United States EPA-600/7-84-021
Environmental Protection
Agency ( _F eb ruary 1984
&EPA Research and
Development
IMPROVED STREET SWEEPERS
FOR CONTROLLING URBAN
INHALABLE PARTICULATE MATTER
Prepared for
Office of Air Quality Planning and Standards
Prepared by
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology, Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22181.
-------
EPA-600/7-84-021
February 1984
IMPROVED STREET SWEEPERS FOR CONTROLLING
URBAN INHALABLE PARTICULATE MATTER
by
Seymour Calvert, Harry Brattin and Sudarshan Bhutra
A.P.T., Inc.
4901 Morena Blvd., Suite 402
San Diego, CA 92117
EPA Contract No. 68-02-3148
EPA Project Officer: William B. Kuykandal
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
-------
ABSTRACT
Dust emissions from paved roads are a major source of urban
inhalable particulate matter. A.P.T. performed an experimental
program to develop design modifications which can be used to
improve the ability of municipal street sweepers to remove
inhalable dust particles from the street.
A commercial regenerative air sweeper was modified. Major
modifications include a charged spray scrubber for fine particle
collection and a gutter brom hood to help contain redispersed
dust particles. The upgraded sweeper was proven to be effetive
in eliminating the dust plume during sweeping and giving a
cleaner street.
11
-------
CONTENTS
Page
Abstract ii
List of Figures v
List of Tables ix
Abbreviations x
Nomenclature xi
Acknowledgment xiii
Sections
1. Summary 1
Introduction 1
Objective 2
Approach 2
Results 3
Conclusions 10
Recommendations 11
2. Introduction 12
Urban Fugitive Dust Emissions 12
Street Sweepers 12
Proposed Method for Sweeper Improvement ... 15
Research Program Outline . . 20
3. Background Information 23
Street Dust Loading and Size Distribution. . . 23
Street Sweepers 34
4. Preliminary Experiments 49
Air Flow Near Broom 49
Redispersed Street Dust Concentration
and Size Distribution 50
5. Sweeper Modifications 59
Tymco Model 600 Street Sweeper 59
Conceptual Design 62
Preliminary Experiments 62
Detailed Design 85
iii
-------
CONTENTS
Page
6. Road Tests 90
Gutter Broom Hood 90
Vent Air Rate 91
Particle Size Distribution and Concentration
at Scrubber Inlet 94
Sweeping Efficiency 106
Scrubber Efficiency 117
7. Recommendations 125
References 126
-------
LIST OF FIGURES
Figure 1.
Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.
Figure 9.
Figure 10,
Figure 11,
Figure 12,
Figure 13,
Figure 14,
Figure 15,
Figure 16,
Page
Street debris accumulation rates
(URS Research Company, 1974) 14
SCAT system for controlling reentrained
dust from paved roads 18
Deposition and removal processes
(Axetell and Zell, 1977) 24
Rate of reentrainment of particulates
from street surfaces (Pitt, 1978) 29
Reentrainment of dust by vehicles as a
function of amount of street debris
(Pitt, 1978) 30
Size distribution of suspended dust
in vicinity of paved streets 32
Particle size distribution near road
(Pitt, 1978) 34
Three wheel broom sweeper 40
Vacuum assisted sweepers 41
Debris picked up vs. brush speed
(Horton, 1968) 43
The effect of pattern on residual
debris (Horton, 1968) 44
The sweeping pattern vs. stiffness for
synthetic broom fibers (Horton, 1968) .... 45
Air velocity near broom of Mobil
sweeper. Wind velocity: 0.5 m/s 51
Particle size distribution of dust dis-
persed by a mechanical street sweeper .... 52
Schematic diagram of sampling equipment
for particle size distribution data 54
Particle size distribution of dust dis-
persed by a vacuum (regenerative air)
sweeper. Data are compared with previous-
ly reported particle size distributions ... 56
v
-------
LIST OF FIGURES (continued)
Figure 17.
Figure 18.
Figure 19.
Figure 20.
Figure 21.
Figure 22.
Figure 23.
Figure 24.
Figure 25.
Figure 26.
Figure 27.
Figure 28.
Figure 29.
Figure 30.
Figure 31.
Figure 32.
Figure 33.
Figure 34.
Page
Comparison of cumulative mass concen-
tration of dust dispersed by street
sweepers (dry sweeping) 57
Side view of TYMCO vacuum
(regenerative air) sweeper 60
Schematic diagram of TYMCO Model
600 regenerative air system 61
Process diagram of improved street sweeper. . 63
Predicted penetration for 300 pm
diameter spray drop scrubbing 67
Predicted penetration for 300 ym
diameter spray drop scrubbing 68
Predicted efficiency for coulombic
attraction and inertial impaction 71
Trajectories for water drops in air 72
Collection efficiencies of neutral
and charged drops for 0.6 umA
diameter particles 73
Experimental apparatus for measuring
particle collection efficiency of a
charged spray scrubber 77
Measured spray scrubber particle
penetrationf Test 1 78
Measured spray scrubber penetration 79
Measured spray scrubber penetration 80
Measured spray scrubber penetration 81
Measured particle charger and
spray scrubber penetration 82
TYMCO street sweeper and trailer 86
Scrubber shell 87
Plan view of the sampling platform 89
VI
-------
LIST OF FIGURES (continued)
Page
Figure 35. Schematic diagram of street sweeper
flow circuit 92
Figure 36. Particle size vs. mass concentration
of dust at scrubber inlet in an in-
dustrial area 95
Figure 37. Particle size vs. mass concentration
of dust at scrubber inlet in a com-
mercial area 96
Figure 38. Particle size vs. mass concentration
of dust at scrubber inlet in a resi-
dential area 97
Figure 39. Particle size distribution of dust at
scrubber inlet in an industrial area 98
Figure 40. Particle size distribution of dust at
scrubber inlet in a commercial area 99
Figure 41. Particle size distribution at scrubber
inlet in a residential area 100
Figure 42. Scrubber inlet dust concentration 103
Figure 43. Particle size distribution 104
Figure 44. Effect of vent air flow on scrubber
inlet dust concentration 105
Figure 45. Uniform airflow vacuum nozzle 108
Figure 46. Air jet nozzle 109
Figure 47. Air jet nozzle performance 110
Figure 48. Test strip layout for street sweeping
efficiency measurement 112
Figure 49. Street dust distribution before and
after sweeping 113
Figure 50. Street dust distribution before and
after sweeping 114
Figure 51. Overall street sweeping efficiency 116
Figure 52. Street sweeping efficiency 118
Figure 53. Street sweeping efficiency 119
VII
-------
LIST OF FIGURES (continued)
Page
Figure 54. Scrubber performance. 121
Figure 55. Scrubber performance 122
Figure 56. Scrubber performance 123
vlii
-------
LIST OF TABLES
Table 1.
Table 2.
Table 3.
Table 4.
Table 5.
Table 6.
Table 7.
Table 8.
Table 9.
Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Page
Dust deposition and removal rates from
paved roads (Axetell and Zell, 1977) 13
Cleaning efficiency of street cleaning
methods (Sartor and Boyd, 1972) 16
Street dust surface loading
(Sartor and Boyd, 1972) 26
Size distribution of street dirt and
dust (Sartor and Boyd, 1972) 27
Vehicle emission rates on paved roads .... 31
Technical features on street sweepers .... 35
Information on street sweepers 38
Sweeper efficiency with respect to
particle size (Sartor and Boyd, 1972) .... 46
Total solids street cleaner removal effect-
iveness by particle size (Pitt, 1979) .... 47
Air flow characteristics of the
TYMCO sweeper 65
Average collection efficiencies of drops
and spray penetrations 75
Data on fine spray nozzles 84
Summary of sampling results from
Los Angeles area 101
Summary of scrubber tests 120
-------
LIST OF ABBREVIATIONS
IPM - inhalable particulate matter
GB - gutter broom
SCAT - Spray charging and trapping
ID - Inside diameter
OD - Outside diameter
PVC - polyvinyl chloride
Sch - pipe schedule
OAQPS - Office of Air Quality Planning & Standards
x
-------
NOMENCLATURE
C = Coulomb
C1 = Cunningham slip factor, dimensionless
dc = collector diameter, cm or urn
d^ = diameter drop, cm or urn
dp = particle diameter, cm or pmA
dpa = aerodynamic particle diameter, umA
fA = gas flow cross-section covered by
spray, fraction
kf = gas dielectric constant
KC = Coulombic attraction parameter,
dimensionless
K = inertial impaction parameter,
dimensionless
Pt = overall particle penetration,
fraction
Ptd = particle penetration for diameter
ndpa", fraction
Qc = collector charge, C
Q_ = particle charge, C
QQ = gas flow rate, crnVs
QL = liquid flow rate, cm3/s
R^ = drop range (distance traveled by drop
relative to gas), cm
ur = gas velocity relative to collector,
cm/s
UQ = gas velocity, cm/s
UGO = initial drop velocity relative to gas,
cm/s
UQ = initial velocity, cm/s
x = coordinate in gas flow direction, cm
-------
dimensionless
NOMENCLATURE (GREEK)
geometric standard deviation,
instantaneous single drop collection
efficiency, fraction
average collection efficiency over
drop range, fraction
particle density, g/cm3
Xll
-------
ACKNOWLEDGMENTS
A.P.T., Inc. expresses grateful appreciation to the
following individuals for their contribution during this study:
Dennis C. Drehrael former Project Officer,
EPA, IERL/RTP
William Kuykendal Project Officer,
EPA, IERL/RTP
Sincere gratitude goes to the City of San Diego and the City
of Anaheim for providing street sweepers and operators during
this project.
-------
Section 1
SUMMARY
INTRODUCTION
Dust emissions from paved roads are a major source of urban
air pollution. Draftz (1978) estimated that 40 to 70% of the
total suspended particulate matter in many urban areas comes from
dust particles redispersed by road traffic. Pitt (1979) esti-
mated that city streets can contribute from 5 to 50 ug/m3 of
particulate matter to the urban air.
The existing types of street sweepers include broomf vacuum,
regenerative vacuum, and water flushing sweepers. These
generally have low-to-moderate efficiency for removing inhalable
particulate matter (IPM) from streets and test data scatter
widely. Buchwald and Schrag (1967) sampled the air inside the
driver's cab and determined that there is a serious exposure to
respirable dust. Closing the windows and the use of a cyclone
separator for dust collection had small effect on respirable dust
concentration in the cab.
It is clear that urban street dirt can cause air pollution,
water pollution from rain water runoff, and an occupational
hazard for the sweeper operator. When evaluated with regard to
IPM removal, each of the conventional types of sweepers has a
significant deficiency, as indicated below.
Type of Sweeper Deficiency
Broom Disperses IPM while sweeping.
Vacuum Either emits IPM or must clean a
large volume of vent air. Gutter
broom disperses IPM into air.
Regenerative Vacuum Disperses IPM while sweeping.
Water flush Moves IPM to gutter and causes
water pollution.
-------
OBJECTIVE
Under E.P.A. contract. A.P.T. has conducted a research and
development program with the primary objective of developing
practical means for improving the efficiency of a suitable street
sweeper for the removal of IPM from urban streets.
A.P.T. evaluated the problem and proposed to develop a
street sweeper which would use a regenerative air flow system,
gutter broom hooding, and a low-volume scrubber on a vent air
stream. The use of a vent air stream would permit a positive
inward flow of air from the sweeping area into the sweeper body
rather than allowing this dusty air to be dispersed. The appli-
cability of the SCAT (Spray Charging and Trapping) system was to
be investigated.
The SCAT system uses combinations of air curtaining, hood-
ing, and spray scrubbing (with or without electrostatic augment-
ation) to capture and retain particles for subsequent disposal.
Specific circumstances dictate the SCAT features which are used
in each case.
Additional objectives of the program were to:
1. Build a modified street sweeper and demonstrate its
capabilities for urban street sweeping.
2. Develop and use methods for evaluating sweeping
efficiency.
APPROACH
The general approach of the program was consistent with the
premise that a suitable scrubber could be designed if one knew
what air flow rate and particle collection capability were re-
quired. In other words, much more was known about scrubber
design than about street sweeping" in terms of IPM control para-
meters. Consequently, the main effort was placed on deter-
mining:
-------
1. Background information on street sweeping.
2. The street sweeper best suited for study.
3. A tentative set of design criteria to use for designing
the experimental street sweeper, including:
a. Air flow rates required to control dust emissions.
b. The total air flow rate which has to be scrubbed.
c. Particle size distribution and particle concentra-
tion in the uncontrolled effluent air.
d. Utilities limitation.
4. Design concepts for the modifications needed for IPM
control.
5. An experimental procedure for determining the
efficiency of removing IPM from urban streets.
A limited effort was expended on charged spray scrubbing and
scrubber design. Some experimental work was done to confirm
earlier experiments and to evalute concepts for atomizers and
drop chargers which could produce smaller charged drops than used
in previous work. The further development of a mathematical
model for particle collection by charged spray also received some
effort.
RESULTS
Literature Search
A literature search and interviews of qualified persons were
conducted to determine:
1. The nature of dirt on paved streets, including informa-
tion on the sources of particulate matter, the methods
of deposition, and various methods of removal.
2. The prevalent types of street sweepers available, their
cost, and the potential for improving them.
3. Methods for sampling and analysis of the IPM on the
street and in the air.
-------
There is abundant literature on street dirt and its contri-
bution to air and water pollution. There is, however, very
little on the subject of street sweeper efficiency as a function
of dust particle size. Brookman and Martin (1979) present an
extensive literature review and recommendations for research
areas. Some of the works which provide useful quantitative
information are those of Axetell and Zell (1977), Cowherd et al.
(1977), Draftz (1978), Pitt (1979), Sartor and Boyd (1972), and
Sehmel (1973).
At the time the research began, broom sweepers were the type
most commonly used. However, the trend appeared to be toward the
use of regenerative vacuum sweepers, so this was the type selec-
ted for the present research.
Street sweeping effectiveness has been determined in several
ways, such as:
1. Ambient air was sampled upwind and downwind from the
street to determine the contribution of dust from the
street (e.g., Cowherd et al., 1977, and Pitt, 1979).
2. Street dirt loading was measured by vacuum-cleaned
sample areas before and after sweeping (e.g., Cowherd
et al., 1977).
3. Street dirt loading was measured by a combination of
sweeping and water-flushing sample areas before and
after street sweeping (e.g., Sartor and Boyd, 1972).
Regenerative Vacuum Sweeper
A regenerative air vacuum sweeper uses a gutter broom to
brush the curb and throw the dirt from the curb toward the cen-
ter. A center plate is used to stop the thrown dirt so that it
piles up at the center of the sweeper path. As the sweeper moves
forward, the debris piles can be taken up by the vacuum pick-up
head, which extends almost the entire width of the sweeper.
Blast orifices in the pick-up head direct air jets at the street
surface to dislodge the dirt, which is then sucked into the
hopper through the vacuum hose. The air and the dirt are separ-
-------
ated in the hopper. The air then recycles back to the blast
orifices through the blower and the pressure hose.
Conceptual Design
Observation of street sweeper operation clearly showed that
gutter broom and leakage from the "pickup head" were the major
sources of IPM emissions from the regenerative vacuum sweeper.
The following design concepts were conceived during sweeper oper-
ation observations. The dust clouds around the gutter broom can
be controlled by installing a hood over the broom and venting to
the hopper. Since the hopper is under vacuum/ an induced flow of
air will convey the contained dust to the hopper.
Dust clouds in the pick-up head area were observed to occur
when the pick-up head travels on uneven street surfaces. Pick-up
head dust clouds can be eliminated by increasing the vacuum in
the vacuum hose, which causes an increase in the inward flow into
the pick-up head.
To balance the air streams from the gutter broom hood and
the pick-up headr a portion of the air from the sweeper blower
must be vented. This vent air stream is cleaned by a scrubber.
Preliminary Experiments
Preliminary experiments were done to obtain desiqn informa-
tion for the vent air scrubber system. Several street dust
samples were taken from the plumes dispersed by broom and vacuum
type sweepers. The representative particle size distribution had
a geometric mean diameter of 4 umA and geometric standard devia-
tion of 2.
Predictions of scrubber efficiency based on the preliminary
information on street dust size indicated that to achieve a
minimum of 90% efficiency with an uncharged spray scrubber system
would be impractical due to high water consumption. The required
efficiency can be achieved with a charged system at a liquid flow
rate of 0.4 1/m3 (3 gal/Mcf) if a fine water spray is used (dd
-------
<100um). Experiments on charged spray scrubbing were conducted
to verify and extend the results of previous studies.
Detailed Design
The basic sweeper was analyzed for air flow and dirt convey-
ance. Extra diicting was needed to transport the vent air from
the recirculating airflow system to the scrubber. Hoods were
designed to contain the dust generated by the' gutter brooms, and
ducting was designed to carry this dirt to an appropriate place
in the hopper.
It was necessary to develop a hood arrangement that allowed
normal broom operation, and also reduced fugitive broom emisions
to an acceptable level. A hood that completely encloses the
broom was built, but was abandoned due to its poor performance.
An abbreviated hood, which is similar to an air curtain distribu-
tion manifold, was later designed to cover the forward portion of
the broom. This hood allows brushing on the curb side and oper-
ates with a high velocity suction air flow similar to a vacuum
cleaner. It also uses the interaction of the rotating broom and
street dirt to capture dust. The captured dust is conveyed to
the main hopper with piping and flexible hose. A damper valve in
the pipe section controls the air flow from the hood to the
hopper.
Suction air flow at the pick-up head was increased by vent-
ing a portion of the air at the pressure hose. In order to
design the vent air scrubber and the gutter broom hood, the
required air flow rate was estimated from industrial ventilation
practice to be about 28 to 56 mVmin (1.000 to 2,000 cfm) vented
through the scrubber.
The sweeper has a blower which the manufacturer rated at 280
to 340 mVmin (10,000 to 12,000 cfm). While its capacity proved
to be only 50% of the rating, it was adequate for venting about
30 mVmin and providing effective dust pickup in the hood.
The scrubber was designed for a maximum flow rate of 56
mVmin (2,000 cfm). A wire and rod type particle charger was
built into the scrubber to pre-charge the particles. Downstream
-------
from the particle charger were four pin type spray nozzles. The
spray was charged by induction with the nozzles maintained at
ground potential and a high voltage grid in front of them.
A 15 cm (6 in.) thick knitted mesh entrainment separator was
designed to remove water drops as well as large solid particles.
The air velocity at the upstream face of the entrainment separat-
or was increased by blocking about'50% of the flow area so that
the cut diameter of the entrainment separator would be 3 gmA.
In an actual system, the scrubber could be located inside
the hopper. Fbr purpose of sampling and easy access, the scrub-
ber was mounted on top of the sweeper and water was drained into
the hopper.
The scrubber needed about 0.4 1/m3 (3 gal./Mcf air handled).
For convenience* this translates to a 0.76 m3 (200 gal.) capacity
to .allow once-through operation, and a reasonable time between
refills. The sweeper came with a 0.12 m3 (30 gal.) tank. An
auxiliary tank with capacity of 0.64 m3 (170 gal.) was built and
mounted on the sampling trailer.
The sweeper came equipped with a piston water pump which can
provide 0.4 1/min (3 gpm) at 3,450 kPa (500 psi). For the first
experimental sweeper, it was assumed that the existing water pump
would suffice and that once-through water use would be
acceptable.
The sampling system provides for the measurement of airflow
from the gutter brooms to the scrubber inlet. Particle size and
concentration were measured with cascade impactors. Sample ports
were provided for the scrubber inlet and outlet at locations at
least 8 duct diameters from upstream transitions or bends. Sam-
ple trains of the type used for EPA Method 5 were used. Scrubber
and gutter broom airflow rates were measured with Venturi meters
and were controlled by dampers.
The positive displacement water pump was calibrated in terms
of auxiliary engine speed. Additional utilities (such as the
electrical power supplies for sampling pumps and scrubber charg-
ing, water tank, transfer pump, sampling personnel station, and
platform for sampling trains) were placed on a trailer pulled by
the street sweeper.
-------
EXPERIMENTS
Vent Air Rate
To minimize the scrubber size and water consumption, the air
flow to be vented through the scrubber should be kept at a mini-
mum. However, the air flow should not be so low that dust puffs
occur around the gutter broom and the pickup head.
The minimum air flow needed to be vented through the scrub-
ber to prevent the occurrence of dust puffs was determined when
the gutter broom hoods were turned either ON or OFF.. When ON,
the air flow in each hood was maintained at 0.17 m3/s (350 acfm),
which was the minimum required flow for satisfactory hood
operation.
It was found that the minimum air flow to be vented through
the scrubber to prevent dust puffs with the gutter broom hoods
OFF was about 0.33 m3/s (700 acfm), and about 0.38 mVs (800
acfm) with the gutter brooms ON.
Street Dust Sampling
Full scale operation of the sweeper and auxiliary equipment
was first done in San Diego to observe the modified components in
use, and to provide more data on the nature of street dirt.
Later, as the apparatus was refined, additional samples were
gathered on other streets, and eventually more were gathered in
heavy industrial areas in Los Angeles.
Sampling results showed that the concentration and particle
size distribution vary greatly from location to location. Samp-
ling results agree with visual observations in that dirtier
streets resulted in higher particle concentrations and larger
particles.
Street samples collected indicate a difference in character-
istic size dirt when neighborhoods are considered. The major
distinction noticed between neighborhoods is the amount of traf-
fic; street dirt is reduced in size by continuous traffic.
-------
The vent rate through the scrubber affects the mass concen-
tration at the scrubber inlet. A higher vent rate resulted in a
higher mass concentration.
Sweeping Efficiency Measurement Method
To determine the sweeping efficiency of the sweeper, a
measurement method is required that measures the amount of dust
that can be dispersed into the ambient air by the mechanisms
actually occurring on street. The mechanisms of street dust
removal are (Brookman and Martin, 1979):
1. Reentrainment (by air currents around moving vehicles).
2. Wind erosion (similar to 1, but due to natural air
currents).
3. Displacement (similar to 1, re-deposition near street).
4. Rainfall runoff.
5. Street cleaning.
The first three mechanisms result in airborne dispersal of
street dirt, so the sampling method should measure the amount of
IPM which can be dispersed by these.
Preliminary experiments were performed to measure street
dust density with a brush-type vacuum cleaner, as performed by
Dahir and Meyer (1974). The dust on the street was first
loosened with a brush, and then taken up and filtered by the
vacuum cleaner. The dust density (mass/unit street area) was
determined from filter weight gain.
This method was found to have several deficiencies. The
vacuum cleaner can remove dust which is deposited deeply in
cracks and is not normally reentrained. Further, the amount of
dust vacuumed increases with each pass of the vacuum nozzle;
therefore, there is no logical end point for sampling.
To simulate the re-entrainment mechanisms, a new method was
developed that uses a vacuum cleaner with a modified pickup
nozzle shaped to create a uniform 97 km/hr air flow field at the
-------
street surface. The dust dislodged by the airflow is sucked into
the vacuum and filtered by the vacuum bag.
Efficiency Data
Street sweeping efficiency was determined by measuring the
street dust concentration before and after sweeping. The street
was divided into strips of equal area for dust concentration
measurements before and after sweeping. Each strip was further
sub-divided into the gutter broom area and the pick-up head area.
Street dust is concentrated in the gutter area. The street
sweeper efficiency were calculated from the measured street dust
concentration before and after sweeping. The overall sweeping
efficiency ranged from 80% to 98% in the gutter broom area and
75% to 90% in the main pick-up head area. Large particles were
removed from the street at higher efficiencies. Particles with
diameter smaller than 2 umA stayed on the ground in the gutter
broom area after sweeping.
The performance of the modified sweeper was compared to that
of a regular sweeper visually at a construction site. The modi-
fied sweeper not only drastically reduced the dust clouds around
the sweeper, it also gave a cleaner street. A cleaner street
will reduce the fugitive street dust emission.
Scrubber Efficiency
A few experiments were done in the laboratory to determine
the collection efficiency of the spray scrubber on the sweeper.
It has an overall collection efficiency of 80% on dust with mass
median diameter of 2.0 umA and geometric standard deviation of
2.0 when the scrubber was un-augmented.
CONCLUSIONS
The feasibility of applying SCAT to the control of fugitive
road dust emissions was proven. A regenerative vacuum sweeper is
ideally suited to the SCAT technique, as it allows a discharge
10
-------
stream of air to be cleaned and recycled into the atmosphere.
This feature reduces the size and power requirements for control
of fugitive emissions of inhalable particle emissions.
Gutter broom dust emissions can be substantially improved
with an advanced, interactive gutter broom hood. Since most of
the dirt is in the gutter, this improvement is very significant.
Additional power requirements for the SCAT system are minimal.
Existing standard equipment pumps, blowers, and auxiliary power
are sufficient to do the job.
RECOMMENDATIONS
A superior street sweeping machine has been developed for
reducing particle emissions from paved streets. This street
sweeper has been subjected to a limited testing program in San
Diego and Los Angeles. Results clearly indicate that the sweeper
can eliminate the dust plume during sweeping and give a cleaner
street. However, additional research work is needed to refine
the design and to demonstrate its capability in improving the
ambient air quality. The following are recommended for further
study:
1. Refine the design of the spray scrubber and incorporate
it inside the hopper.
2. Improve the gutter broom sweeping efficiency for fine
particles.
3. Demonstrate the sweeper on city streets and measure the
improvement in ambient air quality.
4. Use a low pressure drop Venturi scrubber to clean the
vent air instead of a charged spray scrubber.
11
-------
Section 2
INTRODUCTION
URBAN FUGITIVE DUST EMISSIONS
Many metropolitan areas are out of compliance with the
National Ambient Air Quality Standards for particulate matter
largely because of urban fugitive dust emissions. A major urban
fugitive dust problem is the dust emissions from paved roads.
Draftz (1978) estimated that 40 to 70% of the total suspended
particles in many urban areas comes from dust particles
redispersed by road traffic. Pitt (1979) estimated that city
streets can contribute from 5 to 50 ug/m3 of particles to the
urban air.
Table 1 shows the primary dust deposition and removal pro-
cesses and rates for roads. Most of the dust deposition comes
from mud and dirt carryout, such as from construction sites and
heavy industrial traffic. Salting and sanding operations for ice
control also can be a major source of dust. About one-half of
the deposited dust may leave the street as particulate air pollu-
tion by either traffic related reentrainment or wind erosion.
Figure 1 further illustrates the street dirt loading rate
for industrial, residential, and open roads. A street dirt
loading of 50 g/curb-m or less is regarded as a clean street and
a loading of more than 300 g/curb-m a dirty street. As can be
seen, a street in an industrial area becomes dirty about two days
after cleaning. Unless the street is cleaned, the dirt will be
reentrained.
STREET SWEEPERS
Three basic street cleaning methods are currently in use:
12
-------
TABLE 1. DUST DEPOSITION AND REMOVAL RATES FROM PAVED ROADS
(AXETALL AND ZELL, 1977)
Deposition
Process
Mud and dirt
carryout
Litter
Biological
Ice control
compounds
Dustfall
Pavement wear
& decomposition
Typical rate,
kg/curb kmday
28.2
11.3
5.6
5.6
2.8
2.8
Vehicle-related 4.8
(incl. tire wear)
Spills <1
Erosion from 5.6
adjacent areas
Removal
Process
Reentrainment
Displacement
Wind erosion
Rainfall runoff
Sweeping
Typical Rate
kg/curb kmday
28.2
11.3
5.6
14.1
9.9
13
-------
E
i
.a
k.
3
O
^
o>
O
Q
<
O
_i
I-
UJ
LU
OC
H
CO
LIGHT
INDUSTRY
'DIRTY*
STREET
RESIDENTIAL
"CLEAN
STREET
OPEN ROAD
DAYS AFTER CLEANING
Figure 1. Street debris accumulation rates.
(URS. 1974)
14
-------
sweeping, flushing, and vacuuming. The existing types of street
sweepers, which use the above cleaning methods, include broom,
vacuum, regenerative vacuum, and water flushing sweepers. They
generally have low-to-moderate efficiency for removing inhalable
particulate matter (IPM) from streets and test data scatter
widely.
Street sweepers and flushers are used in most urban areas..
Traditionally, they have been designed for removing trash as well
as dirt from roadways. Little effort has been directed at
designing street cleaning equipment for collecting fine dust
particles. As a result, conventional municipal street sweepers
are not effective at removing inhalable particles from the
street. This is illustrated in Table 2 which shows measured
collection efficiencies from broom and vacuum sweepers as well as
street flushers. None of these methods are consistently effi-
cient for fine particle control. Sometimes more fine particles
are left on the street after sweeping because large agglomerates
are broken up by the sweeper.
Technology Needs
There are three major functions involved in any street
sweeping system:
1. Dislodgement and containment of dust particles from the
road surface.
2. Conveyance of dust particles to a dust collection hopper.
3. Prevention of fine particle emissions from the dislodge-
ment apparatus, the hopper, and the conveying gas flow.
Existing sweeper systems need improvement in all three functions.
PROPOSED METHOD FOR SWEEPER IMPROVEMENT
The "Spray Charging and Trapping" (SCAT) system (Yung, et
al., 1981) has been developed by A.P.T. under an earlier EPA
contract as a control method for industrial fugitive emissions.
The SCAT system uses air curtain and/or air jets to contain and
15
-------
TABLE 2. CLEANING EFFICIENCY OF STREET CLEANING
METHODS (Sartor et al., 1972)
Type of
cleaning
Broom
Broom
Broom
Broom
Broom
Broom
Flush
Flush
Flush
Flush
Flush
Vacuum
Vacuum
Vacuum
Vacuum
Percent Removal of Material
Particle diameter range/ gm
<44 44-106 106-841 >841
77*
11
1
63
8
9
38
90
29
1
21
34
79
85
17
-136
- 15
34
80
24
40
-13
90
25
16
25
62
86
2
48
62
11
52
62
23
52
3
-171
7
- 3
69
59
75
60
63
65
45
30
1
28
78
- 2
-54
20
1
71
32
73
65
* Negative sign indicates an increase in particle mass
in a given size range after sweeping.
16
-------
convey fugitive particles to a spray scrubber which could be
charged. This system also has the potential to greatly improve
the performance of urban street cleaning equipment.
Figure 2 shows a schematic of the SCAT system applied to
control urban fugitive dust emissions from paved roads. This
system could be added to existing street cleaning equipment or
could be incorporated in the design of new equipment. The unit
operations involved are discussed below.
Dislodge Dust Particle
Brooms, brushes, scrapers, and other conventional devices
are used to "sweep" the street. This dislodges dirt and fine
dust particles from the roadway and sweeps them towards the dust
pickup location. Many fine dust particles are redispersed into
the air at this time.
It is expected that the brooms will adequately dislodge
particles from the road surface; however, this may not be true
with heavily loaded or muddy streets. To aid in dust
dislodgement, high velocity air or water jets can be used.
Contain Dust
Redispersed dust particles must be contained and collected
within the street sweeper system. The SCAT system uses air
curtains to contain the dust within the sweeping system. Fine
water sprays may be used to suppress, collect, and agglomerate
redispersed dust. Either the drops or the particles (or both)
may be electrostatically charged to increase the drop collection
efficiency. Surfactants may also be added to the water to
improve wetability of the dust particles.
Sprays may be injected into the sweeper air intake, into the
hopper, into the exit duct from the hopper (sometimes a cyclone),
and around the sweeper brooms at ground level. Water drops may
be collected in an entrainment separator at the hopper exit, at
the air intake, or by a squeegee liquid pickup device as in some
small-scale commercial sweepers.
17
-------
r
SWEEP
00
CONTAIN
FAN
CONVEY
CLEAN
AIR INFLOW
T
=t=r=sfl WATER
WATER
TREATMENT
IJ
-*-EXHAUS
DISPOSAL
CONTROL CONCEPT
Figure 2. SCAT system for controlling reentrained
dust from paved roads.
-------
Convey to Hopper
Once the dirt and dust are dislodged and contained within
the street sweeper system/ it is necessary to convey the dust and
dirt to the hopper. The movement of the brooms and the use of
scrapers convey a large amount of dirt and trash to the hopper.
However, this is unlikely to be an effective method of conveying
the inhalable dust which is suspended in the air. Very likely/
it will be necessary to have an induced draft air flow into the
hopper. This will aid in the containment and conveyance of
redispersed dust.
Remove Dust From Air
At least a portion of the conveying air flow must be cleaned
before being released from the hopper to the environment.
Depending on the extent of agglomeration and collection of redis-
persed dust/ it should be sufficient to use a good entrainment
separator for final gas cleaning. Additional sprays may be
necessary at the gas entrance to the entrainment separator in
order to collect dust reentrained in the hopper.
An important option is to recycle a major fraction of the
conveying air to the dust pickup point. There are many potential
advantages to this option:
1. The recycle air may not need to be thoroughly cleaned
because it is contained with the sweeper/SCAT system.
Recycling fines may even increase fine particle
agglomeration in the containment zone of the system.
2. The hopper sprays and entrainment separator will only
handle a fraction of the total gas flow. This will
minimize water and power requirements for final gas
cleanup.
3. The recycle air will be moist/ thereby minimizing eva-
poration of the water sprays and consequently reducing
water consumption.
19
-------
Water Treatment
To keep the system size and power requirements as small as
possible, it is desirable to consider the need for, and economics
of, water treatment and recycle. The water requirement is
closely tied to the fraction of air which is recycled, the amount
and size distribtion of inhalable particles which are
redispersed, the extent of agglomeration of inhalable particles,
the entrainment separator design, and the dust loading and size
distribution in the gas which exits the hopper.
RESEARCH PROGRAM OUTLINE
Objectives
The primary objective of this research and development
program was to develop practical means for improving the
efficiency of a suitable street sweeper for the removal of IPM
from urban streets. The applicability of the SCAT system was to
be investigated.
Additional objectives of the program were to:
1. Build a modified street sweeper and demonstrate its
capabilities for urban street sweeping.
2. Develop and use methods for evaluating sweeping
efficiency.
Approach
The general approach of the program was consistent with the
premise that a suitable scrubber could be designed if one knew
what air flow rate and particle collection capability were
required. In other words, much more was known about scrubber
design than about street sweeping in terms of IPM control
parameters. Consequently, the main effort was placed on
determining:
20
-------
1. Background information on street dirt and street sweep-
ing. A literature search and interview of qualified
persons were conducted to determine:
a) The nature of dirt on paved roads, including
information on the source of particulate matter,
the methods of deposition, and various methods of
removal.
b) The prevalent types of street sweepers available,
their operating principals, their costs, and the
potential for improving them.
c) Methods for sampling and analysis of the IPM on
the street and in the air.
2. The street sweeper best suited for the proposed
approach.
3. Design concepts for the modifications needed for IPM
control. The following were considered:
a) Develop possible approaches for improving perfor-
mance and cost.
b) Evaluate technical and economic feasibility through
design studies.
c) Select best alternatives for further evaluation.
d) Identify areas where more information is needed.
4. A tentative set of design criteria to use for designing
the experimental street sweeper, including:
a) Air flow rates required to control dust emissions.
b) The total air flow rate which has to be scrubbed.
c) Particle size distribution and particle concentra-
tion in the uncontrolled effluent air.
d) Utilities limitations.
5. Perform preliminary expeiments.
a) Design experiments to provide necessary design
information.
b) Build apparatus
c) Conduct experiments
d) Evaluate results
21
-------
6. Fabricate sweeper modifications.
a) Functional equipment.
b) Sampling and measurement apparatus.
7. Conduct performance tests.
a) Develop test method.
b) Perform tests.
c) Analyze data.
A limited effort was expended on charged spray scrubbing and
scrubber design. Some experimental work was done to confirm
earlier experiments and to evaluate concepts for atomizers and
drop chargers which could produce smaller charged drops than used
in previous work. The further development of a mathematical
model for particle collection by charged spray also received some
effort.
22
-------
Section 3
BACKGROUND INFORMATION
A literature search and interviews of qualified persons were
conducted to determine:
1. The nature of dirt on paved streets, including informa-
tion on the sources of particulate matter and the
methods of deposition and removal of dust.
2. The prevalent types of street sweepers available, their
cost, and the potential for improving them.
3. Methods for sampling and analysis of the IPM on the
street and in the air.
STREET DUST LOADING AND SIZE DISTRIBUTION
Street dust loading is affected by many variables, such as
meteorological conditions, vehicle traffic, roadway
configuration, and pavement composition. The street dust
accumulation rate can be obtained by performing a material
balance on the street dust:
Accumulation Rate _ Deposition Rate _ Removal Rate (1)
of debris on the of debris on the from the
street street street
Figure 3 shows various processes for street dust deposition
and removal (Axtell and Zell, 1977). The debris on the street
could come from pavement wear, vehicle related deposition such as
tire and brake lining wear, dust fall, litter, mud and dirt
carryout, erosion spill, biological debris, and ice control com-
pounds. The debris on the street may be removed by reentrain-
ment, wind erosion, displacement, rainfall runoff, and street
sweeping.
Vehicular mud and dirt carryover from unpaved areas, such as
unpaved roads, parking lots, construction sites, and demolition
23
-------
NJ
11 Paveraent wear and decomposition
2 Vehicle-related deposition
3 Dustfall
Litter
5 Mud and dirt carryout
6 Erosion from adjacer.t areas
Spills
[8 biological debris
19 Ice control compounds
DEPOSITION
REMOVAL
1 Reontrainnrent
2 Wind erosion
3 Displacement
4 Rainfall runoff to catch basin
5 Strcat cwocping
Figures. Deposition and removal processes. (Axetell and Zell, 1977)
-------
sites, is the major deposition process. Maximum carryover occurs
in wet weather. According to Roberts et al. (1972, 1974), a car,
driven at 16 km/hr (10 miles/hr) on a wet gravel road, collected
approximately 3.6 kg of mud on tires and underbody, and carryover
on tires from a wet unpaved parking lot averaged about 0.34
kg/vehicle.
Sartor and Boyd (1972) measured street dust surface loading
and particle size distribution at sites in five cities. Their
results are shown in Table 3. Dust loadings were found to depend
on:
1. Time elapsed since the last street cleaning or rainfall.
2. Street surface characteristics: Asphalt streets had
loadings that were 80% higher than concrete-surfaced
streets, and streets in fair-to-poor condition had
loadings about twice as high as streets in good to
excellent condition.
3. Public works practices: Average loadings were reduced
by regular, street cleaning and increased during winter
in areas where sand and salt were applied.
The major constituent of street dust was mineral-like matter
similar to common sand and silt. Typically, 78% of the material
was located within 15 cm from the curb and 85% within 30 cm from
the curb.
It is anticipated that future emissions standards will be
expressed in terms of inhalable particles which pose the primary
threat to human health. Inhalable particles have been defined by
EPA as those smaller than 15 umA (in this report, the micrometer
symbol "umA" is used for aerodynamic micrometer).
The fraction of street dust which is in the inhalable size
range was estimated from the data presented by Sartor and Boyd
(1972) and is presented in Table 4. Approximately 0.5 to 2.5% of
the total mass is inhalable.
Reentrainment and displacement due to vehicular motion and
wind are the major removal processes and account for about 50% of
the street dirt removal (Axetell and Zell, 1977). Both processes
remove dust by resuspending the dust into the air. The
25
-------
TABLE 3. STREET DUST SURFACE LOADING
(Sartor and Boyd, 1972)
Land Use
Mean Initial
Accumulation Rate
ka/km/dav
Residential
Low/old/single
Low/old/multi
Med/new/single
Med/old/single
Med/old/multi
Industrial
Light
Me di um
Heavy
105
126
Dust Loading (kg curb km)
Numerical Weighted
Minimum Maximum Mean Mean
34
9
51
73
40
73
79
68
536
367
339
536
1,947
3,386
367
3,386
339
240
251
121
395
734
251
988
82
Commercial
Central Business
District
Shopping Center
64
17
18
339
181
82
82
82
Overall
98
423
Note: There are two curb-km per street-km.
26
-------
TABLE 4. SIZE DISTRIBUTION OF STREET DIRT AND DUST (Sartor and
Boyd, 1972)
PARTICLE DIAMETER
4,800 ym
2,000 - 4,800 ym
840 - 2,000 ym
246 - 840 ym
104 - 246 ym
43 - 104 ym
30 - 43 ym
14 - 30 ym
4-14 ym
4 ym
MASS LOADING,
kg/curb-km
% <15 ymA
MASS <15 ymA ,*
kg/curb-km
Milwaukee
12.0%
12.1%
40.8
20.4
5.5
1.3
4.2
2.0
1.2
0.5
761
1.2%
9.1
Bucyrus
0
10.1%
7.3
20.9
15.5
20.3
13.3
7.9
4.7
0
389
2.5%
9.7
Baltimore
17.4%
4.6%
6.0
22.3
20.3
11.5
10.1
4.4
2.6
0.9
291
2.4%
7.0
Atlanta
0
14.8%
6.6
30.9
29.5
10.1
5.1
1.8
0.9
0.3
121
0.8%
1.0
Tulsa
0
37.1%
9.4
16.7
17.1
12.1
3.7
3.0
0.9
0.1
93
0.5%
0.5
*for particle density of 2.5 g/cm:
27
-------
difference between the two is that in displacement the dust
settles again nearby.
The accumulation of materials on the street has been found
to level off within a period of three to ten days after a rain-
storm or street cleaning (Sartor and Boyd, 1972). This leveling
off occurs when traffic-related removal rates balance traffic-
related deposition rates. The equilibrium is established more
rapidly with increasing traffic speed.
Pitt (1978) determined the rate of the reentrainment from
street surfaces by measuring the accumulation rate of dust on a
thoroughly cleaned street. He then calculated the reentrainment
rate by assuming that the reentrainment rate of debris from
street is equal the removal rate of debris from streets and that
the dirt deposition rate was constant. The amount of
reentrainment increases with time as dirt deposits on the street
(Figure 4).
Pitt (1978) calculated the emission rate of vehicles from
equation (2) and his data and compared it with the data published
in the literature.
Reentrainment rate = Emission rate of x Frequency of (2)
Vehicles Vehicles
(g/km-day) = (g/km-veh) x (veh/day)
Figure 5 and Table 5 show the results. Not surprisingly, the
emission rates published in the literature vary considerably.
A number of studies have measured the size distribution of
suspended particles in the vicinity of paved streets (Axetell and
Zell, 1977; Cowherd et al., 1977; and Bohn et al., 1978). Figure
6 presents two suspended particle size distributions from the
literature. Based on these size distributions, 90% removal of
inhalable particles would require collecting all particles larger
than about 1 to 4 ymA diameter.
The particle size distribution of dust emitted by the roads
was also measured by Pitt (1978). He measured the concentration
28
-------
CO
CO
13
»»
E
1C
I
.a
i_
a
o
^s
O>
liT
I-
<
DC
z
UJ
Z
UJ
UJ
DC
3500
3000
2500
2000
1500
E 1000
500
I II I
SURFACE CARS/DAY
ASPHALT
OIL &
SCREENS
ASPHALT
8,250
200
2,100
_i i j I i i i_
0 4
1 2 20
28
36
44
52
60
68 76
DAYS PAST CLEANING
Figure 4. Rate of reentrainraent of particulates from street surfaces
(Pitt, 1979).
-------
E
I
o>
CO
CO
<
Q.
DC
<
o
CE
UJ
O.
CO
111
H-
CC
3.6
3.0
2.4
1.8
1.2
0.6
VCLEAN'
STREET
"DIRTY*
STREET
120 180 240 300 360
STREET LOADING, g/curb-m
Figure 5. Reentrainment of dust by vehicles as a function
of amount of street debris (Pitt, 1979).
30
-------
TABLE 5. VEHICLE EMISSION RATES ON PAVED ROADS
Reference
Emission rate
g/veh-km
Axetell and Zell (1977)
Cowherd et al. (1977)
Pitt (1978)
Sehmel (1973)
0.12 -
0.41 -
12.4
8.1
11.2
28
31
-------
K
W
EH
W
H
Q
W
^
U
H
60
50
40
30
20
10
9
8
7
6
5
4 _
3 _
0.5 1
5 10 20 30 40 50 60 70 80
WEIGHT % UNDERSIZE
90
Figure 6. Size distribution of suspended dust in
vicinity of paved streets.
32
-------
of particles upwind and downwind of a road and calculated the
difference. The size distribution (Figure 7) is not log normal
and is widely variable. If a log normal approximation were made,
the mean diameter based on the mass of particles would be
approximately 10 pm. It should be noted that approximately 80%
of the particles in the air are smaller than 30 urn diameter
(Axetell and Zell, 1977). The road emission rate could not be
calculated from their data because of unknown meteorological
conditions.
Roadside Concentration of Particles
Some indication of the roadside concentration can be
obtained from resuspension factors presented by Steward (1964)
and Mishima (1964). The resuspension factor is defined as the
ratio of airborne concentration (weight/volume) to the surface
concentrations (weight/area). Values of the resuspension factors
for vehicular traffic usually range from 10~7 to 10~5 /m. With
a "clean" street surface (particulate loading of 30 g/curb meter,
100 Ib/curb mile) the resulting roadside airborne particulate
concentration from auto traffic may vary from 0.5 to 50 ug/m3.
STREET SWEEPERS
Street sweepers and flushers are used in most urban areas.
Traditionally, they have been designed for removing trash as well
as dirt from roadways. Little effort has been directed at
designing street cleaning equipment for collecting fine dust
particles.
Street flushers flood the street with high velocity water
sprays. This method can be effective in some applications, but
generally it does not remove the inhalable size fraction
effectively. Also, the water requirements and sewer capacity
will limit the general application of this method.
Street sweepers can be sub-divided into 3 categories:
mechanical, regenerative, and vacuum sweepers. Table 6 shows the
33
-------
10.0
w
EH
w
fcH
Q
U
H
EH
40 50 60 70
90 95 98 99
NUMBER % UNDERSIZE
99.8
Figure 7. Particle size distribution near road
(Pitt, 1978).
99.99
34
-------
TABLE 6. TECHNICAL FEATURES ON STREET SWEEPERS
Function
Transfer of debris
from gutter.
Mechanical
Gutter broom kicks
dirt from the curb
to the center plate.
The sweeper moves
forward and the
debris comes into
contact with pickup
mechanism.
Regenerative
Same as mechanical
sweeper.
Vacuum
Same as mechanical
sweeper.
Main pick up.
u>
Ul
The rear broom ex-
tends the entire
width of the sweep-
er. It kicks the
dirt to the con-
veyor/elevator
belt.
The vacuum nozzle
extends the entire
width of the sweep-
er. The debris is
stirred by an air
jet and sucked in-
to the hopper.
The vacuum nozzle
is located on the
right side of the
sweeper and is
about 3 ft. wide.
A full width rear
broom windrows
the debris to the
nozzle.
Transfer of debris
to the hopper.
An elevator or
conveyor belt
Transfer of air
flow.
Same as regenera-
tive sweeper.
-------
TABLE 6 (continued). TECHNICAL FEATURES ON STREET SWEEPERS
Function
External dust con-
trol.
Air clean up.
LJ
a\
Water treatment.
Mechanical
Water is sprayed on
the street to wet
the debris. Water
is also sprayed on
the gutter brooms
and rear broom.
None.
The water is ob-
tained from the
fire hydrant and
is disposed of in
the wet debris.
Regenerative
Water is sprayed on
the street to wet
the debris. Water
is also sprayed on
the gutter brooms,
vacuum pickup head
and in the dust
hopper.
The air velocity
is decreased in
the hopper to drop
"large" particles.
Water is sprayed
on the dust in the
hopper. The air is
sucked into the
fan through a cen-
trifugal separator
and expanded metal
filter. The air is
returned to the main
pickup head.
Same as mechani-
cal sweeper.
Vacuum
Same as regenera-
tive sweeper.
Water is sprayed
in the vacuum
head. The air
clean up is simi-
lar to the regen-
erative sweeper.
Same as mechani-
cal sweeper.
-------
TABLE 6 (continued). TECHNICAL FEATURES ON STREET SWEEPERS
Function
Maneuverability.
Mechanical
Three wheel sweep-
ers are very ef-
fective in going
around parked cars.
Four wheel sweepers
are not as effective
as three wheelers
but can travel at
55 mph. They are
used by cities with
large perimeters.
Regenerative
Available in four
wheels only.
Vacuum
Available in four
wheels only.
OJ
Possible modi-
fication to
remove inhalable
particulates.
Will require a
new fan for fine
dust pick up and
cleaning.
Could use the
existing fan for
dust transfer
pickup and clean-
ing.
Same as regener-
ative sweeper.
-------
technical features on these three types of street sweepers.
Table 7 shows information on commercially available sweepers.
A mechanical sweeper (Figure 8) uses a rotating gutter broom
or brush to dislodge the dirt from the road and move it from the
gutter area into the path of a large cylindrical broom which
rotates to carry the dirt onto a conveyor belt and into the
hopper. Water sprays are sometimes used to suppress the dust.
This type of sweeper is available in several designs, including
self-dumping and three- or four-wheel sweepers. Three-wheel
sweepers are generally more maneuverable, but four-wheel sweepers
can travel at higher speeds when not sweeping.
Vacuum assisted mechanical sweepers (Figure 9) use gutter
and main pickup brooms for loosening and moving street dirt and
debris into the path of a vacuum intake. A large volume of
induced air flow sucks up the loosened particles and conveys them
to the hopper. The debris are saturated with water on entry and
settle out in the hopper. The air leaving the hopper passes
through a coarse filter or screen which removes larger particles
and protects the blower, and then is vented.
In regenerative sweepers, the loosened particles are blasted
with air jets and sucked into the hopper. After cleaning with a
cyclone, the air is recycled to the air jets.
DUST PICKUP MECHANISM
Information obtained from manufacturers indicates that
mechanical broom sweepers represent about 85% of the municipal
sweepers now in use. Street debris from the gutter is moved
toward the center of the sweeper by the gutter broom. The
rotating tips of the broom flick the debris toward the center
where it hits the center plate. The big particles drop on the
street while the small particles are dispersed in the air. The
sweeper moves forward and the debris comes in contact with the
rear broom. As the broom rotates and a fiber in the broom comes
into contact with a particle of debris on the pavement, it must:
38
-------
TABLE 7. INFORMATION ON STREET SWEEPERS
Elgin White Wing
Type
Equipment
Cost
1979
Remarks
Mechanical, 3 wheel, $34,000 Has conveyor instead of
1 gutter broom elevator..
Elgin White Wing
with hydrostatic
drive
Mechanical, 3 wheel, $40,000
1 gutter broom
The broom speed is inde-
pendent of sweeper
speed..
Elgin Whirlwind II
Vacuum, 1 gutter
broom
$64,000 Good for cleaning down-
town area.
Ul
10
TymCo 350
FMC 3AH
FMC 993
Regenerative air $40,000
with 2 gutter brooms
Mechanical 3 wheel $32,000
Mechanical 3 wheel $27,000
Hydraulic drive.. Has
some engine HP avail-
able.
Ecoloter Vacu-Sweep Vacuum
$64,000 Available with diesel
engines only..
Mobil TE-3
Mechanical 4 wheel $44,800With 1 auxiliary engine.
-------
HOPPER
REAR BROOM
ELEVATOR
Figures. 3 WHEEL BROOM SWEEPER
-------
DIRECT THROW
INDIRECT THROW
OVERTHROW
Figure 9. VACUUM ASSISTED SWEEPERS
41
-------
1. Break the particles loose from the street surface and
debris.
2. Transfer kinetic energy to the particle to move it from
the street surface to the conveyor.
3. Direct the particle toward the elevator to avoid
"spills" on the ground or in the air.
The variables affecting these mechanisms are:
1. Broom rpm.
2. Sweeper speed.
3. Broom material stiffness.
4. Sweeping pattern (width of rear broom touching the
street, usually 15 cm).
5. Broom constructionspacing of fibers along the broom
length.
Horton (1968) conducted a laboratory study to determine the
effect of these parameters on the efficiency of a street sweeper.
He found that the efficiency increases as broom rpm increases
(Figure 10). The amount of debris increases as the sweeping
pattern is increased (Figure 11). Limper broom materials like
palmyra require larger patterns than stiff materials like wire
brushes to obtain the same street cleaning efficiency (Figure
12).
STREET CLEANING EFFICIENCY
Street cleaning reduces the surface loading of dust and
thereby decreases the reentrainment rate. The cleaning
efficiency of a street sweeper depends on many conditions, such
as the character of the street surface, street surface dust
loading, particle size, and types of sweeper. Sartor and Boyd
(1972) studied the effectiveness of a mechanical sweeper for
various particle sizes and dust loadings. Pitt (1979) measured
the performance of a mechanical sweeper and a vacuum assisted
mechanical sweeper in different areas in the city of San Jose.
Their results are shown in Tables 8 and 9. The conventional
sweeper is effective in picking up litter larger than 0.63 cm
42
-------
UJ
500
400
300
H
A4
en
ffi
200
100
10
60
20 30 40 50
. % DEBRIS REMAINING
Figure 10. Debris picked-up versus brush speed (Horton, 1968).
70
-------
H
«
CQ
&a
p
D
D
H
cn
W
^
w
60
50
40
30
20
10
1 ,1
'' t 1 '
tLtfci
i ; , j J-j--p -p-(-
--f- p
-|_i __ ^ ^
-4- [' | t ; 'l-|-{- 1 -J--L- --j__|_ I 1 | | LL__-j j i__j_ . -j_
-"Sir:: if:::: Wire Broom
---M-:x|j:^|:^::: Irregular Surfaces
±:;:|||S;;p;pS^;;i:±;:::;|
:44ti:iix:::±i:::?^:-::±::::::||
Wire Broom ^:§:±F::::±[f :
Flat Surface :::±-SS::::S-:-3
:::i::±::S::i*::q
EEEEEE;;Ep::EE^|EE;!;EEEEEEE;EihijE:EEE
+ T- T T T " 35 ~~~
£_- -T 5i SX
- -± --i !s, 3
pj-- -j L_j_ J 1__ . __j_L__ _S^
J 1- LJ .... 4*~~1 p- 1 t !i*-4-
~T~ ~ -i h '-- ..-_ t-_-_l__ «^
---il-i' jt 4-
I^I.X_. ~^~~ --X-1- -.
PT-H--J- 1 H-? IT i rrr i n ! 1 1 1 1 1 p
--:::::-±:-g- Palmyra :-
^-| 4r^;i Broom ::^
1 i (III1
. . JJ riC~T I-T" ' 4"" "-L---
:::::|:^::::±S:i::::::|::+::
i | 1 !|!l 1 I fW l| | J
1 -^'"
, , , . . 1 ,
::|;;l-^l:|::i;:S::|:^:
_ 1 r , 1_|- 13
UJt--r-3^- I r- 1 --
2i r i, . .
_JZ__S^__.^L A. «_^ _.
10
PATTERN, inch
Figure 11. The effect of pattern on residual
debris (Horton", 1968) .
44
-------
o
c
H
W
EH
EH
<
CU
cn
en
H
2
Cfl
14
13
12
11
10
7
6
5
4
3
1
0
8 10 12 14 16 18 20 22 24 26
Figure 12.
STIFFNESS RESILIENCE INDEX
The sweeping pattern versus stiffners
for synthetic broom fibers (Horton,
1968) .
45
-------
TABLE 8. SWEEPER EFFICIENCY WITH RESPECT TO PARTICLE SIZE
(Sartor and Boyd, 1972)
Particle Size (urn) Sweeper Efficiency.
More than 2,000 79
840 - 2,000 66
246 - 840 60
104 - 246 48
43 - 104 20
Less than 43 15
Overall 50
46
-------
TABLE 9. TOTAL SOLIDS STREET CLEANER REMOVAL EFFECTIVENESS
BY PARTICLE SIZE (Pitt, 1979)
Study Area and
Particle Size
Range
<»)
Troplcaaa-Cood
Alpha It
>6370
2000 6370
850 2000
600 850
250 » 600
106 * 50
45 106
<45
11 size*
Reyes-Good
Asphalt
>6370
2000 * 6370
850 * 2000
600 - 850
250 - 600
106 250
45 * 106
<45
11 sizes
Reyes-Oil
ad Screens
>6370
2000 * 6370
850 * 2000
600 * 850
250 * 600
106 250
45 * 106
<45
11 size*
Downtown-Cood
Asphalt
>6370
2000 - 6370
850 - 2000
600 850
250 - 600
106 » 250
45 * 106
<4S
11 sizes
Downtown-Poor
Asphalt
>6370
2000 6370
850 - 2000
600 - 850
250 - 600
106 - 250
45 106
<45
11 sizes
Total Solids Initial Loading
(Ib/curb-islle)
Mean
15
15
21
15
42
50
51
16
220
18
38
54
28
85
83
76
21
400
73
270
270
160
480
380
270
63
2000
14
19
25
14
48
56
57
9.8
240~
89
170
180
85
270
270
230
58
1400
Mia.
9.5
10
13
8.2
19
22
24
7.0
120
6.0
10
16
9.2
39
45
34
13
170
13
77
170
100
320
280
160
40
1200
Max.
36
24
42
42
81
80
70
24
350
27
58
87
44
120
100
100
34
550
120
450
350
200
600
540
380
140
2700
Total Solids Reooval "
(I)
Mean
50
46
47
53
46
41
40
19
43
54
39
35
35
31
26
23
8.3
31
36
24
6.0
4.0
3.3
4.0
3.1
-12
8.1
53
42
39
38
36
33
22
41
34
38
51
42
41
42
39
33
28
40
Kin.
9
28
22
41
14
6
21
-54
Ti
- 8
13
8
12
14
11
-12
-44
14
20
- 5
-16
-10
-16
-20
-30
-47
- 6
Max.
75
68
74
79
63
58
54
64
60
69
5
5
5
4
4
5
48
47
58
47
23
20
18
25
25
24
22
Not. enough saaples were collected to obtain oeanlngful loading ranges.
"Sweepers were 4-wheel mechanical and 4-wheel vacuum
assisted mechanical sweopers.
47
-------
(0.25 in) diameter. However, its efficiency drops sharply as
particle diameter decreases.
STREET DUST SAMPLING
To determine the reentrainment rate or the street sweeper
sweeping efficiency, the street surface dust loading needs to be
measured. Sartor and Boyd (1972) measured the street dust
loading by sweeping and water-flushing the sample area. Cowherd
et al. (1977) and Pitt (1979) used vacuum cleaners to suck up the
dust from the road surface. Pitt (1979) claims that dry vacuum
sampling is capable of removing 99% of the particles from the
street surface.
48
-------
Section 4
PRELIMINARY EXPERIMENTS
Dust is dispersed by the gutter broom and the pickup nozzle
head of the sweeper during sweeping operations. Particle size
distribution of the dust is an important variable required to
design the SCAT system, and is a function of several factors;
amount of dirt on the road, size distribution of dust on the
road, type of road surface, moisture content of the dirt and
vehicle operating conditions. Very little information is
available in the literature on the dust dispersion mechanisms and
the size distribution of the redispersed dust.
To obtain design information for the spray scrubber, the
particle size distribution and concentration of the street dirt
redispersed by the street sweeper brooms was experimentally
determined. The following sections presents the results on dust
dispersement.
AIR FLOW NEAR BROOM
Mechanical broom sweepers represent approximately 85% of the
municipal sweepers now in use. Approximately 65% are 3-wheel
designs and 20% are 4-wheel. For this reason, the initial
decision was to modify a mechanical sweeper and the first series
of experiments were done on a 4-wheel mechanical sweeper (Mobil
model M6).
Street debris from the gutter is moved toward the center of
the sweeper by the gutter broom. The rotating tips of the broom
flick the large particles toward the center and the small
particles are dispersed in the air flow near the broom. The air
flow is due to:
1. Broom rotation
2. Sweeper motion
3. Wind
49
-------
The air flow near the gutter and rear broom was measured
with a hot wire anemometer on a stationary 4-wheel mechanical
sweeper and is presented in Figure 13. The wind velocity was
negligible. The air velocity at the broom surface is almost
equal to the tip speed of the fibers. The velocity decreases
steadily away from the broom. The velocity decreases to 10% of
the maximum within 15 cm from the broom. The gutter broom is
inclined at 10° to the horizontal to obtain a "strike" pattern on
the curb. Near the edge above the ground the air velocity
decreases to 10% of the maximum within 30 cm away from the broom.
REDISPERSED STREET DUST CONCENTRATION AND SIZE DISTRIBUTION
The mass concentration and size distribution of dust re-
dispersed by the street sweeper were determined with cascade
impactors and filters. The ambient air was sampled by cascade
impactors located at the curb before and during the sweeper
passed by. Three nucleopore filters were clamped under the
sweeper between the gutter broom and the rear broom. After
sampling, the filters were washed and the particles were analyzed
by a Coulter counter.
The results from the Coulter counter analyses of the three
filters are presented in Figure 14 along with that reported by
Pitt (1978). The mass median diameter of dispersed dust measured
by A.P.T. ranged from 2.5 to 3 urn physical diameter. Only 5 to
10% of the dust is smaller than 1 urn. No particles larger than
10 urn were redispersed and collected on the filters. The three
filters collected a total of 7.0 mg which corresponds to an
average dust concentration of 313 mg/m3.
The cascade impactors collected insignificant amounts of
dust. This was a result of the short sampling time and windy
conditions.
The experiment was designed to obtain information on the
mass and size of particulate which are redispersed from the road
through the actions of the sweeper brooms. The loading of dirt
50
-------
6.0
,REAR BROOM (120 RPM)
5.0
a) 4.0
*.
E
o
_J
ui
>
-------
100
M
I«
00
o;
<
o.
50
10
0.5
0.3
^
.
\ 1 !
Pi
A-
- : .-'.'-......: -u^ix"
_,.. ; _;>.
V -----
? ; - '; -\. :^;, : ..
.-r-"^ :
-- -
^><
- - !
[ ._
:_.._j
_j _^x
r --_-j _-?j£Ty
-.--- ...-.-.::-.k~_ ;~~;
-_- .:::"-:^:_.:._.:V-^.r
- - -
^_1 * M
-:- n
i ' !
tt * (Mg.'78)
3.T. (Sept. '79)
: ; ' i : : ' :
: : "
. - - i . - . .
" .
^S"
:-;?
^..
-:-_: : i. :-: : :
.-:.:.: rf::-: :
--:: .{..: :.:
-: .---::.'-:-..: ^
g^ff--;-
.: : :.:|....::
r:-"":'_
r r
. .
:.:
^
--^>-
iS^~"
:-.-7
,- ..."
..... j ....
.'.-'.
^^sT
.-..::: :--:
^
:-: :
^ - - -
/**»
i LAw>
-^xf -"
^
;;:f;i-:ri
j r '.
I
, _
1 " !
j :
^
i ' '
-><
:: : :>;?
f^7
X*-
K-F1
Ufep-
. ..
. : ~ :
. : : I : .
::.;k::-
.:j" .::.
! .
1
,
^
:-X:
'
^r T
ix&^i
^^ .
T^f1
DRY
:.-: . :
..""
i \
i I
i
-
.
_f^: :
WE
T STRE
^ ^
DUST BEHIND
SWEEPER
^Dl
BE
'-f SV
k--DRY STREET ':' :':.:
^
STRtE
. : :
.::
- - -i-,-..., -.- i ..
ass % undersize was calculated from
umber % undersize data.
: : f . :.:
5
-i : :
T .: :
10
= ;:;=::;;!:
20 30
'."' !':i:
i:- \
40
*
. . . ! .
I
:
:x.:ii ' 1
50
60
~^~
^4
:-: . 1 .: : .
r-^rf- :;:
.':':-'
1ST -
:HIND
IEEPER-
iC: " :
: : . i : :
"iT7 T_"-"_- ^
^<^:!-:
^j,^^ DUST UNDER
:^lj^ SWEEPER AT 3 L
i LOCATIONS
...
....
. . . . "
i
70
..._-.. .
:: :
: -. r- -\ : -'-
.'.,.". \ .',~ '. -
: : - ~\ : :
- : -. - :
. : 1
'
80 90 95-
MASS % UNDERSIZE
Figure 14. Particle srze distribution of dust dispersed by a
mechanical street sweeper.
52
-------
on the street was not measured and no data were available to
correlate the degree of dispersement by the broom.
Photographs of the dust cloud around the gutter broom, under
the sweeper/ and behind the rear broom were taken to show their
location. Visually, the sweeper brooms appear to be very good at
redispersing the fine dust particles. Some dusty air was
observed to seep into the driver cab. The street side gutter
broom spreads the dirt towards the center of the road surface
after a sweeper pass.
Buchwald (1967) reported an average concentration of 2 x 10'
particles/m3 (dp < 5gm) in the sweeper cab for an "average"
street sweeping condition. The cab was partially enclosed and
the dust seeped through the cracks. The particle concentration
under the sweeper in this experiment was 2.8 x 10lo particles/m3.
The sweeper did not use water sprays to suppress the dust in this
experiment.
Pitt (1978) reported data on the size distribution of dust
dispersed behind a broom sweeper. His results are plotted in
Figure 14. He found that the mean particle size increased when
the operator swept on wet ground.
The redispersed dust measured by Pitt (1978) was larger in
diameter than that measured in the present study. A Coulter
counter was used to analyze the particles collected on the filter
in this study. Agglomerates of dust particles in the air may
have broken down to primary particles during Coulter counter
analysis. Thus, the actual mean particle diameter may be larger
than that shown. Differences in sampling location, street
surface, dust concentration on street may also have contributed
to the difference in measured size.
To determine the size distribution of dust dispersed by the
sweeper as it exists in the air, the sampling system shown in
Figure 15 was built. The system consisted of a pre-cutter to
remove large particles, a cascade impactor (University of
Washington, Mark III) a rotameter, and a portable vacuum pump.
The sample flow rate was measured by the rotameter and was
checked against that calculated from the pressure drop across the
53
-------
Ul
*>.
V i
1
TO
*- EXHAUST
1.
2.
3.
4.
5.
6.
PRECUTTER
CASCADE IM.PACTOR
VALVE
ROTAMETER
PRESSURE GAUGE
PORTABLE VACUUM PUMP
Figure 15. Schematic diagram of sampling equipment for particle size distribution data.
-------
irapactor. The pump power came from a storage battery.
The sampling system was calibrated and tested on the Mobil
broom sweeper. However, before sampling was scheduled to start,
word was received from EPA that they would prefer to modify a
vacuum sweeper because more and more municipalities were switch-
ing to this type of sweeper. For this reason, no additional
sampling was done on the Mobil sweeper. Instead, sampling was
done on a Tymco Model 600 regenerative vacuum sweeper owned by
the City of Anaheim, California (descriptions of the Tymco sweep-
er are presented in the next section).
Air samples were taken near the gutter broom and from the
pressure hose of the Tymco street sweeper. One cascade impactor
was mounted on the centerplate of the sweeper for sampling in the
gutter broom area. Another cascade impactor was mounted on a
plate covering the porthole on the pressure hose and a sampling
nozzle was inserted in the hose to draw air samples. A light
industrial area was chosen for field testing. The dust was not
suppressed by water sprays. The sweeper travelled at
approximately 9 km/hr (5 miles/hr) on the street. The sampling
was conducted on a sunny day, but it had rained on the evening
before the test. The average density of the street dust was
determined in the laboratory with a pyncometer to be 2.46 g/cm3.
Dust was sampled for 15 minutes in the gutter broom area,
and for 5 minutes in the pressure hose. The particle size
distribution of dust sampled are plotted in Figure 16. The
cumulative mass concentration data are plotted in Figure 17. The
data are compared with previous tests and data reported by Pitt
(1979), who measured the particle size distribution with a
particle counter. His data are converted from number percent to
mass percent by assuming a density of 2 and 3 g/cm3 for the
street dust. A wide range of values of mean particle sizes and
cumulative mass concentrations of dust dispersed by street
sweepers have been measured. The differences are mainly due to
variations in the following three parameters:
55
-------
U1
(Ti
P-
M
tH
00
40
30
20
10
5
4
1
- -
-
-
1
. 1
DUST1
VA(
\ \ ' '
ill
i i
....._ .
!
l-r-
! ' i
i i ,
i
j i
1
1
* DISPERSED BY
:UUM SWEEPER
i
r
i _ .
i
_ ..
. _j
X
fl
^~
\
i
f
t-
--
T i :
_
SP^
- j i ; l
i M
i ,
i i i
_ T _
i |
i
' ''":' ' r/7 :
f/ : . ;
/
#fc
X .
\
x
^
' ' i i
1 : : I
' i
, 1 ! .
X^
i _!
l-'i!
i i ! i
Ittt
I ' '
! ! i
1 ; 1 1
V4
- ' ' <
"-<
ill;
1 1 1
i ' i '
f
t ' i '
r| i i .
i^
: I;
, :
: I X
""X^
iiii
-rr~
j '
'J\£
; ; ...
i ' '
4ii
! ' 1
'-S
s
i1:;
i i
^ 3
x
111;
i : i
: i !
i! !
1 1 i
" ' i
! l
n
|;
X
1 i i
I i ;
*" " "^
. : :
x
X-
7 n
: 1 !
i S
:p = 2.0 g/cm3
' 1 i
X
x
-----.'-
/
'
-:- -'-
f ' J
!!i
. ;
; i '
! ' i
i . j
i '
Ix
:;-; li-i
\
-
::!
-
"" "r"
\-.\\
HN
! : i
i
DUST*
SWE
--
DUST
SWE
i
O
n
A
A
* Rained previous night before testing. :-~-
** Calculated from number % to mass % by
assuming a particle density p
i; ' ;-; i 'i ~\ r i ' i -r:'T~T
1 :
i '.
_i_._:
*BEHIN
EPER,
UNDER
EPER (
Run #
1 1
2 [
3 1
4 I
5 I
i :
;.: : '
D BR
PITT
: ' ;
; 1
1
| ;
i i
OOM
(1979)
1
BROOM
A.P.T.)
;
Location
5ressure Hose
Jressure Hose
Jressure Hose
Jnder Sweeper
Jnder Sweeper -
i
0.1 0.2 0.5 1
10 20 30 40 50 60 70 80
MASS % UNDERSIZE
90
95 98 99
99.9
Figure 16. Particle size distribution of dust dispersed by a vacuum (regenerative air)
sweeper.
-------
700
r~Ti I i i
VACUUM TYPE SWEEPER
A.P.T.*
BROOM TYPE SWEEPER
(UNDERSWEEPER)
A.P.T.
UNDER SWEEPER
PRESSURE HOSE ^
BROOM TYPE SWEEPER 1=
(BEHIND SWEEPER)
PITT (1979)
p = 2.0 g/cm3
* Rained mght /..;:;.
i_ _ _t _ j. i* 111- --i
before testing rqr
0.5 1.0 2.0 5.0
PARTICLE SIZE UNDERSIZE
10
20 30
Figure 17. Comparison of cumulative mass concentration of dust dispersed
by street sweepers. (Dry sweeping).
57
-------
1. Water: The mean particle size of dust in these tests is
considerably higher than previously reported data. The
rain on the previous night agglomerated the dust and
may have washed some of the fine dust away.
2. Street type: The size distribution of dust on streets
varies with the street location. The tests were conducted on
different streets which had different types of dirt.
3. Type of sweeper: Broom type and vacuum type sweepers
were used to collect data presented in Figure 16 and 17. The
sweepers have different dust pickup mechanisms which may alter
the particle size distribution.
58
-------
Section 5
SWEEPER MODIFICATIONS
TYMCO MODEL 600 STREET SWEEPER
A Tymco Model 600 vacuum (regenerative air) street sweeper
(Figure 18) was purchased for upgrading under this contract.
Figure 19 shows a schematic diagram of the dust pickup mechanism
of the sweeper. The gutter broom moves the street debris from
the curb toward the center and the blast/vacuum pickup head which
extends almost the entire width of the sweeper. The blower
compresses air and forces it downward through the pressure hose
and into and across the pickup head, creating a full width high
velocity blast that lifts dirt and debris from the street toward
the inlet. The blower creates a vacuum in the hopper that causes
debris to be sucked up through the vacuum inlet and hose into the
hopper, where the air loses velocity and heavy or large debris
fall to the hopper floor. The air is drawn through a screen to
remove paper and leaves, and then enters a centrifugal separator
where large dust particles are removed and thrown into the hop-
per. The air is then sucked into the blower to start another
cycle.
A positive pressure is maintained at the pickup head blast
orifice, and a negative pressure in the pickup head vacuum inlet
and in the hopper. The air flow rate and the static presure in
the system can be varied by changing the blower speed. High
blower speeds are used to pick up heavy materials and low blower
speeds are used to pickup finer materials. Further, the vacuum
in the pickup head can be increased by bleeding air out of the
pressure hose. It is called the leaf bleeder system as it is
used when picking up leaves.
Several nozzles are used in the hopper to wash the hopper
during dumping, and one nozzle is directed at the gutter broom to
suppress the dust clouds. The Tymco sweeper comes equipped with
a 0.12 m3 (30 gal) water tank and a pump with capacity rated at
59
-------
HOPPER
AUXILIARY ENGINE
PICKUP HEAD GUTTER BROOM
SCALE 1 cm^O.5 m
Figure 18. Side View of TYMCO Vacuum (Regenerative Air) Sweeper
-------
DIRECTION OF AIR FLOW
VACUUM
HOSE
-O
HOPPER
1
PICKUP
HEAD
CENTRIFUGAL
SEPARATOR
-FAN
O-
LEAF
BLEEDER
-Jt
PRESSURE
HOSE
BLAST ORIFICE
Figure 19. Schematic diagram of TYMCO Model 600 regenerative air system
61
-------
11.4 1/min at 4.2 MPa (3 gpm at 600 psi).
CONCEPTUAL DESIGN
Figure 20 shows the process design concept for improving the
street sweeper performance. Containment and conveying of the
dispersed dust is already accomplished to a large extent by the
regenerative vacuum system of the Tymco sweeper. However, dust
clouds were observed to occur at the gutter broom and the pickup
head areas, and neither were controlled.
Controlling the dust clouds in the gutter broom area with
sprays is not effective. A better method would be to enclose the
gutter broom and vent the enclosure to the hopper. Since the
hopper is under vacuum, in-bleed air will convey the contained
dust to the hopper.
Dust clouds in the pickup head area were observed to occur
when the pickup head travels on uneven street surfaces, such as
pot holes and pebbles on the street. Pickup head dust clouds are
eliminated by increasing the vacuum in the vacuum hose which
causes an increase in the inward flow to the pickup head.
The vacuum in the vacuum hose could be increased by venting
a small fraction of the recirculating air stream from the pres-
sure hose. Venting from the pressure hose reduces the static
pressure in the pressure hose, which in turn raises the vacuum in
the hopper and the vacuum hose (because the developed pressure of
the blower remains constant).
The vent air contains large quantities of inhalable parti-
cles, so it needs to be cleaned before discharging to the atmos-
phere. In the conceptual design, a charged spray scrubber is
used because it can handle both dry and wet particles, such as
mud.
PRELIMINARY EXPERIMENTS
Preliminary experiments were done to generate design infor-
mation such as the amount of air to be vented and the drop size
62
-------
a\
LO
GUTTER
BROOM
AIR FLOW
o
)W
o
t
HOPPER
VACUUM
HOSE
PRESSURE
HOSE
PARTICLE
CHARGER
SWE£PER
PUMP
PICK UP
HEAD
TO
ATMOSPHERE
AIR
FLOW
t
.BYPASS
,DROP
CHARGER
Jk SPRAY BANK
FROM
FIRE HYDRANT
FRESH
-WATER
TANK
^ENTRAINMENT
SEPARATOR
Sr
SLUDGE
DISCHARGE
LINE
Figure 20. Process diagram of improved street sweeper.
-------
and liquid/gas ratio for the spray scrubber.
Air Flow Characteristics of the Sweeper
The sweeper air flow characteristics had to be measured in
order to determine the air flow rate to be cleaned by the scrub-
ber, and the air flow available to convey dust.
Air flow rates circulating in the sweeper were measured in
the pressure and vacuum hoses. The pickup head was in the sweep-
ing mode during the measurements. Data were collected with the
auxiliary engine speed between 1,500 and 2,000 rpm. A standard
pitot tube was used to measure the velocity pressure.
The air flow rates and static pressures in the pressure
hose, vacuum hose and the hopper are reported in Table 10. Data
indicate that the air flow rate circulating in the sweeper in-
creases with the engine speed. Also, the vacuum in the hopper is
increased when air is bled from the pressure hose. When the leaf
bleeder valve was full open the vent air flow at 1500 rpm was
about 33 AmVmin (1,200 acf m), and the vacuum in the hopper
increased by 7 cm W.C. at an engine speed of 2,000 rpm and by 3
cm W.C. at 1,500 rpm.
The required vent air flow rate was estimated from indus-
trial ventilation practice to be 28 to 56 mVmin (1,000 to 2,000
cfm). Therefore, it seems that the existing blower will suffice.
Spray Scrubber
The dust dispersed by the sweeper is contained and conveyed
to a spray scrubber. For designing the spray scrubber, the dust
was assumed to have a size distribution similar to that measured
under the broom sweeper, i.e. dpg = 4 umA and og = 2.0.
The spray scrubber needed to clean the vent air stream was
established through a series of design studies. As the first
step, the uncharged spray model by Calvert et al. (1975) was
used:
64
-------
Table 10. Air Flow Characteristics of the TYMCO Sweeper
No.
Item
Leaf Bleeder Valve
Closed Open
Engine Speed (rpm) Engine Speed (rpm)
2,000 1.500 2,000 1,500
1 Static pressure in
pressure hose
(cm W.C.) 26.7
2 Static pressure in
vacuum hose (cm W.C.) -16.0
3 Vacuum in hopper
(cm W.C.) -16.0
4 Flow rate in pressure
hose
(Am3/min) 141.6
(NmVmin) 145.2
5 Flow rate in vacuum
hose
(AmVmin) 133.4
(NmVmin) 131.4
6 Flow rate vented at
the leaf bleeder
(AmVmin)
(NmVmin)
16.0 13.3
-8.9 -25.1
-8.4 -23.2
106.6
108.2
97.9
97.0
169.5
165.4
34.8
28.2
6.1
-14.7
-11.2
138.4
136.4
33.1
26.9
65
-------
1.5QL
Ptd = exp / n f& dx (3)
where: Ptd = particle penetration for diameter "dpa", fraction
dd = drop diameter, cm
QG = gas flow rate, cm3/s
QL = liquid flow rate, cm3/s
Rd = drop range or distance traveled by drop relative
to the gas, cm
fA = fraction of gas flow cross-section covered by
sprays, fraction
n = instantaneous single drop collection efficiency
x = coordinate in gas flow direction, cm
A representative set of predictions based on this model is
shown in Figures 21 and 22, plots of particle penetration versus
liquid/gas ratio, with particle aerodynamic diameter as the para-
meter.
Both plots are based on experimentally determined single
sphere collection efficiency (Walton and Woolcock, 1960), 300 urn
diameter water drops, and 50 cm drop range. The drop range is
the distance traveled by a drop relative to the air and is lim-
ited by the drop trajectory or by the dimensions of the scrubber.
Figure 21 is for an initial drop velocity of 30 m/s relative to
the air and Figure 22 is for 20 m/s.
Overall particle penetration, Pt, was predicted for an
uncharged spray scrubber with the assumed inlet particle size
distribution. A water flow rate of about 1.6 1/m3 (12 gal/Mcf)
would be required to attain Pt, = 0.1 with 300 urn drop diameter
and 30 m/s initial velocity. Since the water holding capacity of
a street sweeper is limited, the water flow rate is too high to
be practical. It is clear that the spray scrubber must be
augmented. The approach taken in this study was to use
electrostatic augmentation to enhance the collection of fine
66
-------
30
20
0 0.1 0.2 0.3 0.4 0.5
QL/QG,
Figure 21. Predicted penetration for 300 vim diameter spray drop
scrubbing.
67
-------
UJ
UJ
CL.
O
II
O
ceL
D-
W + W Exper. Effic.
30
20
0.1 0.2 0.3 0.4 0.5
QL/QQ, i/m3
Figure 22. Predicted penetration for 300 urn diameter spray drop
scrubbing.
68
-------
particles (dpa <3 pmA).
The effect of electrostatic augmentation of particle collec-
tion by sprays can be predicted if one can account for the com-
bined influences of electrostatic and inertial deposition.
Nielsen (1974) presented the results of his computations of
predicted particle collection efficiency for single drops for the
cases of:
1. Inertial impaction (NP/ND, Neutral Particle/Neutral
Drop) .
2. Coulombic attraction (CP/CD).
3. Charged particle image force (CP/ND).
4. Charged collector image force (NP/CD).
Nielsen's plots of collection efficiency against inertial
impaction parameter and his electrostatic deposition parameters
can be used to predict the efficiency of a spray drop at various
points along its trajectory. It can be shown that coulombic
attraction is the only mechanism which could cause a significant
increase in the collection efficiency of an electrostatically
augmented spray scrubber. Consequently, we have predicted the
effect of Coulombic attraction on a spray scrubber and compared
the predictions with our experimental results.
Nielsen's Coulombic attraction parameter, KC, is defined as:
. Qc QP C' <4)
Sir2 kf d« dp UG ur
where KC = coulombic attraction parameter, dimensionless
Qc = charge on collector, Coulombs
Qp = charge on particle, Coulombs
C* = Cunningham slip factor, dimensionless
kf = dielectric constant of gas, 8.854 x 10" F/m
dc = collector diameter, m
dp = particle dimeter, m
UQ = 9as viscosity, kg/m-s
ur = gas velocity relative to collector, m/s
69
-------
The relationship between particle collection efficiency,
"Kc", and "Kp", as computed by Nielsen is illustrated in Figure
23. Note that the values of "Kc" are negative, signifying that
the particles and drops are oppositely charged. It can be seen
that the greatest effect of Coulombic attraction occur when "Kp"
is small, which corresponds to small values of relative velocity,
"ur." Since "Kc" is inversely proportional to "ur", it increases
as "K_" decreases, thus intensifying the predicted influence of
Coulombic attraction.
Average Drop Efficiency
The efficiency shown in Figure 23 is an instantaneous value,
while the spray scrubbing model requires accounting for particle
collection over the total path the drop travels. Drop traject-
ories were taken from the computations of Walton and Woolcock
(I960) and are shown in Figure 24, a plot of drop velocity vs.
drop "range" (i.e., the distance traveled by the drop) for drops
of several diameters and with an initial velocity of 30 m/s. A
100 urn diameter drop sprayed into air at a velocity of 30 m/s
would travel 30 cm relative to the air, while a 200 ym diameter
drop would go about 90 cm.
Given the drop velocity at all positions along its range and
the instantaneous efficiency correlation of Figure 23, collection
efficiency for all points on the range can be computed. Figure
25 is a plot of predicted particle (dpa = 0.6 ymA) collection
efficiency vs. drop range for charged particles with charged
drops (CP/CD) and for neutral particles and drops (NP/ND). The
two sets of curves shown are for 100 urn and 250 pm diameter
drops.
The charge level on the drops was computed from experimental
data of Yung et al. (1981), which indicated a charge of approx-
imately 5 x 10" C/g for drops of 200 urn to 500 pm diameter, with
induction charging. For 250 um dia. drops this is about 15% of
the Rayleigh limit, while for 100 um dia. drops it is only 3.7%
of the limit. The particle charge levels were computed for a
70
-------
OH
-------
o
o
LU
O-
o
a:
Q
DROP RANGE, cm
Figure24. Trajectories for water drops in air.
72
-------
O
O
dd = 100 ym
CP/CD
20 30
DISTANCE TRAVELED, cm
Figure 25. Collection efficiencies of neutral and charged drops
for 0.6 umA diameter particles.
73
-------
corona charger with a field strength of 4 x 10s V/m.
Figure 25 shows that the greatest effect of charging occurs
after the drops have slowed to the point where inertial impaction
becomes unimportant. Since the average efficiency is obtained by
integrating over the drop range, it is strongly influenced by how
much of the drop range is considered. In the scrubber configura-
tion envisioned for the street sweeper, the drop range, Rd, is
about 50 cm. This is the distance the drop would go from the
point of atomization to the scrubber wall.
A 50 cm range limit presents no problem for a 100 jjm dia.
drop, whose range would only be 30 cm if its initial velocity
were 30 m/s. For a 250 ym dia. drop the 50 cm range limit
eliminates the most effective part of the drop trajectory. This
fact has obvious importance in the designing of a charged spray
scrubber, as will be discussed later.
Average efficiencies for 100 gm and 250 gm dia. drops were
computed by integration over drop trajectories and are given in
Table 11. Also shown in Table 11 are the computed penetrations
for a spray scrubber with a water/air ratio of 0.4 1/m3. Pene-
trations were computed from:
3 Rd QL
Ptd = exp -
(5)
where Ptd = penetration for a given particle size, fraction
Qd = liquid flow rate, m3/s
QG = 9as flow rate, mVs
n = average collection efficiency over drop range, fraction
Note that the range, Rd used is the smaller of the drop range and
the range limit. Thus, Rd = 30 cm for dd = 100 urn and 50 cm
(limit) for dd = 250 pm.
The influence of drop diameter is clearly shown. Despite
their smaller range, 100 um dia. drops would be much more effec-
tive than 250 um dia. drops.
74
-------
TABLE 11. AVERAGE COLLECTION EFFICIENCIES
OF DROPS AND SPRAY PENETRATIONS
ui
V
ymA
1.0
0.6
1.0
0.6
1.0
0.6
dd>
ym
250
250
250
250
100
100
Rd'
cm
50
50
50
50
30
30
V
m/s
20
20
30
30
30
30
n, %
NP/ND
12.4
1.4
22.2
2.8
28.0
8.6
n, %
CP/CD
15.0
1.8
24.0
3.4
33.0
11.5
Pt @ 0.4
NP/ND
86
98
77
97
60
86
1/m3, %
CP/CD
84
97.8
75
96
55
81
-------
We may also note that higher charge on the drops would have
a large effect. For instance, if the charge level on 100 nm dia.
drops were 10% of the Rayleigh limit, the average efficiencies
for 0.6 umA and 1.0 pmA particles would be 30% and 47%,
respectively. Comparison with the values given in Table 11 for a
lower drop charge level shows the extent of improvement better
charging would cause.
Experiments
Experiments were done on the charged spray scrubber built
under the EPA Contract No. 68-02-3109 (Yung et al., 1981) to
determine the particle capture efficiency. A schematic diagram
of the equipment is shown in Figure 26. Hydrated lime aerosol is
fed into the blower and the particles are charged by a wire and
rod charging grid. Particles are then collected by charged water
drops which are sprayed from pressure nozzles and induction
charged bv means of a grid. The water drops are removed from the
air flow by an entrainment separator and the "clean" air is
vented to the atmosphere. The particle penetration through the
charged spray scrubber is determined by sampling simultaneously
at the inlet and outlet ports with cascade impactors.
Figures 27 through 31 show the experimental penetration
curves. The superficial gas velocity and liquid-to-gas ratio
were 2.9 m/s and 0.4 1/m3 (3 gal/Mcf) for all experiments. The
first two runs (72/09/03 and 72/09/04) used pigtail nozzles and
the other runs used pin type nozzles. Nozzle pressure was about
450 kPa (50 psig). The drop diameter was measured to be 300 ym
and 240 urn for the pigtail and hook type nozzles, respectively.
The charges on particles and drops were -1.1 x 10~* C/g and 5.8 x
10" C/g, respectively (see Yung et al., 1981, for drop diameter
and charge measurement methods).
As can be seen from Figures 27 through 29, charging the
water drops or the particle augmented the collection of fine
particles. Further improvement in efficiency was obtained when
76
-------
POWER SUPPLY
FLOW
STRAIGHTENING
SECTION
\
BLOWER
INLET
SAMPLING
SECTION
\ SPRAY
\ SECTION
PARTICLE
CHARGING
SECTION
\
OUTLET
SAMPLING
SECTION
VENT
OH
PUMP
SUMP
\ENTRAINMENT
SEPARATOR
Figure 26. Experimental apparatus for measuring particle collection
efficiency of a charged spray scrubber .
-------
20.0
10.0
5.0
4.0
3.0
~ 2.0
cc
-------
+J
o
4-
LU
Q.
I RUN NO.
72/10/01
72/10/04 i
ur = 2.9 m/s
b
QL/QQ = 4 x 10"1* m3/m3
BETE P-48 NOZZLE AT 450 kPa
UNCHARGED PARTICLE/UNCHARGED DROP
0.5 1 5
AERODYNAMIC PARTICLE DIAMETER, ymA
Figure 28. Measured spray scrubber penetration.
79
-------
c
o
o
(O
o
II
H-
C£
1.0
0.5
0.1
0.05
0.01
RUN NO. i:-j~s--H- ---
72/10/16> -;-"J
RANGE OF DATA
OBTAINED IN EPA
CONTRACT NO. 68-02-3109
= 2.9 m/s
QL/QQ =4x10'" m3/m
BETE P-48 NOZZLE AT 450 kPa
UNCHARGED PARTICLE/CHARGED DROP
DROP CHARGE LEVEL = 5.8 x 10"7 c/g
AERODYNAMIC PARTICLE DIAMETER, ymA
Figure 29. Measured spray scrubber penetration
80
-------
1.0
0.5
o
M
U
(O
s_
a:
UJ
UJ
Q-
0.1
0.05
0.01
RUN NO.
72/10/25
72/10/23
72/10/2
is-fil UG = 2.9 m/s
p4 Q,/QG = 4 x 10-* m3/m3
Bg BETE P-48 NOZZLE AT 450 kPa
|E| CHARGED PARTICLE/CHARGED DROP
^ PARTICLE CHARGE LEVEL = 1.1 x 10~* c/g
DROP CHARGE LEVEL = 5.8 x 10'7 c/g
0.5 1
AERODYNAMIC PARTICLE DIAMETER,
ymA
Figure 30. Measured spray scrubber penetration
81
-------
u
re
i.
72/10/19^-!:^:fei
= 2.9 m/s
QL/QG = 4 x 10-"
BETE P-48 NOZZLE AT 450 kPa
CHARGED PARTICLE/CHARGED DROP
PARTICLE CHARGE LEVEL = 1.1 x 10"* c/g
DROP CHARGE LEVEL = 5.8 x 10'7 c/g
0.05
0.01
0.5 1 5
AERODYNAMIC PARTICLES DIAMETER, ymA
Figure 31. Measured particle charger and spray scrubber penetration
82
-------
opposite charges were placed on the drops and the particles.
Discussion
While the model predicts no significant effect of charging
for 250 pm dia. drops, experimental results show a large increase
in efficiency for the smaller particles. The discrepancy may be
due in part to the large effects of drop range and drop diameter.
If the drops travel farther than 50 cm before striking a wall and
if the effective average drop diameter is smaller than 250 urn,
the predicted efficiency would be higher.
The overall penetration, which was computed with the experi-
mental grade penetration and the assumed inlet particle size
distribution, was 12.5% for a liquid/gas ratio of 0.4 1/m3 and dd
= 240 urn. The calculated overall penetration was slightly higher
than the target of 90% efficiency. To obtain 90% efficiency,
smaller spray drops (which have higher efficiency) or a higher
liquid/gas ratio must be used. Using smaller spray drops is a
better approach because of limited holding capacity of the water
tank on the street sweeper.
Bench scale experiments on the charged spray scrubber indi-
cated that particle collection efficiency could be increased by
using fine water drops. A survey of single fluid spray nozzles
available in the market was conducted. It was found that the
opposed jet type and pin type nozzles can give fine atomized
sprays. The flow characteristics of these nozzles with the Tymco
piston pump on the sweeper were measured and are presented in
Table 12. The opposed jet nozzle has a higher liquid flowrate
and it generates smaller drop than the pin jet type nozzle.
Drops could be charged by induction by either connecting the
high voltage terminal directly to the nozzle or to a grid in the
proximity of the nozzle. Connecting the high voltage terminal
directly to the nozzle would not be practical for the sweeper
because it is necessary to isolate the pump, the water tank, and
pipes. Charging by the second method is simpler. In order for
the drops to be charged properly, the high voltage grid should be
83
-------
TABLE 12. DATA ON FINE SPRAY NOZZLES
Aux.
Engine
Speed
PIN TYPE NOZZLE
P-48
Mass**
Medi urn
Water Drop
Pressure Flow Rate Diameter
kPa m3/s urn
OPPOSED JET NOZZLE
#22477611
Mass**
Medium
Water Drop
Pressure Flow Rate Diameter
kPa m3/s Urn
1,500
2,000
3,400"
3,400*
3.8 x 10
-5
100
1,378
8.8 x 10
-5
1.3 x
100 2,067 1.7 x 10-
55
50
*Pressure relief valve was set at 3,400 kPa
**0ata provided by manufacturer
84
-------
placed at the location where the liquid sheet from the nozzle
breaks into drops.
Because it can generate smaller drops, the opposed jet
nozzle was initially chosen as the nozzle to be used in the spray
scrubber. Howeverf problems arose in trying to charge the water
drops. The liquid sheet from the nozzle is small and too close
to the nozzle. The high voltage grid cannot be placed close by.
The pin type spray nozzle was finally chosen. This nozzle
can provide 100 ym diameter drops at high pressure (4.2 MPa, 600
psi) .
DETAILED DESIGN
Figure 32 shows the modified sweeper layout. The dust
clouds in the gutter broom area were controlled by enclosing the
broom with a hood and venting to the hopper through 15 cm I.D.
flexible hose and piping. A damper valve in the pipe section
controls the air flow from the hood to the hopper.
Figure 33 shows the dimensions of the scrubber which was
designed for a maximum flow rate of 56 m3/min (2,000 cfm). A
wire and rod type particle charger was built into the scrubber to
pre-charge the particles. The particle charger consisted of two
rows of corona wires. Wire diameter was 0.18 mm (0.007 in.).
The spacing between wires within the same row was 11.4 cm (4.5
in.). The ground electrodes were 2.67 cm O.D. pipe. The cross-
section of the charger was 45.7 cm x 45.7 cm, which was smaller
than the scrubber shell cross-section so that all electrical
connections could be hidden inside the shell.
Four equally spaced pin type nozzles were placed 15 cm
downstream from the particle charger. The spray was co-current
with gas flow and was charged by induction/ with the nozzles
maintained at ground potential and a high voltage ring in front
of each nozzle. The overall length of the spray section was 91.5
cm (3 ft).
A 15 cm (6 in.) thick knitted mesh entrainment separator was
used to remove water drops as well as large solid particles. The
85
-------
SCALE APPROXIMATELY 1 cm - 0.6m
SCRUBBER
AIR FROM
PRESSURE HOSE
AUXILIARY WATER
TANK
00
POWER PLANT
GUTTER BRROM
SAMPLING TRAILER
PICKUP HEAD
Figure 32. TYMCO street sweeper and trailer.
-------
INLET
TRANSITION
PARTICLE
CHARGER
SPRAY
CHAMBER
ENTRAPMENT
SEPARATOR
(DO
\
OUTLET
TRANSITION
H 0.66m ><« 0.3m
0.9m
. 46m ^«0.35m-
Figure 33. Scrubber shell
-------
air velocity at the upstream face of the entrainment separator
was increased by blocking about 50% of the flow area so that the
cut diameter of the entrainment separator would be 3 pmA.
In an actual system, the scrubber could be located inside
the hopper. For purposes of sampling and easy access, the scrub-
ber was mounted on top of the sweeper and water was drained into
the hopper.
Gas flow rate through the scrubber was measured with a
Venturi meter, and was controlled by a damper. The sampling
system provides for the measurement of air flow from the gutter-
brooms to the hopper inlet. Particle size and concentration were
measured with cascade impactors. Sample ports were provided for
the scrubber inlet and outlet at locations at least 8 duct dia-
meters from upstream transitions or bends. Sample trains of the
type used for EPA Method 5 were used. Due to space limitations
on the sweeper, the sample trains were placed on a trailer pulled
by the street sweeper. Additional utilities, such as the power
plant for supplying electricity to the sampling pump, the high
voltage power supplies, auxilliary water tank, and the transfer
pump, were also placed on the trailer. Figure 34 shows the
sampling trailer layout.
88
-------
4.9 m
1:8 m
1.8 m
AXLE
1.2 m
MAXIMUM CARGO
CAPACITY: 1,778 kg
SAMPLE
TRAIN
SAMPLE
TRAIN
-PANEL
--WORK BENCH
WATER TANK
200 gal
TOOL
BOX
POWER
PLANT
BATTERY
1.8TT
Figure 34. Plan view of the sampling platform.
89
-------
Section 6
ROAD TESTS
GUTTER BROOM HOOD
Observation of street sweeper operation clearly showed that
gutter brooms and leakage from the "pickup hood" were the major
sources of inhalable emissions from the regenerative vacuum
sweeper. These emissions need to be contained to prevent disper-
sion into the air. The approach used in the present study was to
enclose the brooms with hoods and to increase the suction in the
pickup hood.
It was necessary to develop a hood arrangement that allowed
normal broom operations and also reduced'f ugitive broom emissions
to an acceptable level. In the first design, the upper half of
the broom was wrapped with rubber sheets. Air in the enclosed
broom was vented through the top and into the hopper. Road tests
of this hood showed that it had poor containment efficiency.
Dust clouds in the gutter broom areas were still visible. This
hood was subsequently abandoned in favor of a new abbreviated
hood.
In the new designf the gutter brooms were not completely
enclosed. A curved, tapered rectangular duct was wrapped around
the forward portion of the broom. The duct was similiar to an
air curtain distribution manifold and had a 2.5 cm (1 in) wide
slot facing the street. The slot was positioned 5 cm (2 in.)
above the street and was slightly offset so that it did not
scrape the road surface when sweeping an uneven street or hit the
curb when the broom was brushing the curb side. This hood oper-
ates with a high velocity suction airflow similar to a vacuum
cleaner, and also uses the interaction of the rotating broom and
street dirt to capture dust. The captured dust is conveyed to
the main hopper with piping and 15 cm I.D. flexible hose. A
damper valve in the pipe section controls the air-flow from the
hood to the hopper.
90
-------
The sweeper with the new vacuum hood was tested on a
commercial street. Visual observations indicated that the hoods
performed satisfactorily with the damper valve fully open. No
dust cloud was visible in the gutter broom area. However, the
hoods were found to be too efficient in picking up street dust.
The hood picked up inhalable particles as well as sand. In about
5 minutes, the 15 cm I.D. hose between the hood and the hopper
was clogged with about 36 kg (80 Ib) of dust.
Since large particles such as sand will be picked up by the
pickup hood of the sweeper, the gutter broom hood should only
collect inhalable particles. The slot width on the gutter broom
hood was subsequently reduced to 1.3 cm (0.5 in) and the air flow
was also reduced so that the hood can accomplish its primary
objective, i.e. to collect inhalable particles existing in the
gutter broom area and convey the particles to the hopper. The
modified gutter broom hood was observed to perform properly.
The minimum volumetric air flow rate required for the gutter
broom hood to eliminate a dust cloud was measured. This was done
by first operating the hood at maximum air flow rate, then
gradually reducing the air flow rate until a dust cloud was
observed to emerge in the gutter broom area. This flow rate was
measured and defined the minimum air flow rate for proper gutter
broom hood operation. It was found that the minimum air flow
rate for each gutter broom hood is 0.17 m3/s (350 acfm) or about
7 to 9% of the total air flow of the sweeper.
VENT AIR RATE
The regenerative air flow of the modified sweeper is shown
in Figure 35. Under normal sweeping conditions (with the leaf
bleeder valve shut) the air from the pressure hose picks up
street debris and returns to the hopper. When the pickup head
travels on uneven street surfaces, dust puffs emerge out of the
pickup head. These dust puffs are eliminated by opening the leaf
bleeder valve. This decreases the pressure on the pressure side
of the blower and increases the suction in the hopper. As a
result, air from the atmosphere flows into the pickup head, elim-
91
-------
VENT AIR
1
ILEAF BLEEDER VALVE
SCRUBBER
PICKUP HEAD -
T
SWEEPER BLOWER
VALVE
LEFT
GUTTERBROOM
HOOD
RIGHT7
GUTTERBROOM
HOOD
T
Figure 35. Schematic diagram of street sweeper flow circuit.
92
-------
inating the dust puffs. The vent air is cleaned by the scrubber
before it escapes into the atmosphere.
To minimize the operating cost and water consumption, the
air flow to be vented through the scrubber should be kept at a
minimum. However, the air flow should not be so low that dust
puffs occur around the gutter broom and the pickup head.
The minimum air flow needed to be vented through the scrub-
ber to prevent the occurance of dust puffs was determined with
the gutter broom hood air flow either on or off. The
experimental procedures with the gutter broom hoods off are as
follows:
1. Close the damper in the hose connecting the gutter
broom hood and the hopper.
2. Set and maintain the auxiliary engine (the one which
runs the blower, the water pump, and the hydraulics) at
a specific speed.
3. Open the damper fully in the duct to the scrubber so
that air is vented through the scrubber.
4. Gradually close the damper to reduce the air flow until
dust puffs are observed to occur in the pickup head
area.
5. Measure the air flow at which dust puffs first occur.
This is the minimum air flow needed to be vented.
6. Repeat the above procedures for other engine speeds.
Similar procedures were used when the gutter broom hoods
were turned on. The air flow in each hood was maintained at 0.17
m3/s (350 cfm), which is the minimum required flow for
satisfactory hood operation.
It was found that the minimum air flow to be vented through
the scrubber with the gutter broom hoods OFF was about 0.33 m3/s
(700 acfm), and about 0.38 mVs (800 acfm) with the gutter broom
hoods ON.
93
-------
PARTICLE SIZE DISTRIBUTION AND CONCENTRATION AT SCRUBBER INLET
The spray scrubber was designed for removing particles hav-
ing a mass median diameter of 4.0 gmA and a geometric standard
deviation of 2.0. This size distribution was determined with a
cascade impactor located underneath the street sweeper. The
particle size distribution might be different at the spray scrub-
ber inlet. Therefore, additional sampling was done at the scrub-
ber inlet to characterize the particles to be cleaned by the
scrubber. Streets were chosen with differing activities in the
San Diego and Los Angeles areas. This sampling was done before
the scrubber was installed, so that scrubber efficiency for these
runs was not measured.
Figures 36 through 41 show the results obtained in the San
Diego area. Particles from residential areas are larger than the
assumed distribution. Particles from an industrial district have
distributions close to design condition; particles from commer-
cial district have more small particles than assumed.
Twelve sampling runs were done in the Los Angeles area and
results are summarized in Table 13. Figures 42 and 43 show the
cumulative mass concentration and particle size distribution,
respectively. The concentration and particle size distribution
vary greatly from location to location. The sampling results
agree with visual observations in that dirtier streets resulted
in higher particle concentration and larger particles.
The air flow rate vented through the scrubber has great
effects on particle size distribution and concentration. Based
on results of experiments performed on the same street, a higher
vent flowrate results in a higher particle concentration and
larger particles (Figure 44).
The spray scrubber was designed for removing 90% of the
particles which have a mass median diameter of 4.0 umA and a
geometric standard deviation of 2.0. A more efficient spray
scrubber is probably needed. There are more submicron particles
in the air stream at the scrubber inlet than assumed.
Street samples collected indicate a difference in character-
istic size dirt when neighborhoods are considered. The major
94
-------
a.
o
a
o
IPPI^ J :
LOCATION: INDUSTRIAL ^
h:- RUN NO. 72-17-9
10 20 30 40 50 100 200 300 400 500
CUMULATIVE MASS CONCENTRATION, mg/DNm3
Figure 36. Particle size vs. mass concentration of dust at scrubber
inlet in an industrial area.
95
-------
LU
I
LlJ
>*
Q
LU
Q
O
1.0 -^
0>5 III ill
10 20 30 40 50 100 200 300 400 500
CUMULATIVE MASS CONCENTRATION, mg/DNm3
Figure 37. Particle size vs. mass concentration of dust at scrubber
inlet in a commercial area.
96
-------
-------
o;
LU
Q
LU
_J
t<
o:
Q.
O
Q
O
01
LU
LOCATION: INDUSTRIAL :
- RUN NO. 72-17-9
10 20 30 40 50 60 70
CUMULATIVE MASS UNDERSIZE, %
80
90
Figure 39. Particle size distribution of dust at scrubber inlet in
an industrial area.
98
-------
o:
«=c
a.
CJ
14
SI
Q
O
o:
30.0
20.0 £ZL
10.0
LOCATION: COMMERCIAL
RUN NO. 72-17-11 :
! RUN NO. 72-17-10
I"" ::.:i-.:: i
10 20 30 40 50 60 70 80 90
CUMULATIVE MASS UNDERSIZE, %
Figure 40. Particle size distribution of dust at scrubber inlet in a
commercial area.
99
-------
30.0
20.0
o:
O
O
10.0
5.0
4.0
3.0
2.0
1.0
n
^L--.^-:- : - I ' ' - i ' -I----!
LOCATION: RESIDENTIAL
^P
MRUN NO. 72-17-12 It
mm
RUN
7_2-i7-i3|»fc.:
:fi-rU-
-. -U-
10 20 30 40 50 60 70
CUMULATIVE MASS UNDERSIZE, %
80
90
Figure 41 Particle size distribution at scrubber inlet in a residen-
tial area.
100
-------
TABLE 13. SUMMARY OF SAMPLING RESULTS FROM LOS ANGELES AREA
Run
Number
Symbol
18/5
18/6
18/7
18/8
18/9
18/10
18/11,
12
Location
(L. A.
Area) Industry
Carson Texaco
Refinery
Wilmington Bridge on
Anaheim
Street
Torrance U.S.
Steel
Torrance Mobil
Refinery
Torrance Coke
plant
Torrance Coke
plant
Torrance U.S.
Traffic
Conditions
Medium-
heavy
Very
heavy
Light-
medium
Heavy
Heavy
Heavy
Light-
medium
Total
Scrubber Mass
Flow Concen-
Rate tration
AmVmin mg/DNm3
20 3,190
20 770
32 3,503
32 684
32 132
32 193
20 386
Mass
Median
Diameter
ymA
10.0
5.2
26
65
42
100
7.7
Geometric
Standard
Dev;ation Comments
g
4.5 Heavy street loading
3.5 Heavy street loading
15.3 Heavy street loading.
U.S. Steel plant was
in the process of be-
ing demolished.
8.1 Medium street loading
7.1 Light-medium street
loading opposite
Mobil refinery on
Crenshaw.
8.7 Light-medium street
loading past the truck
exit from the Coke
plant. Coke visible
on the street.
7.0 Heavy street loading
same street as Run
18.7. Run w/street
sweeper efficiency
test 18/12.
-------
TABLE 13'. CONTINUED
Run
Number
Symbol
18/13
18/14
18/15
18/16
18/17
Location
(L. A. Traffic
Area) Industry Conditions
Vernon Bethlehem Medium
Steel
Vernon Bethlehem Medium-
Steel heavy
Vernon Warehouse Medium
and steel
plant
Vernon Grain Heavy
mill
Vernon Jorgenson Medium
Total
Scrubber Mass
Flow Concen-
Rate tration
ACFM mg/DNm3
710 265
710 362
710 134
1,130 556
1,130 2,186
Mass Geometric
Median Standard
Diameter Deviation Comments
ymA og
17.5 4.9 Medium street loading
13.1 5.0 Medium street loading
15.4 4.3 Medium street loading
170 9.4 Medium-heavy street
loading
47 5.4 Heavy street loading
run w/street sweeper
efficiency test 18/17
o
to
-------
CtL
-------
20.0
10.0
OL
LU
HIGH SCRUBBER FLOW
0.53 m3/s
LOW SCRUBBER FLOWS
0.34 m3/s
is/6 5.2
r 18/11 A 7.7
i 18/5 10.0
- 18/14 O 13.1
18/15 D 15.4
£.p±-_ 18/16 A 170
0.5
5.0 10 20 30 40 50 60 70 80 90
CUMULATIVE MASS UNDERSIZE, %
Figure 43. Particle size distribution.
104
-------
30.0
<:
LOW
AIRFLOW
r 0.34 m3/s
10 20 304050 TOO 200 300 500 1000 20003000
CUMULATIVE MASS CONCENTRATION, mg/DNm3
Figure 44. Effect of vent air flow on scrubber inlet dust
concentration.
105
-------
distinction noticed between neighborhoods is the amount of
traffic; street dirt is reduced in size by continuous traffic.
SWEEPING EFFICIENCY
Street Dust Measurement Method
To determine the sweeping efficiency of the modified sweep-
er, a method is needed which measures the amount of dust that can
be dispersed into the ambient air by the mechanisms actually
occuring on streets. The mechanisms of street dust removal are
(Brookman and Martin, 1979):
1. Reentrainment (by air currents around moving vehicles).
2. Wind erosion (similar to 1, but due to natural air
currents).
3. Displacement (similar to 1, re-disposition near
street) .
4. Rainfall runoff.
5. Street cleaning.
The first three mechanisms result in airborne dispersal of
street dirt, so the sampling method should measure the amount of
inhalable particulate matter which can be dispersed by these.
Preliminary experiments were performed to measure street
dust density with a brush-type vacuum cleaner, as has been done
by Dahir and Meyer (1974). The dust on the street was first
loosened with a brush, and then taken up and filtered by the
vacuum cleaner. The amount of dust was determined from filter
weight gain. Street dust density was calculated by dividing the
weight of the dust by the area vacuumed.
This method was found to have several deficiencies. The
vacuum cleaner can suck up dust which is deposited deeply in
cracks. The dust in cracks is not normally reentrained. Scour-
ing and brushing the street surface to loosen the street dust is
not a naturally occuring dust dispersion mechanism. Further, the
total amount of dust vacuumed increases with each pass of the
106
-------
vacuum nozzle; therefore, there is no logical end-point for
sampling.
Consequently, the above method was abandoned and a new
method developed. The new method uses a vacuum cleaner with a
modified pickup nozzle. The filter in the vacuum cleaner is a
custom-made bag house type filter. Two pickup nozzle designs
were evaluated. The first one was of circular design and is
shaped to create a uniform 97 m/hr (60 mph) air flow field at the
street surface (Figure 45). The dust dislodged by the air flow
is sucked into the vacuum and filtered by the vacuum bag. This
method simulates dust reentrainment due to air currents around
moving vehicles and wind erosion.
A detailed definition of reentrainment would require a com-
plex model of airflow around an automobile. Assuming that the
maximum air velocity caused by the passage of an automobile will
not exceed its velocity, and that 97 km/hr (60 mph) is a reason-
able maximum automobile velocity, the resultant maximum
reentrainment velocity would be 97 km/hr.
Wind speeds are generally much less than 97 km/hr (27 m/s),
so this velocity is conservatively high to represent the influ-
ence of wind erosion.
The displacement mechanism is essentially similar to that
for reentrainment, and due to the same vehicular motion. The
definition differs in that the particles stay in suspension only
long enough to reach the area immediately adjacent to the street.
An air velocity of 97 km/hr would adequately represent this
vehicle-induced mechanism.
Figure 46 shows a sketch of the second pickup nozzle design.
The nozzle had an adjustable inlet air jet. The air jet was
designed to aim the jet at the street surface in such a manner so
as to maximize the dust reentrainment in the air flow.
The inlet jet angle may influence the particle size and
concentration of the reentrained dust, so they are measured for
jet angles of 0°, 45°, and 90° (defined as "0"in Figure 47).
Particle size distribution and concentration of the reentrained
dust were measured at the vacuum hose with cascade impactors.
Figure 47 shows the results.
107
-------
UNIFORM
AIRFLOW
FIELD
/TO VACUUM
A*"COLLECTION
v BAG
STREET
SURFACE
Figure 45. Uniform Airflow Vacuum Nozzle.
108
-------
TO
VACUUM
TYPICAL
UNIFORM AIR FLOW
6.8 cm
Figure 46 . Air Jet Nozzle
109
-------
O 9 = 45°
O 9 = 90°
A fl = 0°
2.0 3.0 5.0 10
PARTICLE DIAMETER, ymA
20 30 40 50
Figure 47. Air jet nozzle performance.
110
-------
The 0° position produced a lower mass concentration and
larger particles than did the 45° or 90° positions. This could
be due to the air flow being parallel to the street surface at 0°
position which does not reentrain dust deposited in crevices. At
other jet angles* the jet stream impinges on the street surface.
Dust deposited in the crevices is blown out and reentrained by
the jet stream.
An inlet jet angle of 45° yielded maximum dust concentra-
tion. Therefore/ in street sweeping efficiency measurements, the
inlet angle of the jet nozzle was set at 45°.
Both nozzles yielded satisfactory results and both were used
in street sweeping efficiency determinations.
Street Sweeping Efficiency Measurement
Street sweeping efficiency is determined by measuring the
street dust concentrations before and after sweeping. The street
was divided into sections of equal area for dust concentration
measurements. Each section was further sub-divided into the
gutter broom area and the pick-up head area as shown in Figure
48. The equal area sections are used to reduce the effect of
localized differences in the street dirt loading. Each "before"
and adjacent "after" section has dirt loading representative of
that local area, and will give accurate street cleaning data.
The sub-division into gutter broom and pick-up head strips is for
separately measuring the performance of these two street cleaning
devices. The gutter area has a higher dirt loading than the
street due to traffic and water run-off. Also, the pick-up head
uses an air-jet mechanism for cleaning the street, while the
gutter broom uses a rotating wire broom. These differences may
result in different street sweeping efficiencies for the gutter
and street area.
Figures 49 and 50 show plots of street dust density before
and after sweeping. Run No. 18/17 was done on a street with on
street parking and Run No. 18/12 on a street without parking.
Street dust was concentrated in the gutter area. Parked cars
increased the dust density.
Ill
-------
r
38 cm
CURB
182 cnrj
"AFTER"
SECTION
"BEFORE"
/SECTION
'*.
60 cm
60 cm
r
GUTTERBROOM
STRIP
PICK-UP
HEAD
STRIP
Figure 48. Test strip layout for street sweeping efficiency measurement.
-------
300
200
100
CM
E
CD
CO
z.
QJ
a
h-
oo
Q
h-
UJ
50
40
30
20
10
9
8
7
6
5
4
^ty7! i
i i '
i 1 [
-1
*«
-
i
1
H
/
\
\
\
\
=
L_
i
!
^CURB
'~~
-M-
t
\
\
\
i
| /
*-|
1
I
r
^
...
H
-^
H : !
'F
r
«i
*
K
B
r-
rN
E
^S
5-
F(
't
Df
\f
^
i
51
.
i
G
B
H
]
r
pft
1
MAIN
. , . i
!
1
i
i
= INCRI
= CO NCI
= TO P,
i f |
«
T
: SWEE
ER SU
/i\i
^\
TH
i \
t i i
PICKUP
:ASE
:NTR
\RKE
-I--7N
IF
E
D [
AT]
D (
1
y
TNR
EPIf
^G
i
I
I
-i
-i
-rr^~
)IRT
[ON
:ARS
_..,.--!
!
]
1
=1
D
3
i
UE
I
^
0
LEFT
G.B. PAT
. 1 !
HOOD
PATH
i ; ^
[
i
s-
i
^
_ 1
4
/
/
; /
i y
j
A
\
f 1
/
-i-i-
-~,
Hi
1
i
1
-i-r-1-
-R
- |
t
i
._
-1 .
UN NO.
8/17
.53 m 3/s
ENTING
1
h
I
i
=
I
-
-J
=
^j
. i
-
1 i
0
1 2 3
DISTANCE FROM CURB, m
Figure 49.Street dust distribution before and after sweeping.
(On street parking)
113
-------
en
UJ
UJ
UJ
en
t
C/J
£:RUN NO.
18/12
tVENTING
to.34 m3/s
| BEFORE SWEEPING
SfeG.B. PATH
:HOOD PATH
DEEP GROOVE
AT STREET/GUTTER
JUNCTION
AFTER SWEEPING
1 2 3
DISTANCE FROM CURB, m
Figure 50.
Street dust distribution before and after
sweeping.
114
-------
The street sweeper sweeping efficiency was calculated from
the results presented in Figures 49 and 50 and it ranged from 75%
to 97%, as seen from Figure 51.
The right hand hood achieved a cleaner street surface than
did the left hand hood even though the right hand hood had a
higher concentration of dirt to clean. In both experiments, the
righthand (curb-side) was maintained at a suction flow rate of
0.27 m3/s (570 acfm) and the left hand (street side) at 0.7 m3/s
(360 acfm). The curb and gutter were made of concrete, which is
smoother than the street surface of asphalt. The dirt is more
easily sucked up in the gutter. The higher cleaning efficiency
for the curbside broom could be due to a higher air flow rate and
different street surface textures.
The fraction of the dust reentrained from the street smaller
than 15 jjmA was measured. Previous investigators (Sartor and
Boydf 1972) have used wire mesh sieves to separate the street
dirt sample into size fractions; but this method has not been
proved to be reliable for particles smaller than 45 pm (325 mesh
sieve) diameter.
A similar method was tried in this study. A sample of dust
was taken from the filter bag of the vacuum and either sieved to
obtain the fraction passing a 325 mesh (45 pm opening) screen or
wet vacuum filtered to obtain the fraction below 25 ym diameter.
This sample was then analyzed for the particle size distribution
with a Coulter counter. The particle size distribution deter-
mined by this method varied greatly even with samples taken from
the same vacuum bag. This indicated that the sample taken was
not representative. In addition, the Coulter counter may have
broken up the agglomerates. Therefore, this method of determin-
ing the sweeper efficiency with respect to particle diameter was
abandoned and a cascade impactor was used.
A sample probe was inserted into the vacuum hose (2 1/2" Sch
80 PVC pipe) of the vacuum cleaner to withdraw an isokinetic
sample of the air stream. The sample passed through a pre-cutter
to remove particles with diameter larger than 35 umA and a
University of Washington cascade impactor for size classifica-
115
-------
CLEANING EFFICIENCY, %
i >
J-P»CJlJ3O
DOOOOOOO
ou
on
C\J
10
0
;!^:
:-;:;. i^1
M 1
.... |
iii^TX
:::.-!:.:.-:-
:--:-|:-.i:.
- -- |
CUR
......
... I
4- - ---
- - t .
" r -
::::L::i".
=r D
3 |fi » i
::j;i::-:r
- - - 1
"r;:4:;-.:
T^s*^
:.:: -..-;
::.-:]:.::
:.-::!- ::
' r!:::r
3 j .
: 1. . . -
;;;;i;;::
'"'.:"::"
...:!;..-.
r
-;-
i LOW
CCCT
- trr 1
SUCT
!"'-'*'--
.. :: ::::
.... i ....
:;":i:;::
: : . : : - - :
^
. . , i . ^^
:-:.: :-;:
SCRUBBER 1
AIR Fl
- - - i - -
:::.---:
.-;;
.ow \
- . . 1 .- . .
.' - - 1 c
i. E
....
:::-.; L
::-~F G
:\ ':
:-.:: ;;.
I:::-!::::
. - . j : .
-: : jp"
: : : I : .
. i . . -
Si*rr>ss
^i-O-,-:
.ow-"
UGH-1
1 -
.;..:!:.-;
DGE OF
EFT HA
UTTERB
-:::":ii-ii
-:-::;j::::
: . :
: . : i . '
ND :
ROOM_
o
LEFT HAND GUTTER BROOM
CIENCY DUE TO INADEQUATE
ION SLOT AIR FLOW
RUN NO. 18/1
RUN NO. 18/1
^ '
7r |.:.:
h-i- -
1 I
\-- -
:..:;;..
. >y
- - <(-
... i ...
: :-: STR
DTP
- L/i r\
; :;LOA
: : LIG
:'. ':'-
EET ':-'-
T ~
DINGii
HT ^L
VY ^.
::-::-j::::"
:;-;i;:;;
;;-:;i-.:;.:
;- ::|;-;.
:: :::
i i ....
: .:
' ... '
i1
J
. ..;
1 . .
-.:.:!:: :.:
' - - - t
._ . ) . ..
|
... .J- ....
- - . 1 .
-..::£::
=
1
DISTANCE FROM CURB, m
Figure 51. Overall street sweeping efficiency.
116
-------
tion. The sample flow rate was kept low to fractionate the
particles into size ranges from 0.9 to 38 ymA diameter.
The cumulative concentration of street dust is plotted
against aerodynamic particle diameter for before and after street
sweeping in Figures 52 and 53. The main pick-up head of the
sweeper had a higher sweeping efficiency because it has air jets
blasting at the road surface to lift the dust off the street. The
gutter broom was not effective in removing street dust less than
2 umA diameter. It seems that the broom does not impart enough
momentum to the smaller particles for them to move over to the
pickup head area or to suspend them in the air, and cannot reach
into the crevices to stir up the dirt.
The performance of the modified sweeper was compared to that
of a regular sweeper visually at a construction site. The modi-
fied sweeper not only drastically reduced the dust clouds around
the sweeper, it also gave a cleaner street. A cleaner street
will reduce the fugitive street dust emission.
The sweeping efficiency of the gutter broom with the hood
air flow on and off was measured in a parking lot. The efficiency
was 57% with hood vent air flow off and 65% with the air flow on.
SCRUBBER EFFICIENCY
The performance of the spray scrubber was determined in the
laboratory by injecting lime dust into the pickup head and simul-
taneously sampling the scrubber inlet and outlet air streams with
cascade impactors. Table 14 shows a summary of the results.
Figures 54 through 56 show the experimental grade penetration
curves for three scrubber operating conditions. The water and
particles were not charged in these experiments because electro-
static augmentation only improves the collection of small parti-
cles, and there were not that many small particles in the street
dust. It can be seen from the un-charged spray scrubber has
adequate collection efficiency at a liquid/gas ratio of 0.68 1/m3
(5.1 gal/mcf) or higher.
117
-------
1000
PICK-UP HEAD AREAH
C_5
0.5
2345 10
PARTICLE DIAMETER, u^
20 30 40 50
Figure 52. Street sweeping efficiency.
118
-------
1000
M-»-J- -M-H
i: ; ' ' --'
H. HOOD
BEFORE
M. HOOD
AFTER v -
G. BROOM I !
BEFORE
i-4 EFFICIENCY
4 GUTTER BROOM
HS|MAIN HOOD
COVERALL 94% 90%:
H<15pmA 40% 89%:
i Run 22-1
I III I 111
2345 10
PARTICLE DIAMETER, ymA
Figure 53. Street sweeping efficiency.
40 50
L19
-------
TABLE 14. SUMMARY OF SCRUBBER TESTS
RUN NO.
22/06
22/07
22/08
22/10
22/11
22/12
AIR FLOW,
RATE
mVrrrin
20.1
20.1
20.1
20.1
32.0
32.0
LIQUID ,
GAS
£/m3
0.68
0.68
0.60
0.60
0.38
0.38
DUST
CONCENTRATION1 ,
mg/DN m3
Inlet
1036.13
1676.66
824.87
726.09
582.24
488.31
Outlet
206.94
313.24
205.38
129.27
157.90
123.03
EFFICIENCY,
%
Overall
80%
81%
75%
74%
73%
75%
<15 ym A
76%
77%
72%
70%
67%
70%
*Test Dust was Lime with MMD = 2.0 \im A and a = 2.
-------
ex.
QC
<
0.
Nozzle Pressure
Liquid/Gas:
Air flow:
1932 kPa
0.38 i/m3
32.0 m3/min
"iRun 22-11
Run 22-12
3.0 4.0 5.0
0.7 1.0
PARTICLE DIAMETER, ymA
Figure 54. Scrubber Performance
10
121
-------
98
T
I
95
90
80
70
60
o;
o
50
₯ 40
UJ
Q.
uj 30
20
10
Nozzle Pressure:
Liquid/Gas:
Air flow:
1932 kPa
0.60 i/m3
20.1 m3/min
Run 22-8
I
I
I (III!
0.7
1.0
2.0 3.0 4.0 5.0 7.0
10
PARTICLE DIAMETER,
Figure 55. Scrubber Performance
122
-------
Q.
UJ
_1
O
98
95
90
80
70
60
50
40
30
20
10
5
__l_ i _J_ I'JLJLJLLJ CT
Nozzle Pressure: 2588 kPa
Liquid/Gas: 0.68 l/m3
Air flow: 20.1 m3/min.
E Run 22-06
0.7 1.0 2.0 3.0 4.0 5.0
PARTICLE DIAMETER, ymA
7.0
10
Figure 56. Scrubber Performance.
123
-------
CONCLUSIONS
The objective of demonstrating the feasibility of applying
"Spray Charge and Trap" (SCAT) to the control of fugitive road
dust emissions was proven. However, it also concluded that a
Venturi scrubber should offer improved performance with less
complexity for this application, A regenerative vacuum sweeper
is ideally suited to a scrubber technique since it allows a
discharge stream of air to be cleaned and recycled into the
atmosphere. This feature reduces the size and power requirements
for control of fugitive emissions of inhalable particle
emissions.
Gutter broom dust emissions can be substantially improved
with an advanced, interactive gutter broom hood. Since most of
the dirt is in the gutter, this improvement is very significant.
Additional power requirements for the SCAT system are minimal.
Existing standard equipment pumps and blowers are sufficient to
do the job.
Even though the current research work was done on a
regenerative air vacuum sweeper, the same technology can be
applied to a mechanical broom sweeper. Modifications to the
mechanical sweeper will be more complicated since a blower needs
to be installed on the sweeper to provide the suction air flow.
The cost for modifying a regenerative air vacuum sweeper was
estimated to be about $2,000 which was about 5% of the sweeper
cost. The cost to modify a mechanical sweeper will be higher
because of the requirement of a blower.
124
-------
Section 7
RECOMMENDATIONS
A superior street sweeping machine has been developed for
reducing particle emissions from paved streets. This street
sweeper has been subjected to a limited testing program in San
Diego and Los Angeles. Results clearly indicate that the sweeper
can eliminate the dust plumes during sweeping and give a cleaner
street. However, additional research would refine the design and
demonstrate its capability in improving the ambient air quality.
The following studies would provide valuable information for
further study:
1. Refine the design of the scrubber and incorporate it
inside the hopper.
2. Improve the gutter broom sweeping efficiency for fine
particles.
3. Demonstrate the sweeper on city streets and measure the
improvement in ambient air quality.
4. Obtain more information on the nature of street dirt
with respect to potential ambient air quality.
5. Use a smaller scrubber than the one presently
installed. Results from this study indicated that an
air vent rate of 14 mVmin (1,000 acfm) was sufficient
to eliminate the dust cloud in the gutter broom area.
It is recommended that a low pressure drop Venturi
scrubber be used instead of a charged spray scrubber
because the wall loss of spray will be high for a small
scrubber.
125
-------
REFERENCES
Axetell, D., and T. Zell. "Control of Reentrained Dust from
Paved Streets." U.S. Environmental Protection Agency, Washington
DC, NTIS. PB 280325, July 1977.
Brookman, E. T., and D. K. Martin. "Assessment of Methods for
Control of Fugitive Emissions from Paved Roads." EPH-600/7-79-
239 (NTIS PB80-139330), Nov. 1979.
Brookman, E. T., and D. K. Martin. "Definition of a Methodology
for the Evaluation of Road Dust Control Techniques." EPA-1ERL
Process Measurements Branch, Research Triangle Park, N.C., April
1979.
Buchwald, H., and K. R. Schrag. "Dust Exposure in Mechanical
Street Sweepers." J. of A.I.H.I., 28, pp. 485-487, 1967.
Cowherd, C., C. M. Maxwell, and D. W. Nelson. "Quantification of
Dust Entrainment from Paved Roadways." U.S. Environmental Protec-
tion Agency, Washington DC, EPA-450/3-77-027, NTIS PB 272613,
July 1977.
Dahir, S. H., and W. E. Meyer. "Bituminous Pavement Polishing."
The Pennsylvania State University, University Park, Pa., Nov.
1974.
Draftz, R. "The Impact of Fugitive Dust Emissions on TSP Concen-
trations in Urban Areas." In Third Symposium on Fugitive Emis-
sions: Measurement and Control, U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-600/7-79-182 (NTIS PB
80-130891), August 1979.
Horton, J. P. "Broom Life Isn't the Most Important Cost..." The
American City, Buttenheim Publishing Co., New York, NY. July
1968.
126
-------
Pitt, R. "Demonstration of Non-Point Pollution Abatement Through
Improved Street Cleaning Practices." EPA-600/2-79-161. August
1979.
Roberts, J. W., A. T. Rossano, P. T. Bosserman, G. C. Hofer, and
H . A. Watters. "The Measurement, Cost and Control of Traffic
Dust and Gravel Roads in Seattle's Duwamish Valley." Paper No.
AP-72-5, Presented at the Annual Meeting of the Pacific Northwest
International Section of the Air Pollution Control Association,
Eugene, Oregon, November 1972.
Sartor, J. D., G. B. Boyd, and W. H. Horn. "How Effective is Your
Street Sweeper", APWA Reporter, April 1972.
Sartor, J.. D., and G. B. Boyd. "Water Pollution Aspect of Street
Surface Contaminants." U.S. Environmental Protection Agency
Publication No. EPA-R2-72-081. 1972.
Steward. The Resuspension of Particulate Material from Surface
Contamination, edited by B. R. Fish, Oxford, England, Pergamon
Press, 1964.
URS Research Company. "Water Quality Management Planning for
Urban Runoff." NTIS PB 241689/LK, December 1974.
Yung, S. C., J.. Curran, and S. Calvert. "Spray Charging and
Trapping Scrubber for Fugitive Particle Emission Control." EPA-
600/7-81-125, NTIS PB82-115304, July 1981.
127
-------
TECHNICAL REPORT DATA .
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-600/7-84-021
2.
3. RECIPIENT'S ACCESSIOf*NO.
4. TITLE AND SUBTITLE
Improved Street Sweepers for Controlling Urban
Inhalable Particulate Matter
5. REPORT DATE
February 1984
6 PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
S. Calvert, H. Brattin, S. Bhutra, and D. Ono
8. PERFORMING ORGX
9. PERFORMING ORGANIZATION NAME AND ADDRESS
A.P. T. , Inc.
4901 Morena Boulevard, Suite 402
San Doego, California 92117
10. PROGRAM ELEMENT NC
11. CONTRACT/GRANT NO.
68-02-3148
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13 TYPE OF REPORT AND PERIOD COVERED
Final; 8/79 - 7/81
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES LERL-RTP project officer is William B. Kuykendal, Mail Drop 61,
919/541-7865.
16. ABSTRACT The report gives results of an experimental program to develop design
modifications that can be used to improve the ability of municipal street sweepers
to remove inhalable dust particles from streets. (Dust emissions from paved roads
are a major source of urban inhalable particulate matter.) A commercial regener-
ative air sweeper was modified. Major modifications included a charged spray
scrubber for fine particle collection, and a gutter broom hood to help contain redis-
persed dust particles. The upgraded sweeper proved effective in eliminating dust
plumes during sweeping and giving cleaner streets.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Pollution
Streets
Brushing
Dust
Pavements
Plumes
Charging
Spraying
13. DISTRIBUTION STATEMENT
Release to Public
b.lDF.NTIFIERS/OPEN ENDED TERMS
Pollution Control
Street Sweepers
Particulate
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (Thispage)
Unclassified
COSATI Field/Group
13B
13H
11G
21B
14G
21. NO. OF PAGES
141
22. PRICE
EPA Form 2220-1 (9-73)
128
------- |