()9 joules for natural gas
used in previous studies. Because of this drastic increase in fuel costs,
which is expected to continue to increase, the future use of afterburners
may become economically unfeasible. Also, these increased fuel costs leave
little doubt as to the economic advisability of spending capital to add
heat recovery devices to afterburner units. In comparing the fuel costs
in Figure B-7 to the capital costs in Figure B-6 for 283 Nm3/min (or 330
actual cubic meters per minute) and 816°C peak temperature, the annual
costs are about eight times the installed capital cost for No. 2 fuel oil
and about four times the installed capital costs for natural gas.
An annual cost of the afterburner has been estimated from the capital
costs and fuel costs by including allowances for depreciation, interest
and taxes, and maintenance. The depreciation was based on straight-line
depreciation for a 15-year life and maintenance was based on a yearly cost
of $3.53/Nm3/min.^.' Interest and taxes were assumed at a current rate of
97<> of capital costs. The annual costs for an asphalt blowing and saturator
application are presented in Figure B-8. These costs are presented for typ-
ical blower and saturators operated at average temperatures and hours per
year.
56
-------
200,000
J2
_o
"o
Q
tQ
= 100,000
o
u
CL
o
U
0
Installation Charge: 100% x Equipment Cost Assumed
I
I
0 200 400 600 800 1000 1200
Gas Volume Flow Rate (Actual m3/niin)
Figure B-6. Estimate Capital Costs for Installed Afterburner Versus
Flow Rate and Temperature Rise
57
-------
600,000
R
- 400,000
c
52
_o
"o
Q
o
U
.0)
D
c
c
200,000
0
Volume Flow Rate: '283 Nm3/Min
Inlet Temperature: 121 °C Blowing, 66 °C Saturator
Costs: No.2 Fuel Oil, $0.10/Liter; Gas, $1.22/109 Joules
Weighted Annual Operating Hours: Saturation,5478 hr/yr;
Blowing, 2080 hr/yr
VBLOWING
800
1000
1200
Peak Afterburner Mixed Gas Temperature °C
Figure B-7. Estimated Annual Fuel Costs for Thermal Afterburners
Without Heat Recovery
58
-------
1.5
e
_o
~o
Q
1.0
a
o
c
<
0.5
0
i
m
/
Inlet Temperature: 121 °C Blowing, 66 °C Saturator
Peak Mixed Afterburner Temperature: 816 °C (~ 99% Control)
Weighted Annual Operating Hours: 2080 Hours Blowing, /
5478 Hours Saturation /
Fuel Costs: No.2 Fuel Oil - $0.10/Liter Feb. 1975
Natural Gas- $1 .22/109 Joules ^ Costs /
Depreciation: 15 Years Straight-Line
Interest & Taxes: 9%
Maintenance: $3.53 x Nm /Min
Fuel Use Efficiency: 80%
Assumption:
No Additional Operating
Labor Required
0 200 400 600 800 1000 1200
•3
Actual Volume Flow Rate (Actual m /min)
Figure B-8. Estimated Annual Costs for Thermal Afterburners
Without Heat Recovery
59
-------
An afterburner peak mixed gas temperature of 816°C was used since this
represents about 99% HC reduction efficiency.
If the saturator and blowing flue gases are combined, the total operating
time of the afterburner might be higher than the 5,478 hr/year for satura-
tor applications, and the cost would increase proportionately (since fuel
is the major cost).
From Figure B-8, note that the annual cost of an afterburner (or after-
burners) for 906 m3/min of gas volume flow rate can run as high as $1 mil-
lion. This cost can be reduced by use of heat exchangers or other heat re-
covery methods. The capital costs will increase but the annual costs will
decrease at a much higher rate. For example, the fuel costs will decrease
as a direct function of the heat recovery, but the annual cost of capital
will increase approximately as an inverse of the heat recovery. The result-
ing annual cost equation can be expressed as follows:
AC = Ax(l - N) + Bx(—i—) + C
1 - N
where AC = annual cost,
A = fuel cost factor,
B = capital cost factor,
C = constant cost factor, and
N = heat recovery ratio.
From the equation and the cost information currently available, the cost
factor C is low and cost factor A is much larger than B . Consequently,
at present the most heat recovery possible up to at least 80% recovery (N =
0.8) results in the least annual cost. The estimated comparative costs for
application of thermal afterburners to control saturator and air blowing
emissions are shown in Tables B-l and B-2.
Catalytic Afterburner Economics
A large amount of economic information on catalytic afterburners is also
available from the previous surveys .^JL^' The capital costs of catalytic af-
terburners are comparable but tend to be slightly higher than for thermal
afterburners because of the cost of the catalyst and because of the more
customized application requirements. The fuel costs are less for catalytic
afterburners and can result in a 43% savings in fuel requirements for 90%
60
-------
Table B-l. ESTIMATE OF THERMAL AFTERBURNER COSTS FOR SATURATORS
Afterburner capacity: 565 ra^/min at 68
Operating time: 5,480 hr/year
Application: Asphalt saturator
A. 957> Hydrocarbon removal efficiency:
Heat recovery (70)
Capital cost (incl. installation)
Annua 1 co s t!L'
Depreciation (15 years)
Interest and taxes (97o)
Maintenance
Fuel: Natural gas ($1.22/
109 joules)
or No. 2 fuel oil ($0.10/
liter; $2.63/109 joules)
Total annual costs:
- with natural gas
- with fuel oil
B. 997» Hydrocarbon removal efficiency:
Heat recovery (7»)
Capital cost (incl. installation)
Annual cost—'
Depreciation
Interest and taxes
Maintenance
Fuel: Natural gas
or No. 2 fuel oil
Total annual costs:
- with natural gas
- with fuel oil
°C inlet
(782°C
0
$ 62,400
$ 4,160
5,620
2,000
$ 11,780
$284,200
611,210
$295,980
$622,990
(816°C
0
$ 65,320
$ 4,360
5,880
2,000
$ 12,240
$297,470
639,750
$309,710
$651,990
temperature
peak mixture
50
$124,800
$ 8,320
11,230
3,440
$ 22,990
$142,100
305,600
$165,090
$3^8,590
peak mixture
50
$130,640
$ 8,710
11,760
3,440
$ 23,910
$148,750
319,900
$172,660
$343,810
temperature)
80
$312,000
$20,800
28,080
4,400
$ 53,280
$ 56,840
122,240
$110,120
$175,520
temperature)
80
$326,600
$ 21,770
29,390
4,400
$ 55,560
$ 59,490
127,940
$116,050
$184,500
a/ Assuming no additional operating labor required.
61
-------
Table B-2. ESTIMATE OF THERMAL AFTERBURNER COSTS FOR AIR BLOWING
Afterburner capacity: 565 m-Vmin at 121°C inlet temperature
Operating time: 2,080 hr/year
Application: Asphalt air blowing
A. 95% Hydrocarbon removal efficiency: (782°C peak mixture temperature)
Heat recovery (%)
Capital cost
Annual costfi/
Depreciation (15 years)
Interest and taxes (9%)
Maintenance
Fuel: Natural gas ($1.22/
10° joules
or No. 2 fuel oil ($0.10/
liter)
Total annual costs:
- with natural gas
- with fuel oil
B. 99% Hydrocarbon recovery: (816°C
Heat recovery (%)
Capital cost
Annual cost3.'
Depreciation
Interest and taxes
Maintenance
Fuel: Natural gas
or No. 2 fuel oil
Total annual costs:
- with gas
- with oil
0
$ 58,010
$ 3,870
5,220
2,000
$ 11,900
$100,290
215,680
$112,190
$227,580
peak mixture
0
$ 60,730
$ 4,050
5,470
2,000
$ 11,420
$104,970
225,750
$116,390
$237,170
50
$116,020
$ 7,740
10,440
3,440
$ 21,620
$ 50,140
107,840
$ 71,760
$129,460
temperature)
50
$121,450
$ 9,100
10,930
3,440
$ 23,470
$ 52,490
112,900
$ 75,960
$136,370
80
$290,060
$ 19,340
26,110
4,400
$ 49,850
$ 20,060
43,140
$ 69,910
$ 92,990
80
$303,640
$ 20,240
27,330
4,400
$ 51,970
$ 21,000
45,150
$ 72,970
$ 97,120
£/ Assuming no additional operating labor required.
62
-------
hydrocarbon reduction efficiency. But, note that operation of catalytic
afterburners at reduction efficiencies of 99% is doubtful based on past
performance and catalyst costs, and note that none have been applied in
the asphalt roofing industry.
CONCLUSIONS AND RECOMMENDATIONS
The information compiled on thermal and catalytic afterburner performance,
fuel requirements, and capital and operating costs has been presented in
this appendix. The available information on thermal afterburners is suf-
ficient and enough experience has been gained in actual use in the asphalt
roofing industry. Follow-up information is needed on actual test evalua-
tions in asphalt roofing plants to verify the recommended peak mixed gas
operating temperature versus total hydrocarbon reduction efficiency.
The catalytic afterburner has not had use in asphalt roofing plants and
has a poor user history. Follow-up for the catalytic afterburner may re-
quire (a) extensive tests of the possibilities of catalyst poisoning and
fouling with asphalt droplets at asphalt roofing plant operating conditions
and (b) evaluation tests to verify performance at 95 to 99% reduction ef-
ficiency with the catalyst after aging.
63
-------
REFERENCES FOR APPENDIX B
1. Danielson, J. A., Air Pollution Engineering Manual, 2nd edition, Los
Angeles County, APCD for EPA, AP-40, May 1973.
2. Rolke, R. W., R. D. Hawthorne, C. R. Cargett, E. R. Slater, T. T.
Phillips, and G. D. Towell, Afterburner Systems Study, Shell De-
velopment Company, EPA-R2-72-062, August 1972, NTIS PB 212560.
3. Herr, G. A., "Odor Destruction: A Case History," Chemical Engineering
Progress. 70_(5) :65-69, May 1974.
4. Ananth, K. P., Study to Support Standards of Performance of New Sources
in the Asphalt Roofing Manufacturing Industry, Interim Report No. 1,
Midwest Research Institute for EPA, Contract No. 68-02-1324, Task 23,
January 1975.
5. Vandegrift, A. E., and L. J. Shannon, Particulate Pollutant System
Study, Volume III - Handbook of Emission Properties, Midwest Research
Institute for EPA, Contract No. CPA 22-69-104, 1 May 1971.
6. Retail Prices and Indexes for Fuels and Utilities, U.S. Department of
Labor, Bureau of Labor Statistics, February 1975.
64
-------
APPENDIX C
EVALUATION OF THE PERFORMANCE OF HEAP (HIGH ENERGY AIR FILTRATION)
SYSTEM FOR CONTROLLING EMISSIONS FROM ASPHALT
SAT URATORS/BLOWERS
65
-------
INTRODUCTION
Control systems presently used in controlling asphalt saturator emissions
may be listed as afterburners (or fume incinerators), high energy air fil-
ters (HEAFs), electrostatic precipitators (ESPs), and wet scrubbers. An
MRI survey of 76 saturators indicates that 28 are controlled with after-
burners, 18 with HEAFs, 10 with ESPs, and 9 with wet scrubbers. The re-
maining 11 are uncontrolled. Because of increasing fuel costs, afterburn-
ers are being replaced either by HEAF units or ESPs.* ESP units usually
require more maintenance because of problems encountered in handling sticky
asphalt aerosols; therefore, HEAF units are being selected more often than
ESPs. Wet scrubbers currently used in the industry are all owned by one
roofing company and have not been utilized by other companies within the
industry. Also, from the available data on one of these wet scrubbers, this
type of control system does not appear to be a potential candidate for best
technology because of the low removal efficiency (*** 7070). Of these four
systems, only afterburners have been used for controlling emissions from
the asphalt blowing operation.
DESCRIPTION OF A HEAF DEVICE
The operation of a HEAF unit is shown in Figure C-l. Solid and liquid par-
ticulate matter to be removed from the process exhaust stream are passed
through a slowly moving blanket of glass fiber material which is disposed
of after use. The thickness and number of fibers in the pad provide a high
degree of impingement resulting in high particulate collection efficiencies,
on the order of 95 to 98%.—' After passing through the glass fiber mat, the
exhaust stream goes through a mist eliminator to capture any entrained
matter.
Glass fiber mats bonded with a phenolformaldehyde resin are normally used
but other fiber materials can also be employed. The filtering mat can vary
between 0.08 to 3.8 cm in thickness with densities between 6.4 and 128 kg/
nr.i' Usually, they are 2.54 cm thick. The fiber diameter is reported to
vary from 1 to 13 urn.—' Filtering velocities cited in Ref. 1 were gener-
ally in the range of 76 to 213 m/min with pressure drops between 15 and
65 cm of water. However, as discussed later, face velocities of 457 to 518
m/min are recommended by the manufacturer.
The filtering pad is in roll form and as the pad becomes exposed to the
gas stream and loaded with particulate, the roll advances to expose a
clean portion. The pad can either be manually advanced or automatically
advanced at a predetermined rate. Automatic advancement is usually employed
and the advancement can be triggered either by a timer or by a solenoid
* Includes the modular multistage units (e.g., "Smog Hog"™).
66
-------
Retaining
Clean Air
Mist Separator
Clean Mat
Dirty Mat
Figure C-l. Operation of a HEAP unit.
67
-------
when the pressure drop exceeds a prescribed limit. The mat uncoils from a
spool, passes over a metal perforated drum, returns to a rewind spool and
is then disposed of. The average life of the exposed portion of the mat
is about 3 hr. Under worst conditions, it can be reduced to about 30 min,
according to the manufacturer. A roll of the filter mat is usually 30 m
in length and is available in widths of 45.7 and 121.9 cm.
In order to obtain collection efficiencies of 95 or 99% with HEAF units
operating on effluent streams containing condensible material, it may be
necessary to provide means for precooling of the gas stream. Cooling is
not generally done by dilution with ambient air because this greatly in-
creases the amount of air to be handled. Discussions with personnel at
plants using HEAF units indicated that gases are precooled at times, usu-
ally with water sprays. One plant which had a HEAF unit equipped with a
water precooling system reported that the saturator exhausts were cooled
from about 67°C to about 54°C. However, it was also claimed that the plant
only used the cooling system when a visible plume was seen from the HEAF
stack. Although the cooling was claimed to be effective against visible
emissions, no data were available to quantify its effectiveness in enhanc-
ing overall collection efficiency of the system.
It appears as though potential water pollution problems and difficulty of
separating oil-water emulsions may be the reason why precooling of exhaust
gas streams to HEAF inlets is not done on a routine basis to enable col-
lection of condensible gases. The manufacturer claims that precooling will
increase operating costs due to costs required for pumping ^0. Also, cool-
ing will increase grain loadings on the mat due to condensation and this
will reduce the mat life. Therefore roofing plants tend to avoid this
operation.
The application of HEAF units in the asphalt roofing industry has its pros
and cons. Among the advantages, one can cite ease of operation, reduced
maintenance and no fuel costs. The disadvantages, however, are its inabil-
ity to control gaseous emissions and odors and the disposal problems with
the used mats. In addition, water pollution may occur if exhaust gas cool-
ing is undertaken with water sprays. Electrical power consumption may also
be significant due to high horsepower blowers used in these units. Plant
personnel at one facility also indicated that the unit was excessively
noisy and it was necessary to install a silencer beyond the demister. How-
ever, this was not observed in the HEAF installations at other roofing
plants that were surveyed.
68
-------
PERFORMANCE
Based on Theory
The performance of high energy air filters can be theoretically evaluated
from a knowledge of the basic mechanisms for particulate capture that are
operative in such systems. Numerous references in the literature indicate
that the removal of particulate matter from gas streams by filtration oc-
curs due to a combination of inertial impaction, interception, and diffus-
sion.3-7/ Additional factors such as gravitational, electrostatic, and
thermal forces, when present, can also exert their influence on particu-
late collection efficiency. For purposes of this discussion we will con-
sider the latter forces to be nonoperative in the HEAF system and further-
more, even if present, their contribution to overall collection efficiency
is usually insignificant compared to impaction and interception. Therefore,
this evaluation will be performed based on the assumption that inertial
impaction, interception, and diffusion are the only forces contributing to
particulate capture in the HEAF system. The combined target (fiber) collec-
tion efficiency will be determined for various particle sizes using indi-
vidual contributions due to each of the three mechanisms under considera-
tion. The overall fractional efficiency curve of the HEAF fiber mat will
be derived using the expression of Bradie and Dickson.^'
As stated above, particulate collection mechanisms considered in this in-
vestigation were inertial impaction, interception, and diffusion. Inertial
impaction occurs as a consequence of the relative velocity between the
particle and the fluid as the fluid streamlines separate to pass an ob-
stacle in the flow field (a fiber, or a previously deposited particle).
Interception or streamline contact with a surface in the flow occurs as
a consequence of finite particle size, and diffusion of particles to sur-
faces of obstacles is a result of Erownian motion. A more detailed discus-
sion of these mechanisms can be found elsewhere ,.£l£'
Assumptions underlying particulate capture by these mechanisms are as
follows:
1. Collecting obstacles situated in the flow are sufficiently far apart
so that the fluid flow in the vicinity of a single obstacle can be repre-
sented by the flow near an isolated obstacle; i.e., flow interference
effects from adjacent obstacles are neglected.
2. The particles approaching a surface do not interact with or distort
the flow to produce additional hydrodynamic lift or drag.
3. The particles always adhere on contact; i.e., bounce, surface migra-
tion, and reentrainment are neglected.
69
-------
The first assumption is required to define the fluid flow field approach-
ing the object. While it is a reasonable assumption for certain very open
fibrous filter geometries, in all cases of interest in operating filters,
deposits of large numbers of adjacent particles present on the substrate
may completely dominate the flow field. However, First and HindsZ/ report
that based on their experiments with HEAF mats there is only a very minor
influence of fiber interference on filtration efficiency. The second and
third assumptions are relatively reliable even in the absence of experi-
mental evidence with this system. Since emissions from asphalt saturators
consist of sticky asphalt particulates, it would seem safe to assume that
they adhere on contact with the fibers.
The inertial impaction parameter 0 can be expressed as
(// = 'p> QDP (1)
where C = empirical correction factor for resistance of gas to movement
of small particles
P = particle density in g/cc (equivalent to asphalt particles)
= 1 g/cc using a sp. gr. of 1.0,
Vo = velocity of particle
= velocity of gas,
D = particle size in microns = 0.1, 0.2, 0.3 . . . 1.6,
u = viscosity of gas in poises = 1.8 x 10"^ poises (for air), and
DJJ = diameter of strand in microns = 4 urn. (Manufacturer reported
a range of 3.683 to 4.19 urn.)
The inertial parameters were calculated for different particle sizes as a
function of velocity and the efficiency of impaction was obtained from
Ref . 5 using the experimental curve of Tj-j. versus ^ ^ . These values are
shown in Table C-l. The interception efficiency is:!/
70
-------
Table C-l. IMPACTION EFFICIENCY AS A FUNCTION OF PARTICLE
SIZE AND PARTICLE VELOCITY*/
Dp (jim)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
1.0
1.2
1.4
1.6
*1
0.0157
0.0628
0.1413
0.2512
0.3925
0.5652
0.7693
1 . 0048
1.57
2.2608
3.0772
4.0192
^2 ^3 y~$r
0.0588 0.0667 0.13
0.2352 0.2668 0.25
0.5292 0.6003 0.38
0.9408 1.0672 0.50
1.47 1.6675 0.63
2.1168 2.4012 0.75
2.8812 3.2683 0.88
3.7632 4.2688 1.0
1.25
1.50
1 . 75
2.0
^/~*h^ 7^3 \
0.24 0.26 0
0.48 0.52 0
0.73 0.77 0.02
0.97 1.03 0.08
1.21 1.29 0.17
1.45 1.55 0.27
1.70 1.81 0.37
1.94 2.07 0.50
0.65
0.78
0.88
0.95
\ \
X2 X3
0 0
0.07 0.09
0.22 0.25
0.45 0.51
0.62 0.68
0.75 0.80
0.86 0.90
0.95 0.97
-
-
-
_a/ Subscripts 1, 2, and 3 refer to velocities of 203 cm/sec, 762 cm/sec,
and 864 cm/sec, respectively. The first value is reported to be a
typical face velocity of PEDCo Environmental Systems in its report
to the EPA (EPA Report No. EPA-650/2-74-101 (October 1974)). The
second and third values are face velocities reported by the HEAP
manufacturer, who claims that they guarantee device performance only
in the range of face velocities provided by these values.
71
-------
» l + R i (2)
(1 + R)
diameter of particle
where R =
diameter of intercepting body (i.e., strand = 4 urn)
These efficiencies were calculated for different particle sizes and are
shown in Table C-2.
The diffusional efficiencies, T^, for different particle sizes were de-
rived from equations in Ref. 5, which relate diffusional efficiency with
Peclet Number, Reynolds Number, and Schmidt Number. These values were at
least two orders of magnitude smaller than TL and TL. Hence they were
assumed negligible.
The combined target efficiency has been expressed by Strauss^/ as:
%C = l ~ (1 " \^1 " V (3)
where TL = efficiency due to impaction, and
T)c = efficiency of interception.
Using the individual efficiencies shown in Tables C-l and C-2, T|JQ has been
calculated for different particle sizes and velocities. These values are
shown in Table C-3. Then, using the expression of Bradie and Dickson,^'
which was developed for a mesh separator, the efficiency can be expressed
as shown in Eq. (4). It has been assumed that this equation is also appli-
cable to the HEAP system.
E = 1 - exp(- Ina
Equation (4) uses the combined target efficiency T]jc instead of the col-
lection efficiency due to any one mechanism alone. In Eq. (4)
a = specific area of pad; surface area of fibers per unit volume
of pad (1.34 x 102cm2/cm2), and
f, = length of pad in the direction of flow (0.63 cm).
72
-------
Table C-2. EFFICIENCY OF INTERCEPTION AS A FUNCTION
OF PARTICLE SIZE
D_ (urn)
F
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
1.0
1.2
1.4
1.6
\
0.0494
0.0976
0.1448
0.1909
0.2361
0.2804
0.3239
0.3667
0.450
0.5308
0.6093
0.6857
73
-------
Table C-3. COMBINED TARGET EFFICIENCY AS A FUNCTION OF
PARTICLE SIZE AND PARTICLE VELOCITY-/
Dp (pm)
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
1.0
1.2
1.4
1.6
1)
c
0.0494
0.0976
0.1448
0.1909
0.2361
0.2804
0.3239
0.3667
0.450
0.5308
0.6093
0.6857
\
1
0
0
0.02
0.08
0.17
0.27
0.37
0.50
0.65
0.78
0.88
0.95
T) 7] T)
I I 1C
23 1
0 0 0.0494
0.07 0.09 0.0976
0.22 0.25 0.1619
0.45 0.51 0.2556
0.62 0.68 0.3659
0.75 0.80 0.4747
0.86 0.90 0.5741
0.95 0.97 0.6834
0.8075
0.8968
0.9531
0.9843
T] T]
1C ' 1C
2 3
0.0494 0.0494
0.1608 0.1788
0.3329 0.3586
0.5550 0.6035
0.7097 0.7556
0.8201 0.8561
0.9053 0.9324
0.9683 0.9810
-
-
-
a/ Subscripts 1, 2, and 3 refer to velocities of 203, 762, and 864 cm/sec,
respectively.
74
-------
The specific area of pad has been computed using a pad density of 31.78
kg/m3, a glass density of 2,400 kg/m3, and a fiber size of 4 um. These
values were obtained from the manufacturer. Also the length of pad was
reported to be 0.63 cm in the compressed state.
Using Eq. (4) the efficiency of the system was found to be 99.98% for a
0.1 um particle at a velocity of 203 cm/sec. Therefore, from basic theory,
it appears as though the device has a very high collection efficiency even
for submicron particulates. The efficiency of the device from field tests
is discussed next.
Based on Field Tests
The performance of high energy air filters has been determined by field
tests on systems controlling asphalt saturators. Table C-4 shows the re-
sults of emission tests conducted by the Los Angeles County Air Pollution
Control District (LAAPCD) on four separate asphalt saturators and related
equipment.2/ The results are based upon stack sampling and analytical pro-
cedures described in detail in the LAAPCD source testing manual.!?./ The
particulate sampling train used by LAAPCD utilized wet impingement fol-
lowed by filtration.
Emission tests results shown in Table C-4 indicate total particulate col-
lection efficiencies of 88 to 94%.2' It is also reported:!' that visible
air contaminants were emitted from three of the four filtration systems.
The opacities, ranging from 5 to 15% white, showed the adverse effect of
condensing organic vapor caused by high effluent temperatures. It is fur-
ther reported that effluent temperatures should not exceed 54°C to assure
compliance with opacity statutes in the LAAPCD area.
Another study reports average particulate collection efficiencies for a
HEAF unit ranging from 71 to 87%. by weight!!/ (see Table C-5). The uncon-
trolled and controlled samples were not measured simultaneously in this
study. Also, sampling of uncontrolled and controlled particulates was
done using different sampling trains. For sampling particulates in uncon-
trolled streams the filter followed the impingers and for controlled
streams the filter preceded the impingers.
Even though field tests suggest maximum efficiencies of only 87 to 94%,
for particulates, it is believed that the unit can achieve higher effici-
encies if the inlet gases to the HEAF unit are sufficiently cooled to con-
dense more of the vapors. Until such field tests are made, it will not be
possible to determine the unit's highest potential for particulate capture.
75
-------
Table C-4. ASPHALT SATURATOR EMISSION TESTS^
9/
Test No.
1
Flow rate, Nm3/min 391
Effluent temperature, °C 48
to atmosphere
Average filter pressure 75.9
drop (cm ^0)
Usable filter width, cm 53
Opacity, % 0
Inlet loading, g/Nm3, 1.23
dry gas
Outlet loading, g/Nm3, 0.129
dry gas
Allowable loading, Rule 52, 0.163
g/Nm-3, dry gas
Total inlet, particulates, 27.4
kg/hr
Total outlet particulates, 3.3
kg/hr
Total particulates, 88
1 1 h.
317 1,200 844
63 68 56
64.8 69.1 65.0
53 95.3 105.4
5-10 10-15 5-10
0.602 0.863 0.897
0.0494 0.0208 0.0188
0.185 0.129 0.140
11.2 14.1 12.0
0.91 1.0 0.73
92 93 94
overall collection
efficiency
Total outlet organics,
kg/hr as carbon
3.9
Gaseous organics, kg/hr
Asphalt product manufacturer Shingles
Saturant temperature, C
215-220
1.1
0.3
Felt
213
2.3
0.2
2.0
Shingles Felt
232-243 243
76
-------
Table C-5. PARTICULATE EMISSION DATA FOR ASPHALT SATURATOR CONTROLLED WITH
Felt Process rates Vent gas— Uncontrolled oarticulate
Weight—' Moist. (metric ton/hr) Flow?/ Temp. Moist.
(kg) (%) Felt Product (Nm3/min) (°C) (%) g/dNm^'
25.0 3.2 5.22 14.5 850 74 1.5 0.64
0.73
0.32
12.3 3.5 1.6 4.0 950 58 1.1 0.53
0.27
0.11
Controlled particulate
kg/metric ton
ke/hr
31
37
16
30
16
7
Felt
5.9
7.1
3.1
18.8
10.0
4.4
Product^/
2.1
2.6
1.1
7.5
4.0
. 1.8
g/dNm^'
0.156
0.183
0.153
0.048
0.037
0.050
kg/hr
8.1
9.7
7.0
2.4
1.9
2.5
kg/metric ton
Felt
1.5
1.9
1.3
1.5
1.2
1.6
Product^/
0.56
0.67
0.48
0.60
0.48
0.63
.a/ Kilograms per 44.6 m2 of felt.
b/ At inlet to control device.
£/ Normal cubic meters per minute corrected to 20° C and 760 mm Hg, dry basis.
,d/ Grams per normal cubic meters (dry).
_e/ Product is the saturated felt. Saturant used is approximately 1.6 times felt weight.
-------
ECONOMICS
Control system costs are generally estimated on the basis of the exhaust
volumes handled. In the roofing industry, exhaust volumes from asphalt
saturators can be affected by the type and efficiency of the hooding ar-
rangement, the type of saturation employed (spray-dip, dip, spray), the
asphalt characteristics, line width, line speed, felt weight, and mois-
ture content of felt. At present there are no available data to quantify
these effects on exhaust volumes. Therefore, costs are indicated for
typical exhaust volumes.
Reference 11 reports the installation, operating, and maintenance costs
for a HEAF unit operating 6,000 hr/year at 850 actual cubic meters per
minute. These are shown in Table C-6. Also included is the cost estimated
for 567 Nm-Vmin, using information provided by the manufacturer. One roof-
ing plant which has two roofing lines and two HEAF units provided detailed
cost data for the HEAF systems. These are shown in Table C-7.
The costs shown in Table C-6, as estimated by MRI, do include the cost in-
volved in cooling the inlet gases to the HEAF unit. Most of the plants do
not generally cool the inlet gases to the HEAF. Cooling will condense some
of the vapors and increase grain loading. This will favorably affect col-
lection efficiency but may also necessitate more frequent mat replacement.
78
-------
Table C-6. ECONOMICS OF HEAF SYSTEM FOR CONTROLLING ASPHALT EMISSIONS
Saturator Blowing
Ref. 11 Estimated by MRI Estimated by MRI
(850 actual m3/min) (567 Nm3/min. 5.480 hr/vear) (567 Nm3/min. 2,080 hr/vear)
Installed cost ($) 234,000^/ 100,000^ 100,000s'
Operating cost ($/year) 22,000 14,OOOliS/ 11,700^2'
Maintenance cost ($/year) 7,000 5,000i/ 9,000£/
Depreciation, interest 16.000^ 15.700^ 15,70Qe/
and taxes ($/year)
Total annualized cost 45,000 34,700 36,400
($/year)
.a/ Cost appears to be very high according to manufacturer and roofing plant personnel (see Table C-7).
b/ Based on 15-year depreciation period.
£/ Unit cost is $64,000 with additional cost of $10,000 for precooling system. Installation cost estimated
to be 35% of unit costs. These estimates are based on information from the manufacturer.
d_/ Operating cost based on 134 kw for providing 66 cm AP and power cost of $0.016/kw-hr ("* $12,000/year).
Operating cost also includes manufacturers estimated mat usage at $0.35/hr (^ $2,000/year).
je/ Fifteen year depreciation, 9% interest and taxes.
.f/ Operating cost based on 134 kw for providing 66 cm AP and power cost of $0.016/kw-hr (*» $4,400/year).
Operating cost also includes mat usage which, because of grain loading approximately 10 times higher
than in saturator emissions, was assumed to be $3.50/hr per footnote _d/ (*** $7,300).
£/ Not including any additional operating labor costs.
hf General maintenance (parts and labor) was estimated to be $7.06/Nm /min based on information from the
manufacturer and from plant personnel (^ $4,000/year). Additional maintenance labor of $1,000/year
for mat replacement on saturator was estimated from data supplied by plant personnel. (Additional
maintenance labor of $5,000/year for mat replacement on blower unit was estimated from the above,
reflecting more frequent replacement due to higher grain loading.)
-------
ft/
Table C-7. ECONOMICS OF HEAF SYSTEMS AT ONE ROOFING PLANT-
Design capacity
Installed cost
Annual operating cost (power)
Annual maintenance
Annual depreciation^-'
Annual mat cost-
Total annualized cost
Unit No. 1
1,330 Nm3/min
$88,152
$34,000^
$1,200
$5,900
$756
$41,856f/
Unit No. 2
850 Nm3/min
$66,826
$23,000£/
$1,500
$4,500
$504
$29, 5 04^
a/ Data was obtained during plant survey trip of the facility in November
1974.
b_/ Unit No. 1 is equipped with a 220 kw blower and a 22 kw booster fan;
operates 8,700 hr/year.
£/ Unit No. 2 is equipped with a 150 kw blower and a 15 kw booster fan;
operates 8,700 hr/year.
d_/ Based on 15 year depreciation period.
e_/ Total mat cost was reported as $1260. Individual costs were estimated
using the same proportion of capacities (i.e., ~3:2).
f/ Does not include interest on investment.
80
-------
REFERENCES FOR APPENDIX C
1. Goldfield, J., V. Greco, and K. Gandhi, J. Air Poll. Control Assoc.,
20(7), July 1970.
2. Filtration Eng., 1(9):9-13, May 1970.
3. Fuchs, N. A., The Mechanics of Aerosols, Pergamon Press, New York
(1964).
4. Davies, C. N., Air Filtration, Academic Press, New York (1973).
5. Strauss, W., Industrial Gas Cleaning, Pergamon Press, New York (1966).
6. Billings, C. E., and J. Wilder, Handbook of Fabric Filter Technology,
Vol. I, Prepared for the National Air Pollution Control Administra-
tion under Contract No. CPA-22-69-38, December 1970.
7. First, M. W., and W. C. Hinds, "High Efficiency Removal of Submicron
Particles by Ultra-High Velocity Aerosol Filtration," Paper No. 74-
139, 67th Annual Meeting of the Air Pollution Control Assoc.,
Denver, June 1974.
8. Bradie, J. K., and A. N. Dickson, in "Entrainment Separators for
Scrubbers - Initial Report," Prepared by S. Calvert et al., EPA
Report No. EPA-650/2-74-119a, October 1974.
9. Netzley, A. B., "Control of Asphalt Saturators by Filtration," Paper
No. 75-66.7, 68th Annual Meeting of the Air Pollution Control Assoc.,
Boston, June 1975.
10. DeVorkin, H., et al., "Source Testing Manual," Los Angeles Air Pollu-
tion Control District, Los Angeles, California (1972).
11. Gerstle, R. W., "Atmospheric Emissions from Asphalt Roofing Processes,"
EPA Document No. EPA-650/2-74-101, October 1974.
81
-------
APPENDIX D
THEORETICAL ANALYSIS OF THE APPLICABILITY OF VENTURI SCRUBBERS
FOR CONTROL OF ASPHALT SATURATOR EMISSIONS
82
-------
Venturi scrubbers have been used in a variety of applications for many years.
Their application to control of particulate emissions from asphalt satura-
tors and air blowing has possibilities, but none have been applied to these
sources. Therefore, a theoretical analysis of venturi scrubbers was carried
out in an attempt to assess the applicability of venturi scrubbers to control
emissions for an asphalt saturator.
In spite of the vast experience gained, there does not exist a reliable de-
sign equation for venturi scrubbers. However, Calvert et al.— have developed
an equation for predicting particle penetration based upon an analysis of par-
ticle and droplet motion in a venturi throat. The following assumptions are
involved in use of this equation:*
1. Particles move with the gas at all times and are collected only by liquid
drops.
2. Average liquid drop diameter is given by the emperical correlation of
Nukiyama and Tanasawa (1938).
*-£* »••(£>"
applicable for UQ < 7,500 cm/sec.,
3. The acceleration of liquid drops may be calculated using the approximation
4. The collection of particles by single drops is due to inertial impaction
only. Therefore, collection is a function of Stokes number, Kp
U p C D2
K =-JLlE -- E (3)
P 9 ^G Dd
* See attached table of nomenclature.
83
-------
5. There is a uniform concentration of particles in any plane perpendicular
to the direction of gas flow.
6. Liquid is not atomized and distributed over the cross section until the
relative velocity between the liquid and gas is:
ur - fuG
atomization
(4)
Here, f is an unknown factor.
Taking into account the continuous change in relative velocity, U as the
droplets accelerate from the point of their atomization to the gas velocity
yielded the following equation:
P.. = exp
_2
55
U
G PL Dd
F (Kpt, f)
(5)
Where F(Kpt> f) is a complicated function of Stokes number and factor f
which for Kpt in the range of 2 to 8 can be approximated by:
F(Kpt, f) = -0.156
(6)
Calvert (see Ref. 1, page 5-122, Eqs. 5.3.6-10) developed an equation to esti-
mate the pressure drop in a venturi by assuming that all energy spent in the
venturi is used to accelerate the liquid droplets to the throat velocity of
the gas.
= 1.03 x ID"3 UG2 -^
^G
(7)
By substituting Eq. (6) in Eq. (5) and using Eq. (3), we get:
= exp
55
Ur pT Dd c PD DD2l
G ^L d (0.156) C P P
9 "G Dd
(8)
84
-------
From Eqs. (7) and (8) we get:
Pt = exp
For air/water system at 20°C, and 1 atm,
liquid density, PT °" 1 gm/cc
particle density, p = 1 gm/cc (assumed).
Cunningham slip correction factor, C = 1 4- 0.162/D
UG = 183 x 10~6 poise
f = 0.25
Using the above, Eq. (9) becomes:
= exp
-1.138 x :UT2 (1
0.162
) D2 AP
(9)
(10)
Equation (10) shows that the penetration of the particles through the scrubber
(Pt = 1 - efficiency) is dependent only on the pressure drop in the venturi
and the particle size. The absence of liquid drop size in Eq. (10) indicates
that it is a relatively unimportant parameter. Since the penetration is pro-
portional to exp (-Dp2) , it is primarily a function of particle size.
Pressure Drop Calculations
R. W. Gerstle^' of PEDCo-Environmental, Inc., has reported particle size dis-
tribution of uncontrolled saturator exhaust. The data were obtained using a
Brinks impactor. The data taken off their graph are tabulated in Table D-l.
In the literature, the density of asphalt particles is given as approximately
equal to one.—' Therefore, particle diameter listed in Table D-l is equal to
aerodynamic particle diameter. The overall penetration, Pt for a control
device can be calculated from Eq. (12).
P*. =
:
-------
Table D-l. TABULATION OF PARTICLE SIZE DISTRIBUTION DATA FOR
ASPHALT SATURATOR EMISSIONS
£E
0.22
0.28
0.3
0.31
0.36
0.41
0.52
0.77
1.50
1 4'°
Extrapolation 10.0
^20.0
52i
0.25
0.29
0.305
0.335
0.385
0.465
0.645
1.135
2.75
7
15
*L
1.172 x 10~3
1.492 x 10"3
1.621 x 10"3
1.895 x 10-3
2.397 x 10'3
3.318 x 10'3
5.924 x 10'3
1.675 x 10"2
9.113 x 10"2
5.705 x 10'1
2.588
Cum 7o w ^
0
1
4
10
20
30
40
50
60
70
90
100
A Mi
0
1
3
6
10
10
10
10
10
10
20
10
A Mi
A log Dp
9.55
100.12
421.34
153.99
177.05
96.88
58.66
34.58
23.48
50.26
33.22
a/ The factor x in the table comes from equation (10) and is defined as
x = 1.138 x ID"2 (1'+ 0.162) Dp2 (11)
D,,
86
-------
Equation (12) for tabulated data (Table D-l) can be written as:
>t = 2 Pt(Dpi) A Mi (13)
Using Eq. (10) and tabulated values of AM^ , the penetration as a function
of size, and the overall penetration, is obtained. The overall collection
efficiency for different values of Ap are given in Table D-2. The Up
at Ap of 1,000 cm H20 and QL/QG of 0.001 is 31,158 cm/sec. At this ve-
locity the ICt calculated from Eq. (3) is 43.3. This shows that the Ap's
obtained in these calculations are probably too low, by a factor of 2, due
to large values of F(Kp£, f) predicted by Eq. (6).
Discussion
The pressure drops needed for the control of asphalt saturator emissions
with particle size distribution as listed in Table D-l are very high. There
are many reasons why we obtained this high estimate of pressure drops. Prin-
cipal reasons are:
1. As can be seen in Table D-l, the median size of asphalt particles is
0.77 urn, which is a fine aerosol.
2. The theory developed by Calverti' considers only the collection by in-
ertial mechanisms. For small particles (< 1 urn) collection by inertia is
small compared to collection by diffusion or electrostatic attraction. Also,
at very high throat velocities (in the present case the velocities reached
are sonic velocities) turbulent agglomeration and subsequent inertial col-
lection can take place.
3. The equations developed for calculating liquid droplet diameter (Eq.
(1)) and the equation to calculate the drag (Eq. (2)) are not valid at high
throat velocities estimated in this calculation.
Considering the above points, it is possible that the theoretical analysis
used herein may not be entirely valid. However, the results do confirm the
expectation that a very high pressure drop venturi scrubber .would be required
for efficient removal of the small particle size emissions from asphalt sat-
urators. In fact, the pressure drops required are much higher than that of
even the highest pressure drop venturies (=>• 254 cm I^O) and the associated
power requirements would be exhorbitant. This may partially explain those
statements of opinion by others that scrubbers are not a viable control
method for this source.
87
-------
Table D-2. OVERALL EFFICIENCY OF REMOVAL FOR ASPHALT SATURATOR
PARTICULATE VERSUS PRESSURE DROP
AP (cm 1^0) Overall efficiency (%)
50 52.5
100 60.7
500 86.0
1,000 95.0
1,500 98.0
1,875 99.0
88
-------
NOMENCLATURE (APPENDIX D)
C = Cunningham slip correction factor
Cp = Coefficient of drag
Dd = Liquid droplet diameter (um)
Dp = Particle diameter (um)
F = Function of 1C, and f
f = Nonuniformity and unknown factor in the atomization and subsequent
acceleration of liquid drops
0.1 to 0.3 for hydrophobic drops
0.5 for hydrophelic drops
Kp = Stokes number
.jt = Stokes number calculated at throat gas velocity
P = Penetration (1 - efficiency)
Pt = Overall penetration
Ap = Pressure drop in (cm l^O)
QG = Volume flow rate of gas (m^/sec)
QT = Volume flow rate of liquid (nrVsec)
Rg = Liquid droplet ReynoldiS number
Vg = Gas velocity at the throat (cm/sec)
U = Relative velocity of particles with respect to droplets (cm/sec)
Pp = Particle density (gm/cc)
PL = Liquid density (gm/cc)
Up = Gas viscosity (poise)
89
-------
REFERENCES FOR APPENDIX D
1. Calvert, S. et al., "Scrubber Handbook, Wet Scrubber System Study," Vol.
1, prepared for EPA/CSL (1972).
2. Gerstle, R. W., "Atmospheric Emissions from Asphalt Roofing Process,"
PEDCo-Environmental, Inc., Report to EPA, NTIS PB 238 445, page 42.
3. Traxler, R. N., Asphalt: Its Composition, Properties and Uses, Reinhold
Publishing Company, New York, page 34 (1961).
90
-------
APPENDIX E
EVALUATION OF ELECTROSTATIC PRECIPITATORS FOR
CONTROLLING EMISSIONS FROM ASPHALT SATURATORS/BLOWERS
91
-------
INTRODUCTION
Electrostatic precipitators (ESP) utilize the forces acting on an electri-
cally charged particle in an electric field to remove the particles from
a gas stream. These forces move the particles to the wall where they can
be collected. In the collection of dry particles, a layer of dust builds
up on the wall and is periodically removed by various methods. In some in-
stances, the wall is wetted by a film of liquid (e.g., water) that contin-
uously runs off, thereby removing the particles. This type of installation
requires further processing steps to remove the particulate from the liq-
uid before the liquid is discharged to the environment or recycled for
further use.
The fundamental advantage of ESP units over many other particulate control
devices is their relatively low energy consumption. In ESP units, the sep-
aration forces are applied directly to the particles instead of to the en-
tire gas stream, as in most mechanical separation methods. High collection
efficiency, low resistance to gas flow, the ability to treat large gas
quantities at high temperatures, and the ability to cope successfully with
corrosive atmospheres account for the wide acceptance and diverse applica-
tions of the electrostatic precipitation process.
THEORY OF ESP OPERATION
Electrostatic precipitators are conventionally classified as single or two
stage units.i' In a single stage unit, the particle charging and collecting
sections coincide as shown schematically in Figure El-A. In a two stage
unit, shown schematically in Figure El-B, the charging and collecting sec-
tions are separate. ESP units can be further classified according to their
geometry, the two most common classifications being flat plate and cylindri-
cal plate configurations. Regardless of classification, the collection ef-
ficiency for design purposes is described by the Deutsch equation (Eq. (6.22)
of Ref. 1) as:
T] = 1 - e- (1)
where 7] = collection efficiency
w = particle migration velocity (m/sec)
A = collecting area of precipitator (m^)
V = gas volumetric flow rate through the precipitator (m-Vsec)
This relation shows that the collection efficiency of a particular device
increases with:
92
-------
COLLECTING
1 PLATES
• •«•••• ••*—CORONA WIRES
A. SINGLE STAGE
CHARGING
SECTION COLLECTION SECTION
COLLECTING
PLATES
HIGH VOLTAGE
ELECTRODE
B. TWO STAGE
Figure E-l. General Classification of Electrostatic Precipitators
93
-------
* increasing specific collection surface (A/V)
* increasing place collection area (A)
* decreasing gas volumetric flow rate (V)
* increasing particle migration velocity (w)
The effect of particle size on collection efficiency is incorporated in
the particle migration velocity. In designing and sizing a unit for a spe-
cific application, the particle migration velocity is determined experi-
mentally through measurements of collection efficiency and the specific
collection surface. When estimated in this fashion, the migration velocity
reflects a number of effects that are not easily modeled in an analytical
theory (e.g., particle reentrainment effects, nonuniform electric fields,
particle shape effects, etc.). To gain insight into the primary factors
governing the migration velocity, theoretical analysis has been employed
and shows that migration velocity is given by (Eq. (6.3), Ref. 1): (see
note)
q En x 107
v - 2-J> (1 + AI 1) (2)
6rr ua a
where w = migration velocity (cm/sec)
q = charge on the particle (coulombs)
Ep = collection section field strength (volts/cm)
u = gas viscosity (poise)
a = particle radius (cm)
AI = Cunningham correction constant
= 0.86 for air at normal temperature and pressure
X = molecular mean free path (cm)
One notes from the above equation that the migration velocity increases
with:
* increasing charge on the particle (q)
* increasing collection section electric field (Ep)
* decreasing gas viscosity (u)
Equation (2), taken by itself, indicates that the migration velocity in-
creases with decreasing particle size. However, the charge acquired by a
particle in the charging section of ESP units depends on the particle size.
Therefore, a brief discussion of particle charging is in order.
Note: Equations taken from Ref. 1 have been converted from esu units to
more practical electrical units.
94
-------
As pointed out in Ref. 1, particles may be charged in numerous ways by
taking advantage of the varied electric activity associated with many phy-
sical and chemical phenomena. Theory and long experience have shown that
the unipolar high-voltage corona discharge is by far the best and most
universally applicable means for achieving very high particle charges for
gas cleaning purposes. Two distinct particle charging mechanisms are active
in the corona field of a precipitator, the most important of which is charg-
ing by ion attachment in the electric field. This is a process that is gen-
erally designated as field or impact charging. A secondary charging process
occurs due to the phenomenon of ion diffusion. In practice, the field charg-
ing process predominates for particles larger than about 0.5 urn diameter,
the diffusion process for particles smaller than about 0.2 urn, while both
are important for particles in the intermediate range of 0.2 to 0.5 urn.
Theoretical relations for the rate of charging by these two mechanisms have
been developed (e.g., Eq. (5.27) for diffusion charging and Eqs. (5.15) and
(5.17) for field charging, Ref. 1). One of the most important differences
between the two mechanisms is that the charge increases continuously with
time of exposure in diffusion charging but reaches a constant value rela-
tively quickly for field charging. This apparent advantage of diffusion
charging cannot be realized in practice, however, because of the large ex-
posure time required. To illustrate this point, calculations were made for
a 0.1 urn diameter particle, assuming an ion density of 5 x 10^ ions/cm^, a
gas temperature of 60°C, a dielectric constant of 2 (typical for oil), and
a charging field strength of 10^ v/cm. The results of these calculations
are presented in Table E-l. One notes that the field charging mechanism
increases the charge rapidly at small values of time but the charge reaches
a constant value after 0.1 sec. The diffusion charging mechanism, however,
increases the charge continuously, eventually surpassing the field charge
(at •* 0.1 sec). However, the full potential of diffusion charging (e.g.,
4.68 times the field charge at t = 1,000 sec) cannot be realized in practi-
cal applications because of the excessive treatment time required.
In order to simplify the remaining discussion, only the field charging
mechanism will be utilized in the remainder of this section. The resulting
expressions should be reasonably accurate down to 0.5 um particles. The
theoretical saturation charge acquired by a particle due to field charg-
ing is given by (Eq. (5.17), Ref. 1):
q = (1 + 2 fcli)JL£ x 10-H (3)
k+2 9
where q = saturation charge on a particle (coulombs)
k = dielectric constant: of the particle
Eo = charging field strength (volts/cm)
95
-------
Table E-l. NUMBER OF ELEMENTARY CHARGES ACQUIRED BY A 0.1 urn PARTICLE
AS A FUNCTION OF TIME BY DIFFUSION AND FIELD CHARGING
Time (sec)
0.001
0.01
0.1
1.0
10.0
100.0
1,000.0
Number of charges
Diffusion charging
0.18
1.09
3.03
5.29
7.58
9.88
12.18
Field charging
0.87
2.17
2.55
2.59
2.60
2.60
2.60
Ratio of diffusion
to field charge
0.21
0.50
1.19
2.04
2.92
3.80
4.68
Combining relations (2) and (3) gives:
E0E a
(4)
When presented in this form, one observes that the migration velocity (as-
suming field charging) and collection efficiency increase with:
* increasing particle size (a)
* increasing charging and collection fields (Eo,Ep)
* decreasing gas viscosity (u)
THEORETICAL CALCULATIONS FOR SATURATOR OPERATIONS
The dielectric constant for petroleum oil is approximately two, which, when
substituted into relation (4) gives (at normal temperatures and pressures):
-, \
w = 8.8425 x 10'7 (1 + 0.86 A
a
(5)
Air, at 60°C, has a viscosity of 2.03 x 10"4 poise and a mean free path of
approximately 0.1 urn. Using these values, introducing a = d/2, and convert-
ing diameter (d) to micrometer units provides:
w = 2.178 x ID"7 (1 + 0<172 ) EoEDd
d ^
(6)
96
-------
This relation is applicable for particles above 0.5 p. diameter. As dis-
cussed previously, for smaller particles, diffusion charging becomes impor-
tant and must be taken into account. A hypothetical example of a modular
ESP collection plate configuration is shown in Figure E-2. Assuming a plate
voltage of 6,500 v, the plate field is 6,500 v/cm. The collection area of
one side of a grounded plate is 0.26 m2 and the total collection area is
12.9 m2. Assuming an air flow of 1.83 m/sec, with the inflow area of 50 x
50 = 2,500 cm2 (= 0.25 m2), the volumetric flow rate is 27.5 m3/min. Thus,
the specific collection surface is:
A/v = ——•-—• = 0.469 min/m
27.5 mj/min
Assuming a charging section field strength of 10,000 v/cm together with the
plate field of 6,500 v/cm, relation (6) becomes:
w (cm/sec) = 14.81 (1 + tjd (7)
d
Substituting the specific collection surface and relation (7) into (1)
gives :
where d is in units of micrometers
According to Figure E-3 (taken from Figure 4.2, Ref. 2), 507» of the parti-
culate mass from saturator operations is contained in particles less than
about 0.8 urn in diameter. Similar particle size data have not yet been found
for the emissions from the blowing operation. Using efficiencies calculated
with Eq. (8), along with the overall size distribution from Figure E-3, Table
E-2 has been prepared. The overall mass collection efficiency is seen to be
94.747o on a single module based on the theoretical calculations. If a second
collection unit were to follow the first module, further increases in effi-
ciency would be realized. Such a configuration would be equivalent to dou-
bling the collection plate area (and the specific collection surface) lead-
ing to:
^^ e-8.24 (l + (9)
The results of calculations with Eq . (9) are shown in Columns 6 and 7 of
Table E-2 with a resulting overall efficiency of 99.27% for tandem units.
97
-------
HIGH VOLTAGE ELECTRODE
(6500V)
1cm
50 cm
Figure E-2. Hypothetical ESP Collection Plate Configuration
98
-------
VO
5.0
4.0
3.0
2.0
g
. 1.0
LU
N
"» 0.8
LU
fe 0.6
Q_
0.4
0.2
J L
_L
J \ L
_L
_L
_L
0.01 0.1 1 2 5 10 20 30 40 50 60 70 80 90 95
PERCENT BY WEIGHT LESS THAN GIVEN SIZE
Figure E-3. Particle Size Distribution in Uncontrolled Saturator Exhaust
97
-------
Table E-2. THEORETICAL OVERALL AND FRACTIONAL EFFICIENCIES FOR SATURATOR OPERATIONS
o
o
(1)
Particle size
range Cp-m)
0-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.06-0.8
0.8-1.0
1.0-2.0
2.0-ofl
Total
(2)
Mean particle
size (H-m)
0.15
0.35
0.45
0.55
0.70
0.90
1.50
—
(3)
Fractional
efficiency
(Ea. 8)
0.7346
0.8835
0.9229
0.9489
0.9725
0.9879
0.9990
1.0
Single module
(4)
Mass fraction
(Figure E-3)
0.05
0.20
0.14
0.06
0.05
0.05
0.10
0.35
1.00
(5)
Mass fraction
removed
(3) x C4)
0.0367
0.1767
0.1292
0.0569
0.0486
0.0494
0.0999
0.35
0.9474
Tandem
(6)
Fractional
efficiency
CEa. 9)
0.9296
0.9864
0.9941
0.9974
0.9992
0.9999
1.0
1.0
modules
(7)
Mass fraction
removed
(4) x (6)
0.0465
0.1973
0.1392
0.0598
0.0499
0.0500
0.10
0.35
0.9927
-------
Based on theoretical calculations, it appears that modular type ESP units
can effectively remove up to 99% of the particulate asphalt emissions from
saturator operations. However, it must be noted that this can be achieved
only for those pollutants that are in particulate form. As pointed out in
this report, uncontrolled emissions from saturators probably contain a sig-
nificant amount of gaseous hydrocarbons. Under these conditions, any col-
lection device that is ineffective in collecting gaseous components (e.g.,
ESPs, HEAF) can achieve a maximum total hydrocarbon efficiency only equal
to the percentage of particulate hydrocarbons. Thus, preconditioning or cool-
ing of the gas stream is required for these devices to reduce the gaseous
hydrocarbon component to an assumed low level in order to achieve 95 to 99%
removal of total hydrocarbons. The gaseous component in asphalt blowing op-
erations is believed to be at least: as much as that found in saturator op-
erations and, therefore, would also necessitate preconditioning or precool-
ing to maximize condensation of gaseous hydrocarbons.
USE OF ESP UNITS IN THE ASPHALT ROOFING INDUSTRY
According to an MRI survey, 10 of 76 saturator installations control their
effluent emissions with ESP units. In an attempt to further define the ap-
plicability of ESP units in the asphalt roofing industry, a number of indi-
viduals and manufacturers have been contacted. One of our first attempts was
to find a more detailed and complete model of ESP operation characteristics
than the simplified approach presented in the previous sections of this re-
port. It was found that Southern Research Institute has developed a comput-
erized model for wire-plate precipitator geometries. As of this writing, the
final report detailing the development of this model has not been completed
for general distribution. Conversations with individuals responsible for the
development of the model indicated that it was best suited for evaluating a
given proposed configuration with known geometry, voltage-current relation-
ship, particle size distribution, and gas flow rate. As such, this model
may prove useful in future design efforts but is not directly useful for
the more general investigation being undertaken in the present study.
Four firms engaged in the manufacturing of electrostatic precipitators were
contacted in an attempt to obtain field test results for typical operations
encountered in the asphalt roofing industry. Only one of these manufactures
a unit that has been utilized to any degree in the asphalt roofing industry.
The various ESP units manufactured by these companies, together with rele-
vant field test data whenever they existed, are discussed in the following
paragraphs.
United Air Specialists, Inc., in Cincinnati, Ohio, manufactures a modular,
two stage unit designated the "Smog-Hog"™. A schematic diagram of this unit
is shown in Figure E-4. Several of these units are presently installed to
control emissions from the saturator process. According to one of United"s
101
-------
MECHANICAL
PRE-FILTER
COLLECTING
CELL
BLOWER
o
ro
DIRTY
AIR
ENTERS
HERE
CLEAN
AIR
EXHAUSTS
HERE
HIGH EFFICIENCY
IONIZER
AFTER
FILTER
ADJUSTABLE
EXHAUST
DIFFUSER
Figure E-4. Schematic of "Smog-Hog
!|TM
-------
representatives, they presently have orders for installation of their sys-
tem in 30 asphalt saturator plants. About half of these were to be installed
by the end of July 1975. This representative said that their units would
not work well in asphalt blowing operations because of the high grain load-
ing (average of 8.4 g/Nm^) but that they worked well on saturators (average
grain loading of 0.80). He believed that a 2.3 g/Nm^ loading was about the
upper limit for acceptable saturator emission control with their units.
In typical installations, they use tandem units to achieve emission con-
trol of particulate hydrocarbons. The units have a 12 KV ionizing section
and 6 KV collection plates. For proper operation, inlet air must be at or
below 60°Co Depending on plant layout, length and insulation of ducts, pre-
cooling is sometimes necessary to achieve this temperature level. A premist
or high pressure spray is normally included to reduce maintenance frequency.
A typical installation including a plenum chamber, prefilters, and tandem
pass units have a capital cost of $88.25/Nm3/min. Operation and maintenance
costs average $0.18/hr/100 Nm3/min capacity. If the loading is especially
heavy caused by fast run, low boiling point asphalt, high moisture, etc.,
then a triple pass "Smog-Hog"™ is necessary. The capital cost of this sys-
tem would be approximately $123.55/Nm3/min. If the temperature exceeds 60
to 66°C at the inlet, then a precooling or condensing device must be uti-
lized. The cost of these units vary, depending on the type of precooler used.
Performance data for two plants are presented in Tables E-3 and E-4. Table
E-3 contains data collected at the Celotex Corporation in Lockland, Ohio,
on two different occasions. Average filterable particulate efficiency for
the two tests is 91.3%. Average total efficiency is 74.37<>. The efficiency
for condensible matter averages 54.2%.
The units installed in this plant: were single-pass "Smog-Hogs"™ which are
less efficient than the usual tandem pass units. Also, the inlet and outlet
sampling were not performed simultaneously, or a considerable loss in ex-
haust gas occurred through the unit, as is evidenced by the large differ-
ences in inlet and outlet flow rates. Also, the average inlet loading is
only 0.066 g/Nm-^ a very low value compared with the previous average of
0.80 g/Nm3.
Tests conducted on a tandem unit installation at the Celotex Corporation
in Fairfield, Alabama, are summarized in Table E-4. The average efficiency
for filterable particulates, condensible matter and total emissions is 98.7,
86.4, and 95.47», respectively. It is interesting to note the higher conden-
sible matter efficiencies in Table E-4 as compared with Table E-3. This is
probably due to the lower gas temperatures for the tests in Table E-4 which
would lead to a higher percentage of condensation and subsequent capture in
the ESP unit. Also, the tests of Table E-3 were performed with simultaneous
103
-------
Table E-3. TESTS AT CELOTEX CORPORATION, LOCKLAND, OHIO
Filterable
particulate
Date
\974
4/9
4/9
4/16
4/16
Unit
No. <
1
1
1
1
Flow
552
428
649
416
Temp.
/ (°c)
61
49
52
49
Sample ,
site p/Mm-*"
Inlet 0.043
Outlet 0.0027
Inlet 0.027
Outlet 0.0054
kg/hr
1.43
0.073
1.10
0.14
Eff.
(7.)
-
94.9
-
87.6
Condensible matter^
g/Nm3
0.034
0.02
0.025
0.018
kg/hr
1.13
0.56
1.02
0.43
Eff.
(%)
-
50.6
-
57.8
g/Nm-'
0.078
0.025
0.053
0.02
Total
kg/hr
2.56
0.63
2.12
0.57
Eff.
(%)
-
75.3
-
73.2
,_, a./ Normal cubic meter per minute, corrected to 20 C, 760 mm Hg--dry basis.
4> b/ Grams per normal cubic meter
£/ Included particulate condensible down to the 10 to 20 C range.
Note: Tests are for single-pass MSmog-Hog"'M units with a fiber roll prefilter.
-------
Table E-4. TESTS AT CELOTEX CORPORATION, FAIRFIELD, ALABAMA (MARCH 4 AND 5, 1975)
Filterable
particulate
Test
No. I
1
1
2
2
3
3
Flow
fNm3/min>2'
587
602
567
585
569
593
Temp.
30
29
34
32
34
33
Sample
site
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
g/Nm3"
0.117
0.00103
0.117
0,00110'
0.011
0.00208
kg/hr
4.13
0.037
4.00
0.03"9
3.75
0.074
Condensible matter^
Eff.
(%) g/Nm3
0.0103
99.1 0.00140
0.071
99.0 0.00906
0.0600
98.0 0.0078
kg/hr
0.36
0.050
2.40
0.322
2.04
0.277
Eff.
(%) g/Nm3
0.128
86.1 0.
0.
86.6 0.
0.
86.4 0.
0025
188
0103
169
0098
Total
kg/hr
4.49
0.087
6.40
0.361
5.79
0.351
Eff.
a)
98.0
-
94.4
-
93.9
ja/ Normal cubic meters per minute, corrected to 20 C, 760 mm Hg—dry basis.
b/ Grams per normal cubic meter.
.c/ Included particulate condensible down to the 10 to 20 C range.
Note: Tests are for tandem "Smog-Hog"™ units with a fiber roll prefilter.
-------
sampling of the inlet and outlet of the control units, providing a higher
confidence in the tests results. The average inlet loading is 0.162 g/Nm ,
again considerably below the previous average.
It is interesting to note the percent condensibles at the inlet in Tables
E-3 and E-4. Table E-3 indicates an average inlet condition with 46.2%
condensibles compared with an average of 26.9% in Table E-4. It is possi-
ble that the filter in the sampling train became fully loaded with subse-
quent penetration of liquid asphalt through the filter that was subsequently
captured in the impingers and reported as "condensible." Further studies to
establish the percent of gaseous hydrocarbons appears needed.
A representative of the American Air Filter Company was contacted to obtain
information relative to their ESP units. They presently manufacture a two
stage ESP unit that they call the Environmental Control Unit (ECU) which
apparently is similar to the "Smog-Hog"™. The ionizer wires are maintained
at 12,400 volts DC and the collection plates are energized to 6,500 volts.
The ECUs are equipped with an internal cold water washing system that can
also dispense various detergents. These units are intended for the control
of visible concentrations of mists and fumes, and the representative be-
lieved they would work satisfactorily on asphalt saturator emissions. How-
ever, the representative's experience with asphalt blowing operations led
him to believe the units would not function effectively in this application
because of the fouling and cleaning problems produced by the tar-like buildup
on the collection plates, insulators, etc. They have not been able to find a
detergent suitable for removing these tar deposits. Equipment capital costs
amount to approximately $31.77/Nm3/min (which seems rather low). Additional
costs for ducting is dependent on plant layout, etc., and are quite variable.
If additional gas conditioners for precooling are needed, they can run $17.65
to $70.60/Nm3/min.
A representative of Fluid-Ionic Systems was contacted relative to their
"Hydro-Precipitrol"™ ESP units. The units are cylindrical and composed of
several concentric wetted wall collection plates and expanded metal ioniz-
ing cages, and would be classified as single stage precipitators. Their
largest unit, 3.8 m in diameter, can handle up to 2,550 Nm^/min of effluent
gases in some applications. The units operate in a voltage range of 38 to
42,000 volts. Although they do not presently have any of their units in-
stalled at asphalt saturator or blowing plants, they do have some test data
collected several years ago. The data, summarized in Table E-5, were ob-
tained on a saturator operation with a small, portable "Hydro-Precipitrol"™
unit comprised of two cylindrical collection plates (76.2 and 50.8 cm diam-
eter) and one expanded metal ionizing cage (63.5 cm diameter). The unit had
a plate length of 1.5 m. Using these dimensions, it is estimated that the
106
-------
Table E-5. TEST DATA OBTAINED FOR A SMALL "HYDRO-PRECIPITROL"™ UNIT FOR A SATURATOR APPLICATION
Filterable
particulate
Test Flow
No._ (Nm-Vmin)
1 37.1
1 48.
2 53.
2 64.
3 55.
3 68.
4 79.
4 102.
8
0
2
0
7
8
5
Temp.
38
26
37
26
33
26
32
26
Sample
site
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
Inlet
Outlet
g/Nm^
0.0460
0.0021
0.0474
0.0032
0.0362
0.0032
0.0339
0.0041
Eff.
kg/hr (%)
0.1025
0.0060 94.1
0.1508
0.0124 91.8
0.1194 -
0.0132 88.9
0.1625
0.0254 84.4
Gondensible matter
g/Nm3
0.0032
0.0009
0.0050
0
0.0069
0.0016
0.0183
0.0007
kg.
0.
0.
0.
0
0.
0.
0.
0.
Eff.
/hr (7o)
0071
0027 62.4
0160
100.0
0227
0066 70.7
0878
0042 95.2
g/_Nm3
0.0492
0.0030
0.0524
0.0032
0.0430
0.0048
0.0522
0.0048
Total
kg/hr
0.1096
0.0087
0.1668
0.0124
0.1421
0.0198
0.2503
0.0296
Eff.
JC2LL '
92.0
-
92.6
-
86.0
-
88.2
-------
unit had a flow area of 0.25 m^ and a collection plate area of 6.1 m^. Re-
ferring to Table E-5, one finds that based on the four tests, the average
filterable, condensible, and total removal efficiencies are 89.8, 82.1, and
89.7%, respectively. Referring to the flow column indicates that there ap-
parently was considerable secondary air leakage into the unit during the
tests. Also, the air temperatures are relatively low, probably due to a
preconditioner incorporated into the ESP unit.
On the average, condensible matter amounted to 16.8% of the hydrocarbon at
the inlet. Once again, the average inlet loading was low, averaging 0.50
g/Nm3 for the four tests. The company would not provide any cost data for
their unit since each potential application has unique features and operat-
ing requirements. They believed a general cost figure would be misleading
and not appropriate without specific tests on a process using one of their
small, portable units.
A representative of the MikroPul Division at United States Filter Corpora-
tion, Summit, New Jersey, was also contacted relative to their "Elektrofil"™
wet electrostatic precipitator. These units are large, the minimum standard
size presently manufactured being a 1,416 Nm3/min. Thus, it does not appear
feasible to utilize the "Elektrofil"™ standard units in asphalt roofing in-
dustry. However, smaller units could undoubtedly become standard if a suf-
ficient market existed. In fact, the company is presently looking into the
possibility of manufacturing modular units. Specially designed smaller units
can be obtained, but at an increased cost per cubic feet per minute capacity.
The unit operates with a continuous flow of fine liquid droplets into the
ionizing section. These droplets become charged and migrate to the collec-
tion plates, forming a continuous film of liquid. The company has had no
applications in the asphalt roofing industry but has had applications of
their units for the treatment of oil and tar emissions.
According to the MikroPul representative, a 1,416 Nm3/min, three stage unit,
constructed with carbon steel, has a capital cost of $229.45/Nm3/min. In-
stallation costs typically add 30% to this cost and operating costs are
0.053 to 0.071 kw/Nm3/min capacity. Water usage amounts to 0.80 to 0.93
liters/Nm3.
When any particulate collection device (e.g., ESP or HEAP) is used to con-
trol the effluent from saturator operations, proper gas preconditioning is
very important for assuring maximum condensation of asphalt vapors. The im-
portance of preconditioning was emphasized by the plant superintendent of
the Bird and Son facilities in Portland, Oregon, who was contacted to ob-
tain information on the operation of their ESP unit. The saturator emissions
from their operation are controlled by a Type B American Air Filter Rotoclone
108
-------
wet centrifugal dust collector and an American Air Filter Type S electro-
cell ESP unit (this ESP unit is no longer being manufactured). The unit
has provided good emission control in this plant but no quantitative data
are available. The superintendent believed that the long duct (about 55 m
long, 99 cm diameter), containing spray nozzles over its entire length, was
very beneficial in achieving good emission control. A similar emission con-
trol system in a competitor's plant has not worked effectively and the only
major difference is the duct length and associated spray nozzle system.
The use of wet electrostatic precipitators on asphalt air-blowing or satura-
tor operations might present a potential water pollution problem if the
hydrocarbon/water emulsion were difficult to "break." It has been assumed
in these discussions that the effluent emulsion could be separated probably
with the aid of emulsion breakers, so that the water could be recycled to
the ESP and the separated asphalt could be returned to the process. However,
if the emulsion were very difficult to break, this might present a potential
water pollution problem that could detract from, or offset, possible advan-
tages of the wet electrostatic precipitators.
ECONOMICS AND COST ESTIMATES
All of the following cost estimates are based on information supplied by
the four manufacturers contacted during the study. Dry precipitator esti-
mates are based on data for the "Smog-Hog"™ and Environmental Control Unit.
Wet precipitator estimates are based on the data supplied for the "Elektro-
fil"™ unit, since no cost data could be obtained for the "Hydro-Precipitrol."™
Saturator costs are based on a gas flow rate of 567 m3/min and 5,480 hr/year
operating time. Estimated efficiencies are based on reported test data. How-
ever, these test data are for relatively low inlet concentrations (0.043 to
0.188 g/NnH) and extrapolation of these efficiencies to higher loadings may
be overly optimistic.
Cost estimates for tandem and triple pass "Smog-Hog"TM units are presented in
Table E-6. Based on the available experimental data, it appears that a tandem
unit can achieve a 95% efficiency at a total annual cost of $18,800. A triple
pass unit should achieve about 99% removal efficiency at an annual cost of
$23,700.
A cost estimate for the Environmental Control Unit is presented in Table
E-7. The company's representative had no test data for this unit controlling
asphalt saturator emissions. However, he believed it would function at about
the same efficiency as their older Model S ESP units (no longer manufactured)
and quoted an efficiency value of 90 to 957o. The estimated total annual cost
for the unit is at least $15,000.
109
-------
Table E-6. COST ESTIMATE FOR "SMOG-HOG"™ UNITS CONTROLLING ASPHALT
SATURATOR EMISSIONS (567 Nm3/min, 5,480 hr/year)
Double (tandem) units Triple pass units
(~ 95% efficiency) (~ 9970 efficiency)
a/
Capital cost" $50,000 $ 70,000
Detergent wash system (~ 10% of 5,000 7,000
capital cost)—
Preconditioned— 10,000 10,000
Subtotal $65,000 $ 87,000
a/
Installation cost (~ 30%)- $20,000 $ 26.000
Total installed cost $85,000 $113,000
Annual costs
Depreciation (15 years) $ 5,700 $ 7,500
Interest and taxes (9%) 7,600 10,200
a/
Operating and maintenance costs" 5,500 6,000
Total annual cost $18,800 $23,700
_a/ Costs estimated on basis of information from manufacturer:
Capital costs - $88.25/Nm3/min for double unit, $123.55/Nm3/min
for triple pass unit plus 10% for detergent wash system.
Operating and maintenance cost - $0.18/hr/100 Nm-Vmin, excluding
any additional operating labor costs.
Installation cost - 30% of total capital cost.
_b/ Information obtained from equipment suppliers and plant operations
indicated that preconditioner system costs may range from $17.65
to $70.50/Nm-Vmin and one supplier's estimate for a 567 NnrVmin
unit was $10,000.
110
-------
Table E-7. COST ESTIMATES FOR ENVIRONMENTAL CONTROL UNIT
CONTROLLING ASPHALT SATURATOR EMISSIONS
(567 NnrVmin, 5,480 hr/year)
Estimated efficiency; 90 to 95%
a/
Capital cost (estimated)- $36,000
Detergent wash system (included - 0 -
in above)
Preconditioned/ 10,000
Subtotal $46,000
c/
Installation cost (% 30%)- $14,000
Total installed cost $60,000
Annual costs
Depreciation (15 years) $ 4,000
Interest and taxes (9%) 5,400
c/
Operating and maintenance costs— 5,500
Total annual cost $14,900
_a/ Capital cost for a single unit was stated by manufacturer
to be $31.77/NnvVmin. It has been assumed that a double
unit would be necessary for 95% efficiency at a cost of
$63.54/Nm3/min.
b/ See footnote _b/, Table E-6.
c_l See footnote at/, Table E-6.
Ill
-------
Presented in Table E-8 is a cost estimate for the "Elektrofil"™ unit (wet
ESP). The total annual cost for this unit is $32,000, not including second-
ary water treatment. This cost is larger than for the other two units, pri-
marily due to the larger capital costs. If the company starts to produce a
line of modular, smaller sized units, the capital costs may drop below the
$229.45/Nm-Vmin value used for the cost estimate. Although specific experi-
mental data do not exist, it appears that the wetted wall ESP type units
may offer the possibility of controlling asphalt blowing operations. The
coating or fouling problems encountered with dry precipitators may be elim-
inated by the water film employed in the wet units. Costs for wet electro-
static precipitators on air blowing could not be estimated but it was as-
sumed that they may be approximately the same as for saturators.
112
-------
Table E-8. COST ESTIMATE FOR "ELEKTROFIL"™ UNITS CONTROLLING
ASPHALT SATURATOR EMISSIONS (567 Nm3/min, 5,480 hr/year)
Estimated efficiency: 95 to 99%
Capital cost ($229 .45/Nm3/min)£/
Detergent wash system
Preconditioner
Subtotal
Installation cost
Total installed cost
$130,000
Not needed
Included above".'
$130,000
$ 39.000
$169,000
Annual costs
Depreciation (15 years)
Interest and taxes (9%)
c/
Operating and maintenance costs-
Secondary water treatment
Total annual cost
$ 11,300
$ 15,200
$ 5,500
Unknown
$ 32,000
£/ Estimate based on information provided by the manufacturer.
b/ Water spray section is provided as part of the installation
so it was assumed that a separate preconditioner section
would not be necessary.
£/ Cost estimated based on information from control equipment
manufacturers (see footnote £/ in Table E-6).
Note: This unit may also be applicable for control of asphalt-
blowing emissions.
113
-------
REFERENCES FOR APPENDIX E
1. White, H. J., "Industrial Electrostatic Precipitation," Addison-Wesley
Publishing Company (1963).
2. Gerstle, R. W., "Atmospheric Emissions from Asphalt Roofing Process,"
EPA-650/2-74-101, October 1974.
3. "Identification of Control Technology for Asphalt Roofing Industry,"
Monthly Progress Report No. 1, 2 June 1975.
114
-------
TECHNICAL REPORT DATA
(Please read Instructions on t/ie reverse before completing)
1. REPORT NO.
EPA-600/2-76-120
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Control Technology for Asphalt Roofing Industry
5. REPORT DATE
April 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Paul G. Gorman
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Midwest Research Institute
425 Volker Boulevard
Kansas City, Missouri 64110
10. PROGRAM ELEMENT NO.
1AB015; ROAP 21AFA-106
11. CONTRACT/GRANT NO.
68-02-1324, Task 35
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final: 4/75-2/76
14. SPONSORING AGENCY CODE
EPA-ORD
is.SUPPLEMENTARY NOTES Task officer for this report is E. J. Wooldridge, Mail Drop 62,
Ext 2547.
he report gives results of evaluations of the technical and economic feasi-
bility of candidate control methods which may be capable of 99% removal of total hy-
drocarbons (HC) emitted from asphalt-saturating and air-blowing operations in asph-
alt roofing plants, sources of HC emissions for which control technology has not been
well characterized. The evaluations were based on information from the literature,
theoretical analyses of control systems, and contacts with equipment manufacturers
and plant operators. An industry survey showed that thermal incinerators or after-
burners are currently the only technique used to control air-blowing emissions. Con-
trol techniques for saturator emissions include afterburners, wet scrubbers, high
efficiency air filters (HEAF's), and electrostatic precipitators (ESP's). Theoretical
analysis of candidate control systems indicated that thermal afterburners, HEAF's,
and ESP's could remove 99% of the particulates, but it is doubtful that wet scrubbers
could achieve 99% removal. Further device evaluation, to identify candidate devices
to be recommended for more research and development, showed that afterburners are
already well developed and should be capable of 99% removal; but they cost much more
than HEAF's and ESP's and fuel availability could constrain widespread use. The
report recommends that pilot scale HEAF's and wet ESP's be tested on an air-blowing
source.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATi Field/Group
Air Pollution
Asphalt Plants
Roofing
Dust
Hydrocarbons
Incinerators
Afterburners
Scrubbers
Electrostatic Precip-
itators
Air Pollution Control
Stationary Sources
Asphalt Saturation
Air Blowing, Particulate
Thermal Incinerators
Wet Scrubbers
HEAF Units
13B
13C
11G
07C
21B
07A
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
125
2C. SECURITY CLASS (Tills page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
115
------- |