EPA-600/2-76-089a
April 1976
Environmental Protection Technology Series
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/2-76-089a
April 1976
TECHNICAL MANUAL
FOR MEASUREMENT OF
FUGITIVE EMISSIONS:
UPWIND/DOWNWIND SAMPLING METHOD
FOR INDUSTRIAL EMISSIONS
by
Henry J. Kolnsberg
TRC--The Research Corporation of New England
125 Silas Deane Highway
Wethersfield, Connecticut 06109
Contract No. 68-02-2110
ROAP No. 21AUY-095
Program Element No. 1AB015
EPA Project Officer: R.M. Statnick
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
•»-<>*•
230 SOUCA Ft- 1
-------
TABLE OF CONTENTS
SECTION
PAGE
1.0 OBJECTIVE i
2.0 INTRODUCTION 2
2.1 Categories of Fugitive Emissions 2
2.1.1 Quasi-Stack Sampling Method 2
2.1.2 Roof Monitor Sampling Method 3
2.1.3 Upwind-Downwind Sampling Method 3
2.2 Sampling Method Selection 4
2.2.1 Selection Criteria 4
2.2.2 Application of Criteria 6
2.3 Sampling Strategies 9
2.3.1 Survey Measurement Systems . 10
2.3.2 Detailed Measurement Systems IQ
3.0 TEST PROGRAM PROCEDURES 12
3.1 Pretest Survey 12
3.2 Test Plan 13
3.3 Upwind-Downwind Sampling Strategies 17
3.4 Survey Upwind-Downwind Measurement System ... 17
3.4.1 Sampling Equipment 18
3.4.2 Sampling System Design 19
3.4.3 Sampling Techniques 22
3.4.4 Data Reduction 31
3.5 Detailed Upwind-Downwind Measurement System . . 31
3.5.1 Sampling Equipment 32
3.5.2 Sampling System Design 33
3.5.3 Sampling Techniques 34
3.5.4 Data Reduction 39
3.6 Atmospheric Tracers 39
3.7 Quality Assurance 43
4.0
APPENDIX
A
ESTIMATED COSTS AND TIME REQUIREMENTS 46
TEST PROCEDURES APPLICATION
111
-------
LIST OF TABLES
TABLE
2-1
3-1
3-2
3-3
4-1
4-2
4-3
FIGURE
3-1
3-2
3-3
3-4
3-5
3-6
4-1
PAGE
Upwind-Downwind Sampling Method Application
to Typical Industrial Fugitive Emissions
Sources
Pre-test Survey Information to be Obtained
for Application of Fugitive Emissions Sampling
Methods 14
Matrix of Sampling System Design Parameters ... 21
Atmospheric Stability Categories 28
Conditions Assumed for Cost Estimation of
Upwind-Downwind Sampling Programs 47
Estimated Manpower Requirements for Upwind-
Downwind Sampling Programs A8
Estimated Costs for Upwind-Downwind Sampling
Programs
LIST OF FIGURES
PAGE
Typical Sampler Locations for Selected
Source Site Configurations 23
Maximum Downwind Sampler Distances 27
9 Q
Maximum Crosswind Sampler Distances ^
Pollutant Concentration Ratios for Crosswind
Locations
Pollutant Concentration Ratios for Vertical
Locations
Vertical Concentration Distribution Factors ... 38
Elapsed Time Estimates for Upwind-Downwind
Sampling Programs
IV
-------
LIST OF FIGURES
(continued)
FIGURE PAGE
^~2 Cost-effectiveness of Upwind-Downwind Sampling
Programs 52
A-1 Portland Cement Plant Site Layout A-3
A~2 Portland Cement Plant Emission Clouds A-9
A~3 Portland Cement Plant Separate Source Clouds . . A-13
-------
1.0 OBJECTIVE
The objective of this procedures document is to present a guide
for the utilization of the Upwind-Downwind Sampling Strategy in the
measurement of fugitive emissions. Criteria for the selection of the
most applicable measurement method and discussions of general informa-
tion gathering and planning activities are presented. Upwind-downwind
sampling strategies and equipment are described and sampling system
design, sampling techniques, and data reduction are discussed.
Manpower requirements and time estimates for typical applications
of the method are presented for programs designed for overall and speci-
fic emissions measurements.
The application of the outlined procedures to the measurement of
fugitive emissions from a Portland cement manufacturing plant is pre-
sented as an appendix.
-1-
-------
2.0 INTRODUCTION
Pollutants emitted into the ambient air from an industrial plant
or other site generally fall into one of two types. The first type is
released into the air through stacks or similar devices designed to
direct and control the flow of the emissions. These emissions may be
readily measured by universally-recognized standard stack sampling tech-
niques. The second type is released into the air without control of
flow or direction. These fugitive emissions usually cannot be measured
using existing standard techniques.
The development of reliable, generally applicable measurement pro-
cedures is a necessary prerequisite to the development of strategies
for the control of fugitive emissions. This document describes some
procedures for the measurement of fugitive emissions using the upwind-
downwind measurement method described in Section 2.1.3 below.
2.1 Categories of Fugitive Emissions
Fugitive emissions emanate from such a wide variety of circumstances
that it is not particularly meaningful to attempt to categorize them
either in terms of the processes or mechanisms that generate them, or
the geometry of the emission points. A more useful approach is to cate-
gorize fugitive emissions in terms of the methods for their measurement.
Three basic methods exist—quasi-stack sampling, roof monitor sampling,
and upwind-downwind sampling. Each is described in general terms below.
2.1.1 Quasi-Stack Sampling Method
In this method, the fugitive emissions are captured in a temporarily
installed hood or enclosure and vented to an exhaust duct or stack of
-2-
-------
regular cross-sectional area. Emissions are then measured In the ex-
haust duct using standard stack sampling or similar well recognized
methods. This approach Is necessarily restricted to those sources of
emissions that are isolable and physically arranged so as to permit the
Installation of a temporary hood or enclosure that will not interfere
with plant operations or alter the character of the process or the emis-
sions .
2.1.2 Roof Monitor Sampling Method
This method is used to measure the fugitive emissions entering the
ambient air from building or other enclosure openings such as roof moni-
tors, doors and windows. The method is especially applicable to situa-
tions in which enclosed sources are too numerous or physically configured
to preclude the application of the quasi-stack method to each source.
Sampling is, in general, limited to a mixture of all uncontrolled emis-
sion sources within the enclosure and requires the ability to make low
velocity exhaust air measurements and mass balances of small quantities
of materials entering and leaving the enclosure through the openings.
2.1.3 Upwind-Downwind Sampling Method
This method is utilized to measure the fugitive emissions from
sources typically covering large areas that cannot be temporarily hooded
and are not enclosed in a structure allowing the use of the roof moni-
tor method. Such sources include material handling and storage opera-
tions, waste dumps, and industrial processes in which the emissions are
spread over large areas. These features are embodied in the typical
-3-
-------
industrial sources and their emitted pollutants listed in Table 2-1.
The upwind-downwind method quantifies the emissions from such sources
as the difference between the pollutant concentrations measured in the
ambient air approaching (upwind) and leaving (downwind) the source site.
It may also be utilized in combination with mathematical models and
tracer tests to define the contributions to total measured emissions of
specific sources among a group of sources.
2.2 Sampling Method Selection
The initial step in the measurement of fugitive emissions at an
industrial site is the selection of the most appropriate sampling method
to be employed. Although it is impossible to enumerate all the combina-
tions of influencing factors that might be encountered in a specific
situation, careful consideration of the following general criteria should
result in the selection of the most effective of the three sampling
methods described above.
2.2.1 Selection Criteria
The selection criteria listed below are grouped into three general
classifications common to all fugitive emissions measurement methods.
The criteria are intended to provide only representative examples and
should not be considered a complete listing of influencing factors.
2.2.1.1 Site Criteria
Source Isolability. Can the emissions be measured separately from
emissions from other sources? Can the source be enclosed?
Source Location. Is the source indoors or out? Does location
-4-
-------
TABLE 2-1
UPWIND-DOWNWIND SAMPLING METHOD
APPLICATION TO TYPICAL INDUSTRIAL FUGITIVE EMISSION SOURCES
I
Ol
I
Industry
Coke
Making
Primary
Aluminum
Primary
Copper
Sand &
Gravel
Source
Coal Handlin
Coal Storage
Charging Ov-
ens
Coking, Door
& Oven Leak
Coke Pushing
Quenching
(Controlled
but some
leaks)
Coke Handlin
Coke Storage
Bauxite Han-
dling &
Storage
Alumina Cal-
cining &
Preparation
Alumina Stor-
age
Mining
Hauling
Tailings Pond
Slag Dump
Quarrying
Truck Hauling
Delivery &
Storage
Rock Transfer
Crushing &
Screening
Drying (Leak-
age)
Product Stor-
age
Product Load-
ine
Ai*s
'roduct
Delivery
Par-
ticulate
Emission
Coal Dust
Coal Dust
Coal Dust
Tars
Coke Dust
Tars
Coke Dust
Tars
Coke Dust
Tars
Coke Dust
Ore Dust
Alumina
Dust
Alumina
Dust
Dust
Dust
Dust
Dust
ust
ust
ust
ust
ust
ust
ust
ust
Gas & Vapor
Emissions
HC.CO.NO
Pyridine
HC,CO,NOX
H2S,NH3,CS2
Phenol
HC,CO,H2S,NH
HCN, Phenol
HC,CO,H2S,NH
HCN,S02,
Phenol
-
-
-
S02
Industry
Electri
Furnace
Steel
Iron &
Steel
Found-
ries
Coal
Asphalt
Source
Scrap & Sinte
Delivery
Lime & Silica
Delivery
Pollution Con
trol Equip-
ment -
Dust Transfer
& Storage
Storage Ponds
for Slagging
Waste Water
Coke, Silica
Sinter Deliv
ery & Stor-
age
Pollution Con
trol Equip-
ment Transfe
& Storage
Mining
Hauling
Storage &
Transfer
Screening &
Crushing
Transfer
Drying
Storage Piles
Waste Transfer
Gravel Deliv-
ery
Asphalt Stor-
age
torage Piles
sphalt Batch-
ing
rier & Blower
eactor Dis-
charge
eactor Charge
roduct Trans-
fer
Par-
ticulate
Emission
Iron Dus
Steel
Dust
Dust
Fume Dus
-
Dust
Dust
Coal Dust
Coal Dust
Coal Dust
Coal Dust
Coal Dust
Coal Dust
Coal Dust
Dust
ust
ars
ust
ust,
Tars
ust,
Tars
ust,
Tars
ust,
Tars
ust,
Tars
Gas & Vapor
Emissions
-
-
H2S,S02
-
-
-
Hydrocarbons ,
CO
ydrocarbons ,
Odors
ydrocarbons ,
Odor
ydrocarbons ,
Odor
rdrocarbons ,
Odor
'drocarbons ,
Odor
'drocarbons ,
Odor
Industry
Coal
Gasifica-
tion
Petroleum
Refining
Source
Coal Delivery
& Storage
Coal Transfer
Waste .Trans-
fer
Scrubber Sol-
ids
Settling Pond
Crude Transfe
Crude Storage
Distilate
Storage
: Distilate
Transfer
Gasoline Stor-
age & Transf ei
hosphate
ertili-
er
Leakage in
Drains &
Sewers
Waste Water
Storage &
Transfer
Process Leaks
Gasoline Ter-
minal Loading
Mining
Storage Piles
Rock Trans-
fer (Truck,
Conveyer)
Settling Pond
jypsum Pile
'roduct Stor-
'roduct Deliv-
ery
Par-
tlculate
Emission
Coal Dus
Coal Dus
Dust
Dust
_
Dust
Tars
ust
Dust
ust
'luorides
Dust
Dust
Dust
Gas & Vapor
Emissions
Hydrocarbons ,
H2S,NH3,S02
Phenol
Hydrocarbons
H2S,S02,NH3
Phenol ,
Hydrocarbons
Hydrocarbons ,
RSH,H2S
Hydrocarbons ,
RSH,H2S
tydrocarbons
Hydrocarbons
Hydrocarbons
Hydrocarbons ,
H2S,HF,
Phenols
H2S,S02,NH3,
Phenols, HF,
Hydrocarbons
H2S,SO2,NH3,
Phenols, HF,
Hydrobarbons
Hydrocarbons
02 f Fluorides
02
Luorides
Luorides
-------
permit access of measuring equipment?
Meteorological Conditions. What are the conditions representative
of typical and critical situations? Will precipitation interfere
with measurements? Will rain or snow on ground effect dust levels?
2.2.1.2 Process Criteria
Number and Size of Sources. Are emissions from a single, well
defined location or many scattered locations? Is source small
enough to hood?
Homogeneity of Emissions. Are emissions the same type everywhere
at the site? Are reactive effects between different emissions
involved?
Continuity of Process. Will emissions be produced long enough to
obtain meaningful samples?
Effects of Measurements. Are special procedures required to pre-
vent the making of measurements from altering the process or emis-
sions or interfering with production? Are such procedures feasible?
2.2.1.3 Pollutant Criteria
Nature of Emissions. Are measurements of particles, gases, liquids
required? Are emissions hazardous?
Emission Generation Rate. Are enough emissions produced to provide
measurable samples in reasonable sampling time?
Emission Dilution. Will transport air reduce emission concentra-
tion below measurable levels?
2.2.2 Application of Criteria
The application of the selection criteria listed in Section 2.2.1
to each of the fugitive emissions measurement methods defined in Section
2.1 is described in general terms in this section.
-6-
-------
2.2.2.1 Quasi-Stack Method
Effective use of the quasi-stack method requires that the source
of emissions be isolable and that an enclosure can be installed capable
of capturing emissions without interference with plant operations. The
location of the source alone is not normally a factor. Meteorological
conditions usually need be considered only if they directly affect the
sampling.
The quasi-stack method is usually restricted to a single source
and must be limited to two or three small sources that can be effectively
enclosed to duct their total emissions to a single sampling point.
Cyclic processes should provide measurable pollutant quantities during
a single cycle to avoid sample dilution. The possible effects of the
measurement on the process or emissions is of special significance in
this method. In many cases, enclosing a portion of a process in order
to capture its emissions can alter that portion of the process by chang-
ing its temperature profile or affecting flow rates. Emissions may be
similarly altered by reaction with components of the ambient air drawn
into the sampling ducts. While these effects are not necessarily limit-
ing in the selection of the method, they must be considered in designing
the test program and could influence the method selection by increasing
complexity and costs.
The quasi-stack method is useful for virtually all types of emis-
sions. It will provide measurable samples in generally short sampling
times since it captures essentially all of the emissions. Dilution of
the pollutants of concern is of little consequence since it can usually
be controlled in the design of the sampling system.
-7-
-------
2.2.2.2 Roof Monitor Method
Practical utilization of the roof monitor method demands that the
source of emissions be enclosed in a structure with a limited number of
openings to the atmosphere. Measurements may usually be made only of
the total of all emissions sources within the structure. Meteorological
conditions normally need not be considered in selecting this method
unless they have a direct effect on the flow of emissions through the
enclosure opening.
The number of sources and the mixture of emissions is relatively
unimportant since the measurements usually include only the total emis-
sions. The processes involved may be discontinuous as long as a repre-
sentative combination of the typical or critical groupings may be in-
cluded in a sampling. Measurements will normally have no effect on the
processes or emissions.
The roof monitor method, usually dependent on or at least influ-
enced by gravity in the transmission of emissions, may not be useful
for the measurement of larger particulates which may settle within the
enclosure being sampled. Emission generation rates must be high enough
to provide pollutant concentrations of measurable magnitude after dilu-
tion in the enclosed volume of the structure.
2.2.2.3 Upwind-Downwind Method
The upwind-downwind method, generally utilized where neither of
the other methods may be successfully employed, is not influenced by
the number or location of the emission sources except as they influence
-------
the locating of sampling devices. In most cases, only the total con-
tribution to the ambient atmosphere of all sources within a sampling
area may be measured. The method is strongly influenced by meteorolog-
ical conditions, requiring a wind consistent in direction and velocity
throughout the sampling period as well as conditions of temperature,
humidity and ground moisture representative of normal ambient condi-
tions .
The emissions measured by the upwind-downwind method may be the
total contribution from a single source or from a mixture of many sources
in a large area. Continuity of the emissions is generally of secondary
importance since the magnitude of the ambient air volume into which the
emissions are dispersed is large enough to provide a degree of smooth-
Ing to cyclic emissions. The measurements have no effect on the emis-
sions or processes involved.
Most airborne pollutants can be measured by the upwind-downwind
method. Generation rates must be high enough to provide measurable
concentrations at the sampling locations after dilution with the ambient
air. Settling rates of the larger particulates require that the sampling
system be carefully designed to ensure that representative particulate
samples are collected.
2.3 Sampling Strategies
Fugitive emissions measurements may, in general, be separated into
two classes or levels depending upon the degree of accuracy desired.
Survey measurement systems are designed to screen emissions and provide
gross measurements of a number of process influents and effluents at a
-9-
-------
relatively low level of effort in time and cost. Detailed systems are
designed to isolate, identify, and quantify individual contaminant con-
stituents with increased accuracy and higher investments in time and
cost.
2.3.1 Survey Measurement Systems
Survey measurement systems employ recognized standard or state-
of-the-art measurement techniques to screen the total emissions from a
site or source and determine whether any of the emission constituents
should be considered for more detailed investigation. They generally
utilize the simplest available arrangement of instrumentation and pro-
cedures in a relatively brief sampling program, usually without pro-
visions for sample replication, to provide order-of-magnitude type data,
embodying a factor of two to five in accuracy range with respect to
actual emissions.
2.3.2 Detailed Measurement Systems
Detailed measurement systems are used in instances where survey
measurements or equivalent data indicate that a specific emission con-
stituent may be present in a concentration worthy of concern. Detailed
systems provides more precise identification and quantification of spe-
cific constituents by utilizing the latest state-of-the-art measurement
instrumentation and procedures in carefully designed sampling programs.
These systems are also utilized to provide emission data over a range
of process operating conditions or ambient meteorological influences.
Basic accuracy of detailed measurements is in the order of +10 to + 50
-10-
-------
percent of actual emissions. Detailed measurement system costs are
generally in the order of three to five times the cost of a survey sys-
tem at a given site.
-11-
-------
3.0 TEST PROGRAM PROCEDURES
This section describes the procedures required to successfully
complete a testing program utilizing the upwind-downwind sampling method
described in Section 2.1. It details the information required to plan
the program, describes the organization of the test plan, specifies the
types of sampling equipment to be used, establishes criteria for the
sampling system design, and outlines basic data reduction methods.
3.1 Pretest Survey
After the measurement method to be utilized in documenting the
fugitive emissions at a particular site has been established using the
criteria of Section 2.2, a pretest survey of the site should be con-
ducted by the program planners. The pretest survey should result in an
informal, internal report containing all the information necessary for
the preparation of a test plan and the design of the sampling system by
the testing organization.
This section provides guidelines for conducting a pretest survey
and preparing a pretest survey report.
3.1.1 Information to be Obtained
In order to design a system effectively and plan for the on-site
sampling of fugitive emissions, a good general knowledge is required of
the plant layout, process chemistry and flow, surrounding environment,
and prevailing meteorological conditions. Particular characteristics
of the site relative to the needs of the owner, the products involved,
the space and manpower skills available, emission control equipment
-12-
-------
installed, and the safety and health procedures observed, will also
influence the sampling system design and plan. Work flow patterns and
schedules that may result in periodic changes in the nature or quantity
of emissions or that indicate periods for the most effective and least
disruptive sampling must also be considered. Most of this information
can only be obtained by a survey at the site. Table 3-1 outlines some
of the specific information to be obtained. Additional information will
be suggested by considerations of the particular on-site situation.
3.1.2 Report Organization
The informal, internal pretest survey report must contain all the
pertinent information gathered during and prior to the site study. A
summary of all communications relative to the test program should be
included in the report along with detailed descriptions of the plant
layout, process, and operations as outlined in Table 3-1. The report
should also incorporate drawings, diagrams, maps, photographs, meteoro-
logical records, and literature references that will be helpful in plan-
ning the test program.
3.2 Test Plan
3.2.1 Purpose of a Test Plan
Measurement programs are very demanding in terms of the scheduling
and completion of many preparatory tasks, observations at sometimes
widely separated locations, instrument checks to verify measurement
validity, etc. It is therefore essential that all of the experiment
design and planning be done prior to the start of the measurement pro-
-13-
-------
TABLE 3-1
PRE-TEST SURVEY INFORMATION TO BE OBTAINED
FOR APPLICATION OF FUGITIVE EMISSION SAMPLING METHODS
Plant
Layout
Drawings:
Building Layout and Plan View of Potential Study Areas
Building Side Elevations to Identify Obstructions and
Structure Available to Support Test Setup ;
Work Flow Diagrams
Locations of Suitable Sampling Sites
Physical Layout Measurements to Supplement Drawings
Work Space Required at Potential Sampling Sites
Process
Process Flow Diagram with Fugitive Emission Points
Identified
General Description of Process Chemistry
General Description of Process Operations Including
Initial Estimate of Fugitive Emissions
Drawings of Equipment or Segments of Processes Where
Fugitive Emissions are to be Measured
Photographs (if permitted) of Process Area Where
Fugitive Emissions are to be Measured
Names, Extensions, Locations of Process Foremen and
Supervisors Where Tests are to be Conducted
Operations
Location of Available Services (Power Outlets, Main-
tenance and Plant Engineering Personnel, Labora-
tories, etc.)
Local Vendors Who Can Fabricate and Supply Test System
Components
Shift Schedules
Location of Operations Records (combine with process
operation information)
Health and Safety Considerations
Other
Access routes to the areas Where Test Equipment/Instru-
mentation Will Be Located
Names, Extensions, Locations of Plant Security and
Safety Supervisors
-14-
-------
gram in the form of a detailed test plan. The preparation of such a
plan enables the investigator to "pre-think" effectively and cross-check
all of the details of the design and operation of a measurement program
prior to the commitment of manpower and resources. The plan then also
serves as the guide for the actual performance of the work. The test
plan provides a formal specification of the equipment and procedures
required to satisfy the objectives of the measurement program. It is
based on the information collected in the informal pretest survey re-
port and describes the most effective sampling equipment, procedures,
and timetables consistent with the program objectives and site charac-
teristics .
3.2.2 Test Plan Organization
The test plan should contain specific information in each of the
topical areas indicated below:
Background
The introductory paragraph containing the pertinent infor-
mation leading to the need to conduct the measurement program and
a short description of the information required to answer that
need.
Ob j ective
A concise statement of the problem addressed by the test
program and a brief description of the program's planned method
for its solution.
Approach
A description of the measurement scheme and data reduction
methodology employed in the program with a discussion of how each
will answer the needs identified in the background statement.
-15-
-------
Instrumentation/Equipment/Facilities
A description of the instrumentation arrays to be used to
collect the samples and meteorological data identified in the
approach description. The number and frequency of samples to be
taken and the sampling array resolution should be described.
A detailed description of the equipment to be employed and
its purpose.
A description of the facilities required to operate the
measurement program, including work space, electrical power,
support from plant personnel, special construction, etc.
Schedule
A detailed chronology of a typical set of measurements or a
test, and the overall schedule of events from the planning stage
through the completion of the test program report.
Limitations
A definition of the conditions under which the measurement
project is to be conducted. If, for example, successful tests can
be conducted only during occurrences of certain wind directions,
those favorable limits should be stated.
Analysis Method
A description of the methods which will be used to analyze
the samples collected and the resultant data, e.g., statistical or
case analysis, and critical aspects of that method.
Report Requirements
A draft outline of the report on the analysis of the data to
be collected along with definitions indicating the purpose of the
report and the audience for which it is intended.
Quality Assurance
The test plan should address the development of a quality
assurance program as outlined in Section 3.7. This QA program
should be an integral part of the measurement program and be in-
corporated as a portion of the test plan either directly or by
reference.
Responsibilities
A list of persons who are responsible for each phase of the
measurement program, as defined in the schedule, both for the
testing organization and for the plant site.
-16-
-------
3.3 Upwind-Downwind Sampling Strategies
The upwind-downwind sampling method, as described in Section 2.1.3,
is used to quantify the emissions from a source to the ambient atmosphere
by measuring pollutant levels in the atmosphere. Upwind measurements
are made within the ambient air approaching the site of the source,
using sampling equipment suitable for the specific emissions to be mea-
sured, to determine the baseline concentration of pollutants in the
air. Downwind measurements are made of the air within the cloud of
pollutants emitted by the source, using sampling equipment similar to
that used for the upwind measurements, to determine the total of the
ambient air and the source's contribution to the concentration of pol-
lutants. The pollutants contributed by the source to the cloud at the
sampling locations are determined as the difference between the measured
upwind and downwind concentrations. Measurement of the wind speed and
direction at the site are combined with the pollutant concentrations at
the sampling locations in diffusion equations to back-calculate the
source strength of the emissions. Section 3.4 and 3.5 describe the
equipment used for sampling, the criteria for sampling system design,
sampling techniques, and data reduction procedures for respectively,
survey and detailed upwind-downwind sampling programs.
3.4 Survey Upwind-Downwind Measurement System
A survey measurement system, as defined in Section 2.3, is designed
to provide gross measurements of emissions to determine whether any
constituents should be considered for more detailed investigation. A
survey upwind-downwind measurement system in its simplest form utilizes
-17-
-------
a single upwind sampler for the determination of the concentration of
the pollutants of concern in the ambient air approaching the source of
the emissions and two or three identical downwind samplers for the de-
termination of the pollutant concentration and distribution in the am-
bient air leaving the source. These data, combined with measurements
of the ambient air wind speed and direction, are used to calculate the
emission rate of the source.
3.4.1 Sampling Equipment
Pollutants that may be measured by the upwind-downwind technique
are limited to those that can be airborne for significant distances,
i.e., particulates and gases. The gross measurement requirements for
survey sampling of particulates are best satisfied by high volume fil-
ter devices to provide data on the average emission rate, particle size
distribution, and particle composition. Particle charge transfer or
piezoelectric mass monitoring devices may be utilized for continuous or
semi-continuous sampling of intermittent emission sources where peak
levels must be defined.
Gaseous emissions in survey programs are usually grab-sampled for
laboratory analysis using any of a wide variety of evacuated sampling
vessels or chemical bubblers. Continuous or semi-continuous sampling
of specific gases may be accomplished using such devices as continuous
monitor flame ionization detectors (for hydrocarbons) and automated
flame photometric devices (for sulfur dioxide).
-18-
-------
3.4.2 Sampling System Design
The number and location of the devices used to collect samples is
extremely important to the successful completion of a survey upwind-
downwind sampling program, especially since the program is designed for
minimum cost and provides for no replication of samples. The design of
the sampling system is influenced by such factors as source complexity
and size, site location and topography, and prevailing meteorological
conditions which govern the distribution of the pollutant cloud in the
ambient atmosphere. Most situations will in general fit into some com-
bination of the following parameters:
Source - Sources may be either homogeneous, emitting a single type
or mixture of pollutants from each and every emission location, or
heterogeneous, emitting different types or mixtures of pollutants
from different locations. The resultant cloud of pollutants will,
for a homogeneous source, be homogeneous. The pollutant cloud for
a heterogeneous source may be either heterogeneous or, as a result
of mixing by suitably directed or turbulent ambient air flow, homo-
geneous . The physical size of a source will determine the extent
of the pollutant cloud and may influence its homogeneity, the prox-
imity of different emissions to each other largely influencing the
degree of mixing in the cloud for a given downwind distance.
Site - Sites in general may be open on level terrain with free
access of ambient air from all sides, partially obstructed by hills
or buildings that interfere with or influence the ambient air flow
either up- or downwind, or located in a valley between hills or
large buildings that influence the air flow both up- and downwind.
Each type of topography will influence the extent and homogeneity
of the pollutant cloud depending on the direction of the wind flow
relative to the obstructions.
Meteorology - The direction of the prevailing wind determines the
basic location of upwind and downwind samplers. It will influence
the pollutant cloud in every instance except that of a homogeneous
cloud at an open level site. In other instances, the wind may be
directed generally across or parallel to obstructing hills or
valleys which may result in channeling, lofting, or swirling of
the air flow across the site that will distort the pollutant cloud.
The homogeneity of the ambient air approaching the measurement
site, while not in the strict sense a meteorological condition,
-19-
-------
may affect the composition and distribution of different pollutants
within the pollutant cloud. Contributions from sources upwind of
the site may result in variations in the pollutant concentrations
in the ambient air passing over the site and thus in the pollutant
cloud as well.
Wind speed, which can affect the cloud's size and distribution,
need not be considered as a governing design factor since it is to
some degree controllable by scheduling to avoid periods of either
excessive wind velocity or calm conditions. Wind speeds within
normal limits are taken into consideration in data reduction cal-
culations.
Table 3-2 presents a matrix of 20 possible combinations of these
parameters (cloud homogeneity, site topography, wind direction and am-
bient air homogeneity). The simplest combination, that of a homogeneous
cloud in an open level site with homogeneous ambient air, would typically
require a single upwind sampler and two downwind samplers located within
the cloud. The complexity of the sampler system design is, in general,
increased by changes in the parameters as follows:
Cloud Homogeneity. A heterogeneous cloud will generally limit the
placement of the downwind samplers to the portion of the cloud
that contains the combined emissions from the various sources. It
may also require the addition of samplers in the cloud to provide
data on the extent of the effects of the heterogeneity and the
consequent variability of the pollutant distributions. This param-
eter will not affect the upwind samplers.
Site Topography. Depending on the relationship of the topography
obstructions and the wind direction, this parameter may affect
both upwind and downwind samplers. Hills and valleys may cause
lofting or depression of the pollutant cloud, requiring sampler
elevation on towers or limiting the downwind distance of samplers
within the cloud. They may also provide funnelling effects that
limit the dispersion of the cloud and restrict the lateral position-
ing of the downwind samplers. Upwind sampler locations may be
restricted by lofting or depression of the ambient air approaching
the site.
Wind Direction. Changes in this parameter alone are not generally
a major factor in the sampling system design. They will dictate
-20-
-------
TABLE 3-2
MATRIX OF SAMPLING SYSTEM DESIGN PARAMETERS
I
N>
Cloud
Homogeneity
Homogeneous
CD
Heterogeneous
(2)
Site
Topography
Open
(1)
Hill
(2)
Valley
(3)
Open
(i)
Hill
(2)
Valley
(3)
Wind
Direction
Not a
Factor
(0)
Parallel to
Hill (1)
Over Hill
(2)
Down Valley
(1)
Across Valley
(2)
Not a
Factor
(0)
Parallel to
Hill (1)
Over Hill
(2)
Down Valley
(1)
Across Valley
(2)
Ambient Air
Homogeneity
Homogeneous
(1)
Heterogeneous
(2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homogeneous
(1)
Heterogeneous
(2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Homo (1)
Hetero (2)
Cloud Homogeneity:
Homogeneous - Sources at site all emitting same pollutants.
- Sources of different pollutants grouped so that
emissions are mixed before sampling.
Heterogeneous - Sources at site emitting identifiably different
pollutants - no mixing before sampling.
Site Topography:
Open - Site on flat terrain ambient air access from any
direction unhindered.
Hill - Site close enough to rise in terrain or large
buildings to cause channeling or lofting Of am-
bient air.
Valley - Site between rises in terrain or large buildings.
Uind Direction:
Parallel to - Wind across site channeled against side of hill.
Hill Usually changes shape of pollutant cloud and
distribution of pollutants within cloud.
Across Hill - Wind across site from or to hill top. Can cause
lofting of depression of pollutant cloud.
Down Valley - Wind across site channeled against sides of hills.
Across Valley - Wind across site from hill to hill.
Homogeneous - Pollutants in approaching air evenly distributed.
Heterogeneous - Pollutants in approaching air measurably different
at points over site. Usually caused by emissions
from nearby upwind external source.
-------
changes in the design in combination with other factors such as
site topography, described above, or the presence of external
sources, which may influence the homogeneity of the approaching
ambient air, described below.
Ambient Air Homogeneity. The presence of external emission sources
that may result in variations in the pollutant concentrations and
distributions in the air approaching a site may require the addi-
tion of samplers both upwind and downwind to ensure that the mea-
surements of the pollutants of interest are not unduly influenced
or masked. Samplers typically are required within and outside of
the external source cloud both upwind and downwind.
Typical sampler locations for selected source site configurations
illustrating some of these effects are sketched in Figure 3-1. The
configurations are identified by a four-digit number referring, in left-
to-right order, to the numbers assigned to 'the parameters identified in
the matrix of Table 3-2. A configuration with a homogeneous cloud emitted
at a valley site with cross-valley wind direction and homogeneous am-
bient air is thus identified as 1321.
3.4.3 Sampling Techniques
Sampling must be scheduled and carefully designed to ensure that
data representative of the emission conditions of concern are obtained.
Effective scheduling demands that sufficient knowledge of operations
and process conditions be obtained to determine proper starting times
and durations for samplings. The primary concern of the sampling design
is that sufficient amounts of the various pollutants are collected to
provide meaningful measurements.
Each of the various sample collection and analysis methods has an
associated lower limit of detection, typically expressed in terms of
micrograms of captured solid material and either micrograms or parts
per million in air of gases. Samples taken must provide at least these
-22-
-------
Wind
1101
2102
1221
2311
Legend
Homogeous cloud source
ItsfD Heterogeneous cloud source
0 External source
A \ Sampler
--•:•:::: Source cloud
^^ External source cloud
Fig. 3-1. Typical sampler legations for selected source site configurations.
-23-
-------
minimum amounts of the pollutants to be quantified. The amount (M) of
a pollutant collected is the product of the concentration of the pollu-
tant in the air (x) and the volume of air sampled (V), thus,
M (micrograms) = x (micrograms/cubic meter) x V (cubic meters).
To ensure that a sufficient amount of pollutant is collected, an ade-
quately large volume of air must be passed through such samplers as
particle filters or gas absorbing trains for a specific but uncontrolla-
ble concentration. The volume of air (V) is the product of its flow
rate (F) and the sampling time (T),
V (cubic meters) = F (cubic meters/minute) x T (minutes).
Since the sampling time is most often dictated by the test conditions,
the only control available to an experimenter is the sampling flow
rate. A preliminary estimate of the required flow rate for any sam-
pling location may be made if an estimate or rough measurement of the
concentration expected is available. The subsitution and rearrangement
of terras in the above equations yields Equation 3-1:
F (cubic meters/minute) = M (micrograms)/x (micrograms/cubic meter)
x T (minutes). (3-1)
This equation permits the calculation of the minimum acceptable flow
rate for a required sample size. Flow rates should generally be adjusted
upward by a factor of at least 1.5 to compensate for likely inaccuracies
in estimates of concentration.
Grab-samples of gaseous pollutants provide for no means of pollu-
tant sample quantity control except in terms of the volume of the sample.
-24-
-------
Care should be taken, therefore, to correlate the sample size with the
requirements of the selected analysis method.
Sampler location is also important in obtaining representative
data. Downwind sampler location is especially critical to ensure that
samples are taken at points known to be within the pollutant cloud at
measurable concentrations. A rough estimate of acceptable downwind
sampler locations may be made utilizing the basic equation for the
diffusion of gases and particulates in the atmosphere from a ground-
level source: x = Q/ifKu, where
X = pollutant concentrations at receptor point, gm/m3
Q = source emission rate, gm/sec
K = product of standard deviations of vertical and
horizontal pollutant distribution, m2
u = wind speed, m/sec
This equation assumes a Gaussian distribution of pollutants in both the
vertical and horizontal directions and no deposition or reaction of
pollutants at the earth's surface.
By rearranging terms, the product of the standard deviations (K),
which are functions of the downwind distance (x) of the receptor from
the source, may be determined as a function of easily estimated or
measured parameters in Equation 3-2:
K - Q/TTXU, (3-2)
Turner, D. Bruce, "Workbook of Atmospheric Dispersion Estimates,"
U.S. Department of Health, Education and Welfare, Public Health Service
Publication No. 999-AP-26, Revised 1969.
-25-
-------
where
Q is estimated from published emission factors,
X is set equal to a selected value related to
the sampling method detection limit and
u is measured at the site.
The maximum downwind sampler distance from the source along the axis of
the wind direction (x) may then be determined from the curves of Figure
3-2, which relate K and x for various atmospheric stability categories.
These categories are listed and explained in Table 3-3.
When suitable x-distances, which may be any distance less than the
maximum determined from Figure 3-2, have been selected, cross-wind dis-
tances (y) perpendicular to the x-axis that will ensure that samples
are taken within the limits of the cloud must be determined. Maximum
cross-wind distances, which are a function of the distribution of the
pollutant concentrations within the cloud, are plotted as a function of
x in the curves of Figure 3-3 for the same atmospheric stability cate-
gories used in determining x. Downwind samplers should in general be
located at two different x-distances within the limits of the maximum
as determined above and at cross-wind y distances less than the maximum
indicated in Figure 3-3 on opposite sides of the wind direction axis.
Upwind samplers should ideally be located on the wind direction
axis just far enough upwind to prevent sampling the backwash of the
pollutant cloud. A minimum upwind distance of x /10, where x is
r max max
determined using x equal to the sampling method's lower detection limit,
will usually be sufficient.
-26-
-------
Atmospheric
stability
category
100
200 400 600 800
Maximum downwind sampler distance from
source along wind direction axis (x) - meters
Fig. 3-2. Maximum downwind sampler distances.
1000
-27-
-------
TABLE 3-3
ATMOSPHERIC STABILITY CATEGORIES
Wind Speed
m/sec
< 2
2-3
3-5
5-6
> 6
Dav*
Solar Altitude*
> 60°
A
A-B
B
C
C
35°-60°
A-B
B
B-C
C-D
D
15°-35°
B
C
C
D
D
Nig
Overcast or
> 50% Clouds
-
E
D
D
D
ht
< 50% Clouds
_
F
E
D
D
*Day is one hour after sunrise to one hour before sunset.
+Solar altitude may be determined form Table 170, Solar
Altitude and Azimuth, Smithsonian Meteorological Tables.
Use neutral class D for overcast conditions at any wind
speed. - Parital cloud cover (60 percent to 85 percent)
will reduce effective solar altitude one division (e.g.,
from > 60°. to 35°-60°) for middle clouds and two divi-
sions (e.g., from >60° to 15°-35°) for low clouds.
-28-
-------
300
200
100
E
I
I 50
a.
•a
o
u
E
3
E
'x
co
10
Atmospheric
stability
category
I
0 100 200 400 600 800
Downwind sampler distance (x) - meters
1000
Fig. 3-3. Maximum crosswind sampler distances.
-29-
-------
To illustrate the application of the equations and curves presented
in this section, assume a source emitting particulates into a four meter
per second wind at an estimated rate of 10 grams per second, and a sam-
pler with a lower detection limit of .001 gram and flow rate of 0.67
cubic meter per minute. For a sampling time of 10 minutes, the required
pollutant concentration, x» at the sampler is x = M/FT, where
M = .001 gram
F = 0.67 cubic meter/minute x 1.5 adjustment factor =
1 cubic meter/minute
T = 10 minutes, and
X = .001/10 = lO"1* grams/cubic meter
The product of the pollutant cloud's standard deviations, K, is found
in Equation 3-2, K = Q/irxu, where
Q = 10_grams/second
X = 10 ^ grams/cubic meter
u = 4 meters/second, and
K = 10/ir x!0ltx4 = 8x 103 meters squared
To measure the emissions during midday with clear skies, Table 3-3
indicates an atmospheric stability category B for the four meter/second
wind. Figure 3-2 for K = 8 x 103 and category B indicates a maximum
sampler downwind distance of 680 meters. Figure 3-3 for x = 680 meters
and category B indicates a maximum cross-wind distance of 145 meters.
Downwind samplers must then be located within the limits of a tri-
angle with an apex at the source, an altitude of 680 meters along the
wind direction axis and a base 145 meters wide on each side of the axis.
The upwind sampler should be located along the wind direction axis
at a minimum distance of x /10 = 68 meters from the source.
ma v
max
-30-
-------
A more detailed description of the application of this method is
presented in the appendix.
3.4.4 Data Reduction
When the sampling program has been completed and the samples have
been analyzed to yield pollutant concentrations in such terms as micro-
grams per cubic meter in the ambient air at each downwind sampling site,
the measured upwind concentrations are subtracted to yield the concen-
tration provided by the source at each sampler. These values are then
back-calculated through known diffusion equations that take into account
the variables of topography and meteorology to produce statistical dis-
tributions of the concentrations within a pollutant cloud generated by
a given source. These calculations yield source strengths of the emis-
sions in such terms as grams per unit time. A library of computer pro-
grams to assist in the performance of the calculations is maintained in
the User's Network for Applied Models of Air Pollution (UNAMAP) at the
Environmental Protection Agency's Research Triangle Computer Center.
Additional programs may be obtained through many environmental consul-
tants.
3.5 Detailed Upwind-Downwind Measurement System
A detailed measurement system is designed to more precisely iden-
tify and quantify specific pollutants that a survey measurement or
equivalent data indicate as a possible problem area. A detailed system
Bulletin American Meteorological Society, Vol. 56, No. 12,
December, 1975.
-31-
-------
is necessarily more complex than a survey system in terms of equipment,
system design, sampling techniques, and data reduction. It requires a
much larger investment in terms of equipment, time, and manpower and
yields data detailed and dependable enough for direct action toward
achieving emissions control. Detailed systems in general employ sam-
pling arrays or networks to measure the concentration and distribution
of specific pollutants in the ambient air approaching and leaving a
source. These actual measurements of the pollutant distribution within
a cloud and the variations in meteorological conditions during the sam-
pling period replace the assumptions utilized in survey sampling sys-
tems. Detailed systems are frequently employed to compare emissions at
different process or operating conditions to determine which conditions
dictate the need for emission control.
The data provided by the sampling arrays are processed in conjunc-
tion with more detailed meteorological data which are taken simultan-
eously to determine source emission rates and ambient distributions in
much the same manner as the simpler survey systems.
3.5.1 Sampling Equipment
The pcLltants to be characterized by a detailed upwind-downwind
sampling system fall into the same two basic classes—airborne partic-
ulates and gases—as those measured by survey systems. Detailed sam-
pling and analysis equipment is generally selected to obtain continuous
or semi-continuous measurements of specific pollutants rather than sim-
ple grab-sampled measurements.
Particulate samples are collected using filter impaction, piezo-
-32-
-------
electric, particle change transfer, light or radiation scattering, elec-
trostatic, and size selective or adhesive impaction techniques. Gases
are sampled and analyzed using flame ionization detectors, bubbler/im-
pinger trains, non-dispersive infrared or ultraviolet monitors, flame
photometry, and other techniques specific to individual gaseous pollu-
tants .
The selection of suitable sampling equipment should be influenced
by such considerations as portability, power requirements, detection
limits and ease of control.
3.5.2 Sampling System Design
The basic criteria reviewed in Section 3.4.2 for the design of a
survey sampling system are generally applicable to the design of a de-
tailed system. The need for replacement of survey assumptions as to
pollutant distribution with actual measured values, however, most fre-
quently requires the design of a sampling array or network that will
provide samples of a distribution at various distances downwind of the
source in both the horizontal and vertical directions. Sampler loca-
tions may generally be determined in the same manner as those for a
survey system. For detailed measurements, each location must provide
for sampling across a section of the pollutant cloud horizontally and/
or vertically. Horizontal distributions may be measured by adding a
number of samplers (usually at least two) at either side of the survey
sampler location at distances estimated to yield significantly differ-
ent pollutant concentrations. Vertical distributions may be measured
by placing a tower of suitable height at each survey sampler location
-33-
-------
and adding samplers over a range of heights on each tower. Combina-
tions of horizontal and vertical distributions may be measured by plac-
ing a grid of horizontally and vertically spaced samplers at each
survey sampler location. Actual numbers of samplers, their spacing,
and heights of towers required must be determined for each location. A
rough guide for estimating the required spacing is presented in Section
3.5.3.
3.5.3. Sampling Techniques
The guidelines presented in Section 3.4.3 for the design and loca-
tion of samplers for a survey system are applicable to detailed systems.
The assumption of a Gaussian distribution of pollutants in the cloud,
sufficient for data reduction in survey systems is reasonable as a rough
guide to locating samplers within the pollutant cloud as in Section
3.4.3, and for the spacing of sampling arrays as outlined below.
The approximate concentration of a specific pollutant within a
cloud in which concentrations vary in accordance with a Gaussian dis-
tribution at a given downwind distance from the source is greatest at
ground level on the wind direction axis of the cloud. Assigning this
concentration the unit value x , the concentration (x) at any cross-
wind distance (y) from die axis is expressed as
X = XQ e
The term F(y) may be expressed as y/y , where y is the maximum cross-
wind sampler distance determined from Figure 3-3. The relationship x/X
0
is plotted .as a function of y/y in Figure 3-4. This may be used to
-34-
-------
1.0
0.8
2
o
'5 0.6
8
4-1
I
2 0.4
0.2
0.5 1.0 1.5
Y/Y_ - crosswind distance ratio
2.0
Fig. 3-4. Pollutant concentration ratios for crosswind locations.
-35-
-------
determine the probable concentration at a sampler location relative to
the concentration at the axis and the concentrations at lateral dis-
tances from that location to assist in the horizontal spacing of sam-
plers in an array.
The concentration in the vertical direction from any ground level
point will decrease as the height, Z, increases in a similar relation-
ship. The ratio of the concentration at the elevated point to that at
ground level, Xi/X» is plotted in Figure 3-5 as a function of Z/Z , where
n ni
Z is a function of the downwind distance from the source and the atmo-
spheric stability as plotted in Figure 3-6. Figure 3-5 may be used to
determine the relative concentrations at elevated points to assist in
the design of sampling towers and the vertical spacing of samplers in
an array or grid.
In general, arrays should be designed to provide data at concentra-
tions approximately two to four times greater or less than the concen-
tration at a selected ground level sampling point. Physical limitations
at the site or very unstable atmospheric conditions will often preclude
the compliance with this design guideline by limiting the available
horizontal positions or by requiring an impractical tower height. In
such situations, the need to adjust the requirements of the guideline
must be recognized and the array designed to compensate for the limita-
tions.
Upwind sampling arrays will generally be less complex than downwind
arrays unless a nearby pollutant source results in a heterogeneous am-
bient air mix. In this case, the guidelines for downwind array design
presented in this section may have to be applied to the upwind array
-36-
-------
I I I I I I I I
0.5 1.0
Z/Z_ - vertical distance ratio
1.5
2.0
Fig. 3-5. Pollutant concentration ratios for vertical locations.
-37-
-------
500
100 —
N
s
t
Atmospheric
stability
category
100 200
800
400 600
Downwind distance (x) - meters
Fig. 3-6. Vertical concentration distribution factors.
1000
-38-
-------
design.
Wind speed and direction should be measured at each sampler or
array location. Pretest survey observations should indicate whether
stratification will occur to a degree which will require wind data at
more than one level.
An example of the application of these guidelines to the design of
survey and detailed systems for the measurement of pollutants at a. Port-
land cement plant is presented as an appendix to this document.
3.5.4. Data Reduction
Samples are analyzed to yield concentrations of specific pollutants
in such terms as micrograms per cubic meter at each sampling site.
Measured upwind concentrations are substituted into appropriate diffu-
sion equations to provide ambient air background concentrations at each
downwind site and the background concentration subtracted from the mea-
sured downwind concentration at each site to yield the source contri-
bution. These values are then substituted into diffusion equations to
back-calculate source strengths in terms of grams per unit time, util-
izing UNAMAP or other available computer programs.
3.6 Atmospheric Tracers
In some instances, prevailing process or meteorological conditions
prohibit the collection of samples containing measurable, clearly de-
fined amounts of a specific pollutant for the back-calculation of source
strengths. In many such cases, the atmospheric tracer method may be
employed to determine a typical distribution of a general class of pollu-
-39-
-------
tant analogous to the pollutant of concern.
The use of tracers should be considered under any of the following
circumstances:
When the pollutant background concentration is either excessively
high or inhomogeneous. This can be caused by significant emis-
sions from external upwind sources.
When the fugitive emissions are of such a complex nature that an
excessive number of downwind vertical profiles are required to
characterize the emissions.
When physical limitations prohibit the installation of adequate
instrumentation for specific pollutant concentration measurement.
When the nature of the specific pollutant prohibits its measure-
ment with acceptable instrumentation or indicates large probable
errors in measurement.
When estimates of fugitive emissions are being made for non-oper-
ating processes or planned operations.
The atmospheric tracer method, which may be considered as a special
detailed system, consists of the introduction into the atmosphere, at
the source site under consideration, of a readily identifiable material
similar in the character of its diffusion in the atmosphere to the pollu-
tant of concern. The quantity released may be controlled to provide
readily measurable concentrations. A detailed downwind measurement
system, designed using the guidelines of Section 3.5, is used to col-
lect samples of the tracer and to determine its dispersion for the known
and controllable source strength. This dispersion will be analogous to
the dispersion of the pollutant of concern and will permit the predic-
tion of pollutant concentrations for a range of source strengths.
-40-
-------
3.6.1 Tracers and Samplers
Both particulate and gaseous atmospheric tracers are in general
use. The most commonly used particulate tracers are zinc-cadmium sul-
fide and sodium fluorescein (urinine dye). The primary gaseous tracer
is sulfur hexafluoride (SF6).
Zinc-cadmium sulfide is a particulate material which can be ob-
tained in narrow size ranges to closely match the size of the pollutant
of concern. The material is best introduced into the atmosphere in dry
form by a blower type disseminator although it can also be accomplished
by spraying from an aqueous or solvent slurry. The zinc-cadmium sul-
fide fluoresces a distinctive color under ultraviolet light which pro-
vides a specific and rapid means of identification and quantification
of the tracer in the samples.
Sodium fluorescein is a soluble fluorescing particulate material.
It is normally spray disseminated from an aqueous slurry solution to
produce a particulate airborne plume, the size distribution of which
can be predetermined by the spraying apparatus. Sodium fluorescein can
be uniquely identified by colorimeter assessment.
Sulfur hexafluoride is a gas which can be readily obtained in ordi-
nary gas cylinders. Sulfur hexafluoride can be disseminated by meter-
ing directly from the gas cylinder through a flow meter to the atmosphere.
The amount disseminated can be determined by careful flow metering and/or
weight differentiation of the gas cylinder.
Particulate tracers are usually sampled with filter impaction de-
vices or, for particles over 10 microns in diameter, the more easily
used and somewhat less accurate Rotorod sampler which collects particles
-41-
-------
on an adhesive-coated U- or H-shaped rod which is rotated in the am-
bient air by a battery-driven electric motor.
Sulfur hexafluoride gaseous samples are collected for laboratory
gas chromatograph analysis in non-reactive bags of such materials as
Mylar.
3.6.2 Tracer Sampling System Design
All of the design guidelines presented in 3.4.2 and 3.5.2 may be
applied to the design of a tracer sampling system as site conditions
dictate. Their application is, in general, simplified since the source
strength may be controlled to provide measurable tracer concentrations
at readily accessible sampling locations.
A single upwind sampler will usually be sufficient to establish
that no significant amount of the tracer material is present in the
ambient atmosphere approaching the source.
3.6.3 Tracer Sampling and Data Analysis
The methods introduced in Section 3.4.3 and 3.5.3 for determining
sampler design and location are fully applicable to tracer sampling.
Like the design guidelines, they may be more easily applied because the
source strength is easily controlled.
The analysis of the data is also simplified since the source strength
is known and no back-calculation is required.
-42-
-------
3.7 Quality Assurance
The basic reason for quality assurance on a measurement program is
to insure that the validity of the data collected can be verified.
This requires that a quality assurance program be an integral part of
the measurement program from beginning to end. This section outlines
the quality assurance requirements of a sampling program in terms of
several basic criteria points. The criteria are listed below with a
brief explanation of the requirements in each area. Not all of the
criteria will be applicable in all fugitive emission measurement cases.
1. Introduction
Describe the project organization, giving details of the
lines of management and quality assurance responsibility.
2. Quality Assurance Program
Describe the objective and scope of the quality assurance
program.
3. Design Control
Document design requirements and standards applicable to
the measurement program as procedures and specifications.
4. Procurement Document Control
Verify that all design specification accompany procurement
documents such as purchase orders.
5. Instructions, Procedures, Drawings
Prescribe all activities that affect the quality of the
work performed by written procedures. These procedures must
include acceptance criteria for determining that these activ-
ities are accomplished.
6. Document Control
Ensure that the writing, issuance, and revision of proce-
dures which prescribe measurement program activities affecting
quality are documented and that these procedures are distrib-
-43-
-------
uted to and used at the location where the measurement program
is carried out.
7. Control of Purchase Material, Equipment, and Services
Establish procedures to ensure that purchased material
conforms to the procurement specifications and provide veri-
fication of conformance.
8- Identification and Control of Materials, Parts, and Components
Uniquely identify all materials, parts, and components
that significantly contribute to program quality for trace-
ability and to prevent the use of incorrect or defective ma-
terials, parts, or components.
9. Control of Special Processes
Ensure that special processes are controlled and accom-
plished by qualified personnel using qualified procedures.
10. Inspection
Perform periodic inspections where necessary on activities
affecting the quality of work. These inspections must be or-
ganized and conducted to assure detailed acceptability of
program components.
11. Test Control
Specify all testing required to demonstrate that applicable
systems and components perform satisfactorily. Specify that
the testing be done and documented according to written proce-
dures, by qualified personnel, with adequate test equipment
according to acceptance criteria.
12. Control of Measuring and Test Equipment
Ensure that all testing equipment is controlled to avoid
unauthorized use and that test equipment is calibrated and
adjusted at stated frequencies. An inventory of all test equip-
ment must be maintained and each piece of test equipment labeled
with the date of calibration and date of next calibration.
13. Handling, Storage, and Shipping
Ensure that equipment and material receiving, handling,
storage, and shipping follow manufacturer's recommendations to
prevent damage and deterioration. Verification and documentation
that established procedures are followed is required.
-44-
-------
14. Inspection, Test, and Operating Status
Label all equipment subject to required inspections and
tests so that the status of inspection and test is readily
apparent. Maintain an inventory of such inspections and oper-
ating status.
15. Non-Conforming Parts and Materials
Establish a system that will prevent the inadvertent use
of equipment or materials that do not conform to requirements,
16. Corrective Action
Establish a system to ensure that conditions adversely
affecting the quality of program operations are identified,
corrected, and commented on; and that preventive actions are
taken to preclude recurrence.
17. Quality Assurance Records
Maintain program records necessary to provide proof of
accomplishment of quality affecting activities of the measure-
ment program. Records include operating logs, test and in-
spection results, and personnel qualifications.
18. Audits
Conduct audits to evaluate the effectiveness of the mea-
surement program and quality assurance program to assure that
performance criteria are being met. . ••'
-45-
-------
4.0 ESTIMATED COSTS AND TIME REQUIREMENTS
Table 4-1 presents a listing of the conditions assumed for esti-
mating the costs and time requirements of upwind-downwind fugitive emis-
sions sampling programs using the methodology described in this document.
Four programs are listed, representing minimum and more typical levels
of effort for each of the survey and detailed programs defined in Sec-
tion 3.3. The combinations of conditions for each program are generally
representative of ideal and more realistic cases for each level and will
seldom be encountered in actual practice. They do, however, illustrate
the range of effort and costs that may be expected in the application of
the upwind-downwind technique except in very special instances.
4.1 Manpower
Table 4-2 presents estimates of manpower requirements for each of
the sampling programs listed in Table 4-1. Man-hours for each of the three
general levels of Senior Engineer/Scientist, Engineer/Scientist, and
Junior Engineer/Scientist are estimated for the general task areas outlined
in this document and for additional separable tasks. Clerical man-hours
are estimated as a total for each program. Total man-hour requirements
and approximately 500 man-hours for a simple survey program and 1500
man-hours for a more complex r.urvey program; and 2800 man-hours for a
simple detailed program and 4500 man-hours for a more complex detailed
program.
-46-
-------
TABLE 4-1
CONDITIONS ASSUMED FOR COST ESTIMATION
OF UPWIND - DOWNWIND SAMPLING PROGRAMS
Parameter
Site
Location
Emmission
Source
Emission
Character
Wind
Measurement
Sample
Sites
Samplers
Towers
Experiments
Estimated
Basic
Accuracy
Survey Program
Simple
Open Area-
Accessible
Well Defined
Steady
External
Source
One Upwind
Two Downwind
3
0
1
± 500%
Complex
Congested-
Limited Access
Complex
Steady
Measured
On Site
One Upwind
Three Downwind
8
4 Low
1
± 150%
Detailed Program
Simple
Open Area-
Accessible
Well Defined
Cyclic
Measured
On Site
Vertical
Arrays-
One Upwind,
Two Downwind
16
4 High
2
± 125%
Complex
Congested-
Limited Access
Complex
Cyclic-
Measured
at Two Levels
Two Measure-
ments On Site
Grid Arrays-
One Upwind,
Two Downwind
30
4 High -
Grids
4
± 75%
-47-
-------
TABLE 4-2
ESTIMATED MANPOWER REQUIREMENTS FOR UPWIND - DOWNWIND
SAMPLING PROGRAMS
00
i
— Estimates in Man-Hemrs
Task
Pretest Survey
Test Plan Preparation
Equipment Acquisition
Field Set-Up
Field Study
Sample Analysis
Data Analysis
Report Preparation
Totals
Engineer/Scientist Tota]
Clerical
Grand Total
Survey Programs
Senior
Engr/Sci
4
8
4
8
24
20
20
16
104
Engr/
Sci
12
12
4
12
48
20
20
16
144
472
40
512
Junior
Engr/
Tech
0
0
12
12
120
36
36
8
224
Senior
Engr/Sci
8
12
4
16
60
60 ;
60 j
64
284
!
|
1
• 1 1 i
Engr/
Sci
24
16
8
48
200
60
60
64
480
1376
120
1496
Junior
Engr/
Tech
0
4
28
64
268
108
108
32
612
Senior
Engr/Sci
Q
12
8
80
120
120
120 |
160 i
628 i
Detailed Programs
Engr/
Sci
24
24
24
140
300
240
240
80
972
J2628
i
200
2828
Junior
Engr/
Tech
0
12
48
100
536 :
96
96
40 ;
1028
Senior
Engr/Sc
12
16
12
120
180
180
180
200
900
i
uomplex
Engr/
Sci
36
32
36
280
560
320
320
200
1784
4240
Junior
Engr/
Tech
16
12
52
240
796
180
180
80
1556
280
4520 !
-------
4.2 Other Direct Costs
Table 4-3 presents estimates for equipment purchases, rentals,
calibration, and repairs; on-site construction of towers and platforms;
shipping and on-site communications for each of the listed programs.
Total costs are approximately $4500 for a simple survey program and
$17,000 for a more complex survey program; and $34,000 for a simple
detailed program and $64,000 for a more complex detailed program.
4.3 Elapsed-Time Requirements
Figure 4-1 presents elapsed-time estimates for each of the listed
programs broken down into the task areas indicated in the manpower es-
timates of Table 4-2. Total program durations are approximately 12
weeks for a simple survey program and 17 weeks for a more complex sur-
vey program; and 21 weeks for a simple detailed program and 41 weeks
for a more complex detailed program.
4.4 Cost Effectiveness
Figure 4-2 presents curves of the estimated cost effectiveness of
the upwind-downwind technique, drawn through points calculated for the
four listed programs. Costs for each program were calculated at $30
per labor hour, $40 per man day subsistence for field work for the man-
power estimates of Table 4-2, plus the other direct costs estimated in
Table 4-3.
-49-
-------
-os-
H P < H CO O W
O P fD i-t p* o ua
H 1 O* (» H- W^TJHP SIPOOSJCOCOC
>• COH-H-XJ M M O CO H- P> H- (D P> H-
f1 H- O M "O (D pJ t! rt f( "d M 3 .a B "d
rt M (D H- O rt (D i-i H- » H- Q- C "d B
TO (D i-l 3 rf Mi i-l C 3H-CT (!) M (D
OQ i-iown OQ i-i i-< 3; 3 re 3
OJOJO H-i-t ft MftoftO^n-
orem os H- na rt ta m
333 P)tn o 0) H-COI-(ITJ
grtrt I-1- 3 H OC C
3MM tc (o n> o>co
M-05 Ort en BMP'
n O o en fH o CD
Co ?r • (D 3 p1 os
rt 1 05 rt P) fD
o T3 ^d re
3 • CD C
en i-t
O
CD
X
re
o
o
-c/>
QT\
00
h~*
o
*»
o
. 0
-0>
u>
vO
o
o
MMNJ Ln Ml— 'VOOO
OOOO O *-JUiOO
OOOO OOO OOOOOO
1— ' NJ I-1 LO NJ J>
OJOi *» (-"OO ONiLnJ^OO
OON O OOO OUiOOO
OOOO OOO OOOOOO
I-1 t-1 00 M U> *- VO
OOOO OOO OOOOOO
OOOO OOO OOOOOO
l-> I-I
MM MVO MOivOOO
vOOOLnM U) O O OOOiOOO
OOOO OOO OOOOOO
OOOO OOO OOOOOO
o
o
CO
rt
H
rt
a
CO
1
nT
n
o
9
*w
re*
X
CO
L J,
|
re
0
o
B
T3
M
re
X
co
^
re
o
OQ
i-i
Co
en
0
re
CD
H-
M
o
OQ
P)
B
CO
M
-------
Weeks
Task 15 10 15 20 25 30 35 40 45
" I I I I I I i I I I i i I llJJIilllllllllillllllllllllil J I
I
01
Pre-test
survey
Test plan
preparation
Equipment
acquisition
Field
set-up
Field
study
Sample
analysis
Data
analysis
Report
preparation
s^
k
^^
mi
•=fe
VTTTi
Simple survey program
Complex survey program
Simple detailed program
detailed program
15
25
35
Weeks
Fig. 4-1. Elapsed-time estimates for upwind-downwind sampling programs.
45
-------
500^
400
300
200
100
Survey program
Detailed program
50
100 150 200
Cost in thousands of dollars
250
300
Fig. 4-2. Cost-effectiveness of upwind-downwind sampling programs.
-52-
-------
APPENDIX A
TEST PROCEDURES APPLICATION
-53-
-------
A.1.0 INTRODUCTION
This appendix presents an application of the upwind-downwind
fugitive emissions measurement system selection and design criteria to a
Portland cement manufacturing plant. The criteria for the selection
of the method and the design procedures for both survey and detailed
sampling systems as presented in Sections 3.4 and 3.5 of this document
are discussed.
-54-
-------
A.2.0 BACKGROUND INFORMATION
The following information relative to the operation of the sub-
ject Portland cement manufacturing plant would ordinarily be gathered
from interviews and observations during a visit to the plant for a pre-
test survey.
Portland cement is made from a mixture of finely ground calcareous
(lime component) and argillaceous (alumina component) materials. The
four major steps for producing Portland cement are:
(1) Obtaining raw materials and reducing their size,
(2) Grinding, blending and homogenization of these materials
to obtain desired composition and uniformity,
(3) Heating to liberate carbon dioxide and burning to form
clinker,
(4) Grinding or fine pulverization of the clinker with
addition of gypsum.
At this location, shown in Figure A-l, limestone is quarried at
the site by dragline buckets, pulverized in a hammer mill, mixed with
water and pumped to raw material storage. Other raw materials are de-
livered to storage by rail. Ball mills reduce first the limestone and
then a limestone-clay mixture to a fine slurry which is stored in con-
crete tanks prior to its introduction to the rotary kilns. The slurry
is dried and burned at about 2700°F to form clinker, which is cooled
and stored in bins until needed for the finish grinding operation where
it is pulverized and mixed with gypsum to produce the finished product.
The cement is stored in silos prior to bagging or transfer to bulk con-
tainer trucks and railroad cars for shipping.
-55-
-------
State road
(3 - lane)
III
Private roads
(2 - lane)
l I I l l l I l I l I I »
Railroad spur line
Fig. A-1. Portland cement plant site layout.
-56-
-------
The plant operates on a three-shift, round-the-clock production
schedule including all operations except shipping and unloading of rail-
delivered raw materials, which are normally carried out only on the
8:00 a.m. to 4:00 p.m. shift. The plant produces about 300 barrels of
finished product per hour, consuming about 600 pounds of raw materials
for each 376 pound barrel produced.
The raw materials and the finished product are essentially dust;
the principal emissions are also dust. The largest contributor is the
kiln used to produce the clinker, where the dried mixture becomes sus-
pended in the combustion gases as dust and is delivered through the
stack to the atmosphere. A multi-cyclone/electrostatic precipitator
combination removes about 95 percent of the dust before it is vented to
the stack. Other sources of dust are the ball mills, materials trans-
fer operations and packaging operations. Hoods at the ball mills and
packing house are utilized to capture and transmit about 85 percent of
their emissions to a bag house. The quarry operation at this plant
contributes little or no dust since the entire process is conducted
with the material in a wet condition.
The EPA estimates for uncontrolled emissions, as published in the
Office of Air Programs Publication AP-42, Compilation of Air Pollutant
Factors, are 15 to 55 pounds from the kiln and 2 to 10 pounds from all
other sources for each barrel of cement produced. If the assumed 95
percent effectiveness of the stack controls is correct, 0.75 to 2.75
pounds per barrel could be transmitted to the atmosphere from the stack.
Assuming that 80 percent of all other emissions are hooded, 0.5 to three
pounds per barrel could be transmitted to the atmosphere as fugitive
-57-
-------
emissions.
The prevailing daytime wind at the plant is from a general easter-
ly direction and averages 10 miles per hour over open, flat, partly
swampy terrain.
-58-
-------
A.3.0 METHOD SELECTION
Selecting the most practical method to measure the amount of
emitted pollutants reaching the ambient atmosphere involves evaluating
the site, processes and pollutants concerned in terms of the criteria of
Section 2.2 as follows:
Site Criteria - the various sources at the site are remote from
one another, both indoors and outdoors, and are not small enough
to be hooded or otherwise enclosed.
Process Criteria - emissions are essentially the same from all
sources at the site with no interfering reactions between emis-
sions or with other constituents in the ambient atmosphere.
The process is continuous and does not entail any limitations
as to the timing of sampling.
Pollutant Criteria - emissions to be measured are particulates
whose generation rate and dilution in the ambient air will pro-
vide measurable concentrations within reasonable distances of
the source.
The site criteria are, in this case, the determining factors in
selecting the measurement method. Since the emissions cannot be con-
tained or directed in any manner, only the upwind-downwind measurement
method may be successfully utilized to determine the plant's contribu-
tion to the particulate concentration in the local atmosphere.
The basic question to be answered by the measurement program is
"Does the rate of particulate emissions from the plant exceed the
accepted regulatory agency standard?" This question can be answered
by a survey program average measurement of the total particulate emis-
sions from the plant, including emissions from the kiln stack. If
the survey program indicates that the plant's emissions do exceed stan-
dards, the question to be answered will then be "What actions are neces-
-59-
-------
sary to reduce emissions to an acceptable rate?" The answer to this
question requires that the rates of the specific sources of the emis-
sions be separately quantified. This will require the increased accur-
acy and extent of measurements of a detailed program. The design of
both systems is described in the following sections.
-60-
-------
A.4.0 SURVEY MEASUREMENT SYSTEM
To determine the total plant contribution of particulates to the
atmosphere, measurement must be made of the approaching ambient air
containing upwind and all background emissions. In this case, a single
upwind sampler located between the kiln building and the road to the
east will include the general ambient background particulates plus the
particulates contributed by traffic on the road. A ground level sampler,
located about 200 meters from the kiln, should provide an accurate mea-
surement. The downwind measurement must include the contributions from
all the sources at the site, which may be considered as emanating from
a line source at ground level with an overlay of emission from the ele-
vated stack, as illustrated in Figure A-2. To ensure that the stack
emission contribution to the cloud is being measured, one downwind sam-
pler is located within the estimated confines of the stack plume and
others outside the stack plume as shown on Figure A-2.
A.4.1 Sampler Location
For a high volume sampler sampling 18 cubic feet per minute, a
desired sample weight of 100 micrograms and a 60 minute sampling time,
the particle concentration required at the sampling point is: (per
Equation 3-1)
X = M/FT = lO"4 (gm)/0.5 (m3/min) x 60 (min)
X = 3.3 x 10"6 (gm/m3)
-61-
-------
*D1
*D3
Stack plume
100 meters
Fig. A-2. Portland cement plant emissions clouds.
-62-
-------
Local emission limitations, promulgated on a process weight basis,
permit 30 pounds per hour of particulate to be transmitted to the at-
mosphere from all sources. In order to measure this total emission rate
in a 10 mile per hour (4.47 m/sec) wind with the proposed samplers, the
product of the standard deviations used to determine the maximum dis-
tance from the source that samplers may be located is found using:
K = Q/irxu (Equation 3-2)
K - 30 x 454 x ^L (fS^/u x 3.3 x lO'6 (g) x 4.47 (^) - 8 x 10*
Table 3-3 indicates the use of an atmospheric stability category B
for clear midday conditions and the wind speed of 4.47 meters per sec-
ond. Figure 3-2 indicates a maximum sampler downwind distance well in
excess of one kilometer for K = 8 x 10^ and category B, so that any
sampler location within one kilometer downwind of the plant will provide
satisfactory measurements. To ensure that the stack emissions are also
adequately measured, one sampler is located along the wind direction
axis through the stack at a distance of 800 meters from the stack, at
point Dl on Figure A-2. Two additional samplers are located within the
fugitive emissions cloud outside the stack plume at points D2, 300
meters from the kiln structure on its wind direction centerline, and D3,
500 meters from the kiln at 100 meters to the south (cross-wind) of its
centerline.
Samples are taken simultaneously at the upwind and three downwind
locations for a one hour period chosen to include activities in all
phases of the process- kiln operation, grinding, packaging and all phases
-63-
-------
of material transfer including bulk product loading and raw material
unloading. The samples are analyzed to determine particulate concen-
trations at the sampler locations, which are then used in computer pro-
grammed diffusion equations to determine the source strengths of the
fugitive and stack emissions.
-64-
-------
A.5.0 DETAILED MEASUREMENT SYSTEM
Assuming that the survey measurements indicate an emission rate
in excess of the local regulations, say 40 pounds per hour, a detailed
system must be designed to more accurately quantify the emissions from
the separate sources at the plant.
The separate sources are identified as individual particulate
clouds on Figure A-3. Their characteristics and schedules are as fol-
lows:
(1) Flotation Tanks - continuous low level emissions. Cloud
usually isolated.
(2) Ball Mill and Slurry Tanks - continuous emissions. Cloud
usually mixed with (3).
(3) Raw Materials Storage - continuous low level emissions,
higher emissions during day shift material unloading opera-
tions. Cloud always mixed with (2).
(4) Packing and Shipping - emission level variable with activity
on day shift only. Cloud always partially mixed with (5).
(5) Finish Grinding Mill - continuous emissions. Cloud partially
mixed with (4).
(6) Stack - continuous emissions.
(7) Materials Transfer - continuous low level emissions as back-
ground to all except (1).
Assuming that the prevailing wind direction remains from the east:
Cloud (1) may be individually measured at any time.
Cloud (2) may be individually measured only when material un-
loading operations are shut down - measurement would be improved
by wetting down raw materials.
Cloud (3) may not be individually measured. Fjnissions may be
quantified by measuring total of clouds (2) and (3) and subtracting
individual measurement of (2).
-65-
-------
Flotation
Y"^\ tanks
Finish
grinding
mill
100 meters
Fig. A-3. Portland cement plant separate source clouds.
-66-
-------
Cloud (4) may not be individually measured. Missions may be
quantified by measuring total of clouds (4) and (5) and sub-
tracting individual measurement of (2).
Cloud (5) may be individually measured when packing and shipping
operations are shut down.
Cloud (6) emissions may be measured by stack sampling at any time.
Cloud (7) emissions may not be individually measured. Their low
level background contribution is present in all clouds measured
except (1).
To measure the source strengths associated with clouds (1) through
(5), a network of individual arrays may be set up as follows:
Array [1] - in cloud (1)
Array [2] - in combined clouds (2) and (3)
Array [3] - in combined clouds (5) and (6)
Samples taken during first shift operations using all three arrays
will provide measurments of the particulate concentrations in cloud
(1), the combined concentrations of clouds (2) and (3) and the combined
concentrations of clouds (4) and (5). Samples should be taken during
materials unloading operations to provide measurements of the maximum
concentrations of cloud (3) and during maximum activity level in the
shipping area to provide measurements of the maximum concentrations of
cloud (4).
Samples taken during second or third shift operations using arrays
[2] and [3] will provide measurements of the particulate concentrations
of clouds (2) and (5).
Stack samples may be taken at any convenient time.
Analyses of the samples will provide particulate concentrations at
-67-
-------
the array locations for each source, which may then be back-calculated
to provide equivalent source strengths, which, with appropriate sub-
tractions described above, will give individual source strengths.
The flotation tanks are located very nearly at ground level and
may reasonably be considered a ground level source. Array [1] may
therefore be composed of only ground level samplers located across the
cloud (1) generated by these tanks.
Raw material storage generates a ground level source cloud (3),
while the ball mills and the slurry tanks generate an elevated cloud
(2). The array [2] used to sample these clouds must then employ both
ground level and elevated samplers located across the portion of the
cloud combining the emissions of both sources.
Packing and shipping operations generate a cloud (4) of both ground
level and elevated emissions, as do the finish grinding mill and clinker
storage in cloud (5). Array [3] must then be composed of ground level
and elevated samplers located across the portion of the cloud combining
the emissions of both clouds.
Assumptions as to the approximate source strengths for each of the
sources are made to provide the starting points for determining array
locations and spacing. Based on the 40 pound per hour rate for the
total emissions indicated by the survey measurements, a source strength
of eight pounds per hour is assigned to each of the sources of clouds
(2) through (5), and four pounds per hour for the sources of clouds (1)
and (7). The conditions of the survey example of Section A.4.1; with a
10 mile per hour wind from the east, an atmospheric stability category
B, a desired sample weight of 100 micrograms and a 60 minute sampling
-68-
-------
time are assumed to apply to the detailed system.
The best locations for the arrays, each within the clouds they are
designed to measure and away from the influences of other clouds, are
shown at Al, A2 and A3 on Figure A-3.
For array [1], located about 250 meters downwind of the source of
cloud (1) in order to avoid the influence of neighboring clouds, the
approximate particle concentration on the wind direction axis at ground
level is determined from Equation 3-2, rearranged as
X = Q/irKu, where
X = concentration (gm/m3)
Q = source strength = 4 (Ibs/hr) = 0.5 (gm/sec)
K = 110 (m2) - from Figure 3-7
u = 4.47 (m/sec), and
X = 3.2 x 10~5(gm/m3)
The required sampler flow rate is determined from Equation 3-1,
rearranged as
F = M/xT, where
F = flow rate (m3/min)
M = sample weight = 10"1* (gin)
X = 3.2 x ID"5 (gm/m3)
T = sampling time = 60 (min), and
F = 5.2 x 10-2 (m3/min)
The cross-wind spacing of the samplers in the array is deter-
mined by assigning a particle concentration desired to be measured at
a sampler location of about 1/2 the concentration at the wind direc-
-69-
-------
tion axis, or, x = 1.6 x 105 (gm/m3), and calculating its ratio to the
calculated concentration on the wind direction axis, x • In this case,
X/X = 0.5. Figure 3-4 indicates a value of 0.91 for the cross-wind
distance ratio y/y , in which y is the desired cross-wind sampler dis-
m r
tance and y is the maximum cross-wind distance determined from Figure
3-3. In this case, y for x = 250 meters and category B is 68 meters,
and y = 62 meters.
Array [1], then, would consist of three ground level samplers lo-
cated 250 meters downwind of the flotation tanks with the central sam-
pler on the wind axis and two samplers 62 meters away, one in each of
the two cross-wind directions. This array will provide measurement
of at least two particle concentrations within the cloud for use in
the back calculation of the source strength at the flotation tanks.
Similar computations may be made for each of the other arrays,
with the addition of a vertical spacing determination using Figures
3-5 and 3-6 in the same manner as Figures 3-3 and 3-4 for the deter-
mination of cross-wind spacing.
-70-
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-089a
2.
4. TITLE AND SUBTITLE Technical Manual for Meas
Fugitive Emissions: Upwind/Downwind Sam
for Industrial Emissions
7. AUTHOR(S)
Henry J. Kolmsberg
3. RECIPIENT'S ACCESSION-NO.
urement of •• "EPO.R,TDATE
Dlinc Method Apri1 1976
r~ 0 6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION. REPORT NO.
9. PERFORMING OR3ANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT NO.
TRC--The Research Corporation of New England 1AB015; ROAP 21AUY-095
125 Silas Deane Highway
Wethersfield, Connecticut 06109
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Labora
Research Triangle Park, NC 27711
15. SUPPLEMENTARY NOTES Proiert
Ext 2557.
11. CONTRACT/GRANT NO.
68-02-2110
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; V75-1/76
14. SPONSORING AGENCY CODE
lory
EPA-ORD
officer for this report is R. M. Statnick, Mail Drop 62,
i6. ABSTRACT ^^Q manuai provides Si guide for the implementation of the Upwind/Down-
wind Sampling Strategy in the measurement of fugitive emissions. Criteria for
the selection of the most applicable measurement method and discussions of general
information gathering and planning activities are presented. Upwind/downwind
sampling strategies and equipment are described. The design of the sampling system,
sampling techniques, and data reduction procedures are discussed. Manpower
requirements and time estimates for typical applications of the method are
presented for programs designed for overall and specific emissions measurements.
The application of the outlined procedures to the measurement of fugitive emissions
from a Portland cement manufacturing plant is presented as an appendix.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Industrial Engineering
Measurement
Sampling
Portland Cements
I8. DISTRIBUTION STATEMENT
Unlimited
b.lDENTIF!ERS/OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Fugitive Emissions
Upwind/Downwind Sam-
pling
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI Field/Group
13B
13H
14B
11B
21. NO. O= =AGES
75
22. PRICE
EPA Form 2220-t (9-73)
-71-
------- |