MRI BS REPORT
                        Urban Fugitive Dust Test Needs

                                            Final Report
                              For Air and Energy Environmental
                                 Research Laboratory (MD-62)
                          U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina 27711

                                       Attn: Charles Masser
                                 EPA Contract No. 68-DO-0137
                                           Assignment No. 8

                                    MRI Project No. 9800-A(8)


                                         September 30, 1991
MIDWEST RESEARCH INSTITUTE 425 Volker Boulevard, Kansas City, MO 64110-2299 • (816) 753-7600

-------
                              MRIBS REPORT
                        Urban Fugitive Dust Test Needs

                                            Final Report
                             For Air and Energy Environmental
                                 Research Laboratory (MD-62)
                          U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina 27711

                                       Attn: Charles Masser
                                EPA Contract No. 68-DO-0137
                                          Assignment No. 8

                                    MR! Project No. 9800-A(8)


                                        September 30, 1991
MIDWEST RESEARCH INSTITUTE 425 Volker Boulevard, Kansas City, MO 64110-2299 • (816) 753-7600

-------
                              DISCLAIMER
      This document has not been released formally by the Office of Toxic
Substances, Office of Pesticides and Toxic Substances, U.S. Environmental
Protection Agency.  It is being circulated for comments on its technical merit and
policy implications.  The use of trade names for commercial products does not
constitute Agency endorsement or recommendation for use.

-------
                                PREFACE


      This report was prepared for the Air and Energy Environmental Research
 Laboratory (AEERL), U.S. Environmental Protection Agency (EPA), under EPA
 Contract No. 68-DO-0137, Assignment No. 8, which is with the Field Studies
 Branch of the Office of Toxic Substances.  Mr. Charles Masser was the requester
 of the work. The report was prepared by  Mr. John S. Kinsey, Dr. Gregory
 Muleski, Ms. Mary Ann Grelinger, and Dr.  Chatten Cowherd, Jr.

 Approved for:

 MIDWEST RESEARCH INSTITUTE
 Paul C. Constant
 Program Manager
September 30, 1991
MRI-OTS\FB600.FNL                                                              III

-------
                                 CONTENTS
 Preface 	iii
 Executive Summary  	ES-1

       1.     Introduction 	   1
       2.     Urban Fugitive Dust Sources  	   3
                   2.1   Open dust sources	   3
                   2.2   Process fugitive sources	   5
       3.     Review of Current Emission Factors 	   7
                   3.1   Generic emission factors for open dust sources  .   7
                   3.2   Uncontrolled emission factors for paved
                         and unpaved roads  	   9
                   3.3   Uncontrolled emission factors for material
                         storage and handling	  16
                   3.4   Uncontrolled emission factors for agricultural
                         operations	  22
                   3.5   Uncontrolled emission factors for construction
                         activities	  23
                   3.6   Uncontrolled emission factors for miscellaneous
                         open dust sources	  24
                   3.7   Uncontrolled emission factors for process
                         sources	  26
       4.     Preparation of Emission Inventories	  27
                   4.1   Emission inventory methodology	  27
                   4.2   Spatial/temporal distribution of correction
                         parameters	  31
       5.     Future Testing Needs	  32
                   5.1   Improvements in existing emission factors	  32
                   5.2   New emission factors	  35
                   5.3   Control efficiency determination  	  39
       6.     Proposed Test Matrix	  47
References	  53

Appendices
       A.     Emission factors for process fugitive sources  	A-1
       B.     General measurement methods for fugitive dust
             emissions	B-1
       C.     Description of MRI dustiness test chamber  	C-1
MRI-OTS\R9800.FNL

-------
                             LIST OF FIGURES
Number                                                              Page

 5-1   Watering control effectiveness for unpaved travel surfaces	   42
 5-2  Emissions quantification requirements for performance evaluation
      of capture/collection system	   45
                             LIST OF TABLES

Number                                                              Page

 2-1   Generic categories of open dust sources 	    4
 2-2   Open dust sources associated with  	    5
 3-1   Correction parameters for AP-42 emission factor models	    8
 3-2   Selection of paved road emission factor  	   10
 3-3   Typical silt content and loading values for paved roads at
      industrial facilities	   11
 3-4   Summary of silt loadings (sL) for paved urban roadways	   13
 3-5   Paved urban roadway classification	   13
 3-6   Typical silt content values of surface material on industrial
      and rural unpaved roads	   15
 3-7   Typical silt and moisture content values of materials at
      various industries	   18
 3-8   Threshold friction velocities—industrial aggregates	 .   20
 3-9   Threshold friction velocities—Arizona sites	   21
 3-10 Emissions increase (AB) by  site traffic volume	   25
 3-11  Emissions increase (AB) by  construction  type 	   26
 6-1   Tentative test matrix for urban fugitive dust sources	   48
VJ                                                                MRI-OTS\R9800.FNL

-------
                           EXECUTIVE SUMMARY
       Regulatory plans for attainment of the National Ambient Air Quality
 Standard for PM10 (i.e., particles < 10 /urn in aerodynamic diameter) require
 reliable emission estimating methods. Significant research needs currently exist
 with respect to emission factors for urban fugitive dust sources that impact
 significantly on PM10 nonattainment areas.  The bulk of the field test data
 supporting currently available emission factors was obtained in the 10-year
 period beginning in 1972.

       This document reviews the reliability of current emission factors and
 inventories and makes recommendations for cost-effective field testing programs
 designed to address the most significant needs.  Emphasis is placed on opened
 dust sources as opposed to process sources of fugitive dust emissions.  Recent
 emission inventories of. urban dust sources show that road traffic , construction,
 and in some cases agriculturally related sources are  often the predominant
 contributors to PM10 emissions in urban areas.

       The reliability of currently available emission factors may be limited for
 several reasons including:

             Insufficient number of source tests
             Limited range of tested parameter values
       •      No directly  applicable source tests, necessitating use of emission
             factors for similar source(s)
       •      No directly  or indirectly applicable source tests, necessitating an
             engineering estimation (e.g., "gap filling" approach)  of emission
             factors

       Because of the  substantial variability of fugitive dust emissions,
 single-valued emission factors are generally inappropriate. Therefore,  emission
 factors for fugitive dust sources have been formulated as predictive equations
 that can  be adjusted to source-specific conditions. Correspondingly larger
 bases  of test data are needed to support the development of predictive emission
factor equations, as compared to single-valued emission factors.

       In some cases,  enlarging the test data base by combining the data for two
 (or more) similar operations may produce a more reliable emission factor
equation representing  both operations.  This was found  to be the case for
MRI-OTS\R9800.FNL
                                                                       ES-1

-------
 materials handling emission factor created by combining the data bases
 supporting the original "batch drop" and "continuous drop" equations. For
 example, one recommendation in this report is that the data bases supporting
 the urban and industrial paved road emission factors be explored in this manner.

       Another significant source of uncertainty in the preparation of urban
 emission inventories is the lack of correction parameter data. This is the case for
 urban paved roads and construction activities. Although a relationship between
 "background" paved road silt loading and average daily traffic has been
 developed, for example, there are many localized conditions such as
 construction  site trackout that may elevate the silt loading by more than an order
 of magnitude above background. Isolating such paved  road segments is critical
 not only to reliably estimating uncontrolled emissions, but also to developing
 effective control strategies.

       A number of testing needs are identified in this report for paved roads,
 construction/demolition, and agriculture. The highest (i.e., A) priority testing
 needs identified in the study include:

             Expansion of paved road silt loading data base to  develop
             corrections with land use, etc.; and

             Development of new emission factors for controlled and
             uncontrolled mud/dirt carryout

       Table 6-1 provides detailed information on these and other tests needed
to improve existing PM10 emission inventories in urban areas.
ES-2                                                             MFU-OTS\R9800.FNL

-------
                                SECTION 1

                              INTRODUCTION
       As more local, state, and regional agencies are required to submit
 regulatory plans for attainment of the National Ambient Air Quality Standard
 (NAAQS) for PM10 (i.e., particles < 10 Mm in aerodynamic diameter), methods for
 accurately determining the emissions from fugitive dust sources will be crucial.
 Consequently, the U.S. Environmental Protection Agency  (EPA) is evaluating the
 methodologies used to inventory these sources in order to provide guidance to
 other regulatory agencies.  In addition, appropriate research and development
 goals for enhancing  existing fugitive dust emission estimation techniques need to
 be identified and prioritized and new factors developed where necessary.

       In this document, Midwest Research Institute (MRI) reviews the state-of-
 the-art with respect to the estimation of fugitive PM10 emissions in urban areas.
 Major urban fugitive  sources are identified in Section 2, current fugitive emission
 factors are reviewed in Section 3, and current methodologies used to inventory
 these sources are discussed in Section 4. In these sections, the emphasis is
 placed on open rather than process sources of fugitive emissions. Future
 research needs are identified in Section 5 and a proposed test matrix is provided
 in Section 6.  The research needs and associated test matrix provided in
 Sections 5 and 6 are based on those  sources or control methods that MRI
 believes to be of greatest importance  in the preparation of urban emission
 inventories and that exhibit the highest level of cost effectiveness. Finally,
 references cited in the report are listed in Section 7.
MRI-OTS\R9800.FNL

-------
                                SECTION 2

                     URBAN FUGITIVE DUST SOURCES
       Fugitive emissions refer to those air pollutants that: (a) enter the
 atmosphere without first passing through a stack or duct designed to direct or
 control their flow; or (b) leak from ducting systems. Sources of fugitive
 particulate emissions may be separated into two broad categories:  open dust
 sources; and  process sources. Of the two, open dust sources are emphasized
 in this document.
 2.1  OPEN DUST SOURCES

       Open dust sources produce fugitive emissions of solid particles when the
 forces of wind or machinery act on exposed materials. Open dust sources
 include industrial emissions associated with the open transport, storage, and
 transfer of raw, intermediate, and waste aggregate materials, and nonindustrial
 sources such as unpaved roads and parking lots, paved streets and highways,
 heavy construction activities, and agricultural tilling.  Generic categories of open
 dust sources are listed in Table 2-1 (Cowherd and Kinsey, 1986).
       The partially enclosed storage and transfer of materials to or from a
process operation do not fit well into either of the two categories of fugitive
particulate emissions defined above. Examples are partially enclosed conveyor
transfer stations and front-end loaders operating within buildings. Nonetheless,
partially enclosed materials handling operations are usually classified as open
sources (Cowherd and Kinsey, 1986).

       The various types  of open dust sources  listed in Table 2-1 can be found
either in an industrial facility or in the public sector.  The mechanisms of dust
formation  and thus the type of controls that can be applied in either case are
essentially the same.  However, both the suitability and cost-effectiveness
associated with a specific control measure can change significantly when applied
in an industrial setting as compared to the same control used for public sector
sources.  Therefore, the control strategies used by public agencies often differ
from those employed by  industrial concerns.
MRI-OTS\R9800.FNL

-------
       Table 2-1.  GENERIC CATEGORIES OF OPEN DUST SOURCES
                  Unoaved Travel Surfaces
                  •  Roads
                  •  Parking lots and staging areas
                  •  Storage piles
                  Paved Travel Surfaces
                  • Streets and highways
                  • Parking lots and staging areas
                  Exposed Areas (wind erosion)
                  • Storage piles
                  • Bare ground areas
                  Materials Handling
                  • Batch drop (dumping)
                  • Continuous drop (conveyor, transfer, stacking)
                  • Pushing (dozing, grading, scraping)
                  • Agricultural tilling
      A number of sources are perceived as single sources when in actuality
they are a series of different dust generating operations confined to the same
locality.  Examples of this type of source include construction and demolition
activities, both of which involve dust generation by various material handling
operations as well as vehicular traffic. Table 2-2 lists the specific sources
associated with construction and demolition activities using the same general
notation indicated in Table 2-1.

      One final source worthy of note is agricultural operations. Agricultural
tilling encompasses those activities associated with soil preparation and soil
maintenance.  Crop harvesting activities also frequently involve fugitive dust
emissions.  The emissions from these operations are generally significant but
usually are not controlled except by work practice modifications. Although
agricultural operations normally are not found in urban areas, transport of
agricultural emissions to urban areas may be significant in some cases.
                                                                 MRI-OTS\R9800.FNL

-------
           Table 2-2. OPEN DUST SOURCES ASSOCIATED WITH
                     CONSTRUCTION AND DEMOLITION
             1.
Construction Sites
• Vehicular traffic on unpaved surfaces
• Storage piles
• Mud/dirt carryout onto paved travel surfaces
• Exposed areas
• Batch drop operations
• Pushing (earth moving)
                   Demolition Sites
                   • Vehicular traffic on unpaved surfaces
                   • Storage piles
                   • Mud/dirt carryout onto paved travel surfaces
                   • Exposed areas
                   • Batch drop operations
                   • Pushing (dozer operation)
                   • Blasting
 2.2  PROCESS FUGITIVE SOURCES

       Process sources of fugitive emissions are those associated with industrial
 operations that alter the chemical or physical characteristics of a feed material.
 Examples are emissions from charging and tapping of metallurgical furnaces and
 emissions from crushing of mineral aggregates.  Such emissions normally occur
 within buildings and, unless captured, are discharged to the atmosphere through
 forced or natural draft ventilation systems.  However, a process source of fugitive
 emissions can also occur in the open atmosphere (e.g., scrap metal cutting).
 Process sources of fugitive particulate emissions typically found in urban areas
 are listed by industry in Appendix A (EPA,  1990).

      Examples of industrial facilities that contain fugitive PM10 sources include
 asphalt plants, cement plants, mineral processing facilities, concrete ready-mix
 plants, power plants, quarries and rock products plants, slag  processing plants,
 steel mills, and grain transfer terminals.
MRI-OTS\R9800.FNL

-------
                                SECTION 3

                REVIEW OF CURRENT EMISSION FACTORS


       The reliability of currently available emission factors may be limited for a
 variety of reasons including:

       •      Insufficient number of source tests

       •      Limited range of tested parameters

       •      No directly applicable source tests, necessitating the use of
             emission factors for a similar source(s)

             No directly or indirectly applicable source tests, necessitating an
             engineering estimation of emission factors

       The following section provides  a brief overview of emission factors
 currently available for urban fugitive dust sources and their limitations. Although
 open sources are  emphasized, emission factors for both types of fugitive
 sources are included.
3.1  GENERIC EMISSION FACTORS FOR OPEN DUST SOURCES

      Emission factors have been published for a variety of open dust sources,
either in AP-42 or elsewhere.  These sources include:  paved and unpaved
roads, materials storage and handling, agricultural tilling, and a number of
miscellaneous sources related to construction and demolition activities.

      Open source emission  factors share many common features.  For
example, models are formulated as empirical expressions that relate variations in
the emission factor (e) to differences in the physical properties (p)  of the material
being disturbed and the mechanical energy (m) responsible for the generation of
paniculate according to the general form:
MRI-OTS\R9800.FNL

-------
                                e = Kpam
(3-1)
      As empirical models, open dust source emission factors have adjustable
 coefficients (K,a,b) that reflect relationships determined from actual open dust
 source testing.  For example, Table 3-1 provides the correction parameters used
 in the emission factor models  published in AP-42 for common open dust
 sources.
              Table 3-1.  CORRECTION PARAMETERS FOR AP-42
                        EMISSION FACTOR MODELS'
Source category
Unpaved roads
Paved roads
Exposed areas
Materials handling
Model parameter
Silt content of a surface material
Mean vehicle speed
Mean vehicle weight
Mean number of wheels
Number of wet days/year
Silt content of surface material
Total surface dust loading
Number of disturbances/year
Erosion potential
Mean wind speed
Material moisture content
Units of measure
Weight %
km/h (mph)
Mg or 1 06 g (tons)
Dimensionless
Dimensionless
Weight %
kg/km2 (Ib/mi2)
Dimensionless
g/m2
m/s (mph)
Weight %
   From EPA. 1990.
      The emission factor equations for open fugitive dust sources were
developed through field testing of representative sources for each source
category.  Field samples were taken to determine emission quantities, surface
material properties (e.g., silt loadings), and appropriate meteorological
conditions.  The data were then analyzed through linear regression for
correlation of emissions with the applicable correction parameters to formulate
each equation.  Independent validation was not possible due to the limited data.
Instead, each equation was cross validated.

      Application of the AP-42 equations is limited though, due to the small data
b:-^e used for each equation.  The equations retain their quality ratings only if
applied within the ranges of parameters used to develop the equation. Use of
the equations for extrapolation outside the ranges specified will result in a lower
8
                                                                 MRI-OTS\R9800 FNL

-------
 quality rating (reduced by one level) of the estimated emissions. Also, to retain
 the quality rating of the equations applied to specific sources, it is necessary that
 reliable correction parameter values for the specific source in question be
 determined.  In the  event that site specific values cannot be obtained, the
 appropriate default  (mean) values may be used, but the quality rating should be
 reduced by at least one level.  The following sections describe the emission
 factor models for open dust sources in more detail.
 3.2  UNCONTROLLED EMISSION FACTORS FOR PAVED AND UNPAVED
     ROADS

       Particulate emissions occur whenever a vehicle travels over a paved
 surface, such as public and industrial roads and parking lots. These emissions
 may originate from material previously deposited on the travel surface, or
 resuspension of material from tires and undercarriages. In general, emissions
 arise primarily from the surface material loading (measured as mass of material
 per unit area), and that loading is in turn replenished by other sources (e.g.,
 pavement wear, deposition of material from vehicles, deposition from other
 nearby sources, carryout from surrounding unpaved areas, and litter).  EPA's
 Compilation of Air Pollutant Emission Factors  (AP-42) indicates that the PM10
 emission factors for paved roads may be written in the general form (EPA,
 1990):
                                e  = A
(3-2)
where A, B, and c are constants and:

       e = PM10 emission factor, mass/vehicle/length of road segment
       s = fractional surface silt content, weight %
       L = total surface loading, mass/area

Surface silt content (s)  is the fraction of material smaller than 200 mesh or 75
in physical diameter. The parameter L is the total dust loading, and the product
sL represents the mass of silt-size dust particles per unit area of the road
surface. As is the case for all predictive models in AP-42, the use of site-specific
values of sL is strongly recommended.  The silt loading should be determined
using the sampling techniques and laboratory analysis procedures described in
Appendix  D of AP-42.
MRI-OTS\R9800.FNL

-------
       Selection of the appropriate emission factor model (i.e., the constants A,
 B, and c) for a given road depends upon:  the value of the silt loading (sL) and
 the average weight of the vehicles traveling on the road.  Table 3-2 describes the
 selection process for paved road emission factors.  It should be noted that for
 purposes of preparing an emissions inventory, this equation would be applied to
 each road segment in an industrial facility or urban area, as the case may be.
 Ideally a road segment is characterized by a narrow range of silt loading.
          Table 3-2. SELECTION OF PAVED ROAD EMISSION FACTOR
Silt loading (sL)
g/m2
sL < 2
sL < 2
sL>2d
2 < sL < 15
sL> 15"
oz/yd2
< 0.06
< 0.06
> 0.06
0.6 < sL <
0.44
> 0.44
Range weight (W)
Mg
W> 4
W < 4
W> 6
W < 6
W< 6
Ton
> 4.4
< 4.4
> 6.6
< 6.6
< 6.6
Applicable PM10 emission factor
g/VKT
220(sL/12)°-3b
2.28 (sL/0.5)°flC
220(sL/12)°3b
220(sL/12)°3b
93
Ib/VMT
0.78 (sL/O.SS)03"
0.0081 (sL/O.OIS)08'
0.78 (sL/0.35)03"
0.78 (sL/0.35)03"
0.33
 •  VKT = Vehicle kilometers traveled, VMT = vehicle miles traveled.
 b  Commonly referred to as the "industrial" paved road model.
 0  Commonly referred to as the "urban" paved road model.
 d  For heavily loaded surfaces [i.e., sL > ~ 300 to 400 g/m2 (9 to 12 oz/yd2)], it is
    recommended that the resulting estimate be compared to that from the unpaved road
    models and the smaller of the two values used.

      The emission rate is determined by multiplying the emission factor of each
road segment by the total vehicle mileage accumulated on the segment over the
averaging time of interest. Totaling the individual emission rates for each road
segment will provide an uncontrolled emission rate for all paved roads.

      The industrial paved road equation (Table 3-2) was developed for
medium and heavy duty vehicles traveling on dry roadways  in 10 industrial sites
(Table 3-3).  The equation (A-rated) was developed from samples ranging as
follows:

            Silt Loading: 2 to 240 g/m2
      •     Mean Vehicle Weight: 6 to 42 Mg

AP-42 also gives a default emission factor of 93 g/VKT with a C rating for use if
silt loadings are heavy.
10
                                                                 MRI-OTS\R8800.FNL

-------
      Table 3-3.  TYPICAL SILT CONTENT AND LOADING VALUES FOR PAVED ROADS AT INDUSTRIAL FACILITIES'
Industry
Copper smelting
Iron and steel
production
Asphaltic concrete
Concrete batching
Sand and gravel
processing
No. of
sites
1
6
1
1
1
No. of
samples
3
20
3
3
3
Silt (wt. %)
Range
15.4-21.7
1.1 -35.7
2.6 - 4.6
5.2 - 6.0
6.4 - 7.9
Mean
19.0
12.5
3.3
5.5
7.1
No. of
travel
lanes
2
2
1
2
1
Total loading x 103
Range
12.9-19.5
45.8 - 69.2
0.006 - 4.77
0.020- 16.9
12.1 - 18.0
43.0 - 64.0
1.4- 1.8
5.0 - 6.4
2.8 - 5.5
9.9-19.4
Mean
15.9
55.4
0.495
1.75
14.9
52.8
1.7
5.9
3.8
13.3
Units"
kg/km
Ib/mi
kg/km
Ib/mi
kg/km
Ib/mi
kg/km
Ib/mi
kg/km
Ib/mi
Silt loading (g/m2)
Range
188-400
0.09 - 79
76-193
11 -12
53-95
Mean
292
12
120
12
70
      From EPA, 1990.
b     Multiply entries by 1,000 to obtain stated units.

-------
       Although public roads tend to have lower surface loadings than industrial
 roads, the fact that these roads have far greater traffic volumes may result in a
 substantial contribution to the measured air quality in certain areas.  In addition,
 many public roads in industrial areas often are heavily loaded and traveled by
 heavy vehicles. In that instance, better emission estimates would be obtained by
 treating these roads  as industrial roads. In an extreme case, a road or parking
 lot may have such a high surface loading that the paved surface is essentially
 covered and is easily mistaken for an unpaved road.

       The emission factor for public paved roads is determined by the decision
 rule discussed above.  Tables 3-4 and 3-5 present summaries of silt loadings as
 a function of roadway classification and the scheme used in AP-42 to classify
 roadways, respectively (EPA, 1990).

       The urban  paved road equation (Table 3-2) was developed  from a base of
 44 emission tests and corresponding silt loadings obtained in five different cities
 (Table 3-4).  The test sites represent a broad range of Eastern and Midwestern
 urban  areas.  However, no correlation with land use or road condition was
 possible. The sampling locations are generally representative of many large
 population centers, except those in the  Southwest. It is also noted that early
 spring loadings may be five or six times higher than the mean loading, due to
 winter  road sanding. Thus, the equation should be used with caution in the
 Southwest and in areas with high winter sand application.

      The range  of silt loadings used in the urban paved road equation is 0.022
 g/m2 to 2.11 g/m2 (Table 3-4).  AP-42 recommends PM10 emission factors based
 on default values  (the mean loadings tested)  if no site specific silt loadings are
 available. These default values, which are typically used in many urban PM10
 emission inventories  in place of site-specific data, are as follows:

            Local streets (< 500 ADT): 5.2 g/VKT

            Collector streets (500-10,000 ADT): 3.7 g/VKT

            Major streets (> 10,000 ADT): 1.8 g/VKT

            Freeways (> 50,000 ADT): 0.21  g/VKT
12
                                                                 MRI-OTS\R9800.FNL

-------
                 Table 3-4. SUMMARY OF SILT LOADINGS (sL) FOR
                           PAVED URBAN ROADWAYS*

City
Baltimore
Buffalo
Granite City,
IL
Kansas City
St. Louis
All
Local streets
Mg/m8)
1.42
1.41
—
—
—
1.41
n
2
5
—
—
—
7
Collector
streets
*0
(g/m2)
0.72
0.29
—
2.11
—
0.92
n
4
2
—
4
—
10
Major
streets/highways
Xgte/m2)
0.39
0.24
0.82
0.41
0.16
0.36
n
3
4
3
13
3
26
Freeways/
expressways
^(g/m2)
_
—
—
—
0.022
0.022
n
—
—
—
—
1
1
   From EPA, 1990. Xg = geometric mean based on corresponding n sample size.  Dash =
   not available. To convert g/m2 to grains/ft2 multiply g/m2 by 1.4337.
               Table 3-5. PAVED URBAN ROADWAY CLASSIFICATION*
Roadway category
Freeways/expressways
Major streets/highways
Collector streets
Local streets
Average daily traffic (vehicles)
> 50,000
> 10,000
500-10,000
< 500
No. lanes
.> 4
> 4
2b
2°
         * From EPA, 1990.
         b Road width > 32 ft.
         0 Road width < 32 ft.
      As is the case for paved roads, particulate emissions occur whenever a
vehicle travels over an unpaved surface.  Unlike paved roads, however, the road
itself is the source of the emissions rather than any "surface loading." Within the
various categories of open dust sources  (i.e., paved roads, storage piles, and
wind erosion), unpaved travel surfaces have historically accounted for the
greatest share of particulate emissions in industrial settings.

      Unpaved travel surfaces are found in rural regions throughout the country,
in industrial settings, and in some limited  cases, urban areas. Roads may be
unpaved because they experience only sporadic traffic and involve considerable
road length, making paving generally impractical. Some industrial roads are, by
MRK3TS\R9800.FNL
                                                                          13

-------
their nature, not suitable for paving.  For example, some roads may be traveled
by very heavy vehicles or subject to considerable spillage from haul trucks.
Other roads may have poorly constructed bases which make paving impractical.
Because of the additional maintenance costs associated with a paved  road
under these service conditions, emissions from these roads are usually
controlled by regular applications of water and/or chemical dust suppressants.

      In addition to roadways, many industries often contain unpaved travel
areas such as around stockpiles, and staging areas. These areas may often
account for a substantial fraction of traffic-generated emissions from industrial
facilities. In addition, these areas tend to be much more difficult to control than
stretches of roadway.

      As was the case for paved roads, unpaved roads may be divided into the
two classes of public and industrial.  However, for the purpose of estimating
emissions,  there is no need to distinguish between the two, because the AP-42
emission factor equation takes source characteristics (such as average vehicle
weight and road surface texture) into consideration (EPA, 1990):
                   e , o.6i  jL             "fi              (3-3)
where:

      e = PM10 emission factor, kg/VKT
      s = silt content of road surface material, percent
      S = mean vehicle speed, km/h
      W = mean vehicle weight, Mg
      w = mean number of wheels (dimensionless)
      p = number of days with > 0.254 mm (0.01 in.) of precipitation per year

      The above equation is rated "A" in AP-42, if used within the range of
correction parameters tested.  As is the case with all AP-42 emission factors, the
use of site-specific data is strongly encouraged.

      Table  3-6 gives the measured silt values for unpaved roads in certain
industries and rural areas. The silt content of the  road surface material should
be determined using the sampling techniques  and laboratory analysis
procedures described in Appendix D of AP-42 (EPA, 1990).
14
                                                                 MRKrrsweoo.FNi

-------
            Table 3-6.  TYPICAL SILT CONTENT VALUES OF SURFACE MATERIAL ON INDUSTRIAL AND RURAL UNPAVED ROADS'
Industry
Copper smelting
Iron and steel production
Sand and gravel processing
Stone quarrying and processing
Taconite mining and processing
Western surface coal mining
Rural roads
Road use or surface
material
Plant road
Plant road
Plant road
Plant road
Haul road
Service road
Access road
Haul road
Scraper road
Haul road
(freshly graded)
Gravel
Dirt
Crushed limestone
No. of sites
1
9
1
1
1
1
2
3
3
2
1
2
2
No. of test
samples
3
20
3
5
12
8
2
21
10
5
1
5
8
Silt (wt percent)
Range
15.9-19.1
4.0-16.0
4.1-6.0
10.5-15.6
3.7-9.7
2.4-7.1
4.9-5.3
2.8-18
7.2-25
18-29
NA
5.8-68
7.7-13
Mean
17.0
8.0
4.8
14.1
5.8
4.3
5.1
8.4
17
24
5.0
28.5
9.6
          • From EPA, 1990.
          NA = Not applicable.
01

-------
      The number of wet days (i.e., precipitation >. 0.254 mm or 0.01 in) per
 year (p) for the geographical area of interest, can be determined from local
 climatic data.  Figure 11.2.1-1 of AP-42 gives the geographical distribution of the
 mean annual number of wet days per  year in the United States.   Maps giving
 similar data on a monthly basis are available from the U.S. Department of
 Commerce National Climatic Center.

      The emission rate is determined by multiplying the emission factor by the
 length of the road segment and the vehicle passes per day.  Adding the
 emission rate from each segment will provide a total unpaved road emission rate.

      The unpaved road  equation was developed from 103 tests performed on
 various plant, haul, and rural roads (Table 3-6).  The equation was developed for
 calculating average annual emissions.  The ranges of parameters sampled are:

            Silt content:  4.3 to 20%
      •     Mean vehicle weight:  2.7 to 142 Mg
      •     Mean vehicle speed:  21  to 64 km/hr
      •     Mean no. of wheels: 4 to 13
3.3  UNCONTROLLED EMISSION FACTORS FOR MATERIAL STORAGE AND
     HANDLING

      Adding aggregate material to a storage pile or removing it usually involves
the dropping of material onto a receiving surface or receptacle.  This can be
performed by one of two methods:  by continuous feed, as exemplified by
conventional fixed stacker reclaimers; or by batch operations as would be
performed by mobile equipment (i.e., front-end loader, truck,  or pan scraper).

      The following equation is recommended in AP-42 for estimating emissions
from transfer operations (batch or continuous drop):
                                           U
                          e = A-(0.0016)       -                     (3-4)
where:
      e = emission factor, kg/Mg
      k = particle size multiplier (dimensionless) = 0.35 for PM10
16
                                                               MRI-OTS\R9800.FNL

-------
       U = mean wind speed, m/s
       M = material moisture content, percent

       Equation (3-4) shows that, for the handling of aggregate materials,
 emissions increase with the ambient wind speed and are reduced as the material
 moisture content increases.

       The equation was developed from 141 field samples of 12 different
 aggregate materials in the five industries shown in Table 3-7. The ranges of
 parameters used to develop the equation are:

             Silt content:  0.44 to 19%
       •     Moisture content: 0.25 to 4.8%
       •     Wind speed:  0.6 to 6.7 m/s

       To estimate the emissions from the loading of material in and out of
 storage piles, one must first obtain a sample of the stockpiled material and a
 record of historical wind data at the location. From the Local Climatological Data
 (LCD) summaries, the mean wind speed can be obtained for the nearest
 recording weather station.  This wind speed should be adjusted from the
 recording anemometer height to represent the  wind speed at the average
 storage pile height using the procedure described in Section 11.2.7 of AP-42.
 Aggregate sampling and subsequent moisture  analysis should then be
 performed on the sample(s) as described in Appendix D of that  document.

       Although in some instances wind erosion of stockpiled material can be
 a substantial source of fugitive PM10, it is usually considered the  least significant
 of the mechanisms responsible for dust emissions from storage  piles.  In most
 industrial settings, fugitive emissions from vehicular traffic and materials handling
 are far greater than the emissions from wind erosion. If vehicular traffic is
 present in and around the storage pile area, those emissions are far greater than
 emissions due to materials  handling.

       The AP-42 emission factor for wind erosion, as a summation over periods
 between disturbances of the erodible surfaces, can be calculated as:
MFH-OTS\R9BOO.FNL
                                                                         17

-------
                    Table 3-7. TYPICAL SILT AND MOISTURE CONTENT VALUES OF MATERIALS AT VARIOUS INDUSTRIES'
oo
Industry
Iron and steel
production





Stone quarrying and
processing
Taconite mining and
processing
Western surface coal
mining
Coal fired power
generation11
Material
Pellet ore
Lump ore
Coal
Slag
Flue dust
Coke breeze
Blended ore
Sinter
Limestone
Crushed
limestone
Pellets
Tailings
Coal
Overburden
Exposed
ground
Coal
Silt (weight %)
No. of test
samples
10
9
7
3
2
1
1
1
1
2
9
2
15
15
3
60
Range
1.4- 13
2.8- 19
2-7.7
3-7.3
14-23
—
—
—
—
1.3- 1.9
2.2 - 5.4
NA
3.4-16
3.8-15
5.1 -21
0.6 - 4.8
Mean
4.9
9.5
5
5.3
18.0
5.4
15.0
0.7
0.4
1.6
3.4
11.0
6.2
7.5
15.0
2.2
Moisture (weight %)
No. of test
samples
8
6
6
3
0
1
1
0
0
2
7
1
7
0
3
59
Range
0.64 - 3.5
1.6-8.1
2.8- 11
0.25 - 2.2
NA
—
—
NA
NA
0.3- 1.1
0.05 - 2.3
2.8 - 20
NA
0.8 - 6.4
2.7 - 7.4
Mean
2.1
5.4
4.8
0.92
NA
6.4
6.6
NA
NA
0.7
0.9
0.35
6.9
NA
3.4
4.5
          • From EPA, 1990.
          h Values reflect "as received' conditions at a single power plant.
          NA = not applicable.  "

-------
                                 e = k     P                           (3'5)
 where:
       e =  PM10 emission factor, g/m2
       k =  particle size multiplier = 0.5 for PM10
       n =  number of disturbances per year
       Pj = erosion potential corresponding to the observed (or probable) fastest
            mile of wind for the period between disturbances, g/m2
 The erosion potential (P) can be calculated as:


                        P = 58 (u*-u;)2 + 25 (u*-ut)                 (3-6)


 where:

       u* = friction velocity (m/s)
       u* = threshold friction velocity (m/s)
       and u* > ut*

       To estimate the particulate emissions from the wind erosion of storage
 piles, one must first obtain a composite surface sample of the stockpiled
 material. Aggregate sampling and determination of the threshold friction velocity
 should be performed according to the procedures described in Appendix D  and
 Section 11.2.7 of AP-42, respectively. From the Local Climatological Data (LCD)
 summaries, the fastest miles of wind can be obtained for the nearest recording
 weather station. Each fastest mile of wind should then be adjusted to represent
 the wind speed that impacts the storage pile surface and an equivalent friction
 velocity determined as described in Section 11.2.7 of AP-42 (EPA, 1990).

       Threshold friction velocities and the associated wind speeds (as measured
 at 10 m) are presented in Tables 3-8 and 3-9 for typical soils and aggregate
 materials, respectively (Cowherd et al., 1988).  Estimates for emissions are
 considered more accurate if site specific data are used.
MFU-OTB\R9800.FNL                                                                1 9

-------
       Table 3-8. THRESHOLD FRICTION VELOCITIES-INDUSTRIAL AGGREGATES
Material
Overburden"
Scoria (roadbed
material)"
Ground coal"
(surrounding coal pile)
Uncrusted coal pile*
Scraper tracks
on coal pile">b
Fine coal dust
on concrete padc
Threshold friction
velocity (m/s)
1.02
1.33
0.55
1.12
0.62
0.54
Roughness
height (cm)
0.3
0.3
0.01
0.3
0.06
0.2
Threshold wind velocity at
10m (m/s)
Z0 =
actual
21
27
16
23
15
11
Z0 = 0.5
cm
19
25
10
21
12
10
 " Western surface coal mine.
 " Lightly crusted.
 c Eastern power plant.
 From Cowherd et al., 1988.
20
                                                                       MW-OTS\R9eoO.FNL

-------
                  Table 3-9.  THRESHOLD FRICTION VELOCITIES-ARIZONA SITES
Location
Mesa - Agricultural site
Glendale - Construction site
Maricopa - Agricultural site
Yuma - Disturbed desert
Yuma - Agricultural site
Algodones - Dune flats
Yuma - Scrub desert
Santa Cruz River, Tucson
Tucson - Construction site
Ajo - Mine tailings
Hayden - Mine tailings
Salt River, Mesa
Casa Grande - Abandoned
agricultural land
Threshold friction
velocity (m/sec)
0.57
0.53
0.58
0.32
0.58
0.62
0.39
0.18
0.25
0.23
0.17
0.22
0.25
Roughness height
(cm)
0.0331
0.0301
0.1255
0.0731
0.0224
0.0166
0.0163
0.0204
0.0181
0.0176
0.0141
0.0100
0.0067
Threshold wind velocity
at 1 0 m (m/sec)
16
15
14
8
17
18
11
5
7
7
5
7
8
From Cowherd et al., 1988.

-------
       The wind erosion equation for storage piles was formulated from field
 testing using a portable wind tunnel. The equation applies only to dry, exposed
 materials with limited erosion potential and is valid only for a time period longer
 than the period between disturbances.
 3.4  UNCONTROLLED EMISSION FACTORS FOR AGRICULTURAL
     OPERATIONS

       Fugitive dust from agricultural operations is suspected of contributing
 significantly to the ambient particulate levels of many urban areas. Such
 agricultural operations include plowing, disking, fertilizing, applying herbicides
 and insecticides, bedding, flattening and firming beds, planting, cultivating, and
 harvesting.  These activities can be classified generically as soil preparation, soil
 maintenance,  and crop harvesting operations.  Dust emissions are also
 generated by  wind erosion of bare or partially vegetated soil.

       The mechanical tilling of agricultural land injects dust particles into the
 atmosphere as the soil is loosened or turned under by plowing, disking,
 harrowing, one-waying, and so on. AP-42 presents a predictive emission factor
 equation (B-rated) for the estimation of PM10 emissions from agricultural tilling
 (EPA,  1990):


                              e = k (5.38)(s)°-e                        (3-?)
where:

      e = emission factor, kg/ha
      k = particle size multiplier (dimensionless) = 0.21 for PM10
      s = silt content (percent) of surface soil (default value of 18 percent)


The above equation is based on limited field testing information cited in AP-42.
Field measurements of mechanical tilling showed that emissions are not
dependent on implement type, speed below 8 to 10 km/hr, or soil moisture.
However,  emissions do vary directly with silt content.  The range for silt content
tested was 1.7 to 88 percent.

      The technique currently used for predicting agricultural wind erosion is
based on  variations of the Wind Erosion Equation developed as the result of
nearly 40 years of research by the U.S. Department of Agriculture to predict total
topsoil losses. This prediction system uses erosion loss estimates that are
22
                                                                  MRK>TS\R9800.FNL

-------
 integrated over large fields and long time scales to produce average annual
 values.

       The modified equation for agricultural wind erosion is of the form
 (Cowherd et al., 1988):


                             e =  kal KC LV                       (3-8)
 where:
       e =   PM10 wind erosion losses of tilled fields, tons/acre/yr
       k =   0.5, the estimated fraction of TSP* which is PM10
       a =   portion of total wind erosion  losses that would be measured as
             suspended particulate (estimated to be 0.025)
       I =    soil erodability, tons/acre/yr
       K =   surface roughness factor, dimensionless
       C =   climatic factor, dimensionless
       L =   unsheltered field width factor, dimensionless
       V =   vegetative cover factor, dimensionless

       The calculation scheme used to derive the various parameters in Equation
 3-8 is quite  complex and thus will not be presented here.  The reader is referred
 to Section 7 of Cowherd et al. (1988) for further details.
 3.5  UNCONTROLLED EMISSION FACTORS FOR CONSTRUCTION
     ACTIVITIES

      As stated in Section 2, construction sites consist of multiple open dust
 sources which include vehicular traffic on unpaved surfaces, storage piles,
 mud/dirt carryout onto paved roads, exposed areas, batch drop operations, and
 earthmoving.  However, only a single valued emission factor (not rated) is
 provided in AP-42.  The single valued AP-42 emission factor of 1.2 tons
 TSP/acre/month for construction sites applies to operations with: (1) medium
 activity level;  (2) silt content about 30%; and (3) semiarid climates. Test data
 were not sufficient to derive a dependence on correction factors.

      In addition to the AP-42 single valued emission factor, PM10 emission
factors have been developed for site preparation activities at a road construction
 project in Minnesota (Kinsey et  al., 1983).  The "gap filling"  PM10  emission factors,
 based on the level of vehicle activity (i.e.,  vehicle kilometers traveled or VKT), are
 (Grelinger et al., 1988):
    *  TSP = Total suspended particulate matter « < 30


MFU-OTS\R9800.FNL                                                               23

-------
            Topsoil removal:  5.7 kg/VKT for pan scrapers
            Earthmoving: 1.2 kg/VKT for pan scrapers
            Truck haulage: 2.8 kg/VKT for haul trucks

The above factors have not been  rated.
3.6  UNCONTROLLED EMISSION FACTORS FOR MISCELLANEOUS OPEN
     DUST SOURCES

      There are a number of PM10 emission factors for miscellaneous open
sources that have been developed especially for the preparation of State Imple-
mentation Plans as "gap filling" techniques. These gap filling emission factors are
generally broad engineering estimates based on extremely limited data or
extrapolation from another type of source.  Therefore, particular caution must be
exercised in the use of these emission factors.  The following information
describes gap filling emission factors for mechanical dismemberment at
demolition sites, demolition debris loading, pushing (bulldozing), and mud/dirt
carryout.

      For demolition sites, the operations involved in demolishing and removing
structures from a site are:

      •     Mechanical or explosive dismemberment
      •     Debris loading
      •     Onsite truck traffic
      •     Pushing (dozing) operations

      For mechanical dismemberment, the AP-42 materials handling emission
factor can be modified for waste tonnage related to structural floor space where
1 m2 of floor space represents 0.45 Mg of waste material (0.046 ton/ft2). The gap
filling emission factor related to structural floor space (using default parameters)
obtained was 0.00025 kg/m2 (Grelinger et al., 1988).

      The emission factor for debris loading is based on two tests of the  filling
of trucks with crushed limestone using a front-end loader, which is part of the
test basis for the AP-42 materials handling  equation. The resulting emission
factor for debris loading is 0.0046 kg/m2 (Grelinger et al., 1988).

      Finally, for pushing (bulldozer) operations at construction and
demolition sites, the AP-42 emission factor equation for overburden removal at
Western surface coal mines can  be used.  Although this equation actually relates
to particulate <15 ^mA, it would be expected that the PM10 emissions from such
operations would be generally comparable. The AP-42 dozer equation is:
24                                                              MRI-OTS\R9800.FNL

-------
                                  _  0.45(s)1-5

                                      W)1'4
(3-9)
 where:

       e =  PM10 emission rate, kg/h
       s =  silt content of surface material, percent (default = 6.9 percent)
       M = moisture content of surface material, percent (default = 7.9 percent)

 and e = 0.45 kg/h (with  default parameters).

       The increase in emissions on paved roads due to mud/dirt carryout has
 been estimated based on surface loading measurements at eight sites (Englehart
 and Kinsey, 1983).  Tables 3-10 and 3-11 provide these emission factors in terms
 of g/vehicle pass which represents  PM10 generated over and above the
 "background" for the paved road tested. Table 3-10 expresses the emission
 factors according to the  volume of traffic entering and leaving the site whereas
 Table 3-11 expresses the same data according to  type of construction.
          Table 3-10. EMISSIONS INCREASE (A£) BY SITE TRAFFIC VOLUME*

Particle
size
fraction1"
<~30 pm
< 10 fim
< 2.5pm
Sites with > 25 vehicle/day
Mean, X
52
13
5.1
Standard
deviation, a
28
6.7
2.6
Range
15-80
4.4-20
1.7-7.8
Sites with < 25 vehicle/day
Mean, X
19
5.5
2.2
Standard
deviation, o
7.8
2.3
0.88
Range
14-28
4.2-8.1
1.6-3.2
   •  A£ expressed in g/vehicle pass (Englehart & Kinsey, 1983).
   b  Aerodynamic diameter.
MRHDTS\R960aFNL
                                                                          25

-------
          Table 3-11. EMISSIONS INCREASE (A£) BY CONSTRUCTION TYPE'

Particle
size
fraction11
< -30 jim
<10^m
<2.5 nm
Commercial sites
Mean, X
65
16
6.3
Standard
deviation, a
39
9.3
3.6
Range
15-110
4.2-25
1.6-9.7
Residential sites
Mean, X
39
10
3.9
Standard
deviation, a
22
5.4
2.1
Range
10-72
2.8-19
1.1-7.3
 ' A£ expressed in g/vehicle pass.
 b Aerodynamic diameter.
3.7  UNCONTROLLED EMISSION FACTORS FOR PROCESS SOURCES

       Fugitive emissions can be generated by many different processes in a
wide variety of industrial facilities.  These emissions are generally associated with
metallurgical operations such as furnace charging and tapping and also include
the dust generated by the processing of stone, wood, grain, and similar mate-
rials. The emissions produced by process sources can have different physical/
chemical characteristics, such as fine metal oxide particles formed by
condensation or solid particulate similar in composition to the parent material.

       Emission factors for process sources are generally expressed as a single
value in terms of mass of pollutant per unit of source operation (e.g., feed or
product).  Appendix A provides the available particulate emission factors
published in AP-42 for process fugitive sources (EPA, 1990).  It should be noted
that sources other than those listed in Appendix A may also exist in urban areas,
but no particulate emission factors are currently available to estimate such
emissions. Moreover, in many cases, available emission factors do not address
the PM10 component.

       Many of the emission factors shown in Appendix A have been assigned a
D or E rating. This indicates that either poor quality data or an engineering
estimate was used to derive the emission factor. It could be concluded,
therefore,  that little real  data exist for many types of process sources  and that
only very limited testing has been conducted.  However, because process
fugitives generally represent a relatively minor contribution  to the total
uncontrolled emissions  from the process, they may not substantially influence the
area-wide  PM10 emissions inventory. Therefore, the remainder of this  document
will  be devoted strictly to open sources.
26
MH1-OTS\H9800.FNL

-------
                                SECTION 4

                PREPARATION OF EMISSION INVENTORIES
       This section discusses current emission inventory techniques associated
 with urban sources of fugitive dust emissions.  These techniques are limited to
 the estimation and spatial distribution of emission factor parameters, and do not
 address determination or apportionment of source extent. Also included in this
 discussion are representative sources of data for the fugitive dust emission
 factors necessary to compile an emission inventory.

       Fugitive dust emissions are estimated  using the applicable emission
 factors (with their input parameters), combined with source extent and control
 efficiencies. Emissions also must be temporally allocated and spatially
 distributed over a grid for modeling purposes, and qualified by particle size.
 Sections 3.1 to 3.6, presented previously, includes a compilation of AP-42 open
 dust source emission factors.

       Typically an emission inventory  plan will be prepared and will include the
 following activities:

       •    Description of the purpose and  type of inventory
       •    Methods to gather source and subsidiary data
       •    Approaches to estimating emissions
            QA Plan
            Final data presentation (including computer files and
            documentation)
4.1 EMISSION INVENTORY METHODOLOGY

      The open dust sources in each grid cell must be identified and coded
according to source type. Records of the local air pollution control agency are
often the starting point for obtaining this information, especially past emission
inventories, even though they may be incomplete.  Occasionally, previously
determined emissions can be adjusted using statistically appropriate scaling
factors to produce current emission estimates.

      The first step is to  identify the study area boundaries and to grid a base
map of the study area. The grid cells may range from 0.5 x 0.5 km to 5 x 5 km,
MRI-OTS\R9800.FNL                                                              27

-------
 depending on the spatial resolution needed. Larger subareas of combined cells
 may be defined within the inventory area for purposes of estimating emissions
 that are relatively consistent across a wide region, such as from wind erosion of
 fallow ground.

       Sources such as landfills should be located by UTM or LATLONG
 coordinates, and then placed into the proper cells.  Sources such as
 construction site activity, road traffic, and agricultural activity are less likely to
 have locational data, and are usually apportioned across an entire city or county
 using surrogate or activity data.

       Emissions from open dust sources are estimated primarily by using the
 factors presented in AP-42, Section 11, Miscellaneous Sources.  This section
 presents emission factors for paved and unpaved roads, aggregate handling,
 wind erosion, and other fugitive dust sources.  Other emission factors for PM10
 from airport runways, burning, construction site preparation, and landfilling  are
 presented in the report "Gap Filling PM10 Emission Factors for Selected Open
 Area Dust Sources" (Grelinger et al., 1988). Two additional EPA documents
 published by Cowherd et al. (1988) and Cowherd and Kinsey  (1986) are also
 helpful in estimating emissions from fugitive dust sources.

      Examples of several sources are discussed to demonstrate how fugitive
 dust emission inventories are conducted. These include three of the highest
 generators of PM10 emissions affecting many urban areas:  road traffic,
 construction activities, and agriculture-related sources.
4.1.1 Road Traffic

      Vehicle traffic on unpaved and paved roads is often the major source of
particulate emissions in urban areas. The procedure to inventory these emis-
sions begins with an analysis of vehicle miles travelled.  State Departments of
Transportation produce either hard copy or computer files of yearly traffic data
on a link by link  (road segment) basis.  Records for each link can contain
starting milepoint, length of subsection, ending milepoint, functional classification,
pavement width, lanes, and average daily traffic (ADT).  The last value is used to
derive the source extent (VMT) which must be classified for paved and unpaved
road functional types. The VMT for each road class is then apportioned into the
gridded cells.
      Silt loading is the only parameter in the paved road emission factor
equation presented in AP-42.  Appropriate silt loading values can be obtained in
two different ways for emission inventory work.  Silt loadings for different paved
road functional classifications  can be obtained from the EPA report "Control of
Open Fugitive Dust Sources"  (Cowherd et al., 1988).  A second option is a
sampling program  to collect representative silt loadings from paved roads in

28                                                                MRI-OTSVRSeOO.FNL

-------
 certain selected cells and apply these values to the remainder of the cells using
 a Krieging technique associated with geo-statistics.

       Emissions per VMT are then calculated for each road classification on a
 per mile basis. The VMT in each cell associated with each functional class is
 then multiplied by the emissions total.

       The parameters of interest are silt contents of unpaved road surfaces,
 vehicle speeds, daily precipitation events per year, and vehicle mean weights
 and wheels.  A series of default parameter values is most often used in emission
 inventory work because of the substantial effort to classify road segments for
 each of the above parameters. The single variable that can be easily obtained
 and that will often be a single value for an entire region is the mean annual
 number of days with rain. This information can be obtained from the National
 Climatic Atlas (and is reprinted in Section  11 of AP-42). The other variables are
 much  more difficult to obtain for each road segment or for each grid cell but can
 be estimated from composite samples and representative field traffic mixes.

       Local government entities that have large computer resources, such as
 GISs, can sometimes access road network files and develop data to assist with
 describing unpaved road segments. In  any  case, the emission inventory effort
 associated with unpaved roads is a particularly labor and computer intensive
 effort because of the large number of parameters associated with the emission
 factor  equation.
4.1.2 Construction

      One approach to estimating construction-related emissions in each cell of
the gridded inventory area is to determine the number of construction permits
issued per month by the governing district, estimate the disturbed acreage per
type of construction site, and apply the AP-42 emission factor of 1.2 tons/mo/
acre disturbed.  In recent urban emissions inventories, MRI has found that
residential, commercial, and industrial construction activities are all seen to be
linearly related to population.  Consequently,  construction emissions from the
governing district can be apportioned across the applicable inventory cells using
population data.

      The figures that can be used for source extent per construction site are:

             Single family residences in high population area:     6/7 acre/mo
             Single family residence in other  area:               6/5 acre/mo
             Multifamily housing (each unit)                      3/10 acre/mo
             Commercial  building (per $ million valuation)        40.7 acre/mo
      •      Industrial building (per $ million  valuation)           44.0 acre/mo
             Institutional building (per $ million valuation)         48.4 acre/mo
MRI-OTSNR9800.FNL
                                                                          29

-------
       The general approach to estimating road construction emissions is to use
the same AP-42 emission factor presented for building construction with the
additional assumptions that:

             The construction of freeways, highways, and city/county roads
             disturbs an area 100, 76, and 64 ft wide, respectively.

             The average period of disturbance is 18 mo, regardless of the type
             of road being constructed.

             All road construction sites are watered to control dust (estimated
             control efficiency of 75%).

       For a specific construction site, the most reliable estimates can be
obtained by considering the separate contributions from  a variety of unit
operations.  These operations could include site preparation, materials handling,
vehicular traffic, mud/dirt trackout onto adjacent paved roads, etc. This
approach will produce a more realistic estimate of emissions, but has three
limitations:

       •     It is very resource intensive.
       •     It requires access to  active construction sites.
       •     The resulting data are not readily extrapolated over the study area.
4.1.3 Agricultural Wind Erosion

      No procedure has been specifically developed for air quality purposes to
predict particulate emissions from wind erosion. Wind blown dust is often esti-
mated by using a modified version of the USDA wind erosion equation that pre-
dicts total soil loss. The source extent is the number of agricultural acres in each
grid cell.  Because the USDA's primary interest is in the loss of farmland, their
equation considers all particles, regardless of size, rather than particle sizes of
interest to air pollution regulatory agencies.  Consequently, the resulting
emissions are then multiplied by an estimated fraction to estimate particulate
matter of the appropriate size  distribution.
4.1.4 Agricultural Tilling

      Currently, the emission factor for tilling emissions consists of an empirical
relationship between emissions and soil silt content. A default silt content value
of 18% is listed in AP-42, Section 11.2.2, and may be used but it will reduce the
quality of the estimate by one level. Silt content values also can be measured in
30                                                                 MRl-OTS\R9800.FNL

-------
 a sampling and analysis program, and would produce a considerably better
 estimate of tilling emissions.

       By utilizing the Soil Conservation Service's computerized soil maps with
 soil horizons classified with size distribution data (but classified by a wet method
 rather than the recommended dry silt content procedure used for dust predic-
 tion), an emission inventory could be somewhat improved over using a default
 sift content for an entire county or region.
 4.2  SPATIAL/TEMPORAL DISTRIBUTION OF CORRECTION PARAMETERS

       Fugitive dust emission factor equations require the following input
 parameters:

       •     Silt content of soil or aggregate
       •     Silt loading on paved surfaces
       •     Moisture content of the disturbed material
       •     Traffic intensity and characterization
       •     Wind speed and precipitation frequency

       In many cases, available input parameter data lack the spatial resolution
 for good emission estimates. Collection of surface dust samples is usually
 desirable to address this need. Or, to improve emission estimates, appropriate
 spatial surrogate data can be used to predict variation in local source conditions.
 A generic procedure for this type of operation begins with spatial data bases that
 contain physical features associated with fugitive dust emissions. These data
 bases could be  used to disaggregate the grossly estimated parameters, such as
 silt loading that results from construction site carryout or salting/sanding for
 apportionment to inventory grids. However, this procedure would require the
 availability of correlations of emission factor parameters  with such parameters as
 population density, soil type, and road classification. Various commercial and
 governmental data bases already are available to begin  this work, such as U.S.
 Census TIGER files and Department of Agriculture soil surveys. Many of the
 relevant data bases can be imported into Geographic Information Systems
 (GISs) for gridded analysis.

      Other efforts associated with  preparation of emission inventories include
the development of sampling strategies and analysis procedures for limited
resource scenarios (dollars, hours, data  bases). Because many cities are finding
it very costly to gather sufficient data on the parent dust for good emission
estimates, efficient sampling methodologies are needed.
MRI-OTSVWaOO.FNL
                                                                         31

-------
                                SECTION 5

                         FUTURE TESTING NEEDS
      Over the past five to ten years, and especially since the 1987
promulgation of the PM10 NAAQS, it has become evident to many regulatory
agencies that nontraditional source control programs may be necessary to attain
the standards.  That is, in addition to the more traditional sources of particulate
matter, attention must now be paid to other industries (e.g., MSW landfills and
road and building construction) as well as public sources such as public roads/
highways, unpaved driveways/parking lots, and landscaping operations. Recent
emission inventories of Lubbock, Tucson, Phoenix, and the Coachella Valley in
California show road traffic, construction, and agriculturally related sources are
often the predominant contributors to PM10 emissions in urban areas.

      Because of the need to examine the newer, nontraditional sources (many
of the emission factor development studies were performed in the mid-1970s to
early 1980s), regulatory agencies have found themselves challenged  in their
attempts to quantify the impacts associated with such sources. The following
discussion outlines research needs to either  improve  or develop emission factors
and improve control efficiency values for open dust sources.
5.1 IMPROVEMENTS IN EXISTING EMISSION FACTORS

      As addressed in Section 3, many fugitive dust sources have unreliable
emission factors.  Additional source measurements are required, therefore, to
improve the emission inventory estimates. The applicability of existing emission
factors to some of the newer sources of interest is often tenuous—for example,
using agricultural tilling emission factors to estimate the impact of lift construction
at landfills.  For other sources, no factors are considered even potentially
applicable.  For some  emission sources studied in the past (such as urban
paved roads), the relatively few available tests make it difficult to determine
variability in the emissions.
32                                                              MRI-OTS\R9800.FNL

-------
 5.1.1  Paved Roads

       Although additional field testing is needed to improve existing emission
 factors, valuable information can be obtained in a very cost-efficient manner by
 reexamining existing data bases.  For example, in keeping with EPA's general
 guidance, only uncontrolled emissions tests were used in the development of
 AP-42 Sections 11.2.5 (Urban Paved Roads) and  11.2.6 (Industrial Paved Roads).
 Inclusion of controlled tests in the emission factor data base would result in a
 100 to 200% increase in the size of the PM10 data base.

       Selection of the appropriate emission factor is not a question of ownership
 (public vs. private) but rather a question  of surface loading and traffic
 characteristics.  Public roads in industrial neighborhoods often have heavy
 surface loadings and are traveled by heavy vehicles.  In that case, the source is
 better characterized as "industrial" rather than "nonindustrial" in terms of
 particulate emissions.  Conversely, some roads within industrial plants have been
 found to exhibit emission characteristics that are far more "urban" than "industrial"
 in nature.  Reexamination of the differences  between industrial and urban paved
 road tests may also reduce the uncertainty in emission inventories noted in
 various EPA documents.

       Unlike many fugitive dust controls, the primary control measures available
 for paved roads seek to reduce silt loading, the independent variable in the
 emission factor equation.  Additional paved road surface sampling is  required to
 enlarge the small existing data base of silt-loadings associated with this major
 source. Paved road silt loading data should be collected to encompass spatial
 and temporal variations, climatology, land-use patterns, and elevated  silt loadings
 created by operations such as sanding and salting. This data base could then
 be coupled with traditional urban planning data to allow better prediction of silt
 loading, without the need for extensive sampling by local agencies.
5.1.2 Agriculture

      Fugitive dust emissions from agriculture-related sources are suspected of
contributing significantly to ambient air particulate levels, even in some major
urban areas. These particulate emissions are generated primarily from two
activities:  (1) wind erosion of open fields and (2) tilling operations.  For the most
part, uncontrolled emission estimates for these sources are based on data from
the 1940s through the mid-1970s, and do not consider modern machinery and
agricultural practices. Moreover, the emission measurements generally do not
incorporate state-of-the-art techniques for  obtaining particle size distribution.

      For example, wind erosion emissions of bare or partially  vegetated
agricultural land currently are estimated based on studies  reported in 1965 and
MRK3TS\R9800.FNL
                                                                          33

-------
 1968.  This prediction system evaluates soil loss of total participate by integrating
 over large fields and long-time scales to produce average annual values.  The
 same equation is used to estimate emissions from:  (1) a single field, (2) a
 regional area such as a valley or county, or (3) an entire State.

       Almost all PM10 emissions resulting from agricultural operations are
 estimated with D-rated factors, the lowest rating usually given to measured
 emission factors.  PM10 emissions must now be estimated cautiously by drawing
 on existing emission factors and particle size distributions for related sources.
 Estimates listed in an interim report, "Gap Filling PM10 Emission Factors for
 Selected Open Area Dust Sources" are still being used by local and State
 agencies, even though this report was intended only for short-term usage
 (Grelinger et al., 1988).  The agricultural sources treated in the gap-filling report
 include:

       •     Agricultural tilling
       •     Agricultural harvesting of cotton
       •     Agricultural harvesting of grain
             Cattle feedlots

       New studies, therefore, are needed to better predict and understand
 particulate emissions from agricultural operations.  These studies include:

       •     Investigation and integration of new size selective emission data
             into EPA recommended emission factors for wind erosion from bare
             and partially tilled ground and  from revised tilling practices and
             environments

       •     Improvement of methods to predict high ambient PM10 episodes in
             urban vicinities, using satellite technology to show soil condition
             (moisture content, bare surface exposure, etc.) and 48-96 hr
             meteorological forecasting of high winds. Short-term control
             measures, such as limited irrigation or specialized plowing, might
             also be studied under such situations.
5.1.3 Other Improvements

      Other emission factor data bases that should be reexamined with respect
to application to other sources include:

      •      Examination of the agricultural tilling and Section 8.24 overburden
             bulldozing factors to determine applicability to MSW landfilling
             operations.  Because of recent interest in inventorying landfill
34                                                                 MRI-OTS\R9800.FNL

-------
             emissions, the need for an emission factor for lift construction  has
             become evident.

       •     Review of the above data bases, as well as tests of pan scrapers,
             to assess applicability to construction site preparation (i.e., earth-
             moving) activities.  Construction sites have recently come under air
             regulatory scrutiny. Although the majority of earth-moving
             emissions are expected to be due to travel  (of pan scrapers or haul
             trucks), the dust-emitting activities often occur on bare earth rather
             than on'Well-defined and well-constructed roads. Consequently,
             there are doubts that the unpaved road equation can adequately
             predict emissions from this source.

       •     Determination of the applicability of wind erosion data bases to
             material losses from vehicles in transit.  Several  agencies have
             expressed interest in transit-related losses, but at present it is not
             known whether these emissions are limited  largely to the first few
             minutes during travel (the so-called "limited  potential") or whether
             emissions continue over the entire duration  of the trip.  It is likely
             that some additional field observations and  measurements will be
             necessary to answer these questions.

       Finally, a standardized emission inventory procedure (and applicable
computer tools) should be developed for State and local agencies.  Currently,
there are few consistent methods being used nationwide. Part of this effort
should be to develop the necessary correlations between surrogate data and the
main parameters associated with fugitive dust emissions. These mathematical
relationships would allow silt content and silt-loading values to be more reliably
ascribed to gridded  cells of the  inventory area, where no measurements are
available.
5.2  NEW EMISSION FACTORS

       For other sources of current regulatory interest, no potentially applicable
emission factors have been identified.  For those sources, suitable field tests
represent the only viable means to obtain applicable factors.  Furthermore, even
for factors that are available,  additional testing provides a means to better
characterize emission sources and their variability.

       This section briefly describes field-testing needs for new fugitive dust
emission factors. The source list is not intended to be exhaustive; virtually every
EPA regional office could list at least one or two other sources of interest within
its jurisdiction.  The sources listed represent those which MRI believes to be the
MR1-OTS\H8800.FNL
                                                                          35

-------
 most important (on the national scale) to provide the greatest immediate
 improvements in urban PM10 inventorying practices.
 5.2.1  Salting and Sanding

       The application of antiskid materials (i.e., deicers and abrasives) to paved
 roads produces a temporary, but substantial, increase in surface silt loading and
 thus PM10 emissions.  Increases in silt loading are a direct function of the quantity
 and quality of the material(s) applied.  For example, an abrasive which is both
 properly sized and highly durable will tend to produce less silt when exposed to
 traffic, which will reduce the potential for PM10 emissions.

       In a recent study (EPA-450/3-90-007), MRI developed criteria for the
 selection of antiskid materials for ice and snow control (Kinsey et al., 1991).
 These criteria were based on an extensive literature review and subsequent
 experimental program conducted in MRI's cold room.  In the study, different
 material samples were exposed to the same simulated traffic conditions, and
 those with the lowest silt generation potential were identified. The selection
 criteria were  based on the characteristics of those materials that exhibited the
 lowest silt generation and thus the lowest potential for PM10 emissions. No direct
 information was (or could be) developed in the study with regard to the actual
 PM10 emissions related to the use of antiskid materials, the changes in surface
 silt loading resulting from such application, or the degree of control that could be
 achieved by compliance  with the material selection criteria.

       Although several field studies have attempted to quantify the adverse air
 quality impact due to the application of antiskid materials, the results of these
 studies are, at best, inconclusive. Additional work is needed to improve the
 current state of knowledge with regard to the PM10 emissions associated with the
 use of deicers and abrasives for ice and snow control.  The recommended
 studies are described briefly below.

       As suggested by Kinsey et al. (1991), additional study of surface-loading
 dynamics is needed to better understand the PM10 impacts associated with
 various antiskid program scenarios. This could be accomplished by a well-
 designed and implemented surface-sampling program performed before,  during,
 and after individual storm events.

       Also, the fate and transport of salts (and other deicers) applied for skid
 control should be studied to determine an applicable emission factor based on
field testing.  A mass balance approach would be useful for this purpose,
 combining  surface and air sampling to determine the PM10 emissions associated
with deicing salts.
36                                                                MRMDTS\R9800.FNL

-------
 5.2.2  Mud/Dirt Carryout

       Carryout of mud and dirt from unpaved areas onto paved public roads
 often accounts for a substantial fraction of paved road silt loadings in many
 areas.  Examples of such unpaved travel areas include:

             Residences (i.e., driveways)
       •     Industrial operations
       •     Parking lots
       •     Construction sites
             Road shoulders

 Once material has been deposited on the paved surface, the material is subject
 to entrainment by passing vehicles.

       Quantification of control efficiencies for preventive measures is essentially
 impossible using the standard before/after measurement approach.  Further-
 more, tracking of material onto a paved road results in substantial spatial
 variation in loading about the access point. This variation  complicates the
 modeling of emission reductions as well as their estimation, although these
 difficulties become less important as the number of unpaved areas in an area
 and their access points become larger.

       Future testing is needed to adequately characterize mud/dirt carryout.
 Past effects have been based on an assumed spatial functional form for road
 loading and application of the urban paved road equation (Englehart and  Kinsey,
 1983).  MRI recommends a field  program be developed that combines extensive
 (in terms of space and time) sampling of deposited material together with
 intensive source testing of emissions from vehicle travel for the areas affected by
 trackout.  Unlike many important nonindustrial PM10 sources, trackout emissions
 represent a source that could be reasonably and successfully simulated under
 controlled conditions.

      Consequently, it is recommended that a comprehensive field study be
 conducted to (a) determine the amount of trackout as a function of source
 parameters (e.g., number of vehicles entering/leaving an unpaved area, unpaved
 surface material properties, spatial/temporal distribution of the additional surface
 loading); (b) characterize the emission strength of the tracked material with  air
sampling; and, (c)  determine the effectiveness of control measures applied to
this source.
MRI-OTS\R9600.FNL
                                                                        37

-------
 5.2.3  Construction and Demolition Activities

       Construction and demolition activities can be broken into generic
 operations such as truck travel over paved and unpaved surfaces or handling of
 bulk aggregate materials, as suggested by recent EPA documents.  Vehicle
 movements are reasonably easy to obtain by knowing the net weight of trucks
 and how much material is to be moved.

       What is not readily available at this time, however, is guidance on the
 design of effective control programs for construction sites.  It is generally
 acknowledged that the most cost-effective control programs for nonindustrial
 sources are necessarily the same as for industrial operations.  However, the
 temporary nature of construction site activities rules out some controls commonly
 found in industry.  For example, because  construction often occurs over short
 periods of time, chemical dust suppressants are often prohibitively expensive.
 Thus, water is often used to control  dust at construction sites, even though
 watering often compounds any mud/dirt trackout problems present at the site.

       A secondary source is created by demolition activities.  Mechanical
 dismemberment and debris loading  of buildings emit  PM10 directly to the
 atmosphere and enhance emissions from paved roads due to deposited
 material.  Further field work should address this secondary source in that an
 accounting of secondary effects may be more important than direct emissions
 from the site.
5.2.4 Abrasive Blasting

      Abrasives from building restoration projects contribute not only direct-
airborne emissions but also represent a secondary impact from paved roads.
Because abrasive blasting often occurs in highly urbanized areas with high traffic
volumes, passing traffic pulverizes and reentrains the deposited material.  It is
recommended that a field program be undertaken to address not only direct
emissions but also to determine the secondary impact from paved roads.

      Based on a recent assessment of outdoor blasting (Kinsey,  1989),
particulate emission factors presented in the literature are of poor quality. In
addition, control data for  abrasive blasting operations are limited at best.  New
field tests are necessary to develop emission factors for both wet and dry
blasting systems typical'of current industry practice. Different types of abrasives
also should be evaluated in the experimental program.
38                                                               MHI-OTS\R9800.FNL

-------
 5.2.5  Agriculture

       Agricultural tilling operations have changed recently due to the two Food
 Securities Acts (FSAs). These Acts mandate conservation tilling that leaves crop
 residues on the soil surface.  Modern agricultural tilling practices, such as use of
 coulters and zone eliminators, may decrease fugitive dust emissions, but test
 measurements are not available to confirm this and to establish the basis for
 revision of the AP-42 emission factors.  In addition, available data are insufficient
 for estimating efficiencies associated with measures to control agricultural wind
 erosion.

       In addition, the following new studies are recommended for agricultural
 sources:

       •     Characterization of reasonable and cost-effective emission control
             practices for wind erosion. Well-designed experimental programs
             should be expanded to develop reliable control  efficiency
             algorithms.

       •     Characterization of the ambient PM10 for agricultural chemical
             constituents, including fertilizers, pesticides, and herbicides. While
             major studies have examined water runoff from fields to which
             agricultural chemicals have been applied, almost no work has been
             done to characterize particulate emissions contaminated with such
             chemicals.

       •     Advancement to a clearer understanding of the  "limited potential"
             and "unlimited potential" erosion processes using  a portable wind
             tunnel to test a variety of agricultural surfaces.  The current
             distinction is based on an estimated threshold friction velocity of
             50 cm/sec, and should be clarified. [Recent analysis of "unlimited
             potential" desert surfaces shows that time duration of a wind storm
             also affects the erosion process and in effect creates a "limited
             potential" surface.]
5.3  CONTROL EFFICIENCY DETERMINATION

      In addition to uncontrolled emission factors, measured values for control
efficiency also are required to develop fugitive emission inventories for urban
areas. These values are needed, not only to determine the impact of historical
control strategies, but also to predict future emission reductions.

      At present, the only open source control efficiencies published  in AP-42
are for unpaved roads and heavy construction activities. Therefore, a number of
MHI-OTS\FB600.FNL
                                                                          39

-------
 guidance documents have been published by EPA that summarize available
 information for open source controls (Cowherd and Kinsey, 1986;  Cowherd et al.,
 1988).  For several of these control techniques, however, the data base is
 extremely limited and in need of improvement.  The following subsections outline
 those control techniques that are in need of additional field testing.
 5.3.1  Road Dust Controls

       Although temporary in nature, the watering of unpaved roads is one of
 the most universally applied controls used throughout the country.  The control
 efficiency of unpaved road watering depends on several factors including:
 (a) the amount of water applied per unit area of road surface area;  (b) the time
 between applications; (c) traffic characteristics throughout that period; and
 (d) prevailing meteorological conditions during the period of interest.  While
 several studies have estimated or investigated watering efficiencies, few have
 specified all the above factors.

       Two different techniques have been used to predict the efficiency of road
 watering. In the first approach, an empirical model for the performance of
 watering was developed from 14 tests performed in four states during five
 different summer and fall months.  The model for average control efficiency is
 (Cowherd and  Kinsey, 1986):


                            C =  100 - °'8 P d *                      (5-1)
where:
      C = average control efficiency, percent
      p = potential average hourly daytime evaporation rate, mm/h
      d = average hourly daytime traffic rate, vehicles/h
      i  = application intensity, L/m2
      t = time between applications, h

The potential average hourly evaporation rate (p) in the above equation can be
determined using the procedure outlined by Cowherd et al. (1988) with the other
parameters determined from on-site observations.
40                                                                MRI-OTS\R9800.FNL

-------
       The range of values used to derive the above expression was limited to
 the following:

       •     Average potential evaporation rate: 0.042 to 0.26 mm/h
             Average traffic:  23 to 98 vehicles/h
       •     Application intensity:  0.2 to 1.9 L/m*
       •     Time between applications:  1.8 to 4.5 hours

       As an alternate method, Figure 5-1 can also be used.  This figure is
 adapted from 11 field tests conducted of heavy vehicle travel over a coal pile
 (Cowherd et al., 1988). Measured control efficiencies did not correlate well with
 either time or vehicle passes after application.  However, this is believed to be
 due to reduced evening evaporation when the tests were conducted.  Also,
 surface samples taken throughout the test period showed a strong correlation of
 moisture with control efficiency.

       As can be seen from the above discussion, additional field data are
 needed to improve the prediction of watering control efficiency.  This study
 should address both heavy and light duty vehicles traveling over road surfaces
 with varying silt contents and  traffic volumes.  Also, other meteorological factors
 such as insolation, wind speed, relative humidity, etc., should be characterized to
 improve model prediction. Finally, a control decay  history is also necessary for
 watering,  as was done for chemical treatments, to determine the average control
 efficiency for the period between applications  (Cowherd et al., 1988).

      As stated in Section 5.2.2 above, the increase in paved road surface silt
 loading due to mud/dirt carryout is a substantial contributor to urban PM10
 problems. At present no  efficiency data are available for carryout controls.
 Potential control techniques to be evaluated for mud/dirt carryout include:
 cleaning of vehicle tires and underbodies before entering a paved road, paving
 or chemically treating access to site exit, or semicontinuously cleaning of the
 paved road at the site access point.

      Several approaches could be used to assess the efficacy of  mud/dirt
 carryout controls.  The most cost-efficient method would be to conduct extensive
 sampling of paved road loadings at "controlled" and "uncontrolled" construction
 sites coupled with intensive emissions testing. The data generated in this
 manner would give a good relative indication of control efficiency and would also
 be useful for the determination of baseline emissions prior to the implementation
 of control  strategies  for a  particular urban area.
MFU
-------
100
    0
                                                                   95
2468
  Surface  Moisture  Content %
   Figure 5-1. Watering control effectiveness for unpaved travel surfaces.
42
                                                            MRI-OTS\R9800.FNL

-------
 5.3.2  Controls for Materials Handling

       Although the emissions from materials handling are generally minor
 compared to other open dust sources such as roads, additional control
 efficiency data are needed for wet suppression, full and partial enclosure, and
 capture and collection. Each will be discussed below.

       Wet suppression refers to the application of water, a water solution of
 chemical agent, or micron-size foam to the surface of the dust-generating
 material to prevent (or suppress) the liberation of fine dust. The efficiency of
 control is dependent on a number of factors including:  the type of material
 handled, liquid application rate (L of water per Mg of material handled), droplet
 size, water distribution, and point(s) of application.

       All of the available data for wet suppression systems (including both water
 sprays and foam) were originally published by Cowherd and Kinsey (1986). The
 control efficiency values presented in that document were derived from
 personnel sampling of dust concentrations during the handling of coal or sand.
 Little information was available in the way of control application parameters.

       To determine the control efficiency of wet suppression, a well-designed
 and implemented field study is needed.  This study should address a wide range
 of material characteristics, as well as the key operational parameters  listed
 above. To limit influences of outside sources, the test location should be isolated
 to the  extent possible.  Material samples should also be collected for moisture
 content to determine emission rate for various levels of water application.

       A second control technique needing additional data involves the use of
 full or partial enclosures (including porous wind fences). Enclosures are a low
 maintenance control alternative used in many materials handling and processing
 applications to reduce the incident wind speed thus limiting the escape of
 airborne particles to the atmosphere. Essentially no control efficiency data are
 available for the use of enclosures.  This lack of data is  due to the fact that the
 fugitive emissions must be determined either before and after installation of an
 enclosure on the same source or with and without an enclosure on two
 essentially identical sources.

       To quantify the control efficiency of exposure profiling, air sampling could
 be conducted before and after the installation of a temporary enclosure(s)
 around a typical source(s).  Wind speed and particle flux measurements could
 be made to determine the controlled and uncontrolled emission rate  as a
function of wind speed  reduction.  A source which is relatively small and located
 near ground level would be  required to make sampling  possible.  Also, different
types of operations handling a range of materials should be evaluated to provide
a suitable predictive model of control efficiency.
MHMDTS\R9800.FNU
                                                                          43

-------
       The control efficiency of capture and collection is a direct function of
 both the capture efficiency and the collection efficiency of the system as shown
 in Figure 5-2. Hood capture efficiency is dependent on a number of key design
 and operational parameters including: airflow rate, hood-to-source distance,
 capture velocity, hood cross-sectional area, material  induced airflow, etc.

       Cowherd and Kinsey (1986) found only limited data for capture efficiency
 of hood systems applied to metallurgical operations. These data were obtained
 indirectly using a tracer gas (SF6).  Otherwise, no other data was available in the
 open literature.

       Additional data is definitely needed to characterize the control efficiency of
 capture and collection systems that are used on many large material handling
 systems (e.g., coal fired power plants).  To meet this objective, a mass balance
 approach should be used to determine the m4, m2, and m3 emissions shown in
 Figure 5-2.  Sampling of m2 and m3 emissions can be performed using standard
 stack sampling techniques, whereas the m4 emissions would be determined
 downwind of the source by exposure profiling.  Again, a range of hood design
 and operational parameters should be tested along with a variety of material
 characteristics.
44                                                                MRI-OTS\R9800.FNL

-------
           Not         	/
         Captured    /"rfi-    •*	     Collection
                                    .      Device
                         Captured  \	


                         ^^^
                   Capture
                    Device








Uncontrolled
                                                        \

                                                        /
Control Efficiency (%) -

where m-j = rh2 + rh4
                                                     Captured
                                                      but not
                                                     Collected
                                                 x 100
Figure 5-2.  Emissions quantification requirements for performance
            evaluation of capture/collection system.
MRMDTS\B9800.FNL
                                                                       45

-------
                                 SECTION 6

                         PROPOSED TEST  MATRIX
       Based on the data gaps identified in Section 5 above, a tentative test
 matrix has been prepared for planning purposes. This matrix (Table 6-1) outlines
 the individual test series to be conducted, the types of measurements to be
 made, and the number of locations, sources, and sampling runs to be per-
 formed.  The recommended geographic location(s) of the tests to be performed
 are also provided, as applicable.

       As shown in Table 6-1, the proposed testing has been categorized in two
 ways: by test objective; and by test priority. Test objective indicates whether the
 series is intended:  (a) to improve existing uncontrolled emission factors; (b) to
 develop new uncontrolled emission factors; or (c) to improve control efficiency
 estimates. Test priority is expressed as a relative ranking (A = highest) of all test
 series within the matrix indicating the overall need for implementation.

       The priority ratings assigned in Table 6-1 were based on the relative cost-
 effectiveness of each test series to obtain the most useful information for the
 least  cost. For example, testing which involves only surface material sampling
 and analysis was generally given a higher rating than more costly source
 sampling, when a test-based emission factor is already available.   Other factors
 influencing test priority included the expected typical impact of the individual
 sources  (or control methods) on the total PM10 emissions inventory as well as the
 extent and reliability of available data.

       Table 6-1 also specifies the measurement method to be used in each test
 series.  Since the reader may not be familiar with some of these techniques,
 Appendix B outlines general measurement methods for fugitive particulate
 emissions and Appendix C provides  similar information on use of the MRI
 dustiness test chamber.

       Finally, several'factors must be noted with regard to the information
 provided in Table 6-1 which affects its future use. First, the testing needs
 identified in Section 5 are preliminary in nature. Therefore, by necessity, the
testing proposed in Table 6-1 reflects a "first cut" analysis and should be used
 accordingly.
MRI-OTS\FB600.FNL
                                                                         47

-------
Table 6-1.  TENTATIVE TEST MATRIX FOR URBAN FUGITIVE DUST SOURCES"
Test
objective
Improvements to
current AP-42
emission (actors




Develop new test-
based emission
factors
Test
priority
(A = highest)
A
B
B
B
C
A
Description of test series
to be conducted
Expand silt loading data
for paved roads and
develop correlations with
land use, etc.
Collect additional source
test data for "low till"
farming practices to
update current PM10
emission factor and
quantify associated air
toxics (e.g., pesticides).
Collect additional source
test data for pan
scrapers and truck
haulage in construction
sites.
Reevaluate existing
paved road data base to
combine data from
"industrial" and "urban"
roads as well as
"controlled" testing.
Update current emission
factor model for
agricultural wind erosion.
New emission factor for
mud/dirt earn/out onto
paved urban streets.
Type of
measurements to
be made
Surface sampling
Source testing
coupled with
surface sampling
Source testing
coupled with
surface sampling
N/A
N/A
Extensive surface
sampling coupled
with intensive
source testing
General test
method(s)b
Vacuuming
Exposure profiling,
grab sampling,
and selected
chemical analyses
of collected
paniculate
Exposure profiling
and grab sampling
N/A
N/A
Vacuuming of
paved road(s) near
access point(s)
and exposure
profiling of paved
road(s)
No. of test
locations0
10
3
3
N/A
N/A
4
No. of sources
per location
10
2
2
N/A
N/A
1 access point
x 3 points in
time
No. of test
runs per
source
1
3
3
N/A
N/A
3
Test locations and other
factors Influencing sampling
Sampling In one city per
EPA Region. May delete
certain Reglon(s) If adequate
data already exist One
sample composite/road
evaluated.
Tests conducted at 1
California site, 1 Texas site,
and 1 Iowa site. Two types
of implements or operations
tested.
Tests conducted at 1
Southwestern site, 1 Middle
States site, and 1
Southeastern site.
Data analysis only
No work proposed at
present— track USDA
research.
All testing conducted locally
at 2 construction sites and 2
Industrial facilities at 3
different times during
construction project or
spring, summer, and fall for
Industrial sites.

-------
                                                Table 6-1 (Continued)
Test
objective
Develop new test-
based emission
(actors




Test
priority
(A=hlghest)
B
B
C
C
C
Description of test series
to be conducted
Evaluation of surface
loading dynamics for
application of antiskid
abrasives.

New emission factor for
application of delcing
chemicals.

New emission factor for
handling of soil in active
construction sites.

New emission factor for
debris loading at
demolition sites.
New emission factor for
abrasive blasting.

Type of
measurements to
be made
Extensive surface
sampling before,
during, and after
selected storm
events
Extensive surface
sampling
combined with
Intensive source
testing
Emission sampling
of end loaders and
shovels plus
material grab
sampling
Emission testing of
typical building
debris
Emission testing
coupled with
determination of
heavy metals
General test
method(s)b
Vacuuming
Vacuuming,
exposure profiling,
and selected
chemical analyses
of collected
participate
Exposure profiling,
grab sampling,
and "dustiness"
testing of soils
from other
geographic areas
Exposure profiling
N/A
No. of test
locations"
4
1
1
1
N/A
No. of sources
per location*1
1 road
segment x 3
points in time
1
2
1
N/A
No. of test
runs per
source
3
3
3
3
N/A
Test locations and other
factors Influencing sampling
Testing conducted at 1
Northeastern site, 1 Middle
States site, 1 Northern Plains
site, and 1 Western
Intermountain site. Samples
collected 3 different times
during winter months.
Source testing of 1 local site
once/season and surface
sampling 6 times during
winter.
Source testing of 1 local site
and analysis of soil samples
form 9 other cities for
"dustiness Index."
Source testing at 1 local site
for end loaders only.
No work proposed at
present— track Federal
Highway study.
CO

-------
en
o
Table 6-1  (Continued)
Tesl
objective
Improvements In
control efficiency
estimation




Test
priority
(A=highest)
A
B
C
C
C
Description of test series
to be conducted
Mud/dirt carryout
controls
Watering of unpaved
roads.

Wet suppression of
materials handling

Passive enclosures
Capture/collection
systems
Type of
measurements to
be made
Extensive surface
sampling coupled
with Intensive
source testing
Emission testing to
determine control
decay history
Emission testing
and material
sampling
Wind velocity
measurements
coupled with
source testing
Concurrent
emissions testing
at 3 locations
General test
method(s)b
Vacuuming of
paved road(s) near
access polnt(s)
plus exposure
profiling of paved
road (s)
Exposure profiling
and surface grab
sampling
Exposure profiling
or quasi-slack
sampling coupled
with grab sampling
and HjO analyses
Warm wire
anemometry and
exposure profiling
EPA Method 5
upstream and
downstream of
dust collector plus
exposure profiling
of uncaptured
emissions
No. of test
locations0
2
2
1
1
1
No. of sources
per location
3 emission
levels x 3
points In time
2 roads x 3
application
rates
2 materials x 2
handling
devices
2 materials x 3
emission
levels
2
No. of test
runs per
source
3
Multiple
3
3
3 tests x3
points
Test locations and other
factors Influencing sampling
All testing conducted locally
at 1 construction site and 1
Industrial facility. Testing for
uncontrolled emissions plus
paving of access point and
semlcontlnuous cleaning of
paved street
Test locations = 1 local site
plus 1 Western site.
Testing at local site on
handling of crushed stone
and coal using stacker and
belt conveyor.
Testing at one Plains state
site for two different
aggregate materials.
Sampling of uncontrolled
emissions plus two levels of
wind speed reduction.
Same site and materials
used for enclosures above.
                •  Matrix prepared from "first-cut" analysis of testing needs. Specific test plans (and associated budget) required for each test series.

                b  See Appendix B for description of general measurement methods and Appendix C for determination of "dustiness index."

                0  Geographic locale In contiguous United States.

                **  Sources or emission points to be tested.

-------
Specific test plans must be prepared prior to any type of field activity to more
closely define the scope (and associated budget) of each test series, the precise
methodology to be used, and applicable Quality Control/ Quality Assurance
protocols.  Also,  good test design calls for controlled experimental conditions in
order to collect meaningful data. Thus, in every case possible, it is preferred to
use sites, sources, and controls which are under the direct supervision  of the
testing organization rather than using the more traditional "take what you can
get" approach to source sampling.
MRI-OTS\R9800.FNL
                                                                           51

-------
                                SECTION 7

                               REFERENCES
 Cowherd, C. and J.S. Kinsey (1986).  "Identification, Assessment, and Control of
 Fugitive Particulate Emissions."  EPA-600/8-86-023, U. S. Environmental
 Protection Agency, Research Triangle Park, NC, May.

 Cowherd, C. et al. (1988). "Control of Open Fugitive Dust Sources."  EPA-450/3-
 88-008, U. S. Environmental Protection Agency, Research Triangle Park, NC,
 September.

 Englehart, P.J. and J.S. Kinsey (1983). "Study of Construction Related Mud/Dirt
 Carryout."  Final Report, EPA Contract No. 68-02-3177, Work Assignment No. 21,
 U. S. Environmental Protection Agency, Research Triangle Park,  NC,  July.

 Grelinger, M.A.  et al. (1988).  "Gap Filling PM10 Emission  Factors for Selected
 Open Area Dust Sources." EPA-450/3-88-008, U. S. Environmental Protection
 Agency, Research Triangle Park, NC, March.

 Kinsey, J.S. et al. (1983).  "Study of Construction Related Dust Control." Final
 Report, Contract No. 32200-07976-01, Minnesota Pollution Control Agency,
 Roseville, MN, April  19.

 Kinsey, J.S. (1989).  "Assessment of Outdoor Abrasive Blasting."  Interim Report,
 EPA Contract No. 68-02-4395, Work Assignment No. 29, U.S. Environmental
 Protection Agency, Research Triangle Park, NC, August.

 Kinsey, J.S. et al. (1991).  "Guidance Document for Selecting Antiskid Materials
Applied to Ice- and Snow-Covered Roadways." EPA-450/3-90-007, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC, July.
MRI-OTSXR9800.FNL
                                                                        53

-------
 U.S. Environmental Protection Agency (1980). 'Technical Procedures for
 Developing AP-42 Emission Factors and Preparing AP-42 Sections."  Office of Air
 Quality Planning and Standards, Research Triangle Park, NC, April.

 U.S. Environmental Protection Agency (1990). Compilation of Air Pollutant
 Emission Factors, Volume I:  Stationary and Area Sources.  AP-42 (with
 Supplements A-C), Fourth Edition, Office of Air Quality Planning and Standards,
 Research Triangle Park, NC, September.
54                                                                MRI-OTS\R9800.FNL

-------
                          APPENDIX A






        EMISSION FACTORS FOR PROCESS FUGITIVE SOURCES
MRI-OTS\R9800.FNL

-------
                               APPENDIX A
         EMISSION FACTORS FOR PROCESS FUGITIVE SOURCES
      The emission factors currently published in AP-42 for process fugitive
sources typically found in urban areas are provided in Table A-1 (EPA, 1990).
The individual factors are provided by industry and source category in the order
presented in AP-42. Also shown is the rating of each factor according to
published criteria (EPA, 1980).  For other comments on Table A-1, the reader is
referred to Sections 2.2 and 3.7 of the main text.
MRI-OTS\R9800.FNL
                                                                      A-3

-------
Table A-1. AP-42 PARTICULATE EMISSION FACTORS FOR PROCESS FUGITIVE SOURCES
AP-42
Section
5.3
5.8
5.14
5.22
6.4
6.5
7.1
7.2
7.3
Industry category
Carbon black
Hydrofluoric acid
Printing ink
Lead alkvl manufacture
Grain elevators
Grain processing
Beer making
Primary aluminum production
Coke manufacturing
Primary copper smelters
Process source
Oil furnace process
Spar handling
Pigment mixing
Sludge pits
Drying
Cleaning
Drying
Spent grain drying
Bauxite grinding
Anode baking furnace
Prebake cell
Vertical Soderberg stud cell
Horizontal Soderberg stud cell
Coal crushing
Coal charging
Coke pushing
Roaster calcine discharge
Reverb furnace matte tapping
Reverb furnace slag tapping
Converter
Converter slag and copper blow
Anode furnace
Slag cleaning furnace
AP-42 Emission factor*
0.10 kg PM,JMq carbon black
3.0 kg TP/Mg fluorospar
1 kg TP/Mg pigment
0.6 kg Pb/Mg product0
0.4-0.6 kg TP/Mg grain handled
1.5 kg TP/Mg grain handled
0.15-3.6 kg TP/Mg grain received
2.5 kg TP/Mg grain handled
NA
NA
1.45 kg PM10/Mg aluminum produced
6.0 kg TP/Mg aluminum produced
1.55 kg PM10/Mg aluminum produced
0.055 kg TP/Mg coal charged
0.004 kg PM10/Mg coal charged
0.25 kg PM10/Mg coal charged
1 .3 kg TP/Mg ore concentrate
0.074 kg PM10/Mg ore concentrate
0.028 kg PM10/Mg ore concentrate
2.2 kg TP/Mg ore concentrate
2.1 kg PM10/Mg ore concentrate
0.25 kg TP/Mg ore concentrate
4 kg TP/Mg ore concentrate
Emission factor
rating"
C
E
E
B
B
B
D
D
C
A
D
D
E
D
B
D
D
B
D
B
B

-------
                                                         Table A-1 (Continued)
      AP42
     Section
      Industry category
         Process source
       AP-42 Emission factor*
Emission factor
    rating*
       7.5
Iron and steel mills
Sinter machine windbox
Sinter discharge
Blast furnace casthouse
Hot metal desulfurization
BOF charging
BOF melting/refining
BOF tapping
Hot metal transfer
Electric arc furnace melting/refining
Electric arc furnace charging,
tapping & slagging
Open hearth furnace
melting/refining
Leaded steel teeming
Unleaded steel teeming
Scarfing
0.83 kg PM10/Mg sinter
0.016 kg PM10/Mg sinter
0.15 kg PM10/Mg metal
0.10 kg PM10/Mg metal
0.14 kg PM10/Mg metal
14.25 kg TP/Mg steel
0.21 kg PM10/Mg metal
0.028 kg TP/Mg steel
11.02kgPM10/Mgsteel
0.7 kg TP/Mg steel

8.76 kg PM10 metal

0.405 kg TP/Mg steel
0.035 kg TP/Mg steel
0.05 kg TP/Mg metal feed
      D
      C
      C
      E
      E
      B
      E
      B
      D
      C
                                                                                                                         A
                                                                                                                         A
                                                                                                                         B
       7.6
Primary lead smelters
Ore mixing/pelletizing
Sinter machine leakage
Sinter return handling
Sinter discharge, crushing &
screening
Blast furnace pouring
Blast furnace slag cooling
Zinc fuming furnace vents
Dross kettle
Reverb furnace leaks
Silver retort building
Lead casting
1.13 kg TP/Mg lead produced
0.34 kg TP/Mg sinter processed
4.5 kg TP/Mg sinter processed
0.75 kg TP/Mg sinter processed

0.47 kg TP/Mg lead produced
0.24 kg TP/Mg lead produced
2.3 kg TP/Mg lead produced
0.24 kg TP/Mg lead produced
1.5 kg TP/Mg lead produced
0.9 kg TP/Mg lead produced
0.44 kg TP/Mg lead produced
      E
      D
      E
      E

      D
      E
      E
      D
      D
      E
      E
>
cn

-------
>
CD
Table A-1 (Continued)
AP-42
Section
7.7
7.8
7.9
7.10
Industry category
Primary zinc smelters
Secondary aluminum
processing
Secondary copper
smelting/alloying
Gray iron foundries
Process source
Roasting
Sinter plant windbox
Sinter plant discharge/screens
Retort building
Casting
Crushing/screening/bailing
Shredding/classifying
Rotating barrel dross furnace
Dry milling
Reverb furnace chlorine
demagging
Reverb furnace refining
Sweating furnace
Crucible furnace
Furnace charging/tapping
Magnesium treatment
Inoculation
Pouring/cooling
Shakeout
Cleaning/finishing
Core making/baking
AP-42 Emission factor*
Nil
0.12-0.55 kg TP/Mg zinc slab
0.28-1.22 kg TP/Mg zinc slab
1-2 kg TP/Mg zinc slab
1 .26 kg TP/Mg zinc slab
NA
NA
NA
NA
266 kg PM10/Mg chlorine feed
1.3 kg PM10/Mg aluminum processed
7.25 kg TP/Mg aluminum processed
0.95 kg TP/Mg aluminum processed
NA
0.9 kg TP/Mg metal
1.5-2.5 kg TP/Mg metal
1.03 kg PM1(/Mg metal
1.12kg PM10/Mg metal
8.5 kg TP/Mg metal
0.6 kg TP/Mg metal
Emission factor
rating*.
E
E
E
E
D
D
C
C
—
E
D
D
E
D
D

-------
                                                    Table A-1 (Continued)
 AP-42
Section
               Industry category
                                      Process source
                                         AP-42 Emission factor"
                                     Emission factor
                                         rating"
  7.11
Secondary lead processing
Battery breaking/crushing
Sweating furnace charging/tapping
Smelting furnace charging/tapping
Kettle refining
Casting
NA
0.8-1.8 kg TP/Mg material charged
4.3-12.1 kg TP/Mg material charged
0.001  kg TP/Mg material charged
0.001  kg TP/Mg material charged
                                                                                                                    E
                                                                                                                    E
                                                                                                                    E
                                                                                                                    E
  7.12
Steel foundries
Charging/tapping
Backcharging
Alloying
Oxygen lancing
Slag removal
NA
NA
NA
NA
NA
  7.13
Secondary zinc processing
Crushing/screening
Reverb sweating
Rotary sweating
Muffle sweating
Kettle (pot) sweating
Electric resistance sweating
Sodium carbonate leaching
Melting furnace
Alloying retort distillation
Retort and muffle distillation
Casting
Distillation/oxidation
Retort reduction
2.3 kg TP/Mg product
0.63  kg TP/Mg product
0.45  kg TP/Mg product
0.54  kg TP/Mg product
0.17  kg PM10/Mg product
0.25  kg TP/Mg scrap
NA
0.0025 kg TP/Mg product
NA
1.18  kg TP/Mg product
0.0075 kg TP/Mg product
NA
NA
E
E
E
E
E
E
                                                                                                                    E
                                                                                                                    E

-------
>
00
Table A-1 (Continued)
AP-42
Section
7.14
7.17
8.1
8.3
8.4
8.10
8.11
8.14
Industry category
Storage battery production
Miscellaneous lead products
Asphaltic concrete plants
Bricks and related clay
products
Calcium carbide
manufacturing
Concrete batching
Glass fiber manufacturing
Gypsum manufacturing
Process source
Grid casting
Paste mixing
Plate stacking/burning/assembly
Lead reclaim furnace charge/tap
Dry formation
Melting furnace
charqinq/tappinq/castinq
Hot side
conveying/classifying/mixing
(puqmill)
Raw material grinding/screening
Tap fume vents
Furnace room vents
Primary/secondary crushing
Mixer loading
Raw material mixing/weighing
Crushing/batch charging
Primary/secondary crushing and
screening
Plaster bagging
Pin mixer
AP-42 Emission factor*
NA
NA
NA
NA
NA
NA
NA
2.66 kg PM10/Mg brick produced
NA
13 kg TP/Mg of calcium carbide
NA
0.02 kq TP/Mg material processed
0.3 kg TP/Mg material processed
Nil
NA
NA
NA
Emission factor
raiingb
—
—
—
E
C
E
B
—

-------
                                                          Table A-1 (Continued)
AP-42
Section
8.15
8.19.1
8.19.2
8.23
Industry category
Lime manufacturing
Sand and gravel processing
Crushed stone processing
Metallic minerals processing
Process source
Primary crushing/screening
Secondary crushing/screening
Pulverizing/screening
Calciner
crushing/pulverizing/screens
Primary/secondary crushing
Primary/secondary crushing (dry)
Primary/secondary crushing (wet)
Tertiary crushing (dry)
Primary crushing
Secondary crushing (high moisture
ore)
Tertiary crushing
AP-42 Emission factor*
NA
NA
NA
NA
0.009 kg TSP/Mg material processed
0.0085 kg PM10/Mg feed
0.009 kg TSP/Mg feed
0.93 kg TSP/Mg feed
0.02-0.004 kg PM10/Mg primary feed
0.012 kg PM10/Mg primary feed
0.08-0.001 kg PM10/Mg primary feed
Emission factor
rating*
—
D
D
D
E
C
D
E
>
CD

-------
>
o
                                         Table A-1 (Continued)
       Ap-43
      Section
     Industry category
         Process source
       AP-42 Emisstof} faetbi^
                                                                                                   mte$i0r> factor
      10.3
Plywood veneer & layout
operations
Log debarking
Log sawing
Veneer lathing
Plywood cutting/sanding
0.012 kg TP/Mg logs processed
0.175 kg TP/Mg logs processed
NA
0.05 kg TP/m2 plywood produced
E
E
        PM,0 = particles < 10 ^m in aerodynamic diameter (/jmA). 1 Mg = 10" gm.
        TP = total paniculate matter regardless of size.
        TSP = total suspended particulate matter • particles < 30
        Per EPA, 1980.
        Assumed to be all PM10 emissions.

-------
                          APPENDIX B


   GENERAL MEASUREMENT METHODS FOR FUGITIVE DUST EMISSIONS
MRI-OTS\R9800-FNL
                                                         B-1

-------
                               APPENDIX B


   GENERAL MEASUREMENT METHODS FOR FUGITIVE DUST EMISSIONS


 B.1  MASS EMISSION MEASUREMENT METHODS

       Fugitive particulate emission rates and particle size distributions are
 difficult to quantify because of the diffuse and variable nature of such sources
 and the wide range of particle sizes involved (including particles which deposit
 immediately adjacent to the source). Standard source testing methods, which
 are designed for application to confined streams under essentially steady-state,
 forced-flow conditions,  are not suitable for measurement of fugitive emissions
 unless the plume can be drawn into a forced-flow system.

       For field measurement of fugitive mass emissions, four basic techniques
 have evolved:

       1.    The quasi-stack method involves capturing the entire particulate
            emissions stream with enclosures or hoods and applying
            conventional source testing techniques to the confined flow.

       2.    The roof monitor method involves measurement of particulate
            concentrations and airflows across well-defined building openings
            such as roof monitors, ceiling vents, and windows, followed by
            calculation of particulate mass flux exiting the building.

       3.    The upwind-downwind method involves measurement of upwind
            and downwind particulate concentrations utilizing ground-based
            samplers under known meteorological conditions, followed by
            calculation of source strength (mass emission rate) with
            atmospheric dispersion equations.

      4.    The exposure profiling method involves simultaneous, multipoint
            measurements of particulate concentration and wind speed over
            the effective cross section of the plume, followed  by calculation of
            net particulate mass flux through integration of the plume profiles.

      5.    The wind  tunnel method involves the use of a portable open-
            floored wind tunnel for in situ  measurement of emissions from
            representative surfaces under predetermined wind conditions.

Each of these methods is discussed below.
MRI-OTS\R9800.FNL
                                                                      B-3

-------
 B.1.1  Quasi-Stack Method

       In practice, the quasi-stack method (Kolnsberg et al., 1976) converts a
 fugitive emission source to a conventional ducted source.  Because it is usually
 impractical to enclose an open dust source or capture its entire emissions
 plume, the quasi-stack method is generally limited in applicability to process
 sources.

      The quasi-stack method qualifies as a sound methodology only if the
 enclosure or hood is capturing the entire emissions stream without affecting the
 emission rate.  In addition, an accepted  sampling technique must be chosen to
 quantify the emission rate,  taking into consideration the special problems
 associated with highly fluctuating emissions.
 B.1.2  Roof-Monitor Method

       The roof monitor method (Kenson and Bartlett, 1976) is similar to the
 quasi-stack method in that it utilizes the building ventilation system to direct the
 emissions stream to the sampling location.  Usually this method is practical only
 for high temperature processes which produce buoyant plumes.

       The roof-monitor method qualifies as a sound methodology only if flows
 and concentrations can be adequately characterized within building discharge
 openings. Also, it must be shown that plume interference from other sources in
 the same building is not occurring.  Finally, as with the quasi-stack method, the
 special problems associated with highly fluctuating emissions (and, in the case of
 natural ventilation, fluctuating ambient winds) must be dealt with.
B.1.3 Upwind-Downwind Method

      The basic procedure of the upwind-downwind method (Kolnsberg, 1976)
involves the measurement of particulate concentrations both upwind and
downwind of the pollutant source. The number of required upwind sampling
instruments depends on the isolability of the source operation of concern (i.e.,
the absence of interference from other sources upwind). Although at least five
downwind particulate samplers must be operated during a test, increasing the
number of downwind instruments improves the reliability in determining the
emission rate by providing better plume definition.  In order to reasonably define
the plume emanating from a point source, instruments need to be located at a
minimum of two downwind distances and three crosswind distances.  The same
sampling requirements pertain to line sources except that measurements at
multiple crosswind distances are not required.
B-4                                                              MFU-OTSXFWeoo.FNL

-------
       After the concentration(s) measured upwind are subtracted from the
 downwind concentrations, the net downwind concentrations are input to
 standard dispersion equations (normally of the Gaussian type).  The dispersion
 equations are used to back-calculate the source strength (i.e., particulate
 emission rate) required to generate a pattern of downwind concentrations. A
 number of meteorological parameters must  be recorded concurrently for input to
 these dispersion equations.  The minimum parameters that must be recorded
 on-site are wind direction and speed.
 B.1.4  Exposure Profiling Method

       In much the same manner as the quasi-stack and roof monitor methods,
 exposure profiling (Cowherd et al., 1974) uses the same basic concept as
 conventional (ducted) source testing. The difference is that in the case of
 exposure profiling, the ambient wind directs the plume toward the sampling
 array.  The passage of airborne particulate matter immediately downwind of the
 source is measured directly by means of simultaneous multipoint sampling of
 particulate concentration and wind velocity over the effective cross section of the
 fugitive emissions plume.

       For measurement of nonbuoyant fugitive emissions, profiling sampling
 heads are distributed over a vertical network positioned immediately downwind
 (usually about 5 m) from the source.  Particulate sampling heads should be
 symmetrically distributed over the concentrated portion of the plume representing
 about  90% of the total mass flux (exposure). A vertical line grid of at least three
 samplers is sufficient for measurement of emissions from line or moving point
 sources, while a two-dimensional array of at least five  samplers is  required for
 quantification of fixed virtual point  source emissions.  At least one  upwind
 sampler must be operated to measure background concentration, and wind
 speed on-site must be measured concurrently.

       Unlike the upwind-downwind method, exposure profiling employs a mass-
 balance calculation scheme rather than requiring  indirect calculation through
 application of a generalized atmospheric dispersion model. The mass of
 airborne  particulate  matter emitted by the source  is obtained by spatial
 integration of distributed measurements of particulate  flux,  after subtraction of the
 background contribution. The exposure is the point value of the flux
 (concentration of airborne particulate accumulated over the measurement
 period).
MRI-OTS\R9800.FNL
                                                                       B-5

-------
 B.1.5  Wind Tunnel Method

       Finally, the wind tunnel method (Cuscino et al., 1983) employs a portable
 pull-through wind tunnel with an open-floored test section placed directly over
 the surface to be tested.  Air is drawn through the tunnel at controlled velocities.
 The exit air stream from the test section passes through a circular duct fitted with
 a sampling probe at the downstream end. Air is  drawn through the probe by a
 high-volume sampling train.  This technique enables precise study of the wind
 erosion process with minimal interference from background sources.
 B.2 PARTICLE SIZING

       A number of different methods have been used for the determination of
 particle size in fugitive dust studies.  These methods include high-volume
 impactors and cyclones; dichotomous samplers; modified inlets for high volume
 air samplers; and optical and electron microscopy.  Each is briefly described
 below.

       High-volume cascade impactors with glass fiber (or other) substrates may
 be adapted for sizing of fugitive particulate emissions.  A cyclone preseparator
 (or similar device) is necessary to remove coarse particles which otherwise
 would be subject to particle bounce within the impactor, causing fine particle
 bias.  The cyclone can also be operated without the cascade impactor if a single
 size cut is desired.  The sampling intake should be  pointed into the wind and  the
 sampling velocity adjusted to the mean local wind speed by fitting the intake with
 a nozzle of appropriate size.

      The EPA version of the dichotomous sampler, which is virtually free of
 particle bounce problems, can also be useful for quantification of PM10
 concentrations.  However, this device operates at a low flow rate, yielding only
 relatively small quantities of sample in a 24-h period. Thus, an analytical balance
 of high precision is required to determine mass concentrations below and above
 the fine particulate (2.5 /imA) outpoint. This outpoint falls at the minimum in the
 typical bimodal size distribution of atmospheric particulate.

      Modifications to the standard  high-volume sampler have also been
 designed to capture particulate matter smaller than  10 ^mA.  These units are
 much less wind-sensitive than the dichotomous sampler but do not produce a
 cutpoint at 2.5 jinnA.

      Microscopy is another particle sizing technique used in fugitive dust
studies. Optical or light microscopes, transmission electron microscopes (TEMs),
and scanning electron microscopes  (SEMs) have been used  in particle sizing.
Optical microscopy  has proven useful in determining particle  sizes greater than
B-6                                                              MRI-OTS\R9800.FNL

-------
0.25 p.m in diameter. An electron microscope is needed to size particles greater
than 0.001  /im in diameter.

      Of the many microscopic techniques available for sizing particles by their
physical dimensions, the most common is the projected area technique. The
particle diameter is set equal to the diameter of an equivalent circle with the
same area as the projected area of the particle. A minimum of 300 particles is
usually required to determine a distribution of about nine size categories by this
method. Because such work is long and tedious when performed manually,
automation of the process is now common. Automatic image analysis for optical
microscopy and computer-controlled scanning electron microscopy (CCSEM)
are examples.
B.3  REFERENCES FOR APPENDIX B

Cowherd, C. Jr., et al. (1974). Development of Emission Factors for Fugitive
Dust Sources. EPA-450/3-74-037, U.S. Environmental Protection Agency,
Research Triangle Park, NC, June.

Cuscino, T., et al. (1983).  Iron and Steel Plant Open Source Fugitive Emission
Control Evaluation. EPA-600/2-83-110,  U.S. Environmental Protection Agency,
Research Triangle Park, NC, October.

Kenson, R. E., and P. T. Bartlett (1976). Technical Manual for the Measurement
of Fugitive Emissions:   Roof Monitor Sampling Method for Industrial Fugitive
Emissions. EPA-600/2-76-089b, U.S. Environmental Protection Agency, Research
Triangle Park, NC, May.

Kolnsberg, H. J.  (1976). Technical Manual for Measurement of Fugitive
Emissions: Upwind/Downwind Sampling Method for Industrial Emissions.
EPA-600/2-76-089a, U.S. Environmental Protection Agency, Research Triangle
Park, NC, April.

Kolnsberg, H. J., et al. (1976). Technical Manual for the Measurement of Fugitive
Emissions: Quasi-stack Sampling Method for Industrial Fugitive Emissions.
EPA-600/2-76-089C, U.S. Environmental Protection Agency, Research Triangle
Park, NC, May.
MRI-OTS\R9800.FNL

-------
                            APPENDIX C



            DESCRIPTION OF MRI DUSTINESS TEST CHAMBER
MRI-OTS\R9800.FNL
                                                            C-1

-------
                               APPENDIX C
             DESCRIPTION OF MRI DUSTINESS TEST CHAMBER
      The MRI dustiness test chamber is a bench-scale device used to generate
and sample airborne particulate resulting from the pouring of bulk material in
the chamber. Because dustiness testing has been shown to produce unit
emissions comparable to large scale operations, it may be used to bracket
projected emissions from the field transfer of bulk materials.
C.1  Test Chamber Configuration

      The basic design of the chamber includes the following features for the
sampling of total particulate (TP — approximate particle diameter < 75 //m):

      •    Test chamber dimensions of 20 cm x 20 cm x 50 cm high
      •    Cup with 0.27 L capacity used to hold, then pour the sample
           material 25 cm to the floor of the chamber
      •    Vibratory apparatus attached to the cup for smooth pouring at a
           cup rotation rate  of 0.8 rpm
      •    Chamber air exchange flowrate of 8.3 L/min with exhaust through
           a filter cartridge mounted in the chamber lid
      •    A 47-mm diameter, mixed cellulose ester filter (MCEF) with a pore
           size of 0.8 um for particulate collection
      •    Two side vents into the chamber, designed to direct make-up air
           into the lower portion of the chamber
      •    A Matheson No. 604 rotameter for control and monitoring of the
           sampling flowrate

      The chamber  lid can be modified to accept four personal samplers (37-
mm filter cassettes  plus 10-mm Dorr Oliver cyclone) with approximate cutpoints
of 3.5 pmA for the determination of respirable particulate (RP) matter.  In this
alternate configuration, four personal samplers are mounted symmetrically in
the modified lid. Each 37-mm filter cassette is located on top the lid with the
10 mm nylon cyclone extending down inside the chamber. The flowrate of
each personal sampler is 2.07 L/min to give a total air exchange rate of 8.3
L/min (equivalent to dustiness tests for TP).
                                                                  MRI 10/10/91

-------
      Even though a 2.07 L/min flowrate is slightly above the recommended
1.85 L/min for a personal sampler, the ACGIH manual on "Air Sampling
Instruments" states that "For non-fibrous test aerosols, including mica and
silica, there was essentially no change in the mass collected on the filter for
flowrates between 1.3 and 2.65 L/min.  As the flowrate increases, the aerosol
mass entering the cyclone increases proportionally, but so apparently does the
collection efficiency."
C.2   Test Conduct

      Each dustiness test generates suspended mass that is captured on filters
mounted in the chamber lid. The following steps represent a typical test
scenario for sampling both total and respirable paniculate resulting from the
pouring of sample material in the chamber.

      1.    Record the masses of material to be poured in the chamber

      2.    Conduct triplicate  tests for each material with:
             - Standard lid and 47-mm filter sampler for TP
             - Modified lid and four 37-mm personal samplers with cyclones
              for RP

      3.    Gravimetrically analyze each filter and record tare and final weights

      4.    If required, analyze filter substrates for other components.  For
            example, analyze for asbestos fibers  by transmission electron
            microscopy (TEM) or for crystalline silica by  NIOSH Method 7500.
            In the latter instance, ash the four 37 mm filters from each test for
            one Method 7500 analysis.
C.3   Test Calculations

      The net weights of particulate matter resulting from the above tests are
used to calculate the dustiness index (or emission factor -- mg of suspended
particulate per kg of material poured). The PM35/TP fraction is obtained by
comparing the combined catch on the four 37-mm filters with the catch on the
single 47-mm filter. The PM10 fraction can be estimated using a log-normal
distribution with reference points at an approximate largest particle size of 75
    and at 3.5 //m.
                                                                   MRI 10/10/91

-------