EPA-600/2-76-122
April 1976
Environmental Protection Technology Series
TECHNICAL MANUAL FOR PROCESS SAMPLING
STRATEGIES FOR ORGANIC MATERIALS
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research 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-600/2-76-122
April 1976
TECHNICAL MANUAL
FOR PROCESS SAMPLING STRATEGIES
FOR ORGANIC MATERIALS
by
W. Feairheller, P. J. Marn, D.H. Harris, and D. L. Harris
Monsanto Research Corporation
P. O. Box 8 (Station B)
Dayton, Ohio 45407
Contract No. 68-02-1411, Task 11
ROAPNo. 21ACX-094
Program Element No. 1AB015
EPA Project Officer: L.D. Johnson
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
-------
TABLE OF CONTENTS
Page
I. Introduction 1
II. Preliminary Methods - Level I 9
2.1 Gas Matrix Sampling 10
2.1.1 Proposed Level I Sampling System 12
2.1.1.1 Probe Tips and Probe Systems 16
2.1.1.2 Particulate Collection 17
System
2.1.1.3 Organic Collection System 20
2.1.1.4 Impinger System 26
2.1.1.5 Umbilical Cord, Control Unit 27
and Pump
2.1.1.6 Pressure Drop Evaluation of 27
Level I Sampling System
2.1.2 Gas Matrix Sampling - Fugitive Emis- 31
sions
2.1.2.1 Particulate Emissions 32
2.1.2.2 Vapor Emissions 34
2.2 Liquid Stream Sampling 36
2.2.1 Level I Sampling Schemes 39
2.2.1.1 Grab Sampling Techniques 40
2.2.1.2 Field Concentrated Sample 42
Collection
2.2.2 Level I Liquid Flow Measurements 47
2.3 Solid Sampling Strategies 48
2.3.1 Level I Sampling of Bulk Material 52
ill
-------
TABLE OF CONTENTS - Continued
Page
2.3-2 Level I Sampling of Moving Solid 53
Materials
2.3.3 Material Collected by Control Devices 55
III. Level II and Level III Sampling Strategies 57
3.1 Gas Matrix Sampling Strategies 58
3.1.1 Stationary Source Methods 59
3.1.1.1 Solvent Scrubbing 62
3.1.1.2 Condensation Techniques 63
3.1.1.3 Use of Porous Polymer Adsor- 65
bents in Sampling Gas Streams
3.1.1.4 Adsorption on Chemical Sub- 76
strates and Silica Gel
3.1.1.5 Sampling in Particulate Laden 77
Streams
3.1.1.6 Procedures for Low Molecular 80
Weight Materials
3.1.2 Fugitive Emission Methods 83
3.1.2.1 Particulate Emissions 85
3.1.2.2 Organic Vapor Emissions 87
3.2 Liquid Sampling Schemes 88
3.2.1 Sampling Systems 89
3.2.2 Flow Measurement Techniques 99
3.2.2.1 Procedures for closed, filled 101
systems
3.2.2.2 Flow Measurements in Open 103
Pipes
iv
-------
TABLE OF CONTENTS - Continued
Page
3.2.2.3 Control of Plow Proportional 107
Sampling
3.2.3 Preservation of Samples 113
3.3 Solid Sampling Procedures 116
3-3.1 Sampling of Bulk Material 11?
3.3.2 Sampling from Moving Streams 118
IV. References 121
V. Appendix 127
-------
LIST OF FIGURES
Figure Page
1 Schematic of Level I Gas Matric Sampler 13
2 Separator cyclone 19
3 Prototype high volume porous polymer adsorber 22
4 Possible design for production high volume 23
adsorber
5 Possible design for thermostated high volume 24
adsorber
6 Pressure drop data for various sampling 29
systems
7 Integrated gas sampling train, 60
(Solid lines showing normal arrangement,
Dotted Lines - alternate arrangement,
evacuating chamber around bag.)
8 Sampling train for aldehydes 64
9 Porous polymer vapor sampling method 66
10 Alternate porous polymer system for organic 68
vapors
11 Diagrams of the sampling systems used in the 79
Battelle train evaluation study
12 Porous polymer and thermal gradient sampling 82
train
13 Kaiser tube configuration 84
14 Sampler positions; S0 - background; 86
Sj, S2, S3, S4 - sample collectors
15 Schematic of the CVE Sampler 95
vii
-------
LIST OP FIGURES - Continued
Figure Page
16 Schematic of the Manning S^IOOO Sampler 98
17 Open pipe flow measurement method9 104
18 California pipe flow method9 106
19 Typical V-notch and Rectangular Weirs11 108
20 Design of Parshall Flume 109
viii
-------
LIST OF TABLES
Table Page
1 Components of Level I Sampling System 14
2 Detector-Column Options for the GLC Separation 6l
of Various Classes of Organic Materials27
3 Organic Substances Detected from Paint and 69
Polymer Curing Ovens by Porous Polymer Ad-
sorption and GC Mass Spectrometric Analysis
4 Physical Properties of Porous Polymers 71
5 Volume of Sample Required for Determination of 90
the Various Constituents of Industrial Water10
6 Samplers for Wastewater Collection 9^
1 Methods of Flow Measurement and Their Applica- 100
tion^5
8 Methods of Measurement of Flow in Closed Pipes 102
9 Sample Preservation and Maximum Holding Periods9 115
ix
-------
SECTION I
INTRODUCTION
The purpose of this technical manual is to describe the sam-
pling approaches for conducting an environmental assessment
survey and control technology studies on the feed, product
and waste streams associated with the production of organic
materials. The intent is to provide selected "state-of-the-
art" sampling techniques that can be employed to obtain sam-
ples of materials for both chemical analysis and biological
testing. The methods presented can be used by all groups in-
volved in this type of sampling, although modifications may
be required for particular sites. These methods are presented
as extensions of current art and should be viewed as reason-
able approaches that will be subject to refinement as further
advances are made.
The following description taken in part from the two guide-
line documents on environmental assessment sampling and analy-
sis programs defines a portion of our area of interest.1*2
"An environmental source assessment contains (1) a systematic
evaluation of the physical, chemical, and biological charac-
teristics of the streams associated with a process; (2) pre-
dictions of the probable effects of the streams on the environ-
ment; (3) the prioritization of the streams; and (4) identifi-
cation of any necessary control technology program. An en-
vironmental source assessment program addresses, to the
-------
maximum extent possible, the identification of potential prob-
lems with pollutants for which specific standards have been
set and materials that are suspected to have deleterious
effects on the environment. The ultimate goal of an environ-
mental source assessment, then, is to insure that the waste
streams from a given process scheme are "environmentally
acceptable" or that adequate technology exists for control.
The overall strategy incorporates three "levels" (phases) of
information development. In general, the intent of each
phase is the measurement of the mass flow rates of primary
pollutant classes either out of the envelope containing the
process or out of the plant of which the process is one part.
The strategy which has been developed includes characteriza-
tion of feed streams to provide a rough material balance, and
to determine if an effective control approach may be through
feedstock modification. The characterization also extends
to the product streams whenever they may directly affect the
environment at the next step of usage.
Level I is structured to produce a cost effective information
base for prioritization of streams and for planning of any
subsequent programs. It seeks to provide input data to sup-
port evaluation of the following questions:
a. Do streams leaving the process have a finite
probability of exceeding existing air, water,
or solid waste standards or criteria?
b. Do any of the streams leaving the process con-
tain any classes of substances that are known
or suspected to have adverse environmental
effects?
c. Into what general categories (classes) do these
adverse substances fall?
2
-------
d. What are the most probable sources of these
substances?
e. Based on the adverse effects and mass out-
put rates, what is the priority ranking of
streams?
f. For streams exhibiting potential environ-
mental effects, what is the basic direction
that control strategies are likely to follow?
The Level I measurement program provides information on physi-
cal and chemical composition of process streams. It provides
information on the in vitro cytotoxicity and mutagenicity of
the streams and stream components as well. In the chemical
analysis program information concerning the elemental composi-
tion and the distribution of chemical substances according to
conventionally, but broadly, defined classes is sought.
This program provides data that permit both the identification
of existing broad problems and the evaluation of the possible
adverse environmental effects of the streams. The measurement
techniques employed in Level I do not attempt the quantitative
determination of compliance with existing standards, but in-
stead provide results that can be used both for semiquantita-
tive evaluation of process compliance and for planning subse-
quent sampling and analysis programs.
Such subsequent programs are effected through Level II and
Level III sampling and analytical studies. Briefly, the
Level II measurement program is an extensive qualitative,
semiquantitative approach that seeks to identify specific
substances that exist in any streams having a significant
environmental problem. At the conclusion of Level II, the
physical, chemical, and biological characteristics of the
-------
stream should be well understood in a qualitative, semiquanti-
tative sense under the normal operating conditions of the
process. Level III is a quantitative study of the effect of
process variables on the emission rates of specific substances
that have been identified in Level II as appropriate "indica-
tors" of the environmental impact of streams. Both Level II
and Level III provide the information necessary for the devel-
opment of control technology - an area of equal importance to
the emission assessment of the source.
It is desirable to provide measurement methods that will in-
sure comparability of information from a wide range of proc-
esses. At Levels II and III, much of the work will be very
process- and stream-specific. Present indications are that a
reasonably specific set of procedures can be defined for
Level I studies."
Based on these definitions, the current state of the sampling
art would usually provide sufficient samples for the sensi-
tivity of the available analytical techniques. However, the
additional requirement to provide samples not only for chemi-
cal analysis but also for biological effect studies indicates
that a larger quantity of samples (approximately 500 mg) is
necessary. For health studies, fractionation of particulate
matter in size ranges of 3 to 10 ym and less than 3 ym (based
on the inhalable size ranges of particulates) is desirable and
sufficient samples should be collected in each size range for
chemical analysis and toxic effect evaluation. However, the
collection of these samples in a short time period implies
that assessment programs will involve sampling with higher
volume equipment (mass flow) than the currently popular 1 cfm
commercial sampling trains.
-------
This manual presents approaches that can be used by persons
who are basically knowledgeable in sampling. This means that
experience in the techniques is required.
There will, of course, be occasions in which changes in de-
sign and operating conditions may be indicated and in such
situations experience will be helpful. The methods presented
are advances in the "state-of-the-art", not the ultimate solu-
tion to all situations. These techniques provide a high
probability of success if used with understanding and respon-
sibility.
In an organic sampling effort, problems can be anticipated
due to the labile nature of the compounds. Vapor pressure
effects can be very serious, and as a result, pressure and
temperature control of the sampling system must be considered
carefully. Many organic process streams will contain a wide
range of chemical compounds, and any one stream may contain
all three physical states; solids, liquids, and gaseous. Or-
ganic process reactions seldom give 1005? yield of the desired
product, and the many side reactions that occur can produce a
wide range of chemical compounds.
The adsorption of organic compounds on inorganic substances
is very common, and adsorption-desorption processes can occur
within the sampling system. Reactions can occur during sam-
pling, and as a result, the compounds that are subsequently
identified in analysis may not be the compounds actually
present in the original process stream.
For the most part, moisture sensitive materials will not be
found in exhaust stacks from processes as the materials will
react before they can be sampled. However, if the actual
process streams are sampled, sufficient moisture may not be
-------
present and, if these are to be collected unchanged, precau-
tions must be taken.
As for example, anhydrides and acid chlorides would form acids
upon hydrolysis. In sampling these materials, it is essential
that moisture laden air be excluded from the apparatus.
Compounds such as aldehydes can undergo air oxidization to the
corresponding alcohol. Formaldehyde is difficult to sample
due to polymer formation and the fact that such low molecular
weight species may not be collected in total in the adsorber
approach.
Examples of unstable compounds are peroxides, dienes, and
some polycyclic aromatic compounds. Peroxides can decompose
spontaneously and under the proper conditions with violence.
The conditions within the sampling train could easily pro-
mote this decomposition - the peroxide would be collected and
held on the filter and other surfaces and exposed to the
source gas at elevated temperatures during the run. Dienes
and other similar unsaturated materials can easily decompose
into smaller molecules if the temperature of the train is
high enough.
Many compounds including polycyclic aromatic compounds are
sensitive to UV light and if exposed, can change chemical
composition before analysis. It is essential that all com-
pounds be protected from light during all phases of sampling
and analysis.
There are many classes of compounds that can easily polymer-
ize and these polymers could be formed during sampling and/
or analysis. Examples of such materials are vinyl chloride,
vinylidene chloride, styrene and acrylate esters. If this
-------
polymerization occurred in the sampling system, the material
most likely could not be recovered during the cleanup and
analysis steps.
The comments and problems outlined briefly here are mentioned
to alert those planning to sample organic process streams
that it is extremely important to understand the process being
sampled and to the necessity of being fully aware of what can
happen during a sampling and analysis program. Valid results
can be obtained, but not without thought as to what is in-
volved. In general, this manual is directed to those who are
experienced in emission sampling. However, included in Ap-
pendix I, is a description of some of the considerations re-
quired for a successful sampling program and material that
has been helpful on past projects.
-------
SECTION II
PRELIMINARY METHODS - LEVEL I
The objective of a Level I sampling effort is to obtain large
quantities of sample from a process source in a relatively
short period of time with a minimum of manpower. The sample
that is obtained can be used to determine the classes or
groups of organic compounds that are present and to establish
the in vitro cytotoxicity and mutagenicity of the stream and
stream components. In order to accomplish this objective,
methods are defined in this manual for sampling gaseous, liq-
uid and solid phase systems. The methods are intended to ap-
ply to many different process feedstock, waste and product
streams without major modification. Based on the results of
the Level I preliminary survey sampling and analysis program,
the requirements for a Level II or Level III program can be
defined, and for these studies, the most useful of several
optional techniques can be selected.
During the Level I sample acquisition program, process data
must be acquired in order to relate the analytical data ob-
tained to the emission rate of the processes. At the minimum,
the flow rates and temperatures of each sampled stream must
be established. Beyond this, data on the feed rates of raw
materials and production rates of the products should be ob-
tained from operation logs and if possible by measurement in
order to define emission factors, and to aid in the evalua-
tion of the sampling program with respect to later Level II
9
-------
and Level III efforts. Regardless of the level of the sam-
pling program, the accuracy and precision of the flow rates
and production parameters should be consistent with the ac-
curacy and precision of the sampling technique.
In the following pages of this Section, the sampling methods
are defined for Level I surveys of the organic processes in
which the sample is contained in a gaseous matrix, an aqueous
or non-aqueous liquid matrix, or as a bulk solid. The pro-
cedures described are "state-of-the-art" or logical exten-
sions of the "state-of-the-art" and as a result should be
highly reliable for meeting the Level I sampling objectives.
2.1 GAS MATRIX SAMPLING
The nature of the organic components in a gas stream depends
on the physical properties of both the compounds of interest
and the gas stream. The organic components can exist in the
solid, liquid, or vapor phase at the condition of the gas
stream. As the component is removed from the source, the
sampling system temperature and pressure can change the physi-
cal state of the component. As a result, one is confronted
with the problem of sampling condensed, condensible and vapor
forms with the possibility of phase changes occurring through-
out the system. Existing sampling techniques provide a means
of collecting condensed and condensible species (filter cy-
clone or impinger collection), but the collection of vapor
species presents a particular challenge to the design of ef-
ficient sampling systems.
The detection and quantitative measurement of organic compo-
nents generally requires a concentration step to attain the
required analytical detection limits. This is accomplished
10
-------
with existing systems by sampling for long periods of time and
concentrating the condensed or condensible material for this
time period on the filters or in the impingers of the sampling
system. Some type of concentration step will be required for
those materials which exist in the vapor state. This concen-
tration step should collect the material without chemical
change and in a form that can be conveniently delivered to the
remote analysis location.
The proposed system should collect sufficient material for bi-
ological effect studies (a minimum of 50 mg) in a time period
of three to five hours with a field team of two to three per-
sons. This requirement eliminates low flow rate sampling sys-
tems (0.02 mVm, 0.75 cfm) and suggests that sampling flow
rates of 0.08-0.14 m3/m (3-5 cfm) would be desirable.
In addition to the ability to collect large quantities of mate-
rial, the Level I sampling system for organic materials should
provide the capability of also collecting inorganic materials
for a complete evaluation of the source, and provide for size
fractionation of particulate matter corresponding to respira-
ble and non-respirable size intervals. These provisions add
a great deal of complexity to the sampling system.
Based on this approach, the Level I system ideally should have
a sampling flow rate of 0.1-0.15 m3/m (3-5 cfm), provide size
fractionation, combine organic and inorganic sampling proce-
dures and yet be of suitable size and weight to be handled eas-
ily by several men. A system to accomplish all of these re-
quirements is nearing completion. In order to provide the ra-
tionale for such a system, it is necessary to rely on the cur-
rent state-of-the-art of organic sampling, based primarily on
experience with 0.02 m3/m (0.75 cfm) flow rate systems and de-
velopments in inorganic sampling with 0.14 m3/m (5 cfm) units.
11
-------
Under EPA contract 68-02-1393, a technical manual has been
prepared for the process measurements of trace inorganic ma-
terials.3 The sampling method for Level I described in this
referenced manual is based on a high volume stack sampler4
and some of the design features of that sampler can be applied
to the organic process stream sampler.
In addition, a sampling system has been designed and con-
structed which has three cyclones in series prior to the fil-
ter.5 This study provides the necessary background for the
proposed size fractionation system for organic sampling.
2.1.1 Proposed Level I Sampling System
In order to provide the entire range of capability desired of
a Level I sampling system, it is necessary to combine the
triple cyclone approach for size fractionation of particulate,
the organic adsorber and the inorganic trace element impinger
system. This combined system has not been field tested, but
based on laboratory and field pressure drop studies, it is
reasonable to assume that such a combined system can be made
operational for field use. At present a design and prototype
construction study is underway under an EPA contract.6
A schematic diagram of the proposed Level I sampling system
is shown in Figure 1, and the major component specifications
shown in Table 1. The various components and available op-
tions that require further description are discussed in the
following sections.
12
-------
FILTER
STACK T. C.
GAS CONDITIONER &
'MOISTURE COLLECTOR
<
k__
X
. /
V,
1
1 . 1
HHr ™x
HEATED PROBE
PITOTAP
MAGNEHELIC GAUGES
POROUS POLYMER
ADSORBER
TYPE SPITOTTU&E
1
OVEN T. C. / SAMPLE T. C. 130V TEMP
SENSOR UMBILICAL
AUTO CONTROL VALVE
EM I AUTOMATIC
CONTROL
MODULE
VACUUM GAUGE
GAS METER ING
SYSTEM
10 cfm VACUUM PUMP
Figure 1. Schematic of Level I Gas Matrix Sampler
-------
Table 1. COMPONENTS OP LEVEL I SAMPLING SYSTEM
Component
Subcomponent
Material of Construction
Specifications
Probe
(stack temp, below
1000°P)
3, 5 or 10' lengths
Particulate Removal
System
Organic Adsorber
Impingers
Probe Tips
Probe Liner
Probe Heater
Temperature Monitor
Sheath
Pltot Tube (optional)
Stack Temperature Monitor
Cyclones - 3
Filter Holder
Filter
Oven System
Temperature Monitor
Holder for Adsorber
Adsorber Material
Impinger #1
Impinger *2 and #3
Impingers IU and #5
316 Stainless
316 Stainless
Resistance Heater
Type K Thermocouple
316 Stainless or Inconel
316 Stainless
Type K Thermocouple
316 Stainless
316 Stainless
Quartz (A. D. Little Co.)
Type A Fiber glass (optional)
Teflon (optional)
As required
Type K Thermocouple or bi-
metallic contactor
316 Stainless Steel
TENAX OC (Enka N.V.,
Holland)
Olass or Lexan
Glass or Lexan
Glass or Lexan
Size Range 1/8-3A" Join to Probe liner - Swagelok Connector
3/il" I.D. Suggested, 1/2" I.D. Optional
Capability of heating liner to 500°F
Compatible to Readout Device or Controller
Sized to contain liner, thermocouple, heater, insulation, and
possible stack thermocouple and pltot tube
Type S, with coefficient between .80 and .85
Compatible with Readout Device
Aerotherffl "stub" cyclones - nominal cut points of 10, 3, and
1 yra.
To hold 1*2 am. filters, 1/2 or 3/4" Connectors, no teflon
coating. Stainless support screen, "0" ring seal if not In
contact with gas stream.
Sized to holder, retain particles greater than 0.3
Sized to contain cyclones, and filter, capable of 260°C, insu-
lated electrically and thermally.
Compatible to Readout Device, and/or Temp. Controller,
Size approximately 90 mm I.D., 90 mm bed depth, Stainless
retainer screens, 1/2 or 3/&" Inlet and outlet connections
60-80 mesh size, preconditioned at 200°C for 2>t hours under
He flow.
Multltip tube of Aerotherm design or enlarged impinger of modi-
fied Smith Greenbei'g design - contains 3M H202 or kK NHi,OH
Modified Smith Greenberg of enlarged design, contains 0.2M H202,
0.2M HN03 and 0.02M AgNOj or 0.2M (NH,,)2 S20S and 0.02M AgN03
Contains silica gel if dry test meter is used, these implngers
connected in parallel.
-------
Table 1 (cont'd). COMPONENTS OF LEVEL I SAMPLING SYSTEM
Component
Subcomponent
Material of Construction
Specifications
Meter Section and
Pump
Umbilical Cord
Temperature Display
Temperature Control
Timer
Manometers
Dry Test Meter (optional)
Orifice
Pump
Vacuum Gauge
Valves
Pitot Lines
Sample Lines
Power leads
Thermocouple
Optional
Optional
Stainless Steel
Oilless Carbon Vane Type
Brass, TPE
Polyethylene or Teflon
Polyethylene or Teflon
3 wire grounded
Type K
Digital Thermometer, thermocouple switch, Readout of Stack,
probe, oven, adsorber, exit gas, dry test meter inlet and out-
let temperatures.
The probe and oven can be controlled with temperature controller
{to 260°C). A number of such units are available.
Elapsed time Indicator - record to .01 minutes and up to 999
minutes.
Inclined manometers or magnehelic gauges for readout of pitot
tube iP and orifice meter AH. Read Intervals 0.01 inches of
water up to 1" water and 0.1 inch above 1 inch of water,
Range - 0-6 or 0-10" water, or metric equivalent.
Dry Test Meter (Rockwell) of 7 cfm capacity. This unit may be
eliminated if calibrated orifice meter and differential pres-
sure readout are employed for total volume determination.
Design depends on desired AH reading at various flow rates.
Should provide AH of 6-10" of water at 7.5 cfm. Several
ranges could be provided for different flow rate ranges as
found in Aerotherm unit.
Capable of 10 cfm at 0" Hg Oast 1022, equivalent - requires 3/1)
H.P 1725 RPM motor such as Doerr D272X.
0-760 mm of Hg vacuum.
Shut off valve and control valve. Shut off valve - Whltey B-3F8
or equivalent. Temperature limitations -20 to 70°C. Control
valve - Whitey B-7RF6 (0.312" orifice) needle valve or equiva-
lent.
Umbilical cord should be 25-50' lengths with quick connects on
lines, grounded plugs and sockets on power lines and Type K
thermocouple connectors.
Optional Components - Air Cooled Probes
Water Cooled Probes
Cooling Coil
- Stainless liners with stainless or Inconel Jacket
Stainless liners with stainless or Inconel Jacket
Stainless coil and traps to be employed prior to Tenax adsorber.
Thermostated Adsorber Tenax adsorber enclosed in water cooled Jacket - include ice bath
and circulation pump to maintain temperature of adsorber. Sug-
gested temperature 50°C.
-------
2.1.1.1 Probe Tips and Probe Systems
The probe tips and probe for the high volume sampling system
are essentially identical to those units used in 0.75 cfm sam-
pling systems. All probe tips should be constructed of 316
stainless steel with round cross sections and knife edges.
The probe tip should connect to the probe liner with a Swage-
lok type connector to provide a leakfree seal. For the most
part, the tubing in the sampling system is 1/2" O.D.; there-
fore, this size is acceptable, however, the use of 3/4" O.D.
probe liner and tip connector could be considered to reduced
system pressure drop.
The probe liner should be well insulated not only to retain
heat from the heating element, but also to provide for safe
handling of the probe sheath and eliminate the possibility
of electrical shock. The probe liner, heater, probe thermo-
couple, stack thermocouple, and pitot tube if used, can be
enclosed in a single sheath to provide protection for the
units and to ease the task of closing the space between the
sampling port and the probe.
Ideally, the sampling gas temperature leaving the probe should
be a few degrees above stack temperature to prevent condensa-
tion of water and condensible organic materials. At the min-
imum, the temperature should be above the dew point tempera-
ture of the gas and, if no water is present, it should be
maintained at 50°C (122°P).
For high temperature sources, where glass or quartz are in-
adequate, stainless steel probe liners must be used. For
high temperature sources (above 320°C, or 600°F), air-cooled
or water-cooled probes are necessary. As the temperature of
various sources reach the 500-600°C range, the thermal
16
-------
decomposition of organic molecules occurs and much of the
emitted organic species may be gaseous and of relatively low
molecular weight. Condensation in the probe becomes less of
a problem, but collection of gaseous species must be empha-
sized.
The TRW manual on inorganic sampling suggests that a liner
made from a thermally stable polyimide (e.g., Kapton, Trade-
mark of Du Pont Co.) be placed inside the stainless or glass
probe liner. This liner is not recommended for the combined
organic-inorganic Level I system. If at a later time, it
can be shown that the liner is inert in organic sampling, and
does not have an effect on the results of biological testing,
it may be useful. Stainless steel (3l6), pyrex or quartz
seem the best materials at present for both inorganic and or-
ganic sampling and are also suitable for biological testing.
2.1.1.2 Particulate Collection System
This portion of the sampling train consists of the cyclones,
filter, and oven components. The original cyclones as dis-
cussed in the triple cyclone sampling manual5 have been re-
designed to fit in an enlarged heated oven. These "stub"
cyclones were expected to provide the basic particle size
cuts of the original larger units. Based on the data pre-
sented in the triple cyclone sampling at ambient temperature
(21°C), the D50 values of the cyclones were closer to the
desired 10, 3 and 1 ym cut points at 3 scfm than they were
at 5 scfm.
The "stub" cyclones were evaluated using 12, 10, 8, 5, 3,
1.75, 0.27 and 0.5 pro mono-dispersed aerosol of ammonium
fluorescein.7 At 4 aefrn, the D50 values for the nominal 10,
17
-------
3 and 1 urn cyclones were found to be 12 vm, 5 vm and 0.45 urn,
respectively. Thus, at the original design flow rate of 5
acfm, the D5o values would be approximately 10.5, 4.5 and
0.4 urn. The design of the 3 vim (middle) cyclone, which is
typical of all three units, is shown in Figure 2.
At the present time, the experimental "stub" cyclones are
constructed of 304 stainless steel, and it is understood that
units made from 316 stainless steel will be available within
a few months. Blueprints of the cyclone design for those who
desire to construct their own will be available from the EPA.
Any such units constructed should be made of 316 stainless
steel as this material has been evaluated from a toxicologi-
cal viewpoint whereas the 304 variety has not been studied
as yet.
A filter holder for 142 mm diameter filters would be accepta-
ble for the system, however, it might be preferable to elimi-
nate any coatings such as Teflon, applied to provide a non-
stick surface for particulate. It is possible that abrasion
of this coating surface could occur and thus add to the or-
ganic composition of the collected material. The rate of
abrasion would depend on the nature of the particulate mate-
rial.
Quartz filters would be the most desirable, as these materi-
als have the low metals content needed for inorganic analysis.
Experimental quartz filter that were at one time available
from A. D. Little Co. were excellent, however, commercial
quartz filters, in general, have poor mechanical properties.
Fiber glass filters may be useful, although it has been found
that some commercial lots have varied metals content. This
variation is apparently not a problem in the spectrograde
fiber glass filters (Gelman Instrument Co.), however, we do
18
-------
Figure 2. Separator cyclone
19
-------
not know how this material effects the biological testing re-
sults. Teflon filters are also a possibility, but the in-
creased pressure drop experienced with this filter material
may create problems with the completely assembled sampling
train.
2.1.1.3 Organic Collection System
The main component of the organic collection section of the
train is the Tenax GC adsorber. Several options must be con-
sidered here, based on the temperature and moisture content
of the effluent gases. For relatively low temperature stacks,
approximately ambient to 50°C (122°F), it is not necessary to
provide a gas cooler, and a very simple adsorber trap can be
used. For mid-range temperatures, 50°C (122°F) to about 150°C
(300°F), a water-cooled adsorber is suggested. .For higher
temperatures, a cooling coil-water condenser followed by a
thermostated adsorber is required. At this point these tem-
peratures are somewhat arbitrary and may be modified with ex-
perience. In each case, it is assumed that probe and oven
temperatures are above stack temperature, or at least high
enough to prevent condensation in the probes and cyclones or
on the filter.
The cooling coil can be a coil of stainless steel (316) with
a water collection bottle, arranged so that there is minimal
contact between the liquid and the cooled gas. It is sug-
gested that the coil be made of 3/4" I.D. stainless tubing.
The water collector should be designed to cause a minimal
pressure drop. The Aerotherm gas precooling coil (HVSS-470)
is not recommended in the complete sampling system as it has
been reported to have a pressure drop of 100 to 152 mm. of
mercury.
20
-------
It would appear that the major contribution to the pressure
drop is from the cyclone-type water collector, however, the
1/2" I.D. tubing may also be a factor. By employing 3/4" I.D.
tubing and a Tee connection at the lower end connected to a
receiver, the pressure drop could be reduced somewhat.
Aerotherm has recently designed a multi-dip tube type of im-
pinger with a water-cooled inner jacket, that may be very
useful as a precooler assembly between the heated filter and
the organic adsorber. This device will add complexity to the
sampling system in that it requires a pump and cooling water
reservoir. The multi-dip tube design will aid considerably
in the reduction of system pressure drop, but the fact that
the device is glass and the sampled gases come in contact
with the condensed liquid, may create problems.
The organic adsorber for the system is a scaled up version of
the unit used successfully in 0.02 m3/m sampling systems. The
adsorber used in lower flow rate systems had a bed depth of
^90 mm with a 30 mm I.D., containing about 5 grams of Tenax GC,
In order to operate at higher flow rates, the area of the ad-
sorber is increased corresponding to the increase in flow
rate. Thus, to operate at 0.1-0.15 m3/m, the I.D. should be
about 90 mm, and with the 90 mm bed depth, approximately 40
grams of Tenax GC is required. A prototype of the adsorber
constructed of acrylic plastic as shown in Figure 3 has been
used to evaluate bed settling and pressure drop effects. For
routine use, 316 stainless steel is the material of choice.
A proposed design is shown in Figure 4. This unit is suita-
ble for low temperature sources. For mid-temperature opera-
tion, a water jacket is required to control the temperature
at 50-60°C. A possible design is shown in Figure 5.
21
-------
Figure 3.
Prototype high volume porous
polymer adsorber
22
-------
90 MM
-90 MM
Figure 4. Possible design for production
high volume adsorber
23
-------
OUTLET
THERMOSTATIC
CONTROLLED
BATH
90mm
-90mm-
INLET
Figure 5. Possible design for thermostated high volume
adsorber
-------
The Tenax GC used in the organic adsorber is a material typ-
ically used as a gas chromatographic substrate. In the sam-
pling train application, the material is maintained at a
fairly low temperature in an effort to enhance the adsorption
characteristics of the material. In the chromatographic ap-
plication, the material is packed into a column. The packed
column is operated at higher temperature and with helium flow,
the Tenax acts as a selected adsorption-desorption material
to separate various organic compounds. This same type of ad-
sorption-desorption phenomenon could occur in the sampling
train if the temperature of the Tenax was not maintained at
50-60°C. Further desorption could occur quite readily if the
sampling system were operated under high vacuum conditions.
The organic absorption characteristics of the Tenax adsorber
is, therefore, dependent on maintaining the suggested temper-
ature and system pressures between 0 and 380 mm. of mercury
vacuum, and even at this condition some volatile materials
will be desorbed from the material. For this reason, it is
necessary to employ a supplemental method, which will be dis-
cussed in a later section, for the detection of low molecular
weight materials.
The current practice on organic sampling is to extract Tenax
with pentane in a Soxhlet extractor to remove the organic ma-
terial. With proper extraction and conditioning the Tenax can
be reused. In the combined organic-inorganic sampling system,
the porous polymer most likely would trap some of the As, Sb,
Se, Cd, Hg, Cl, F, and Pb compounds that would pass through
the filter and would be collected in the impingers of an in-
organic-only system. This not only means that an additional
component must be analyzed, but also that there is a reasonably
good chance that the Tenax would be destroyed during the analy-
sis. Even if this were necessary, for a Level I evaluation,
-------
the cost of the Tenax is rather small compared to the cost of
sampling and analysis when one considers the information gath-
ered.
2.1.1.4 Impinger System
The impinger section of the sampling train is designed for
the collection of inorganic components that can pass through
the cyclones, filters and Tenax adsorber. Reference should
be made to the inorganic sampling and analysis manual for
specific details.3 However, in order to design a Level I
system for both organics and inorganics some modification of
the system is necessary. As stated in the last section, some
inorganic components will most likely be found in the adsorber.
Although this will complicate the analysis procedures, it will
permit a change in the impinger system to reduce the pressure
drop problems.
All impingers used in the system should be of the straight
tube design rather than of the Smith-Greenberg type. As a
result, the pressure drop in each impinger will be dependent
on the liquid level above the inlet tube and will be in the
order of inches of water. The nozzle approach of the Smith-
Greenberg type impinger is restrictive and adds considerably
to the pressure drop.
As will be shown in Section 2.1.1.6, where pressure drop informa-
tion is discussed, the silica gel impinger, required if a dry
test meter is used, adds considerable pressure drop to the
system. Two silica gel impingers in parallel can reduce the
pressure drop by splitting the flow between the two units.
This can be done by providing a Tee in the line ahead of the
impingers and then combining the flow after the impingers and
before the umbilical cord to the control unit.
26
-------
2.1.1.5 Umbilical Cord., Control Unit and Pump
The umbilical cord, control unit, and pump from the Aerotherm
high volume sampler can be used, or similar units can be con-
structed. For Level I sampling, a simplified control unit
can be built that will provide the minimum requirements. In
brief, this can consist of an inclined manometer to monitor
the pressure differential across the final cyclone, digital
readout thermometer and thermocouple switch, proportional
types of probe and oven controllers, and a timer. A dry test
meter can be eliminated if a calibrated orifice meter and
elapsed time are used/to provide volume data. A manometer or
magnihelic gauge for pitot tube pressure differential is not
required. For the most part, sampling would not be done iso-
kinetically, but rather at a specific rate to maintain the
desired flow rate and hence particulate size cut points
through the cyclones. If isokinetic sampling were attempted
in a system with a wide variation of flow rate profile, the
cut points of each cyclone would be varied each time the sam-
pling rate was adjusted. A simplified control unit, there-
fore, would provide the essentials for Level I sampling at a
saving of weight and cost, but it could not be used for com-
pliance or similar testing studies.
2.1.1.6 Pressure Drop Evaluation of Level I Sampling System
During the process of developing the Level I sampling system
concept, a major concern was the pressure drop of the various
components and complete system. Since the proposed system
has not been field tested, there is no assurance that the
system will perform as expected. However, Aerotherm has been
assembling a sampling train for use on a coal gasification
system which included a water-cooled probe, triple cyclones,
27
-------
filter, and an inorganic-type irapinger system. This provided
the opportunity to evaluate the pressure drop of this system
and, in addition, to add a proposed prototype of the Tenax GC
adsorber and measure its contribution to the pressure drop of
the system. As a result of this study, it is felt that the
system will operate as expected although it could potentially
cause some problems in very dirty sampling situations.
The results of the pressure drop studies on both the high vol-
ume system and the 0.02 m3/m system are shown in Figure 6. The
data plotted in this figure represent the pressure drop in mm
of Hg of various components on the complete system against the
flow rate in scfm. Curves D and E represent the complete 0,02
mVm system with and without the adsorber but with Type A fil-
ters and charged impingers. With the adsorber in place, flow
rates in the 0.008-0.017 m3/m range are possible, and at these
flow rates the initial pressure drops are low enough to permit
collection of particulate before exceeding the leak check
pressure drop of 15" or 380 mm of Hg vacuum.
Curves A, B and C taken from the data in the TRW report on the
Series Cyclone Sampling Train show the pressure drop of this
system.5 Curve B provides the data on the standard Aerotherm
impinger system, charged with liquid as required in the in-
organic system. These impingers would create too great of a
pressure drop for reasonable operation. Curve C represents
data on the cyclone system fabricated by TRW, while Curve A
is for the complete TRW-cooler system. Because of the high
pressure drop of the impingers, TRW employed the cooler system
and substantially reduced the pressure drop across the entire
train. Based on these data the system could operate in the
0.04 to 0.14 mVm (1.5 to 5 cfm) region.
28
-------
600
500
400
300
200
•s 100
°- 90
< 80
70
60
50
40
30
20
' LEAK CHECK PRESSURE
10
.1 .1.5 .2.2.5.3 .4 .5.6.7.8.91.0 1.5 2 2.5 3 456 78910
FLOW RATE (SCFM)
A. TRW TRAIN WITH COOLER F. INORGANIC SYSTEM-AEROTHERM
B. AEROTHERM IMPINGER ONLY ( STUB CYCLONES, LARGE IMPINGERS )
C. TRW TRIPLE CYCLONES ONLY G. INORGANIC SYSTEM WITH ORGANIC
D. 0.75 CFM SYSTEM (NO ADSORBER) ADSORBER.
E. 0.75 CFM SYSTEM WITH ADSORBER H. AEROTHERM -TRIPLE CYCLONES AND GLASS
GLASSWARE - EMPTY, NO FILTER
Figure 6. Pressure drop data for various sampling systems
29
-------
The Aerotherm system designed for the study of particulate
emission from the coal gasification facility was employed to
determine pressure drop data as a complete system. Various
portions of this system were evaluated. The results of sev-
eral of the measurements are shown in Curves F, G and H.
Curve P represents the pressure drop of the complete Aero-
therm system, with a 1/2" probe tip, the triple stub Type A
fiber glass filter, and the charged enlarged impinger systems.
Curve G represents the same system with the prototype Tenax
GC adsorber (enlarged unit) in place. These data indicate
that a flow rate of 0.085 m3/m (3 cfm) is feasible, but 0,11
m3/m (4 cfm) would exceed the nominal leak check vacuum.
The data show that a major source of pressure drop across the
sampling train is the impinger system. A straight tube im-
pinger (modified Smith-Greenberg) containing water can show a
pressure drop equivalent to the height of the water above the
dip tube and thus would cause a small pressure differential.
At a flow rate of 0.031* m3/m (1.2 cfm) the pressure drop
would be in the order of 7 mm of Hg for a 10 cm water height.
The 0.02 mVm Smith-Greenberg impinger without liquid indi-
cated a 25 mm Hg pressure differential. When filled with
liquid the differential was 33 mm Hg. Thus, if at all possi-
ble, straight tube impingers are desirable from a pressure
viewpoint. Normal impinger operation employs a final silica
gel impinger to reduce the moisture content ahead of the dry
test meter. If one silica gel impinger is used, this adds
109 mm Hg to the pressure drop. However, by using two silica
gel impingers in parallel, this can be reduced to 4l mm Hg.
Thus, for the combined cyclone-organic-inorganic system, it
is suggested that all impingers be of the straight-tube der-
sign and the silica gel, if used, should be in two parallel
impingers. The use of straight-tube impingers will reduce
30
-------
the collection efficiency for inorganic materials in the im-
pingers, however, it can be assumed that the inclusion of the
Tenax GC adsorber in the system should trap these materials
with reasonable efficiency, and the impingers would trap the
material not collected in the Tenax.
The data indicate that the proposed Level I sampling system
could be used successfully in the field to collect both inor-
ganic and organic materials and provide for size fractionation
of particulate matter. An 0.085 mVm (3 cfm) flow rate is
quite possible and 0.11 m3/m (4 cfm), feasible. This latter
flow rate will provide approximately the desired size fractiona-
tion of particulate.
2.1.2 Gas Matrix Sampling - Fugitive Emissions
The material that is blown from bulk storage and conveyors
by the wind is of much lower concentration than generally
observed in sampling of stationary or confined sources. Am-
bient air particulate concentrations are sampled by high vol-
ume samplers, and this type of device can be used to collect
fugitive dust emission for organic analysis.
Another source of organic fugitive emissions is from various
plant operations where volatile organic vapors can be released
into the air. Sources of these vapors can include transfer
operations, leaking valves and fittings, and open areas such
as settling ponds. Such materials will not be collected by
high volume samplers and are generally present at such low
concentrations that sampling is quite difficult.
31
-------
2.1.2.1 Particulate Emissions
The accepted procedure for either ambient air particulate or
fugitive dust sampling is to establish a network of high vol-
ume air samplers around the parameter of the area, which, for
fugitive emissions, would be the plant property lines. With
the network of samplers, the dust can be collected by one or
more of the samplers, regardless of changes in the wind di-
rection and velocity. By relating the wind direction to the
sampler location, it can be determined which samplers received
emissions from the source and which were receiving "background"
dust.
A much simpler system can be used for a Level I sampling pro-
gram, provided several assumptions are made regarding the
meteorological conditions during sampling. In general, if
there is no frontal activity in the area of the plant to be
sampled, fairly stable weather conditions can be found in the
periods from 9-11 am and 1-3 pm. Before 9 am, there is often
an inversion; between 11 am and 1 pm, the winds may shift;
and between 3 pm and dark, another inversion may take place.
For the most part, winds are typically in the 3-15 mph range
(90% of the time) which is quite suitable for sampling. Still
air (0-3 mph) is typically encountered 5% of the time and is
too slow for sampling. Winds higher than 15 mph occur about
IQ% of the time and also present sampling problems. Winds
generally increase during the day and are highest in the af-
ternoon, falling again near dark.
The fugitive emission method for Level I sampling consists of
two or three high volume sampling units, one for the background
sample and one or two for the collection of the emissions.
Sampling should be done only during stable conditions (9-11 am
and 1-3 pm). The background sampler is located upwind from
32
-------
the source and the emission samplers downwind from the source.
Sampling duration should be as long as possible and is depen-
dent on the concentration of dust, wind velocity, type of ma-
terial, etc. If a major wind shift is noted during the inter-
val, the emission sampler should be moved to maintain the
downwind location with respect to the source.
A major problem with this procedure is that it is unlikely
that sufficient sample will be collected for biological test-
ing, however, the filter can be analyzed chemically and a
portion of the material generating the dust source can be col-
lected for both analytical and biological programs.
The type of material collected by this procedure is expected
to have a high concentration of inorganic material and a low
concentration of organic matter. Volatile organic material
will be lost due to the high flow of air across the collected
particulate. High molecular weight organic material (such as
polycyclic organic materials) would be collected by the fil-
ter and the detection of these materials would be an impor-
tant part of a Level I effort.
A method for measuring the wind velocity and direction will
be required for this method. These data, along with the flow
rates of the samplers should be recorded routinely during the
sampling period. Simple meteorology stations are available
from a number of commercial sources or such devices can be
constructed that will fit the requirements of the program.
In addition to the ambient air type of fugitive emission dis-
cussed above, TRC (The Research Corporation) under EPA Con-
tract 68-02-1815» in their technical manual on fugitive emis-
sions have identified two other types of fugitive sources -
those from roof monitors, and those from quasi-stack sources -
33
-------
which differ mainly in the methods required to obtain samples.
Emission from any part of a building, such as windows, doors,
open roof sections, vents, fans, etc. into the air is con-
sidered in the roof monitor sampling approach. The quasi-
stack sampling approach can be employed at any location where
a stack-like structure can be built to contain the source.
The roof monitor approach for a Level I study employs high
volume particulate sampling systems located within the build-
ing adjacent to the openings so as to collect the emission
just prior to its entrance into the atmosphere. In addition
to the sampler, a flow measurement device (such as a hot wire
or rotating vane anemometer) is required at each opening sam-
pled to determine the gas velocity exiting through the open-
ing. The sampler is positioned within the emission "cloud"
and the sample collected. The flow device, positioned so it
is not influenced by the sampler, and the area of the opening
provide a measurement of the emission rate.
The quasi-stack approach is suitable for sources that can be
isolated by use of temporary ducting or enclosures. Such
sources could include material transfer operations, leaking
system components, and open evaporation or fabrication areas.
This method generally requires the enclosure, a fan and an
exhaust stack or duct. Sampling can be done with any of the
typical stack type gas matrix methods and flow measurement
devices. For our purpose, the high volume sampling train de-
scribed earlier should be employed to collect both particu-
late and gaseous organic materials.
2.1.2.2 Vapor Emissions
The organic vapor emission in the vicinity of a process, even
though readily detectable by odor, is generally quite low in
34
-------
concentration. The ambient air methods for detecting such
fugitive emissions are not suitable for a Level I program
where both chemical analysis and biological testing is to be
done.
The procedures typically employed for ambient air hydrocarbon
determination can be used (1) to obtain a total hydrocarbon-
less-methane value, and (2) if concentrations are high enough,
to provide levels of individual components. These procedures
employ a Plame lonlzation Detector (FID) chromatograph. This
detector combination is very sensitive to hydrocarbons but
the materials are destroyed in the flame during analysis and
as a result are not retained for other analytical or biologi-
cal procedures. As this method is more applicable to Level
II or III procedures, it will be discussed in more detail
later in this manual.
The other possible procedures for organic vapor analysis in-
clude the use of charcoal tubes and porous polymer collection.
At present, the charcoal tube method is being used extensively
for NIOSH sampling programs even though there are some defi-
nite problems with the removal of vapors from the charcoal.
The NIOSH sampling procedures use a small charcoal tube and
a battery powered air pump to draw ambient air through the
sample tube. This procedure is generally used for quantita-
tive analysis of specific materials using sensitive instru-
mental analysis methods. The method is not employed for gen-
eral identification schemes as required by Level I studies.
For the Level I program, the procedure suggested for ambient
vapors is an extension of the Level I stationary source sam-
pling system. In the Level I stationary source system there
is a filter and organic vapor adsorber that can be used in-
tact for ambient air samples. By removing the probe and
35
-------
cyclone assembly (although this could be left intact for size
fractionation of air-borne particulate) the sampling train can
be operated at 0.11 ra3/m near the fugitive emission source
to collect the organic vapor materials. This flow rate is an
order of magnitude lower than the typical flow rate of high
volume ambient air samplers and, therefore, longer sampling
periods will be required to provide the required sample. The
benefits of this approach are that the sampler is available
already for the Level I program and that the methods for han-
dling the sample and completing the analysis are established.
Scaling up the organic vapor adsorber to a larger size does
not appear practical as long as Tenax or a similar porous
polymer is used. This method, at present, is only a concept
and has not been tested. There is a reasonable chance that
it can provide a sample, at the very minimum for chemical anal-
ysis, and hopefully for biological testing as well.
2.2 LIQUID STREAM SAMPLING
The sampling strategies required for the determination of the
organic content of liquid streams appear initially to be rath-
er simple, however, there are a number of problems that must
be understood if a truly representative sample is to be ob-
tained. As in gas matrix sampling, a sample is extracted from
a flowing system, and the determination of an emission rate
requires measurement of the mass flow rate of the liquid ma-
trix. In contrast to gas matrix sampling, the liquid may
fill only a portion of the duct or pipe. The choice of sam-
pling locations in a liquid sampling situation can be depend-
ent on existing facilities such as manholes or sewer lines,
although these locations may not be ideal. Again, in contrast
to gas matrix streams, the optimum location for liquid sam-
pling is at a point of turbulent flow. At such locations,
36
-------
suspended or floating material should be more uniformly mixed
and less stratified. Liquid streams at plant locations can
consist of either water-based (aqueous) or organic-based (non-
aqueous) streams. Nonaqueous systems such as in in-plant
ducts and transfer lines are often single phase systems and
thus will be the easiest to sample. The organic content of
aqueous streams can consist of dissolved, suspended, settled,
or floating materials. This type of situation makes the col-
lection of a representative sample a more difficult task.
In this manual, we will restrict our discussion to the sam-
pling of well-defined liquid streams as contained in pipes,
ducts or similar systems. No discussion is presented dealing
with sampling large bodies of water such as lakes or rivers
that would receive outflow from plant sites. This latter
type of sampling is much too broad in scope and varied in de-
tail to be covered adequately here.
A Level I program should provide large quantities of material
in a relatively short period of time for both chemical and
biological testing. The time period must be determined, how-
ever, by an analysis of the plant operation and how this af-
fects the acquisition of a representative sample. If the
operation is relatively uniform, then a suitable sample can
be collected at any time, but if the operation is varied,
sampling over a longer period (eg., 24 hours) may be required
in order to obtain an "average" sample. Sample preservation
is an important consideration. Different materials will re-
quire different treatment. This is especially true if both
inorganic and organic constituents are to be determined. For
Level I, one general preservation procedure is desirable.
The detection and quantitative measurement of organic mate-
rials in a water stream can require a field-implemented
37
-------
preconcentration procedure. This is especially true for
aqueous systems where the concentration levels are expected
to be low since shipment of large quantities of water is not
practical. This type of approach can be employed for dis-
solved and suspended materials providing that a suitable ad-
sorbing medium and removal solvent is used that does not in-
terfere with the biological and chemical analysis.
Organic matter in aqueous streams can consist of,both soluble
and insoluble (floating or particulate) compounds. Typically,
soluble compounds fall into three categories. The first is
materials which are energy sources for microbiota and are
considered biodegradable. The second type includes those
materials which are toxic to aquatic biota (and humans) and
are of major interest in the sampling strategies. The third
type are those materials which are neither toxic nor utilized
by aquatic biota. The insoluble matter in aqueous streams
includes inorganic materials such as process materials, sand,
silt or clay, and organic materials such as process material
and living or previously living matter. Such materials are
in the form of a suspension in turbulent flowing streams, but
tend to settle in slow moving streams or pools, and the re-
sulting two-phase system complicates the sampling program.
As part of the Level I sampling scheme, it must be assumed
that little information is available on the concentration
levels of the organics materials and a strategy is used which
will provide suitable samples over a wide concentration range.
In addition, the potentially low concentration of organic ma-
terial in aqueous streams requires that extreme care be exer-
cised to prevent contamination from any source.
The sample size requirement for typical wastewater analysis
indicates that volumes up to 1 liter usually are sufficient
38
-------
for anlon and cation analysis, physical tests, and dissolved
gases. However, the determination of oily material can re-
quire volumes up to 5 liters just to obtain a weighable quan-
tity of extracted material. This quantity could be suffi-
cient for chemical analysis but not for additional biological
testing.
2.2.1 Level I Sampling Schemes
At the present time, the procedures that have been published
in the Federal Register are concerned with the analytical
methods to be employed in water sampling programs. Such pro-
cedures are given in the October 16, 1973 Federal Register.3
Sampling methods are outlined in various ASTM procedures in-
cluding D510-68 "Standard Methods of. Sampling Industrial Wa-
ter," D1192-70 "Equipment for Sampling Water and Steam,"
D860-J4 (1972) "Sampling Water From Boilers" and D1496-67
(1972) "Sampling Homogeneous Industrial Waste Water." Sever-
al other sources of sampling methods include "Standard Methods
for the Examination of Water and Wastewater"9 and the EPA pub-
lication "Handbook for Monitoring Industrial Wastewater".10
A review article by Rabosky and Koraido that covers both sam-
pling and flow measurement can be found in Chemical Engineer-
ing. 11 All of the above references and most of the informa-
tion that is readily available from other sources are con-
cerned mainly with aqueous streams.
The wide variation in organic component concentrations in
liquid streams suggests that two methods are required for a
Level I Sampling Program. The grab sampling approach will
be appropriate for process streams and effluent streams where
the concentration of the organic material is known or, based
on the presurvey, expected to be high. It is difficult to
39
-------
define exactly what is meant by a high concentration, however,
as a general rule, if the components present can be easily de-
tected by the analytical procedures in 1-5 liters of liquid,
then this procedure should be used. However, if the concen-
tration is expected to be low and trace organic compounds are
of interest, than a concentration procedure is required. Grab
sampling does require the transportation of large quantities
of liquid, maintained at a low temperature (for preservation
of the sample), and this presents some logistics problems.
As the concentration levels are unknown at the beginning of
Level I program, it would be logical to collect both grab and
field-concentrated samples.
2.2.1.1 Grab Sampling Techniques
For a Level I study, a grab-type liquid sample rather than a
composite sample should be suitable. It must be understood
that a grab sample can only represent the conditions present
when the sample was taken, but if the process is relatively
constant, or if a time is chosen that would indicate possible
worse case conditions, the sample would be acceptable. For
a Level I grab sample, mechanical collection would be an un-
necessary complication, and therefore, manual collection of
the grab sample will suffice.
The following list provides some of the more important cri-
teria for collection of grab samples from aqueous streams:
(1) Sample containers should be of Teflon with
Teflon-lined screw caps. Alternatively, if
Teflon is too expensive, glass bottles (wide-
mouth) may be used. Polyethylene or poly-
propylene bottles should not be used for
-------
organic sampling due to possible contamina-
tion. One- to two-liter bottles are con-
venient.
(2) All glass bottles and caps should be cleaned
with chromic acid cleaning solution, concen-
trated HC1, tap water, and 3 rinses of dis-
tilled water. Caps and Teflon bottles should
be cleaned with detergent instead of the
cleaning solution, followed by tap water and
distilled water.
(3) A site for sampling should be located where
turbulent flow exists so that solid matter
is well suspended in the medium.
(4) While one- or two-liter bottles are indicated
for handling convenience, it is suggested
that at least 2 liters and preferably 5 liters
be collected.
(5) Immediately after sampling, bottles should be
cooled to about 4°C, placed in the dark and
shipped to maintain this condition. Analy-
sis should be done as soon as possible, and
the time between sampling and analysis noted.
The mechanics of the grab sample collection will vary with the
physical layout of the site. If the liquid stream is readily
accessible, the sample bottles could be filled with the liq-
uid directly. This type of approach could also be used for
nonaqueous sampling of process streams by employing sampling
taps if such taps are available for plant quality control
analysis. In this case, it must be determined if the temper-
ature and other conditions of the stream are compatible with
the sampling vessel.
-------
If the physical layout of the site is such that direct sam-
pling is not feasible, then other methods of withdrawing the
sample will be required. One such device is a long-handled
stainless steel dipper that can be lowered into the stream to
extract the sample. Another device is the Sirco Uniscoop
Liquid Sampler which is similar to the dipper but has a manu-
ally operated ball valve which can be opened to admit a sam-
ple after the device is in the liquid stream. Other possible
schemes would require a hand- or motor-operated pump to with-
draw a sample. The choice of hand operation or one of the
various motor-driven units would depend on the distance from
the operator to the stream and hence the hydrostatic head of
liquid required to reach the sample container. While a num-
ber of automatic samplers are available, the use of such de-
vices generally would be more appropriate in Level II or III
sampling programs.
2.2.1.2 Field Concentrated Sample Collection
Based on the current state of the sampling art, the Carbon
Adsorption Method (CAM) provides a procedure that can be used
for the identification of classes of organic compounds present
in an aqueous stream. Although this method has been employed
for a number of years, there are a number of problems that
must be considered.
Two carbon adsorption methods have been employed depending on
the nature of the stream to be sampled. The high flow method,
used for clean drinking water systems, collects the organic
components in 19,000 liters of water at a flow rate of 1 1/m
over a 14-day period. This method is not suitable for streams
with a high suspended solids content. The low flow CAM em-
ploys a flow rate of 120 ml/min for typically a one-week
42
-------
sampling period to collect organic components from a total
of 1200 liters of water. Because of the high flow rate, the
collection efficiency of the high flow system is not as good
as in the low flow system. Both methods suffer from the lack
of complete recovery of the organic components by solvent ex-
traction of the carbon.
The carbon from the CAM procedures are extracted with chloro-
form and the carbon chloroform extract (CCE) is evaporated
and weighed to determine the total organic contamination in
the water. In the Dec. 24, 1975 Federal Register12, some of
the problems of the CCE as an indicator of the health effect
of organic chemicals are discussed. First, the CCE accounts
for only a fraction of the organic components. Second, there
is a serious question as to the reliability of the method in
identifying toxic materials. Third, there are no data avail-
able for establishing a specific level for the CCE on a ra-
tional basis. Finally, chloroform is suspected of possessing
carcinogenic properties.
The CAM sample can also be subjected to an ethanol extraction
procedure. This is termed the carbon alcohol extract (CAE).
It would be expected that additional compounds that have lit-
tle or no chloroform solubility could be removed from the car-
bon, with this solvent. Typically, both extracts are reduced
to dryness and the weights used to obtain the organic content.
As a possible variation, both the CAE and CCE could be em-
ployed for chemical analysis without complete solvent evapo-
ration to provide some information on the nature of compounds
present in the water.
The many problems associated with carbon adsorption and sub-
sequent incomplete removal of the adsorbed species has prompted
investigation into other substrates. Polyurethane foam13^
43
-------
XAD-2 resin14, and porous polymers15 have been tried on a
limited basis. For the most part, the applications of these
materials have been restricted to concentration steps of
existing water samples prior to analysis and not to field
use as a sampling substitute for the CAM technique.
The application of porous polymer materials for the collection
of organic species of air samples has developed to the point
where it is used routinely and good separation techniques for
removing the materials from the polymer are available. The
technique of employing porous polymer to remove organic mate-
rials from water samples in the laboratory prior to analysis
appears to be very successful. A very desirable Level I tech-
nique, therefore, would be to develop an on-site sampling de-
vice that employs this approach. At present, this is beyond
the state-of-the-art, however, it is conceivable that the
porous polymer could be used in place of the carbon in a mod^
ified CAM sampling system. The basic system would consist
of a probe to reach into the liquid stream, a porous polymer
filled adsorber and a pumping system. There are a number of
unanswered questions and additional research is required to
determine: (1) the diameter and length of the adsorber,
(2) the flow rate for efficient collection, (3) length of
time required to collect sufficient sample without break-
through of the compounds of interest, (4) the pumping require-
ments of the system, and (5) the collection efficiency of the
porous polymer for the organic compounds of interest in a
water matrix. The existing evidence clearly indicates that
the method has a great deal of promise, but at the present
time the uncertainties tend to preclude the recommendation
of this technique as the prescribed Level I approach. As a
result, it is suggested that the established CAM technique,
with its inherent problems, be employed, but that the
-------
feasibility of the porous polymer method should be investi-
gated and hopefully, in time, it could replace the older pro-
cedure .
By employing the CCE from a CAM sample as a measure of the
approximate organic content, it is possible to calculate the
time requirement for sampling. Typically, clean ground waters
have CCE values of about 25-50 yg/liter, and drinking water
with a CCE value above 200 yg/liter has objectional odor and
taste.16 For nonpolluted water, the CAE value may be less
than 100 yg/liter, and in polluted water the value could be
200-300 yg/liter. For fresh industrial pollution, the CCE
typically exceeds the CAE, however, for other types of water,
the CAE may be several times higher than the CCE. If we as-
sume that the CCE for a typical plant effluent ranges from
100 yg to 1000 yg/liter, at the typical flow rates of 0.25
gal/min (^1.0 liter/minute) approximately 0.7 to 7 days would
be required to adsorb 1.0 gram of chloroform-extractable mate-
rial. Further, if it is assumed that the CAE value is the
same as the CCE value, 0.35 to 3-5 days would be required to
obtain 1.0 gram of total extractable organics. This would
indicate that a high flow technique would definitely be re-
quired for Level I sampling.
The suggested approach for the field concentration sample
collection would, therefore, be to employ the high flow pro-
cedure, using flow rates of 1 liter/min and sampling for
periods of four to twenty-four hours. As this period is
shorter than the typical 14 days and less than 19,000 1. are
sampled, there is less likelihood that clogging of the fil-
ter with suspended material will occur during this shorter
time period. The procedure will depend, however, on much
more complete removal of the adsorbed organic species from
the charcoal in the analytical phase of Level I sampling.
-------
Details of the method can be found in the "Standard Methods
for Water and Waste Water"9 under Procedure 139, "Organic
Contaminants". A review of the technique and some of the
modifications that have been attempted can be found in Chap-
ter 11 of the "Water and Water Pollution Handbook".16 In
brief, the carbon adsorber is a 3" I.D. pyrex tube 18 inches
long with end caps to fit 3/4" stainless steel pipe. The end
caps are also provided with 40 mesh stainless steel screen to
contain the carbon adsorber. Two mesh sizes of carbon, 4 x
10 mesh and 30 mesh, are required for the adsorber. The 4 x
10 mesh is added to the tube to a depth of 4.5 inches and 30
mesh to a depth of 9 inches. The remainder of the tube is
filled with 4 x 10 mesh. The carbon adsorber is compacted by
gentle tapping. All pipe joints in the connections to the
source should be made with stainless steel pipe (rather than
the galvanized pipe specified in the methods) and Teflon-type
tape, avoiding other types of pipe joint compound. To col-
lect the sample, the water flow is started, slowly at first,
then up to the 1 liter/min flow rate after several minutes.
This flow rate is then continued for the specified time inter-
vals. It would not be essential to monitor the total volume
for a Level I screening study, but if possible a meter would
be useful in determining the concentration of organic mate-
rials. After the specified time interval, the carbon adsorb-
ing unit should be sealed with stainless end caps and the
unit held at about 4°C for shipment to the laboratory. Prom
this point on, the procedure for analysis would be similar to
the extraction procedures employed for the porous polymer
from air sampling, presented in the Analytical Methods manual
prepared by Battelle Laboratories17 rather than by the pro-
cedures for the CCE and CAE in "Standard Methods"9.
46
-------
The carbon adsorber and other sampling components can be
easily constructed, however, they are also available from
commercial sources such as General Metal Works, Cleves, Ohio.
The affinity of carbon for organic materials requires that a
great deal of care be exercised in the use of the CAM tech-
nique to prevent contamination of the samples. The carbon
must be of high purity and must not be exposed to any organic
vapor during manufacture, shipping, or laboratory handling.
Ideally, the laboratory where the adsorbers are handled should
have ventilating systems that are isolated from any organic
sources and all other laboratories. It is essential that car-
bon blank samples be carried through any operations and analy-
sis in an identical fashion as the samples.
2.2.2 Level I Liquid Flow Measurements
Accurate measurements of the flow rate of streams in pipes
and ducts are rather complicated and generally not required
for a Level I study. As quantitative data are limited in a
Level I study, only approximate emission rates would be pos-
sible even if reliable flow data were obtained. It is sug-
gested, therefore, that an approximate flow be determined if
this can be done in a simple manner.
Since the requirement for the examination of waste streams
by the plant personnel for internal or state agency use is be-
coming more common, it is likely that flow measurement devices
will be found at many plant sites. Whenever possible, exist-
ing facilities should be used.
In the event that flow measurements must be made on a Level I
study, methods should be chosen that are low in cost and that
will not require the installation of complex hardware.
47
-------
Simple procedures that can be employed include: (1) reading
of inlet water flow meters, if such a meter is directly re-
lated to the effluent stream of interest, (2) capacity of
pumps in the system, (3) a bucket and stopwatch, to time the
interval required to fill a bucket of known capacity. (4) the
use of floating objects or dyes, by timing the interval re-
quired for the object or dye to travel the known distance
from one point to another. The accuracy of these methods
varies from fair to good and no one procedure is appropriate
to all systems. Other procedures that are described later
under the Level II and III approaches can also be employed.
2.3 SOLID SAMPLING STRATEGIES
Defining a Level I sampling scheme for solid materials is not
a simple task. The sampling of bulk materials is concerned
with (1) the raw materials and fuels that are used in the
process, (2) the materials that are transferred or produced
in the process, and (3) material that may be collected as
solids in control devices such as baghouses, electrostatic
precipitators and cyclones. Each of these situations pre-
sents a different problem in the acquisition of a representa-
tive sample.
Raw materials and product solids can be found in large piles,
as in open storage, or contained in hoppers, silos, or rail
and highway vehicles, or in flowing streams including pneu-
matic transfer lines, process conveyor systems and bucket
type elevators. The choice of the sampling approach is de-
pendent on many factors such as accessibility, the physical
layout of the system, the physical properties of the stream
and the degree of accuracy and precision required in the sam-
pling, flow estimation, and subsequent analysis. Fortunately
48
-------
for the purposes of Level I programs, many facilities main-
tain some type of quality control program on both raw mate-
rials and products and thus suitable Level I sampling sites
may be available.
In general, the collection of material from control devices
is often ignored, unless a material balance type of program
is being conducted, or the material represents a useful prod-
uct. Most often the material is simply removed and disposed
of, either to another plant or as waste. As a result, the
material in a baghouse or ESP or from a cyclone may not be
accessible. Special arrangements with plant personnel must
be made to obtain such samples.
The basic procedures for sampling solid materials can be ob-
tained from various ASTM procedures. These techniques are
usually specific for certain types of non-organic materials,
however, the procedures can be applied to almost any type of
bulk or solid sampling situation. "Standard Method of Sam-
pling Coke for Analysis" ASTM D346-35,18 while designed, for
gross samples, provides the techniques for the alternate
shovel and coning and quartering methods and defines the loca-
tion of sampling points from the exposed surface of transpor-
tation vehicles.
ASTM D1799-65 "Sampling Packaged Shipments of Carbon Black,"19
provides a procedure for sampling bagged or cartoned materials,
Method C183-71 "Sampling Hydraulic Cement,"20 provides a meth-
od for obtaining samples from bulk storage at a point of dis-
charge and also provides the design of a slotted tube sampler
for hopper sampling. A similar slotted tube sampler is de-
scribed in D2617-72 "Sampling Particulate Ion-Exchange Mate-
rials".21 Several mechanical sample dividers and riffles
-------
which can be used to collect or reduce the size of a sample
are shown in D2013-72 "Preparing Coal Samples for Analysis".22
Three other ASTM reports of interest in sampling are E105-58
"Probability Sampling of Materials"23,*El4l-69 "Acceptance of
Evidence Based on the Results of Probability Sampling"24, and
E122-72 "Choice of Sample Size to Estimate the Average Quality
of a Lot or Process"25.
The methods required for sampling of solid organic materials
can be, for the most part, adaptations of these procedures.
The equipment and methods, however, must be tailored to the
vessel or site to be sampled as well as the material.
The problems associated with this type of sampling is more
dependent on the physical nature of the material rather than
the chemical composition. Materials in powder forms exhibit
many unique properties. If such powder is compacted by vi-
bration and pressure, it can exhibit some of the character-
istics of a solid body. If the powder contains well-rounded
particles that are graded in size, it can flow like a liquid.
If the powder is very fine, it can aerosolize to form a gas-
like system. Solid material in a moving system tends to seg-
regate by particle size. Thus the job of obtaining a repre-
sentative sample is complicated. Segregation will occur when-
ever one part of the powder moves relative to other parts.
The large particles generally will be found throughout the
samples, but the finely divided material will tend to be col-
lected at the bottom of the conveyor belt or other moving
stream. In any sampling situation where the material is
moved, the sample must include the full cross-section of the
material in order to obtain both coarse and fine material.
50
-------
When solid material is poured into a heap, the larger parti-
cles tend to be found at the rim of the heap. The center of
the heap may be completely devoid of large particles. This
effect can also occur in a hopper or bin. When the bin is
opened, the initial flow may consist of only fine material.
In practice, this effect is minimized by employing steeply
angled hopper designs.
Another effect that can be observed with powdered materials
is termed "bridging", and this effect is readily noted in
vertical pipes. The particles can "bridge" across an opening
that may be as large as ten times the particle diameter.
When this occurs, flow will cease. This effect is well-known
in plant design engineering. However, it is an effect that
also can occur during sampling if the sampling equipment is
designed with openings that are too small.
Another interesting property of a powder is that the flow
rate of a powder through an orifice is independent of the
head of solid above the orifice. This is in contrast to
liquids where the flow rate is directly dependent on the hy-
drostatic head.
Fugitive emissions can be encountered when solid materials
are stored in the open or during transfer of solid materials
from one vessel to another in such a manner that air currents
blow across the material. This situation has been discussed
under the gas matrix sampling strategies.
The concentration of the organic components in a solid sample
can vary from trace concentrations to 100$. As a result, the
sampling procedures must include a consideration of the con-
centration levels expected, the sensitivity of the analytical
procedures, and mass requirements for biological testing
51
-------
programs. This complicates the specification of a sampling
method, and some judgment will be required on the part of the
team members to collect sufficient sample for all required
procedures.
The wide variation in possible sources to be sampled requires
that a description of many possible methods be considered and
that the choice of which method is to be used depends on the
situation. In the Level I approach, these samples are mainly
of the "grab" type. As a result, the restrictions inherent
to this type of sampling must be considered.
2.3-1 Level I Sampling of Bulk Material
For this type of situation, equal portions of material at
random locations should be removed from the pile. These por-
tions can then be composited to obtain a single sample. If
at all possible, each portion should represent the full depth
of the material so that the effect of powder segregation can
be eliminated. Suitable sampling devices include a pipe
borer, slotted tube (thief), or auger, the choice depending
on the physical condition (such as moisture content, particle
size, degree of agglomeration, and hardness) of the material.
If the resulting composited sample is larger than necessary,
the sample for analysis can be reduced in quantity by the al-
ternate shovel, and coning and quartering methods. The
sample should be packaged in clean glass or polypropylene
bottles for shipment. For this type of sample, plastic bot-
tles should be satisfactory unless there is some reason to
believe that the material can be contaminated by plasticizer
or other components present in the plastic.
-------
The slotted tube type sampler can also be used to obtain sam-
ples from packaged materials. In this case the sampler should
be inserted diagonally through the bag of material. A similar
approach is suitable for sampling from bulk shipments con-
tained in cars and trucks. The usual approach is to take
three samples from well-distributed points in the bulk of ma-
terial. Devices for sampling packaged materials and from
bulk shipment vehicles are given in the ASTM procedure for
"Sampling Hydraulic Cement".20
2.3-2 Level I Sampling of Moving Solid Materials
The sampling procedure specified for a moving solid stream
is dependent on the nature of the design of the transport
system. The simplest system is an open conveyor belt where
material such as coal, aggregate, or simpler material is
transported from a pile to another vessel. A critical point
to remember is that segregation is going to occur, therefore,
a sample must include a complete cut from the surface of the
solid through to the conveyor belt. The initial approach is
to employ the "stop belt" procedure, which involves stopping
the conveyor and removing a sample of the full cross-section
of the solid. This method is practical only if it will not
disrupt the production schedule of the plant. The important
criteria for this sample are: (1) the full cross-section cut
with parallel sides is removed from the stream, and (2) the
distance between the parallel faces should not be less than
three times the diameter of the largest piece of material.
If the "stop belt" procedure cannot be used, it is not recom-
mended that a grab sample be removed from a moving belt. Not
only would this method yield an unrepresentative sample due
53
-------
to the segregation effect, but also this method is an unsafe
procedure, and safety must be a prime consideration around
moving machinery.
The best procedure for removing a sample from a moving belt
is moving a cutter device entirely across the stream at a
uniform speed. The speed of this cutter should be designed
to minimize disturbance of the material and thus reduce the
amount of segregation. The major problem with this approach
is that it requires a cutter device that is designed for the
specific velocity of the stream of interest, and the cutter
is often built into the system to be sampled. Thus, this
degree of complexity is not usually appropriate for a Level
I study. If such a device is available for quality control
purposes, it should be used. The criteria for this device
are given in ASTM D223^-72 "Collection of a Gross Sample of
Coal".26
The collection of a sample from the discharge point of a bin
or hopper can be accomplished for a Level I program if access
is available. The procedure involves moving a pan across the
discharge so as to accept the entire cross-section of the ma-
terial, without overflowing the container. It is desirable
to collect three random samples from each hopper, avoiding
the initial discharge. Less segregation will be apparent in
nearly full bins. At times this procedure will require the
construction of the sampling system and in this case is not
appropriate for a Level I program.
There are also a number of reciprocating cutters, rotating
cutters, and riffle type devices that are employed for sam-
pling. These devices must be installed in the system and
thus are not appropriate for Level I sampling programs. If
they are available, however, they should be used.
-------
2.3.3 Material Collected by Control Devices
For certain types of environmental and process studies, the
material collected by the baghouse, ESP units, or cyclones
are of significance in order to completely characterize a
process. The material deposited in the baghouse or ESP can
be sampled by grab techniques during a shutdown. It is sug-
gested that material be collected from a number of areas in
order to obtain an "average" sample. Cyclone devices provide
two sources of solid material, both of which are entrained
in a moving air stream. The fine material emitted from the
outlet can be sampled by typical gas matrix sampling methods,
while the heavier material from the bottom of the unit is
more typically sampled by moving stream solid samples or gas
matrix procedures, depending on the air/solid ratio.
55
-------
SECTION III
LEVEL II AND LEVEL III SAMPLING STRATEGIES
At the completion of the Level I sampling and analysis pro-
gram, it is expected that basic information on the physical
composition of the process stream will have been obtained,
the classes of chemical compounds present will have been
identified, and the material collected will have been evalu-
ated to determine its cytotoxicity and mutogenicity proper-
ties. At this point, an evaluation of the data would indi-
cate if sufficient data to assess the source is available.
Based on this evaluation, a decision can be made to terminate
the study of this source or to continue to the more complex
Level II or Level III programs.
This chapter discusses methods that can be employed if Level
II or Level III studies are indicated. To review, a Level II
program is an extensive qualitative, semiquantitative approach
to identify specific substances that exist in streams having
a significant environmental impact. At the conclusion of
Level II, the physical, chemical and biological characteris-
tics of the stream should be well understood under "normal"
operating conditions. Level III is a quantitative study of
the effect of process variables on the emission rates of
specific substances that have been identified as appropriate
indicators of the environmental impact of the process streams.
The indicators to be chosen are identified in a Level II
study. As a result, a Level II evaluation must be done before
57
-------
a Level III study or, if potential indicators can be identi-
fied before sampling. Level II and Level III sampling pro-
grams can be done concurrently.
In the previous chapter of this manual, Level I sampling
methods were specified for each of the various sampling situ-
ations. It is expected that these methods will be used with-
out major deviations in an effort to standardize the proce-
dures over a wide variety of source types. In the Level II
and Level III approach, more flexibility is required in the
methods; the methods should be specific for the material to
be analyzed and for the site to be sampled. As a result,
there is considerably more judgment required in the selection
of sampling methods. Therefore, a number of potential pro-
cedures are presented in this manual that have been or may be
used. The choice of methods will depend on the objectives of
the sampling program, the nature of the chemical compounds,
the physical layout of the sampling site, and the accuracy
and precision desired in the data.
In the Level III program, detailed information on the process
and process variables must be obtained during the sampling
effort. This requires complete cooperation between the plant
personnel and the sampling teams, and in addition, the abili-
ty of the sampling group not only to sample the emissions but
also to measure the process conditions as they vary during
the program.
3-1 GAS MATRIX SAMPLING STRATEGIES
Organic air emissions occur in the condensed, readily condens-
ible, and vapor forms. Of these, only the vapor form presents
a serious challenge to the design of efficient sampling and
-------
concentration approaches. In general, the condensed organic
species can be removed by filtration of an integrated sample
over a period of time sufficient to satisfy the detection
limit of the analytical procedure. Similarly, readily con-
densed species can be sampled employing trapping procedures
at ice or dry-ice temperatures with care taken to avoid loss
of material by microfog (submicron aerosol) formation. Trace
organic vapor species generally require a concentration step
to attain the required detection limits. The most frequently
employed concentration techniques are solvent scrubbing, con-
densation (cryogenic trapping), adsorption on activated car-
bon, chromatographic equilibration, chemical reactions, and
chromatographic column trapping.
3.1.1 Stationary Source Methods
By far the simplest technique to obtain organic emission sam-
ples is grab sampling or some similar approach in which the
sample is collected slowly over a time period. The sample is
drawn from the source with a glass sampling bulb or into an
inert bag (such as Tedlar or Teflon) as shown in Figure 7 and
the entire contents are analyzed for organic materials. Typ-
ically, a particulate filter is used prior to the sample ves-
sel; this will also remove condensed species. As the vapor
at duct temperature is usually cooled in the collection device,
material will condense on the walls of the container and thus
may not be removable for the analytical device. The quantity
of collected material is small and not sufficient for deter-
mination of low levels of materials or for biological testing.
Reactions can occur within the sample and the results of the
analysis may indicate materials that were not in the gas
stream. Leakage is a major problem with these devices. This
approach might find utilization as a secondary collection
59
-------
AIR-COOLED 11
CONDENSER '
PROBE
RIGID
CONTAINER
Figure 7-
Integrated gas sampling train,
(Solid lines showing normal arrangement,
Dotted Lines - alternate arrangement,
evacuating chamber around bag.)
-------
procedure for low molecular weight gases that could easily
pass through other collection devices, but generally the pro-
cedure Is not recommended.
One other application for this type of sampling is in the
use of an on-site chromatograph with a flame ionization de-
tector. Suitable units are commercially available that are
truly "stack portable" and thus provide for the analysis of
methane and total hydrocarbon measurements on-site or with
the proper columns and detectors for specific classes of
volatile compounds.
Such on-site chromatographs can provide semiquantitative
analysis of specific compounds provided that calibration data
are obtained using the proper columns and detectors. Com-
pound identification is accomplished by the use of retention
times. This technique, while not providing absolute identi-
fication, is usually reliable. Examples of compound classes
that can be sampled and analyzed by this technique along with
the detector and typical column are given in Table 2.
Table 2. DETECTOR-COLUMN OPTIONS FOR THE GLC SEPARATION OP
VARIOUS CLASSES OF ORGANIC MATERIALS27
Class
Detector
Column
Aliphatic Hydrocarbons
Sulfur containing
Alcohols
Acetates
Chlorinated Compounds
CO, C02, N2, CH^
Flame Ionization
Flame Photometric
Flame Ionization
Flame Ionization
Flame Ionization
Thermal Conductivity
61
Chromosorb 102
UCON-50 on
Chromosorb T
5$ Carbonwax on
Chromosorb WHP
5$ Carbonwax on
Chromosorb WHP
10% DC200 on
Chromosorb WHP
Molecular Sieve
-------
This list is not all-inclusive; many other useful columns and
detector combinations can be found in the literature.
3.1.1.1 Solvent Scrubbing
Solvent scrubbing for organics is achieved using an impinger
train containing a solvent system or combination of solvents
which will trap the desired emissions. The train is often
held at ice temperature to enhance collection efficiency and
minimize loss of the desired components. Depending on the
concentration of the emission, the flow rate and the sampling
time, the solvent must be reduced in volume to concentrate
the pollutants before analysis. Evaporation of the solvent
runs the risk of significant losses in the more volatile com-
ponents of interest.
This type of approach was employed in the earliest studies
for organic materials. Typical particulate systems were em-
ployed to collect the inorganic particulate and condensed
organic materials on the filters. Those materials that were
in the vapor phase at the filter temperature passed on through
the filter. Those organic materials that could condense at
the impinger temperature were collected in the impingers of
the train. The actual gas temperature would depend on cool-
ing efficiency within the impingers and could be as high as
21°C (70°F) and still meet the method criteria. Materials
that condense in the 21-100°C (70-212°P) temperature range
may be collected, but low boiling point materials and those
with high vapor pressures would most certainly be lost. The
procedures require that the impingers be filled with water.
The organic materials will be found in the water, the acetone
rinsings of the impingers and later during laboratory work
in the water or chloroform-ether extract of the water phase.
62
-------
Results with this procedure were very disappointing. Often
a major portion of the collected material was found to be
silicone grease used to vacuum seal the impinger glass-to-
glass joints.
Another approach was to employ different collection media in
the impingers, such as toluene, xylene, methylene chloride,
etc. Silicone grease could be eliminated in favor of water
seals, Teflon seals, Spectrovac grease, or ground ball joint
seals. The approach may be quite useful for specific organic
materials if the collection efficiency of the impinger solu-
tion is known for the compounds of interest. It may be the
best approach for some materials, such as aldehydes, which
present difficulty in the more generalized sampling schemes.
The sampling train for aldehydes shown in Figure 8 is typi-
cal of impinger collection systems.
3.1.1.2 Condensation Techniques
Use of condensation techniques is one of the least desirable
approaches since (a) collection efficiencies are poor and
vary significantly with physical and chemical properties of
the substances being collected; (b) condensation of water
with attendant trap plugging and hydrolysis of collected or-
ganics can occur; and (c) aerosols (microfog) can form and
not be trapped unless electrostatic precipitators are used.
If significant amounts of moisture are present, as is often
the case with combustion processes, incineration, or absorber
vent gases, the trap will contain a two-phase system which
will require special handling before analysis. Cryogenic
trapping at temperatures sufficient to condense oxygen or
nitrogen requires the use of special equipment to carry out
analyses.28
63
-------
ON
GLASS WOOL
STACK
THERMOCOUPLE
READ OUT
VARIABLE
TRANSFORMER
FLOWMETER
Figure 8. Sampling train for aldehydes
-------
3.1.1.3 Use of Porous Polymer Adsorbents in Sampling Gas
Streams
Potentially, the most attractive method for collecting and
concentrating organic substances from fugitive or stationary
emission sources employs the adsorption and/or partitioning
properties of materials normally used in gas chromatographic
analysis to retain organic substances selectively while re-
moving the major diluent gases, such as air, nitrogen and
water vapor. By proper selection of materials which retain
little water, separation of organic substances from water
vapor can be accomplished even in samples taken in humid at-
mospheres. Various types of chromatographic materials have
been used. They include charcoal, molecular sieves, liquid
phases on solid supports (e.g., Dexil 300 GO on Chromosorb
AW HMDS and Silicone oil DC 200 on Chromosorb), and porous
polymers.
The retentive characteristics, varied polarity, high-thermal
stability, and low affinity for water of porous polymers,
suggest that these materials might be the best media for ef-
ficiently collecting and enriching organic substances, How-
ever, the varied nature of the emission sources requires an
evaluation of the limiting properties before specific appli-
cations can be defined.
A very simple sampling system using the porous polymer ap-
proach is shown in Figure 9. This system is suitable for
higher molecular weight hydrocarbon components (C7 and above)
The sampling time can be short — 5 to 15 minutes at flow
rates of 0.075 cfm to collect sufficient material for analy-
sis by GLC/MS using thermal desorption. Typical dry test
meters are not very accurate for yielding the total gas vol-
ume at such low flow rates.
65
-------
ROTOMETER
STAINLESS STEEL PROBE
SOURCE
n
POLYMER
PACKED
TUBE
D
VALVE
GAS METER
FLEXHOSE
Figure 9- Porous polymer vapor sampling method
-------
The sampling system as shown In this illustration consists
of a 1/4 in, or 3/8 in. stainless steel tube 6 in, long
packed with porous polymer bead material. This sampling
tube is attached to a pump (Model MP^-155 Metal Bellows Corp.),
a rotameter with valve (Dwyer 0*-5 SCPH), and a gas meter
(Type AI-,-110 American Meter Co.). The pump is required be-
cause there is an appreciable pressure drop due to the packr-
ing in the tube.
A similar train, shown in Figure 10 would replace the pump
with an evacuated cylinder to eliminate pump fluctuations in
the sample lines, The cylinder used for Preon refrigerant
of about 0.023 m3 volume is convenient and should be equipped
with a thermocouple or other temperature readout and a vacuum
guage. A glass sampling bulb is used to back-up the polyr
mer packed tube to permit at least qualitative identifica-
tion of low molecular weight gaseous species. Typical flow
rates of 100^200 cc/min for twenty minutes will provide 2
to 4 liters of sampled gas. The temperature and pressure
differential in the cylinder before and after collection pro-
vides the necessary volume of gas sampled. A partial list
of some of the organic materials detected by this procedure
from point system and polymer curing ovens is given in Table
3.
Recently. Tenax GC, a new, more polar, and more thermally
stable porous polymer, has become commercially available.
This system is based on 2,6~diphenyl*-p-phenylene oxide, Other
polymers that have been used in laboratory tests but that are
not widely used or are not commercially available, are polyr-
imides, polyamides, polyacrylates, and phosphonated or hal-
ogenated resins,
-------
V,
STACK WALL
STAINLESS STEEL PROBE
CO
POLYMER PACKED TUBE
250 ML
DOUBLE ENDED
FLASK
TEMPERATURE GAUGE
II
FLOWMETER
0
EVACUATED CYLINDER VACUUM PUMF
Figure 10. Alternate porous pplymei? systern £
-------
Table 3. ORGANIC SUBSTANCES DETECTED PROM PAINT AND
POLYMER CURING OVENS BY POROUS POLYMER AD-
SORPTION AND GC MASS SPECTROMETRIC ANALYSIS
cr\
Methanol
Ethanol
Isopropanol
2-Ethoxyethanol
Isobutanol
n-Butanol
C5 Alcohols
n^Propanol
2-Methylbutanol
Ethyleneglycol monoethyl ether
2^ (2r-ethoxyethoxy) ethanol
Formaldehyde
Acetaldehyde
Acrolein
Acetone
MethylethyIketone
Diethylether
Butylacetate
Saturated Hydrocarbons
2-Ethoxyethylacetate
Chloroform
Methylene chloride
Cyclohexane
DimethyIcyclohexane
Benzene
Toluene
Xylenes
Styrene
Methylstyrene
Dimethylstyrene
C3 Alkylbenzenes
Ck Alkylbenzenes
Cit Substituted Styrene
Trichloroethane
Dichloroethylene
Carbon disulfide
Isopropylbenzene
Phenol
Benzaldehyde
-------
Five groups of porous polymers are potentially usable as sor-
bents for collecting and concentrating organic compounds from
stack emissions. These are:
(1) Porapak series (Waters Assoc., Inc.)
(2) Chromosorb Century series (Johns-Manville
Products Corp.)
(3) XAD Resins (Rohm and Haas Co.)
(4) Tenax GC (Enka, N.V., the Netherlands)
(5) Polyimides
Note: Some XAD-type resins are marketed by Johns-
Manville Products Corp. as the Chromosorb
Century series; e.g., Chromosorb 102 is XAD-2.
A limited amount of information is available which directly
compares the chromatographic properties of these materials.
Retention indices obtained under similar operating conditions
are reported for two groups, namely, the Porapak, and the
Chromosorb Century series.
In general, the retention characteristics of the porous poly-
mers are influenced by both gas-solid and gas-liquid mechan-
isms. The pore size distribution and micropore volume, the
nature of the polymer, and the surface activity all influence
the adsorption, diffusion, and partitioning processes. Al-
though specific retention indices are not available for all
porous polymers, certain physical property data and a rela-
tive ranking of polarity can describe the relative retention
characteristics. These data are shown in Table 4.
70
-------
Table 4. PHYSICAL PROPERTIES OF POROUS POLYMERS
Type
Porapak P
Porapak Q
Porapak R
Porapak S
Porapak N
Porapak T
Chromosorb 101
Chromosorb 102
Chromosorb 103
Chromosorb 104
Chromosorb 105
Chromosorb 106
Chromosorb 107
Chromosorb 108
XAD-1
XAD-2
XAD-4
XAD-7
XAD-8
XAD-11
Tenax GC
Surface Area
(mVg)
110
840
780
670
437
450
30-40
300-400
15-25
100-200
600-700
-
—
-
100
300
784
450
140
69
(UNK)
Ave. Pore Diam.
fA)
150
75
76
76
-
91
3000-4000
85
3000-4000
600-800
400-600 '
-
-
-
200
90
50
90
235
352
Temp. Limit
(°C)
250
250
250
250
190
190
275
(325)*
250
(300)*
275
(300)*
250
(275)*
250
(275)*
250
(275)*
250
(275)*
250
(275)*
200-250
200-250
200-250
200-250
200-250
200-250
Monomer
Composition
STY-DVB
EVB-DVB
Vinyl pyrollidone
Vinyl pyridine
Vinyl pyrollidone
Ethyleneglyco-
dimethylacrylate
STY-DVB
STY-DVB
Cross-linked
PS
ACN-DVB
Polyaromatic
Cross-linked
PS
Cross-linked
acrylic ester
Cross-linked
acrylic ester
STY-DVB
STY-DVB
STY-DVB
Acrylic
Ester
Acrylic
Ester
Amide
Diphenylphenylene
oxide
STY-styrene; DVB-divinylbenzene; PS-polystyrene; ACN-acrylonitrile
*Maximum temperature for short duration
71
-------
The major problems related to the use of porous polymers as
sorbents for collecting organic compounds from industrial
emission sources are:
(1) Displacement of Volatile Compounds - The dis-
persement of volatile organic species by less
volatile organic substances is a major problem
when using porous polymers. High molecular
weight compounds are more readily retained
than low molecular weight substances.
(2) Irreversible Adsorption or Poor Desorption Ef-
ficiencies ^ As derived from the information
sources and general commercial literature, the
most pertinent data to use for porous polymers
as adsorbents relate to the chemical classes
that cannot be desorbed from the resins. Gen-
erally, the adsorption characteristics of most
resins are adequate. However, some chemical
classes are irreversibly adsorbed or are de-
sorbed slowly over a relatively long period.
In this respect, we are referring to the re-
moval of material from the polymer by appli-
cation of heat and a carrier gas flow.
The resins and associated chemical classes
that will provide potentially poor thermal
desorption efficiencies are as follows:
Glycols - Complete adsorption
on Chromosorb 103
Some tailing on Pora-
pak Z, R, and S. Se-
vere tailing on Pora-
pak QS
72
-------
Nitriles - Severe tailing on Chromo-
sorb 103
Nitroparaffins - Severe tailing on Chromo-
sorb 103
Amines and - Severe tailing on Chromo-
diamines sorb 101 and 102 Porapak N,
P, Q, R, S, T. Some tail-
ing on Porapak QS
Anilines - Severe tailing on Porapak
N, Q, S, T, QS
Some tailing on Porapak R
Carboxylic acids - Complete adsorption on
Chromosorb 103
Severe tailing on Porapak S.
Some tailing on Chromosorb
102 and Porapak Q
Alcohols - Some tailing on Porapak N.
Branch-chain broadening on
Chromosorb 101, 102, 103
and Porapak T
Many of these problems can be eliminated by solvent
extraction, and this procedure is generally pre-
ferred. The solvent must not react with the porous
polymer but must still have sufficient solvent
power to remove the materials of interest quanti-
tatively. In addition, the solvent should not inter-
fere with the analysis scheme. For the most part
aliphatic hydrocarbons such as pentane or hexane
will not react with the Chromosorb or Tenax mate-
rials and are easily removed to concentrate the
73
-------
sample prior to analysis. These materials are
generally poor solvents, but continuous extraction
procedures (Soxhlet extraction, etc.) minimizes
this problem.
(3) Chemical Reaction of Sorbates with Sorbents and
Production of Artifact Species - Porapak Q and
Chromosorb 102 were found to react with N0229
oxygen.30 The reaction with N02 yields NO, water
and nitrated aromatic rings of the polymer plus the
possible increased olefinic unsaturation and/or
oxidation of the polymer. Oxygen reacts with the
resin above 100°C to depolymerize part of it, pro-
ducing carbonyl compounds. As a result, the use of
this type of polymer for adsorption of organics
should be restricted to temperatures of about 60°C
or lower.
In general, polystyrene-type materials suffer from
oxidation and thermal fragmentation at temperatures
above 250°C.
Change in Sorption Properties of Porous Polymers -
The reactions discussed above in (3) undoubtedly
influence the sorption properties of porous polymers.
The displacement phenomena indicated above in (1)
also point out potential problems related to physi-
cal adsorption changes at collection temperatures
below l4o°C, where the physical adsorption mechanism
for compound retention predominates with the styrene
or divinyl benzene systems.
Also, problems may be experienced when using porous
polymers under high-humidity, high-temperature con-
ditions. Although Chromosorb 101, 102, and 103,
and Porapaks N, P, Q, QS, R, S, and T are hydrophobic,
-------
and Tenax GC has little affinity for water, some
water can be adsorbed on the resins. It has been
reported that Porapak Q retains up to 3-^ ug H20/g
of polymer at 110°C.31
At present it is generally assumed, in a qualita-
tive sense, that the interaction mechanism for ad-
sorbates on porous polymers is a combination of
both adsorption and partitioning, especially at
higher temperatures. Below the glass transition
temperature (T ), absorption of organic vapors by
&
porous polymers occurs through very complex proc-
esses. Amorphous polymers would be expected to ad-
sorb organic vapors to a much greater extent if
they were in a rubbery state rather than a glass
state. Data suggest that surface adsorption mech-
anisms should predominate for organic molecules at
temperatures below 1^0°C.
(5) Retention Capacity of Porous Polymers - Pore size
determinations for Porapak P and Q indicate that a
large proportion of very small pores exist in these
resins, particularly Porapak Q. As a result, a
large portion of the "N2" surface area reported by
the manufacturers may not be available to the more
bulky organic molecules. Chromosorb 101 has rela-
tively large pores compared to Porapak P and Q.
Estimates of "available" surface area to organic
molecules were made by Gearhart and Burke32 for
Chromosorb 102, Porapak P, and Porapak Q. The basis
for their estimates was the measurement of free
energy changes for molecular probe-adsorbent inter-
actions. By relating these measurements for ben-
zene, cyclohexane, cyclohexene, hexane, hexene,
75
-------
raethylene chloride, and chloroform, estimates of
"available" surface area were computed. Chromo-
sorb 101 was used as a norm for comparison since
it probably has the greatest available surface
area. The reported apparent surface areas for
Chromosorb 102, Porapak P, and Porapak Q are 95,
37, and 133 m2/g, respectively. These estimates
represent 33-1%» 27.1%, and 20.2$ of the manufac-
turer's reported surface areas.
(6) Thermal Stability of Sorbent - Thermal stability
of the porous polymer sorbent is critical princi-
pally from the standpoint of the optimum tempera-
ture for desorption. If relatively high molecular
weight materials (e.g. MW 1^0) are to be measured,
desorption temperatures as high as 290-300°C may
be required. Obviously, lower molecular weight
materials will be desorbed at lower temperatures.
The choice of sorbent for a particular sorbate
will depend in large part on the temperature needed
if the organic materials are removed by gas flow
and thermal desorption.
3.1.1.4 Adsorption on Chemical Substrates and Silica Gel
Sample collecting and concentration techniques based on ad-
sorption on activated carbon have been used extensively.
Activated charcoal has been shown to quantitatively remove
an extremely broad range of organic contaminants from air.
The National Institute of Occupational Safety and Health
(NIOSH) has promulgated a general procedure for sampling and
analysis of organics in work place atmosphere.33'31* This pro-
cedure is based on adsorption of the organics on activated
76
-------
charcoal and desorption with carbon disulfide followed by
subsequent analysis by gas chromatograph. The current NIOSH
procedure for vinyl chloride in in-plant atmospheres employs
this method. While the adsorption process is quantitative,
the recovery of the collected components is usually incom-
plete and variable.35 The charcoal may also serve as a
catalyst to promote alteration of the sample,36'37 and it is
extremely subject to adsorption of water vapor, which limits
the adsorption capacity and can displace the desired organic
components. Desorption by heating requires high temperature
(up to 400°C) and is accompanied by chemical changes due to
pyrolysis of the organic species and thermally enhanced re-
actions between the components.
Silica gel has been used for collecting three-carbon and
higher molecular weight hydrocarbons. The collection ef-
ficiency for lower hydrocarbons^ such as ethylene, from air,
has been demonstrated to be poor even when trapping at dry
ice acetone temperatures.38
3.1.1-5 Sampling in Particulate Laden Streams
The porous polymer materials have also been used in particu-
late-laden streams as a supplemental collection medium fol-
lowing a filter or cyclone collection device. As mentioned
earlier, low flow rate particulate systems have been used
for collection of the particulate in the probe and filters
and the volatile organics in the impinger liquid. A more
efficient procedure is to employ the porous polymer in place
of or in addition to the impinger collection and thus elimi-
nate some of the differences mentioned earlier.
77
-------
One of the early attempts at this approach consisted of a
probe, heated filter, ice cooled impinger and a porous poly-
mer trap containing Tenax GC. In general, the results indi-
cated that more organic material was collected by the ad-
sorber than by the impinger collectors.
A modification of this technique uses organic liquids in the
impingers that precede the porous polymer adsorber. This
procedure is generally unsatisfactory because the organic
liquid vapors are trapped in the adsorber. The analysis of
the collected species is then complicated by the presence of
a large quantity of impinger solvent and a small quantity of
other collected organic material.
A more appropriate position for the adsorber would be between
the filter used to collect particulate and the impingers.
The collected particulate could be both inorganic and organic,
depending on the filter temperature and nature of the emis-
sion source. The porous polymer adsorber will collect or-
ganic vapors, but temperature becomes an important factor.
In streams with high water content, water can condense if the
temperature of the adsorber is below the dew point tempera-
ture of the gas. If the temperature of adsorber is too high,
organic material will not be collected or can be driven off.
As new stack gas reaches the adsorber, an equilibrium will
result and the adsorption of the organic materials could be
less than quantitative.
Peter Jones et al of Battelle39 reported the results of an
organic sampling study that compared these various systems.
Block diagrams of the systems are shown in Figure 11. System
4 collected the most material, even more than system 3 if
all impinger, filter, and organic absorber contents were
added together. Systems 1 and 2, based on the analysis of
78
-------
PROBE
( 1) EPA TRAIN
PROBE
< 2 ) HIGH-VOLUME SAMPLER
PROBE
(3)MODIFIEDEPATRAIN
STANDARD
IMPING ERS
OVERSIZE
IMPINGERS
LINEAR MASS
FLOW METER
STANDARD
IMPINGERS
ADSORBENT
SAMPLER
I
1
PROBE
ADSORBENT
SAMPLER
IMPINGERS
|
!
( 4) ADSORBENT SAMPLING SYSTEM
Figure 11. Diagrams of the. sampling systems used in the
Battelle train evaluation study
-------
organic material collected, were not as efficient as systems
3 and 4. The differences between systems 2 and 4, were ap-
parently caused by the S02 content of stack gases. With low
S02 content, ,the results were in closer agreement, indicating
that "aqueous S02" may chemically alter some of the collected
materials.
Based on this work and other recent work, it appears that
reasonable results could be obtained using a modified particu-
late sampling system by placing the porous polymer adsorbent
after the filter and before the impinger train.
It should be stressed that this system has been used for mate-
rials of relatively high molecular weight and lower volatility
than many of the compounds that would be of interest in gen-
eral organic sampling schemes. At this time, little is known
about the accuracy of the techniques since the "true" con-
centration is not known. It is known, however, which system
collects the greater portion of material and it can be as-
sumed that the true value is being approached. In other words
there is good knowledge of the precision of particulate sam-
pling, but very little data on the accuracy.
3.1.1.6 Procedures for Low Molecular Weight Materials
The preceding sections of this chapter have dealt with the
collection of organic materials using procedures applicable
to materials with molecular weight above Cy, and several pro-
cedures have been mentioned for collecting low molecular
weight materials. One such procedure was use of a gas sam-
pling bulb after a porous polymer adsorber. Another sug-
gested the use of a Tedlar or Teflon collection bag for an
integrated gas sample.
80
-------
At present, there are many unanswered questions as to the
collection efficiency of porous polymers for low molecular
weight materials. If water can pass through a Porapac or
Tenax adsorber, then there is reason to suspect that methanol,
ethanol, and other low molecular weight alcohols may not be
collected quantitatively. Fortunately, these materials are
not considered highly toxic, and some loss may not be signif-
icant.
One possible approach for the collection of low molecular
weight materials is to employ a cryogenically-cooled thermal
gradient tube patterned after that described by R. E. Kaiser.^
The thermal gradient tube will isolate organic emissions that
slip through the primary porous polymer sampling tubes.
The sampling train utilizing the porous polymer and Kaiser
tubes is shown schematically in Figure 12. An evacuated tank
is used as a sampling gas driving force when a positive pres-
sure flow stream is not available. The valve and calibrated
flowmeter upstream of the tank adjusts the flow rate in the
general region of 100 to 200 cc/minute. Sampling conducted
over a 20-minute period provides a total sampled volume of
2 to 4 liters of effluent gas. The cryogenic thermal gradient
tubes are fitted with thermocouples, and the porous polymer
adsorption tubes are constructed of stainless steel and stan-
dard Swagelok fittings. A condensation trap (0°C) is posi-
tioned before the polymer tube to trap aerosolized water and
water vapor.
Some details of the sampling system includes the following:
- The evacuated cylinder is an 0.02 m3 (0.83 ft3)
Freon tank fitted with a thermocouple and a 3-in.
vacuum gauge to permit calculation of total vol-
ume sampled.
81
-------
oo
ro
TEMPERATURE GAUGE
PRESSURE GAUGE
.THERMOCOUPLES
HEATED SECTION
POLYMER PACKED TUBE
VACUUM PUMP
STAINLESS STEEL PROBE CONDENSER
EVACUATED CYCLINDER
LIQUID NITROGEN DEWAR
COMPRESSED NITROGEN
Figure 12. Porous polymer and thermal gradient sampling train
-------
- The liquid nitrogen dewar is a four-liter Nalgene
dewar flask (Cat. #4150) constructed of a double
wall of highly crosslinked polyethylene.
- The condensation trap in a midget impinger cooled
by an ice-salt mixture to about -10°C,
- A photograph of the Kaiser tube is shown in Figure
13. The inner tube of the concentric heat ex-
changer is a 20-cm length of 3/16 in. stainless
steel tubing and the outer tube is a 9-cm length
of 5/16 in. stainless steel tubing.
- The porous polymer tube is a 7 inch length of
1/4 in. stainless steel tube fitted with Swage-
lok fitting and plugs.
The flow of nitrogen through the jacketed thermogradient tube
is adjusted so that the entering nitrogen flow is near liquid
nitrogen temperature (-l60°C). Under normal operating con-
ditions, the thermocouple designated TC 2 will register
about -100°C.
This technique is considerably more complex than desired for
field sampling work. However, the method does collect the
volatile compounds that may not be collected by the ambient
temperature porous polymer sampling system.
3.1.2 Fugitive Emission Methods
In Section 2.1.2 of this manual, procedures were specified
for the sampling of particulate and vapor fugitive emissions,
which are consistent with the Level I criterian of obtaining
large quantities of material in short periods of time. In
83
-------
Figure 13. Kaiser tube configuration
-------
a Level II or Level III sampling program, it can tie assumed
that more effort will be required to obtain the required sam-
ple. In the case of particulate sampling for organic mate-
rials, the additional time does not change the basic sampling
concept of using high volume samplers. However, a more re-
liable sampling network can be employed. For fugitive vapor
sampling, a wide variety of techniques can be employed that
are more dependent on the chemical and physical properties
of the material to be sampled.
3.1.2.1 Particulate Emissions
For Level I sampling for particulate-emissions it was sug-
gested that a one or two high volume samplers collect the
emission sample and another sampler collect a background sam-
ple. This procedure is suitable if there is little change
in the wind direction. In contrast, a network of high volume
samplers completely around a source is relatively unaffected
by wind direction, but such a network is too complex for even
Level II or Level III studies.
In a number of fugitive emission projects, MRC has employed
a diffusion model and a five-sampler network that can be used
with a Level II or Level III sampling program. Details of
this diffusion model are presented in Appendix B.
However, a compromise system of five samplers provides a
very satisfactory Level II or III sampling procedure. If
four samplers are deployed downwind from a fugitive source
as shown in Figure 14, small changes in wind direction dur-
ing the course of sampling will have little effect on the
collection of sample and thus the potential of repositioning
the samplers as might be required in the Level I approach is
-------
Angle of Wind
B
E
Source
Resultant
Wind
Direction
D
Figure
Sampler positions; S0 - background;
Sj, S2, S3, Sij - sample collectors
86
-------
avoided. The two samplers that are directly downwind from
the source provide some indication of how quickly particulate
matter is falling out of air. This provides a simpler system
than placing samplers on a tower in order to obtain informa-
tion on the dispersion of the fugitive emission plume.
The five sampler system requires little more effort than does
the Level I system, but does provide for a great deal more
flexibility in operation and variation in wind conditions -
both direction and velocity.
The roof monitor and quasi-stack procedures described in
Level I are equally applicable to Level II or III programs.
3.1-2.2 Organic Vapor Emissions
In the preceding sections several methods are described that
can be employed for Level II and Level III fugitive organic
vapor emissions. The flame ionization detector chromatograph
with suitable separation columns described for stationary
source emission is suitable for ambient vapor concentration
levels. This sensitivity can be increased by using precon-
centration procedures, such as porous polymer tubes, prior
to injection into the chromatograph. Organic sulfur com-
pounds should be analyzed on-site with a flame photometric
gas chromatograph.
Other techniques such as grab sampling (glass or Tedlar bag)
impinger or trapping procedures can be used. A number of
procedures have been developed for the in-plant atmosphere
study for organic solvents and many of these procedures can
be used directly for the sampling of low level fugitive emis-
sions. One source of such methods is the series of procedures
87
-------
available from the National Technical Information Service,
Springfield, Va.41 At present, there are ten sets with 6 to
16 sampling and analytical procedures in each set. While
these methods are generally specific for an individual com-
pound, there is the possibility that a given procedure can
be used for many other compounds in the same homologous series,
The other stationary source methods, porous polymer and Kaiser
tube collection, can be used in ambient atmospheres by in-
creasing the sampling time.
The high volume sampling systems have also been adapted for
vapor sampling. For this purpose the usual 20 x 25 cm filter
mat is replaced by a two-stage sampling head. This sampling
head contains a 10 cm circular fiber glass filter, followed
by a 5 cm x 10 cm vapor trap containing polyurethane foam or
coated glass beads. The polyurethane foam has been used suc-
cessfully for polychlorinated biphenyls42, and glass beads
coated with cottonseed oil have been used for air-borne pesti-
cide sampling.43 Very recently, it has been observed that
many high volume samplers contain a capacitor to reduce spark-
ing, and contamination of samples has been traced to a poly-
chlorinated naphthalene from this electrolytic capacitor.
Apparent concentrations of this material were found as high
as 1 ng/m3; high enough to prevent determination of pesti-
cides, PCB's and other trace organic materials.
3.2 LIQUID SAMPLING SCHEMES
In aqueous stream sampling, it is imperative that a Level I
study be conducted before either a Level II or Level III pro-
gram is begun. The Level I effort would identify what
classes of compounds are of major importance and also provide
88
-------
approximate levels of this material. It is not the intent of
the Level I effort to provide quantitative data, however, from
the Level I samples one can determine how much sample will be
required for the more specific individual compound identifica-
tion and quantification. As the analytical procedures to be
employed in Level II and Level III studies will generally be
rather sensitive, it will not be necessary to collect the
large quantities of sample specified in Level I unless ad-
ditional bio-testing is required. Regardless, sufficient sam-
ple must be available to satisfy the analytical separation
and detection limits. The volume requirements for the stan-
dard analysis of wastewater are presented in Table 5.
In general, Level II and Level III sampling program? will pro-
vide more sampling time than a Level I study. As a result,
mechanical means of collecting samples are preferred, and a
considerable number of suitable commercial samplers are avail-
able.
3.2.1 Sampling Systems
A major requirement of a sampling system is that it collect
a representative sample. Grab samples provide neither average
concentrations nor variations in concentration. The average
concentration is best determined by a composite sample col-
lected over periods ranging from hours to weeks. Variations
in the composition are determined by individual samples taken
at specified intervals. The intervals between the samples
must be small enough to show these variations. Another prob-
lem with processes that show a wide variation is that the
flow rate through the source may also vary. The question
that must be answered is whether sampling should be done in
relation to time or in proportion to the flow.
89
-------
Table 5. VOLUME OF SAMPLE REQUIRED FOR DETERMINATION OF
THE VARIOUS CONSTITUENTS OF INDUSTRIAL WATER10
Volume of
Sample,3 ml
Volume of
Sample,3 ml
PHYSICAL TESTS
*Color and Odor 100 to 500
Torrosivity ... flowing sample
*Electrical conductivity 100
*pH,electrometric 100
Radioactivity 100 to 1000
'Specific gravity 100
*Temperature flowing sample
*Toxicity 1000 to 20 000
Turbidity 100 to 1000
CHEMICAL TESTS
Dissolved Gases:
tAmmonia, NH3 500
tCarbon dioxide, free
C02 200
tChlorine, free C12 200
fHydrogen, H-, 1000
fHydrogen suffide, H^S 500
tOxygen,02 "• 500 to 1000
tSulfur dioxide, free SO2 100
Miscellaneous:
Acidity and alkalinity 100
Bacteria, iron 500
Bacteria, sulfate-reducing 100
Biochemical oxygen demand 100 to 500
Carbon dioxide, total CO2
fincludingC03"JiCO3',
and free) 200
Chemical oxygen demand
(dichromate) 50 to 100
Chlorine requirement 2000 to 4000
Chlorine, total residual C12
(including OC1', HOC1,
NH2C1, NHC12, and free) 200
Chloroform - extractable
matter 1000
Detergents 100 to 200
Miscellaneous:
Hardness 50 to 100
Hydrazine 50 to 100
Microorganisms 100 to 200
Volatile and filming amines 500 to 1000
Oily matter 3000 to 5000
Organic nitrogen 500 to 1000
Phenolic compounds 800 to 4000
pH, colorimetric 10 to 20
Polyphosphates 100 to 200
Silica 50 to 1000
Solids, dissolved 100 to 20 000
Solids, suspended 50 to 1000
Tannin and lignin 100 to 200
Cations:
Aluminum, A1+++ 100
f Ammonium,NH^
Antimony, Sb+++ to Sb+++++ • • • 10°
Arsenic, As+++ to AS+++++ 100
Barium, Ba"1"1" 100
Cadmium, Cd++ 100
Calcium, Ca++ 100
Chromium ,Cr+++ to Cr+-|"l"f++ .
Copper, Cu++
flron, Fe++ and Fe+++
Lead, Pb++
Magnesium, Mg++
Manganese, Mn^to Mn+++~H"H"
Mercury, Hg+ and Hg++ 100
Potassium, K+ 100
Nickel, NH+ 100
Silver, Ag+ 100
Sodium, Na+ 100
Strontium, Sr++ 100
Tin, Sn++ and Sn++++ 100
Zinc,Zn++ 100
100
200
100
100
100
100
to 1000
500
to 1000
to 1000
to 1000
to 1000
to 1000
to 1000
to 4000
to 1000
to 4000
to 1000
to 1000
to 1000
to 1000
to 1000
to 1000
to 1000
to 1000
to 1000
to 1000
90
-------
Table 5 (cont'd.). VOLUME OF SAMPLE REQUIRED FOR DETERMINATION OF
THE VARIOUS CONSTITUENTS OF INDUSTRIAL WATER10
Volume of
Sam pie ,a ml
Anions:
Bicarbonate, HC03 100 to 200
Bromide, Br" 100
Carbonate. CO3" 100 to 200
Chloride, CK 25 to 100
Cyanide, Cn" 25 to 100
Fluoride, FT 200
Hydroxide, OH" 50 to 100
Iodide, I' 100
Nitrate, N03" 10 to 100
Nitrite, N02 150 to 100
Phosphate, ortho, PO4~",
HP04",H2P04- 50 to 100
Sulfate, SO4", HS04" • •'. 100 to 1000
Sulfide, S", HS" 100 to 500
Sulfite, SO3~, HSO3' 50 to 100
a Volumes specified in this table should be considered as a guide for the approximate quantity of
sample necessary for the particular analysis. The exact quantity used should be consistent with the volume
prescribed in the standard method of analysis, whenever the volume is specified.
* Aliquot may be used for other determinations.
t Samples for unstable constituents must be obtained in separate containers, preserved as prescribed,
completely filled and sealed against all exposure.
91
-------
Since the concentrations of organic materials that may be
toxic is relatively low, it is suggested that the best approach
is to obtain a composite type of sample of at least 24 hours
duration. Several such composites may be required if the con-
centration of materials is expected to be low. Compositing
in the basis of flow would be the most desirable, particularly
if higher flow rate can be related to increased production
activity.
The choice of an appropriate sampling system for this type of
program is not a simple one. In contrast to most of the waste-
water sampling, a sampler is required that can be used in a
wide variety of conditions and not designed for a particular
source. Some of the requirements for a system are as follows:
(1) The sampler must be portable.
(2) The sampler should be able to take individual
samples, with the option of a composite sample.
(3) Plow-proportional sampling with time-proportional
option.
(4) The sampler should be refrigerated so that col-
lected samples can be maintained at 4°C.
(5) The sampler should be designed to prevent con-
tamination of samples either from the sampler
or from previously collected samples (sampling
lines).
(6) The unit should be self-contained and operate
without connection to an external power source
for at least 24 hours. Operation on 115 volt
lines should also be possible.
92
-------
(7) It is imperative that the sampler can function
over a 24 hour period with only intermittent
attention after the initial set up.
(8) The sampler should be versatile enough and
provide enough options to make it useful for
both organic and other types of sampling pro-
grams .
(9) Sample containers should be of glass or Teflon
to prevent contamination of the samples.
(10) The sampler should be weather-proof to the ex-
tent that it could operate under all types of
conditions and should be explosion-proof for
safety in operation.
(11) The sampler should be equipped with pumps
capable of extracting samples from streams
that are up to 20 feet below the sampler lo-
cation.
There are a number of samplers that are commercially avail-
able that can meet most of these criteria. A number of such
samplers and their capabilities are listed in Table 6. The
information shown is from sales literature and data presented
in an EPA report by Harris and Keffer44 and as a result may
not completely cover all of the available models and manu-
facturers .
It is of interest to note that Harris and Keffer indicated
that in their opinion they have had the best success with
the Quality Control CVE unit and a schematic of the sampler
is shown in Figure 15. They did not have any of the Manning
units or the newer ISCO units. Laboratory examination of
93
-------
Table 6. SAMPLERS FOR WASTEWATER COLLECTION
Name
Sirco Automatic
Quality Control
Quality Control
Isco
Isco
Isco
Manning
Manning
Sigmamotor
Sigmamotor
Brailsford
Brailsford
Brailsford
Hants
Sirco
N-Con
N-Con
Model No.
NW3-8
CVE
CVE-II
1680
1580
1580R
S3000
SitOOO
WA-2
WD-2
EV-1
DU-1
EP-1
3B
MKU57
Scout
Surveyor
Composite or
Sequential
sequential
composite
composite
sequential
composite
composite
composite
sequential
composite
composite
composite
composite
composite
composite
composite
composite
composite
Sample proportional to:
_. Time Flow
yes
external
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
yes
no
external
external
external
external
external
external
external
no
no
yes
VP S
y c o
no
no
yes
no
yes
Purge
Refrigerated Peature_
add ice separate
lines
no yes
no yes
no yes
no yes
(ice)
yes yes
no yes
no yes
no no
no
no
— no
no
--- no
T— T- yes
yes
gravity
Self-contained
no
yes
yes
(4 days)
yes
(1 day)
yes
(18 hra.)
no
yes
(120 samples)
yes
(120 samples)
no
yes
yes
yes
yes
manual
yes
yes
no
Sample Bottles
glass
glass
glass
P.P. or glass
glass or PE
glass or PE
glass or P.P.
P.P.
glass or PE
glass or PE
glass
glass
glass
glass or PE
glass
glass or PE
glass or PE
-------
VACUUM SYSTEM
BLOW DOWN
SOLENOID VALVE
BLEED & DRAIN VALVE
115V INPUT
Figure 15. Schematic of the CVE Sampler
-------
the ISCO 1580 and Manning S3000 and S4000 units indicates
that these instruments should perform well.
Harris and Keffer1*1* state that the flow compositing of waste-
water with appreciable solids content is not justified and
further flow-proportional sampling should be restricted to
well treated effluents with no visible solids. In this work,
the long-term dependability of the sampler is of the utmost
importance, and they had the option of choosing a particular
sampler for a particular site. For the type of sampling dis-
cussed in this manual, a sampler is needed that can be used
for short periods over a wide variety of conditions and sites,
Therefore, while dependability is important, a sampler should
be chosen that has many options and modes of operation and
that can be operated to collect many types of samples under
a variety of conditions.
The ISCO 1680 or Manning S4000 should provide the necessary
capabilities, but these units have not been tested in the
field nor have all other possibilities been examined. These
two units are sequential samplers. Manual compositing would
provide a composite sample if desired. The Manning unit can
also be programmed to collect partial bottles. In this way,
small composites of a number of individually collected sam-
ples can be obtained. For example, samples equivalent to
1/4 of a bottle could be collected each 15 minutes. Each
full bottle would represent a one-hour composite and in 24
hours, 24 one-hour composites would have been collected.
Both units can receive sampling signals from an external de-
vice and so could be programmed on flow proportion or any
other porameter. The Manning sampler employs a vacuum pump,
a decided advantage over the peristaltic pump in the ISCO
for streams containing high levels of suspended matter. It
would appear that glass bottles are not an option to the
96
-------
standard polypropylene bottles for the Manning unit, a decided
disadvantage for organic samples, but they are available for
the ISCO sampler. The Manning unit can be ice-cooled, while
ISCO has an optional refrigerator available for this sampler.
A schematic of the Manning sampler is shown in Figure 16.
It is thus quite clear that each sampler is better suited for
certain applications, and the proper selection of a unit must
be based on the nature of the sampling program.
Under certain sampling conditions, it would be desirable to
employ a concentration-type device In the field. For this
approach the CAM in either the high-flow-rate or low-flow-rate
versions may be useful. In a Level II or III type of study,
some of the disadvantages of the low-flow system would be
eliminated since considerably more sampling time would be
available.
One possible approach that may be useful is to employ a porous
polymer such as Tenax GC In place of the carbon as the organic
adsorber. The Tenax has a low affinity for water and if the
passage of large quantities of water through the material does
not reduce its ability to absorb organic material, it may be
possible to directly substitute the Tenax for carbon in the
CAM-type adsorber. This type of approach has been employed
in laboratory separations of organics from water, so there is
evidence that the technique has promise in field sampling.
The use of Tenax is an advantage in that generally organics
can be removed from this material more readily than from car-
bon. Also, the technique that will be employed in the han-
dling of the Tenax in gas matrix sampling can be employed on
the water-collected samples once the water is removed. The
Tenax technique as a field concentrator will require further
97
-------
FILL SENSOR
COMPRESSOR
CONTROLS
INLET HOSE
VALVE
SAMPLE SIZE
ADJUSTMENT
MEASURING
CHAMBER
r
STEPPING
MOTOR
SPOUT ROTARY UN I ON
DISCRETE
SAMPLE BOTTLES
J
] VALVE
'"I
COMPOSITE
INTAKE
Figure 16. Schematic of the Manning S^OOO Sampler
98
-------
development, and possible loss of low molecular weight vola-
tile organics must be investigated.
3.2.2 Flow Measurement Techniques
The analysis of samples collected in any sampling program at
Level II or Level III will provide quantitative data in terms
of the concentration in a unit volume. However, in order to
assess the seriousness of the emission of a toxic material
and relate this to process variables, it is necessary to de-
termine the total flow of a system in a given time. The se-
lection of the proper measurement method will depend on a
variety of factors such as cost, accessibility of the source,
distances involved, and the characteristics of the stream.
In a monitoring program, it is much more feasible to set up
a permanent system for measurement of flow rate, but in the
assessment type of program, it is necessary to evaluate a
trade-off of cost against the reliability of the data.
In an organic sampling program, there are several types of
flow that must be considered; flow in open partial-filled
channels as in waste streams and flow in completely filled
ducts as in process lines or incoming water supplies. In
the partially filled stream, the flow can be fairly constant
with respect to velocity and depth, or it can be quite varied
over a time period. In the closed, completely filled streams,
it is usually possible to employ methods that will relate
some easily measured parameter such as pressure to the flow
rate. Rabosky and Koraido45 summarized the methods of flow
measurement as listed in Table ?•
99
-------
Table 7. METHODS OF FLOW MEASUREMENT AND THEIR APPLICATION1*5
H
O
o
Pump capacity and operation
Floating objects
Dyes
Salt dilution
Orifice meter
Weirs and flow recorders
Parshal Flume
Venturi meter
Magnetic flow meter
Flow nozzles
Pltot tube
Rotameter
Plow Range
Measurement
Small to large
Small to large
Small
Small to large
Small to medium
Small to medium
Small to medium
Small to large
Small to large
Small to large
Small to large
Small to large
Small to large
Small to medium
Small to medium
Cost
Low
Low
Low
Low
Low
Low
Low
Medium
Medium
High
High
High
Medium
Medium
Medium
Ease of
Installation*
N/A
Fair
N/A
N/A
N/A
N/A
N/A
Fair
Difficult
Difficult
Pair
Pair
Fair
Fair
Fair
Accuracy of Data
.Fair
Excellent
Good
Good
Fair
Fair to average
Fair
Excellent
Good to excellent
Excellent
Excellent
Excellent
Excellent
Good
Excellent
where Joints can be dis-
connected
Lines where water is being
pumped
Open channels
Pipe flow and open channels
Pipe flow and open channels
Pipe flow
Open channels
Open channels
Pipe flow
Pipe flow
Pipe flow
Pipe flow
Pipe flow
*N/A - Not Applicable
-------
3.2.2.1 Procedures for Closed, Filled Systems
A number of standard methods are available for the measure-
ment of parameters that can be related to flow rate in this
type of stream. A summary of these methods and the situations
where they can be employed are given in Table 8. Of the
methods listed, the water meter and rotameter are restricted
to certain types of materials. The venturi meter, flow noz-
zle, orifice meter, pitot tube and magnetic flow meters are
more generally useful, but all require special pipe, pipe
sections, or specific pipe designs. As a result, all such
devices will require installation during a process inter-
ruption. The pressure-sensing devices, with the exception of
the pitot tube, will restrict the flow to varying degrees.
A high solids content creates problems with all pressure de-
vices. The magnetic flow meter eliminates the flow restric-
tion and solids problems but requires that the liquid be a
conductor. Thus it is not appropriate for nonaqueous streams.
Fortunately, the flow rate in many closed-type filled ducts
is part of the information required in the control of chemi-
cal processes. The required flow rate information can often
be obtained from the process logs and instrument data, how-
ever, it is essential that the instruments and/or flow de-
vices be recalibrated on or about the time a sampling program
is initiated. Arrangements for such recalibration should be
made as part of the pretest site survey at the plant location,
since scheduling of such recalibration may be a problem and
may only be possible during a plant shut-down.
101
-------
Table 8. METHODS OF MEASUREMENT OF FLOW IN CLOSED PIPES
Method
o
IV)
Water Meter
Venturi Meter
Plow Nozzle
Orifice Meter
Magnetic flow
Meter
Pitot tube
Rotometers
Parameter
Measured
Total flow
Pressure
Pressure
Pressure
Induced
Voltage
Pressure
Flow Rate
Application
fresh water
most materials
Waste with
moderate solids
low solids
content
high solids
content
straight pipe
sections,
low solids
clean liquid
Relationship to flow rate
direct
Q - 0.98AK/H (a)
Q - CAK/H (b)
Q - CAK/H (b)
Voltage is proportional
to to flow
depends on configuration
direct
Remarks
not suitable of other materials
accurate, free of solid accumulation,
expensive
between venturi meter and orifice cost
and effect of solids
low cost, high pressure loss, not
suitable for high solids waste streams
can be used in pipes down to 0.1 Inch.
Accuracy increases with Increased flow.
No pressure loss.
economical in large pipej sollda cause
plugging, requires straight pipe 15 to
50 times the diameter.
suitable for only clean liquids, such
as process lines, simple inexpensive,
relatively accurate, must be kept clean.
(a) Q = flow rate (ftVsec), 0.98 coefficient, A = throat area (ft2), K •* i-^'/d )** where dz = throat diameter (ft),
d: = dia. of inlet pipe (ft), H » differential head (ft of water)
(b) Q,A,K,H as above, C • coefficient, value depends on design of device
-------
3.2.2.2 Flow Measurements^ in Open Pipes
The methods that are employed for measurement of flow in open,
partially filled pipes range from relatively simple calcula-
tions to fairly sophisticated flow sensing procedures that
can signal a sampling device. While the first type is suit-
able for estimating flow for Level I studies, the latter type
is required when one is sampling in proportion to the flow
rate.
In the simple case of a horizontal or sloped pipe in which
the discharge falls into the air, it is possible to calculate
the flow rate by the determination of the vertical and hori-
zontal displacement of the liquid after it leaves the pipe
(Figure 17). In this case the flow is determined by the re-
lation:
Q = 1800AX
/Y
where
Q = the flow rate (gallons/min)
A = wet cross-section area of the liquid in the pipe (ft2)
X = distance between the end of the pipe and a vertical
guage point (ft)
Y = vertical distance from the water surface in the pipe
to the intersection of the water surface with the
vertical guage (ft)
Another method is called the California pipe method. For this
procedure it is necessary to have a horizontal pipe of diameter
103
-------
ADJUSTABLE NUT SO THAT
X AXIS IS PARALLEL TO SEWER
ANDY AXIS IS VERTICAL'
X
\\
Vv
B = (DI STANCE FROM BOTTOM OF PI PE TO.
SURFACE OF FALLING LIQUID)
FOR SLOPED
SEWERS OR PIPES:
CO
+
o
tl
OPEN-PIPE FLOW MEASUREMENT -THIS DEVICE, ADJUSTED TO THE
SLOPE OF A SEWER AND CALIBRATED, CAN THEN BE CLAMPED TO
THE SEWER OUTFALL.
Figure 17. Open pipe flow measurement method9
104
-------
d at least 6 diameters long with free discharge at the open
end (Figure 18). The information required is the diameter
of the pipe "d" and the distance between the top of the pipe
and the level of the liquid "a". The calculation of the
flow rate is given by the equation:
Q = TW
where:
Q = flow rate (gal/min)
T = 3900 (l-a/a)1.88
and W = d2'**8
with both a and d expressed in feet
With this type of measurement, it is possible to employ water
level devices to continuously determine the value of "a". In
this way changes in the flow rate can be recorded, or the
water level device can be employed to operate, through the
proper electrical circuits, a flow proportional sampler.
It is also possible to measure flow from the wet cross sec-
tion area and the average velocity of the liquid because the
flow is equal to the area times the average velocity. The
average velocity can be measured by pitot tubes and other
types of velocity meters. Floating objects can be employed
for measuring the surface velocity and by assuming the aver-
age velocity of the waste stream is 85% of the surface veloc-
ity.
For the most accurate measurements of flow, some type of re-
stricting device such as a weir is constructed across the
105
-------
H
O
ON
OPEN END
HOSE
MEASUREMENTS NEEDED FOR
CALIFORNIA PIPE FLOW METHOD
INCLINED PIPES SHOULD BE CONNECTED TO
A HORIZONTAL LENGTH OF PIPE BY HOSE.
Figure 18. California pipe flow method9
-------
pipe so that the liquid flows through a specific size opening
and the flow is proportional to the depth of liquid at a
given distance upstream from the weir. Most of the devices
used for controlling flow-proportional sample require a weir
or other such device to be installed in order to maintain a
range of depths behind the weir. One major problem with a
weir is that solid in the waste stream tends to accumulate
behind the weir and can influence the measurements. The de-
vice is constructed with a pipe section and thus is considered
as a permanent installation. Plastic weirs which can be in-
stalled as a temporary device in a half-open pipe have re-
cently become commercially available. Typical weirs are of
the 60- or 90-day V-notch or rectangular design, as shown in
Figure 19, and the data required for their use in flow measure-
ment is available in most hydraulics handbooks.
One of the most widely used devices for flow measurement is
the Parshall flume, a specially shaped channel consisting of
a converging section, a throat, and a diverging section, as
shown in Figure 20. As with the weir, data are available to
relate the head of water above the floor in the converging
section to the flow rate. Plastic flumes are available that
can be installed in existing pipes. A major advantage of
the flume is that it is self-cleaning, and suspended materi-
als do not tend to accumulate behind the flume.
3.2.2.3 Control of Flow Proportional Sampling
Many of the samplers discussed in the sampling section can
accept a signal from an external device to permit the col-
lection of flow proportional samples. The signal can be ob-
tained from continuous flow measuring devices that typically
are used in conjunction with a weir or flume. Since the
107
-------
A.
^90°
Z>2.5H
'2.5H
- WATER LEVEL
y//////////////////////////////////////////////^
Figure 19.
Typical V-notch and Rectangular Weirs1}
2 . «4 8 5/2
(A. Q = 2.*\Q H , ; B. Q = 1.^3 H ;
3/2
C. Q = 3.33 LH )
108
-------
STILLING-^
WELLS
SCONVERGING-
F SECTION:
THROAT
SECTION
DIVERGING
"SECTION"
/ WATER SURFACE \
STANDING
WAVE X ir^rn
\
\
V ASLOPE 1/4
Figure 20. Design of Parshall Flume
-------
water level varies just upstream of the weir or flume, a sen-
sor such as a float or pressure-sensitive device can provide
an electrical signal for continuous recording of the liquid
depth. Most of these devices have built-in capability for
converting the flow depth data into a flow rate, and, in ad-
dition, can be set to trigger the sampling device at some
preset liquid level.
There are a large number of recording devices commercially
available and they employ a number of different methods to
determine the liquid level. Details of a number of such de-
vices, using three different approaches, are outlined below,
and they represent the types of equipment available.
ISCO Model 1*170 - This flow meter senses water
levels with a float upstream from a weir .or flume
which has known flow vs head characteristics, and
then integrates the water level with time to mea-
sure the passing volume. The float, of the type
typically used in a water closet, is available to
cover either a six-inch or twelve-inch height
range. The Model 1^70 employs a specially de-
signed calibration disc for each specific flume
or weir shape and this disc must be replaced if
the flow meter is used with a different type of
weir. If required, the manufacturer will supply
special discs for non-standard weirs or flumes
or open pipes. The flow meter can be obtained
with the required interface to trigger the sam-
pler at a set liquid level.
Fisher and Porter Model 10F1275 - This unit em-
ploys a boat-shaped plastic float to sense
liquid level. The basic unit is available in
110
-------
many options and provides only depth informa-
tion. Conversion to flow is accomplished from
tables relating depth to the type of weir or
flume.
ISCO Model 1700W - ISCO will soon release this
flow meter. It employs a different principle
for operation than the Model 1*170. In the new
model, a tube is placed in the liquid upstream
from the flume or weir and air is bubbled slow-
ly through the tube. The flow meter senses the
pressure required to maintain the bubbling and
relates this pressure to the liquid depth. In
contrast to the machined disc used with the
Model 1470, the Model 1700W employs a trans-
parent plastic disc with darkened areas and a
photocell arrangement to convert the depth to
flow rate. A separate plastic disc is required
for each type of weir or flume, however, the
discs are easily changed for a new flow situ-
ation. Although this device has not been field
tested as yet, the prototype demonstrated seemed
to operate reliably. This device can be employed
to trigger a sampler at a preset liquid level.
Manning F-3000 Series - The Manning Dipper Plow-
meter has many features that indicate it would
be very useful in a source assessment type of
program. The unit is relatively small and can
be set up by one man in about 10 minutes. The
flow meter works on a dipper principle and, with
an internal switch, can be used for obtaining
flow in round pipes, flumes or weirs (and other
types of channels on request). The dipper
111
-------
measures the liquid level by sensing the sur-
face of the liquid and does not enter the
liquid. A thin non-corrosive probe is lowered
on a wire controlled by a motor. When the
probe makes contact with the surface, a micro-
ampere circuit is completed through a con-
ductive liquid to a ground return. The signal
reverses the motor, raising the probe above
the surface. After 5 seconds, the probe is
lowered, contacts the surface and reacts again.
The changes in the cable length during this
dipping action are translated into rotation of
a measuring wheel and hence to an electronic
servo system.
The flow-proportional output from the servo
system can be used to record flow on a chart
and provides an input to a totalizing circuit.
In order to obtain the actual flow rate, it
is necessary to multiply the reading over a
time period from the totalizing circuit by
the maximum flow/minute of the system. While
the system does not give the total flow di-
rectly, it is a simple matter to obtain this
data. The unit is battery powered and can
supply flow data for a period up to 7 days on
a chart powered by a spring-driven 8-day
clock. The flow meter includes capability
of controlling an automatic sampler on a flow
proportional basis.
As indicated earlier, sequential or composite sampling can be
done on either a time basis or flow proportional basis. Time-
based samples can give a distorted picture, overemphasizing
112
-------
the nighttime readings when low flow and processing activity
occurs and underemphasizing the higher flow and processing
of the daylight hours. This is not critical in around-the-
clock processing situations, however, most industries show
considerable variation in daily activity. The flow-measuring
devices described have a sensor that can activate a sampler
when the effluent stream reaches a predetermined level.
In general, there are two methods that are employed for taking
the sample, the time constant-volume variable method (TcVv),
and the time variable-volume constant method (TvVc). Many
samplers can operate on either principle, however, the TvVc
method provides the same quantity of liquid each time and
thus eliminates the possibility of collected samples that are
too small for analysis during low flow periods. This is es-
pecially important in sequential sampling. In operation, the
TvVc method provides for shorter intervals between samples
during periods of high flow (and possible high process activ-
ity) and thus may detect occasional non-periodic discharge of
materials that may not be detected on a TcVv basis.
3.2.3 Preservation of Samples
Once the sample is collected, it is essential that it be
maintained for storage and shipment in such a way that it is
representative of the stream to be investigated. Complete
and unequivocal preservation of samples is a practical im-
possibility because complete stability of each and every com-
pound can never be achieved. The preservation techniques
are intended to retard chemical and biological changes prior
to analysis.
113
-------
Some of the potential pitfalls in handling the samples in-
clude:
(1) Breaking of glass containers during han-
dling and shipment
(2) Improper cleaning of glass containers
(3) Use of cap liners other than Teflon (this
material eliminates plasticizer or ad-
hesive contamination)
(4) Biological degradation
(5) Loss of metals and other materials to the
container walls
(6) Possibility of expansion if frozen and
breakage of the container
(7) Potential loss of low molecular compounds
if head space in container is not mini-
mized
(8) Chemical changes in components due to time
delays between sampling and analysis
For this program, our major interest is the organic constitu-
ents in the sample. However, as has been noted many times,
it would be desirable to obtain a sample for other analyses
as well. In order to provide an understanding of the com-
plexity of the problem, recommended sample preservation and
maximum holding period data are given in Table 9 for a
variety of measured parameters.46 This list includes many
of the parameters that are typically measured to characterize
waste water samples. It is obvious that a wide variety of
preservation methods are employed.
-------
Table 9. SAMPLE PRESERVATION AND MAXIMUM HOLDING PERIODS9
Parameter
Preservative
Acidity-Alkalinity
Biochemical Oxygen Demand
Calcium
Chemical Oxygen Demand
Chloride
Color
Cyanide
Dissolved Oxygen
Fluoride
Hardness
Metals, Total
Metals, Dissolved
Nitrogen, Ammonia
Nitrogen, Kjedahl
Nitrogen, Nitrate-Nitrite
Oil and Grease
Organic Carbon
PH
Phenolics
Phosphorous
Solids
Specific Conductance
Sulfate
Sulfide
Threshold Odor
Turbidity
Refrigeration at
Refrigeration at
None required
2 ml H2SOi, per liter
None required
Refrigeration at 4°C
NaOH to pH 10
Determine on-site
None required
None required
5 ml HN03 per liter
Filtrate: 3 ml 1:1 HN03 per liter
40 mg HgCl2 per liter - 4°C
1)0 mg HgCl2 per liter - 4°C
40 mg HgCl2 per liter - 4°C
2 ml H2SOk per liter - 4°C
2 ml H2SOi, per liter (pH 2)
Determine on-site
1.0 g CuSO.,/1 + H3POi» to pH 4.0 - 4°C
i(0 mg HgCl2 per liter - 4°C
None available
None required
Refrigeration at 4°C
2 ml Zn acetate per liter
Refrigeration at 4°C
None available
Maximum
Holding Period
24 hours
6 hours
7 days
7 days
7 days
24 hours
24 hours
No holding
7 days
7 days
6 months
6 months
7 days
Unstable
7 days
24 days
7 days
No holding
24 hours
7 days
7 days
7 days
7 days
7 days
7 days
7 days
115
-------
In general, the preservation methods can be summarized into
four types: (1) use of HgCl2 to prevent bacterial action;
(2) use of acid (either HN03 or H2S04) to prevent bacterial
action, form soluble salts, or to form complexes; (3) use of
alkali to form salts of volatile compound; and (4) refrigera-
tion to prevent bacterial action, to retard chemical reactions,
and to reduce loss of volatile compounds.
For the purpose of this study and the possibility that other
analyses will be done on the collected sample, it is sug-
gested that preservation at 4°C is the best compromise. It
should be noted, however, that this is not the preferred
method for samples to be analyzed for trace metal content.
In addition, filtering the sampling presents an additional
problem. It is generally recommended that samples for or-
ganic analysis not be filtered because the organics are fre-
quently adsorbed on the surface of the suspended particles.
Instead, filtering on-site followed by stabilization of the
filtrate with HNOs is suggested for trace metal analysis.
This will present a major problem if analysis for both types
of materials is desired, especially for Level II or Level III
programs. At this time there is no obvious solution to the
problem. However, since the major interest in this manual
is organic materials, the refrigeration method is suggested
and the loss of some metal compounds must be accepted.
3.3 SOLID SAMPLING PROCEDURES
In the Level I approach to solid sampling, the specified
techniques were basically "grab" samples taken from conven-
ient locations or employed existing sampling devices if they
were available from the plant quality assurance programs.
There was no attempt to install sampling devices since these
116
-------
installations would not fall within the Level I cost and man-
power structure.
For the Level II and Level III program there is increased
emphasis on the acquisition of a representative sample in
order to identify specific organic compounds and to relate
these compounds to the process conditions. In this context,
some of the available types of mechanical samplers can be
employed to obtain the samples. If the analysis information
is essential to the program, the installation of these de-
vices can be justified.
As indicated in the Section 2.3 on Level I sampling, many of
the procedures are available from the ASTM methods. All of
the references listed in the Level I section are appropriate
for the higher level studies.
3.3.1 Sampling of Bulk Material
The methodology for Level II and III sampling is basically
identical to that employed at Level I. The only difference
is that more sampling locations within the source are es-
sential to establish a pattern of variation and trends in
the composition of the material. Along with the larger sam-
ple, mechanical methods of compositing and reducing sampling
size would be employed instead of alternate shovel and cone
and quartering techniques. ASTM procedure D2013-72 "Pre-
paring Coal Sampling for Analysis"22 describes a sample di-
vider (riffle) and several rotating devices to permit sample
reduction. Diagrams of these devices, indicating the princi-
ples of operation, are shown in that procedure. These types
of devices generally eliminate the bias in sample size experi-
enced in the alternate shovel and cone and quartering tech-
niques.
117
-------
The objective of a Level III study is to evaluate the effect
of process variables on the emissions and composition of ma-
terial. This will require that sampling must be performed
on the process material to establish the degree of variabil-
ity to be expected under "normal" operating conditions. Un-
less the degree of variability is known, it would be impossi-
ble to establish if the process variables were causing a
change in the material composition. There is no question as
to the ability of the analytical methods to measure small
variations in composition. Therefore, it is essential that
the sampling should be done in a manner that such variations
represent process changes rather than sampling anomalies. A
number of individual samples should be obtained and analyzed
separately to provide both an average value and the variabil-
ity of the composition instead of trying to obtain an "average"
sample.
3.3.2 Sampling From Moving Streams
The Level I approach for moving streams was based on the
"stop belt" grab sampling procedure. This procedure could
equally apply to Level II and III studies within the restric-
tions of grab sampling techniques. As in bulk sampling, ad-
ditional individual samples would be desirable and these sam-
ples should be subdivided by riffles or other devices to ob-
tain a single sample for analysis.
Level II and Level III programs also present the possibility
of installing mechanical devices such as a ram, rotating cut-
ter or slip streams to provide periodic samples. These de-
vices must be designed for the specific site. As a result,
the device and its design must be discussed in detail with
plant management personnel.
118
-------
The ram or straight line sampler will provide the most accu-
rate sample. The design is such that entire cross-sections
of the material in the stream is removed each time the sampler
operates and an equal time is spent in each portion of the
stream. In general, the sampler operates at right angles to
the stream, although some of these devices operate at an
angle to minimize disturbing the sample and thus reduce the
amount of segregation that can take place.
119
-------
SECTION IV
REFERENCES
1. Dorsey, J. A., C. H. Lockmiiller, L. D. Johnson, and R. M.
Statnick. "Guidelines for Environmental Assessment Sampling
and Analysis Programs Level I." Environmental Protection
Agency, Research Triangle Park, N.C. 9 March 1976.
2. Idem. "Guidelines for Environmental Assessment Sampling
and Analysis Programs - Historical Development and Strategy
of a Phased Approach." Environmental Protection Agency,
Research Triangle Park, N.C. 9 March 1976.
3. TRW Systems Group. "Technical Manual for Process Measure-
ments of Trace Inorganic Materials," (Draft Copy) Environ-
mental Protection Agency. Contract 68-02-1393. February
1975-
4. High Volume Stack Sampler. Aerotherm Division of Acurex
Corporation, Mountain View, California, 9^042.
5. TRW Systems Group. "Fabrication and Calibration of Series
Cyclone Sampling Train." Environmental Protection Agency.
Contract 68-02-1412. Task Order No. 7. April 1975.
6. Aerotherm Division of Acurex Corporation. "Development of
a Source Assessment Sampling System." Environmental Pro-
tection Agency. Contract 68-02-2151*.
121
-------
7. Gushing, K. M. , W. B. Smith, and H. B. Nicholas. Southern
Research Institute. Correspondence to D. B. Harris (EPA).
24 September 1975.
8. Federal Register 3^, No. 199, p. 28758-28760 (1973).
9- "Standard Methods for the Examination of Water and Waste-
water." American Public Health Association, New York, N.Y.,
10019- Thirteenth Edition (1971).
10. "Handbook for Monitoring Industrial Wastewater." Environ-
mental Protection Agency. Technology Transfer Series.
11. Rabosky, J. G., and D. L. Koraido. Chem. Eng. January 8,
1973- P- 111-120.
12. Federal Register ^0, No. 248. p. 59566-59588. (Dec. 24,
1975).
13. Gesser, H. D., A. B. Sparling, A. Chow, and C. W. Turner.
J. Am. Water Works Assoc. 6j>, p. 220-1 (1973).
14. Burnham, A. K., G. V. Culder, J. S. Fritz, G. A. Junk,
H. J. Svec, and R. Virk. J. Am. Water Works Assoc. 65.
p. 722-5 (1973).
15. Mieure, J. P., and M. W. Dietrich. J. Chromatogr. Sci. 11.
p. 559-70 (1973).
16. Andelman, J. B., and S. C. Caruso. Water and Water Pollu-
tion Handbook. Vol. 2. Ed. by Craccio, L. L., Marcel
Dekker, Inc, New York. (1971).
122
-------
17. Jones, P. W., A. P. Graffeo, R. Detrich, P. A. Clarke, and
R. J. Jakobsen. Technical Manual for Analysis of Organic
Materials In Process Streams. EPA 600/2-76-072. U. S.
Environmental Protection Agency, Washington D.C. March,
1976.
18. Standard Method of Sampling Coke for Analysis. ASTM Desig-
nation: D 3^6-35. In: 1973 Annual Book of ASTM Standards.
Philadelphia, American Society for Testing and Materials,
1973.
19- Standard Method for Sampling Packaged Shipments of Carbon
Black. ASTM Designation: D 1799-65. In: 1973 Annual
Book of ASTM Standards. Philadelphia, American Society
for Testing and Materials, 1973.
20. Standard Methods of Sampling Hydraulic Cement. ASTM Desig-
nation: C 183-71. In: 1973 Annual Book of ASTM Standards.
Philadelphia, American Society for Testing and Materials,
1973-
21. Sampling Particulate Ion-Exchange Materials. ASTM Desig-
nation: D 2617-72. In: 1973 Annual Book of ASTM Standards
Philadelphia, American Society for Testing and Materials,
1973.
22. Standard Method of Preparing Coal Samples for Analysis.
ASTM Designation: D 2013-72. In: 1973 Annual Book of
ASTM Standards. Philadelphia, American Society for Test-
ing and Materials, 1973-
23. Probability Sampling of Materials. ASTM Designation:
E 105-58. In: 1973 Annual Book of ASTM Standards. Phil-
adelphia, American Society for Testing and Materials, 1973-
123
-------
24. Acceptance of Evidence Based on the Results of Probability
Sampling. ASTM Designation: E 141-69. In: 1973 Annual
Book of ASTM Standards. Philadelphia, American Society
for Testing and Materials, 1973.
25. Standard Recommended Practice for Choice of Sample Size to
Estimate the Average Quality of a Lot or Process. ASTM
Designation: E 122-72. In: 1973 Annual Book of ASTM
Standards. Philadelphia, American Society for Testing
and Materials, 1973-
26. Standard Method for Collection of a Gross Sample of Coal.
ASTM Designation: D 223^-72. In: 1973 Annual Book of
ASTM Standards. Philadelphia, American Society for Test-
ing and Materials, 1973.
27. Application Information Data. Analytical Instrument Devel-
opment, Inc. Avondale, Pa.
28. Rasmussen, R. A. Amer. Lab. December 1972. p. 55.
29. Trowell, J. M. J. Chromatogr. Sci. £. p. 253 (1971).
30. Neumann, M. G., and S. Morales. J. Chromatogr. 7£.
P- 332 (1972).
31. Gough, T. A., and C. F. Sampson. J. Chromatogr. 68.
p. 31 (1973)-
32. Gearhart, H. L., and M. F. Burke. J. Chromatogr. Sci. 11.
p. 411 (1973).
33. National Institute of Safety and Health. PfCM #127.
-------
34. Mueller, P. X., and J. A. Miller. Amer. Lab. May 1974.
35- Jennings, ¥. G., and H. E. Nursten. Anal. Chem. 39.
p. 521 (1967).
36. Altshuller, A. P. Advan. Chromatogr. 5_. p. 229 (1968).
37. Adams, D. P., R. K. Kappe, and D. M. Jungrath. Tappl 43.
p. 602 (I960).
38. Altshuller, A. P., T. A. Bellar, and C. A. Clemens. Am.
Ind. Hyg. Assoc. J. 23. p. 164 (1962).
39. Jones, P. W., R. D. Glammar, P. E. Strup, and T. B. Stan-
ford. Presented at the 68th Annual Meeting of the Air
Pollution Control Association, Boston, Mass. June 15,
1975.
40. Kaiser, R. E. Anal. Chem. 45. p. 965 (1973).
41. Stanford Research Institute. NIOSH Analytical Methods
for Standards Completion Program. Prepared for National
Institute for Occupational Safety and Health. PB-245-850,
851, 852. PB-245-935. PB-246-148, 149, 150, 151, 152,
153.
42. Bidleman, T. P., and C. E. Olney. "High Volume Collection
of Atmospheric Polychlorinated Biphenyls." Bulletin of
Environmental Contamination and Toxicology, Vol. 11, No.
5. 1974. pp. 442-447-
43. Compton, B., P. P. Bazyalo, and G. Zweig. Field Evaluation
of Methods of Collection and Analysis of Air-borne Pesti-
cides. EPA Contract CPA 70-145. Environmental Protection
Agency, Research Triangle Park, N.C. May 1972
125
-------
44. Harris, D. J., and W. J. Keffer. "Wastewater Sampling
Methodologies and Flow Measuring Techniques." Environmental
Protection Agency, Region VII. EPA 907/9-74-005 (1974).
45. Rabosky, J. G., and D. L. Koraido. Chem. Eng. January 8,
1973. P. 111-120.
46. "Handbook for Monitoring Industrial Wastewater." Environ-
mental Protection Agency. Technology Transfer Series.
126
-------
TABLE OF CONTENTS - APPENDIX
Page
A. General Information 129
A-l. Preliminary Considerations 129
A.1.1 Preliminary Background Search 133
A. 1.2 Preliminary Survey 132
A-2. Planning The Sampling Program 134
A.2.1 Objectives 135
A.2.2 Equipment Requirements 136
A.2.3 Manpower Requirements 136
A.2.4 Job Scheduling 137
A.2.5 Pretest Briefing 138
A. 2.6 Equipment Checkout 139
A.2.7 Logistics 139
Appendix A References 150
B. Fugitive Source Diffusion Model 151
Appendix B References l6l
127
-------
APPENDIX A
GENERAL INFORMATION
The purpose of the manual is to present sampling procedures
that can be employed in the collection of samples of organic
material in gas, liquid or solid process streams and emis-
sion sources.
A.I PRELIMINARY CONSIDERATIONS
The success of any sampling mission depends upon the effort
spent in obtaining in-depth background information on the
processes involved and in a comprehensive preliminary survey.
Table 1 lists some of the industrial processes which could
serve as point sources of organic emissions.1 The emission
characteristics are presented in terms of composition, humid-
ity, acid content, temperature, pressure, and flow rate.
While this table presents only a cursory view of the emission
sources, it can serve as a frame of reference for identifica-
tion of potential problem areas prior to the preliminary sur-
vey. Sources that have reactive emissions (NO , S02, acids,
.A.
oxidizing atmospheres), elevated temperatures, and high water
loadings would have to be approached with caution to assure
that the final analysis were truly indicative of the emission
composition.
129
-------
Table A-l. CHARACTERISTICS OP POTENTIAL ORGANIC EMISSION SOURCES
U)
o
Potential Organic Composition
Emission Sources Part.
Storage Tanks
Unloading Facilities
Chemical Reactors
Non-Catalytic
Catalytic
"luldlzed Ded X
Fixed Bed X
Moving Bed X
Distillation Column
Flash Separator
Filters
Pressure Leaf Y
filters *
Rotary Vacuum _
Filters A
Nutsohe Filters X
Horizontal Plate _
Filters
Tubular Filters X
Bag Filters X
Mixers
Grinders X
Crushers X
Scrubbers
Dryers
Counter-Current ..
Dryer
Rotary Drum ~
Dryer
Vacuum Rotary Dryer X
Spray Dryers
Screeners X
Vacuum Jets
NO,, SOfc CO VC I
X
X
X X
X X
X X
X X
X X
x x
X
X
X
X
X
X X X X
X
X
X
X X X X
x x x x
X X X X
x x x x
X
X
XX X
lumldlty (I RH)
0-20
0-20
0-20
0.20
0-20
0-20
0-98
0-98
0-98
0-98
0-98
0-98
0-98
0-98
0-20
0-20
0-20
80-95
0-95
0-95
0-95
0-95
0-20
95-99
Temperature,
Add Content °F
X
X
X
X
X
X
X
X
X
X
X
X
x
x
X
X
x
x
X
x
-51
-51
-20
0
0
0
80
-18
70
70
70
70
70
70
32
32
32
60
100
100
100
100
32
270
- 300
- 200
- 1000
- 300
- 300
- 300
- 250
- 300
- 150
- 150
- 150
- 150
- 150
- 150
- 90
- 90
- 90
- 150
- 300
- 300
- 300
- 300
- 100
- 390
Pressure ,
0-2
0-2
0-1500
0-50
0-50
0-50
0-50
0-50
0-10
0-80
0-10
0-10
0-10
0-10
0-2
0-2
0-2
0-10
0-20
0-20
0-35
0-20
0-2
25-200
Plow Rate,
scfm
<100
<100
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
>10,000
100-10,000
100-10,000
100-10,000
100-10,000 .
100-10,000
>10,000
-------
Table A-l (Cont'd). CHARACTERISTICS OF POTENTIAL ORGANIC EMISSION SOURCES
potential organic
Emission Sources
Waste Incinerators
Utility Boilers
Pneumatic Conveyors
Conveyor Belts
Extruders
Pellitizers
Paint Spray Booths
Ovens
Blenders
Cyclones
H
^ Extraction Towers
H
Flares
Baggers
Loading Facilities
Cooling Towers
Settling Ponds
Evaporators
Leaching Vat
Cookers
Refrigeration
Machines
Composition
Part. NOX SQx 66 HC Humidity (% RH)
X X X X X
X X X X X
X X
X X
X X X X
X
X X
X X X X X
X X
X X
X
X X X X X
X X
X X
X
X
X
X
X X X X X
10-95
10-95
10-30
0-20
0-20
0-20
0-20
0-50
0-20
0-20
0-90
10-95
0-20
0-20
10-95
10-95
0-95
0-95
0-95
0-10
Temperature,
Acid Content °F
x 500 -
500 -
X 10 -
32 -
100 -
100 -
60 -
500 -
32 -
70 -
X 70 -
1500 -
100 -
x -51 -
32 -
x 32 -
X 100 -
X 100 -
X 100 -
-50 -
1500
1500
90
90
350
200
100
1500
90
150
300
3000
200
300
100
80
200
200
300
32
Pressure,
psig
0-5
0-5
0-20
0-2
0-2
0-2
0-5
0-2
. 0-2
0-20
0-50
0-2
0-2
0-2
0-2
0-2
0-50
0-2
0-2
50-300
Flow Rate,
scfm
>10,000
>10,000
>10,000
<100
<100
<100
>10,000
>10,000
100-10,000
10,000
100-10,000
>10,000
<100
<100
100-10,000
100-10,000
100-10,000
100-10,000
100-10,000
<100
-------
A.1.1 Preliminary Background Search
The first phase of a sampling mission consists of an in-depth
gathering of information relevant to the industry to be sam-
pled. Information is sought which describes the processes
in operation in the industry, the nature of the chemical re-
actions occurring, whether or not the process has been pre-
viously determined as a source of vapor, particulate solid,
or liquid emissions, and the likely emission characteristics
(i.e., information analogous to that presented in Table 1).
Any sampling history of similar industries is important and
can shed light upon troublesome emission sources. Only after
such a literature search should the sampling group elect to
conduct a preliminary survey; in addition, it is strongly
recommended that individuals thoroughly familiar with the
background material and sampling requirements conduct the
survey.
A.1.2 Preliminary Survey
Included in the end of Appendix A is a preliminary survey
form which shows the type of information to be gathered on
the survey.2 In the majority of cases it is expected that a
visit to the location will be necessary. While this form has
been used mainly for stationary source sampling, it includes
much of the information required for multi-media sampling
program, and so data suitable for water and solid sampling
have been added. The type of information required can be
summarized as follows:
I. General Data basic information on the source, in-
cluding the nature of expected pollutants and the process
from which they are emitted.
132
-------
II. Detailed Description concerning the process, in-
cluding the raw materials and fuels. Effects of external
conditions such as weather and operating schedules are also
noted.
III. Sampling Site detailed description of the location
of the site and the geometries which will concern the sam-
pling crew.* Different data are required if the emphasis is
on stack sampling (vapor and particulate), as opposed to
liquid sampling or solid sampling. Note that some prelimi-
nary measurements may be necessary if they are not already
available from the source, particularly on the water content
of gas streams and the flow parameters of gas and liquid
streams. The recommended methods of sampling are to be in-
cluded as are the listings of specific raw materials and
products which may have to be collected to complete the eval-
uation of the source.
IV. Safety Checklist an itemization of needed equipment
and where it is to be obtained. This includes specific
equipment needed to set up the test site for sampling. Of
equal importance is the determination of the needs of per-
sonnel safety equipment so that the sampling team can conduct
the program without exposure to toxic chemicals or undue
physical risk.
V. Plant Entry details specific instructions for accom-
plishing the sampling tasks with minimum friction between
sampling and plant personnel. Plant requirements may include
needs for special insurance coverage, special clothing and
shoes, or instructions in safety rules at that location. The
*The choice of the sampling site will depend on the nature
of the study to be conducted and the end use of the data
generated by the study.
133
-------
inter-corporate agreement section should detail any confi-
dentiality or special clauses that have been negotiated with
regard to testing schedules or data control. It has been
our experience that considerable time is often required to
obtain the cooperation of the plant management in source
evaluation programs. Since many of the compounds of potential
interest are very toxic, some management officials feel that
they have nothing to gain and much to lose if the evaluation
study is conducted. It should be stressed to management that
the plant being studied has the opportunity to obtain valuable
information long before it is required by compliance programs
and thus time can be gained for the development of control
strategies.
VI. Sample Handling itemizes the methods and the loca-
tion of facilities in which the samples can be processed be-
fore shipping or other transfer.
VII. Level of Effort presents the costs to management
regarding the sampling mission.
VIII. Field Test Schedule largely self-explanatory, but
care should be taken to refer to Section II regarding length
of operating day, weather effects, operating week, and the
time requirements for cleanup.
A. 2 PLANNING THE SAMPLING PROGRAM
The background data gathering and the presurvey visit provide
the information required on the industry such as process or
processes of interest, location of sources of samples whether
gaseous, liquid or solid, process conditions that will effect
sampling, and safety considerations. The next step is to
134
-------
plan the sampling program. To do this it is necessary to
define clearly the objectives of the program.
A.2.1 Objectives
To establish the objectives of the sampling program it is
necessary to define why the testing is to be done and what
specific information is required. For the purpose at hand,
it is assumed that a major reason for the sampling and analy-
sis scheme is to determine if the industry represents a seri-
ous toxic emission source for which control strategies may
be required in the future. This implies that samples must be
collected for chemical analysis and often for health effect
/
studies. This objective puts different constraints on the
sampling program than if the objective were standard setting
or compliance sampling.
The objectives of any sampling program must be clearly defined
and understood for a number of reasons. First, the overall
purpose of the test must be known so that the methods to be
used can be determined. Second, the number and kind of sam-
ples required must be identified. Each specific sampling
point must be determined and the kind and number of samples
to be obtained from each identified. Special conditions of
the process that are required for sampling at each point and
the time requirement for sampling must be specified. Third,
the methods to be used to collect samples and data must be
specified and any modification or alterations to these meth-
ods must be pointed out. Finally, the methods for sample
handling and analysis must be given and any deviation from
normal procedures indicated.
135
-------
Once the objectives of the program are outlined, the equip-
ment and manpower needs can be determined. The information
taken at the presurvey can be used, along with the knowledge
of the objectives and requirements of the job, to set the re-
quirements for equipment and manpower.
A.2.2 Equipment Requirements
The task of making up an equipment list should be assigned
to one of the most experienced sampling personnel. He needs
to have a knowledge of the equipment inventory of the sam-
pling organization and must be able to mentally visualize
each sampling action. Equipment lists of an entire inventory
are quite often made up in advance so that necessary items
can be checked as needed for the specific job at hand. The
entire job then must be thought through, taking each sample
at each location into consideration and checking necessary
equipment on the list. At this point, some idea of manpower
and manhours needed will already have been formed and the se-
quence of sampling is being established. Some equipment can
be reused at a different location by scheduling the sampling
at that location later in the program rather than having two
pieces of duplicate equipment in the field. One of the prime
considerations is to reduce as much as possible the equipment
load, while providing sufficient spares so that equipment that
is prone to breakage or malfunction can be quickly replaced
without destroying the continuity of the sampling program.
A.2.3 Manpower Requirements
The list of required manpower is obviously dictated by the
requirements of the sampling program and the scope of the
136
-------
job. Amplifying factors, however, may cause a change in ac-
tual manpower requirement. For instance, a stack or process
duct sampling procedure may require only two men if it is
close to the ground, but additional men are required if the
site is over one hundred feet above the ground. If liquid
samples are to be collected at widely separated locations,
this also may add team members. The two man crew could easily
escalate to four or more due to physical or logistics problems
of the job. Conversely, manpower requirements could be less
than the typical due to the reverse of the adverse factors.
An example is a situation where one man could operate two
simultaneous flow control meter stations by alternating read-
ing times. In any event, each job will have to be evaluated
on its own merits or disadvantages, and a great deal of judg-
ment is required.
The educational level of required personnel is based, obvi-
ously, on an individual's experience, however, general guide-
lines provide for an Engineer as crew chief, an Engineer or
high level technician as Team Leader, and lower level techni-
cians as team members. This personnel mix is, of course,
flexible and can be altered to fit the situation.
A.2.4 Job Scheduling
The job must be scheduled, at least in theory, down to the
last detail. This requires contact with everyone involved
in the sampling program as well as internal sampling and'ana-
lytical personnel. The plants under study must be contacted
to confirm their state of preparedness and to ensure that a
shutdown will not conflict with the program. Officials of
the government, when they are involved, must be notified of
the program dates. Contractors, such as scaffolding erectors
137
-------
or port welders, must be arranged for, if necessary, and must
be contacted in sufficient time prior to the job to do their
work. Internally, analytical departments must be cognizant
of impending sample arrival, and travel and living arrange-
ments must be made for the sampling team members.
A.2.5 Pretest Briefing
An effective means of relaying information to the sampling
crew is a pretest briefing. This is a meeting of all sam-
pling and analytical personnel involved in the program. Dur-
ing the meeting all personnel are given the objectives and
the schedule of the Job, and specific Job assignments are
made. Quite often process data are discussed along with sam-
pling methods, sampling locations, number and kinds of sam-
ples, and sample handling procedures. The equipment list is
discussed and suggestions of the crew are reviewed. This lat-
ter point, the suggestions of the crew, is an important step
and can serve two purposes. First, it can remind the Job
organizer of any forgotten items. Second, it promotes under-
standing of the objectives of the program by the crew. Fi-
nally, the pretest briefing brings all personnel together so
that essential information can be exchanged. For instance,
both analytical and sampling personnel become more familiar
with what the organizer requires, analytical personnel become
more acquainted with the scope of the program and the required
analysis load, and sampling personnel become aware of special
problems in areas such as process requirements and sample
handling.
After the field portion of the project is completed, it is
often beneficial to have a post-test debriefing with the same
personnel to discuss problems that occurred. This meeting
138
-------
Is especially beneficial when, during the course of the job,
the objectives were altered or for some reason certain sam-
ples could not be obtained.
A.2.6 Equipment Checkout
A thorough equipment checkout should be made just before the
crew leaves for the job. During the check, each piece of
equipment should be tested for functionality and calibrated
if required. This procedure should be followed unless a pro-
gram of check, repair, and calibration on all equipment re-
turned from each job has been established. If this latter
procedure is used, all equipment in the storage area will
have been checked prior to storage\
A.2.7 Logistics
One person, usually the crew chief, should be made responsi-
ble for logistics. In this capacity, he will be in complete
control of the movement of all supplies, equipment, and per-
sonnel to and from the job site. His responsibility is to
see that all materials that need to be packed and shipped
will arrive at the job site at the proper time. He makes
sure that the proper equipment to do the job is transported
to or acquired at the job site in time for use according to
the schedule. He is also responsible for the movement of
field personnel to the job site. When the job is completed,
he becomes responsible for moving supplies, samples, equip-
ment, and personnel to the next job location.
139
-------
PRELIMINARY SURVEY DATA SHEETS
I. Name of Company Date of Survey
Address Phone
Name of Contacts
EPA Personnel
Purpose of Test
Phone
Pollutants to be Measured
Process Type
Portion of Process to be Sampled
Description of Control Equipment
-------
II. Process Description
Raw Materials
Fuel Burned _
Products
Operations Performed
Operating Cycle
Check: Batch Continuous Cyclic_
Timing of Batch, or Cycle
Best time to test
Abnormal Conditions Affecting Testing
Weather
Shutdowns
Length of Operating Day
Length of Operating Week_
Other
-------
III. Sampling Site
A. Site Description^
1. Stack
Shape
Diameter
inches Rectangular
Stack length_
Distance before ports
Equlv. Diameters
ft
ft
_Diameters
Diameters
Distance after ports
Device_
ft
Diameters
Wall Thickness
Material
2. Ports
Already Installed^
Obstructions
Device_
inches
Size
Distance to Platforms_
Supports Required
Working Environment
3. Stack Conditions
No. of Traverse Point s_
Stack Velocity
Ambient Temp
Stack Temp_
Stack Temp
Static Pressure _
Moisture Content
Materials Present
inches of water
142
-------
Hazards
Modifications: Extensions
Straightening Vanes_
Scaffolding
Platforms
Other
4. Access to Site
Height above ground ft Distance to Van ft
Stairs? Elevators? Ladder?
Problems
B. Methods to be Employed (circle appropriate method)
EPA 1 2 3 4 5 6 7 8 9 10 11 12 13A 13B
14 15 Asbestos Be Hg.
Other ASTM PTC27 HC Instrumental (specify)
Shell
Others not specified
Raw Materials and Products to be Collected and Frequency
Fuel
Input Materials
Products
143
-------
ESP Dust
Water Samples
Other
-------
IV. Safety Checklist
A. Personnel Protection Equipment (check if required)
Item
Safety glasses
Goggles
Side shields
Face shields
Hard hats
Ear plugs
Safety shoes
Life belt
Ladder climbing
device
Ground fault
interrupt
Grounding clamp.
Plant
MRC
/
/
J
B. Test Site
1. Ladders
General condition_
Cage
Comments
2. Scaffold-platforms
General condition
Toebcards
Comments
Item
Dust masks
Vapor masks
Air purifying
Air supply
Air packs
Chem. res't clothes
Heat res't clothes
Chem. res't gloves
Heat res't gloves
First aid
Fire extinguisher
Plant
MRC
Rest stops
Special belts_
Guardrails_
Screening
3. Smoking restrictions
4. Vehicle traffic rules
-------
5. Evacuation procedures_
6. Alarms
7- Hospital location
Phone
8. Emergency numbers
-------
V. Plant Entry
A. Plant Requirements_
B. MRC Agreement
C. Potential Problems
VI. Sampling Handling
Method
Cleanup
Location
Analysis
Location
Shipment
Where to How
Comments*
* Comments to include need for special bottles (Weaton, etc.)
and other special handling required.
147
-------
VII. Level of Effort
Cost Item Estimated Manhours
1. Planning and Administration
2. Travel
3. Setup and Cleanup
4. Field Testing
5. Laboratory Analysis
6. Report Preparation
7/8. Subcontracting
9- Presurvey
10. Sub-total Manhours
11. Cost per Manhour
12. Cost of Labor (line 10 times
line 11)
13. Other Direct Costs (Itemized)
Total Cost (line 12 plus line 13)
148
-------
VIII. Field Test Schedule
>. Time
Day^X^
Sunday
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
AM
PM
149
-------
APPENDIX A
REFERENCES
1. Anon. "Prioritization of Sources of Air Pollution." Mon-
santo Research Corporation. EPA Contract 68-02-1320. 31
July 197*1.
2. Anon. "Preliminary Survey Form." Developed by Monsanto
Research Corporation for use on EPA Contract 68-02-1404.
150
-------
APPENDIX B
FUGITIVE SOURCE DIFFUSION MODEL
The most frequently used model to describe the behavior of
pollutants emitted from a source point is the Gaussian dif-
fusion model, or Pasquill-Gifford equation. This model can
estimate the concentration of a pollutant at a chosen point
at a given distance downwind from the source for a given at-
mospheric stability and pollutant emission rate. A series
of calculations using various combinations of atmospheric
stabilities, locations, and sampling times will indicate the
range of air quality changes expected to occur around a plant
for each pollutant.
Open source sampling efforts must utilize the diffusion model
in reverse. Normal use is to predict (estimate) concentra-
tions surrounding a point source of known strength. To calcu-
late a non-point (open) source strength, several concentra-
tion readings are taken and the source strength is then de-
termined.
For a number of years estimates of concentrations were calcu-
lated either from the equations of Sutton1 with the atmo-
spheric dispersion parameters C , C , and n, or from the
equations of Bosanquet2 with the dispersion parameters p
and q.
151
-------
The general form of the equation of Pickian diffusion, allow-
ing for any spatial variations in the eddy diffusivities, is:
ix. + ulx + vlx + wlx = _a _
3t 8X 3y 9Z 9X x 8X / 3y \ y 3y 3z \ Z 3z
where x is the concentration, K , K , and K are eddy dif-
x y z
fusivities in x, y, and z directions, respectively, and u, v,
and w are wind components in the x, y, and z directions. Al-
though a concentration distribution is given by the solution
of Equation (1) with certain initial and boundary conditions,
it is difficult to obtain any meaningful solution directly
from Equation (1). The reasons are that the eddy diffusiv-
ities change greatly with stability of the atmosphere, and
that the boundary conditions vary with meteorological con-
ditions and surface conditions of the ground.
A solution which was more useful than the solution of the
Pickian diffusion equation was obtained by Sutton1 by intro-
ducing the statistical theory of turbulent diffusion which
had been suggested by Taylor. 3 In his dispersion formulas
the eddy diffusivities K , K , and K are replaced by the
x y z
empirical parameters C , C , C , and n. For the two-dimen-
x. y z
sional case of Equation (1), Bosanquet obtained a meaningful
solution. This is called K-theory and a further development
of this theory has been carried out by many meteorologists.
The latter type of solution is predominant in most particular
urban air pollution computer models being developed and used.
Hay and Pasquill1* presented experimental evidence that the
vertical distribution of spreading particles from an elevated
point was related to the standard deviation of the wind ele-
vation angle,
-------
incorporating standard deviations of Gaussian distributions:
a for the distribution of material in a plume across wind in
«y
the horizontal direction, and a for the vertical distribu-
tion of material in the plume. These statistics were related
to the standard deviation of wind azimuth angle, a., and wind
elevation angle, OE, calculated from wind measurements made
with a bi-directional wind vane (bivane). Values for these
diffusion parameters, a and a , based on field diffusion
tests were suggested by Cramer, et al.9 Hay and Pasquill
also presented a method for deriving the spread of pollutants
from records of wind fluctuation. Pasquill further proposed
a method for estimating diffusion when such detailed wind data
were not available. This method expressed the height and
angular spread of a diffusing plume in terms of more commonly
observed weather parameters. Suggested curves of height and
angular spread as a function of distance downwind were given
for several "stability" classes. Gifford10 converted Pas-
quill's values of angular spread and height into standard
deviations of plume concentration distribution, a and a .
Pasquill's method, with Gifford's conversion incorporated, is
the commonly accepted calculation method in use today.
Advantages of this system are that (1) only two dispersion
parameters are required, and (2) results of most diffusion
experiments are now being reported in terms of the standard
deviations of plume spread.
In the coordinate system considered here the origin is de-
fined at ground level at or beneath the point of emission,
with the x-axis extending horizontally in the direction of
the mean wind. The y-axis is in the horizontal plane per-
pendicular to the x-axis, and z-axis extends vertically.
The plume travels along or parallel to the x-axis. Figure B-l
illustrates the coordinate system.
153
-------
Ul
4=-
x,-y,Z)
x,-y,0)
Figure B-l. Coordinate system showning Gaussian distributions in
the horizontal and vertical directions.11
-------
The concentration, x> of gas or particles at point x, y, z
from a cjjnt inuous poinjt source with an effective height, H,
is given by Equation (2).
exp
The notation used to depict the concentration is x(x,y,z;H).
H is the height of the plume centerline when it becomes es-
sentially level, and is the sum of the physical stack height h
and the plume rise AH. The following assumptions are made:
the plume spread has a Gaussian distribution in both the hori-
zontal and vertical planes, with standard deviations of plume
concentration distribution in the horizontal and vertical of
a and a , respectively; the mean wind speed affecting the
plume is u; the uniform emission rate of pollutants is Q; and
total reflection of the plume takes place at the earth's sur-
face, i.e., there is no deposition or reaction at the surface.
Any consistent set of units may be used. The most common is
X in g/m3, Q in g/sec, u in m/sec, and a , a , H, x, y, and
z in meters. The concentration x is a- mean over the same time
interval as the time interval for which the o's and u are
representative. The values of both a and a are evaluated
in terms of the downwind distance x, conventionally by graphi-
cal methods. Curves have been fitted to these graphs which
give excellent agreement.
For concentrations calculated for a source of ground level
with no plume rise (H=0), Equation (2) simplifies to:
Q exp I - =• I *-
ire a u
y z
155
-------
Where ground level concentrations (z=0) are to be calculated
further simplification results:
X(x,y,0;0) =
And, when the concentration is to be calculated along the
centerline of the plume (y=0), the equation reduces to:
x(x,0,0;0) = irq Q u (5)
For sampling open sources (those which are not emitted through
a stack or vent), whether it is a single operation or several
located in one defineable area, MRC uses a downwind array of
samplers to determine source strengths or emission rates.
Such an array is shown in Figure B-2. It is felt, and borne
out by experience, that this technique will give at least
three "good" concentration measurements and allows for minor
shifts in wind direction. Ideally, the mean wind direction
will lie along the path to samplers S, and S3, but any shift
will be detected by either S2 or S4. Concentration values
at sampler S0 are subtracted from the others to determine the
effect of the source on downwind concentrations.
For the arrangement shown in Figure B-2, let the origin be de-
fined at the source and all remaining points in the usual
Cartesian coordinate system. Let 6 be the angle of the mean
wind speed. Then to find the value of any point Y. perpen-
dicular to the wind direction centerline, the following compu-
tation is made.
mi = tan 6
156
-------
WIND AZIMUTH
METEOROLOGICAL STATION
Figure B-2. Sampling arrangement
157
-------
For point S. with coordinates (x., y.)
m
m2 - —
Xi
The angle a is found from
a = arc tan
The lateral distance Y. is
Y, = (sin a) /x,* + y,* (6)
and the downwind distance X. is
X± = (cos a) /x±* + y±* (7)
With an open source of some magnitude, the use of the con-
tinuous point source diffusion equation is often insufficient
to describe the emission rate, and other variations on the
Gaussian diffusion model must be utilized. If an open source
has a large width to it, a line source model may better de-
scribe the emission rate.
Concentrations downwind of a continuously emitting infinite
line source, when the wind direction is normal to the line,
can be expressed by rewriting an equation of Sutton.1
2Q
X(x,y,0;H) = —±— exp I - ± ( S_ }' I (8)
where QT is the source strength per unit distance in g/sec-m.
LI
Note that the horizontal dispersion parameter a does not ap-
«7
pear in this equation, since it is assumed that lateral
158
-------
dispersion from one segment of the line is compensated by
dispersion in the opposite direction from adjacent segments.
Also, y does not appear, since concentration at a given x is
the same for any value of y.
Concentrations from infinite line sources when the wind is
not perpendicular to the line can be approximated. If the
angle between wind direction and line source is <}>, Equation
(8) is modified to:
2Q
x(x,y,0;H) = — = exp
(9)
for 0 <. <. 45°C.
In dealing with diffusion of pollutants in an area having a
number of sources, there may be too many sources to consider
each source individually. An approximation can be made by
combining all the emissions in a given area and treating this
area as a source having an initial horizontal standard devia-
tion of plume spread, cr . A virtual distance x can then be
found that will give this standard deviation. Values of x
J
will vary with stability. Then equations for point sources
may be used, determining a as a function of x + x .
«y j
This procedure treats the area source as a cross-wind line
source with an initial normal distribution, a fairly good ap-
proximation for the distribution across an area source. The
initial standard deviation for a square area source can be
approximated by a = S/4.3, where S is the length of a side
c/
of the area. The vertical dispersion coefficient az is
treated in the regular fashion as having originated from the
source rather than a virtual point.
159
-------
The choice of model for source strength calculation depends
on several factors. Sampler locations, personal observations
and engineering judgment predominate. MRC routinely calcu-
lates emission rates by all three models - point, line, and
area source - for open sources since the calculation process
is computerized. This enables the data to be checked for
agreement between samplers, realistic orders of magnitude for
the emission rates, and experiential judgments. Meteorologi-
cal and plant operating conditions may cause an open source
to behave in a different mode on different days.
While the procedure described requires considerable computa-
tion, it does provide a method for collecting particulate
samples for organic analysis with a reasonable field effort.
The equipment requirements of five samplers and a meteorology
station is practical and a team of several persons sampling
for periods up to a week can obtain the required samples.
160
-------
APPENDIX B
REFERENCES
1. Sutton, 0. G. A Theory of Eddy Diffusion in the Atmosphere,
Proc. Roy. Soc., A, 135. p. 1^3-65 (1932).
2. Bosanquet, C. H. and J. L. Pearson. The Spread of Smoke
and Gases from Chimneys. Trans. Faraday Soc., 32. p. 12*49-
1263 (1936).
3. Taylor, G. I. Diffusion by Continuous Movement. Proc.
London Math Soc., 20. p. 196 (1921).
4. Hay, J. S., and F. Pasquill. Diffusion from a Fixed Source
at a Height of a Few Hundred Feet in the Atmosphere. J.
Fluid Mech., 2. p. 299-310 (1957).
5. Cramer, H. E. A Practical Method for Estimating the Dis-
persion of Atmospheric Contaminants. Proc. First Natl.
Conf. on Appl. Meteorol. Amer. Meterol. Soc. (1957).
6. Cramer, H. E., F. A. Record, and H. C. Vaughan. The Study
of the Diffusion of Gases or Aerosols in the Lower Atmo-
sphere. Final Report Contract AF 19(604)-1058 Mass. Inst.
of Tech., Dept. of Meteorol. (1958).
161
-------
7. Cramer, H. E. A Brief Survey of the Meteorological Aspects
of Atmospheric Pollution. Bull. Amer. Meteorol. Soc., 40,
H. p. 165-T1 (1959a).
8. Cramer, H. E. Engineering Estimates of Atmospheric Dis-
persal Capacity. Amer. Ind. Hyg. Assoc. J., 20, 3- p. 183-
9 (1959b).
9. Hay, J. S., and P. Pasquill. Diffusion from a Continuous
Source in Relation to the Spectrum and Scale of Turbulence.
pp. 3*45-65 in Atmospheric Diffusion and Air Pollution.
Edited by F. N. Frenkiel and P. A. Sheppard. Advances in
Geophysics, 6, New York, Academic Press. 471 pp. (1959).
10. Gifford, F. A. Uses of Routine Meteorological Observation
for Estimating Atmospheric Dispersion. Nuclear Safety, 2,
4. p. 47-51 (1961).
11. Turner, D. B. "Workbook of Atmospheric Dispersion Esti-
mates." U. S. Dept. of Health, Education & Welfare, Natl.
Air Pollution Control Administration, Cincinnati, p. 5
(1969).
162
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-122
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
Technical Manual for Process Sampling Strategies
for Organic Materials
5. REPORT DATE
April 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W. Feairheller, P.J.Marn, D.H.Harris, and
D. L. Harris
8. PERFORMING ORGANIZATION REPORT NO.
MRC-DA-512
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
P. O. Box 8 (Station B)
Dayton, Ohio 45407
10. PROGRAM ELEMENT NO.
1AB015; ROAP 21ACX-094
11. CONTRACT/GRANT NO.
68-02-1411, Task 11
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 PERIO
Task Final; 7/75-1/76
IOD COVERED
14. SPONSORING AGENCY CODE
EPA-ORD
is.SUPPLEMENTARY NOTESproject officer for this manual is L.D. Johnson, Mail Drop 62,
Ext 2557. V '
16. ABSTRACT,
The manual describes sampling approaches for conducting Level I, n, and
in environmental source assessment surveys of the feed, product, and waste streams
associated with the production of organic materials. Level I provides large quanti-
ties of sample in a short time period for both analysis of the chemical classes of com-
pounds present and biological testing programs. Level II is a more detailed qualita-
tive and quantitative chemical analysis of the organic components. Level HI is a
quantitative study of the effect of process variables on the emission rates of specific
organic materials. The manual: provides specific methods to be used in Level I to
obtain samples from stationary sources, fugitive emission sources, and process
and waste streams (including gas, liquid, and solid phases); and provides the current
state-of-the-art, an extension of the state-of-the-art sampling metjods that are
available for application to Level n and m studies. The manual is directed to those
who are basically experienced in sampling techniques and will be required to apply
these methods in source assessment programs.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
ir Pollution
Sampling
Organic Compounds
Qualitative Analysis
Quantitative Analysis
dustrial Processes
,
ia.
Air Pollution Control
Stationary Sources
Environmental Assess-
ment
Fugitive Emissions
Process Streams
13B
14B
07C
07D
13H
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
172
2O. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (3-73)
163
------- |