EPA-340/1-77-004
March 1976
               PRELIMINARY EVALUATION
              OF AIR POLLUTION ASPECTS
                       1  OF THE
                   DRUM-MIX PROCESS
                 U.S. ENVIRONMENTAL PROTECTION AGENCY
                       Office of Enforcement
                     Office of General Enforcement
                      Washington, D.C.  20460

-------
                                   EPA-340/1-77-004
 PRELIMINARY EVALUATION
OF AIR POLLUTION  ASPECTS
             OF  THE
     DRUM-MIX PROCESS
               Prepared by
               JACA Corp.
             506 Bethlehem Pike
        Fort Washington, Pennsylvania 19034
          in partial fulfillment of Task 8
            Contract No. 68-02-1356
          EPA Project Officer:  Kirk Foster
               Prepared for

       U.S. ENVIRONMENTAL PROTECTION AGENCY
            Office of Enforcement
          Office of General Enforcement
            Washington, O.C. 20460

               March 1976

-------
       This report was furnished to the U.S.  Environmental Protection
Agency by JACA Corp., Fort Washington, Pennsylvania, in partial fulfill-
ment of Contract No. 68-02-1356, Task No.  8.   The opinions, findings, and
conclusions expressed are those of the contractor and not necessarily
those of the Environmental Protection Agency.  Any mention of products
or organizations does not constitute endorsement by the U.S.  Environmen-
tal Protection Agency.
                                   11

-------
                             Table of Contents









Section                                                               Page




          ACKNOWLEDGMENTS	   i:Li




          LIST OF EXHIBITS	    iv




  1       INTRODUCTION  	     1




  2       THE DRUM-MIX PROCESS  	     3




  3       EMISSIONS	    16




  4       EMISSION CONTROL TECHNIQUES 	    20




  5       EMISSION DATA ANALYSIS  	    29




  6       EMISSION FACTORS  	    40




  7       SOURCE TESTING	    43




  8       FINDINGS AND RECOMMENDATIONS	    46




          REFERENCES	    49




          APPENDIX A   Manufacturers of Drum-Mix Plants 	    A-l




          APPENDIX B   Sampling Train Modification	    A-2
                                    111

-------
                              ACKNOWLEDGMENTS
          The assistance obtained from many individuals from EPA and sev-
eral State air pollution control agencies during the conduct of this evaluation
is gratefully acknowledged.   JACA Corp.  wishes to especially thank Mr.
Kirk Foster of EPA for his supervision and coordination with the various
regional offices of the agency.   We also wish to express our appreciation
for information provided by a number of manufacturers of the drum-mix pro-
cess contacted by us during this study.
                                     IV

-------
                           LIST OF EXHIBITS



Figure                                                                Page

  1        Schematic of Shearer Type Drum-Mix Plant                     6

  2        Drum-Mix Plant With Separate Asphalt Injection               8

  3        Typical Drum-Mix Moisture Content and Aggregate Mix
           Temperature Profiles                                         9

  4        Venturi Scrubber                                            25

  5        Efficiency vs. Size for Typical Venturi      '               27

  6        Uncontrolled Emissions from Drum-Mix Process                32

  7        Emissions With Dry Mechanical Collectors                    34

  8        Emissions With Wet Scrubber Controls                        35

  9        Emissions With Venturi Scrubber                             37

 10        Typical Drum-Mix Exhaust Flow Rates                         41

 11        Relationship Between Condensibles and Type of Control       45
Table

  1        Number of Drum-Mix Plants By Site                           15

  2        Composition of Asphalt Hot-Mix Emissions from Truck
           Loading of Product                                          18

  3        Particle Size. Distribution Before and After
           Primary Collection                                          23

  4        Emission Factors for the Drum-Mix Process                   42

-------
                                 Section 1




                               INTRODUCTION









       New Source Performance Standards for the asphalt concrete industry




were published on March 8, 1974 (39 FR 9308), pursuant to Section 111 of




the Clean Air Act (42 USC 1857 et. seq.).  These standards are applicable




to sources whose construction or modification commenced after June 11, 1973.




       There has been a new process development of significance in asphalt




concrete production technology since the promulgation of the standards.  A




new production process called the "drum-mix" process (also known as "drum




dryer", "turbulent mass") has gained increased commercial acceptability in




the industry and now constitutes an important portion of new asphalt con-




crete plants.  It is estimated that 30% of new asphalt concrete plant con-




struction over the past 3 years is  of the drum-mix type.




       Although various versions of the drum-mix process have been in exis-




tence for a number of years, its significant use in the production of asphalt




concrete is a recent phenomenon.  There are at least eight manufacturers of




such plantr (See Appendix A).   Based on information gathered during this




evaluation, it is estimated that there are at present approximately 130




to 150 asphalt concrete plants in the U.S. using the drum-mix process.  New




Source Performance Standards are applicable to between 50 and 70 percent of




these plants.  It was the intent of this preliminary evaluation to rely on




existing data rather than develop new data through extensive plant inspec-




tions and emission testing.  Although this approach would be susceptible to




data inadequacies since there was no control over the source test reporting,




it was felt that such an approach was commensurate with the modest time and




funds available for the task.

-------
       EPA regional offices, state air pollution control agencies and




manufacturers of drum-mix plants were therefore the chief sources of data.




Seventy emission tests were obtained and screened for methodology, calcu-




lations, isokinetic conditions, etc.  Sixty-three tests were found accep-




table for inclusion in the analysis contained in Section 5.  Even these




tests, however, often inadequately described process materials, control




equipment operating parameters, and process operating conditions.  One




drum-mix plant was tested by the contractor using EPA Method 5 with some




equipment changes.

-------
                                 Section 2



                           THE DRUM-MIX PROCESS






       There are approximately 4000 asphalt concrete plants in the United



States of the familiar "conventional" type described in several publica-


      345
tions. ' '   The salient features of the newer drum-mix process can best



be described by comparing it with the way in which asphalt concrete is



produced in these "conventional" asphalt plants.



       The conventional process begins with the conveying of a pre-determined



mixture of different sized cold aggregates from separate storage bins into



an inclined rotary drum which drys the aggregate  by counter-current flow



interaction with combustion gases from a burner mounted at one end of the



drum.  The dried, heated aggregate is then transported by a hot elevator



to a set of vibrating screens located over storage bins where it is sized



and stored.  Pre-designed quantities of the sized}dried aggregate are



weighed  and fed into a pugmill where it is mechanically mixed with heated



asphalt to produce the desired finished product.   The mixing of the aggre-



gate with "he asphalt is accomplished by either a batch or continuous pro-



cess.  Thus the drying and heating of the aggregate, and its mixing with



aj.phnlt are carried out in separate stages in the conventional asphalt plant.



A majority of the emissions are from the drying and heating stage in the form



of entrained particles, the remainder coming from vents from the mixing



tower which is nearly totally enclosed.



       In the drum-mix process, the aggregate is dried, heated and mixed



with asphalt in the same vessel -- a specially designed rotary drum dryer.



This obviates the need for a separate mixing tower with screens, weigh



hopper and pugmill, thereby reducing plant capital costs and improving

-------
portability.  These are two of the advantages cited by drum-mix process

manufacturers in selling their equipment.

       The major equipment differences can be shown in table form as follows:
       Conventional Plant

       Cold Storage bins § hoppers
       usually with vibratory feeds

       Load cells sometimes used
       Dryer with less sophisticated
       flight design $ counter cur-
       rent flow - no asphalt injec-
       tion

       Hot elevator

       Hot screens

       Weigh  Hopper

       Pug Mill

       Storage silo and conveyor
       optional but usually found
       in continuous process
Drum-Mix Plant
Same
Load cell nearly always
used

Dryer with sophisticated
flight design, parallel flow
and asphalt injection
Not required

Not required

Not required

Not required

Storage silo § conveyor re-
quired
       The different versions of the drum-mix process can be classified

in two ways:   the manner in which the material flows with respect to the

flow of gases, and the point at which asphalt is introduced into the drum.

       The majority of the designs currently marketed in the U.S. utilize

a parallel - flow dryer, where the flow of material and hot gases is in

the same direction.  The hottest flame and gases exist at the charging

end of the drum, where the aggregate is at its coldest temperature.   It

is felt that in this manner, asphalt is best protected from oxidation

by moisture.   Another characteristic of the parallel-flow dryer is that

lower aggregate discharge temperatures result.

                                    4

-------
       In the counter-flow dryer design, the aggregate and asphalt are com-

bined at the inlet end of the drum, and the drying-mixing process proceeds

toward the burner end of the drum, where the mixture is discharged before it

comes into direct contact with the flame.  It is estimated that the counter

flow dryer design accounts for only 5-10% of the drum dryer market, and

plants using this design are generally lower in capacity than the parallel-

flow dryer.  Typical capacities of a counter-flow dryer are 40-50 tons per

hour, whereas those of the parallel-flow dryer range from 100 to 600 tons

per hour.

       Parallel-flow dryers which comprise 90% of known drum-mix plants

can be divided into two general types based on the point of introduction

of the asphalt.

       In the Shearer process, the aggregate and asphalt arrive in the mixer

at the same time, alongside a stainless steel firebox which shields the

mixture  from direct contact with the burner flame.  A chute then discharges

the mixture into the next section of the drum where the flight design pro-

duces a  mixing action without developing a full curtain of material through

the flame.  In the following section of ^ he drum, the flight design causes

a full curtain of material to develop, where mixing action takes place in

?n atmosphere of steam and hot gases.  The finished mix is discharged at

the end  of the drum mixer onto a conveyor where it is transferred into a

heated storage silo for delivery into trucks.  This process is shown sche-

matically in Figure 1.

       Another version of the Parallel-flow design introduces the aggre-

gate separately into the dryer drum.  Drying of the aggregate begins imme-

diately  in direct contact with the burner flame.  A full curtain of ag-

gregate  is developed in the first section of the dryer drum:  In the next

section  of the drum, out of direct contact with the flame, adjustable
                                    5

-------
  Heated

  Storage

   Silo
                          Conveyor
                      To Exhaust Fan
                            t i
 Lħ
  ITU
Finished Product
to Trucks
Asphalt
Pump
                                      Rotary Drum
                                                           Asphalt
                                                           Storage
                                                           Tank
                                                                                    Aggregate Storage Bins
                                 Burner and
                                 Turbo-Compressor
                                                           able Speed
                                                       Conveyor Belts
                                                  Figure  1

                                   SCHEMATIC OF SHEARER TYPE DRUM-MIX PLANT

-------
spray bars coat the aggregate with hot asphalt.  Mixing is completed fur-




ther along in the drum and the mixture is discharged as described previously.




This process (shown in Figure 2) can be further described by the moisture




content and mix temperature changes in the dryer as shown in Figure 3.  This



shows that the bulk of the moisture is removed during Phase 1 where the ag-




gregate is in the burning zone.  Asphalt is generally introduced during




Phase 2 where it is assumed* that moisture trapped deeper into the aggregate




surface begins to vaporize.  The escape of this moisture through partial




coating of the asphalt, in Phase 3, produces violent foaming, which is said




to increase the uniformity of the asphalt layer coat on the stone particles.




Phase 4, at the far end of the drum, sees a rapid increase in the mix tem-




perature after the moisture has escaped from it.




       The extent to which these drum-mix plants will be found in new asphalt




concrete plant construction depends on the resolution of a number of questions




concerning product quality, efficiencies of operations, and the nature of




emissions from the drum-mix system as opposed to what has been called the




"conventional system."  The nature of drum-mix system emissions is discussed




in subseq. ;nt sections of the report, while the product and efficiency




questions will be briefly covered here.  It is not the intention of this




trif.f discussion to present an authoritative, quantifiable analysis of the




two systems since that was not the primary purpose of the project.  Jleadily




available data was used which was not verified in all cases although some




effort was expended to resolve obvious ambiguities.  Information frequently




advanced by manufacturers, users, and state personnel which may have a




bearing on the growth of the market and hence the amount of effort EPA should




devote to enforcement related activity in this area is presented in reportorial




fashion with limited analysis of the information advanced by the various




sources.

-------
                                                                                    Aggregate Storage Bins
                                                          Asphalt
                                                          Storage
                                                           Tank
                                          Adjustable
                                          Spray Bars
Finished Product
to Trucks
                                                                                 Burner  and
                                                                                 Turbo-Compressor
Variable Speed
Conveyor Belts
                                                Figure 2

                            Drum Mix Plant with Separate Asphalt Injection

-------
280

260

240

220
J2  200
   180

   160


   140
               PHASE 1
DRUM MIXER LABORATORY  SIMULATION
DENSE GRADED LIMESTONE MIX:
MOISTURE AND TEMPERATURE
PROFILES
    PHASE_2     PHASE  3         PHASE 4
                                               -i-REMOVE FREE WATER
                                               -=-DEVELOP INTERNAL
                                                    VAPOR PRESSURE
                                               -T VIOLENT FOAMING
                                               ~ RAPID TEMP. RISE
                                                    u
                                                    £
                                                    o
                                                    o
                                                    I
                                                    H
                                                    CO
                                                    M
                             TIME THROUGH DRUM, (MINUTES)
                                       Figure 3
        TYPICAL  DRUM-MIX MOISTURE CONTENT AND AGGREGATE MIX TEMPERATURE PROFILES
       Source:   Reference  1

-------
       Feed

       Since the drum-mix process does not use hot screens to control the

aggregate blend, more careful control of cold aggregate gradation is neces-

sary.  Usually three or four cold aggregate storage bins are employed with

variable speed conveyor belts from each bin, sometimes coupled with variable

gates that feed the aggregate onto a main conveyor belt, where the aggre-

gate weight is monitored by means of a load cell.  Some plants monitor the

aggregate weight on each belt conveyor being fed by the storage bins.

       The rate of asphalt feed is controlled either manually or automatically

to maintain the proper ratio of asphalt to aggregate.  The trend in asphalt

plants of both the conventional and drum-mix type is to make greater use of

automatic control.  Where automatic control is used a frequent technique

employed is to feed the signal from the aggregate load  cell(s) back through

a control loop which actuates a pump to feed the heated asphalt into the

dryer drum.

       The combination of aggregate feed rate and size control, and asphalt

injection rate, allows the operator to change both the production rate and

product mix throughout the cycle.



       Asphalt

       Penetration grade asphalts are used in the drum-mix process, often

in conjunction with proprietary chemicals to insure proper coating and

adhesion.



       Product

       Asphalt concrete top mixes can be produced at a temperature of 210

to 220 F, compared to 300 to 325 F in the conventional process.  At these

discharge temperatures, the mixture contains from 1 to 3% moisture,
                                   10

-------
compared to 1/4% or less in conventional mixtures.  Manufacturers of drum-



mix plants claim that this higher moisture assists in laying the product;in




that during the field compacting operation, it acts as a lubricant.  Moisture




equilibrium of the layed product from a drum-mix plant is attained by a




loss of moisture while the layed product from a conventional plant reaches




moisture equilibrium by a gain in moisture from the surroundings.
       Fuel
       The burners are usually fired with fuel oil although liquid propane




can also be used when available.




       Since the asphalt concrete produced by drum-mix plants is at a lower




temperature, fuel savings are claimed from the process.  A conventional




plant uses an average of 2 gallons of fuel oil to produce one ton of mix.




A drum-mix plant under the same conditions of aggregate moisture and ambient



temperature is reported to require an average of 1 1/2 gallons of fuel oil




per ton of mix.








       Process/Product Considerations




       Major commercial advantages of the drum-mix process over the con-




ventional process as cited by drum-mix manufacturers are overall lower




capital costs and increased portability due to the elimination of the mixing




tower.  This portability advantage is reflected in the high number of por-




table plants.  It is estimated that more than 70% of the drum-mix plants




currently in operation are portable.  This portability poses secondary




control problems when wet collection systems are employed.  Each new site




must have an appropriate water supply available, and must have a proper




water disposal facility to assure that applicable water pollution control





                                  11

-------
regulations are being met.  From an enforcement viewpoint the ability to




move the plant could increase record keeping and inspection problems es-




pecially when different state jurisdictions are involved.




       Lower operating costs arising from fuel economy, and lower maintenance




costs due to generally fewer system components are also cited.  On the other




hand, process controls to ensure desired product quality are more complex




with the drum-mix process, and the ability of the process to detect changes




in product quality and to correct for it is limited because the proportioning




of different sized aggregates and asphalt is carried out prior to their



entry in the drum.  In a conventional plant, the aggregate is monitored




at the point of introduction into the dryer and before entering the mixer




from hot bins which store the dry aggregate according to size.  In conventional




batch plants finer control is accomplished because each individual batch can




be mixed after preweighing all the ingredients for that individual batch.




The amount of fine material (frequently classified as aggregate passing a




200 mesh screen) in the product is also similarly capable of better control




in a conventional plant where the fines can be metered back into either the




hot elevator or the hot bin from the dry collector (cyclone and/or fabric




filter).




       The major market for asphalt concrete is highway construction, where




local, state and federal highway departments exercise careful product quality



control.  Highway specifications for asphalt concrete include sizing of




aggregate, proportions of asphalt to aggregate, coating of asphalt, tem-




perature and final moisture content of the product depending upon the grade




(e.g. sub-base, base, top, curbing), or use of the product.  Each asphalt




concrete plant doing business with the highway department is approved,



based on the product quality and the ability to control it.  The product





                                  12

-------
 quality  control  capability of the drum-mix process is being carefully in-

 vestigated by  state transportation agencies with varying degrees of accep-

 tance.   Many Eastern States have given only conditional acceptance as for

 example  in Pennsylvania,.where approval of a drum-mix plant for highway

 work  is  conditional, and determined on a case-by-case basis.

       Two other limitations of the drum-mix process from a production

point of view were often encountered.   The drum mix process design lends

itself to long, continuous production runs with the same mix.   Consequently,

it is less suited for applications where the demand for different types

 of mix is random, even on a given day, as in the case of many plants loca-

 ted near urban areas, especially in the East.  The dwell time of the mix

 in the dryer is on the order of 5-7 minutes, and by the time a change is affec-

 ted at the input proportions of the aggregate, as much as a half hour may be

 required before a new steady state is reached for the production of a dif-

 ferent mix.  In the conventional plant, on the other hand, a product change

 can be accomplished within a matter of minutes by simply changing the pro-

 portion  of different sized, dried aggregates in the pugmill mixer, prior

 to making a change in the dryer input gradation.

       Many plants buy aggregate from different sources depending upon need

 and price.  The size specification generally covers a range, and within the

 designated range there is considerable variability in size characteristics

 of aggregate occurring between sources, depending upon type of rock, crush-

 ing and  screening sequence, etc.  Such variations in the source of aggre-

 gate  cannot be easily accommodated with a drum-mix plant, because a new

 setting  of dryer input controls is required to produce the same product.

       In areas West of the Mississippi, with less population and highway

 concentration, the  ability of a plant to move from one construction site to

 another, an advantage of the drum-mix, apparently outweighs the above  limi-
                                 13

-------
tations.  The majority of drum-mixer plants thus  exist  in the  Western half



of the United States as can be seen from Table 1  which  shows known plant



location by state.  It should be noted that the plants  can be  moved  so that



locations shown are subject to change.
                                   14

-------
                     Table 1

        NUMBER OF DRUM-MIX PLANTS BY STATE
Arizona                         5
California                      1
Colorado                        5
Georgia                         2
Illinois                        1
Indiana                         4
Iowa                            4
Kansas                          9
Louisiana                       2
Maryland                        1
Michigan                        2
Minnesota                       7
Mississippi                     2
Montana                         1
Nebraska                        1
Nevada                          3
North Dakota                   11
Ohio                            1
Oklahoma                        4
Oregon                         10
Pennsylvania                    4
South Dakota                    4
Texas                          12
Utah                            5
Virginia                        1
Washington                      3
Wisconsin                       3
Wyoming                         3
                        15

-------
                                 Section 3

                                 EMISSIONS



       Sources of air pollution from the drum-mix process include both fu-

gitive and stack emissions.  In both instances, the source, nature and mag-

nitude of the emissions are considerably different from their counterparts

in the conventional process.  This difference is attributable to the marked

difference in the processing techniques.

       The sources of emissions of the conventional and drum-mix plant

are shown in the following table:
     Conventional
                                   c
Fugitive emissions from stockpiles, cold
feed bins and conveyors

Fugitive emissions from finished product
discharge to trucks, or to storage

Stack emission from scavenger ductwork
to hot elevators, hot screens, bins,
weight hopper, mixer

Stack emissions from dryer
Drum-Mix
 Same
 Finished product conveyor to silo
 None
 Stack emissions frm drum-mix
     Fugitive emissions from stockpiles, cold feed bins and the conveying

machinery prior to the introduction of the aggregates into the drum-mix,

are, as in the case of the conventional process, dependent of the moisture

content, size of aggregate and ambient conditions.  The conveying of finished

product from the drum-mix to the storage silo produces fugitive emissions

similar to those in conventional plants with continuous processes which

are generally equipped with storage silos.
                                    16

-------
      Stack  emitted  dust  from  scavenger ductwork  connected to emission
 sources  such  as  the hot  elevator, screens, bins  and mixer are eliminated
 in  the drum-mix  process.  Any fugitive emissions from these sources are
 also  eliminated  in  the case of drum-mix plants.
      Fugitive emissions  from  the  finished product  in discharge to trucks
 or  storage  from  conventional  plants have been  characterized in a study done
 by  the Asphalt Institute and  the  Exxon Research  and Engineering Company,
 and are  summarized  in Table 2.  These data identify the various components
 in  the fugitive  emissions from the handling of the finished product.
       As in the  conventional  plant,  stack emissions  are the major air pol-
lution source in  the drum-mix  plant.   The new source  performance  standards
apply to  the stack emissions in the form of emission  concentration and
opacity.   The opacity standard, however,  also applies to the entire process
subsequent to the introduction of the aggregate in the drum, and therefore
covers fugitive emissions from that point on.
       Both particulate as well as gaseous components are present in the
stack emissions from a drum-mix operation.   The particulate emissions gen-
erally include mineral,  hydrocarbon and carbonaceous  matter.  Aggregate
dust entrained during the drying-mixing action in the drum is the source
of the mineral matter,  while the hydrocarbon and carbonaceous matter results
primarily from the exposure of asphalt to various degrees of oxidation in
the drum.  This,  as well as the combustion of the fuel,  also accounts for
the gaseous  emissions  in the  stack.  Section 5 includes  data on emission
tests of plants with varying production capacity and  different degrees
of control.
     The  test data indicate that uncontrolled stack emissions from drum-
mixers .are significantly less  than those  from the conventional plant,  by
almost an order of magnitude.   The simultaneous drying and mixing of the

                                   17

-------
                                   Table 2

      COMPOSITION OF  ASPHALT HOT-MIX EMISSIONS FROM TRUCK LOADING OF PRODUCT
                            (Conventional  Plants)
Sample Location .

Number of Samples
Edison, N. J.
Greensboro, N. C.

        2
Non-Visible Components (ppm)

  Carbon monoxide (CO)                             4-6                 3-4
  Nitrogen dioxide (N02)                          <0.1               .05-.08
  Sulfur dioxide (S02)                             <2                 <0.5
  Hydrogen sulfide (H2S)                        <0.2-1.5              <0.2
  Carbonyl sulfide (COS)                          <0.2                <0.2
  Mercaptan (RSH)                                 <0.2                <0.2
  Aldehydes (ECHO)                                <0.1               0.3-0.4
  Phenol (00H)                                     <1                  <1
  Ozone (0^)                                      <0.1                 	
  Methane (CH^)                                    2-3                 2-3
  Non-methane Hydrocarbons  (C2-Cg) (NMH)           <1                  <1
  Volatile organic compounds (Cy-C-^) (VOC)   .   0.5-1.5             0.5-1.0

Particulates (mg/m^)

  Total particulates                             2.6-7.2             0.5-5.7
  Benzene solubles                               0.3-2.8             0.2-5.4
  Polynuclear aromatics  (total), max.            0.00034             0.00016
  Nickel (Ni), max.                              0.000005            0.00004
  Vanadium  (V), max.                             0.00008            <0.0001
  Cadmium (Cd)                                     	              <0.00005
  Lead  (Pb)                                        	              <0.00005
NOTE;  Where the less than  (<) values are indicated, the numbers represent the
       sensitivity of the sampling or testing procedure used.  If the component
       is present at all, it is below the value shown.
   Source:   Reference 2
                                    18

-------
aggregate with asphalt in the drum tends to trap a large portion of air-




entrained mineral particles in the asphalt spray resulting in "balls" which




further breakdown, coating the surface of the aggregate.  Manufacturers




claim that this considerably reduces the amount of mineral dust carried




over by the exhaust gases, and promotional literature from manufacturers




prominantly advertises this attribute.  However, a reduction in the uncon-




trolled emissions does not generally make it any easier for a drum-mixer




to meet the new source performance standards as will be discussed in Section 5.




     An increase in the asphalt-related emissions is generally found in the




drum-mix exhaust, as compared to the exhaust from a conventional plant.



In a conventional plant, this type of emission is vented into the exhaust




from the enclosed mixer, accounting for approximately 5 to 10 percent of




the total exhaust flow rate directed to the control device.  The genera-




tion as well as entrainment of asphaltic emissions is therefore limited.




In the drum-mixer, on the other hand, the asphalt is exposed to the total




exhaust in a turbulent fashion, thereby tending to increase the entrain-




ment of asphaltic products.




     While the uncontrolled mineral dust is generally less in a drum-mix




plant, and the asphalt-related emissions greater, than a conventional




plant, the amounts are a function of process design and operating variables




of a facility.
                                  19

-------
                              Section 4

                      EMISSION CONTROL TECHNIQUES



       Control techniques applied to existing drum-mix process plants have

varied from state to state because of the varying stringency of regulations

for existing plants which are included in the State Implementation Plans.

       In the case of plants that should fall under the New Source Perfor-

mance Standards, another preliminary study for EPA' has indicated that

reporting of such new plants may suffer appreciable omissions so that it

can be expected that some plants, that should be classified as new, and hence

under federal regulations, may be classified under SIP.  Such plants will

then be frequently operating at levels that meet the SIP, but which may not

meet the NSPS.

       Controls at the 63 plants reported in the various test records reviewed

included the following:



     Type of Control               No. of Plants       Percent of Plants Reported

None (May have knockout boxes           14                        22
      as an integral part of
      the output ducting)

Cyclones or multicyclones                7                        11

Low energy wet scrubbers                24                        38

Venturi scrubbers                       18                        29

Baghouses                                0                         0

Electrostatic precipitators              0                         0



       An important word of caution concerning the above table and the

description of control devices that follow:  This preliminary evaluation


                                  20

-------
of air pollution aspects of the drum-mix process did not involve develop-




ment of primary data.  Instead, it is based on information available in




stack test reports that were available to us, sometimes only after they




were several times removed from the actual testing organization.  In almost




all cases the stack test reports gave no or only meager data on the type




of control system being used.  For example, in the case of wet systems the




liquor flow rate and pressure drop was not usually presented.  Details




of stack sprays insofar as their location, type and number of nozzles and




liquor flow rate was not stated.  It was therefore necessary to ascribe



pressure drops and some operating characteristics to some of the vague




descriptions given in the reports based on experience with conventional




reports.




       Fourteen reports indicated no controls used.  We believe that in




these cases there may have been a knock-out box which was never mentioned




in the report.  Such boxes operate on the principal of abruptly changing




the direction of the gas stream so that inertial force of the heavier par-




ticles overcame the entrainment forces of the gas stream and they are




essentially "knocked out" of the stream.  This phenomenon is an undesirable




effect where it inadvertently occurs in poor duct design causing duct wear




and pile-up of material.  The technique has only low efficiency, operating




best when the size distribution of particles is heavily toward larger and/




or more dense particles.



       Primary cyclones of the type generally used with conventional plants




are more efficient that the knock-out boxes in that they remove between




50 and 70% by weight of the entrained dust from conventional asphalt batching




plants.  Such cyclones operate on the centrifugal force principle.  Practical




limitations on their use usually involve re-entrainment problems and unworkable






                                  21

-------
high pressure drops.  The overall efficiency of the cyclone depends on




particle size, geometric design of the cyclone and pressure drop.   In most




practical situations involving mineral dust, pressure drops are on the




order of 2 to 4 inches of water.  Table 3 shows the particle size distri-




bution before and after entering a cyclone collector on a conventional




asphalt plant.




       Stack sprays were mentioned in the reports, but as previously stated




there was no detail on the type nozzles, location, and liquor rates.  The




techniqe however is to inject water droplets into the exit stack relying




on impaction and agglomeration of the dust particles to produce particles




either heavy enough to overcome the upward force of the gas stream so that




they settle to a collection sump, or of sufficient mass to be removed by




baffle plates upon which particles can impinge, or by centrifugal forces




applied by mechanical means subsequent to the introduction of the spray.




       Another type of wet collector frequently encountered in conventional



plants and what we think the testers generally meant when they referred to



"wet scrubber" in their test reports on the drum-mix plants used in this




preliminary 'investigation was a dynamic scrubber which incorporates a wet




fan as an integral part of the unit.  A spray of water is either directed




toward an impeller type fan or is fed axially in the case of paddle-wheel




fans.  The mechanism is mainly one of impingement of dust particles on the




wetted rotating blades.  The purpose of the sprays is to keep the fan blades




wet and to flush away the collected dust.  This technique also relieves




the abrasion and condensation buildup which would otherwise accumulate on




the blades. Our experience with these devices on conventional plants indicates



that they usually operate at 5" to 10" water gauge.
                                  22

-------
                               Table 3

                      PARTICLE SIZE DISTRIBUTION

                  BEFORE AND AFTER PRIMARY COLLECTION
               FROM DRYER AND VENT

            Size/C     % Less Than
  FROM PRIMARY COLLECTOR

Size/*     % Less Than
5
10
15
20
25
30
55
40
45
19.5
30.5
38.2
45.1
50.1
55.5
60.0
64.0
67.5
5
10
15
20
25
30
35
40
45
78.00
96.40
97.50
97.80
97.90
98.03
98.20
98.28
98.40
Source:  Air Pollution Control Technology and Costs In Nine Selected
         Areas, Industrial Gas Cleaning Institute
                                  23

-------
       Those installations which exhibited the best degree of control were




venturi scrubbers.  Again, the reports did not adequately describe them




nor note the pressure drop or liquor flow rate.  Such venturi scrubbers




could fall into several categories as shown in Figure 4, but their operating




principles are essentially similar.




       These scrubbers, with pressure drops in the range of 20 inches of




water, or greater, generally are capable of reducing emissions of conven-




tional plants to below the required NSPS of 90 mg/scm (0.04 gr./dscf).




       The venturi scrubber is a high efficiency wet collector that operates




by impinging particulates on atomized water droplets.  The effective mass




of the particle thus increased, cyclonic separation is then possible.




       As the particle laden gas enters the device, a constriction reduces




the cross-sectional area of the gas stream, thereby increasing the stream




velocity.  This correspondingly increases the velocity of the particles




relative to the formerly stationary water droplets that were introduced at




the apex of the constriction.  Increasing the relative speeds heightens the




probability that a particle will impinge upon the water droplet.  As the




dust-laden water droplets leave the venturi constriction, they further




agglomerate due to deceleration.  The gas stream then passes through a cyclonic




separator, which removes the larger, heavier particles formed during the




agglomeration phase.




       Venturi scrubbers can achieve collection efficiencies in excess of




99%.  Variables affecting efficiency include pressure drop, water injection




rates, venturi design, and particle concentration and size.  Collection




efficiencies improve with higher pressure drop, attainable by increasing




the throat velocity by constricting the throat and to a lesser extent by




increasing the water injection rate.  Pressure drops will probably be 20"





                                  24

-------
                                 Figure 4

                            VENTURI SCRUBBER
                   Venturi scrubber may feed liquid through  jets  (a)
                   over a weir (b), or swirl them on a shelf (c).
Source:
Control Techniques for Particulate Air Pollutants. USDHEW 1969
                                 25

-------
or more for most venturies while water injection rates normally encountered




will nominally be 6 and 10 gallons of water per minute per 1000 acf of gas.




Efficiencies fall rapidly at injection rates below this range; rates in




excess of 10 gallons of water per minute 1000 acf of gas produce lesser




increases in collection efficiencies.




       Greater particle concentration also improves collection efficiency.




Assuming the number of water droplets formed in the system is constant,




the frequency of particle collisions is increased when more particles are




introduced into the system.  Figure 5 shows a nominal collection efficiency




and particle size relationship for a typical venturi scrubber.  Note that




the efficiency is greater than 97% for particles larger than 1.5 microns;




note too that the efficiency falls sharply for particles less than 1 micron




for a fixed set of conditions.




       Disadvantages of venturi scrubbers include high operation costs




associated with producing high pressure drops, and also the need for large




quantities of water which entails elaborate recycling of alkaline, acidic




or  adoriferous water.  This would require the use of settling basins which




also present a problem of solid waste disposal when they must be dredged.




       Advantages of venturi scrubbers include their relatively low initial




cost, and their ability tİ partially control $be hydrocarbon emissions




from the drum-mix operation.




       No drum-mix plants using fabric filter controls were encountered




in  this preliminary investigation.  The asphaltic emissions from the drum-




mix process as well as mineral particles coated with asphalt are difficult




to  control with  fabric filters because of sticking and blinding of the filter




medium.  Two sources have reported that they were experiencing plugging




even  using  cyclones from  these emissions.

-------
                                                        Figure 5


                                        EFFICIENCY  vs.  SIZE FOR TYPICAL VENTURI
IS)
         
-------
       Electrostatic precipitators were also not encountered.  The low gas

volumes associated with conventional asphalt concrete plants and the high

fixed costs for the portion of the precipitator that develops the high

voltages needed generally renders such devices economically less attractive

than other solutions.  They can however attain collection efficiencies of

over 99% in conventional plants with proper gas conditioning.*  Wet electro-

static precipitators that do not require rapping, but remove the captured

particles by a film of water on the collector wall might reduce some of the

sticking problems.  Even with the somewhat higher gas volumes found in the

drum-mix plants, however, they may not be economically attractive, and they

may impose some plant portability restrictions.

       Important trends are discernable in the development of this relatively

young process technology.  One manufacturer has reported to us that they

are trying to affect certain changes in the process variables (such as

flight design through sections of the drum, rotational speed, slope, and

the point at which the asphalt is sprayed in the drum) whereby asphaltic

emissions could be reduced, even at the cost of increasing the mineral

particulate emissions, so that a fabric filter could be used for the control

of process emissions.
  Stack Test performed by JACA Corp. for compliance with Pennsylvania
  Regulations
                                  28

-------
                              Section 5



                        EMISSION DATA ANALYSIS








      .. The preceding sections provide general background of the drum-mix




process, and control techniques employed.  Quantitative information about




the emissions from the drum-mix process, collected during this study, is




presented here.




       It is important for the reader to keep in mind the sources and




limitations of the data which are described below.  The analysis is based




on information obtained from regional EPA offices, state air pollution




control agencies, independent stack testing companies and drum-mix manu-



facturers.  One counter flow drum mix plant was tested using Method 5 with




slightly modified equipment.




       The test information was obtained in various forms.  While most of the




information analyzed here is based on actual emission test reports; in




other cases, it was necessary to utilize aggregated summaries of test results




in others.  In one instance only the combined results of 31 tests were made




available to us by a drum-mix manufacturer.




       Thorough analysis and comparison of data contained in information




gathered from such diverse sources is hampered by two limitations:  (1)




the various modifications of the basic EPA sampling train used and approved




in various parts of the country does not permit full comparison of data




results from all tests and, (2) adequate information on source conditions




and description of control equipment and parameters was usually unavailable.




For example, some tests report particulate concentrations based on the front




half catch, some of the 'total' catch, and some do not report the  'condensible'




fraction.  None of the reports give important details of control equipment





                                  29

-------
configuration and only a few specified production rates, liquor rates and




pressure drops on wet scrubbers.




       Another limitation is that the source test data had to be judged




as to its acceptability on the basis of broad yardsticks such as isokineticity,




thoroughness in reporting the various stack-test related variables, although




a majority of the tests were conducted for compliance purposes and were




presumably accepted by the relevant government agency.




       As a result of the data gathering phase, emission tests results




were obtained from 70 different plant tests (not including 31 test runs




from plants where only the aggregate data were made available to us by a




manufacturer).  Of these 70 different plant tests we decided to exclude a




total of seven tests, 2 because the tests were run outside the isokinetic




range, 3 due to a paucity of information available (including the only two




counter current plants) and 2 because they were repeat runs on the same




plants included in the analysis.




       Of the 63 tests found to be acceptable for inclusion in the analysis,




14 were on uncontrolled plants, 7 were on plants with dry mechanical con-




trols such as cyclones and multicyclones, 24 were from plants with scrubbers




of the spray, impingement or wet fan type, and 18 were on plants with ven-




turi scrubbers of varying pressure drops.




       The 63 tests gave a total of 158 independent runs for analysis, of




which 108 reported only "front-half" results, 7 reported only the "total"




particulate matter, and 43 reported both front-half as well as total par-




ticulate concentrations.




       The emission concentrations obtained from the above tests are reported




below, in order of increasing level of controls, for parallel flow plants




which account for more than 90% of known drum-mix plants.(the one counter






                                  30

-------
current flow plant is plant D of Figure 6).   Due to the large variability




associated with these results, they have been plotted, in Figures 6, 7, 8




and 9, on a logarithmic scale.  Plants identified with letters are on the




horizontal axis, and information regarding control device, production rate




and capacity, where available, is indicated below each plant.  Where




available, maximum, minimum and arithmetic average concentrations are shown




for each plant tested.  The percentages shown above the test results for




some plants are percentage opacity readings reported in the test records.




       The opacity data suffers from many of the problems encountered in




the test reports.  There was no data sheet showing number of readings,




specific time, position of the observer, atmospheric conditions, etc.




Information on opacity in those reports where it was mentioned was almost




parenthetical, a brief statement that the opacity was a particular percentage.




       Figure 6 shows the range of particulate emission concentrations for




uncontrolled plants.  (Figures 6 through 9 include results from parallel-




flow plants except for plant D of Figure 6).




       The combined results from 31 tests on 9 uncontrolled plants (Figure 6)




with capacities varying from 300 to 350 tons per hour, at maximum moisture




removal rates, shows an extremely wide range of particulate concentration,




a maximum grain loading of 40.5 gr./scfd, a minimum of 0.14 gr./scfd, a




mean of 6.19 gr./scfd with a standard deviation of 8.2 gr./scfd!  The other




five plants all have much lower average readings and the ranges of plant




B § F are more like those encountered in conventional plants.  Of the three




plant reports with opacity data only plant E was outside of NSPS.



       The significant variation in uncontrolled emissions from plant to




plant suggests that process variables might be one of the principal causes.




Plant D for example was known to be a counter current flow plant as described




                                  31

-------
4-1
i—i
CTJ
C
o
4-1
4-1
(J
1/1
-a

I/I
c
•H
nJ
f-i
CJ3
      50
      1C
     1.0
     0.5
     0.1




     .05

     .04
     .01

 Plant
                                  Figure 6

               UNCONTROLLED EMISSIONS FROM DRUM-MIX PROCESS
                   I
                  Maximum
                 ^Average
                  Minimum
                                    T
                            %=opacity
                  T
          I
                  10%
                                                   30-35%
                                                  I       I
          Federal Standard =0.04
 Avg.
 Production
 Rate,  tph
                  200
175
 Capacity,
  tph	
 60


100
166


200
                                                              95


                                                             200
Combined
Data from
31 tests,, 9 plants

 300-350


 300-350
                                     32

-------
on page 5 while all others are parallel flow plants.   The other plants are



made by a variety of vendors and plants have a considerable range of process




settings - much more so than do conventional plants.   Some vendor companies




indicated that they were conducting research on product characteristics




and emissions as a function of several process variables including:








       •    Point of injection of asphalt




       •    Flight design




       •    Aggregate mix




       •    Moisture content




       •    Temperature gradient through the dryer




       •    Drum rotational speed




       •    Rate of production




       •    Temperature of mix




       •    Type of Asphalt







       The analysis of the effect of these parameters was not attempted in




this work, and would entail quite a large study because of the number of




different suppliers and the present state of design flux.




       Emission concentrations from seven plants controlled with dry mechanical




collectors are shown in Figure 7.  The results fall between a high of 3.0




to a low of 0.03 grains/scfd.  Plant L for example just met the NSPS stan-




dard of 0.04 gr./scfd.  The mean of all the results was 0.853 gr./scfd, with




a standard deviation of 1.16!  Of the three plants reporting opacity readings




only one was within NSPS.



       Results of  emission tests from 24 plants controlled with wet scrubbers




as described in Section 4  are shown in Figure 8.  Seven out of twenty-four




plants were within NSPS.   Excluding plants N and T for which only the total





                                 33

-------
                                                   Figure 7

                                   EMISSIONS WITH DRY MECHANICAL COLLECTORS
      5.0
(4-1
i—I
n)
X
o
o
I/)
tn
C
rt
      1.0
0.5
      0.1
      .05
      .04
      .01
 Maximum
 Average  %
 Minimum
   I
= Opacity
                Federal  Standard = 0.04
                                                                      1
                                                                    30%
                                                                 ğħ
                                                                   ħH
                                                         15%
                                                                                     30-40%
                                  HE
                               3H
        Plant
      Average
      Production
      Rate,tph

      Capacity
         tph
               G
            Cyclone
              380
              450
   H
Cyclone
 250
 450
                                          Multi-
                                          Cyclone
 500
 500
               J
            Cyclone
360
500
   K
Multi-
Cyclone


 170


 200
   L
Multi-
Cyclone


 250


 450
                                      M
                                   Cyclone
                                                                                       160

-------
l/J
en
                                                                  Figure 8
                                                EMISSIONS WITH WET  SCRUBBER CONTROLS
          1 .0
          .04
                       1     1    I    1
                     1 — |
                    . j  I |
                     |  |
Maximum
Average
Minimum
= Opacity
                          '    '
                    1 LJ '
                                         10°;

a - Scrub.-'
b - IV ct !•' i
c - Stack
* Include; •
                                                                                                                       i b 1 es
                                                                          10°,,
                                                                     15-20%
                                                                                                           10%
                                                                           Federal Standard = 0.04

                                                                            0-10%
                                                   1     1    1    1    1     1    1    1    1
                                                                                 1    1    1     1
          Plant    N*    0    p .   Q    !<   S    T*    U    V    W    AA    BB    CG   DD    BE    FF    OG  ;1H    II    JJ    '.-'    '.I,    MM  NN
          Controls b,ab    aaaaa    a    a    a     b     aaaaaa     a     a     a     a     a   a
                  c
          Average

          i'rou.   210  5S5  425  ?   215  350   400   250   225   ?    450   110   225  278   392   336   200 528   400   350   400   125   240 540
          Kate,tph
         Capaci t v  •
           toh   dOO  600  TOO  V   400  400   700   600   600   ?    600
                                                                           ?  400  400  400   ?  600  400  400  430    '   400 600

-------
 catch  results were  available,  the  values  range  from 0.394 to 0.017, with a




 mean of .094 gr./scfd,  with  a  standard  deviation of 0.089 gr./scfd.  Of the




 twelve plants reporting opacity readings,  two were greater  than  20 percent.




        Figure 9 shows  the results  of emission tests on  18 plants with ven-



 turi scrubber controls.  The pressure drops  across the  venturi scrubber




 axe only known  for  three plants.   These are  shown in  inches water gauge




 in the lower right  of  the data plot.  The emission values vary from a



 maxium of 0.191 to  a low of  0.005  grains/dscf.  The mean of these values




 is 0.0557 and the standard deviation  is 0.052.  Of the  eight plants reporting



 opacity readings, all  were within  NSPS.




        The data (excluding the one countercurrent flow  plant) shows that



 the percentage^plants  passing  NSPS was:








        •   Uncontrolled                                               0%



        •   Dry Mechanical (Multicyclone)                              14%




        •   Wet Scrubber (Per  the  broad description in  Section 4)      29%



        •   Venturi Scrubber (Pressure  drop  known in  only 3 instances) 50%








       Attempts to relate  emissions concentrations with  production rates,



or production rate as  a fraction of the plamt capacity  for  each  of the




four sets of data presented  abow  failed to  show any  consistent  pattern.



       Beari ng  in mind the nearly  total lack of process and control setting




data, the t< ;sts reported shows:








        1.    Emissions  associated with drum-mixers  show a very large



             variability.  Process factors discussed previously may be



             responsible for these differences.





                                   36

-------
                                                               Figure  9
                                                 EMISSIONS  WITH VENTURI SCRUBBER
     04
~.   .005
.00!
1'iaiu    00
Control  -
Average  i:5
ProJ.
Rate, tph

Capacityjoo
tph
                      Maximum
                      \voraye
                      Mi n imuin
               = ilp;ic i ty
                                                                                         1 0°,,

                                                                                                       ' = 25-30 ''K.I'.
                                                                                                                           ta'idai\! =
                                                                   i—I A  P=  11 • ••:.<;.
                         QQ
                                                      UU
                                          VV     WW
                                         • Vonturi
                                                                                                      Bl
                                                                                                                    PI
                          160    230    300    275
200    250    500    500    500
                                                      400    dOO    174    500
                                                      500    600    200    600
                                                                                 380     300     210    450    230    300      400   200
                                                                                         400     220    600    400    400      400   250

-------
       2.    No uncontrolled parallel flow plant met NSPS




       3.    The single countercurrent flow uncontrolled plant marginally




            met NSPS,  but we are advised that such designs have inherent




            limitations in production capability,  and are not expected to




            be a significant part of the market




       4.    All but one of the mechanical collector controlled plants failed




            NSPS on grain loading as did two out of three of these plants




            reporting opacity readings




       5.    All eight plants reporting opacities which were controlled by




            venturi scrubbers were within NSPS opacity limits.








       Although an extensive hydrocarbon emission analysis was not under-




taken as a part of this study, several of the particulate emission tests




contained various hydrocarbon analysis.  The results are described below:








       •    Analysis of the particulate catch  from a test performed on




            Plant K, controlled by a multicyclone, showed that 21% of the




            filterable particulate was hydrocarbon.  The test lab assumed




            this material was asphalt.




       •    Total hydrocarbon (THC) analysis were run on Plant II controlled




            by a venturi scrubber.  Tests were performed by chromotography




            and the results for the three test runs were 163 ppm, 501 ppm




            and 112 ppm.




       •    A more extensive hydrocarbon analysis was performed on samples




            from Plant U controlled by a wet scrubber:
                                  38

-------
            THC = (methane equivalent)  - 41.5ppm




                  methane   -   .6 pprn




                  ethane    - 25.6 ppm



                  x         - 16.4 ppm




                  propane   -   .6 ppm




                  propylene -  1.0 ppm




            THC = measured on two other dates on the same plant were




as follows, but no explanation was given for the large differences from




the other dates:








                       2/75 - 7830 ppm




                       4/75 - 1187 ppm








       •    Plant R controlled by wet scrubber showed THC as 13.5 lbs./hr.,




            90% of which was methane and 5% ethane.  The test lab stated




            that no reactive hydrocarbons were found in measurable quan-




            tities.








       This aspect of the emissions will also require further testing and




analysis, once an acceptable method of hydrocarbon analysis is developed.
                                  39

-------
                              Section 6




                           EMISSION FACTORS









       The large variation in emission concentrations from drum-mix plants




with varying degrees of control shown in Section 5 makes the task of assigning




an emission factor for a given degree of control difficult.  The emission




factor estimates developed here, therfore, should be used with this limitation




in mind.  For purposes of assigning emission rate values to a plant in air




quality studies, NEDS applications, etc., however these factors are reasonably




appropriate.




       Figure 10 shows the typical range of flow rates as a function of




drum-mix production rates.  They were developed from theoretical calculations




based on a 5 percent moisture remove, 70°F inlet air, 350°F aggregate mix




and exhaust temperatures.  The higher limit of the range for a given pro-




duction rate is associated with a lean burning condition (ratio of flow




to stoichiometric flow = 2.5) and the lower limit with a richer burning




condition (ratio of flow to stoichiometric flow = 1.5).  This range of




conditions are  typically found in drum-mix plants.  This analysis yields




a value of 7000 to 4000 dscf air requirement per ton, respectively.   (Analysis




of available test data essentially confirms this range).




       The emission concentration  (mean values) found fres Section 5  for




each degree of  control, when multiplied by the air requirement determined




above yields a  range of emission factors shown in Table 4.
                                  40

-------
                                       Figure 10



                          TYPICAL DRUM-MIX EXHAUST FLOW RATES
     140,000
rt
ai
o
x
tu
         120
         100
          80
          60
          40
      20,000
1
                   acfm range

                   at 350°F
                                                              dscfm range
100         200         300



        Production Rate, tph
                                                             400
                                               500
600
                                          41

-------
                              Table 4
                       EMISSION FACTORS FOR
                      THE DRUM-MIX PROCESS*
   Degree of Control

   Uncontrolled

   Cyclone or Multi-cyclone

   Wet Scrubber (Stack spray, wet fan,
                 dynamic scrubber)

   Venturi Scrubber
    Particulate
 Emission Factor,
Ibs/ton of product

     3.6 - 6.2

     .49 - .85

     .05 - .09


     .03 - .06
The reader must bear in mind that there are different process designs
and that control parameters such as pressure drop, liquor flow, etc.
were generally not included in reports from which these data were taken.
                               42

-------
                              Section 7




                            SOURCE TESTING








       Difficulties in source testing drum-mix exhausts by means of the




EPA train have been reported.  The chief reason for this is the clogging




of the filter in the front half of the test train with asphaltic emissions,




which prevent isokinetic flow rates through the train because of high pressure




drop created at the filter.




       Source testing personnel who have had experience with drum-mixers




were contacted during this study, and the conclusion is that clogging of




the filter diminishes as the degree of control increases.  Tests on plants




with no controls or with dry mechanical controls asphaltic emissions appear




to be the cause of filter blinding since plants with wet controls were




reportedly not as difficult to test in this regard.



       Based on JACA's test experience on an uncontrolled drum-mixer, and




the experience of EPA's Emission Measurement Branch on the source testing




of asphalt roofing plant exhausts, the following techniques are suggested




to minimize the clogging problem and to reduce the need to frequently change




the filter during a run.








       •    A loosely packed portion of glass wool, inserted into the top




            half of the filter holder causes the sticky, asphaltic material




            to adhere to it, without causing excessive pressure drops through




            the train (See Appendix B).




       •    A cyclone and flask inserted in the hot box between the probe




            tube and the filter holder retains bigger particles, thereby




            reducing the build up on the filter medium.





                                  43

-------
       •    Methylene chloride is the preferred solvent for use in the




            sample recovery operations.








       Another potential area of concern was whether, due to the presence




of asphaltic matter in various stages of oxidation, the process emissions




contain large amounts of matter which escapes the filter but gets caught




in the impingers as "condensibles" matter.




       This aspect was studied by analyzing the ratio of "condensible" to




"front-half" catch for a total of 43 runs for which such data was available.




Figure 10 shows how this ratio varies with degree of control.  For uncon-




trolled plants, the condensible catch is on an average 35% of the front




half catch, but as the degree of control increases, this ratio decreases.




For drum-mix plants where the test reports cited venturi control devices,




this ratio reduces to 10%, although one anomalous result was found here with




a ratio of 83%.
                                  44

-------
                                                       Figure 11

                              RELATIONSHIP BETWEEN CONDENSIBLES AND TYPE CF CONTROL
in
        Type of Control
               No Control
               Mechanical
             Wet Scrubber
         Venturi Scrubber
                                                 10
20
30
40
                                                        Percentage Ratio:  Condensible
                                                                           Front Half

-------
                              Section 8




                      FINDINGS 5 RECOMMENDATIONS









       Drum-mix asphalt concrete plants are likely to constitute an impor-




tant fraction of new asphalt concrete plant construction, falling within




the New Source Performance Standards.  There are at present approximately




150 drum-mix plants occuring in 27 states in the United States of which




it is estimated that 50-70% fall under the provisions of NSPS.  At least




eight companies are engaged in the manufacture of drum-mix plants.




       Some product use restrictions imposed by state departments of trans-




portation at present inhibit the market acceptance of the process, but




further product test evaluations and experience by highway officials may




reduce this barrier.  The fact that eight manufacturers including four




prominent in conventional plants have entered the field indicates an ex-




panded use of the drum-mix process.




       This source can be typified as a growing one with relatively young




process and control experience.




       Although uncontrolled emissions are less than those from a conven-




tional plant of comparable production capacity, they clearly exceed NSPS.*




These uncontrolled plant emissions also generally exceed 100 tons per year.




Furthermore, this potential of exceeding 100 tons per year is enhanced




because the production capacities of drum-mix plants often exceed those of




conventional plants.




       An analysis of 63 stack tests which included 158 test runs showed




high variability of emissions at each level of control which may be related
  One Teported countercurrent flow plant met NSPS emissions.





                                 46

-------
to a variety of process designs marketed by different manufacturers and




continuing refinements in design, in addition to the normal causes of such




variations such as raw material, and condition of a given type of control.




Control results improved from mechanical (14% meeting NSPS) through wet




scrubber (29%) to venturi scrubbers (50%).   Baghouse and precipitator con-



trols were not encountered in this study.




       The ratio of "condensibles" to the "front half" catch averaged




35% for uncontrolled plants and decreased to 10% for venturi scrubbers.




This means that a considerable amount of emissions are not measured by




Method 5 under the NSPS since the back half of the measurement train that




reports condensibles is not used.  Secondly it may mean that water pollution




problems could be encountered in a drum-mix plant using wet collection




techniques unless appropriate ponds or closed loop recycling is used.








Recommendations








       Better primary data are needed if firm, accurate information on




emissions and control efficiency on drum-mix plants is desired.  This pre-




liminary study pointed out the shortcomings of relying on published test




reports.  To provide meaningful primary data it is necessary that future




reports include more data on the model and type of drum-mix plant, details




on the control device(s) including as a minimum design feature, pressure




drops, and liquor flow.  Information in the test reports as to the type




of raw material and the size gradation should also be included.




       We would recommend a two-pronged approach in developing these data:
                                  47

-------
       1.    New drum-mix plants being tested by either the federal  govern-
            ment or the states should be provided with data sheets  or asked
            to report the data noted above.   If convenient, federal rep-
            resentatives should witness the  tests
       2.    An effort should be launched to  develop better data from the
            tests already reported.   While it is difficult if not impossible
           .to identify some factors that held during the test such as
            pressure drop, liquor rate, and  raw material gradation, it may
            be possible to obtain better information on the type drum-mix
            and a better description of the  controls, and the design parameters
            of the control system.

       There is significant current  R§D effort by manufacturers in  improving
the product and air'pollution characteristics of the drum-mix process.
(One goal  is to make fabric filters  feasible).  This indicates the  need for .
continuing attention by EPA as these R§D efforts are reflected in equipment
and emission changes.
       Method 5 testing of drum-mix  asphalt  concrete plant emissions is
feasible with minor modifications to the sampling train.
                                 48

-------
                              REFERENCES
Number
             R. W. Beaty and B. M.  Bunnel  "The Manufacture of Asphalt
             Concrete Mixtures in the Dryer Drum" presented at the annual
             meeting, Canadian Technical Asphalt Association,  November 1973.

             "Asphalt Hot-Mix Emission Study".   The Asphalt Institute,
             Research Report 75-1,  March 1975.

             "Background Information For New Source Performance Standards".
             EPA, 450/2-74-003 (APTD-1352C) February 1974.

             "Inspection Manual For Enforcement of New Source  Performance
             Standards:  Asphalt Concrete Plants", Contract 68-02-1356,
             Task 2, JACA Corp., June 1975.

             "Air Pollution Engineering Manual".  EPA, AP 40,  2nd Edition,
             May 1973

             "Group Buying to Reduce Air Pollution Costs for Small Plants",
             JACA Corp. for Conservation Foundation, August 1972.

             "Study of Selected Potential Problem Areas in the NSPS Sur-
             veillance of the Asphalt Concrete Industry", JACA Corp.,  for
             EPA, March 1976.
                                 49

-------
                     Appendix A

         MANUFACTURERS OF DRUM-MIX PLANTS
Aedco, Inc.
13333 U.S. Highway 24 West
Fort Wayne, Indiana  46804

Astec Industries, Inc.
P.O. Box 2787
Chattanooga, Tennessee  37407

Barber-Greene Company
Aurora, Illinois  60507

Boeing Construction Equipment Co.
P.O. Box 3707
Seattle, Washington

CMI Corporation
P.O. Box 1985
Oklahoma City, Oklahoma  73101

Iowa Manufacturing Company
Cedar Rapids, Iowa  52401

Stansteel Corporation
5001 S. Boyle Avenue
Los Angeles, California  900.58

Portec, Inc.
Minneapolis, Minnesota  55414
                                          Type of Flow
Counter
Parallel
Parallel
Parallel
Parallel
Parallel
Parallel
Parallel
                        A-l

-------
                              Appendix B




                      SAMPLING TRAIN MODIFICATION








       Desiccate a quantity of Pyrex glass wool for 48 hours.   Using large




tongs, loosely pack the top of the glass filter holder with the glass wool.




Remove and weigh the wool to a constant weight.  Repack the glass wool into




the filter holder top and assemble the filter holder.




       During the sample recovery procedures in the laboratory, remove the




glass wool from the filter holder and place it on clean, tared pyrex dish.




Desiccate the glass wool for 48 hours and re-weigh it  to a constant weight.




Include the net weight gain in calculating the particulate emission rate.




       If any tarry residue is trapped on the filter holder top, rinse it




with methylene chloride and include the washings with  the probe and nozzle




washes.
                                  A-2

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.

   EPA-340/1-77-004
                              2.
                                                            3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE

  Preliminary Evaluation of Air Pollution  Aspects of the
  Drum-Mix Process
                5. REPORT DATE
                  Issue:  March  1976
                6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS

  JACA Corp.
  506  Bethlehem Pike
  Fort Washington, PA  19034
                                                            10. PROGRAM ELEMENT NO.
                11. CONTRACT/GRANT NO.

                    68-02-1356
12. SPONSORING AGENCY NAME AND ADDRESS
 Environmental Protection Agency
 Division of Stationary Source Enforcement
 Washington,  DC  20460
                                                            13. TYPE OF REPORT AND PERIOD COVERED
                                                                Final
                14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
  This report focuses  on the air pollution aspects of a process of recent practical
  application in the asphalt concrete  industry, called the  Drum-Mix process.   In-
  cluded in this report  is a description  of the drum-mix process,  factors affecting
  its  use in new asphalt concrete plant construction, its air emission potential,
  and  applicable emission control techniques.   Data from emission tests on uncon-
  trolled and controlled drum-mix plants  are analyzed, and  emission factors for
  various levels of control are reported.   Also included in this report is a dis-
  cussion on the ways  to overcome sampling problems particular to the drum-mix
  exhaust.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                              c. COS AT I Field/Group
  Asphalt Concrete  Plants
  Air Pollution Control
    Emission  Factors
    Emission  Testing
    Drum-Mix  Process
13. DISTRIBUTION STATEMENT

  Release unlimited
   19. SECURITY CLASS (ThisReport)
     Unclassified
21. NO. OF PAGES
   55
                                               20. SECURITY CLASS (Thispage)
                                                 Unclassified
                                                                          22. PRICE
EPA Form 2220-1 (9-73)
B-l

-------