An error occurred while trying to OCR this image.
-------
United States
Environmental Protection
Agency
Air Pollution Training Institute
MD20
Environmental Research Center
Research Triangle Park NC 27711
EPA 450/2-80-004
September 1980
Air
APTI
Course 435
Atmospheric Sampling
Student Manual
Prepared By:
M. L.Wilson
D. F. Elias
R. C. Jordan
O. G. Durham
Revised By:
K.C. Joerger
B.M. Ray
Northrop Services, Inc.
P.O. Box 12313
Research Triangle Park, NC 27709
Under Contract No.
68-02-2374
EPA Project Officer
R.E. Townsend
United States Environmental Protection Agency
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
US Environmental Protection Agenc*
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12tn
Chicago. II 60604-3590
-------
Notice
This is not an official policy and standards document. The opinions findings, and
conclusions are those of the authors and not necessarily those of the Environmenta
Protection Agency. Every attempt has been made to represent the present state of
the art as well as subject areas still under evaluation. Any mention of products or
organizations does not constitute endorsement by the United States Environmental
Protection Agency.
Availability of Copies of This Document
This document is issued by the Manpower and Technical Information Branch Con-
trol Programs Development Division, Office of Air Quahty Planning andI Stand,ds,
USEPA It was developed for use in training courses presented by the EPA Air Pollu-
tion Training Institute and others receiving contractual or grant support from the
Institute. Other organizations are welcome to use the document for trammg purposes.
Schools or governmental air pollution control agencies establishing training programs
mav receive single copies of this document, free of charge, from the Air Pollution
Tuning Institute, USEPA, MD-20, Research Triangle Par,, NC 27711. Others ^
obtain copies, for a fee, from the National Technical Information Service, 5825 Port
Royal Road, Springfield, VA 22161.
n
-------
•
s
°
POLLUTION TRAINING INSTITUTE
MANPOWER AND TECHNICAL INFORMATION BRANCH
CONTROL PROGRAMS DEVELOPMENT DIVISION
OFFICE OF AIR QUALITY PLANNING AND STANDARDS
I7»e .4ir Pollution Training Institute (1) conducts training for personnel working on the
development and improvement of state, and local governmental, and EPA air pollution control
programs, as well as for personnel in industry and academic institutions; (2) provides consulta-
tion and other training assistance to governmental agencies, educational institutions, industrial
organizations, and others engaged in air pollution training activities; and (3) promotes the
development and improvement of air pollution training programs in educational institutions
and state, regional, and local governmental air pollution control agencies. Much of the pro-
gram is now conducted by an on-site contractor, Northrop Services, Inc.
One of the principal mechanisms utilized to meet the Institute's goals is the intensive short term
technical training course. A full-time professional staff is responsible for the design, develop-
ment, and presentation of these courses. In addition the services of scientists, engineers, and
specialists from other EPA programs, governmental agencies, industries, and universities are
used to augment and reinforce the Institute staff in the development and presentation of
technical material.
Individual course objectives and desired learning outcomes are delineated to meet specific pro-
gram needs through training. Subject matter areas covered include air pollution source studies,
atmospheric dispersion, and air quality management. These courses are presented in the
Institute's resident classrooms and laboratories and at various field locations.
R. Alan Schueler
Program Manager
Northrop Services, Inc.
a.
James A. Jahnae
Technical Director
Northrop Services, Inc.
Jean jLochueneman
Chief, Manpower fr Technical
Information Branch
ill
-------
Table of Contents
Page
Chapter 1. An Introduction to Atmospheric Sampling .................... 1-
Objectives of Air Monitoring.
Sampling Train Design.
1- 1
1- 3
Chapter 2. Basic Gas Properties and Mathematical Manipulations .......... 2- 1
Temperature ........... : ........................................
Pressure [[[ "
Ideal Gas Laws .................................................. ^
Gas Density [[[ ^~14
Standard Conditions for Atmospheric Sampling ....................... z'lb
Reynold's Number ............................................... ^-17
Summary of Useful Equations ...................................... 2-20
References [[[ "
Units of Measurement
References
Definitions [[[
Chapter 3. Air Measuring Instruments ................................. 3~ |
Introduction to Air Movers ........................................ 3- 1
Air Mover Selection Criteria ....................................... 3" i
Pumps [[[ J- b
Ejectors [[[ 6'IA
Liquid Displacement ............................................. •* ™
Evacuated Flasks ................................................. *~l*
Flow Rate Control ................................................ ^-15
Flow Rate for Sampling ........................................... f~|°
Summary [[[ "
References [[[ J- °
Air Measuring Instruments ........................................ £ ^
Calibration [[[ ^"
Volume Meters .................................................. ~~^*
Rate Meters [[[ ''Jj
Variable Area Meters ............................................. ;>^°
Velocity Meters .................................................. jj"4"
Summary [[[ "
Sample Problems ................................................ *~™
References [[[ ^"49
Chapter 4. Particulate Sampling ....................................... 4' |
Introduction [[[
Principles of Inertial Collection .....................................
Types of Inertial Sampling Devices .................................. 4' 5
-------
Page
Chapter 5. Gaseous Sampling 5- 1
Principles of Absorption 5- 1
Determination of Collection Efficiency 5- 4
References 5- 9
Principles of Adsorption 5-10
References 5-22
Selection and Performance of Wet Collector Media 5-23
References 5-29
Principles of Grab Sampling 5-30
References 5-36
Principles of Freezeout Sampling 5-37
References 5-43
Chapter 6. Generation of Standard Test Atmospheres 6- 1
Introduction 6" 1
Static Systems 6" 1
fi 7
Dynamic Systems °" '
References 6"19
Preparation of Zero Air 6-20
References °~3*
Chapter 7. Standard Methods for Criteria Pollutants 7- 1
Introduction '" 1
References 7'10
Chapter 8. Continuous Air Monitoring Instrumentation 8- 1
Introduction 8" 1
Coulometric Instruments 8- 2
Second Derivative Spectroscopic Instruments 8-5
Flame Photometric Instruments 8- 7
Fluorescence Instruments 8- 9
Chemiluminescence Instruments 8-11
Ultraviolet Photometric Instruments 8-14
NDIR Instruments 8-16
References 8'20
Chapter 9. Design of Surveillance Networks 9- 1
Elements of Surveillance Networks 9- 1
References °" °
Chapter 10. Statistical Techniques Employed in Atmospheric Sampling 10- 1
Introduction *"" *
Data Plotting 10' *
Least Squares Linear Regression 10-9
Measures of Central Tendancy 10-11
Geometric Mean 10-12
Measures of Dispersion 10-13
Distribution Curves 10-15
Lognormal Distributions 10-18
References and Additional Reading 10-19
-------
Page
Appendix 1. Theory and Calibration Procedures for a Rotameter Al- 1
Nomenclature Al- 2
Description of a Rotameter Al - 2
Development of Flow Equations
Common Practices in the Use of a Rotameter for Gas Flow Measurement. . Al - 6
Appendix 2. Federal Register Reference Methods. ........ -... • • .•••••• • A2'*
A-Reference Method for the Determination of Sulfur Dioxide in the
Atmosphere (Pararosaniline Method) •
B- Reference Method for the Determination of Suspended Particulates
in the Atmosphere (High Volume Method)
C-Measurement Principle and Calibration Procedure for the
Continuous Measurement of Carbon Monoxide m the Atmosphere
(Non-Dispersive Infrared Spectrometry)
D- Measurement Principle and Calibration Procedure for the
Measurement of Ozone in the Atmosphere J " ' '
E-Reference Method for Determination of Hydrocarbons Corrected
, A/" 1 o
for Methane • '
F- Measurement Principle and Calibration Procedure for the
Measurement of Nitrogen Dioxide in the Atmosphere (Gas Phase ^ ^
Chemiluminescence) • • • • ' ,
G-Reference Method for the Determination of Lead in Suspended
Particulate Matter Collected from Ambient Air A/-Z3
Appendix 3. Conversion Factors
and Useful Information A3- 1
Vll
-------
Figures
Figure
Page
1 - 1 Significant harm levels established by the
Environmental Protection Agency 1 - 2
1 - 2 Typical sampling train 1' 3
2- l(a) Temperature relationships 2- 2
2- l(b) Relationship of the absolute temperature systems 2- 2
2- 2 The manometer and mercurial barometer 2- 4
2- 3 Fortin barometer 2"
2- 4 Blow-up of Fortin barometer 2~ 5
2- 5 Blow-up of Vernier scale 2" 6
2- 6 Aneroid barometer 2" '
2- 7 Mechanical pressure transducer 2~ 8
2- 8 Electrical pressure transducer 2~ 8
2- 9 Absolute-atmosphere-gage pressure relationship 2'10
2-10 Velocity gradient 2"16
2-11 Viscosity nomograph for air 2'18
2-12 Viscosity nomograph of various gases at 1 atmosphere 2-19
3-1 A sampling train 3"
3- 2 Pressure profile for basic sampling apparatus 3- 4
3- 3 Positive displacement pump 3' 5
3- 4 Classification of positive displacement pumps 3- 6
3- 5 Centrifugal pump 3" ^
3- 6 Piston pump 3~
3- 7 Diaphragm pump 3' 8
3- 8 Centrifugal pump operation 3- 8
3- 9 Characteristic curve for a positive displacement pump 3-10
3-10 Characteristic curve for a centrifugal pump 3-10
3-11 Pump comparison 3-11
3-12 Pressure-flow relationship for metal bellows pumps for vacuum.... 3-12
3-13 Ejector operation 3~13
3-14 Liquid displacement 3-14
3-15 Flow rate control by diversion 3-15
3-16 Flow rate control mechanism 3-16
3-17 Spirometer 3'2^
3-18 Orthographic and cross-sectional views of a 5-ft3 spirometer 3-22
3-19 Water displacement bottles 3-23
3-20 Soap-bubble meter • • 3-24
3-21 Soap-bubble meters (a) with bubble breaker and capability
of handling vacuum at (2) or pressure at (1); (b) one
capable of handling only pressure at (3) 3-25
3-22 Mercury-sealed piston volume meter 3-27
3-23 Calibrator console, front view 3-28
3-24 PVC piston for mercury-sealed piston volume meter 3-28
3-25 Wet test meter 3"29
3-26 Setup for calibrating a wet test meter against a spirometer 3-30
3-27 Calibration of wet test meter (WTM) against a
mercury-sealed piston 3-30
3-28 Calibration of wet test meter with displacement bottle 3-31
3-29 Principle of gas flow through the roots meter 3-32
3-30 Dry gas meter 3'3*
3-31 Dry test meter *'|™
3-32 Working mechanism of dry test meter 3'35
3-33 Orifice meter 3~36
Vill
-------
Figure PaSe
3-34 Typical orifice meter calibration curve 3-37
3-35 Venturi meter 3-38
3-36 Rotameter 3-39
3-37 Gas stagnation against an object 3-41
3-38 Components of total pressure 3-41
3-39 Standard pitot tube 3-43
3-40 S-type pitot tube 3-44
3-41 Problem 4: Pump capacity 3-48
4- l Respiratory collection of particulates 4- 2
4- 2 Bimodal distribution of particles in the atmosphere 4- 2
4- 3 Characteristics of particles and particle dispersoids 4- 3
4- 4 Paniculate collection by impaction 4- 4
4- 5 Annular impactor 4- 6
4- 6 Cascade impactor schematic diagram 4- 7
4- 7 Capture efficiencies of a cascade impactor 4-8
4- 8 Andersen sampler schematic diagram 4- 9
4- 9 Andersen sampler 4-10
4-10 Collection efficiency of Andersen sampler 4-11
4-11 Diagram of modified Andersen impactor sampler and shelter 4-12
4-12 Collection efficiency of modified Andersen sampler (3 ftVmin). . . . 4-12
4-13a Cross-sectional view of hi-vol Andersen impactor 4-13
4-13b Expanded view of hi-vol Andersen impactor 4-13
4-14 Collection efficiency of hi-vol Andersen impactor (20 ftVmin) 4-14
4-15 Multi-slit high-volume cascade impactor 4-15
4-16 Penetration vs. particle size 4-16
4-17 Lundgren type inertial collector 4-17
4-18 Cyclone sampler 4-18
4-19 Air centrifuge 4-19
4-20 Schematic diagram of virtual impactor —critical
impaction parameters are noted 4-20
4-21 Diagram of a dichotomous sampler 4-21
4-22 Aerosol inlet for dichotomous sampler 4-22
4-23 Dichotomous sampler 4-23
4-24 Expanded view of dichotomous sampler 4-23
4-25 Flow schematic of control module 4-24
4-26 Efficiency of collection as a function of size 4-26
4-27 Respirable percentages of different particle sizes 4-30
4-28 Respirable dust sampler 4-30
4-29 Particle shattering 4-31
4-30 Hi-vol sampler 4-37
4-31a Hi-vol sampler with shelter 4-38
4-31b Air flow of hi-vol sampler in shelter 4-38
4-32 The effect of relative humidity on the weight of glass fiber
filters at 75°F 4-39
4-33 The effect of relative humidity on the weight of atmospheric
particulates at 75 °F 4-40
4-34 Hi-vol cartridge assembly 4-42
4-35 Summary of hi-vol filter handling procedures 4-43
4-36 Orifice calibration unit 4-44
4-37 Diagram of orifice calibration set-up 4-46
4-38(a) Orifice calibration curve 4-47
4-38(b) Visifloat calibration curve 4-47
4-39 Hi-vol setup with flow transducer 4-51
-------
Figure PaSe
4-40 Hi-vol sampler with pressure taps 4-52
4-41 Respirable retention vs. particle diameter 4-54
4-42 Respirable particulate matter curves for a
polyurethane foam collector 4-55
4-43 Suspended particulate matter ratios on a concentration basis 4-56
4-44 Respirable/total particulate matter ratios for selected pollutants. . . 4-56
4-45 Mass and percentage composition of size-fractionated St. Louis
aerosol samples from August 18 to September 7, 1975 4-58
4-46 Seasonal trends in concentrations of suspended
particulate matter 4-58
4-47 Long range trends in concentrations of suspended
particulate matter 4-59
4-48 Filtration mechanisms 4-67
4-49 Cellulose fiber filters 4-70
4-50 Glass fiber filter characteristics 4-72
4-51 Mixed fiber filter characteristics 4-73
4-52 Membrane filter media 4-74
4-53 Filter manufacturers code 4-75
4-54 Membrane filter solubility characteristics 4-75
4-55 Effect of glass fiber pH on concentration observed with 24-hr.
standard high volume sampling in Anderson, CA,
July-Aug. 1972 4-76
4 56 Initial filtration efficiency and flow behavior of the Lundgren
impactor as a function of the after filter media 4-77
4-57 Initial efficiency of filtration of particle >0.3/t diameter via
Royco particle counter 4-79
4-58 Filter impurity levels (ng/cm2) for various filters 4-79
5- 1 Solubility of selected gases in distilled water at 20 °C 5- 2
5- 2 Absorption device adapted from an Erlenmeyer flask 5- 5
5- 3 Typical fritted-glass absorbers 5- 6
5- 4 Absorption sampling devices 5- 7
5- 5 Two types of impingers 5- 8
5- 6 Gas adsorption isotherms 5-13
5- 7 Adsorption of gases on one gram of charcoal at 15°C 5-14
5- 8 Typical surface areas of adsorbents 5-16
5- 9 Assembled sampler and shelter with exploded view of the
filter holder 5'19
5-10 High speed organic vapor collector 5-20
5-11 Desorption of pollutants from aTenax-GC cartridge 5-20
5-12 Dynamic enrichment on adsorption column 5-21
5-13 Absorption coefficient of gases at 20°C 5-23
5-14 Influence of temperature on solubilities of gases in water 5-24
5-15 Influence of pressure on solubility of CO2 in various solvents
at-59°C 5'24
5-16 Performance curves — commercially available absorbers 5-26
5-17 Ideal and observed solubilities at 20°C 5-28
5-18 Vacuum tube 5~30
5-19 Vacuum flask 5"31
5-20 Gas-displacement collector 5-32
5-21 Liquid displacement collector 5-33
5-22 Aspirator bottle 5-34
5-23 Inflation sampler 5-35
5-24 Freezeout unit 5"3'
5-25 Bath solutions 5"38
5-26 Freezeout equipment for atmospheric samples
(horizontal sampling train) 5-39
X
-------
Figure Pa§e
5-27 Freezeout equipment for atmospheric samples
(vertical sampling train) 5'39
5-28 Freezeout unit showing packing material 5-40
6- 1 Rigid chamber used for producing standard gas mixture 6-4
6- 2 Rate of decay of SO2 concentration in a bag 6- 6
6- 3 Carbon monoxide sample deterioration with time in bags
of various materials "' '
6- 4 Calibration of two permeation tubes 6- 9
6- 5 Permeation rate vs. temperature for four gases 6-10
6- 6 Components and flow of a typical permeation system 6-10
6- 7 Permeation rates of some typical compounds through FEP Teflon. . 6-11
6- 8 Some materials used to construct a permeation tube 6-11
6- 9 Nitrogen dioxide permeation device 6-12
6-10 SO2 permeation tube 6'12
6-11 Gravimetric calibration apparatus 6-!3
6-12 Typical strip chart readout from an in situ gravimetric apparatus. . 6-14
6-13 Single dilution system 6-15
6-14 Sketch of a system for making double dilutions 6-16
6-15 Method for injecting liquids and gases into a test atmosphere 6-17
6-16 Method for injecting liquids into a test atmosphere 6-18
6-17 Method for injecting liquids into a test atmosphere 6-18
6-18 Typical composition of clean, dry air near sea level 6-20
6-19 Materials used in producing zero air up to SOf/min 6-22
6-20 Typical properties of adsorbents 6"23
6-21 Adsorption of gases by carbon 6-24
6-22 Molecular sieve adsorption characteristics 6-24
6-23 Molecular sieves —Linde type 6-25
6-24 Effective sorption capacities of molecular sieves 6-25
6-25 Comparative efficiency of various drying agents 6-26
6-26 Comparative efficiencies and capacities of various solid
desiccants in drying a stream of nitrogen 6-27
6-27 Summary of cold bath solutions 6-29
7. 1 National ambient air quality standards 7- 2
7- 2 Performance specifications for automated methods 7- 3
7- 3 Pararosaniline interferences '- 4
7- 4 Manual SO2 sampling train for 30 minute to 1 hour sampling 7-4
7- 5 Assembled sampler and shelter 7- 6
7 - 6 Orifice calibration unit '" '
7- 7 Schematic diagram of a typical BAKI calibration system 7- 8
7- 8 Schematic diagram of a typical UV photometric
calibration system '" 9
8- 1 Coulometric titration of iodine 8- 3
8- 2 Coulometric titration of bromine 8-4
8- 3 Schematic of optics employed in a second derivative
spectrometer 8- 6
8- 4 Flame photometric detector 8- 7
8- 5 Energy levels in fluorescence emission 8-10
8- 6 Compound specific chemiluminescence detector 8-12
8- 7 Emission spectrum of ozone/nitric oxide 8-13
g- 8 Ultraviolet absorption of ozone 8-15
8- 9 Flow and components of ozone detector using
UV photometry 8-16
-------
Figure PaSe
8-10 Components and gas flow of an NDIR instrument 8-17
8-11 Overlap of carbon monoxide and broad band absorption curves. . . 8-18
8-12 "Negative filter" analyzer 8-18
8-13 Absorption in the detector of a "negative filter"
NDIR analyzer 8-19
10- 1 SO2 levels 10-2
10- 2 Frequency table 10-2
10- 3 Pollution concentration frequency polygon 10- 3
10- 4 Pollutant concentration histogram or frequency
distribution curve 10-4
10- 5 Histogram of percent frequency distribution curve 10- 4
10- 6 Cumulative frequency table 10- 5
10- 7 Cumulative frequency distribution curve 10- 5
10- 8 Relative frequency distribution showing: Curve A and B both
centrally located; Curve B being more disperse than Curve A
and the skewness of Curve C 10- 6
10- 9 Logarithmic transformation 10- 7
10-10 Logarithmic frequency table 10- 7
10-11 Normalized data plot vs. non-transformed data 10- 8
10-12 Linear regression curve 10- 9
10-13 Example of nonsymmetrical distribution of data
(median vs. mean) 10-12
10-14 Dispersion characteristic curves 10-13
10-15 Gaussian distribution curve "normal curve" 10-16
10-16 Characteristics of the Gaussian distribution 10-16
10-17 Frequency vs. concentration of SO2 10-18
Al-1 Rotameter Al- 2
A1-2 Forces acting upon a rotameter float Al- 3
Al-3 Test setup for calibrating a rotameter Al- 6
A1-4 Rotameter calibration curve Al- 7
A1-5 Arrangement of sampling components Al- 7
A1-6 Family of rotameter calibration curves Al- 8
Al-7 A universal calibration curve for a rotameter Al- 9
A1-8 Predicting calibration curves from the universal calibration curve. Al-11
Al-9 Calibration curves predicted from universal calibration curve Al-12
xil
-------
Chapter 1
An Introduction
to Atmospheric Sampling
Objectives of Air Monitoring
National Ambient Air Quality Standards (NAAQS) fur sulfur dioxide, total
suspended particulates (TSP), carbon monoxide, ozone, nitrogen dioxide,
nonmethane hydrocarbons, and lead have been established under the authority of
the Clean Air Act (as amended August, 1977). There are two types of ambient air
quality standards: primary standards and secondary standards. As stated in Section
109 of the Clean Air Act, primary standards are those, "the attainment and
maintenance of which are requisite to protect the public health." The section fur-
ther defines a secondary ambient air quality standard as, "a level of air quality
attainment and maintenance of which is requisite to protect the public welfare
fro -i any known or anticipated adverse effects associated with the presence of such
air pollutants in the ambient air." (1)
These primary and secondary ambient air quality standards must be achieved
throughout the United States and its possessions. In order to meet the standards,
states are required to develop and implement air pollution control strategies
through the mechanism of State Implementation Plans (SIPs). The implementation
plans may contain control strategies such as industrial and urban zoning, the
development or expansion of mass transit systems and vehicle inspection and
maintenance programs, and the establishment of stationary source emissions stan-
dards for specific industrial categories. Ambient ai monitoring plays a vital role in
the development and evaluation of these control st aiegies. During the development
phase, air quality data is used to determine if an a <>a is attaining or not attaining
the air quality standards. This determination is cr ical for the proper designing of
control strategies for the area in question because ittainment areas usually require
less stringent control measures than nonattainmen areas. Furthermore, air quality
data may be used to generate or validate compute, models of air pollution disper-
sion which are then used in the development of control strategies. After the
implementation plans containing their various control strategies have been put into
force, further ambient air quality monitoring is required in both attainment and
nonattaiment areas. For areas that have achieved attainment, further monitoring is
necessary to assure that attainment is maintained. Additional monitoring is
required in nonattainment areas for evaluating progress towards reaching
attainment. (2)
In addition to the requirement that the primary and secondary National
Ambient Air Quality Standards be achieved and maintained throughout the
country, the Clean Air Act also stipulates that no significant deterioration of
1-1
-------
existing air quality will be allowed in any portion of any state. In order to comply
with this provision, it is necessary to determine the impact on the already existing
air quality of an area by a planned new emissions source. The estimated emissions
contributed by the new source must be compared to the already existing air quality
to ascertain whether the new source would significantly deteriorate the present air
quality. Ambient air quality monitoring data is used to establish the preexisting air
quality of the area in question.
It should be recognized that the overall goal of an air quality monitoring net-
work is the protection of human health and welfare. The aforementioned moni-
toring objectives should assure the attainment of this goal under ordinary cir-
cumstances. However, abnormal meteorological conditions, such as temperature
inversions, which cause poor air pollutant dispersion may result in the formation of
pollutant levels that could cause significant harm to public health. These signifi-
cant harm levels are listed in Figure 1-1. Air pollution emergency episode plans
are utilized to prevent pollutant levels from reaching concentrations which would
cause significant harm to the health of persons. The plans specify incremental
reductions of source emissions based on up-to-date ambient air monitoring data. (3)
Pollutant
SO2
Paniculate matter (TSP)
SO2x paniculate
CO
Ozone
NO2
Concentration
2620
1.0
1000
490 XlO3
57.5
50
86.3
75
144
125
1200
0.6
3750
2.0
938
0.5
Units
/ig/m3
ppm
Mg/m3
0«/m')«
rng/m3
ppm
mg/m3
ppm
mg/m3
ppm
/ig/m3
ppm
Mg/m3
ppm
/tg/m3
ppm
Averaging time
24 hr.
24 hr.
24 hr.
8 hr.
4 hr.
1 hr.
1 hr.
1 hr.
24 hr.
Figure 1-1. Significant harm levels established by the Environmental Protection Agency.
1-2
-------
Sampling Train Design
Most atmospheric sampling techniques make use of a sampling train similar to the
block diagram in Figure 1-2. Air containing the pollutant of interest enters the
sampling train and passes through a sample collection device. This device either
physically or chemically removes the pollutant from the air stream, holds the
polluted air for later analysis, or allows the pollutant to be analyzed simultaneously
with the collection. Ideally, only the pollutant of interest is collected. This is
seldom the case, however; hence, interferences must be considered when
measurements are made.
Many sampling techniques use collection devices that remove the pollutant from
die air for later analysis. Wet chemical methods, such as the pararosaniline method
for sulfur dioxide (SO2), remove the pollutant from the air and hold the pollutant
by means of a chemical reaction for later analysis. In the pararosaniline method,
the sample collection device is a bubbler containing an absorbing reagent. The
High Volume method for total suspended particulates (TSP) uses a filter as the
sample collection device. In either case, the pollutant is held by the collector for
analysis by a contaminant detector. The contaminant detector in the pararosaniline
method is analysis by colorimetry; for TSP High Volume sampling, it is a
gravimetric balance.
Contaminant
detector
Air in •
Sample
collection
device
A
_ .
Air mover
A
*
Flow
measuring
device
^MMH^H
A
i
i
i
i
i
- —
• Air out
Figure 1-2. Typical sampling train.
1-3
-------
Instrumental methods such as chemiluminescence for nitrogen dioxide (NO2) and
ozone (O3) combine the sample collector and the contaminant detector into one
device.
Some methods of air sampling are used that collect a volume of polluted air for
later analysis. These methods usually make use of grab or integrated sampling
using containers made of stainless steel, glass, pliable plastic or simple hypodermic-
syringes. Many factors are important in considering what material should be used
when grab sampling (4); many uses have been found for plastic bags. In this type
of sampling, the plastic bag, syringe, stainless steel or glass container, is the sample
collection device. Contaminant detection is usually by instrumental analysis.
Mechanisms used to move air through the sample collection device and measure
the quantity of air are integral parts of sampling trains. Air movers are usually
motor driven pumps. When motor driven pumps are not practical, ejectors.
displacement methods, and evacuated flasks can be used. Some typical flow
measuring (or controlling) devices for sampling are: rotameters, mass flow meters,
and critical and subcritical orifices. It is often necessary to determine how long an
air mover has pulled a certain flow rate to determine the volume sampled
(uolume — flow rate X time); therefore, the time sampled is data that is recorded.
Placement oi the flow measuring device is dependent on what device is used. Air
movers and flow measuring devices are usually placed after sample collection
devices to avoid contamination of the air stream.
Materials used in a sampling train (at least to as far as the collector) must be
sufficiently inert to the pollutant of interest so as not to interfere with collection. If
the air mover and flow measuring device must be placed before the collection step,
then parts of those devices contacting the air stream must be inert. Glass,
Teflon"', and stainless steel are generally considered to be nonreactive. Hence,
these materials have been used extensively in sampling train construction. However,
the materials listed as being generally nonreactive can become reactive if used to
sample the wrong environment. For instance, glass could not be used if a sampling
train was being built to monitor hydrofluoric acid. Even if a material is considered
nonreactive, significant wall loss can occur if sampling lines are too long.
Interference with the measurement of an air pollutant is not the only considera-
tion important to the selection of sampling train materials. Care must be taken to
protect sampling train components from damage caused by the sampled air or pro-
ducts of the measurement system. If a rubber diaphragm pump is to be used in
conjunction with an analyzer that measures nitrogen oxides by the reaction of
nitric oxide (NO) with ozone (O3), then a charcoal filter must be used before
the pump to remove excess ozone. If ozone is allowed to contact the rubber
diaphragm of the pump, the rubber will deteriorate. Other sampling train com-
ponents that usually need protection are rotameters and small orifice meters. In
this case, a filter and moisture trap are placed in front of the orifice or rotameter
to prevent them from becoming clogged.
1-4
-------
References
Note: Some references may not be referred to in the text. They are supplied for
additional reference.
1. Title 40, Code of Federal Regulations, Part 50, pp. 4-6. July 1, 1979.
2. Air Monitoring Strategy for State Implementation Plans. EPA-450/2-77-010.
June 1977.
3. Title 40, Code of Federal Regulations, Part 51, pp. 53-151. July 1, 1979.
4. Schuette, F. J. Plastic Bags for Collection of Gas Samples. Atmospheric
Environment 1:515-519, 1967.
1-5
-------
Chapter 2
Basic Gas Properties and
Mathematical Manipulations
Temperature
The Fahrenheit and Celsius Scales
The ranee of units on the Fahrenheit scale between the freezing and boiling point
of water at one atmosphere (atm) pressure is 180 (212°F-32°F = 180°F); on the
Celsius or Centigrade scale, the range is 100 (100°C-0°C = 100°C). Therefore,
each Celsius degree is equal to 9/5 or 1.8 Fahrenheit degrees. To be able to con-
vert from one system to the other, the following equations can be used:
(Eq.2-1) °F=1.8°C + S2
(°F-32)
(Eq. 2-2) °C fy~
Where: °F = degrees Fahrenheit
°C = degrees Celsius or degrees Centigrade
Absolute Temperature
Experiments in which a gas volume is determined as a function of temperature (at
a constant pressure) yield results similar to the data presented in Figure 2-l(a). The
solid portion of each line represents the gaseous state. If each line is extrapolated
(dashed portion of line) to a volume of zero, they all intersect at a common
temperature (-273.15°C or -459.67°F). This is the temperature at which a gas,
if it did not condense, would theoretically have a volume of zero. This temperature
(-273 15°C or -459.67°F) is called absolute zero. Another temperature scale,
developed by and named after English physicist Lord Kelvin, begins at absolute
zero and has temperature intervals equal to Centigrade units. This absolute
temperature scale is in units of degrees Kelvin (°K). A similar scale was developed
to parallel the Fahrenheit scale and is called the Rankine scale (°R). The following
formulas can be used to convert temperatures to their respective absolute scale.
(Eq. 2-3) °K=°C + 273.16
(Eq. 2-4) °R=°F + 459.67
2-1
-------
- 300 4-200 -150 - 100 - 50 0
-273.15°C t(°C)
Figure 2-l(a). Temperature relationships.
50 100
ahrenhe
212°F
32°F
Absolute zero -459.6°F
180
1000C
3en igrade
0°C
491.6
'fahrenheit-
degrees
-273°C
00
671.6°R
Absolute
491.6°R
273
centigrade-
degrees
0°R
373 °K boiling point of water
at 1 atm pressure
Freezing point of water
273 °K at 1 atm pressure
0°K Absolute zero
Figure 2-1 (b). Relationships of the absolute temperature systems.
Relationship of the absolute temperature systems are shown graphically in
Figure 2-l(b). The symbol "T" will be used throughout this manual to denote
absolute temperatures and the "t" will be used to indicate Fahrenheit or Centigrade
degrees. The absolute temperatures are always in volume calculations involving
temperature and pressure.
2-2
-------
Pressure
Definition of Pressure
A body may be subject to three kinds of stress: shear, compression, and tension.
Fluids are unable to withstand tensile stress; hence, they are subject to shear and
compression only. Unit compressive stress in a fluid is termed pressure and is
expressed as force per unit area.
Pressure
Metric
grn/cm*
English
lb,/ in2 (psi)
Pressure is equal in all directions at a point within a volume ot fluid and acts
perpendicular to a surface.
Barometric Pressure
Barometric pressure and atmospheric pressure are synonymous. These pressures are
measured with a barometer and are usually expressed as inches, or millimeters, of
mercury. Standard barometric pressure is the average atmospheric pressure at sea
level, 45° north latitude at 35°F and is equivalent to a pressure of 14.696 pounds-
force per square inch exerted at the base of a column of mercury 29.921 inches
high (in the English System). In the metric system, standard barometric pressure is
equivalent to a pressure of 1033.23 grams-force per square centimeter exerted at
the base of a column of mercury 760 mm high. Weather and altitude are responsi-
ble for barometric pressure variations.
Torricelli Barometer
The Torricelli or mercurial barometer was first used by one of Galileo's students,
Torricelli, in 1643. A mercurial barometer is made by sealing a tube, about 32
inches long, at one end. The tube is filled with mercury. It is then inverted and
placed into a container that is partially filled with mercury. The mercury in the
tube will fall until the weight of the mercury in the tube is equal to the force of the
air pressure on the mercury in the container. As shown in Figure 2-2, the
manometer and the mercurial barometer work on the same principle —atmospheric
pressure being measured with reference to a vacuum.
2-3
-------
Vacuum
h
Air
pressure
\
\
w
t!/>ii
'—
•'•
)
' In
Air
pressure
\
Air
pressure
\
Manometer
A
Barometer
B
Figure 2-2. The manometer and mercurial barometer.
Fortin Barometer
Since the mercurial barometer is the most accurate measurement (calibration
uncertainty of 0.001 to 0.03% of reading) of atmospheric pressure, it is still in wide
use today. The most common modified version of the mercurial barometer is the
Fortin type shown in Figure 2-3
Height of
mercury
Mercury
reservoir
Adjustment for
vernier scale
Adjustable
vernier scale
Reservoir level
adjustment screw
Figure 2-3. Fortin barometer.
2-4
-------
The height of the mercury column in a Fortin barometer is measured from the
tip of the ivory index point (see the enlargement in Figure 2-4) to the top of the
mercury column. The mercury leyel in the glass cylinder (ambient-yented cistern) is
adjusted until the ivory index point just pricks the surface of the mercury. This is
done by turning the datum-adjusting screw. Then the \ernier scale is adjusted
until the bottom of it is even with the top of the mercury meniscus. After the \er-
nier scale is adjusted, the height of the mercury column is read.
Glass cylinder
ambient-vented
cistern
Datum-adjusting
screw
I\or\ index
point
Leather bag
FigTare 2-4. Blow-up of Fortin barometer.
-------
A typical vernier scale is shown in Figure 2-5. The barometric pressure indicated
in the figure is determined in the following way:
The bottom of the vernier scale indicates not only the integer
component of the barometric pressure, but also the tenths
components, in this case, 29.9. The hundredths component is
indicated by the match between the outer scale and the vernier, in
this case, 0.04. The readings are totaled to determine the barometric
pressure: 29.9 + 0.04 = 29.94 in Hg. The equivalent metric reading is
76.05 cm.
inches
Scales match
Bottom of
vernier scale
Adjustable
vernier scale
Mercurv
column
Figure 2-5. Blow-up of Vernier scale
Aneroid Barometer
The aneroid barometer is usually not as accurate as a Torricelli barometer.
However, aneroid barometers are more widely used because the\ are smaller, more
portable, less expensive, and easier to adapt to recording instrumentation than are
Torricelli barometers.
The aneroid barometer usually consists of a metal chamber, bellows, or sylphon
cell (accordian-like) that is partially evacuated. A spring is used to keep the metal
chamber from collapsing (see Figure 2-6). The width of the chamber is determined
by the balance between the spring and the force exerted by the atmosphere. The
width of the chamber is indicated by a pointer and scale that can be calibrated to
2-6
-------
read directly in units of pressure (i.e., millimeters or inches of mercury, etc.). The
pointer movement can be amplified by using levers. Read-out systems can vary
from visual scales to recording devices. The combination of an aneroid barometer
and an automatic recording device is called a barograph.
Spring to prevent
collapse of diaphragm
Scale
Metal post
Partially evacuated
metal diaphragm
Figure 2-6. Aneroid barometer.
Pressure Transducers
A transducer is a device that is activated by power from one system and supplies
power in some other form to a second system. Conventional pressure transducers
use an elastic element that converts the energy from a pressure differential into a
displacement of a mechanical device. An example of a mechanical pressure
transducer is shown in Figure 2-7. Other pressure transducers convert the
mechanical displacement into an electrical signal. An example of an electrical
transducer is shown in Figure 2-8. Electrical pressure transducers have become very-
popular because the signal is easy to measure, control, amplify, transmit, and
record.
2-7
------- An error occurred while trying to OCR this image.
-------
Gage Pressure
Gages indicate the pressure of the system of which they are a part relative to
ambient barometric pressure. If the pressure of the system is greater than the
pressure prevailing in the atmosphere, the gage pressure is expressed as a positive
value; if smaller, the gage pressure is expressed as a negative. The term "vacuum"
designates a negative gage pressure.
The abbrevatiori "g" is used to specify a gage pressure. For example, psig, means
pounds-force per square inch gage pressure.
Absolute Pressure
Because gage pressure (which may be either positive or negative) is the pressure
relative to the prevailing atmospheric pressure, the gage pressure, added
algebraically to the prevailing atmospheric pressure (which is always positive), pro-
vides a value that is called "absolute pressure." The mathematical expression is:
(Eq. 2-3)
P=Pb
Where: P- absolute pressure
Pb = atmospheric pressure
Ps = SaSe pressure
Note: P, Pb, and ps must be in the same units of pressure before they can be added
(i.e., all must be in inches of mercury, mm of mercury, etc.).
The abbrevation "a" is sometimes used to indicate that the pressure is absolute.
For example, psia, means pounds per square inch absolute pressure.
Equation 2-3 allows conversion of one pressure system to the other. Relationship
of the pressure system is shown graphically in Figure 2-9 using two typical gage
readings, 1 and 2. Gage reading 1 is above the prevailing atmospheric pressure,
and, hence, is expressed as a positive value. Gage reading 2 is below the prevailing
atmospheric pressure and, therefore, is expressed as a negative value. Gage reading
3 has both sides open to the atmosphere, hence, the gage pressure is zero.
2-9
-------
Open to atmosphere
Pressure
source
Open to atmosphere
Gage pressure = 0
Gage reading 2
Vacuum
source
Open to atmosphere
Atmospheric pressure
Figure 2-9. Absolute-atmospheric-gage pressure relationship.
Example Problems
Problem 1:
The primary pressure gage of a regulator attached to a compressed nitrogen
cylinder indicates a reading of 2000 psig. An aneroid barometer mounted on the
wall indicates that the atmospheric pressure is 14.2 psi. What is the absolute
pressure inside the tank?
Solution:
P=
P= 2014.2
pg= 14.2 + 2000
Problem 2:
A water manometer is used to measure the pressure inside an evacuated flask. The
water manometer indicates that the evacuated flask has a vacuum of 26 inches of
water. A nearby Fortin barometer indicates that atmospheric pressure is 752.6 mm
Hg. What is the absolute pressure inside the flask?
2-10
-------
Solution:
Before pK and Pb can be added to give P, both must be in the same unit of
pressure. It is most common for pg to be converted to the same units as Pb since Pb
is in much larger units.
Since 1 inch of Hg= 13 inchs H2O (Hg is 13 times denser than water)
/ W linchHg \/25.4mm\ cn . „
26 inches H20 - — - ^— . = 50.8mmHg
\ / \13 inches H2O/\ 1 inch /
Now pg and Pb can be added. pg is negative because the evacuated flask is at a
vacuum or below atmospheric pressure.
.6 mm Hg+(-50.8 mm Hg)
P=701.8 mm Hg
The Concept of Pressure-Head
Pressure-head is the height of a column of fluid required to produce a given
pressure at its base.
The relationship between pressure and pressure-head is:
(Eq. 2-4) P~Qfh( — J
Where: p = pressure, force/area
Qf= density of fluid, mass /volume
g = local acceleration due to gravity, length/time2
gc - dimensional constant
h —pressure-head in terms of Qf, length
Pressure-head may be expressed in terms of any fluid that is convenient: e.g., Hg
or H2O.
Dalton's Law of Partial Pressure
When gases or vapors (having no chemical interaction) are present as a mixture in
a given space, the pressure exerted by a component of the gas-mixture at a given
temperature is the same as it would exert if it filled the whole space alone The
pressure exerted by one component of a gas-mixture is called its partial pressure.
The total pressure of the gas-mixture is the sum of the partial pressures.
Expressed mathematically:
Pto.al =
Where: P total — total pressure exerted by the system
p,= pressure of each component of the system
Zp,=pl+pS+. . . +pn
. +pn
(L: means "sum of")
2-11
-------
The Ideal Gas Laws
Ideal gases are gases whose molecules do not attract one another and which occupy
no part of the whole volume. Although there are no gases which have these proper-
ties, real gases, which deviate very slightly from ideal gas behavior under ordinary
temperature and pressure conditions, may be considered to be ideal gases.
Boyle's Law
Boyle's Law states: when the temperature (T) is held constant, the volume (F) of a
given mass of an ideal gas of a given composition varies inversely as the absolute
pressure, i.e.:
Va—(at constant T)
Where: a-proportional to.
One can see that, as the pressure on a gas system increases, the volume of the
gas system will decrease and vice versa.
Charles' Law
Charles' Law states: when the pressure (P) is held constant, the volume (V) of a
given mass of an ideal gas of a given composition varies directly as the absolute
temperature (T), i.e.,:
VaT (at constant P).
In other words, as the temperature of a gas system increases, the volume will also
increase and vice versa.
The Law for Ideal Gases
Both Boyle's and Charles' Law are satisfied in the following equation:
(Eq. 2-5) PV = nRT
Where: P = absolute pressure
V~ volume of a gas
T —absolute temperature
R = universal gas-constant
n - number of moles of a gas.
2-12
-------
A mole of a substance is the substance's molecular weight expressed in mass
units. Where the substance's molecular weight is the sum of the atomic weights of
the atoms which compose the substance.
We know that
m
n= —
M
Where: m = mass of a gas
M = molecular weight of a gas
therefore:
J
M
The units of R depend upon the units of measurement used in the equation.
Some useful values are:
(1) R = 0.082 (£) (atm) (°KYl(g-mole)'1
(2) fl = 62.4(£)(mm Hg)(0K.)~l(g-moleYl
Where the units are: V (£)
m(g)
M (g/g-mole)
T(°K)
P(atm for (1) or
mm Hg for (2))
Different values of R can be obtained by utilizing the appropriate conversion
factors.
Molar Volume (V)
One mole of any gas at 273°K and 760 mm Hg will occupy 22.414 liters. This con-
stant is obtained from the ideal gas law. From equation 2-5:
If: P=760 mm Hg
n — 1 mole
R = 62.4 (f) (mmHg) (i
T=273°K
V = V (molar volume)
PxV=nxRXT
(760 mm Hg) (V) = (l g-mole)(62A )(273°K)
g-mole °K
(1)(62.4)(273) f
760
F=22.414£ = V
2-13
-------
Therefore 1 mole of an ideal gas at 273°K and 760 mm Hg occupies 22.4141'. In
other words, the molar volume (V) of an ideal gas at 273 °K and 760 mm Hg is
22.474 I'/mole.
At EPA standard conditions (760 mm Hg, 298°K) one mole of any gas will
occupy 24.46£. The volume per mole constant for any gas at a given pressure and
temperature is called the molar volume and is symbolized by V.
Gas Density
Gas density can be determined by rearranging Equation 2-5 and letting density
PV=nRT=™RT
M
m PM
V RT
Where: Q = density
P= absolute pressure
M = molecular weight
T = absolute temperature
R = universal gas constant
Another method of determining density is by utilizing the fact that there are
24.46 liters per g-mole at 298°K and 760 mm Hg.
In the relationship, Q = m/V: if V is in terms of molar volume, V ((/g-mole of
a gas at STP), then m must be in terms of molecular weight, M (g/g-mole). So
Q = M/V at a given temperature and pressure.
298 P
M-
corrected to standard temperature
and pressure conditions (STP)
Where: Q=gas density (g/Q
M= molecular weight (g/ g-mole)
24.46 = molar volume (I/g-mole) at standard conditions
298 = temperature (°K) at standard conditions
T= temperature (°K) at actual conditions
760 = pressure (mm Hg) at standard conditions
P = pressure (mm Hg) at actual conditions.
2-14
-------
Standard Conditions for Atmospheric Sampling
To be able to compare gas sampling data collected by various agencies and other
organizations, all gas volumes must be corrected to a set of predetermined ("stan-
dard") conditions. For atmospheric or ambient sampling, these conditions are:
25°C or 298°K, and 760 mm Hg
The equation used to correct volumes sampled to standard conditions is:
"
/ P, \/298°K\
= 7T A—r—)
mm Hg/\ 1\ /
Where: V2 — volume of gas at 2nd conditions or at P2 and T2,
FI = volume of gas at 1st conditions of PI and Tlt ('
TI = initial temperature of gas, °K
Tz— final temperature of gas, in this case = 298°K.
P, = initial pressure of gas, mm Hg
P2= final pressure of gas, in this case = 760 mm Hg
O.S9=
760
Standard conditions for Temperature and Pressure are abbreviated STP.
Origin and Definition of Viscosity
Viscosity is the result of two phenomena: (a) intermolecular cohesive forces and
(b) momentum transfer between flowing strata caused by molecular agitation
perpendicular to the direction of motion. Between adjacent strata of a flowing
fluid a shearing stress results that is directly proportional to the velocity gradient.
(Figure 2-10). Viscosity is often defined as resistance to flow.
The relationship of these forces is shown in Equation 2-8.
(Eq. 2-8) S^=^
dy
Where: gc = dimensional constant
T = unit shearing stress between adjacent layers of fluid
— = velocity gradient
dy
JJL — proportionality constant (viscosity)
2-15
-------
mmmmmmmmmmmmmmmmmmmmmm KL* «••••••••••• mm ••••••
v + gv ^ / \
^ M i
Figure 2-10. Velocity gradient.
The proportionality constant, /*, is called the coefficient of viscosity, or merely,
viscosity. It should be noted that the pressure does not appear in Equation 2-8,
indicating that the shear (T) and the viscosity (/*) are independent of pressure.
(Viscosity actually increases very slightly with pressure but this variation is
negligible in most engineering problems (2).)
Kinematic Viscosity
Kinematic viscosity is defined according to the following relationship:
(Eq. 2-9)
Where:
v — kinematic viscosity
/i — viscosity of the gas
Q = density of the gas (note the absence of dimensions of force)
Liquid Viscosity Versus Gas Viscosity
Liquid Viscosity
In a liquid, transfer of momentum between strata having different velocities is
small, compared to the cohesive forces between the molecules. Hence, shear stress
(T) is predominantly the result of intermolecular cohesion. Because forces of
cohesion decrease with an increase in temperature, the shear stress decreases with
an increase in temperature. Equation 2-8 shows that shear stress is directly propor-
tional to the viscosity. Therefore, liquid viscosity decreases when the temperature
increases
2-16
-------
Gas Viscosity
In a gas, the molecules are too far apart for intermolecular cohesion to be effec-
tive. Thus, shear stress is predominately the result of an exchange of momentum
between flowing strata caused by molecular activity. Because molecular activity
increases with temperature increases, the shear stress increases with a rise in the
temperature. Therefore, gas viscosity is increased when the temperature increases.
Determination of Viscosity of Gases
The viscosity of a gas may be found accurately from the following formula:
Where: fi = viscosity at temperature T (°K)
fi° = viscosity at 0°C and prevailing pressure
T = absolute prevailing temperature (°K)
n-an empirical exponent (n =0.768 for air).
The viscosity of air and other gases at various temperatures and at a pressure of
1 atmosphere can be determined from the nomograph in Figures 2-11 and 2-12, or
from Equation 2-10. The unit of the viscosity coefficient is the poise:
1 poise = gm/cm. sec. A centipoise (cp) is equal to 10~2 poise.
Reynold's Number
Definition
A typical inertial force per unit volume of fluid is
A typical viscous force per unit volume of fluid is
The first expression divided by the second provides the dimensionless ratio known
as Reynold's Number:
(Eq. 2-11) NRe= . = ial force
viscous force
Where: Q — density of the fluid (mass/volume)
v = velocity of the fluid
gc = dimensional constant
L = a linear dimension
fji = viscosity of the fluid
N Re = Reynold's Number
2-17
-------
The larger the Reynold's Number, the smaller is the effect of viscous forces; the
smaller the Reynold's Number, the greater the effect of the viscous forces.
The linear dimension, L, for flow through tubes and ducts is a length
characteristic of the flow system. It is equal to four times the mean hydraulic
radius, which is the cross-sectional area divided by the wetted perimeter. Thus for
a circular pipe, L = diameter of the pipe; for a particle settling in a fluid medium,
L — diameter of the particle; for a rectangular duct, L = twice the length times the
width divided by the sum; and for an annulus such as a rotameter system,
L = outer diameter minus the inner diameter.
Viscosity of air at 1 atmosphere *
Temperature
Degree centigrade Degree fahrenheit
- 100 ^
--100
0-
100-
200-^
300-E
400-:-
500-
600-
700
800
900
1000-
-0
•100
— 200
— 300
— 400
-500
— 600
•700
800
•900
•1000
•llii'i
1200
•1300
1400
1500
•1600
•1700
• 1800
Viscosity
ceiitipoises
-0.1
- 0.09
- 0.08
-0.07
0.06
-0.05
-0.04
^0.03
-0.02
0.01
0.009
0.008
0.007
0.006
0.005
(l)centipoise
(10) -2 gm
cm - sec
(10) -2 poise
2.09(10) ~5
lbf - sec
ft*
2.09(10) -5
slug
ft - sec
6.72(10) -4
">m
ft - sec
*Perry, J. H. Chemical Engineer's Handbook, McGraw-Hill Book Co., New York, 1950.
Figure 2-11. Viscosity nomograph for air.
2-18
-------
a,
u
.036
100
200
500
600
300 400
Temperature, °F
Figure 2-12. Viscosity nomograph of various gases at 1 atmosphere.
700
2-19
-------
Laminar and Turbulent Flow
Laminar Flow
In laminar flow, the fluid is constrained to motion in layers (or laminae) by the
action of viscosity. The layers of fluid move in parallel paths that remain distinct
from one another; any agitation is of a molecular nature only. Laminar flow occurs
when Reynold's Number is less than 0.1 for particles settling in a fluid medium.
Turbulent Flow
In turbulent flow, the fluid is not restricted to parallel paths but moves forward in
a haphazard manner. Fully turbulent flow occurs when Reynold's Number is
greater than 1000 for settling particles having diameters greater than 40/Mm in air
with a 60% RH at 20 °C.
Summary of Useful Equations
Temperature
Where:
°C= (°F- 32)71.8
°F=1.8°C + S2
°K= °C + 273
°F = degrees Fahrenheit
°C = degrees Centigrade or Celsius
°K = degrees Kelvin
Pressure
1 std atm = 29.92 in Hg
= 760 mm Hg
Where: P = absolute pressure
Q = density
h = pressure head or height
g = gravitational acceleration
gc — dimensional constant
Subscripts
g = gage
f=fluid
b = barometric or atmospheric
2-20
-------
Ideal Gas Law
PV=nRT
= ™RT
M
_ 0.08205 (liters)(a;m)
(g-mofe)(°K)
62.4 (liters)(mm Hg)
1 g-mole = 22.414 liters at 273 °K and 760 mm Hg (molar volume)
Where: P= absolute pressure
V = volume
m = mass
M — molecular weight
R = gas constant
T = absolute temperature
n = number of g-moles of a gas
Gas Density
Where: Q ~ density
P= absolute pressure
R = gas constant
T = absolute temperature
M = molecular weight
Viscosity, /*
PM
RT
ft »sec.
1 poise — 1
g
cm»sec
1 cp= 10"2 poise
2-21
-------
Reynold's Number
_ LVQ _ inertialforce
;
/j, viscous jorce
Where: Q = density of the fluid (mass/volume)
v — velocity of the fluid
gc = dimensional constant
L-a linear dimension
\i = viscosity of the fluid
NRe = Reynold's Number
References
1. J. K. Uennard. Elementary Fluid Mechanics. New York: John Wiley and Sons,
Inc., 1947.
2. M. B. Lemon and M. Ference. Analytical Experimental Physics. Chicago: The
University of Chicago Press, 1946.
3. J. H. Perry. Chemical Engineers Handbook. New York: McGraw-Hill Book
Co., Inc., 1950.
4. Robert P. Benedict. Fundamentals of Temperature, Pressure, and Flow
Measurements. New York: John Wiley and Sons, Inc., 1969.
5. William H. Nebergall, Frederic C. Schmidt, and Henry F. Holtzclaw, Jr.
General Chemistry, Lexington, MA: Reytheon Education Company, 1968.
2-22
-------
Units of Measurement
Recommended Units (Reference 1)
"At the present time, personnel engaged in the study of air pollution are con-
fronted with a multitude of confusing and conflicting units of expression. A search
through the literature has shown a wide variation in the methods of reporting data.
Many of the units of expression are carry-overs from other fields, such as water
pollution studies and industrial hygiene surveys. While these methods of expression
are not incorrect, their application to air pollution studies is often misleading."
This section of the manual covers the units presently being used and those
recommended for the more commonly measured air pollution parameters.
The recommended units were selected so that the reported values would be small
whole numbers in the metric system. If possible, the reported units should be the
same as those that are actually measured. For example, weight should be reported
in grams or milligrams, and volume in cubic meters. The measured value should
never be multiplied by large numbers to extrapolate to extremely large areas or
volumes. If this is done, the resulting values are misleading. For example: to report
paniculate fallout on a weight per square mile basis, the area actually sampled,
which is about 1 square foot, would have to be extrapolated to a square mile by
multiplying the measured results by almost 28,000,000. Reporting the results on
the basis of a square mile is misleading, because we are saying that the one square
foot that we sampled is representative of a square mile surrounding this sampling
site. This we know, in most cases, is not true.
When reporting results, the type of sampling instrument should be described,
and when volumes of air are sampled, the temperature and pressure at the time of
the sampling should be reported.
Particle Fallout
Units Presently In Use
• Tons per square mile per month
• Tons per square mile per year
• Pounds per acre per month
• Pounds per acre per year
• Pounds per thousand square feet per month
• Ounces per square foot per month
• Grams per square foot per month
• Grams per square meter per month
• Kilograms per square kilometer per month
• Grams per month per 4-inch or 6-inch jar
• Milligrams per square inch per month
2-23
-------
Recommended Units
Milligrams per square centimeter per time interval as mg/cmVmo, or mg/cmVyr.
Ranges Reported
0.5 to 135 mg/cmVmo.
Outdoor Airborne Particulate Sampling
Units Presently In Use
• Milligrams per cubic meter
• Parts per million by weight
• Grams per cubic foot
• Grams per cubic meter
• Micrograms per cubic meter
• Micrograms per cubic foot
• Grams per cubic foot
• Pounds per thousand cubic foot
Recommended Unit
Micrograms per cubic meter at standard temperature and pressure.
Ranges Expected
10 to 5000 micrograms per cubic meter.
Gaseoiis Materials
Units Presently In Use
• Milligrams per cubic meter
• Micrograms per cubic meter
• Micrograms per liter
• Parts per million (ppm) by volume
• Parts per hundred million
• Parts per billion (ppb) by volume
• Parts per trillion (ppt) by volume
• Ounces per cubic foot
• Pounds per cubic foot
• Grams per cubic foot
• Pounds per thousand cubic foot
Recommended Unit
ppm or ppb by volume
Ranges Reported
Parts per trillion to parts per million.
2-24
-------
Standard Conditions for Reporting Gas Volumes
Units Presently In Use
• 760 millimeters Hg pressure and 20°C
• 760 millimeters Hg pressure and 0°C
• 760 millimeters Hg pressure and 65 °F
• 760 millimeters Hg pressure and 25°C
• 700 millimeters Hg pressure and 0°C
• 700 millimeters Hg pressure and 20 °C
• 30 inches of mercury pressure and 65 °F
Recommended Units
760 millimeters Hg pressure and 25 °C.
Particle Counting
Units Presently In Use
• Number per cubic meter of gas
• Number per liter of gas
• Number per cubic centimeter of gas
• Number per cubic foot of gas
Recommended Units
Number of particles per cubic meter of gas
Range Reported
10 million and above particles per cubic meter.
Particle Count In Sedimentation Devices
(both horizontal and vertical)
Recommended Units
Number of particles per square centimeter per time interval.
Temperature
Units Presently In Use
• Degrees Celsius or Centigrade
• Degrees Fahrenheit
Recommended Unit
Degrees Celsius or Centigrade (°C)
2-25
-------
Time
It is recommended that time be measured on the 0000 to 2400 basis to eliminate
the possible confusion that results from two 12-hour (a day) sections. Example:
1:00 a.m. = 0100, 1:00 p.m. = 1300, 12:00 midnight = 0000.
Pressure
Units Presently In Use
Atmospheric Pressure
• Atmospheres
• Millimeters of mercury
• Inches of mercury
Sampling Pressures
• Millimeters of mercury
• Inches of mercury
• Millimeters of water
• Inches of water
Recommended Unit
Millimeters of mercury (mm Hg)
Sampling Rates
Units Presently In Use
• Cubic meters per second
• Cubic meters per minute
• Cubic feet per second
• Cubic feet per minute
• Liters per second
• Liters per minute
• Cubic centimeters per second
• Cubic centimeters per minute
Recommended Units
• Cubic meters per minute
• Cubic centimeters per minute
Ranges Reported
Cubic centimeters per minute to 3 cubic meters per minute. The selection of cubic
meters, or cubic centimeters, will depend on the sampling equipment used: the
units chosen should give small whole numbers.
2-26
-------
Visibility
* Mile.-, a..-.! ii utii is;-, ol a m»le
* Kilometers and (."actions ot ki'omeu-1,%
Recommended Onit
Kilometers
Summary «/ iiacipnimended Units
* Particle fallout. Milligrams per squaie < entmu-'ei p": Um;-,iL.-
• Outdoor airborne ^articulates. Microgntrns pt.i cubsc -ve".''" a:
* Gaitou*, material. ppm or ppb by volume
* Ga.-, volumes repoi led at SIT (760 millimeu-i:, i'g pj ...,.:-• ;•--'
* Pur tick count on ^edimentatnni devices. Numbe:- ot <;t".i- '<• (
lemimeter per time interval.
• Temperature. Centigrade scale.
* Time 0000 to 2400 hours per day.
• Pressure. Millimeters of mercury
* Sampling rates Cubic meteis per minuU . - .
• Visibility. Kilometeis,
It is believed (hat adoption and use o? «ii' i\ *.omi.:e:, j\ ,\ s,^-; ,
will lesult in more uniform reporting a.id will leniov^ inuJi ./! • i>-, ,<»utn
that is now found in reports of air pollution.
Conversions
Until a untorm system of units has been agieed upon and adheied to m the
scientific liteiature, conversion from one unit ol expression to another will be a
necessity,
Sample Paniculate Problem:
Convert 89 0 tons pei square mile to (be iei oinmended uirii-, s nnlligi ams p<-i
square- ce.ntnnetei)
Solution.
tons '^000 Ib l!/i.6x 10' mi; »'.i h ;i
89.0 - - x - - - ^ - "
rnr' to.i Ib iM'oO)'!'! ' ',)"., V/
-------
Sample Gas Problem:
Discussion:
The expression parts per million is without dimensions, i.e., no units of weight or
volume are specifically designed. Using the format of other units, the expression
may be written:
parts
million parts
"Parts" are not defined. If cubic centimeters replace parts, we obtain:
cubic centimeters
million cubic centimeters
Similarly, we might write pounds per million pounds, tons per million tons, or
liters per million liters. In each expression, identical units of weight or volume
appear in both the numerator and denominator and may be cancelled out, leaving
a dimensionless term.
An analog of parts per million is the more familiar term "percent." Percent can
be written:
parts
hundred parts
To convert from part per million by volume, ppmv, (jnf/0 it is necessary to know
the molar volume at the given temperature and pressure and the molecular weight
of the pollutant.
At 25 °C and 760 mm Hg, one mole of any gas occupies 24.46 liters.
Convert the following:
. \ 2.5 ppm by volume of SO2 was reported as the atmospheric concentration.
* a' what is this concentration in micrograms (fig} per cubic meter (m3) at 25 °C
and 760 mm Hg?
b. What is the concentration in fig/m3 at 37 °C and 752 mm Hg?
Solution:
Let parts per million equal fit/? then 2.5 ppm = 2.5^/f. The molar volume at 25°
and 760 mm Hg is 24.46f and the molecular weight of SO2 is 64.1 g/mole.
C 24.46 id /imole m3 m3
b. (24.46 ,Q = 25.73
V mm Hg/
2.5 ^ x 1 jonole x 64.1 fig x J000_f = 6.2 X 103 ^ at 37°C, 752 mm Hg
( 25.73 \d fimole m3 m3
This sample problem also points out the need for reporting temperature and
pressure when the results are presented on a weight to volume basis.
*since, at STP 1 mole of a gas occupies 24.46 liters, 1 /imole = 24.46 fit
2-28
-------
Problems
1. Convert the following:
a. 68 °F - °C (answer 20°C)
b. 28 °C-°K (answer 301 °K)
c. 29.03 in Hg—mm Hg (answer 737.3 mm Hg)
2. An ideal gas occupies a volume of 2000 mf at 700 mm Hg and 20 °C.
What is the volume of the gas at STP? (answer 1874 mf)
3. If a concentration of carbon monoxide (CO) is noted as 10 ppm, what is this
concentration in terms of /ig/m3 at STP? (CO = 28 g/mole)
(answer 11,440 /ig/m3, 11.4 mg/m3)
4. Ambient air was sampled at a rate of 2.25 liters per minute for a period of
3.25 hours at 19°C, 748 mm Hg. What volume of air was sampled at STP?
(answer 441 Q
5. Convert 1000 /ig/m3 SO2 at STP to ppm. (SO2 = 64 g/mole)
(answer 0.38 ppm)
Tables to use in this task appear in Appendix C of this manual.
References
'. i t-iraglio, F. P., Sheehy, J. P., and Manganelli, R. M. Recommended Units of
Expression for Air Pollution. J. Air Poll. Control Assoc. 8:220-222, 1958.
2. Weast, R. C., and Astle, M. J. Handbook of Chemistry and Physics, Boca
Raton, Florida: Chemical Rubber Publishing Co., 60th edition, 1979.
2-29
-------
Air at EPA Standard
Conditions
Air Pollution
Arrester
Aspirator
Atmosphere, The
Atmosphere, An
Breathing Zone
Chimney Effect
Collection Efficiency
Collector
Cloud
Condensate
Condensation
Condensoid
Contaminant
Count Median Size
Definitions
Air at 25°C and 760 mm Hg (29.92 in. Hg).
The presence of unwanted material in the air. The
term "unwanted material" here refers to material con
centrations present for a sufficient time and under cir-
cumstances to interfere significantly with comfort,
health, or welfare of persons, or with the full use and
enjoyment of property.
A term for an air cleaning device.
Any apparatus such as a squeeze bulb, fan. pump, or
venturi, that produces a movement of a fluid by-
suction.
The whole mass of air surrounding the earth and being
composed largely of oxygen and nitrogen.
A specific gaseous mass, occurring either naturally or
artificially containing any number of constitutents and
in any proportion.
That location in the atmosphere at which persons
breathe.
A phenomenon consisting of a vertical movement of a
localized mass of air or other gases due to temperature
differences.
The percentage of a specified substance retained by a
gas cleaning or sampling device.
A device for removing and retaining contaminants
from air or other gases. Usually this term is applied to
cleaning devices in exhaust systems.
A visible dispersion occupying a discrete portion of
space, with apparent boundaries.
Liquid or solid matter formed by condensation from
the vapor phase. In sampling, the term is applied to
the components of an atmosphere that have been
isolated by simple cooling.
The process of converting a material in the gaseous
phase to a liquid or solid state by decreasing
temperature, by increasing pressure, or both. Usually
in air sampling only cooling is used.
The particles of a dispersion formed by condensation.
Unwanted material.
A measurement of particle size for samples of par-
ticulate matter, consisting of that diameter of particle
such that one half of the number of particles is larger
and half is smaller
2-30
-------
Density
Diffusion, Molecular
Dispersion
Dispersoid
Diurnal
Dust
Dust Fall
Dust Loading
Droplet
Efficiency
Efficiency, Fractional
Ejector
Emissions
Emission Mixture
Flocculation
Flowmeter
The mass per unit volume of substance.
A process of spontaneous intermixing of different
substances, attributable to molecular motion and
tending to produce uniformity of concentration.
The most general term for a system consisting of par-
ticulate matter suspended in air or other gases.
The particles of a dispersion.
The term means recurring daily. Applied to (variations
in concentration of) air contaminants, diurnal indicates
variations following a distinctive pattern and recurring
from day to day.
A loose term applied to solid particles predominantly
larger than colloidal and capable of temporary suspen-
sion in air or other gases. Dusts do not tend to floc-
culate except under electrostatic forces; they do not dif-
fuse but settle under the influence of gravity. Deriva-
tion from larger masses through the application of
physical force is usually implied.
See Particle Fall.
An engineering term for "dust concentration," usually
applied to the contents of collection ducts and the
emission from stacks.
A small liquid particle of such size and density as to
fall under still conditions, but which may remain
suspended under turbulent conditions.
The ratio of attained performance to absolute perfor-
mance, commonly expressed in percent.
The mean collection efficiency for specific size fractions
of a contaminant. Commonly this term has been
applied to the performance of air cleaning equipment
towards paniculate matter in various size ranges.
A device that uses a fluid under pressure, such as
steam, air, or water, to move another fluid by develop-
ing suction. Suction is developed by discharging the
fluid under pressure through a venturi.
The total of substances discharged into the air from a
stack vent, or other discrete source.
The total mixture in the outside atmosphere of emis-
sion from all sources.
Synonymous with agglomeration.
An instrument for measuring the rate of flow of a fluid
moving through a pipe or duct system. The instrument
is calibrated to give volume or mass rate of flow.
2-31
-------
Fly Ash
Fog
Freezing Out
Fume
Gas
Grab Sample
Impaction
Impactor
Impingement
Impingement, Dry
Impingement, Wet
Impinger
The finely divided particles of ash entrained in flue
gases arising from the combustion of fuel. The particles
of ash may contain incompletely burned fuel. The term
has been applied predominantly to the gas-born ash
from boilers with spreader stoker, underfeed stoker,
and pulverized fuel (coal) firing.
A loose term applied to visible aerosols in which the
dispersed phase is liquid. Formation by condensation is
usually implied; in meterology, a dispersion of water or
ice.
See Sampling, Condensation.
Properly, the solid particles generated by condensation
from the gaseous state, generally after volatilization
from melted substances, and often accompanied by a
chemical reaction such as oxidation. Fumes flocculate
and sometimes coalesce. Popularly, the term is used in
reference to any of all types of contaminant, and in
many laws or regulations with the added qualification
that the contaminant have some unwanted action.
One of the three states of aggregation of matter,
having neither independent shape nor volume and
tending to expand indefinitely.
See Sampling, Instantaneous.
A forcible contact of particles of matter, a term often
used synonymously with impingement.
A sampling device that employs the principle of impac-
tion (impingement). The "cascade impactor" refers to a
specific instrument employing several impactions in
series to collect successively smaller sizes of particles.
The act of bringing matter forcibly in contact. As used
in air sampling, impingement refers to a process for the
collection of paniculate matter in which the gas being
sampled is directed forcibly against a surface.
The process of impingement carried out so that par-
ticulate matter carried in the gas stream is retained
upon the surface against which the stream is directed.
The collecting surface may be treated with a film of
adhesive.
The process of impingement carried out within a body
of liquid, the latter serving to retain the paniculate
matter.
Broadly, a sampling instrument employing impinge-
ment for the collection of paniculate matter. Com-
monly, this term is applied to specific instruments, the
"midget" and "standard" impinger.
2-32
-------
Impinger, Midget
Impinger, Standard
Isokinetic
Mass Concentration
Mass Median Size
Mist
Month
Odor
Odor Concentration
Odor Unit
Odorant
Orifice Meter
Particle
A specific instrument employing wet impingement,
using a liquid volume of 10 ml and a gas flow of
0.1 cu. ft. per min.
(Note: SeeJ. R. Littlefield, E. L Feicht, and H. H.
Schrenk, "Midget Impinger for Duot Sampling," U.S.
Bureau of Mines. Report of Investigations 3360 (1937).)
A specific instrument employing wet impingement,
using a liquid volume of 75 ml and a gas flow of
1 cu. ft. per min.
(Note: See L. Greenburg and G. W. Smith, "A New
Instrument for Sampling Aerial Dust, U.S. Bureau of
Mines, Report of Investigations 2392 (1922). See also
T. Hatch, H. Warren, and P. Drinker, Journal
Industrial Hygiene, No. 14, p. 301 (1932).)
A term describing a condition of sampling, in which
the flow of gas into the sampling device (at the opening
or face of the inlet) has the same flow rate and direc-
tion as the gas stream being sampled.
Concentration expressed in terms of mass of substance
per unit volume of gas or liquid.
A measurement of particle size for samples of par-
ticulate matter, consisting of that diameter such that
the mass of all larger particles is equal to the mass of
all smaller particles.
A loose term applied to dispersions of liquid particles,
the dispersion being of low concentration and the par-
ticles of large size. In meteorology, a light dispersion of
water droplets of sufficient size to be falling.
For reporting analyses of outdoor air on a monthly rate
results are calculated to a base of 30 days.
That property of a substance affecting the sense of
smell; any smell; scent; perfume.
The number of unit volumes that a unit volume of
sample will occupy when diluted to the odor threshold.
Unit volume of air at the odor threshold.
Odorous substance.
A flowmeter, employing as the measure of flow rate the
difference between the pressures measured on the
upstream and downstream sides of the orifice (that is,
the pressure differential across the orifice) in the con-
veying pipe or duct.
A small discrete mass of solid or liquid matter.
2-33
-------
Particle Concentrations
Particle Fall
Particle Size
Particle Size Distribution
Precipitation, Electrostatic
Precipitation,
Meteorological
Precipitation, Thermal
Precipitation, Ultrasonic
Precipitator, Electrostatic
Precision
Pressure Static
Concentration expressed in terms of number of par-
ticles per unit volume of air or other gas.
(Note: On expressing particle concentration, the
method of determining the concentration should be
stated.)
A measurement of air contamination consisting of the
mass rate at which solid particles deposit from the
atmosphere. A term used in the same sense as the older
terms Dust Fall and Soot Fall but without any implica-
tion as to nature and source of the particles.
An expression for the size of liquid or solid particles
expressed as the average or equivalent diameter.
The relative percentage by weight or number of each
of the different size fractions of paniculate matter.
A process consisting of the separalion of paniculate
matter from air or other gases under the influence of
an electrostatic field.
The precipitation of water from the atmosphere in the
form of hail, mist, rain, sleet, and snow. Deposits of
dew, fog, and frost are excluded.
A process consisting of the separation of paniculate
matter from air and other gases under the influence of
a relatively large temperature gradient extending over a
short distance. In the "Thermal Precipitator" (a
sampling instrument), the air or gas is drawn through a
narrow chamber across which extends a heated wire,
paniculate matter being deposited upon the adjacent
collecting surface.
A process consisting of the separation of particulate
matter from air and other gases following agglomera-
tion induced by an ultrasonic field.
Apparatus employing electrostatic precipitation for the
separation of particles from a gas stream. The
apparatus may be designed either for sampling or for
cleaning large volumes of gas.
The degree of agreement of repeated measurements of
the same property, expressed in terms of dispersion of
test results about the mean result obtained by repetitive
testing of a homogenous sample under specified condi-
tions. The precision of a method is expressed quan-
titatively as the standard deviation computed from the
results of a series of controlled determinations.
The pressure of a fluid at rest, or in motion, exerted
perpendicularly to the direction of flow.
2-34
-------
Pressure, Velocity
Pressure, Total
Pressure, Gage
Probe
Rotameter
Sample, Integrated
Sample, Continuous
Sampling
Sampling, Condensation
That pressure caused by and related to the velocity of
the flow of fluid; a measure of the kinetic energy of the
fluid.
The pressure representing the sum of static pressure
and velocity pressure at the point of measurement.
The difference in pressure existing within a system and
that of the atmosphere. Zero gage pressure is equal to
atmospheric pressure.
A tube used for sampling or for measuring pressures at
a distance from the actual collection or measuring
apparatus. It is commonly used for reaching inside
stacks and ducts.
A device, based on the principle of Stoke's Law, for
measuring rate of fluid flow. It consists of a tapered
vertical tube having a circular cross-section, and con-
taining a float that is free to move in a vertical path to
a height dependent upon the rate of fluid flow upward
through the tube.
A sample obtained over a period of time with (1) the
collected atmosphere being retained in a single vessel,
or (2) with a separated component accumulating into a
single whole. Examples are dust sampling in which all
the dust separated from the air is accumulated in one
mass of fluid; the absorption of acid gas in an alkaline
solution; and collection of air in a plastic bag or
gasometer. Such a sample does not reflect variations in
concentration during the period of sampling.
Withdrawal of a portion of the atmosphere over a
period of time with continuous analysis or with separa-
tion of the desired material continuously and in a
"linear" form. Examples are continuous withdrawal of
the atmosphere accompanied by absorption of a com-
ponent in a flowing stieam of absorbent or by filtration
on a moving strip or paper. Such a sample may be
obtained with a considerable concentration of the con-
taminant but it still indicates fluctuations in concentra-
tion that occur during the period of sampling.
A process consisting of the withdrawal or isolation of a
fractional part of a whole. In air or gas analysis, the
separation of a portion of an ambient atmosphere with
or without the simultaneous isolation of selected
components.
A process consisting of the collection of one or several
components of a gaseous mixture by simple cooling of
the gas stream in a device that retains the condensate.
2-35
-------
Sampling, Continuous
Sampling, Instantaneous
Sampling, Intermittent
Series Collection
Settling Velocity
Smog
Smoke
Soot
Sorbent
Sorption
Specific Gravity
Temperature, Absolute
Sampling without interruptions throughout an opera-
tion or for a predetermined time.
Obtaining a sample of an atmosphere in a very short
period of time such that this sampling time is insignifi-
cant in comparison with the duration of the operation
or the period being studied.
Sampling successively for limited periods of time
throughout an operation or for a predetermined period
of time. The duration of sampling periods and of the
intervals between are not necessarily regular and are
not specified.
An operation involving the use of two or more collec
tors joined in a series.
The terminal rate of fall of a particle through a fluid
as induced by gravity or other external force; the rate
at which frictional drag balances the accelerating force
(or the external force).
A term derived from smoke and fog, applied to exten-
sive atmospheric contamination by aerosols, these
aerosols arising partly through natural processes and
partly from the activities of human subjects. Now
sometimes used loosely for any contamination of air.
Small gas-borne particles resulting from incomplete
combustion, consisting predominantly of carbon and
other combustible material, and present in sufficient
quantity to be observable independently of the presence
of other solids.
Agglomerations of particles of carbon impregnated
with "tar," formed in the incomplete combustion of
carbonaceous material.
A liquid or solid medium in or upon which materials
are retained by absorption or adsorption.
A process consisting of either absorption or adsorption
or both.
The ratio of the density of the substance in question to
the density of a reference substance at specified condi-
tions of temperature and pressure.
(a) Temperature measured on the thermodynamic
scale, designated as degrees Kelvin ( °K). (b)
Temperature measured from absolute zero ( —273.15°C
or -459.67°F). The numerical values are the same for
both the Kelvin scale and the ideal gas scale.
2-36
-------
Vapor The gaseous phase of matter which normally exists in a
liquid or solid state.
Volume Concentration Concentration expressed in terms of gaseous volume of
substance per unit volume of air or other gas usually
expressed in percent or parts per million.
Week For reporting analyses of outdoor air on a weekly rate
results are calculated to a base of 7 consecutive 24 hour
days.
Year For reporting analyses of outdoor air on a yearly rate
twelve 30-day months are to be used.
2-37
-------
Chapter 3
Air Measuring Instruments
Introduction to Air Movers
The primary purpose of an air mover in the sampling process is to create a flow of
air that will allow the contaminant in the air to be analyzed directly or to be cap-
tured by a collection device for subsequent analysis. Collection devices include
filters, impingers, and impactors (1). Air movers range in capacity from a few
cubic centimeters per minute (cmVmin) of air up to tens of cubic meters per
minute (mVmin). In operational complexity air movers range from a squeeze bulb
to a multistage pump.
Relationship of Air Movers to Other Sampling System Components
Air movers are vital components of sampling trains used for the sampling and
analysis of air for its pollutant content (see Figure 3-1). Several of the considera-
tions governing the selection of an air mover for a particular application will be
mentioned in this discussion.
Contaminant
detector
Air in
Sample
device
A
Flow
measuring
device
Air out
Figure 3-1. A sampling train.
3-1
-------
Air Mover Classification
There are many parameters by which air movers can be evaluated and classified.
Classification of air movers can be made according to their function, capacity,
"driving" force, principle of operation, among other parameters. The classification
scheme often depends to a great extent on the classifier. The actual types of air
movers 10 be discussed are pumps, ejectors, liquid displacers, and evacuated flask,.
Air Mover Selection Criteria
Since ihere is no one "typical" air sampling train, equipment selection is an impor-
tant aspect of any airborne sampling scheme, meaning that air mover selection
must be made along with the selection of other sampling train components. Here
are some of the many factors to be considered before selection will be discussed.
The itemizing of the considerations is not necessarily in order of their importance.
Pollutant Concentration and Sampling Time
The required flow rate of the air mover is often determined by the sensitivity of the
analysis method, which in turn sets the minimum quantity of the contaminant
required for analysis (2). For example, if the analysis procedure required 10 /*g of
material to obtain the desired precision, and if the air contains 1 jig7™3 of that
material, a sample size of 10 m3 would be required, assuming 100% collection effi-
ciency. The required flow rate of the air mover can be calculated by determining
the available sampling time and the sample volume required. For example, if the
above-mentioned 10 m3 of air had to be moved (sampled) in a period of 10 hours,
an air mover of 1 mVhr capacity could be used. In many instances it is more feasi-
ble to sample for longer periods of time, and sometimes at higher rates, than those
required. As long as the sample size is greater than the minimum detectable and
less than the concentration that theoretically saturates the collection medium, this
may be of no consequence.
Sampling Rate Required
In some instances it may be necessary to sample at some required flow rate to
ensure efficient sample collection. An example is the collection of a sample by
using a chemical reaction. The reaction kinetics may depend upon the length of
time the contaminant is in the presence of the reacting substance, and the length
of contact time is dependent on the sampling rate. Other examples include the col-
lection of particulates by use of impingers or impactors. The particle size collected
will be dependent on the approach velocities, and these in turn are dependent
upon the sampling rates.
3-2
-------
Physical and Chemical Nature of Air to be Sampled
The operation of the air mover selected must be compatible with the physical and
chemical characteristics of the air and contaminant to be sampled. Air of a cor-
rosive or abrasive nature can create problems with the air mover unless the internal
parts are nonreactive with the sample air.
In instances where sampling is to be done in an environment of a potentially
explosive nature, approved explosion-proof air movers or explosion-sealed air
movers should be selected. A completely sealed electric motor for a pump (1), in
conjunction with non-sparking metal parts, is an example of an explosion-sealed air
mover.
Portability of Air Mover
Much air sampling is performed at temporary sampling sites. When sampling
trains must be moved from location to location it is important that each
component be as portable and light-weight as possible. At permanent sampling
locations air mover portability is not such an important consideration.
The power source for the air mover may be the limiting factor in its portability,
especially when the power source must be AC line voltage.
Air Mover Noise
Since air sampling may be required in areas having noise restrictions, the noise
produced by an air mover must be considered. Also, the suppression of air mover
noise has aesthetic values. Noise levels may be considered in relation to the length
of the sampling period, that is, for short periods high noise levels may be accep-
table. Many commercially available air movers have special sound-adsorbing liners,
vibration-reducing cushioning material, or mufflers as noise-reducing components
of the mover itself.
Air Mover Maintenance
It is an established fact that man-made equipment, whether air sampler com-
ponents or computers, will not be maintenance-free. Maintenance considerations
are of special importance for equipment to be used in the field, because valuable
time can be lost in transit between the repair shop and the sampling location.
There are three particular points to consider about air mover maintenance: (a) air
mover parts that might need repairing should be easily accessible; (b) the complete
repair should not be very time consuming; and (c) air mover parts should not be
overly expensive. Many costly breakdowns can be avoided if the proper preventative
maintenance schedule is used.
3-3
-------
Resistance
Flow resistance considerations may dictate the selection of an air mover of a par-
ticular type (see Figure 3-2). The sample collection device will offer some resistance
to the flow of air through it; therefore, the air mover must be able to overcome
this resistance so that efficient sample collection can be accomplished.
+ P
Filter
(sample collection
device)
1. Pressure loss through sampling lines due to friction
2. Pressure drop across filter
3. Pressure increase through pump
4. Pressure drop through meter
Figure 3-2. Pressure profile for basic sampling apparatus.
Constant Flow Rate
It is often desirable to collect a sample at a constant flow rate. Variation in sample
conditions or other external influences can make this difficult. For example, the
resistance of a filter can be expected to increase with sample 'buildup," thereby
decreasing the flow rate through the sample apparatus. Poor voltage regulation can
result in variable motor speeds leading to variable sampling rates (3). However, some
commercially available air movers feature constant flow rates despite varying
sampling conditions; others require flow regulation devices in order to have con-
stant flow rates. Some flow regulation mechanisms will be discussed later in the
chapter. Another factor of importance in some applications is the ability to vary
the flow rate of the air mover and then to maintain constancy at the selected rate.
This topic will also be discussed in more detail later.
3-4
-------
Pumps
Pumps have been defined as devices that raise or transfer fluids (2). Since air is a
fluid, a pump that moves air either raises it to another level or transfers it to
another location. In air sample collection, the air is transferred from one location
through, or into, a sample collection device.
Classification of Pumps
Categorization of pumps is a difficult task because of the many variables. This sec-
tion will discuss pumps in terms of two broad classes based on flow variation with
pressure. These two major classes will be subdivided into categories according to
specific principles of operation: (a) positive displacement pumps and (b) centrifugal
pumps.
Positive Displacement Pumps
Positive displacement pumps are often characterized by a linear relationship
between the suction pressure and pump capacity (see Figure 3-3). This indicates
that AQ/Ap (AQ is the change in flow rate; Ap is the pressure drop across the
pump) is a constant value. Figures 3-3 and 3-5 are representative of characteristic
curves of pumps to be discussed later.
Positive
displacement
>s
*J
°u
Q.
u
Pressure (p)
Figure 3-3. Positive displacement pump.
The name positive displacement arises from the fact that the inner parts of these
pumps are movable and tight-fitting, and the air is displaced through them by the
movement (displacement) of these tight parts (4). Figure 3-4 indicates a further
subdivision of positive displacement pumps, this division being made according to
the principle of operation. Reciprocating pumps are characterized by fixed casings
containing movable pistons that work only forward-and-backward or up-and-down,
and by the pressure of suction and discharge valves.
3-5
-------
Principle of Operation
Reciprocating
Rotary (not discussed
in this manual)
Type of Pump
piston
plunger
diaphragm
gear
lobe
vane
screw
rotary plunger
Figure 3-4. Classification of positive displacement pumps.
The operation of some of the specific types of reciprocating pumps will be
discussed later in this section.
Centrifugal Pumps
Centrifugal pumps are representative of pumps other than positive displacement
pumps. Centrifugal pumps do not have a straight line relationship between suction
pressure and capacity, thus AQ/Ap is not constant (see Figure 3-5).
u
a
O-
Centrifugal
Pressure (p)
Figure 3-5. Centrifugal pump.
A centrifugal pump moves fluids by a centrifugal force created by a wheel,
called an impeller, revolving in a tight casing (4). Some additional examples of
pumps other than positive displacement pumps are: (a) turbine pumps, (b) propeller
pumps, and (c) screw drag pumps. Each of these exhibit the same general pressure-
capacity relationship as the centrifugal pump.
3-6
-------
Positive Displacement Pump Operation
Some of the positive displacement pumps previously classified will now be discussed.
Piston Pumps (Reciprocating)
The principle of operation of a piston pump is that air is drawn into a chamber or
cylinder on the suction stroke of a piston and then is pushed out on the discharge
stroke (5), as illustrated in Figure 3-6. On the suction stroke the suction valve is
open, allowing air to flow in; on the discharge stroke the suction valve closes and
the discharge valve opens, allowing air to flow out. An internal combusion engine
is an example of a piston pump. Piston pumps vary in complexity of operation
from manually operated ones to models with many working parts.
0
I/////////////////
Piston
1 1
c^ ;
\ 1 '
[/////A///V/////
Chamber
Suction
' valve
t
Discharge
[ valve
Figure 3-6. Piston pump.
Diaphragm Pumps (Reciprocating)
The operation of a diaphragm pump is very similar to a piston pump in principle.
The piston (plunger) in a diaphragm pump does not move in a tightly-fitted
chamber as in the piston pump, but is attached to the center of a circular
diaphragm, the outer edge of which is bolted to a flange on the pump casing (6).
The diaphragm may be made of metal (7) or some soft material such as Teflon®
or neoprene (12). The most important characteristic of the diaphragm material is
its flexibility and resistance to reaction with the air being moved. The up-and-
down motion of the plunger is permitted by diaphragm flexibility without the
rubbing of one part on another (see Figure 3-7). On upward movement of the
plunger, air flows into the pump through a suction valve. Downward movement of
the plunger closes the suction valve and the air is forced through a discharge valve,
perhaps located in the plunger itself. An automobile fuel pump is an example of a
diaphragm pump.
3-7
-------
Diaphragm
Piston
Discharge valve
Diaphragm
Suction valve
Figure 3-7. Diaphragm pump.
Centrifugal Pump Operation
Centrifugal pumps (or fans) employ centrifugal force to move air. The simplest
form of this type of pump consists of an impeller rotating in a volute ("snail s
shell") casing (see Figure 3-8). The rotation of the impeller creates a decreasing
pressure at the impeller "eye", causing air to be drawn into the pump. Air drawn
into the center of the impeller is "picked up" by the vanes and accelerated to a
high velocity by rotation of the impeller. It is then discharged by centrifugal force
into the casing and out the discharge nozzle.
Suction
X
/
/
Discharge
nozzle
Impeller
eye
Impeller
Volute
Figure 3-8. Centrifugal pump operation.
3-8
-------
Centrifugal pumps encountered in air sampling can be divided primarily into
three ategorie^: (a) radial flow, (b) axial flow, and (c) mixed flow. Centrifugal
pump may also be classified into single-stage or multi-stage. Single stage indicates
a pump in which the total head is developed by one impeller; multistage indicates
a pump having two or more impellers acting in series in one casing (13).
"Driving Forces" for Pumps
All pumps have at least one common characteristic - they have movable parts. The
movement of the part, is the basis for the transfer of the particular fluid_ot
intelt For the parts to move there must be a "driving force." Driving forces for
pumps can be categorized as: (a) manual, and (b) motors.
Some air movers in the general category of pumps can be operated manually. A
hand-operated hypodermic syringe (piston pump) and a tiro pump using a toot
pedal -e examples. The hand-operated, portable MSA Midget Impmger Sampler
has seen much L. It is operated by a hand-cranked, 4-cylinder pump that draws
-air through a small glass nozzle at a relatively high velocity. A relatively constant
flow rate can be obtained by use of this sampler. Two obvious disadvantages of
manually-operated pumps is that only small sample volumes can be collected, and
that sampling time is limited because of personnel requirements.
Electric motors operated by commercial power, motor-generation sets, or by bat-
teries are all used for driving air sampling pumps. When batteries are to be used as
the driving force, several factors should be considered, among which are the motor
power requirements and the required length of sampling. The length of the samp-
line time is important in relation to the life of the battery.
In instances where constant flow is required it is important that the driving force
for the pump be constant and not affected by environmental factors. If the driving
force is variable, measures may have to be taken to try to regulate ,, For example,
a x >ltage regulator may be required in conjunction with an electric motor that is
driving an air pump where variable voltage power sources are encountered.
Characteristic Curves
Pumps perform differently under different conditions; therefore, -characteristic
curves "showing the relationships between the various conditions affecting their
performance, are usually supplied by the manufacturer. The characteristic curves
of most interest in air sampling are those indicating the pressure-flow relationshlp.
See Figures 3-9 and 3-10.
3-9
-------
It
•~ I
-------
Selecting A Pump
It is evident that often a choice must be made between pumps. The selection may
be limited to certain categories of air movers due to the sampling rate required
On the other hand, if flow rate is not critical, a wide variety of air movers may be
applicable necessitating a more involved evaluation and choice. A comparison of
some of the advantages and disadvantages of certain types of pumps is contained
in Figure 3-11. A comparison of this type, although it is not necessarily complete in
all considerations, should be of value in responsible pump selection.
Piston pump
(reciprocating)
Diaphragm pump
(reciprocating)
Centrifugal pump
1. Can operate at high
suction pressure
2. Can be metered
1. Wide range of
capacities
2. No seal required
3. Good in continuous
operation
1. Small capacity
2. Seal required between
piston & piston chamber
3. Working parts such as
check valves and piston
rings may cause
difficulties
4. Pulsating flow
5. Moderate maintenance
1. Limited materials
of construction
2. Operation at limited
suction pressures
3. Pulsating flow
4. Periodic diaphragm
replacement
5. Moderate maintenance
1. No small capacities
2. Turbulence
3. Operational noise
1. Large range of
capacities
2. No close clearance
3. Can obtain high
suction heads by
multistages
4. Light maintenance
Figure 3-11. Pump comparison (1, 2, 7).
There are many features of commercially available pumps that may or may not
warrant consideration. Some features may have direct applicability for certain uses,
others may provide flexibility making the pumps more generally useable. In this
discussion only gages and continuous operation capability will be considered.
3-11
-------
Pump Gages
Many pumps have inlet vacuum gages and/or outlet pressure gages. These gages,
upon proper calibration, can be used to determine the approximate flow rate
through the pump. The flow rate determination can be made by use of the pump's
characteristic curve for the pressure-flow relationship (See Figure 3-12, or by direct
reading if the gage is calibrated in terms of cfm on its dial).
•ss
w °i
2.0 4.0 6.0 8.0 10.0 12.0 14.0
Vacuum, inches of Hg
1.6
1.4
1.2
1.0
.8
.6
.4
.2
0
5 10 15 20
Vacuum, inches of Hg
25
Figure 3-12. Pressure-flow relationship for metal bellows pumps
for vacuum.
Continuous Operation Capability
Pumps are often required to operate continuously for long periods of time (hours to
days) at high vacuum. Some pumps such as rotary or diaphragm pumps used in
ambient air monitors are capable of continuous operations at high vacuum, while
others would tend to "burn" themselves out. The importance of this capability
would be dependent on the sampling time required.
3-12
-------
Ejectors
Another classification of air movers is known collectively as ejectors. Ejectors are
also referred to as aspirators.
Principle of Operation
As depicted in Figure 3-13, ejectors operate according to the jet principle (1). At
the nozzle the pressure head of the driving force is converted into a high velocity
stream. The passage of the high velocity stream through the suction chamber
creates a decreased pressure (vacuum), thus drawing air into the chamber itself.
The incoming air is mixed with the high velocity driving force mixture and can be
ejected against moderate pressure through the diffuser.
Nozzle
Suction
chamber
Diffuser
Driving
Air out
Air in
High
velocity
driving
force
Figure 3-13. Ejector operation.
Driving Forces
The pressurized fluid that is converted into a high velocity jet stream in an ejector
may be of several types.
Some examples of fluids used are water, steam, compressed air or CO2, and
other gases such as freon (1, 7).
Sampling Use of Ejectors
Using ejectors as sampler air movers is adequate for lower sampling rates. The
sampling rates vary from a few liters per minute up to several cubic feet per
minute. The flow rate through an ejector can be regulated to some extent by
adjusting the nozzle opening. A limitation to the use of ejectors is that the
pressurized driving force may have a time limit on its available effectiveness, (i.e.,
If a pressurized gas cylinder is used, it may last only for a limited amount of time.)
3-13
-------
Liquid Displacement
Air movers that operate according to the principle of liquid displacement incor-
porate two sampling train components into one entity. In this case the liquid
displacement unit serves as the sample collection device and the air mover together
or may provide the air moving capability for a second device which may act as the
sample collection device.
Principle of Operation
Gravity flow of liquid from a container creates a vacuum within the container; thus
drawing air into the container to fill the displaced volume (see Figure 3-14).
Air in
Closed
valves
Liquid
Open
valves
Figure 3-14. Liquid displacement.
Sampling by Use of Liquid Displacement
Liquid displacement is usually used only for grab sampling. The sample volume
capacity is limited to the size of the liquid container. Some examples of liquids
used are water, mercury (Hg), and organic solvents.
Evacuated Flasks
Air moving by use of evacuated flasks is another example of having the sample col-
lection device and the air mover combined into one unit.
Principle of Operation
A flask is evacuated by a vacuum pump to a very low pressure that must be deter-
mined. The flask is sealed and transported to the sampling location. When a valve
is opened on the flask, the surrounding air moves into the flask because of the
pressure differential. On closing the valve, the sample is confined for subsequent
analysis.
3-14
-------
Sampling by Use of Evacuated Flasks
Evacuated flasks are usually used only for grab sampling; consequently, only
relatively small sample volumes can be collected. Careful consideration should also
be given to the possibility of flask "implosion" when glass containers are used, and
appropriate protective means should be undertaken.
Flow Rate Control
Control by Diversion
The principle of flow diversion is simply that the air moved by the air mover is not
all passed through the sample collection device. As depicted in Figure 3-15, a
"bleed" valve control in the sampling train allows the variation of the actual flow
through the sample collection device. The position of the flow measuring device is
such that it measures only the flow passing through the sample collection device.
Flow
Measuring
Device
Air in
Sample
Collection
Device
. "Bleed
\
" valve X
/
V
Air
Mover
Air out
t
Air in
Figure 3-15. Flow rate control by diversion.
Resistance Control
In most air moving devices the flow rate decreases as the resistance it must over-
come increases. Examples of this are depicted in Figures 3-9 and 3-10, which show
sample characteristic curves for several pumps. The flow rate of the air mover can,
therefore, be regulated by controlling the resistance it must overcome. A common
method of control is to partially close a valve in the intake line, thus creating a
greater resistance. Reproducible flow control can be accomplished by using needle
valves with resettable marking.
3-15
-------
Driving Force Control
It may be possible to control the air mover driving force, thereby controlling the
rate of air flow. Adjustment of the nozzle opening on an ejector can be considered
as a driving force control, because it affects the velocity of pressurized gas stream.
The velocity of the gas stream, in turn, affects the suction pressure.
Another example of driving force control is electric motor speed regulation. Th.s
can be accomplished on some pump motors by use of a variable transformer, which
controls the amount of power sent to the motor. These variable transformers art-
known by several names, such as variacs and powerstats.
Flow Rate for Sampling
After a particular flow rate has been selected for sampling, and after it has been
set for the sampling train, it is usually necessary to maintain the flow at exactly
that rate (14).
Need for Control
A variation of the desired flow rate can be caused by a variation in the air mover
driving force or a variation in the resistance to air flow. The resistance to air flow
is equivalent to the head the pump must overcome (suction head). Some sampling
conditions affecting flow variation have been mentioned above. Flow variation
during sampling may affect more than the determination of the volume of air
sampled; it may also affect the performance of the sample collection device. For
example, the absorption rate of a bubbler may be altered by a variation of the flow
rate throught it.
*
Control Mechanisms
Many flowrate control mechanisms operate by keeping the effective resistance that
the pump must overcome at a constant value (7, 10, 11). In other words, as
depicted in Figure 3-16, the pressure drop (A/?) from the environment being
sampled to the pump intake is held constant.
Air in
Sample
collection
device
Flow
measuring
device
Flow
regulator
-
Pump
Air out
Ap
APl
Apz
Ap3
Ap«
Ap4
•HH-
Figure 3-16. Flow rate control mechanism.
3-16
-------
Sources of pressure drop in a system of this type include:
• the pressure drop across the sample collection device (A/?,),
• the pressure drop across the flow measuring device (/
• the pressure drop across the flow regulator (Api),
• the pressure drop due to friction in connecting lines
Friction losses and flow measuring losses are usually considered as being
constant.
The resulting relationship can be shown in Equations 3-1, 3-2, 3-3, 3-4.
(Eq. 3-1) Ajb = Ap\ + Ap2 + A^3 + £A/?4
(Eq. 3-2) Ap2 + £A/?4 = constant
(Eq. 3-3) Ap — Ap2 — £A/?4 = Ap adjusted
(Eq. 3-4) A^J adjusted = Ap\ + Api
Initially in the sample collection, the sample collection device resistance (A/>,) is
low; therefore, the regulator would have to offer a higher resistance (Aps). As
sampling proceeds, the sample collector resistance (A/),) increases; thus
automatically lowering the flow regulator resistance (A/?,) and keeping Apad, at a
constant value. The regulator resistance control may be actuated by several
mechanisms, usually a pressure drop directly related to time rate.
Another type of control mechanism accomplishes control by varying the pump
motor speed. As the pressure drop across the sample collection device increases, a
switching arrangement increases the pump motor speed thus drawing a constant
air flow.
Summary
The air mover in a sampling train is certainly an important component. There are
many factors for consideration involved in the selection of an air mover. Air movers
can be classified according to their principle of operation: pumps, ejectors, liquid
displacement, and evacuated flasks being some of the categories. There also may
be further subdivisions of the general categories of air movers. In many instances
the ability to change the flow rate of the air mover and then to keep it constant at
that rate are important.
3-17
-------
References
,,n do, V. C. and Harris, William B. Air Movers. Air Sampling Instruments.
-.. ,i ;c an Conference of Governmental Industrial Hygienists, fourth edition.
p. B 1-1, 1972.
2 V t< hell, Charles B. Which Pump and Why? Fluid Flow in Practice, chpt. 4.
\( u \ ork: Reinhold Publishing Corporation, 1956.
3. Robson, Charles D., and Foster, Kirk E. Evaluation of Air Paniculate
Campling Equipment. American Industrial Hygiene Association Journal.
23-404, Sept.-Oct. 1962.
4. Babbitt, Harold E., and Doland, James J. Water Supply Engineering, 5th edi-
tion, chpt. 15. New York: McGraw-Hill Book Company, Inc. New York.
". Salvato, Joseph A. Environmental Sanitation, p. 138. New York: John Wiley
and Sons Inc., 1958.
b Kristal, Frank A., and Annett, F. A. Pumps, chpt. 1. New York: McGraw-Hill
Book Company, Inc., 1953.
Stern, Arthur C. Air Pollution, vol. 1, chpt. 1. New York: Academic Press,
1962.
Reprinted from the Standards of the Hydraulic Insti u'e, tenth edition.
Copyright 1955 by the Hydraulic Institute. 122 Eas, 4-nd St. N.Y., N.Y.
10017.
>. Daughu-iy, A. B. and Ingersoll, A. C. Fluid Mechanics, chpt. 18. New York:
McGraw-Hill Book Co., Inc., 1954.
Harrison, Walter K., Jr., Nader, John S. and Fugman, Frank S. Constant Flow
Regulators for the High Volume Air Sampler. American Industrial Hygiene
Association Journal. 21:115, 1960.
, ,. Schmidt, A. C., and Wiltshire, L. L. A Constant Flow Suction Unit —for
Aerosol Sampling Work. American Industrial Hygiene Association Quarterly,
June 1955.
1 '. Lintak, B. G. Instrument Engineers Handbook, vol. 1. Philadelphia: Chilton
Book Company, 1969.
1 in- Randolph Company. Catalog No. 38., March 1971.
Stem, Arthur C. Air Pollution, vol. Ill, pp. 161-2. New York: Academic
Press, 1976.
i ,. i ivlor, R D Report UCRL-51239, TID-4500, UC-41. Lawrence Radiation
Lab. University of California. Levermore, California, 1972.
3-18
-------
Air Measuring Instruments
Introduction
The accuracy and precision of a given method for the determination of the concen-
tration of an air pollutant is based on two factors:
• the accuracy and precision of the sampling method;
• the accuracy and precision of the analytical method.
Examine the following term:
Mass of Pollutant \
Volume of air sampled /
You can see that the "fig" (mass of pollutant) term is a result of the analytical
process, while the "M3" (volume of air sampled) term is a result of the sampling
process. Not only is the mass of pollutant important in the final calculation of
jUg/M3, but the volume of air sampled is also important. We will concentrate in
this section on the measurement of the volume of air sampled.
The accuracy and precision of the sampling method depends upon these factors:
• the environmental conditions during sampling,
• the collection efficiency of the sampling method,
• the accuracy and precision of the flow rate measurement, and
• any interferences.
The determination of the volume of air sampled, V, generally involves a
measurement of flow rate, (), and sampling time, t.
(Eq. 3-5) V=Qxt
The flow rate during atmospheric sampling can be measured by a variety of air
measuring devices. These devices and their calibrations will be discussed in this
chapter.
Types of Air Measuring Devices
Air measuring devices can be broadly classified into three categories:
• volume meters,
• rate meters,
• velocity meters.
Volume meters measure the total volume, V, of gas passed through the meter
over some specified time period. If the time period, t, is measured with a timing
device, flow rate can be calculated by:
3-19
-------
Rate Meters measure the time rate of flow through them. Flow rate is measured
through some property of the gas.
Velocity meters measure the linear velocity, u, of a gas in a duct. Volumetric
flow rate can then be calculated by measuring the cross-sectional area, A, of the
duct through which the gas is flowing, by:
= A Xu
Calibration
Air that is to be sampled often is moved at a known rate over a known time
period. The determinative process used to establish this known flow rate and
known time period is a form of calibration. Remember that it has been said before
that V— Qxt, or that volume sampled, V, is the product of flow rate, Q, and time
t. The calibration process applied to both the flow rate and time allows the
accurate determination of volume.
The frequency with which calibration occurs depends upon a number of condi-
tions. Some of these are:
• instrument use —
what are the conditions under which the instrument is used?
• instrument users —
how many different people use the instrument? what are the qualifications
of the people?
• instrument characteristics —
how often does the instrument require calibration under controlled
laboratory conditions? how sensitive is the instrument?
The basic equipment required for calibrating air flow measuring instruments
include a standard meter, an air mover, and often a source of constant power.
Standard meters are of three types:
• primary standards
• intermediate standards
• secondary standards.
Primary standard meters are those whose volumes can be determined by
measurement of internal physical dimensions alone. The measured internal dimen-
sions are regular, and accuracies better than ±0.30% can be achieved.
Intermediate standards are those standards that cannot easily be calibrated by
measuring physical dimensions, but accuracies of ±1—2% can be achieved.
Intermediate standards are calibrated against primary standards. Secondary stan-
dards are those calibrated against primary or intermediate standards under known
conditions of gas type, temperature, and pressure. Accuracies less than 5% are
achievable.
3-20
-------
Volume Meters
Volume meters measure the volume of gas passing through the meter. When
coupled with a timing device (like a calibrated stopwatch) flow rate (volume/time)
can be calculated. There are seven volume meters that are in common use in air
sampling and analysis.
The Spirometer (or "Bell Prover")—Primary Standard
The spirometer consists of a cylinder of known volume, closed at one end, with the
open end submerged in a circular tank of fluid (Figure 3-17). The cylinder can be
opened or closed to the atmosphere by a valve. As the cylinder is lowered into the
water, the water displaces the air and causes it to be discharged from the cylinder;
the rate of discharge can be regulated.
Floating
bell
Counter-
weights
Vertical
displacement
scale
- s Pointer
Inlet or outlet
fixture
'M-- Water
Figure 3-17. Spirometer.
The volume of the cylinder is determined from its dimensions. A counter weight
and cycloid counterpoise allow pressure differentials across the spirometer as low as
0.02 inches of water (Figure 3-18).
3-21
-------
_Cycloid counterpoise
Isolating liquid
41 Gas
Shut-off lever
Figure 3-18. Orthographic and cross-sectional views of a 5-ft1 spirometer. (5)
The volume of air passed through a spirometer is given by the following formula:
(Eq. 3-6) V= —
Where: V= volume of air passed through spirometer
ir = a constant = 3.14
d = diameter of bell
h = vertical displacement of bell
The fluid in the spirometer should be at the same temperature as the room. This
is to ensure that the fluid and the air will be in thermal equilibrium and thereby
minimize spirometer fluid evaporating into the air. This thermal equilibrium also
simplifies volume corrections since temperature is constant during the calibration
procedure. This is true for water, but some spirometers use oil. The real impor-
tance of thermal equilibrium is that the air displaced from the bell must be at the
same temperature as the room for volume calculations. The pressure inside the bell
is also brought into equilibrium with room conditions.
Once the volume of air is determined using room conditions, a conversion to
standard conditions must be made to determine the true volume of air that has
passed through the spirometer. This conversion to standard conditions is made
using the following formula (see Equation 2-7):
P, \/298°K\
2~ '
760 mm Hg T,
3-22
-------
The spirometer is usually calibrated by the manufacturer against an NBS "cubic-
foot" bottle. If an owner suspects that his spirometer is in error he can check the
calibration with an NBS certified "cubic-foot" bottle or by a "strapping"
procedure.
The "strapping" procedure involves the measurement of the dimensions of the
bell with a steel tape and subsequent calculation of the volume. Experienced per-
sonnel routinely obtain accuracies of +0.2%, when calibrating a spirometer by the
"strapping" procedure. Nelson gives a very detailed procedure for performing the
"strapping" procedure (5). Manufacturers of spirometers include example calcula-
tions in the literature accompanying their instruments.
Flow rates can be measured by timing the volume of air passing to or from the
spirometer and determining the rate of flow. The spirometer is simple, inexpensive,
dependable and is used almost solely as a primary standard for calibration of other
types of flow and volume measuring devices. Because the spirometer can be pro-
duced in large sizes it has typically been used to calibrate roots meters, which are
positive displacement meters.
Displacement Bottle—Primary Standard
The displacement bottle consists of a bottle filled with water (or some other liquid)
and a tube through which air can enter the bottle (Figure 3-19). As the liquid in
the bottle is drained or siphoned out, air is drawn in, to take the place of the
volume of liquid lost. The volume of gas sampled is equal to the volume of liquid
displaced. The fluid in the displacement bottle should be in thermal equilibrium
with the room temperature. This equilibrium will ensure no liquid evaporation
from the bottle water to the air, and simplify volume corrections for T and P. The
volume of displaced liquid can be measured with a graduated cylinder or Class A
volumetric flask, depending on how accurately the volume needs to be measured.
Accuracy can range from 1-5% depending on what measuring device is used.
Air
Air
Water
out
Water
out
Marius bottle Siphon bottle
Figure 3-19. Water displacement bottles.
3-23
-------
Again, once the volume of air has been determined at room conditions, it should
be converted to the volume at standard conditions. This conversion uses the rela-
tionship previously stated in Equation 2-7.
Frictionless Pistons—Primary Standards
The soap-bubble meter and the mercury-sealed piston are two frictionless pistons
that will be discussed in this section. Accurate and convenient measurement of
flows between 1 and 1000 ml/min can be made with a soap bubble meter.
Mercury-sealed pistons are available that can accurately measure flows from
100 cmVmin to 24,000 cmVmin.
Soap-Bubble Meter
A bubble meter consists simply of a cylindrical glass tube with graduated markings,
usually in milliliters. Inverted burets are often used as soap-bubble meters (Figure
3-20); however, burets cannot be used with anything other than a vacuum source.
Simple bubble meters (Figure 3-21) can be purchased, although the basic design
can be made conveniently by a competent glass blower.
To vacuum source
Inverted buret
Stopcock
Moving bubble
Soap solution
Figure 3-20. Soap-bubble meter.
3-24
-------
Gas out (2)
Bubble breaker
Gas out
Gas in (1)
Graduated tube
Soap bubble
Rubber bulb
Soap solution
Rubber bulb
Gas in (3)
(a)
Figure 3-21. Soap-bubble meters (a) with bubble breaker and
capability of handling vacuum at (2) or pressure
at (1); (b) one capable of handling only pressure at (3).
The soap-bubble meter is one of the simplest of primary standards. The inside walls
of the tube are wetted with a soap solution. A bubble is formed by touching the tip
of the buret to the soap solution, as in Figure 3-20, or by squeezing the rubber
bulb until the soap solution is raised above the gas inlet (Figure 3-21).
Either a vacuum at the top or slight positive pressure at the bottom of the tube
moves the bubble (a frictionless piston) up the tube. By timing this movement and
noting the volume traversed by the bubble, over the measured time span,
volumetric flow rate can be calculated.
The volume measured by a soap-bubble meter must be corrected for two condi-
tions. First, if the room temperature and pressure are different from standard
atmospheric conditions the volume must be corrected by the relationship previously
stated (see Equation 2-7):
(Eq. 3-7)
'298°KN
1760 mm Hg A Tl
3-25
-------
Secondly, the measured volume can be slightly larger than the actual volume
because water from the bubble evaporates into the gas behind the bubble. If the
gas behind the bubble has a relative humidity greater than 50%, the error is small.
If the gas is dry, the error can be large and must be corrected by the following
formula:
(Eq. 3-8) Vc=Vmra,
Where: Vc = corrected volume
Vmea< ~ measured volume
Ph - atmospheric pressure (mm Hg)
Pw = vapor pressure of water at the room temperature (mm Hg)
Note that Pb and Pw must have the same units.
Vapor pressure tables for water can be found in almost any chemistry handbook.
Soap-bubble meters can be calibrated by measuring the dimensions of the tube;
iowever, poor control on glass dimensions in manufacturing make this inaccurate.
The bubble meter is usually calibrated by filling the tube with a liquid (water or
mercury for example), draining the liquid from the top graduation to the bottom
graduation. The volume or weight of the collected liquid can be measured. With
proper corrections for temperature this calibration is accurate. The soap-bubble
meter should only be used to measure volumes between graduations that have been
calibrated.
The bubble meter is used almost exclusively in laboratory situations for calibra-
tion of other air measuring instruments. In average laboratory conditions the soap-
bubble meter is accurate to about + 1 % depending on how accurately it is
calibrated. Accuracy decreases for flows below 1 ml/min and above 1 f/min mainly
because of gas permeation through the bubble. A detailed description of the
calibration and use of bubble meters has been published by Levy (6). Increased
accuracies have been reported for bubble meters fitted with automatic sensing
devices that start and stop a timer (7).
Mercury-Sealed Piston
If a bubble meter is unsuitable, an electronically actuated mercury-sealed piston
may meet the need. Although the mercury-sealed piston is expensive, its accuracy
(±0.2% for time intervals greater than 30 seconds) and simple operation make it
an extremely useful tool.
The mercury-sealed piston consists of a precision-bored, borosilicate glass
cylinder with a close fitting polyvinyl chloride piston (Figure 3-22). The piston and
cylinder wall are sealed with a ring of mercury that stays in place because of its
high viscosity and the closeness of the fit between the cylinder and piston. Gas
entering the solenoid valve is vented until the measurement cycle is actuated. When
the measurement cycle is started the solenoid valve closes the vent, allowing gas to
enter the cylinder.
3-26
-------
Thermometer
Vo
i
Me
<
lu
:hl
rci
Ga
f\
1£
me graduations
• —
Polyvinyl
oride piston v^
s
Manometer
— pSk —
^*^l
^ ^
\
---.
\
_rz
c
F
J
i
JL
LGas
Movable
proximity coil
H
Precision-bore
glass cylinder
Stationary
proximity coil
Solenoid valve
Leveling screws
Figure 3-22. Mercury-sealed piston volume meter.
A timer is started and stopped as the mercury seal passes the lower and upper
proximity coils (metal detectors). The volume displaced can be set by adjusting the
upper proximity coil. The volume is corrected to standard conditions using the
pressure drop across the piston (usually <3 inches of water). The measured time
and the corrected volume can be used to calculate volumetric flow. The system
shown in Figure 3-23 has a reported accuracy of ±0.2%.
Calibration of the mercury-sealed piston volume meter is usually performed by the
manufacturer. The borosilicate glass cylinder is bored to a precise diameter. The
inside diameter is air gaged at least every inch to check for consistency. Before the
instrument is sent out it is compared to a standard-mercury-sealed piston volume meter
that is traceable to NBS. If a multi-cylinder unit is purchased then the cylinders
must be aligned. One cylinder is chosen to be correct and all others are aligned
with the set screw located on top of the piston (Figure 3-24), which changes the
displaced volume slightly.
The mercury-seal can be broken by erratic movement of the instrument. For this
reason the mercury-sealed piston instrument is used as a primary standard in
laboratory settings. Mercury-sealed piston volume meters are available for accurate
flow measurement over a wide range (100 cmVmin to 24,000 cmVmin).
3-27
-------
Timer
Glass
cylinder
Upper
proximity
coil
(moveable)
PVC piston
Lower
proximity
coil (fixed)
Control
panel
Figure 3-23. Calibrator console, front view (Brooks model 1051).
(Courtesy Brooks Instrument Division, Emerson Electric Co.)
Set screw
Groove
Figure 3-24. PVC piston for mercury-sealed piston volume meter.
3-28
-------
Wet Test Meter—Intermediate Standard
The wet test meter consists of a series of inverted buckets or traps mounted radially
around a shaft and partially immersed in water (Figure 3-25c). The location of the
entry and exit gas ports is such that the entering gas fills a bucket, displacing the
water and causing the shaft to rotate due to the lifting action of the budget full of
air. The entrapped air is released at the upper portion of the rotation and the
bucket again fills with water. In turning, the drum rotates index pointers that
register the volume of gas passed through the meter (Figure 3-25b).
After the meter is leveled, the proper water level is achieved by using the filling
funnel, fill cock, and drain cock (Figure 3-25a) to bring the meniscus of the water
in touch with the tip of the calibration index point. The calibration gas should be
passed through the meter for one hour to saturate the water with the gas. The
water in the meter should be at the same temperature as the surrounding
atmosphere. If any water is added, sufficient time must be allowed for complete
equilibration.
Calibration point
for water level
Water manometer
Figure 3-25. Wet test meter
Drain cock
Rotating partitioned drum
Gas outlet
Direction of
rotation
Water
level
Gas inlet
(c)
3-29
-------
Once the water level is set and the meter is equilibrated, the wet test meter is
ready for calibration (or use if it is already calibrated.) An accurate calibration of
a wet test meter can be done with a spirometer (Figure 3-26).
. Oil-temperature thermometer
Oil manometer /-Gas-temperature thermometer
Control valve
| To vacuum
Wet test meter
Figure 3-26. Setup for calibrating a wet test meter against
a spirometer. (Ref. 5)
The wet test meter can also be calibrated against a mercury-sealed piston as
shown in Figure 3-27.
Gas
Mercury-
sealed
piston
Gas
Wet test meter
Figure 3-27. Calibration of wet test meter (WTM)
against a mercury-sealed piston.
Enough gas is drawn through either system (Figure 3-26 or 3-27) to turn the wet
test meter at least three revolutions and to significantly move the spirometer drum
or piston This measurement is made several times. Atmospheric pressure and
temperature, and the temperature and pressure differential for both the wet test
3-30
-------
meter and calibrating device, are needed to correct the volume to standard condi-
tions (taking pressure differentials into consideration). Since both the calibration
device and the wet test meter are measuring gas saturated with water vapor in
Figures 3-26 and 3-27, there is no need to correct for water vapor effects.
If a wet test meter is used to measure a dry gas stream, a significant error is
introduced if the measured volume is not corrected to dry conditions. This correc-
tion is the same as for the soap-bubble meter correction:
Vc=Vn
A simple calibration check can be performed using a displacement bottle as
shown in Figure 3-28. After all the water is thermally equilibrated, the wet test
meter is properly set up, and the drain tube of the displacement bottle is filled, the
pinch clamp is opened allowing 2 liters of water to drain into a 2-liter class A
volumetric flask. The corresponding wet test meter readings are taken. This is
repeated several times (usually 4).
WTM
Displacement
bottle
2 liter class A
volumetric
flask
Figure 3-28. Calibration of wet test meter with displacement bottle.
The calibration data can be used to:
Option 1: Draw a multipoint calibration curve for flow for Figures 3-26 and
3-27 set ups.
Option 2: Adjust the calibration index point so that the meter volume is correct.
Option 3: Calculate a correction factor for the wet test meter.
Wet test meters should check within ±0.5% if option 2 is used.
3-31
-------
All volumes measured by a wet test meter should be corrected to standard condi
lions by Equation 2-7.
Wet test meters are used as transfer standards because of their high accuracy
(less than + 1%). Because of their bulk, weight, and equilibration requirements
they are seldom used outside a laboratory setting. Wet test meters are useful for
laboratories that need an accurate standard yet do not have the funds or space lor
a spirorneter or mercury-sealed piston. Wet test meters can be used to measure tlo.v
rates up to 3 rev min, at which point the meter begins to act as a limiting orifice
and obstructs the flow. Typical ranges of wet test meters are 1, 3, and 10 C/rev.
Roots Meter—Intermediate Standard
The Roots meter is a positive displacement rotan type meter toi measuring \uliune
flow. It is suitable for handling most types of clean, common gases. It is not
suitable for handling liquids, and its operation can be impeded by excessive par-
liculates carried in the gas stream.
Roots meters consist basically of two oppositely rotating impellers of two-lobe or
"Figure 8" contour, operating within a rigid casing (Figure 3-29). The casing is
arranged with inlet and outlet gas connections on opposite sides. Impeller contours
are mathematically developed and accurately produced, and are of such form that
a continuous seal without contact can be maintained between the impellers at all
positions during rotation. To accomplish this, the correct relative impeller positions
are established and maintained by precision-grade timing gears. Similar seals exist
between the tips of the impeller lobes and the two semicircular parts of the meter
casing. As a result of this design, the gas inlet side of the meter is always effectively
isolated from the gas at the outlet side of the impellers. Consequently, the impellers
can be caused to rotate by a very small pressure drop across the meter.
Inlet
Position 1 Position 2 Position 3
Figure 3-29. Principle of gas flow through the Roots meter.
3-32
-------
The rotation of the impellers is in the direction indicated in Figure 3-29, and as
each impeller reaches a vertical position (twice in each revolution) it traps a known
specific volume of gas between itself and the adjacent semicircular portion of the
meter casing at A and B (Figure 3-29, Position 2). Thus, in one complete revolu-
tion the meter will measure and pass four similar gas volumes, and this total
volume is the displacement of the meter per revolution.
The displacement of volume of the roots meter is precisely determined by the
manufacturer, both by calculation and by testing it using a known volume of air or
other gas. Roots meters are usually calibrated against large spirometers prior to
shipment. Users do not usually have a way to calibrate roots meters and must
depend on the supplied calibration data. Volumetric accuracy of the Roots meter is
permanent and nonadjustable (except for linkage adjustment), because its
measuring characteristics are established by the dimensions and machined contours
of nonwearing fixed and rotating parts.
The revolutions of the impellers are indexed with the meter reading calibrated in
a volume unit (i.e., ft3 or m3). Units are sold that have temperature compensation
devices but corrections to standard temperature and pressure conditions are easily
made with the previously mentioned formula (see Equation 2-7) (pressure drop
across the roots meters should be taken into consideration):
/ P, \/298°K\
The symbol P,, in this instance, is the atmospheric room pressure (Pb, in mm
Hg) minus the pressure drop across the roots meter Ap, in mm Hg.
(Eq. 3-10) P, = P»-A/»
The metering unit is magnetically coupled to the impellers. The entire counting
unit is enclosed in a plastic cover. The cover also holds an oil that lubricates the
metering device. The proper oil level is set by the inscribed oil level lines on the
ends of the plastic covers. The user of a roots meter must be careful not to severely
tilt the roots meter when oil is in the plastic cover as this can force oil into the
impeller casing. If the oil gets into the impeller casing, flushing with kerosene can
remove the oil.
Although roots meters are widely used in industrial applications, they have been
used almost exclusively as the standard for high-volume sampler flow rate in
atmospheric sampling applications.
Dry Test Meter — Intermediate Standard
Dry test meters are an improvement over the more common dry gas meters (Figure
3-30). Dry gas meters (a secondary standard) are most commonly used in residen-
tial and industrial settings to measure gas flow (e.g., natural gas). The dry test
meter (an intermediate standard) works on the same principle as the dry gas meter,
3-33
-------
(a secondary standard) but a different indexing method (read out) makes it more
accurate (usually ± 1 to 2% when new). The dry test meter shown in Figure 3-31
shows the new readout mechanism.
Figure 3-30. Dry gas meter.
(Courtesy of Western
Precipitation Division,
Joy Manufacturing Company)
Figure 3-31. Dry test meter.
3-34
-------
The interior of the dry test meter contains two or more movable partitions, or
diaphragms attached to the case by a flexible material so that each partition may
have a reciprocating motion (Figure 3-32). The gas flow alternately inflates and
deflates each bellows chamber, simultaneously actuating a set of slide valves that
shunt the incoming flow at the end of each stroke. The inflation of the successive
chambers also actuates, through a crank, a set of dials that register the volume of
gas passed through the meter.
Gas
Gas
Sliding
valves
Figure 3-32. Working mechanism of dry test meter.
The dry test meter is calibrated against a spirometer, mercury-sealed piston, or
displacement bottle similar to the wet test meter. One big advantage of the dry test
meter over the wet test meter is that no correction for water vapor is needed. If the
dry test meter is off calibration by more than 2% it can be corrected by adjust-
ment of the meter linkage. If linkage adjustment cannot correct the problem, then
the dry test meter must be returned to the manufacturer for repairs.
Dry test meters are used in the field as well as laboratory calibrations. Since the
dry test meter does not contain water, it is lighter and easier to use than the wet
test meter. Also, the dry test meter is more rugged than the wet test meter.
Accuracy of the dry test meter does, however, worsen with age.
3-35
-------
Rate Meters
The most popular devices for measuring flow rate are the rate meters. Rate meters
measure, indirectly, the time rate of the fluid flow through them. Their response
depends on some property of the fluid related to the time rate of the flow.
Variable Pressure Meters—Head Meters
Head meters are those in which the stream of fluid creates a significant pressure
difference that can be measured and correlated with the time rate of flow. The
pressure difference is produced by a constriction in the stream of flow causing a
local increase in velocity.
Orifice Meter—Noncritical—Secondary Standard
An orifice meter can consist of a thin plate having one circular hole coaxial with
the pipe into which it is inserted (Figure 3-33). Two pressure taps, one upstream
and one downstream of the orifice, serve as a means of measuring the pressure
drop, which can be correlated to the time rate of flow. Watch jewels (9), small
bore tubing, and specially manufactured plates or tubes with small holes have been
used as orifice meters. The pressure drop across the orifice can be measured with a
manometer, magnehelic, or pressure gage.
Upstream *• —^--g^^—' — *• Downstream
* _ _ —-•• jS"*?_-^~* -J*~ —- — >
Pressure taps
Figure 3-33. Orifice meter.
Flow rates for an orifice meter can be calculated using Poiseuille's Law; however,
this is not done for practical use. Instead the orifice meter is usually calibrated
with either a wet or dry test meter or a soap bubble meter. A typical calibration
curve is shown in Figure 3-34.
3-36
-------
1
O
Figure 3-34. Typical orifice meter calibration curve.
Calibration curves for orifice meters are nonlinear in the upper and lower flow
rate regions and are usually linear in the middle flow rate region.
Orifice meters can be made by laboratories with a minimum of equipment. They
are used in many sampling trains to control the flow. Care must be exercised to
avoid plugging the orifice with particles. A filter placed upstream of the orifice can
eliminate this problem. Orifice meters have long been used to measure and control
flows from a few ml/min to 50 f/min.
Orifice Meter—Secondary Standard
If the pressure drop across the orifice (Figure 3-33) is increased until the
downstream pressure is equal to approximately 0.53 times the upstream pressure
(for air and some other gases), the velocity of the gas in the constriction will
become acoustic or sonic. Orifices used in this manner are called critical orifices.
The constant 0.53 is purely a theoretical value and may vary. Any further decrease
in the downstream or increase in the upstream pressure will not affect the rate of
flow. As long as the 0.53 pressure relationship exists, the flow rate remains constant
for a given upstream pressure and temperature, regardless of the value of the
pressure drop (Figure 3-34). The probable error of an orifice meter is in the
neighborhood of 2 percent.
Only one calibration point is needed for a critical orifice. The critical flow is
usually measured with a soap-bubble meter or a wet or dry test meter. Corrections
for temperature and pressure differences in calibration and use are made with the
following formula:
/ P v T \ V&
/ "l * 1 2 \ '*•
(Eq-3-11) &=&(lT^F.)
3-37
-------
Where: Q=flow
P = pressure
T - temperature in °K
1 = initial conditions
2 = final conditions
The same formula can be used to correct orifice meter flows to standard condi-
tions by substituting Pz = 760 mm Hg and T2 = 298°K. Note the square root func-
tion of T and P. Any time that rate meters are corrected for T and P, this square
root function is needed.
Critical orifices are used in the same type of situations as noncritical orifices.
Care must also be taken not to plug the orifice.
Venturi Meter—Secondary Standard
The vetituri meter consists of a short cylindrical inlet, an entrance cone, a short
cylindrical throat, and finally a diffuser cone (Figure 3-35). Two pressure taps, one
in the cylindrical inlet and one in the throat, serve to measure the pressure drop.
There is no abrupt change of cross section as there is with an orifice, thus the flow
is guided both upstream and downstream, eliminating turbulence and reducing
energy losses. Venturi meters are, of course, more difficult to fabricate. The
probable error of a venturi is 1 percent.
Pressure taps
Figure 3-35. Venturi meter.
The venturi meter is calibrated in the same manner as the orifice meter. The
calibration curve generated plots pressure drop across the venturi versus flow rates
determined by the standard meter.
Variable Area Meters
The variable area meter differs from the fixed orifice; the pressure drop across it
remains constant while the cross-sectional area of the constriction (annulus)
changes with the rate of flow. A rotameter is an example of a variable area meter.
3-38
-------
The Rotameter—Secondary Standard
The rotameter consists of a vertically graduated glass tube, slightly tapered in bore,
with the diameter decreasing from top to bottom, containing a float of the
appropriate material and shape (Figure 3-36). The fluid to be measured passes
upward through the conical tube, which is inserted in the flow circuit.
Section A A
Float•
Tapered tube
B-
Annulus
Float
Section B B
Figure 3-36. Rotameter.
A specially shaped float, with a diameter slightly greater than the minimum bore
of the conical tube, is carried upward by the passage of the fluid until it reaches a
position in the tube where its weight is balanced by the upward forces due to the
fluid flowing past it. A variable ring or annulus is created between the outer
diameter of the float and the inner wall of the tube. As the float moves upward in
the tube, the area of the annulus increases. The float will continue to move
upward until a pressure drop across the float, which is unique for each rotameter,
is reached. This pressure drop across the float is constant regardless of the flow
rate. A measure of the flow is noted by the float position on a vertical scale com-
pared with a calibration chart.
The flow rate through a rotameter can be calculated from the tube diameters,
float dimensions, float composition, and gas characteristics; this is not commonly
done for calibration purposes. A detailed development of the flow equations for
rotameters is contained in the appendix of this manual. Manufacturers generally
provide accurate calibration curves for rotameters; it is advisable, however, to
calibrate a rotameter under its operating conditions.
3-39
-------
Most rotameters are used and calibrated at room temperature with the
downstream side at atmospheric pressure. Corrections for pressure and temperature
variations can be made using the previously mentioned formula:
(Eq.3-12) Q.2
If a gas is measured with a different density from the calibration gas the flowrate
can be corrected using the following formula:
/62V/2
(Eq. 3-13) ^1 =
-------
The pilot tube actually measures the velocity pressure (Ap) of a gas stream. Gas
streamlines approaching a round object placed in a duct, flow around the object
except at point "P+" where the gas stagnates and the stagnation pressure (P+) is
found (Figure 3-37 and Figure 3-38a).
The static pressure in a gas stream is defined as the pressure that would be
indicated by a pressure gage if it were moving along with the stream so as to be at
rest or relatively "static" with respect to the fluid. The static pressure can be
measured as shown in Figure 3-38b.
The difference between the stagnation pressure (P+) and the static pressure (Ps) is
the velocity pressure differential (Ap). This is shown in Figure 3-38c.
Figure 3-37. Gas stagnation against an object.
VK/
(a) Stagnation
pressure
vr^
i
Ap
t
(b) Static
pressure
(c) Velocity
pressure
Figure 3-38. Pitot tube pressure components.
3-41
-------
Bernoulli's Theorem relates pilot tube velocity pressure (A£) to gas velocity in the
following equation:
(Eq. 3-15) v=KpCp
Where: v= velocity of the gas stream, ft/sec
T = absolute temperature, °R (°F+46Q}
P= absolute pressure, in. Hg
M= molecular weight of the gas, lb/lb-mole
Ap = velocity pressure, in. H2O
,, 1 I ft2 in. He Ibs/lb-mole , , , .
K = constant 85.49|/^ - 2 -- for the above dimensions
\ sec"- in. H2O °R
C,, = pitot tube coefficient, dimensionless
Pilot tubes are used extensively in ventilation work to measure air flow in ducts.
Literature sources describe pitot tubes in detail (2, 11). The standard and S-type
pilot tubes are the most commonly used.
Standard Pitot Tube— Primary Velocity Standard
The standard pitot tube (Figure j-^) consists of vv» >• > iceniric tubes. The center
tube measures the stagnation or impact pressure while the static pressure is
measured by the holes located on the side of the outer tube. The pitot tube must
be placed in the flowing air stream so that it is parallel with the streamlines. The
velocity pressure differential (Ap) can be measured with a U-tube manometer, an
inclined manometer, or any suitable pressure-sensing device. Only velocities greater
than 2500 ft/min can be measured with a U-tube manometer, while flows as low as
600 ft/min can be measured with a carefully adjusted inclined manometer. Stan-
dard pitot tube velocity pressures are typically 0.14 inches of water at 1500 ft/min
and 0.56 inches of water at 3000 ft/min.
The standard pitot tube was first calibrated against an orifice meter using
Bernoulli's Theorem. Repeated calibrations proved that different standard pitot
tubes have the same characteristic flow calibration. If the static pressure holes are 6
outer tube diameters from the hemispherical tip and 8 outer tube diameters from
the bend (Figure 3-39), then the Cp value in the previously mentioned formula is 1.
\PM /
Standard pitot tubes can be used to measure linear velocity in almost any situa-
tion except in particulate-laden gas streams. The particulates will foul the carefully
machined tip and orifices. The velocity of gas streams with high paniculate con-
centrations can be measured better with an S-type pitot tube.
3-42
------- An error occurred while trying to OCR this image.
-------
S-Type Pitot Tube
The S-type pilot tube consists of two identical tubes mounted back to back (Figure
3-40). The sampling end of the tubes are oval with the openings parallel to each
other. In use, one oval opening should point directly upstream, the other directly
downstream. The tubes should be made of stainless steel or quartz if they are used
in high temperature gas streams. The alignments shown in Figure 3-40 should be
checked before use or calibration as this may cause variations in the calibration
coefficient (C,,).
1.05 Dt
-------
Mass Flow Meter—Secondary Velocity Standard
Mass flow meters work on the principle that when a gas passes over a heated sur-
face, heat is transferred from this surface to the gas. The amount of current
required to keep the surface at a constant temperature is a measure of the velocity
of the gas. Since the amount of heat transferred depends on the mass and velocity
of the gas, these meters measure mass flow rate.
Atmospheric sampling applications of the mass flow meter are usually limited to
the measurement of volumetric flow. Since these devices measure mass flow directly
they should be calibrated against a primary, intermediate, or secondary volumetric
standard. The standard meter flow is corrected to standard conditions and com-
pared to the mass flow rate measured. No corrections for temperature and pressure
need to be made to the mass flow meter readings. Calibration must be done with
the same gas as will be measured in use because different gases have different
thermal properties.
Mass flow meters are most often used for flow measurement or as calibration
transfer devices in the field and laboratory. Their insensitivity to temperature and
pressure make them a useful tool for standard conditions measurement.
Summary
The calibration and use of flow measuring devices is basic to the measurement of
air pollutants. Most atmospheric measurements require the accurate knowledge of
either the total volume of air or the flow rate. Many devices exist to aid in the
measurement of volume and flow rate. Three types are most commonly used in
atmospheric sampling —volume, rate, and velocity meters. Standard devices exist at
the primary, intermediate, and secondary level.
Sample Problems
Problem 1. Volume Conversion to Standard Conditions
A volume of 20 m3 was drawn from a spirometer at 20 °C and 700 mm Hg. What
was the standard volume drawn?
Where: F2 = volume at condition 2
Vi = volume at condition 1 = 20 m3
Ft = pressure at condition 1 = 700 mm Hg
p2 = presure at condition 2 = 760 mm Hg
T, = temperature at condition 1 = 20°C + 273 = 293 °K
T2 = temperature at condition 2 = 25 °C + 273 = 298°K
3-45
-------
Problem 2. Orifice Conversion When Used
at Other Calibration Conditions
An orifice was calibrated at 21 °C and 760 mm Hg. It is to be used to calibrate a
sampler at conditions of 25 °C and 700 mm Hg. The flow at field conditions is
0.85 mVmin (30 CFM): (a) What is the flow rate at calibration conditions? (b) What
is the flow rate at standard conditions?
Where:
Q2 = Flow 2
Q, = Flow 1
P2 = Press 2
PI = Press 1
T2 = Temp 2
T, = Temp 1
Part a Part b
__ (cal. condt.)
0.85 MVmin
760 mm Hg
700 mm Hg
21° + 273° = 294°K
25° + 273° = 298°K
(STD condt.)
Qj, in Part a
760 mm Hg
760 mm Hg
298 °K
294 °K
(a) The flow rate at calibration conditions:
/700 mmHgx294°K\
MVmin)
= 0.81 MVmin
V760 mm Hgx2980K;
(b) The flow at standard conditions: (using orifice data at calibration
conditions and the flow derived in part a)
PiX298°K
760mm Hg XT, ,
760mmHgx298°K
760mmHgx294°K
Q=0.82 MVmin
3-46
-------
Problem 3. Conversion for Different Gas Used with a Rotameter
A rotameter was calibrated with air at 0°C and 760 mm Hg. The rotameter is now
to be used to add helium as a carrier gas at 0°C and 760 mm Hg. If the flow
reading at point X on the rotameter corresponded to 28.3 liters/min (1 ftVmin)
when air was used, what flow will point X correspond to when helium is used?
(Density of helium = 0.1785 g/t and for air = 1.2929 g/l
Where: Qj =flow with gas 2
Qi =floui with gas 1 = 28.3 f/min
o, = density of gas 1 = 1 .2929 g/g= (air)
Q2 = density of gas 2 = 0.1785 g/f (helium)
0.1785 g/f
= 28.3 (7.2431)**
(£., = 76.2 liters/min
Problem 4. Pump Capacity
Can a pump with the following capacity curve (Figure 3-41) be used to drive
critical orifices that have flow rates of (a) 50 cc/min and (b) 250 cc/min at
criticality? (Hint: atmospheric pressure = 30 in Hg)
For an orifice to be critical Pdown«ream^0.53
Pdavnstream
------- An error occurred while trying to OCR this image.
-------
References
1. Lapple, C. E., et al. Fluid and Particle Mechanics, 1st edition. Newark,
Delaware: University of Delaware, 1956.
2. Ower, E. and Pankhurst, R. C. The Measurement of Air Flow, 4th edition.
Elmsford, N.Y.: Permagon Publishing Company, 1966.
3. Encyclopedia of Instrumentation for Industrial Hygiene, University of
Michigan, Institute of Industrial Health, Ann Arbor, Michigan, 1956.
4. A.S.M.E. Research Publication: Fluid Meters, Their Theory and Application,
4th edition. New York: The American Society of Mechanical Engineers, 1937.
5. Nelson, G. O. Controlled Test Atmospheres, 1st edition. Ann Arbor, MI: Ann
Arbor Science, 1972.
6. Levy, A. The Accuracy of the Bubble Meter Method for Gas Flow
Measurements./ Sci. Instr. 41:449, 1964.
7. Frisone, G. J. A Simple and Precise Soap-Bubble Flow Meter. Chemist-Analyst
54:46, 1965.
8. Lodge, J. P. et al. The Use of Hypodermic Needles as Critical Orifices in Air
Sampling./ Air Poll. Cont. Assoc. 16:197, 1966.
9. Brenchley, D. L. Use of Watch Jewels as Critical Flow Orifices./ Air Poll.
Cont. Assoc. 22:967, 19r/2.
10. Corn, M. and Bell, W. A Technique for Construction of Predictable Low-
Capacity Critical Orifices. Am. Ind. Hyg. Assoc. J. 24:502, 1963.
11. Industrial Ventilation, 10th edition. Ann Arbor, MI: Edwards Brothers, 1968.
3-49
-------
Chapter 4
Particulate Sampling
Introduction
Paniculate matter is one of the more noticeable forms of air pollutants. Conse-
quently, particulates have been monitored more extensively and for a longer time
than other pollutants. Particulate monitoring has been performed mostly by filtra-
tion and gravity-induced settling for many years. Studies showing that the human
respiratory tract is size selective of particulates led to the development of samplers
for "respirable" particles. Most "respirable" paniculate samplers have relied on
inertial methods to collect samples. Precipitator samplers (both thermal and elec-
trostatic) are now of limited use in airborne paniculate sampling.
The major prerequisite in the design of any sampling train is the ability to
collect a representative sample. Particulate sampling trains for concentration, size
distribution, and chemical composition (among other parameters) have varied and
unique requirements. Optimal methods for each type of sampling are still
debatable. The discussion of particulate sampling in this chapter will be limited to
five collection mechanisms (gravity, filtration, inertial, inertial-filtration combina-
tions, and precipitators) and the advantages and disadvantages of each mechanism
will be discussed.
Principles of Inertial Collection
Introduction
Particle size distribution studies are of interest because of the adverse health effects
of particles in certain size ranges. Very large particles (10 /xm or larger) entrained
in inhaled air are removed in the nose and throat, and thus do not reach the
lungs. Very small particles (smaller than 0.1 jmi) pass into the lungs but are
exhaled and not retained in the lungs. The particles in the 0.1 10 fim size range
are generally thought to be respirable. Figure 4-1 shows the depth of penetration
into the lungs of particles of various size ranges.
It should be noted that particulates in the air seem to be distributed bimodally
with regard to particle size (6). This distribution appears to have peaks at about
0.4 /am and 10 /im (Figure 4-2). The larger particles are those that appear
naturally (dust, pollen, etc.). The smaller particles are usually marimade
(anthropogenic) consisting of products of combustion or process losses (see Figure
4-3). Hence, most manmade particles are in the respirable range and are usually
the particles that carry health hazards (i.e., heavy metals, polynuclear aromatics,
etc.).
4-1
-------
30 microns
9.2—30 microns
5.5—9.2 microns
3.3—5.5 microns
Trachea & primary
bronchi
2.0—3.3 microns
Secondary
bronchi
I
1.0 — 2.0 microns
0.3 — 1.0 microns
0.1 — 0.3 microns
terminal
bronchi
Alveoli
Alveoli
/V f>
k.u
. \^
Figure 4-1. Respiratory collection of particulates.
Usually
natural particles
Generally
man-made
particles
4/t
I/*
Particle diameter
Figure 4-2. Bimodal distribution of particles in the atmosphere.
4-2
-------
CHARACTERISTICS OF PARTICLES AND PARTICLE DISPERSOIDS
Pirtete Diameter, microns {fit
liny)
00001 0001 001 01 1
Equivalent
SIZM
Electromagnetic
Wave*
Technical
Definitions
Common Atmospheric
Dispersoids
Typical Particles
and
Gas Dispenoidt
Method! for
Particle Size
Analysis
Types of
Gas Cleaning
Equipment
Terminal
Gravitational
Settling'
for spheres,
L sp. gr. 2.0 J
Particle Diffusion
Coefficient*
cm '/sec
DlSperSOKrS
Soil
1 aim
In Water
at
25' C
In Air
at 25'C
latm
In Water
at
25'C
1
Liquid
3 K
ingstrOm Units, A
Wterberg tx International Std Classification S
adopted by Internal Soc Soil Sci Since 1934
Oi COi
C.H.
Gas
CO H,0 HCt C*HIO
'Molecular diameters calculated
from viscosity data at O'C
Reynolds Number
Settling Velocity,
cm/sec
Reynolds Number
Settling Velocity,
cm/sec
h«
** X
0 1,C
ystem
(1mm) (1cm)
10 100 1.000 10,000
5,000 1,2
00 10,000 2,500
Theoretical Mesh
(Used very mfrequen
|
Visible
et M* — »*«j-Near infrared *
Solar Radiation
+ Rosin Smoke >
* Tobaccc
*—— Carbon Bla
-—Zinc Oxide F
' Silica* '
Aitken
Nuclei
w-SeaS
Combustion
Nuclei
uses H
— Ultram crosoc
urgical Dusts and
•-Ammonium Chlond
Ch H- C°n
,0 «,*tx<\'
625 nnn
W "^J"
1 1 n i
•« Far In
Du
« 66 36 20
Tyler Sen
100 48 28
u 60 40 20
US Sere
100 50 1 30
, ill
Irared •
st -™
Spray
— »^« — Fine Sand— *^*- Coarse Sa
and Fog — *+-*Mist«HOrzzte** —
k — Fertilizer, Ground L mestone— >
Dust »
Fume-***- — Cement Dust *
Su If uric !
Concentrator M st
ck Sulfunc Mist ' ' '
•*• ••-- Paint Pigments H
jme — H h*- Insecticide Dusts -*
H—J-— Ground Talc
-— -Spray Dred Milk »
* Alkali Fume H
~H h* Milled
-• — Flotation Ore
Plant
Spores'1
-• — Pollens »
Flour -I
i Beach Sand
H
j M
alt Nuclei-H I-.— Nebulizer Drops-*i i*— t Hydrau ic Nozzle
ILung Damaging. Pneumatic ,
'"" Dust ' \ * Nozze Drops
Red Blood Cell Diame er (Adults) 7 5/i iO 3/J
K j Bacteria H-Human Ha rn
Electroformed
ope1 — — H* — i -Microscc
r* — j- Cen nfuge— \ H*
*ay Diffraction ^
Sieves
— M
— Elutnation
Turbid metryit-— - — j- -- H
-i-
Scanners
-Nuclei Counter-j »** Electrical Conductivity *•
Ultrasonics _|_ ^L
(very limited ind
h
r
L0~" 10" "3 1
235
10T, 5lcr'!5
i i i 1 1
10"6 1 10"6
, i,.,,i-.-i .__ii44J
>ii ,11
stnal application) ' '
^
— __ C|0
-Liquid Scubbers
h Collectors— —
H
V*- Common Air Filters — *•
Thermal Precipitation ^,
(used only
or sampling)
Electrical Precipit
0' 1D 10~93 10 S3 10'^ 10"^
10-'
, : r I
Iff" 10 "10^°
10"" 10"'
III II
II II 1 1 1 1 I
235 2 1 S
10"' 10"° 10" 3
10"" 10 b
10-'
J ' - ' i1 Jl Jl
— Mechanical Se
10"^ 10"* 10 J
io^f?io-;M
10 " lO'l 10"'
\0'\ 10") 10°
2] b 1 ?
10 33 10 j 10 j
10' 10"' 1C"' 10-
, . . 1 1 , , , I III
10-'
1 1 1 1 1
i « 3 2 6
-1 — , 1 . — 1 M i 1
H
10 6 3
>enMesh 1
14 8 4 X
12 6 3
*nMesh j
••Microwaves (Rad
nd -»•*•« Grave
_ i
Drops — f
"Furnishes average
diameter but no s
distribution
"'"'"Size distribution rr
obtained by speci
cal bration
c._
Machine Too s (M
.a* ing
sarators — w
10 s 102
1 1 1
io°3 io1 3 10
2 3 ' ». 3
| II 1
, 3 , l°~"
ii i h , i
'£ V
T
•f I-
V
lar, etc)
particle
ze
ay be
1
rs, etc)-
j ,
10' ,
1
6789
, 10""
1
412
S1 ' .'
SoTlnclSSv1 00001 0001 001 01 1 10 100 1,000 10,000
values given for air llm)*l (1mm) llcm)
but not included lor wit.r Particle Diameter, microns (p.) POEKKD !< c E UPPIC
RiprlntKl Iron Slinlord HeMarch Institute Journal. Third OuiiW. 1961 Single copus 8-1/2 by 11 inches tree or £10 per Hundred 20 by 26 inch wall chart SIO each Both charts available from Oepl 300
SRI Intemationd 333 Ravenswood Ave. • Menlo Park, California 94025 • (415) 326-6200
Printed in USA M214 • 2M • 361 -7806 i,
Figures 4-3. Characteristics of particles and particle dispersoids.
4-3
------- An error occurred while trying to OCR this image.
-------
Analysis of Inertial Collection Mechanisms
A complete analysis of the inertial collection of particles in aerosols includes an
analysis of the forces acting on a particle in a rapidly moving aerosol stream. The
collection itself should also be examined. Collection is the phenomenon involved
when a particle strikes and becomes attached to an obstacle in the aerosol
trajectory.
The net force on a particle in a moving aerosol stream is the result ol several
forces acting on the particle. The net force can be expressed in the equation:
(u — v) —• 7T du
(Eq. 4-1) - V- --- ' mg + F. = m —
z dl
Where: ~u — particle velocity
~v = velocity of air stream
z = "mobility"
in = particle mass
g~= gravitational acceleration
F, = electrical force
The net force (F= ma) on the particle is represented by m(du7dt).
The term - (u-~v) . ••>,,•, ^v.ion for the fluid resistance of the air to the par-
ticle flow in which z, the mobility, is a function of the particle diameter and the
physical properties of the air. The term for the fluid resistance is valid under the
assumption that viscous forces predominate in the particle flow through the air
stream, that is, the particle is equivalent to a sphere of small diameter and the par
tide velocity is small. The term mg represents the gravitational force^on the par-
ticle. The electrical forces acting on the particle are represented by F..
In the inertial collection process, the particle is removed from the air stream as a
result of a sudden change in the fluid resistance force. While the particle is
traveling in the air stream prior to passing through the jet, the particle velocity, IT
is nearly the same as the fluid velocity ~v. As the air stream approaches an obstacle,
the air velocity iTundergoes a sudden change as the fluid spreads and flows past the
obstacle. The particle follows the trajectory described in the equation. At this point
the gravitational and electrical forces on the particle are negligible in comparison
with the momentum of the particle. Whether the particle will reach the surface of
the collector depends upon the proximity of the starting position to the axis of flow
through the center of the collector, the nature of the streamlines along which the
particle passes, the configuration of the flow system, and the size and shape of (he
collector (17).
Types of Inertial Sampling Devices
The inertial collection process is subdivided into two main types, impaction and
impingement. The distinction is made by the manner in which the sample material
is retained in the sampling device.
4-5
-------
Impaction Devices
Impaction devices collect and retain particles from an aerosol stream on a surface.
The collecting surface is removed from the instrument and the sample analysis is,
in many cases, performed directly on the collecting surface. Particle adhesion is
caused primarily by electrostatic attraction and by molecular surface phenomena
called Van Der Waals forces (1). Some loss of large panicles occurs with high
aerosol velocities. It is believed that in the case of small particles, (several
micrometers or less), nearly all of those striking the colic-cling surface are retained
on the surface. The collection surface in many impacticm devices is coated with a
thin film of oil or light grease to aid in particle retention. In some devices, reten-
tion is aided by passing the incoming particles through a /.one of moisture
saturated air; moist particles adhere more readily to a collection surface. Coating
of the plates and water saturation of the particles affect the calibration of an
impactor and must be accounted for if the impactor is 10 be used for panicle M/e
distributions.
Single Stage Impactor
A single stage impaction sampler contains one jet, or orifice, and one collection
surface. The earlier, simpler devices were this type, and many of these collected a
.ample by the operation of a simple attached pump. An example of a single stage
impactor is the Staplex Annular Impactor shown in Figure t-5. It was designed
specifically for selectively sampling dense plutonmm particles in aerosols generated
by various operations in the production of fissionable material.
Figure 4-5. Annular impactor (courtesy of the Staplex Co.).
4-6
-------
Cascade Impactor
A widely used impaction device is the "cascade" impactor shown in Figure 4-6.
This device consists of several impaction stages arranged in series. Each successive
stage contains a smaller jet placed closer to the collection surface than in the
previous stage. This arrangement causes an increase in the aerosol velocity and a
greater deflection angle at each stage through the sampler. The result is a higher
collection efficiency for particles of decreasing size at each stage through the
impactor.
Air path
Jetl
7 mm
Jet 4
0.6 mm
3
Inches
Figure 4-6. Cascade impactor schematic diagram.
4-7
------- An error occurred while trying to OCR this image.
------- An error occurred while trying to OCR this image.
-------
Figure 4-9. Andersen sampler.
(Courtesy Andersen Samplers and Consulting Service)
4-10
------- An error occurred while trying to OCR this image.
-------
Vacuum
pump
5-stage
Andersen
Sampler
Removable
rain
shelter
Air inlet
Spring clamps
4-inch backup
filter holder
Orifice
Magnehelic
pressure gage
mounted on shelter
Vented
motor shelter
Figure 4-11. Diagram of modified Andersen impactor sampler and shelter.
Figure 4-12 shows the collection efficiency ot the modified Andersen samplei lor
each stage as well as the overlapping of particle si/e collection that occurs with ail
irnpacior collectors. The dotted lines represent an extrapolation of the curves that
were not obtained experimentally (7).
100
0.1 0.2 0.3 0.4 0.5 0.8 1.0 1.5 2,0 3.0
Particle size
Figure 4-12. Collection efficiency of modified Andersen Sampler (3 ftVmin).
4-12
-------
High Volume Andersen Cascade Impactor
A further modification of the Andersen Sampler has led 10 the High Volume
(hi-vol) Andersen Sampler, which can be used at flow rates as high as 0.57 in*,'min
(20 ft* min). This sampler consists of five stages with the typical Andersen per-
forated discs. Following the discs are gaskets and collection surfaces (Figure \ 13).
Impactor jet
Aerosol deposit Impaction surface
Back-up
To hi-vol motor hi-vol filter Screen support
Hi-vol
Figure 4-13a. Cross-sectional view of hi-vol Andersen impactor.
^^^M^*^^
Sneed
ball handle
Sm^ Flat washer
n-w4— h^ llisr 1
\-~* — Collec
i — i — rt~~p^ — Disc 'i
i i^l K Disc 5
washer
(5 places)
tion surface
1
Pressure tap
Interface plate
,»••••«
<
,•• »»^
f
•
•
•
•
•
•
V
*•*••••*!
• •*.••.(
••'•••ed
• •.••••<
•>.:•—
(symmetrical •• ••}
about center) ••
Gasket
(symmetrical
about center)
Figure 4-13b. Expanded view of hi-vol Andersen impactor.
4-13
-------
"lie collection surfaces, usually a glass fiber filter or aluminum foil (2), are sup-
ported by the plates themselves, i.e., the collection surface for plate 1 is supported
by plate 2 and so on. To allow unrestricted aerosol flow, the collection surface has
holes that line up with the orifices of the plate supporting it. The five stages are
held together by a central bolt and are aligned by four evenly spaced rods. The
impactor is mounted on a hi-vol sampler. An 8" X 10" filter mounted on the hi-vol
is the backup filter for the impactor. Figure 4-14 shows the collection efficiency of
the hi vol Andersen impactor operated at 20 fWmin.
Theoretical efficiency
for Anderson hi-vol
sampler
(20 cfm)
Particle size (microns)
Figure 4-14. Collection efficiency of hi-vol Andersen impactor (20 ftVmin).
Multiple-Slit High Volume Cascade Impactor
Another ot the size fractionating paniculate samplers utilizes a typical high-volume
motor blower unit with an adapter comprised of four stages with successively
smaller slit openings. Behind each slit is a collection plate tor retaining particles.
The fifth stage of this impaction device uses a typical hi vol filtci for collection of
small particulates.
4-14
-------
The air is drawn through the slits and over the collet -lion-surfaces. The slils
become increasingly thinner and the velocity increases ai each slage so particles ol
decreasing size are impacted on each successive surface. The collection plates are
5.1 cm (2 in.) wide, 12.7 cm (5 in.) long, and made ol aluminum sheets 2.5 mm
(1 10 in.) thick. The actual collection surface is made of glass fiber or other
materials typically used in hi-vol sampling lo permit the same type of ( liemical
analyses to be done.
The distance from the slit opening to the collet lion surface becomes smaller with
each stage to increase the collection efficiency (set Figuie I 15).
Jet to plate
spacer
Tie bolt
Pressure
tap
Filter
Impaction
plate
Rubber
spacer
Figure 4-15. Multi-slit high-volume cascade impactor.
One advantage of ihis impactor unit is lhai a will li> O\ shows the percent penetra-
tion, which is related to the collection efficiency, ol dilleient sized particles and the
impactor stage. The greater the percent penetration, the lower the collection
efficiency.
4-15
------- An error occurred while trying to OCR this image.
-------
Top section view
Belt drive
Motor
Side section view
Figure 4-17. Lundgren type inertial collector.
Impingement Devices
Impingement deuces differ from impaciois in thai ilu- jei and .sinking surface ate
immersed in a collecting fluid such as water. The panicles that are removed from
the aeio.sol stream are wetted by and retained in the fluid. Most impinge-t.s in use
aie variations of the instrument developed by Greenbiug a.id Smith (1). The
Greenburg Smith type of impingei is constructed as a glass cylinder wiih a small
concentric glass tube insert The jet and impingemen, plate snuciute. also made u
glass, arc' attached to one end of the tube1, which is immersed in i he collecting
tluid. These impingers collect a sample1 ai flow uiu-s ot about D.I to 1 cubic teci
per minute depending on the- mode of operation.
Impingers are most commonly used in colleciing a. mosphei ic samples lot dusts,
misls, and tumes in the1 e\aiuation of Occupational health ha/.aids. In addition 10
the collection of soluble- gases and pariiculaies. tlie (.ircnbuig Smith impmgei is
an efficient collector of insoluble* particles greaiei ban '\vu mici ome-ters in
diameier (3).
Centrifugal Separation Devices
Centrifugal separation is a variation of the menial collection piocess in which pai
ticlc-s are removed from an air stream by the' cenisifugal force1 created by moving
an aerosol rapidly thiough a circular path.
There are several types of sampling devices employing the principle of
centrifugal separation.
4-17
-------
Cyclone Samplers
Samplers called cyclones, or miniature cyclones, are small versions ot the large
cyclones used in air cleaning applications. The cyclone shown in Figure 4-18 con-
tains no moving parts and is designed so that air drawn through the cyclone will
move in either a circular or a helical path of decreasing radius through the device
in order to increase the collection efficiency of small particles. The gas stream
surges through the cyclone and the particles are separated at the inside surface ot
the cyclone wall by the centrifugal force created. In most cyclone samplers the par-
ticles stick to the walls or drop into a collection space be-low the cyclone chamber.
Outlet
Inlet
Inlet
Cap for
sample removal
Figure 4-18. Cyclone sampler.
Most available cyclones are not efficient collectors of particles with diameters less
than 2 or 3 micrometers, depending on the particle density. However, small
cyclones can be designed to collect particles below 1 micrometer in diameter. The
collection efficiency of a cyclone sampler is strongly dependent upon flow rate.
Spiral samplers are similar to the cyclone in having no moving parts. The air
follows a spiral path and the particles are deposited on the outside wall of the
spiral on a removable collection surface.
4-18
-------
Air Centrifuges
In an air centrifuge device for aerosol sample collection, the air passes between two
concentric cones. The inner cone is rotated by a motor and draws air in at the
narrow upper end of the cones and exhausts the air at the large end of the cones
(see Figure 4-19). The sample is collected on the inside surface of the outer cone,
which may be lined with a removable collection surface. The inner rotating cone,
in one model, is channeled in the form of a spiral and the outer cone spins with
the rotor cone (1); laminar flow is maintained in the channels in this manner. The
particles are deposited on the wall of the outer cone in the stationary boundary
layer where they are protected from shear (16). The speed of rotation of the rotor
cone can be varied, and it is possible to collect particles with effective diameters as
small as 0.03 micrometers (1).
Inlet
Sample
collection cone
Collection
surface
Figure 4-19. Air centrifuge.
Air centrifuges differentiate between particles according to size. The larger
particles have a shorter length of deposit than do the smaller ones. The length of
deposit refers to the maximum distance along the sample deposit on the collection
surface which includes particles of a given size. All particles of the same size are
not removed at one point. Each particle size exhibits a maximum penetration
distance into the collection path.
Thus, some small particles are deposited with the large particles so that analysis
by subtraction is necessary when studying size distribution (16).
4-19
-------
Virtual Impactors—Dichotomous Sampler
In the virtual impaction method, instead of having the larger particles impacted
onto a solid collection surface, they are impacted into a slowly pumped void and
collected on a filter later down stream. Thus, particles larger than the 2.5 micron
cut-point impact through a non-existent, or "virtual" surface. As pictured in Figure
4-20, the smaller ("fine") particles follow the streamlines of the faster ilow (Ff);
whereas, the larger ("coarse") particles are impacted into the slower flow (Fc).
A small fraction of the fine particulates are impacted and collected with the coarse
particles. Corrections for this can be easily made for both the fine and coarse par
ticulates with the following equations (21):
(Eq. 4-2)
(Eq. 4-3)
Where-
fm,f.
r/r
c, =
t(f,+f.
d and c, = the atmospheric concentrations of the fine and coarse
particle fractions, respectively in ng/m*
nil and m, = the masses collected on the fine and coarse particle
filters, respectively in fig
f, and f, = the flow rates through the fine and coarse particle
filters, respectively in m'^/min
t-the sampling time m minutes (normally 120 mm.)
Fine flow
(ff)
Total flow
(J -I
Fine flow
>:-^:!f*r^-:'^"--:••• pm F'netl
"-. ^4^:c-..-,-.-^.::.- i» oy
•:'• • -^ • "1'>>j«
-------
The most recent use for virtual impaction has been in EPA's efforts to develop a
paniculate sampler that will separate and collect paniculate matter above and
below a 3.5 /tin aerodynamic diameter. These efforts have led to the development
of the dichotomous sampler. The dichotomous sampler obtains its name from us
funcdon of sampling fine (<3.5 jim aerodynamic diameter) and coarse (>3.5 jtm
aerodynamic diameter) on separate filters (see Figure 4-21). The sample inlet
design shown in Figure 4-22 rejects particles with an aerodynamic diameter greater
than 20 /*m. This inlet design has been shown to be insensitive to wind direction.
Virtual impactors have several advantages over conventional impactors (20):
• Problems associated with conventional impactors, panicle bounce and
reentrainment, non-uniform deposition, cumbersome sample handling, are
essentially eliminated.
• Uniform deposition onto a filter medium is ideally suited for photo-excited
X-ray fluorescence (a non-destructive chemical analysis technique) and beta
gage mass measurement.
• No grease or oil is needed to improve collection etliciency.
• Choice ol collection media (i.e., filter material) can be made to eliminate
interferences.
From aerosol inlet
Fine particles
less than 3.5 microns
Coarse particles
greater than 3.5 microns
Filter
cassette
Fine
particle
filter
Inlet tube
Virtual
impactor
nozzle
• Virtual
impactor
receiver tube
Filter
cassette
Coarse
particle
filter
37 mm dia.
Filter
holder
0.9 CMH
0.1 CMH
Figure 4-21. Diagram of a dichotomous sampler.
4-21
-------
Measurements have shown that particle loss is essentially eliminated in the 1 to
2.5 and 4.5 to 20 /xm ranges. The loss curve peaks at 5 percent for particles in the
vicinity of 3.5 /an.
Figures 4-23 and 4-24 picture a manual dichotomous sampler that is presently
being marketed. An automatic version has also been developed that changes the
fine and coarse filters at pre-set time intervals.
Flow
Shield
Bug screen
16 x. 16 mesh
Flow to dichotomous sampler
Figure 4-22. Aerosol inlet for dichotomous sampler.
The filter material used in both the coarse and tine sections is high porosity
TeflonR) with a 1 /«n pore size. This filter material was chosen because of its
several advantages over other types (22):
• Collection efficiency for particles above 0.01 /un greater than 99 percent;
• Extremely stable mass for high gravimetric accuracy;
• Negligible tendency to absorb or react with gases (therefore, low artifact
formation of nitrate and sulfate);
• Minimal impurities to interfere with analyses for chemical and elemental
species;
• Low mass per unit area (desirable for gravimetric, X-ray fluorescence, and
iS-gage measurements).
These Teflon"1 membrane filters will not support themselves; therefore, they are
bonded to a polyethylene net or a thin annular polyester ring. The polyester ring
seems to be the best choice. Investigations are being made to develop a Teflon"
filter with a 2 to 5 /mi pore size with high collection efficiency. These new filters
would allow longer sampling periods and greater loading without clogging.
4-22
-------
Filter holders
Sampling head
Sample inlet
Virtual impactor
Flow control unit
Connections
for flow control units
Figure 4-23. Dichotomous sampler.
Bottom plate of
sampling head
Initial jet
High flow
fine particles
Fine particle
filter holder
Low flow
coarse particles
Coarse particle
filter holder
Figure 4-24. Expanded view of dichotomous sampler.
4-23
-------
A constant flow rate must be maintained through the dichotomous sampler if it
is to be used for quantitative procedures. The use of a differential pressure
regulator in the exhaust seems to be the most rugged and promising technique to
control flow. This method of flow control is used in at least one commercially
available instrument (see Figure 4-25).
Icmh
^-^JU-
0.9^1JlJ^
Flow
riu*v
_? controller
( *) UUffllB p
T
1
cmh Fine Inkt
particle fiher
filter „
r4
Coarse VAC C'*)
particle
filter
n m r"
T
-,
Jj
O '
L
§1
>•
J(- h
16.7
p ~ (pm
(1 cmh)
m^l
^Filter/ SjT^"1 Filter/,
sursre Thomas surge \_
tan! CA 727 tank A
_ Flow
vacuum pump
selector
| 1
— (VV>J Flow
PV-X
i 171 meter
r low uitiA.*
selector
J valve
P
set
port
valve
1.67 tpm
0.]
cmh
nir-1_r
Inlet
filter
(0.1 cmh)
Flow meter
Figure 4-25. Flow schematic of control module.
Respirable paniculate sampling is becoming a major objective of atmospheric
monitoring. The dichotomous sampler may possibly replace high volume sampling
because of its ability to sample for fine, coarse, and total particulates, without
many of the problems associated with hi-vols.
Collection Efficiency
Though the principle of collection is the same for all types of inertial sampling
devices, the parameters are somewhat different for the basic single stage or cascade
impactor than for the air centrifuge or cyclone. The discussion in this section is
confined to the jet type of impactor, and this discussion may not be applicable
without modification to other types. In this section discussions of particle size refer
to the equivalent aerodynamic particle size or diameter (Stokes' diameter) unless
otherwise stated. The equivalent diameter is the diameter that a unit density par-
ticle of spherical shape would have if it behaved the same as the particle being
studied (9).
4-24
-------
Definition of Collection Efficiency
Particle collection efficiency for inertial sampling devices has been defined in
several ways by various people. The collection efficiency of an impaction device is a
function of several parameters, making the theoretical and empirical expressions
that have been derived somewhat difficult to use for specific applications. Ranz
and Wong (17) showed that the efficiency of impaction was a function of an
inertial parameter /, defined as follows:
• 2
;>
(Eq. 4-4)
Where: v,, = initial velocity of aerosol through jet
Dr = effective particle diameter
d.. = width of jet
/j. = air viscosity
Q,, = particle density
and C was defined as, "an empirical correction factor for the resistance fluids
opposed to the movement of small particles, and is dimensionless. For air at not
ma I room temperature and pressures
D,,
"The physical significance of / may be taken as the ratio of the stopping
distance, in this case the ratio of the distance of particle penetration into still air
when given an initial velocity, V,,, to the diameter or width of the aerosol jet" (1).
The collection efficiency of an impaction device can be defined as the ratio of
the number of particles striking an obstacle to the number that would strike if the
stream lines were not deflected (3).
For the purposes of this discussion, the collection efficiency will simply be
defined as the fraction of the particles in an incident aerosol stream that are
retained on the collection surface of the sampling device. The efficiency of impac-
tion can be plotted as a function of /, the inertial impaction parameter, to give an
efficiency curve for an impaction device. Studies on the calibration of cascade
impactors are available (23).
Impactor Performance Characteristics
Instead of plotting efficiency of impaction against /, u can be plotted against par-
ticle size for a given device, since / is a function of particle size and several other
parameters. The efficiency of impaction when plotted against particle size follows a
sigmoid (or S-shaped) curve such that there is a minimum size below which no par
tides are collected and a maximum size above which all particles are collected. For
a well designed impactor, the range between these two sizes is sufficiently narrow
that a functional size separation is made (16). The most effective way to describe
the ability of an impactor stage to separate particle sizes is to show the efficiency of
collection as a function of size, as shown in Figure 4-26 (16).
4-25
-------
u
1.0
0.8
0.6
' 0.4
0.2
2 3
Size in microns
Figure 4-26. Efficiency of collection a^ ., tu. itiua ol size.
Where the efficiency function crosses the 0.5 collection efficiency point is the 1),,,
01 50% cutoff size. ' Larger particles are collected with more efficiency and
smaller particles penetrate with more efficiency.
The inverse of the collection curve is the penetration curve. 1'he size distribution
of paniculate on any stage is determined by the pem-muion curve ot the stage
ahead and its own collection curve. To get the actual si/e distribution on the stage.
the efficiency and penetration values are multiplied into the si/.e distribution curve
of the incoming aerosol. The reverse operation, when the size distributions on the
stage are known, gives the distribution of the incoming aerosol (16). The 50%
cutoff value does not depend upon the size distribution of the aerosol being
supplied to the impactor as would be true of the MMD (mass median diameter).
The mass median diameter is that diameter above which and below which exist
50% of the collected paniculate mass. It is truly valid as a characteristic of a
sampler only when the aerosol sampled by the sampler has the same size distribu
lion as the aerosol used tor calibration. Since this restiiction is not true of the
efficiency curve, this curve is now considered by many to be the best description of
sampler performance (16).
4-26
-------
Properties of Aerosols Affecting Inertial Collection Efficiency
For a given collection device several properties of the aerosol affect the efficiency of
collection.
Particle Size
For any device there is a minimum particle size below which no particles are
collected and a maximum particle size above which all particles are collected or
removed from the sampling stream. Above 50 micrometers some particle loss occurs
by impaction on the walls of the sampling device. Panicles and droplets greater
than about 200 micrometers seldom are found on the sample collection area but
are shattered or lost on the walls. Particles below about 0.5 micrometers in
diameter are difficult to collect with impaction devices because their momentum is
not appreciably different from that of the air, and thev aie not readily deflected.
Most impacting devices have a high collection efficiency for particles with
diameters greater than 0.5 micrometers. Collection efficiencies of nearly 100 pet
cent aie claimed for several impaction devices foi particles in the si/e range from
0.5 to 10 micrometers (3).
Particle Density
Impaction devices are efficient collectors of high density panicles: they have been
used to selectively sample work atmospheres for plutomum dusts. Sub micrometei
particles of high density can be efficiently collected, since they are equivalent
unit density particles of much greater size.
to
Aerosol Fluid Properties
The physical properties of the gas in which the paniculate1 is dispersed will affect
the collection efficiency of an inertial collector. The most important factor in this
category influencing inertial collection efficiency is mveisely pioponional to the
viscosity, so that a factor affecting the viscosity, such as temperature, will in turn
affect the collection efficiency.
Properties of the Collecting Device Influencing Collection Efficiency
Impactors and impingers are designed for sampling a \anety of aerosols. The
efficiency of collection for a certain aerosol can be opinm/ed by design
considerations.
4-27
------- An error occurred while trying to OCR this image.
-------
Some Applications of Inertial Sampling Devices
Gross Sampling
Under certain conditions an inertial sampling device may be the most appropriate
to use in collecting a gross sample of an aerosol. Conditions such as high
temperature or humidity may eliminate the use of other types of sampling devices.
Inertial devices are efficient samplers of hot, moist aerosols and droplets that are
likely to be encountered in stack sampling, as the moisture aids in the sample
retention on the collection surface.
Size Distribution Studies
In many cases where impaction is used for aerosol sampling, the particle size
distribution is of interest. This can be studied by microscope sizing and counting.
Autoradiographic studies have been performed in conjunction with microscopic
analysis of radioactive aerosols. When microscopic analysis is used, it is desirable to
collect the sample directly on the viewing surface. The loss of particles or alteration
of the sample in transferring from collecting surface to viewing surface is thus
avoided. Many impaction devices use microscope slides as collection surfaces.
Theoretically, particles with diameters less than about 0.25 micrometers cannot
be resolved using light microscopy. Electron microscopy can be used for size studies
of particles less than 0.25 micrometers in diameter. An impactor sample for elec-
tron microscope sizing is collected directly on a fine mesh, carbon-covered, copper
viewing grid (14).
The classification produced by impactors has certain advantages for microscopic
determinations of size distributions when there is a large variation in particle size.
Furthermore, this classification provides a technique for obtaining mass distribution
curves. The impactor must be calibrated to determine the smallest particles col-
lected at each stage or, preferably, the size of particles which are sufficiently small
that they are collected with only 50% efficiency at each stage. The dust whose par-
ticle size distribution is to be determined is then collected with the impactor. The
weight of material collected at each stage is determined and the results are plotted
as cumulative weight distribution curves. The particle diameter corresponding to
the 50% fraction is called the Mass Median Diameter (MMD)
Respirable Dust Sampling
To study the hazard to humans from inhaled dust particles, it is desirable to know
which portion of an aerosol actually penetrates and is retained in the lower portion
of the lung. To relate airborne concentrations to the amount actually deposited in
the lower portions of the lung, it is necessary to sample by a technique that collects
only the segment of an aerosol that is deposited in the lower portion of the
respiratory system. "Respirable dust" has been defined as that portion of the
inhaled dust deposited in the non-ciliated portions of the lung (3). A recent discus-
sion of "respirable" dust sampling has been published by Lippmann (24). Samplers
4-29
-------
have been designed that closely follow the deposition characteristics of the human
lung. Figure 4-27 lists the respirable percentage of several particle sizes obtained
from experiments performed on humans (7).
Size in microns
particles of unit density
>10
5
3.5
2.5
< 2
Percent
respirable
0
25
50
75
100
Figure 4-27. Respirable percentages of different particle sizes.
A respirable dust sampler must have a collection efficiency equal to the percen-
tage respirable for the various particle sizes. Some respirable dust samplers utilize
inertial principles in their design. A typical sampler is constructed with two stages:
a first stage consisting of an inertial device (cyclone collector) that simulates the
upper respiratory tract in the removal of large particles; and a second stage con-
sisting of a membrane filter that collects the smaller particles that pass through the
first stage (see Figure 4-28). The first stage must have a collection efficiency curve
for particle size that allows the respirable fraction, or a close representation
thereof, of the various particle sizes to pass through. When the sampler is operated
at the proper How rate, the portion collected on the membrane filter should repre-
sent the dust deposited in the lower respiratory tract. High volume cascade impac-
tors and the newly developed dichotomous sampler are examples of samplers used
in monitoring networks for respirable dust sampling.
Figure 4-28. Respirable dust sampler (courtesy of Union Industrial Equipment Corp.)
4-30
-------
Limitations and Sources of Error
in Inertial Collection (Disadvantages)
Inherent Errors in Sample Collection
In many cases it is impossible to characterize a system without altering the system
by the act of measurement itself. In other words, do the particles collected in an
impactor sample fairly represent the dust as it is suspended in the air? There are
-veral inherent sources of error in the impaction process.
Particle Shattering
Large particles (greater than 200 micrometers) and agglomerates are readily
shattered upon impaction, and, at the high velocities attained in some impaction
devices, particles with diameters as small as two or three micrometers can be
shattered. In studies where the number of particles per unit volume of air is of
interest, shattering of particles upon collection results in erroneously high results.
In size distribution studies, there will appear to be fewer large particles and more
small particles than actually exist in the aerosol. Figure 4-29 is a picture taken
from an electron microscope of a shattered fly ash particle.
•. • «• •/.* • •* ^y, *\ •*.*"•* *c~
•
Figure 4-29. Particle shattering.
4-31
-------
Particle Bounce, Re-entrainment of Particles, and Wall Loss
At high impaction velocities a small fraction of the particles collected may be
re-entrained in the air stream. This occurs most often with fragments of large
particles that have shattered upon striking the collection surface. Some of the
pieces of the shattered particles may be iost from the sample by impacting on tin-
walls of the instrument. A few of the large particles may impact directly on the
walls of the instrument.
If the instrument is not thoroughly cleaned prior to use, particles that have
impacted on the walls in previous sampling may become re entrained in the aerosol
stream within the instrument and be collected on the collection surface.
Particulates can also become re-entrained because they bounce when they strike
the collection surface. Dzubay, et. al. experimentally showed that the ". . .nature
of the impaction surface can have a severe distorting effect on the size distribution
which one measures with a cascade impactor" (25).
Limited Sample Quantity
The small quantity of sample collected also restricts the choice of analytical
methods to those with high sensitivity. Care must be laken to preserve all sample
material intact, since with only a few micrograms of sample, the loss of any par-
ticulate becomes significant.
Some types of inertial samplers, such as cyclones, are not so limited in this
respect. Some of these can collect gram quantities of particulai.es.
Nonrepresentative Sample Collection
If jet impactors are used in sampling high concentrations of particles or mists, the
time of sample collection must be short in some applications. If too much material
is collected, the sample will be useless for size and mass distribution.
Sample Loss in Collection
If too much particulatc material collects on the sample collection surface, subse-
quent particles that impact may be lost by re-entrainment when they strike par-
ticles already collected instead of the collection surface.
A phenomenon called "ghost depositing" can occur when particles bounce off the
collection area and are redeposited by eddy currents a few millimeters on either
side of the sample.
When aerosols containing mists or droplets are being sampled, care must be
taken to avoid collection of too much material. If too many droplets are collected,
they will merge on the collection surface. The individual drops will be lost for size
analysis and some material may be lost by run-off.
4-32
-------
Poor Particle Resolution for Size Analysis
Particles that collect close to and on top of each other will introduce error in con-
centration and size studies through the inability to distinguish between individual
particles and clumps of particles when examined optically. However, if a represen-
tative portion of the collected material is properly remounted in a preparation,
these problems can be minimized.
Errors Associated with the Calibration of Collection Devices
When impactors are used to collect aerosol samples for particle size or mass
distribution studies, the instrument must be calibrated. The particle collection
efficiency as a function of the desired parameter, (e.g., size or mass) must be deter-
mined. One method of calibration is to empirically determine the collection
efficiency in the laboratory. In this procedure a known quantity of a monodisperse
aerosol (a gas stream with only 1 size particle suspended in it) is passed through the
impactor. The amount of material collected on each stage is weighed. By varying
the size of the monodisperse aerosol the cutoff points for each stage can be deter
mined. Detailed studies of theoretical collection efficiencies for several types of
impaction devices have been performed (3). The efficiency curves that have been
derived for several types of impactors show varying agreement with experimental
curves. These curves are usually given in terms of an inertial collection parameter
or equivalent particle size such as the one in Figure 4-10. If a particular curve is to
be used in a sampling application, the user should be satisfied that the curve is
valid for that particular application. Most commercially available inertial samplers
are provided with a curve of collection efficiency. In most cases the methods used
to obtain these curves will not be known to the user. Care should be exercised in
the application without specific knowledge of the manner in which the curves were
obtained (23).
Errors in Sample Analysis
Investigations of radioactive aerosols, and aerosols in general, in which the sample
is collected with an inertial collection device may utilize any of several kinds of
sample analysis. The analyses that are performed on samples collected by impactors
may also be used in conjunction with other sampling methods and are discussed
elsewhere. Mention of the error associated with the analytical procedure to be used
is made because it should be considered in the overall error assessment of the
sampling procedure. The information desired in an investigation may require the
use of an inertial collection device, which in turn may limit the choice of analytical
procedure. The reverse could also occur, i.e., the information desired such as par-
ticle size distribution, may dictate the use of an impaction device to collect the
sample. One should be aware of the limitations and error associated with each
analytical procedure that is considered.
4-33
-------
Summary
Inertial sampling devices have found numerous applications in the study and
monitoring of atmospheric aerosols. Though not used as frequently as some other
types of sampling devices, inertial samplers are superior for certain applications.
The investigator can choose from a number of available devices with different per
formance characteristics. A wide range in cost and degree of sophistication are
exhibited by these devices. An understanding of the principles involved as well as
the limitations and sources of error is a prerequisite for the achievement of valid
results in any investigation using inertial samplers.
References
1. Balzer, J. L. Inertial Collectors. Air Sampling Instruments for Evaluation of
Atmospheric Contaminants. American Conference of Governmental Industrial
Hygienists, 1972.
2. Burton, R. M.; Howard,]. N.; Penley, R. L.; Ramsay, P. A.; and
Clark, T. A. Field Evaluation of the High-Volume Particle Fractionating
Cascade Impactor—A Technique for Respirable Sampling. Paper #72-30,
presented at 65th Annual Meeting of the Air Pollution Control Association,
June 18-22, 1972.
3. Green, H. L. and Lane, W. R. Paniculate Clouds: Dusts, Smokes, and Mists.
D. Van Nostrand Inc., 1957.
4. Gussman, R. A.; Sacco, A. M. and Ladd, R. E. Design and Calibration of a
High Volume Cascade Impactor. Paper #72-27, presented at the 65th Annual
Meeting of the Air Pollution Control Association, June 18-22, 1972.
5. Hertz, M. B. Size Segregating Mass Sampling Comparisons of Single, Two-
Stage, and Multistage Sampling. Presented at the 13th Conference on Methods
in Air Pollution and Industrial Hygiene Studies, October 30-31, 1972.
6. Hidy, G. M. et al. Summary of the California Aerosol Characterization Experi-
ment. Paper #74-119, presented at the 67th Annual Meeting of the Air Pollu-
tion Control Association, June 9-13, 1974.
7. Hu, J. N. An Improved Impactor for Aerosol Studies —Modified Andersen
Sampler. Environ. Sci. Tech. 5: 251-253, March 1971.
8. Lee, R. E., and Goranson, S. National Air Surveillance Cascade Impactor
Network. I. Size Distribution Measurements of Suspended Particulate Matter in
Air. Environ. Sci. Tech. 6: 1019, 1972.
9. Lippman, M. Review of Cascade Impactors for Particle Size Analysis and a
New Calibration for the Casella Cascade Impactor. Industrial Hygiene Journal
20, 406, 1959.
10. Lippman, M. and Harris, W. B. Size Selective Samplers for Estimating
"Respirable" Dust Concentrations. Health Physics 8: 155, 1962.
11. Lundgren, D. A. An Aerosol Sampler for Determination of Particle Concentra-
tion as a Function of Size and Time./. Air Poll Control A,ssoc. 17: 225, 1967.
4-34
-------
12. Magill, P. L.; Holden, F. R.; and Ackley, C. editors. Air Pollution Handbook.
McGraw-Hill, 1956.
13. Miller, K. and de Koning, H. W. Particle Sizing Instrumentation. Paper
#74-48, presented at the 67th Annual Meeting of the Air Pollution Control
Association, June 9-13, 1974.
14. Peterson, C. M. Aerosol Sampling and the Importance of Particle Size. Air
Sampling Instruments for Evaluation of Atmospheric Contaminants. American
Conference of Governmental Industrial Hygienists, 1972.
15. Picknett, R. G. A New Method of Determining Aerosol Size Distributions from
Sampler Data. Journal of Aerosol Science 3: 189, 1972.
16. Stevenson, H. J. R. Sampling and Analysis of Respirable Sulfates. Division of
Air Pollution, U.S. Public Health Service.
1 7. Ranz, W. E. and Wong, J. B. Impaction of Dust and Smoke Particles on
Surface and Body Collectors. Journal of Industrial and Engineering Chemistry
44: 1371, 1952.
18. Fuchs, N. A. The Mechanics of Aerosols. New York: McMillan Co., 1964.
I'1 <:in. | G. and Kurz, J. L. High Volume Air Sampling. Pollut, Eng., p. 30,
ja.i. 19'.j.
) Lo'i Billy W.; Jaklevic, J. M.; and Goulding, F. S. Dichotomus Virtual Impac-
j. loi I.,ir. •• Scale Monitoring of Airborne Particulate Matter. In Fine Par-
ticles, pp. .311-350, B. Y. H. Liu, ed. New York: Academic Press, 1976.
Jl. Dzubay, T. G.; Stevens, R. K.; and Peterson, C. M. Application of the
Dichotomus Sampler to the Characterization of Ambient Aerosols. In: X-Ray
Fluorescence Analysis of Environmental Samples, p. 99, T. G. Dzubay, ed.
Ann Arbor, MI; Ann Arbor Science, 1977.
22. Dzubav T. G , and Stevens, R. K. Dichotomus Sampler —A Practical
Approach to Aeiosol Fractionation and Collection. EPA-600/2-78-112,
June, 1978.
.. > Swdi./, D. B. Demon, M. B.; and Moyers, J. L. On Calibrating of Cascade
Impaciors. Am "id. Hyg. Assoc. J. 34: 429, 1973.
-1 Lippmann, Morton Respirable Dust Sampling. Am. Ind. Hyg. Assoc. J.,
~\ <). it \, il, 1970.
25. Dzuba;, 1 G I lines. L. E.; and Stevens, R. K. Particle Bounce Errors in
Cascade Impactors. Atmos. Environ. 10: 229, 1976.
4-35
-------
High-Volume Air Sampling
Introduction
When air pollution control agencies attempt to determine the nature and
magnitude of air pollution in their communities and the effectiveness of their con-
trol programs, they collect samples of suspended and, sometimes, settlcable par
ticulate matter.
Several different sampling techniques and devices filtration, electrostatic and
thermal precipitation, and impaction may be employed to collect suspended par-
ticulate pollutants from ambient air. Of the various techniques, filtration has been
found to be the most suitable for routine air sampling. The so-called, high volume
(hi-vol) sampler is generally accepted as the instrument of choice for this purpose.
Approximately 20,000 high vols are operating at Federal, State, and local air-
pollution control agencies, industries, and research organizations for either routine
or intermittent use. The number is growing rapidly as air monitoring programs
expand.
The Environmental Protection Agency has designated the High-Volume method
as the reference method for total suspended particulars (TSP) (21)- Certain situa-
tions e.g., sampling for State Implementation Plans (SIP) and Prevention of
Significant Dcterroration (PSD), require the organization responsible for sampling
to use the reference high-volume method when determining ISP.
Development of the High-Volume Sampler
In 1948 Silverman (16) developed an aerosol collector that consisted of a household
vacuum sweeper motor encased in an airtight sheetmetal housing adapted to hold a
4-inch diameter filter. Provision was made for measuring the air flow through the
system Because this sampler operated at a much higher flow rate than other
available samplers, it was identified as a high volume sampler and the designation
persists Replacement of the thin sheet-metal motor housing with one of cast
aluminum by the Staplex Company* in the early 1950s improved the Silverman
sampler Adoption of a stainless steel filter holder to accomodate an 8 by 10-inch
filter permitted 24-hour operation of the sampler and collection of a much larger
sample of paniculate matter than previously possible (see Figure 4-30).
Although the Staplex sampler performed satisfactorily, it was decided after
several years of experience that a more rugged sampler was needed to meet the
requirements of a large scale sampling network operation. Accordingly, rn 1957 a
new high-volume sampler (11), developed in collaboration w,th General Metal
Works' was introduced. The new sampler used a different motor that was com-
pletely enclosed in a cast aluminum housing, eliminating the sheetmetal and rub-
ber components used in the Staplex sampler thus simplifying brush and motor
replacement. This sampler is widely accepted. Other high-volume samplers, some
of which embody slight modifications, have been marketed recently.
•Mention of company or product by name is for identification and information purposes only and
does not constitute endorsement by the Environmental Protection Agency.
4-36
-------
Figure 4-30. Hi-vol sampler.
Equipment
Sampler-Shelter Combination
The sampler and its shelter should be considered as a single, functioning unit (see
Figure 4-3la). The shelter must provide protection for the sampler, and at the
same time allow unrestricted access of ambient air from all directions without
direct impingement of particles on the filter. A high-volume sampler with a
7 x 9-inch exposed filter area operated in a standard shelter at 50 to 60 cubic feet
per minute (1.4 to 1.7 cubic meters per minute) collects particles up to 100
microns in size and uniformly distributes the sample over the filter surface. The
standard peak roof of the shelter, which acts as a plenum above the filter, is placed
to provide a total opening area of slightly more than the 63-square-inch filter area,
thereby permitting free flow of air into the plenum space (see Figure 4-31b). Any
deviation from the size of the opening to the filter or the volume of air filtered per
unit time will affect the particle size range collected. Distribution of particles on
the filter may also be affected.
4-37
-------
Figure 4-31a. Hi-vol sampler with shelter.
Hi volume
sampler in
shelter
Figure 4-3Ib. Air flow of hi-vol sampler in shelter.
4-38
-------
Filter Media
Choice of filter media is influenced by the objectives of the sampling program and
the characteristics of the sampler to be employed. Results of a comprehensive study
of the characteristics of different types of filter media were published in 1961 by
Lockhardt and Patterson (7). An excellent discussion of filter media and filtration
sampling is presented in Air Sampling Instruments (23).
Glass fiber filters, although not perfect in all respects, have been found to meet
most of the requirements for routine particulate sampling. Such filters have a col-
lection efficiency of at least 99.9 percent for particles of 0.3 micrometers and
larger, low resistance to air flow, and low affinity for moisture, all of which are
distinct advantages during sampling. However, in order to eliminate possible
weight errors due to small amounts of moisture, both clean filters and samples
should be equilibrated at 20° to 25 °C at a relative humidity below 50 percent for
24 hours before weighing. Figure 4-32 shows the effect of moisture on the weight of
glass fiber filters. Figure 4-33 shows the effect of moisture on the weight of par-
ticulate matter.
2,506
2,505
2,504
a
^
be
. 9
w *•"'
&C
'5
503
2,502
2,501
2,500
Gelman type A
lot 8005
0
20 40 60 80
Relative humidity, percent
Figure 4-32. The effect ol , dative humidity on the weight
of glass fiber filters at 75°F.
4-39
-------
600
500
be
3 300
be
200
100
Fly.
20 40 60 80
Relative humidity, percent
;, , fii'i i ;i relative humidity on the weight
of atmospheric particulates at 75°F.
100
Samples collected on glass fiber filters are suitable for analysis of a variety of
organic pollutants (13, 19, 20, 25, 26) and a large number of inorganic con-
taminants (18, 21, 27) including trace metals and several nonmetallic substances.
Glass fiber filters are excellent for use in monitoring gross radioactivity (6, 15). It
must be pointed out, however, that because of filter composition, satisfactory
analyses for silica, calcium, sodium, potassium, lead, zinc, and other materials
present in substantial amounts in the filter are not possible. Recent work has pro-
duced an acceptable glass fiber filter for metals analysis (29). Paniculate samples
can be analyzed for a variety of organic pollutants including benzo(a)pyrene, am-
monia (from ammonium salts), nitrates and sulfates, (however, artifact formation
may be a problem) fluorides, antimony, arsenic, beryllium, bismuth, cadmium,
chromium, cobalt, copper, iron, lead, manganese, molybdenum, nickel, tin,
titanium, and vanadium. A random but statistically significant sample of new
filters should be analyzed to determine what the filter blank concentration is and
4-40
-------
whether or not it is high enough to interfere with a particular analysis. It is wise to
obtain this information before purchasing large numbers of filters to avoid poten-
tial problems caused by high blanks. Results of studies of the analyucal aspects of
glass fiber filters in air pollution were reported by Pate and Tabor (10) and
Little (30).
While glass fiber filter material has been dominant in the measurement of total
suspended paniculate, numerous applications have been found for cellulose fillers.
Cellulose filtering media has relatively low metal content making it a good choice
for metals analysis by neutron activation, atomic absorption, emission spectroscopy,
etc Conventional high-volume samplers usually have to be modified to use cellulose
filters because cellulose filters clog rapidly causing flow to sometimes decrease by as
much as a factor of two during a one-day sample (31, 32). Other disadvantages of
cellulose are its irreversible absorption of water (33) and enhanced artifact forma-
tion of nitrates and sulfates. These disadvantages can usually be overcome by using
a control blank filter. Spectro-quality grade glass fiber filters have sufficiently low
background metal contents to make them acceptable for metal analysis, if cellulose
cannot be used.
Precautions in Filter Handling
After the desired filter material has been chosen, it should be inspected visually for
holes and tears that may cause uneven loading, loss of paniculate, and failure
during the sampling period. The aid of a light table may be used to better deter-
mine the quality of the filter paper. A number should be assigned to each filter to
create better data handling. Care should be taken not to tear the filter when affix-
ing the identification number.
Before the weighing of the filter paper, the filter should be equilibrated in a
conditioning environment (20° to 25°C not to vary more than ±3%, relative
humidity less than 50% not to vary more than ±5%) free of acidic or basic gases
that might react with the filter media. The analytical balance used for the
weighing should be calibrated with standard weights between three to five grams
(the average clean filter weight should be in this range). The fillers should be taken
directly from the conditioning chamber to the weighing area 10 minimize the risk
of contaminating the filters. Then they should be weighed 10 the nearest milligram,
and the weight and the number of each filler recorded. The filler must not be
folded or creased before use as this may establish enoneous flow patterns during
sampling.
To install a clean filter in the sampler, the wing nuts are loosened and the
faceplate is removed. The filter should be centered and the gasket should be placed
so that equal spacing of the filter is held by the gasket. The faceplate is then
replaced and the wing nuts tightened. The gasket should not be tightened to the
point where filter damage might occur. Hi-vol filter cartridge assemblies similar to
the one pictured in Figure 4-34 make installation of filters easier. Installation and
removal of a filter can be performed inside a building thus eliminating handling
problems due to lack of space and windy conditions.
4-41
-------
Figure 4-34. Hi-vol cartridge assembly.
When removing the filter, it should be checked for holes or failures and foreign
material (such as insects). Caution should be observed to minimize filter damage
when removing the faceplate and the gasket. If the border appears fuzzy or non-
existent, there may be an air leak around the gasket. The sample should be
discarded and the gasket checked. After it has passed the visual test, the filter
should be folded in half lengthwise so that sample touches sample. This minimizes
sample loss during transport to the lab. If the filter has a nonuniform border, the
paper should stilf be folded so that sample touches sample. Usually the filter is then
put into an envelope and a folder and taken to the lab for weighing and analysis.
The filter should be put into the conditioning atmosphere for 24 hours before
weighing to insure the same humidity and temperature conditions are available for
the"final weighing as were used for the initial weighing. See Figure 4-35 for a sum-
mary of filter-handling procedures.
Documentation of all these steps is important foi legal purposes and to ensure
data reliability. The operator who starts the sampler is responsible for recording
the following information (17).
• Filter number
• Station number and/or address
• Sampler number
• Starting time
• Initial flow rate
• Date and initials
• Summary of conditions that may affect results (e.g., meteorology, construction
activities, etc.).
4-42
-------
The operator who removes the sample is responsible for recording the following
information
• Stop time and elapsed time (if available)
• Final flow rate
• Date and initials
• Summary of existing conditions that may affect results
This documentation is important for calculation of the pollutant concentration
in the atmosphere.
Activity
Filter selection
and collection
efficiency
Filter
identification
Filter
equilibration
Filter weighing
Filter handling
Acceptance limits
1. 0.3 fim diameter
particle col-
lection efficiency
>99%
2. No pinholes,
tears, creases, etc.
Identification number
in accordance with
specifications
Equilibration in con-
trolled environment
for no less than 24
hours. Room: con-
stant humidity
chamber employing
a saturated chemi-
cal solution to
afford an RH of
less than 50% and
constant within
+ 5%. Average
temperature
between 20 and
25 °C with variation
less than ±3%.
Determine indicated
filter weight to
nearest milligram
within 30 sec.
after removal from
equilibration
environment.
Filter in protective
folder and envelopes
undamaged.
Frequency or
method of
measurement
1 . Manufacturing
should furnish
proof of DOP test
(ASTM-D2986-71)
2. Visually check
each filter with
aid of light table
Visually check each
filter
For each sample
observe:
• room or chamber
conditions
• equilibration
period
Gravimetric
Visually check each
filter.
Action if
requirements
not met
1 . Reject shipment
2. Return to
supplier
Identify properly
or discard filter
Repeat equilibra-
tion step after 48
hours or more at
ambient
conditions
Reweigh after 48
hours or more at
ambient con-
ditions and repeat
of equilibration
step
Repack un-
damaged filters,
discard damaged
filters.
Figure 4-35. Summary of hi-vol filter handling procedures.
4-43
-------
Calibration
Most hi-vol samplers use a rotameter (visi-float), flow recording device, or magne-
helic gage for measuring air flow. Since either a fraction of the total sampled air
passes through the measuring device or a pressure differential is measured, calibra-
tion procedures must be performed against a known air flow. A calibration orifice
unit has been designed for these procedures and is pictured in Figure 4-36.
Figure 4-36. Orifice calibration unit.
(Courtesy General Metal Works)
This calibration unit consists of a tube 7.6 cm (3 in.) inside diameter and
15 9 cm (6K in.) long. A static pressure tap is located 5.1 cm (2 in.) from one
end The end of the tube farthest from the pressure tap is flanged to an outside
diameter of about 10.8 cm (4& in.)- This will accomodate the resistance plates
that accompany the calibration unit. The other end of the tube consists of a metal
plate over the air inlet with a hole 2.9 cm (1 1/8 in.) in diameter. This hole is the
orifice on which the calibration procedures are based. Five resistance plates are
provided with the calibration unit representing the resistance of tillers with varying
4-44
-------
paniculate loading. The resistance plates have 5, 7, 10, 13, and 18 holes. The
plate with 18 holes represents a clean filter; the other plates represent a filter with
increasingly heavy dust loading.
To calibrate the orifice unit, a primary standard meter is required. The calibra-
tion unit is attached to the inlet of a positive displacement meter (rootsmeter), and
a hi-vol motor blower unit is attached to the outlet (see Figure 4-37). A series of
steady flows are drawn through the orifice unit for each resistance plate. The air
flow is recorded along with the corresponding pressure differential from the
manometer attached to the pressure tap of the orifice unit. Placement of the orifice
unit before the primary standard reduces the inlet pressure to this meter below
atmospheric. To compensate for this, a second manometer is attached to an inlet
pressure tap of the primary standard. After recording barometric pressure, the true
volume of air drawn through the primary standard is calculated from the following
relationship (24).
(Eq. 4-6) VK=
" /
Where: V '„ = true air volume at barometric pressure, m1
PI,- barometric or atmospheric pressure, mm Hg
Pm- pressure drop at inlet of the primary standard, mm Hg
V, „,,,„ = volume measured by primary standard, m1
Then determine the true air flow rate by:
V
Where: Q= true air flow rate (nWmin)
V , = true air volume (m3)
T = time of flow (min)
A plot of the orifice manometer pressure reading (in. of water) versus the true
air flow rate is generated (Figure 4-38a). This is used as the calibration curve for
this orifice unit. This orifice calibration should remain unchanged over a period of
time unless the unit is damaged.
Once the orifice unit has been calibrated, it can be used to calibrate high-
volume samplers. The orifice unit should be attached to a high-volume motor
blower with the 18-hole resistance plate in place. A manometer should have one
end connected to the orifice unit with the other end open to the atmosphere. The
sampler should be switched on and allowed to run for 5 minutes to set the motor
brushes.
4-45
-------
Thermometer
Mercury
manometer
Barometer
Calibration
orifice
Positive
displacement
meter
(rootsmeter)
Hi-vol
motor
Resistance
plates
Figure 4-37. Diagram of orifice calibration set-up.
The reading indicated by the rotameter, continuous recording device, or
magnehelic is recorded along with the corresponding orifice manometer pressure
value. This is done for all five resistance plates. The orifice manometer pressure
value is converted to flow in cubic meters per minute using the orifice calibration
curve. A plot of the flow rate versus the rotameter, continuous recorder, or
magnehelic reading is plotted and the best fit curve, having not more than one
point of inflection, is used as the calibration curve for the high-volume unit
(Figure 4-38b).
4-46
-------
1
S 8
c '.7
& '6
.5
.4
New conditions
-P; = 630 mm Hg; T, = 299 °K
Original conditions
P, = 755 mm Hg; T, = 299°K
.1 .2 .3 .4.5.6.7.8.91.0 234 5678910 20 3040506080100
Ap in inches H2O
Figure 4-38a. Orifice calibration curve.
c
'g
"^
w
5
c
&
3.0
2.0
1.0
10
20
Visifloat calibration curve
30
40
50
60
70
80
Visifloat reading
Figure 4-38b. Visifloat calibration curve.
4-47
-------
When using an orifice calibration unit to calibrate a hi-vol sampler, corrections
must be made to the indicated flow if the actual atmospheric temperature and
pressure vary from the calibration conditions*. The corrected flow may be
calculated as follows (17):
Where Q., = corrected flow rate, m' min
Qj — uncorrected flow rate, read from orifice calibration curve,
m' min
T, -absolute temperature at orifice calibration conditions, °K
T-, = absolute temperature when calibrating the ^ampler, °K
P} — barometric pressure at orifice calibration conditions, mm Hg
P',= barometric pressure when calibrating the sampler, mm Hg
(Figure 4-38(a) shows new curve drawn to reflect differing conditions of
temperature and pressure).
Standard High-Volume Sampling Procedure
The following procedure is recommended for high volume air sampling
1. Install the sampler in the standard shelter located at a suitable site;
2. Remove the face plate, place a filter in the center of the filter holder and
replace the face plate securing it in place by tightening the four wing nuts to
equalize the pressure and prevent air leakage;
3. Close the roof of the shelter, turn on the sampler and allow it to run for
about 5 minutes; then take a flowmeter reading;
4. At the end of the sampling period (24 hours), take one final flowmeter
reading, then stop the sampler. If the sampler has been shut off auto-
matically, turn on the sampler and let it run for 5 minutes to set the brushes
before taking an air flow measurement;
5. Carefully remove the filter, fold in half along the longer axis making sure
that sample touches sample and place in a protective folder for
transportation;
6. Filter number, starting time, ending time, initial and final flowmeter
readings, sampler location, and weather conditions should be recorded for
each sample collected;
7. Determine the total volume of air sampled by converting flowmeter readings
to actual air flows in cubic meters per minute by referring to the sampler
calibration curve, then multiplying the average of the initial and final flow
rates by the length of the sampling period in minutes; this volume should be
corrected to standaid EPA conditions.
*As a guideline, when pressure differs by as much as 15% and temperature (°C) differs by 100% or
more from the calibration conditions only a 15% error can be expected. Smith, F., Wohlschlegel,
P. S., and Rogers, R. S. C. 1978. Investigation of Flow Rate Calibration Procedures Associated
with the High Volume Method for the Determination of Suspended Particulates. EPA report no.
600/4-78-047.
4-48
-------
8. Calculate the mass concentrations of suspended particles using the following
equations:
(Eq. 4-8)
(nif — m, \
(Eq. 4-9) TSp ={-——] XIO*
Where: TSP=mass concentration of suspended paniculate,
m, — initial weight of filter, g
mf-final weight of filter, g
ff—final flow rate, mVmin
f, = initial flow rate, mVmin
t = time, min.
V = air volume sampled, m3
106 = conversion of g to fig.
Filters from hi-vol samplers can be analyzed much more thoroughly than is possi-
ble from a simple mass concentration determination. After a filter has been
weighed, it can be cut into sections for subsequent analysis. Destructive techniques
that might be applied are organic solvent extraction, acid extraction, and aqueous
extraction.
Atmospheric Organic Analysis
An organic solvent extraction is made with a portion of the filter to determine the
amount of organic aerosol present in the ambient air. A solvent such as benzene or
chloroform is used in conjunction with a Soxhlet extractor to remove organics from
the filter. This organic fraction can be further treated for analysis by infrared or
ultraviolet spectrometry.
Atmospheric Metals Analysis
Acid extraction is performed on another segment of the filter paper. The filter can
either be placed in an acid bath for several hours or it may be extracted in either a
Soxhlet or Bethge apparatus. Several acid mixtures have been suggested for extrac-
ting the metals from the filter, yielding solutions of soluble salts. Any filter matter
that disintegrates in the acid can be removed by filtration or centrifugation. After
further preparation, the soluble metal solution can be analyzed by a number of
methods, including atomic absorption spectrophotometry, emission spectroscopy,
and polarography.
Water Solubles Analysis
A third portion of the filter is extracted into deionized, distilled water. This extrac-
tion will solubilize sulfates, nitrates, and other water-soluble anions and cations.
The extracted material is filtered. Water-soluble species can then be analyzed using
known methods.
4-49
-------
Other Analyses
The filter paper that remains is saved to be used for more nonroutine analyses, or
for any other method if it must be repeated because of some error. These other
analysis methods might include nondestructive neutron activation analysis, x-ray
fluorescence analysis, electron microscopic analysis, etc.
It should be noted that the above-mentioned analysis techniques are not required
for all filters. Considerable thought must be used to decide what pollutants should
be determined from the filters and what method should be used for subsequent
analysis for obtaining the most meaningful results.
Sampling Accuracy and Precision
The limits of accuracy and precision of any sampling method must be understood
for proper interpretation of data obtained using that method. Factors influencing
the accuracy and precision of high-volume sampling include sampler operating
characteristics, accuracy of calibration, filter characteristics, location of sampler,
nature and concentration of particulates and gases in air being sampled, and the
temperature and humidity of the air.
Accuracy
Accuracy may be defined as the extent to which a measurement agrees with the
true value of the quantity measured. In sampling for suspended particulates, there
is no standard method for determining true values; consequently, accuracy as
defined above cannot be measured. Howevei, there are several situations that are
known to lead to inaccurate measurement of suspended paniculate concentrations.
When sampling is conducted over a 24-hour period, the final air flow rate is
usually lower than the starting flow rate, the difference being greater when the
particulate level is high and/or the paniculate matter has a strong tendency to
plug the filter medium. The change of flow rate with time cannot be predicted and
can only be determined by taking air flow readings at intervals throughout the
sampling period. Since it would be impractical to take such a series of readings for
each sample collected, the currently accepted procedure is to compute the volume
of air sampled, by assuming a linear decrease of flow during the sampling period,
i.e., multiplying the average of the initial and final flow rates by the length of the
sampling period. Air flow measurements obtained by averaging initial and final air
flow rates have been compared with those obtained by use of several air flow
measurements taken during the sampling period. The results indicate that
reasonably accurate flow measurements can be obtained by the use of accurate flow
rates. When sampling is conducted in urban areas with high particulate levels,
errors of as much as 10% may occur. Fortunately, such incidents are few in
number. In urban areas with low particulate levels and in nonurban areas the
change in air flow over the normal 24-hour sampling period is relatively small;
therefore, the sampling rate is practically constant and the volume of air sampled
may be accurately determined.
4-50
-------
In some communities inaccurate air flow measurements may result from frequent
and sizeable changes in the operating voltage. Use of a constant voltage
transformer between the sampler and the power line will ensure against excessive
voltage changes.
Modification of the high-volume sampler to permit a continuous recording of air
flow during the sampling period provides a more accurate measure of the volume
of air sampled. This modification includes the incorporation of a fixed orifice into
the motor housing, use of a pressure transducer to monitor the air flow, and a
chart recorder to make a continuous recorder of the air flow during the sampling
period (Figure 4-39).
Figure 4-39. Hi-vol setup with flow transducer.
Chahal and Hunter discuss the use of an orifice meter as a more accurate means
to measure hi-vol flow rate (see Figure 4-40) (34). Pressure drop across the orifice
could be monitored with a pressure transducer for a continuous record of flow.
This modification can be performed after the purchase of a hi-vol. If the adapted
hi-vol is to be used for Federally required monitoring, permission must be asked for
and given for this sampler to be used.
4-51
-------
Filter support (screen wire)
Pressure
taps
Hi-vol motor
Figure 4-40. Hi-vol sampler with pressure taps.
Non uniformity in sampling rate is a more serious problem troin the standpoint
of the representativeness of the sample. To accurately measure the average
pollutant level over any sampling period, sampling must be done at a uniform rale.
Obviously, a 24-hour sample collected with starting and ending flows of 60 and
40 cfm will not accurately represent the average paniculate concentration over the
sampling period; the final result will be under- or overestimated, depending on
whether the paniculate level was higher or lower at the start than at the end. For-
tunately, it has been found that, except for in the most heavily populated areas
and during prolonged stagnation periods, the drop in air flow rate is minimal and
for all practical purposes the sampling rate may be considered essentially uniform
for the whole sampling period. However, the largest divergence from uniformity of
sampling rate is found in the more heavily polluted areas where accurate informa-
tion is greatly needed. Non-uniformity of sampling rate and inaccuracies due to
nonlinear decrease in air flow can best be eliminated by the use of constant flow
samplers, which will be described later.
It has also been shown that the total paniculate collection is dependent on the
flow rate chosen (9). In a test where identical lu-vols at flow rates between 30 and
60 cfm were tested side by side, the hi-vol running at 60 cfm obtained a paniculate
value 1% higher than the 50 cfm hi-vol. The 40 cfm hi-vol obtained a value of
3.8% higher than the 60 cfm hi-vol. The 30 cfm hi-vol obtained a value 7.4%
higher than the 60 cfm hi-vol. This difference may be attributed to the more
efficient trapping of small particles at lower flow rates by glass fiber fillers (1).
Other errors associated with hi-vol particulate measurements are:
• Filters have been shown to gain weight during idle or nonsampling periods
(35, 36);
• The roof design causes the hi-vol to be wind-direction sensitive;
• Nitrates and sulfates (artifacts) are formed on the filter by reaction (both glass
fiber and cellulose) during and after sampling causing erroneously high par-
ticulate concentrations (41, 42).
4-52
-------
Precision
Precision can be thought of as reproducibility of results obtained by measurements
made under the same conditions, at the same time, following identical procedures.
The precision of high-volume sampling can be demonstrated by using data col-
lected in a recent investigation of the comparative collection efficiencies of glass
fiber filters from two different sources.
Four carefully calibrated and matched high-volume samplers were set up at the
corners of an 8-foot square on the roof of a building located in the Mill Creek
Valley of Cincinnati. All samplers were started at the same time and had identical
initial air flow rates; sampling was terminated at the same time on all samplers.
Each day two samplers were fitted with the Gelman Type A filters and two with
MSA 1106 BH filters, the samplers being randomly selected each day.
During the study, 40 samples were collected on each of the two types of filters.
Average suspended particulate concentration measured was 101 //g mf with the
Gelman fillers and 100.5 /xg m:) with MSA filters. Thus, one ran conclude thai the
collection efficiencies of the two filters are identical, and, thereiore, the study
results can also be used to evaluate the precision of high-volume sampling. Results
for the 20-day period for each of the samplers averaged 97.3, 101.3, 101.9, and
100.7 /(g/rn1 for an overall average of 101.3 /iig in*. A careful and detailed
statistical analysis of the data showed no significant differences between samplers.
This study thus demonstrated the precision of high volume sampling when samplers
have been accurately calibrated and properly operated. However, this was a
controlled experiment and does not necessarily represent the hi vol precision in
actual use.
To insure the quality of data reported from high-volume sampler networks, an
accurate, quick, and easy audit technique is needed. One such device, a Reference
Flow device (ReF device) has been developed by the National Bureau of Standards
under contract to the EPA. This device provides a quick, easy, and accurate
technique for flow calibration audits. The ReF device is a modified orifice designed
for ease of placement onto a high-volume sampler without dissassembling the
sampling unit.
For a more detailed description of the ReF device and audit procedure, see the
companion manual to this one, the Laboratory and Exercise Manual for course
435, EPA no. 450/2-80-005, p. 2-27, 2-35.
Constant Flow Samplers and Devices
As mentioned earlier one of the largest errors in hi vol measurements is in the
sample flow rate. Combining inaccurate, varying flows with the fact that par
ticulate concentration varies during the sampling period can cause relatively large
errors in TSP measurements. Therefore, constant flow through the hi-vol sampler
should be maintained during the sampling period to ensure accurate TSP
measurements. Constant flow control could also facilitate the use of membrane
filters 10 measure heavy metals with the hi-vol sampler. The following regulalion
systems are commercially available to help achieve constant hi-vol flow.
4-53
-------
Mass Flow Regulation
This method of flow regulation uses a constant temperature thermal anemometer
sensor to measure mass flow in the neck of the hi-vol sampling head. Electronic
feedback circuitry adjusts the motor speed to maintain a constant mass flow. Since
mass flow is controlled, the volumetric flow rate is maintained at standard condi-
tions. Thus, flow variations caused by temperature, pressure, line voltage, and par-
ticulate loading are all compensated for by this system. Flow rates can be selected
over a wide range allowing the same regulator to be used for standard reference
method sampling or for a high-volume cascade impactor. This unit can be retro-
fitted to any hi-vol.
Pressure Sensitive Regulators
A second type of regulator uses a pressure sensitive switch connected to the exhaust
of the sampler to activate a motor-driven variable transformer that controls the
speed of the blower motor. Thus, as the sample builds up during sampling, with a
corresponding increase in resistance to air flow and lowering of exhaust pressure,
the motor speed will be automatically increased to create a higher vacuum and
maintain a constant flow rate. Such a device is sensitive and provides a uniform
flow. This flow regulator, which operates in the 5 to 50 cfm range, is distributed
either as a separate unit or as a component of one constant flow sampler.
Sampling of Respirable Particulates
Samples collected by means of the standard high-volume sample-glass fiber filter
combination represent a wide range of particle sizes from 100 micrometers down to
those in the submicrometer range. While such samples may be adequate for the
definition of total suspended particulate pollution, they are not suitable for the
evaluation of potential health effects because only those particles smaller than
10 micrometers are of significance in this respect (Figure 4-41). Various techniques
are available for use with the high-volume samplers to permit selective collection of
the fraction that is commonly referred to as "respirable particulates" (see Figure 4-27).
A simple technique for respirable particulate sampling, which has not gained
wide use, was developed and reported by Roesler (12). This method requires a sim-
ple modification of the sampler faceplate to permit the insertion of a i/£-inch sheet
of 66 ppi polyurethane foam in front of the glass fiber filter, plus the installation of
baffles around the perimeter where the air enters the shelter. The polyurethane
foam removes practically 100 percent of the larger particles that pass through and
smaller (respirable ones) are collected on the filter (Figure 4-42). Using this pro-
cedure, a study was made of the relationship between the respirable and total
suspended particulates in Five cities. The average ratio of respirable to total par-
ticulate matter ranged from 0.48 (St. Louis) to 0.586 (Cincinnati) (Figure 4-43).
Detailed analyses of samples (Figure 4-44) indicate that by far the greater portion
of benzene-soluble organics, sulfates, nitrates, and lead (three of four cities) is
found in the respirable fraction. This has gained considerable significance in the
investigation of health effects of air pollution.
4-54
------- An error occurred while trying to OCR this image.
-------
City
Chicago
Philadelphia
Cincinnati
St. Louis
Washington
Minimum
0.400
0.425
0.542
0.375
0.415
Maximum
0.648
0.924
0.678
0.677
0.540
Average
0.494
0.547
0.586
0.477
0.490
Figure 4-43. Suspended particulate matter ratios on a concentration basis.
(Ratio = respirable fraction/total collection)
Pollutant
Benzene-soluble
Nitrates
Sulfates
Iron
Lead
Manganese
Chicago
0.69
0.64
0.84
0.32
0.47
0.43
Cincinnati
0.79
0.62
0.96
0.41
0.62
0.62
Philadelphia
0.77
0.53
0.84
0.39
0.69
0.50
St. Louis
0.88
0.56
0.87
0.33
0.68
0.44
Figure 4-44. Respirable/total paniculate matter ratios for selected pollutants.
If air is drawn through a properly designed elutriator at a suitable rate, then
through a glass fiber filter on a standard high-volume sampler, the larger particles
settle out in the elutriator and the smaller ones arc collected on the tiller. This
device permits a rough separation of particulates into the respirable and non
respirable fractions but is too cumbersome for use in most situations.
When the high-volume sampler is operated with a cyclone separator of suitable
size connected ahead of the filter, a separation into respirable and nonrespirable
fractions is accomplished. In this instance, using the proper flow rate and cyclone
separator, the particle size distribution of the sample collected on the filter closely
resembles that of the particles taken into the human respiratory tract.
A recent study used hi-vols with cyclones that were capable of attaining 50%
particle retentions at 3.5, 2.5, 1.5, and 0.5 /*m aerodynamic diameters (37). Since
the cyclones removed the majority of the particles above these diameters, the hi-vol
backup filter could be used as a measure of the particles below the cut points.
Cascade impactors are available for use with high volume samplers. Cascade
impactors provide data for a complete size distribution or respirable particles. Two
types of high-volume cascade impactors are currently sold. One type uses
multiorifice plates with round orifices and operates at 20 ftVnnin while the other
type uses multislit plates and operates at 40 fWmin. For a complete discussion of
high-volume cascade impactors refer to the section in this chapter on "Impaction
Devices.'
4-56
-------
Respirable particulate sampling by the high-volume attachments has shown
many disadvantages. Variability in the polyurethane foam made results inconsis-
tent. Errors normally associated with cascade impactors (i.e., particle bounce and
shatter, wall losses, etc.) and the inaccessibility of collected particulate for analysis
made them undesirable for routine field sampling of respirable particulates.
Cyclones and elutriators were too bulky for routine sampling. Because of these and
other disadvantages EPA has spent much time and effort developing an acceptable
sampler for respirable particulates. A dichotomous sampler was developed and
shows great promise as a routine sampler for respirable and total particulates (for a
complete discussion of the dichotomous sampler refer to the section in Chapter 4,
"Dichotomous Sampler").
Sampling for respirable particulates is already becoming a widely accepted pro-
cedure. Once the procedures have been standardized and adopted for routine
sampling, the study of the nature of respirable particulates will become an integral
part of all State and local sampling programs.
Maintenance
Although the motors used in the Silverman type of high-volume samplers were
originally designed for use in household vacuum sweepers, their performance as air
sampling devices has far exceeded what one would expect from a component of a
household appliance. Modifications to the air mover have made them even more
dependable. Naturally, brush and motor life have been considerably shortened due
to the excessive demands of air sampling for longer continuous operation of the
sampler. Brushes wear out and have to be replaced at frequent intervals, and new
motors must be installed when brush wear becomes excessive. Motor and brush life
can be extended substantially without seriously decreasing the amount of sample
collected if the sampler is operated at a slightly reduced voltage by use of a "Buck
or Boost" transformer. Many constant flow manufacturers claim increased brush
life because the voltage is "bucked" by the controller. Occasional cleaning and/or
replacement of the flowmeter, flowmeter tubing, and face plate gasket, and
replacement of the rubber gasket between the motor and sampling head, complete
the maintenance requirements.
Application of High-Volume Sampling
Samples and data obtained by high-volume air sampling may serve many purposes,
several of which are discussed below.
Nature and Magnitude
The operation of high-volume samplers at strategic locations in a community will
result in the collection of samples collectively representing the particulate pollution
of the community. Quantitative analysis of the samples provides information on the
ambient concentrations of the various pollutants. Figure 4-45 (22) illustrates the
type of data obtained, which helps to define the nature and magnitude of par-
ticulate pollution. Such information also provides clues as to the existence of poten-
tial health hazards.
4-57
-------
Urban"
Average
cone
Si
S
K
Ca
Ti
Fe
Zn
Br
Pb
Fine
29 ng/m3
1.00%
12.50
0.40
0.70
1.10
1.40
0.35
0.33
2.20
Coarse
22 /ig/m3
8.00%
1.40
1.20
8.20
2.00
4.80
0.20
0.16
0.60
Rural"
Fine
26 ng/m3
0.50%
12.60
0.30
0.50
< 0.10
0.30
0.13
0.06
0.51
Coarse
15 ftg/m3
4.00%
0.90
0.90
4.20
0.20
1.30
0.15
0.04
0.11
"Located at the Missouri Botanical Garden in St. Louis.
"Located in an agricultural area in Illinois, 40 km south of St. Louis.
Fine = particles less than 3.5 microns in diameter.
Coarse = particles greater than 3.5 microns in diameter.
Figure 4-45. Mass and percentage composition of size-fractionated
St. Louis aerosol samples from August 18 to September 7, 1975.
Trends
Data obtained in a high-volume sampling program of many years duration may be
used to predict trends and the air pollution levels in a community. Detection of
trends by the examination of data on a sample basis is difficult due to the wide
ranges of values; however, application of sophisticated statistical smoothing tech-
niques offers a solution to this problem (18). Seasonal trends may be detected by
limited smoothing (Figure 4-46), whereas more drastic treatment is required to
demonstrate long range trends (Figure 4-47). Likewise, a properly designed
sampling program may be used to monitor the effectiveness of an air pollution con-
trol program. The same statistical treatment of the data serves this purpose also.
Yearly trends for national air quality and emissions are published by the Office of
Air Quality Planning and Standards of the Environmental Protection Agency.
u
Is
O,~J.
"8 *
5
o
H
500
400
300
200
100
0
I
Winter
Spring
Summer
Fall
1979
Figure 4-46. Seasonal trends in concentrations of suspended paniculate matter.
4-58
-------
, 1 1, 38, 39). Moie lecem
woik has requned kilogram quantities ol organic inalciials, which have- been
obtained by extiacting low blank used lilteis from the ventilating systems ol large
office buildings.
4-59
-------
The hi-vol can be- used lor the above* purposes but it inherently lacks the abihiv.
to provide continuous data on paniculate concemiation. Nephelomen v, one- ol the
more- acceptable methods that gives a continuous paniculate concentration, uses
the1 light scattet ing properties of aerosols. Kiet/sc limai ( If)) found ihai
nephe-lometers coneUued we'll with ISI* measuic-mc-nis made In die- irleieme In v •>!
mc-ih'id h'uitlu'i woik is needed to develop a samplei in.). ,u < uia-elv measuies pai
iic'ula e c oncc-nti ai ion on .1 teal-time basis.
Summary
rbe high volume- samplcM has be-comc- the' most \videiv used >ooi loi momtoimg
paniculate1 an pollution. It is a letw cost, portable. casiK in.iim ,nned, and
reasonably piecise sampling device1. ImprovcMnc-nts in sample-i pet lot mance- have
resulted liom incorporation ol automatic timcMs. liou i.ii<' lecouleis, consiant liow
device's, and si/.e1 se-paration de-vice's, i i o the1 basic svsteni Aiihongh In vols have-
bc'C't) the' accepted standaid in p.n t iculale monnoimg seveial ne-eds lemam m he
liilc'd. One- need is loi a samplei dial sepaiaie-s iesj>n;d)ie ,md nonie'spn able- p.n
ncles loi gi.uiinc'ine and chc'inical analysis. 1 tie Ju n i > niMis sample-' ma\ n-piacc-
the hi-vol for these paniculate measurements. Another need is for instrumentation
that provides "real-time" paniculate measurements.
Sample Problem
l)e-ie'i mine- i he- pan te le mass e once-ni i .11 ion h om . iie- |. .1 >\\ i ;ig I li.ei and I lo\\
i ate- data
Mass ol Hiliei :
j » i > 1 O i >
lie-lot e- _ _ .. __ _> to- g
All CM . ..._ _ .. - . ' 1 >-• g
Flow Rate-
Si ,111 _ _ .._ - . I 70 in1 mm
F,nisi! _ .. _____ .1 11 m' mm
1
Sun. __ - ... Mi'huglu S I 1 "/ }
1-inish . _ . _ Midmgh. IS \1 . 1
Solution:
1.70+ 1.11 . ...
AI't'xive Sampling Rdle — -: !..),).> m mi.;
*•/
Samblnw Volume ^
mn.
M <>j Collected Material = 3. 455 - 3. 182 --= ().27:i ^
— "— )-^— 121. 9 X 10 !' e/m3 X 106 ^
2239.2 in1 S %
Concentration--- 122 ^g, 'in1
4-60
-------
Questions
1. A high volume filter weighed 2.1600 grams beloie Campling. The filler was
installed al midnight on 12-25-77. The inilial flow rale was 1.50 m' inin. and
ihe final flow was 1.32 m'Vmin. The sampler was shut down at midnight on
12-26-78. The filter weighed 2.4508 grams alter sampling. Wluu was the 1'SP
concentration? (Ans.: 143 /ig/m'1)
2. Why is it necessary 10 sample for respirable parlieulaies.-'
3. Give several reasons why constant flow regulation ol high volume sampleis is
nee essai v.
t. Paniculate concentrations by the high volume method haw leceiuly been
doubted because ol sevcial sampling enors. Whai aie some ol , hese euois/
References
1. Cohen, A. L. Dependence of Hi-Vol Measurements on Airflow Rate. Environ.
Set. Tech. 7: 60-61, January 1973.
2. Sawicki, E., and Golden, C. Ceramic Alumina in the HPLC Determination of
Benzo(a)pyrene in Air Paniculate Material. Analytical letters, 9: 957-973, 1976.
3. Eisenberg, W. C. Fractionation of Organic Material Extracted from Suspended
An Particulate Matter Using High Pressure Liquid Chromotography./ourna/ of
Cliromatographic Science 16: 145, April 1978.
4. Harrison, W. K.; Nader, J. S.; and Fugan, F. S. Constant Flow Regulators for
the High-Volume Air Sampler. American Industrial Hygiene Association Journal
22: 115, 1960.
5. l.ee, M. L., et al. Gas Chromatography/Mass Spectrometric and Nuclear
Magnetic Resonance Determination of Polynuclear Aromatic Hydrocarbons in
Airborne Particulates. Analytical Chemistry 48: 1566, 1976.
6. Lockhan, L. B., Jr., and Patterson, R. L., Jr. Intercalibration of the Major
North American Networks Employed in Monitoring Airborne Fission Products.
X R L Report 6025, U.S. Naval Research Laboratory, Washington, B.C., 1963.
7. Lockhart, L. B., Jr.; Patterson, R. L., Jr.; and Anderson, W. L. Characteristics
of Air Filter Media Used for Monitoring Radioactivity. N R L Report 6054, U.S.
Naval Research Laboratory, Washington, D.C., 1964.
8. McKee, H. C., et al. Collaborative Testing of Methods to Measure Air Pollutants
I. The High-Volume Method for Suspended Particulate Matter.
/ Air Poll. Control Assoc. 22: 342, 1972.
9. Miller, K., and deKoning, H. W. Particle Sizing Instrumentation. Paper #74-48,
presented at the 67th Annual Meeting of the Air Pollution Control Association,
June 9-13, 1974.
10. Pate, J. B., and Tabor, E. C. Analytical Aspects of the Use of Glass Fiber Filters
for the Collection and Analysis of Atmospheric Particulate Matter. American
Industrial Hygiene Association Journal 23: 144, 1962.
11. Robson, C. D., and Foster, K. E., Evaluation of Air Particulate Sampling Equip-
ment. American Industrial Hygiene Association Journal 24: 404, 1962.
4-61
-------
12. Roesler, J. R. Application of Polyurethane Foam Filters for Respirable Dust
Separating./. Air Poll. Control Afsoc. 16: 30, 1966.
13. Sawicki, E., et al. Benzo(a)pyrene Content of the Air of American Communities.
American Industrial Hygiene Association Journal 21:443, I960.
14. Sawicki, E., et al. Polynuclear Aromatic Hydrocarbon Composition of the
Atmosphere in Some Large American Cities. American Industrial Hygiene
Association Journal 23: 137, 1962.
15. Setter, L. R.; Zimmer, C. E.; Licking, D. S.; and Tabor, E. C. Airborne Par-
ticulate Beta Radioactivity Measurements of the National Air Sampling Network
1953-1959. American Industrial Hygiene Association Journal 22: 19200, 1961.
16. Silverman, L., and Viles, F. G. A High-Volume Air Sampling and Filter
Weighing Method for Certain Aerosols./. Indust. Hyg. and Toxicol. 30: 124,
1948.
17. Quality Assurance Handbook for Air Pollution Measurement Systems, vol. II,
Ambient Air Specific Methods, Section 2.2, EPA-600/4-77-027a.
18. Spirtas, R. Personal Communication.
19. Tabor, E. C. Pesticides in Urban Atmospheres./ Air Poll. Control Assoc. 15:
415, 1965.
20. Tabor, E. C.; Hauser, T. R.; Lodge,]. P.; and Burtschell, R. H. Characteristics
of the Organic Paniculate Matter in the Atmosphere of Certain American Cities.
A MA Arch. Indust. Health^: 58, 1958.
21. Air Pollution Measurements of the National Air Sampling Network — Analysis of
Suspended Particulates, 1957-1961, Public Health Service Publication No. 978,
Washington, D. C., 1962.
22. Stevens, R. K. and Dzubay, J. G. Dichotmous Sampler—A Practical Approach to
' AerosolFractionation and Collection, p. 13, EPA-600/2-78-112, June, 1978.
231 Air Sampling Instruments. American Conference of Governmental Industrial
Hygienists, Cincinnati, Ohio, 1972.
24. Reference Method for the Determination of Suspended Particulates in the
Atmosphere (High Volume Method). Federal Register, September 14, 1972 or 40
CFR 50 Appendix B.
25. Gordon, R. J. Distribution of Airborne Polycyclic Aromatic Hydrocarbons
Throughout Los Angeles. Environ. Sci. Tech. 10: 370, 1976.
26. Ketseridis, G.; Hahn, J.; Jaeniche, R.; and Junge, C. The Organic Consti-
tuents of Atmospheric Particulate Matter. Atmos. Environ. 10: 603, 1976.
27. Paciga, J. J., and Jervis, R. E. Multielement Size Characterization of Urban
Aerosols. Environ. Sci. Technol. 10: 1124, 1976.
28. Sugimas, A. Sensitive Emission Spectrometric Method for the Analysis of Air-
borne Particulate Matter. Anal. Chem. 47: 1840, 1975.
29. Gelman, C., and Marshall, J. C. High Purity Fibrous Air Sampling Media.
Am. Ind. Hyg. Assoc. J. 36: 512, 1975.
30 Arthur D. Little, Inc. Development of a High Purity Filter for High
Temperature Particulate Sampling and Analysis, EPA-650/2-73-032 (Nov.
1973).
31. Dams, R.; Heindrychx, K. A High Volume Air Sampling System for Use with
Cellulose Filters. Atmos. Environ. 7: 319, 1973.
4-62
-------
32. Neustadter, H. E., et al. The Use of Whatman-41 Filters for High Volume Air
Sampling. Atmos. Environ. 9: 101, 1975.
33. Demuynck, M. Determination of Irreversible Absorption of Water by Cellulose
Filters. Atmos. Environ. 9: 523, 1975.
34. Chahal, H. S. and Hunter, D. C. High Volume Air Sampler: An Orifice Meter
as a Substitute for the Rotameter./. Air Pollut. Control Assoc. 26: 1171, 1976.
35. Bruckman, L., and Rubino, High Volume Sampling: Errors Incurred During
Passive Sample Exposure Periods./. Air Pollut. Control Assoc. 26: 881, 1976.
36. Chahal, H. S., and Romano, D. J. High Volume Sampling: Effect of Wind-
borne Paniculate Matter Deposited During Idle Periods./. Air Pollut. Control
Assoc. 26: 885, 1976.
37. Bernstein, David M., et al. A High-Volume Sampler for the Determination of
Particle Size Distributions in Ambient Air./ Air Pollut. Control Assoc. 26,
1069, 1976.
38. Dong, Michael; Locke, D. C.; and Ferrand, Edward High Pressure Liquid
Chromatrographic Method for Routine Analysis of Major Parent Polycyclic
Aromatic Hydrocarbons in Suspended Paniculate Matter. Analytical Chemistry
48: 368, 1976.
39. Fox, M. A., and Staley, S. W. Determination of Polycyclic Aromatic
Hydrocarbons in Atmospheric Particulate Matter by High Pressure Liquid
Chromatography Coupled with Fluorescence Techniques. Analytical Chemistry
48: 992, 1976.
40. Kretzschmar, J. G. Comparison Between Three Different Methods for the
Estimation of the Total Suspended Matter in Urban Air. Atmos. Environ. 9:
931, 1975.
41. Spicer, C. W., et al. Sampling and Analytic Methodology in Atmospheric Par-
ticulate Nitrates. Environmental Protection Agency, Report No. 600/2-78-067.
42. Contant, R. W. Aerosol Research Branch Annual Report FY 1976/76 A.
4-63
-------
Evaluation of Filter Media
Introduction
There has been a tremendous increase in the use of filter media lot collecting air-
borne paniculate matter; in the fields of air pollution and environmental radiation
surveillance, it is the primary means of paniculate sampling. Many circumstances
have contributed to this situation, among which are the low cost and simplicity of
filter sampling. Topics to be discussed in this section include the basic advantages
and disadvantages of air sampling with filters, filtration iheoiv. and some ol the
criteria necessary for the selection of a filter media to be used in a specific
sampling program. Although the term filter media can be extended to cover a
large number of media, such as filter thimbles and granulai beds, this discussion
will be confined to the more' common media available in sheet form commonly
referred to as filter papeis.
General Considerations
Advantages
There are several advantages of filter sampling for particulaies over other methods.
A primary one is the feasibility of handling large volume tales of flow. Some dust
sampling instruments, such as midget impingers and thermal ptecipitators, do not
have ibis capability for large flowrates. Also, after collection, the tillered sample is
usually readily available for direct observation.
The number of sizes of filters available has pioved to be anotliei advantage. By
changing the size of the paper, the volume of air sampled can be varied while still
maintaining the same linear flowrate through the filter. The- selection of sues also
allows filter holders to be designed for use in a variety of situations. This is a
definite advantage when the sampling space confines limit accessibility. I'he
variability among filters extends beyond the matter of si/.c- alone. Appropriate
fillers can be obtained thai are capable of sampling ovei a wide range of
environmental conditions of temperature, humidity, and dust loadings. Further,
fillets can be found thai are adaptable to analysis schemes tanging from
microscopic examination to elaborate chemical separation sche
eines.
Disadvantages
The use of filter media for sampling is not without iis difficulties, one of which is
related to the variation in physical and chemical properties among a quantity of
any given filter papers. In the case of an impaction instrument, once its operating
characteristics have been determined these should remain relatively fixed. On the
other hand, in sampling with filters, the paper is changed between each sample
collection. Although filters can be obtained that minimize the variability between
individual filters of a specific type, it is more common to find significant dif-
ferences in performance, particularly between different lots ol filters. Smith and
Nelson (13) have prepared guidelines for a quality assurance piogram, which, if
followed, will enable the user to control the effects on sampling and analysis
resulting from inconsistencies between filters.
4-64
-------
Another case in which filters are at a disadvantage is in conjunction with selec-
live particle sizing. A cascade impactor can, theoretically, be designed so that par-
ticles of different'size ranges can be collected on different stages of the impactor.
Such a size separation can hardly be done with filter media, although some
gradations are possible (17).
The Theory of Filtration
The most common misconception about air filtration is that it is primarily a sieving
mechanism. If this were the case, only particles larger than the pore size would be
trapped and the theory of collection would be relatively simple. Actually, outside of
the kitchen strainer, very few media could be classified as pure mechanical
strainers. In air filtration a number of mechanisms contribute to the collection of
particulates. The degree to which each one contributes is a function of a number
of parameters, discussed below. For a more complete discussion of filtration theory-
see R. D. Cadle's book on The Measurement of Airborne ^articulates (Wiley
Inteiscience, 1975).
Diffusion
The collection ol particulai.es on a filter by diffusion depends on a particle concen-
tration gradient between the filter and the air passing between the fibers. The
highly concentrated particles in the air stream diffuse to the filter fiber where the
concentration is near zero. The diffusion theory further postulates that when the
particle comes in contact with the filter it remains there. The contribution made
by diffusion depends on the transit time of particles through the filter, with a
longer time resulting in greater diffusion contribution. The amount of diffusion
will then depend on linear rate of flow, filter thickness, size of particle, interfiber
distance, and particle concentration in the air. The effect of diffusion increases as
particle size approaches molecular dimensions.
Direct Interception
Direct interception can be considered as that part of the filter collection
mechanism that is analogous to mechanical straining. The interception mechanism
takes place when a particle following its air movement streamline comes within a
distance from the filter material which is equal to, or less than, the particle radius
so that it comes in contact with the filter medium. As with simple straining, this
type of collection predominates where the particles are greater in diameter than the
interfiber distance, or pore size. The effectiveness of direct imeiception increases
with increasing particle size.
Inertial Collection
As a particle is carried by an air stream it possesses a certain amount of inertia,
depending on its mass and velocity. When a sufficient inertial force has been
established, the particle will, as the air stream turns, leave its streamline and con-
tinue on its previous path. If the inertia of a particle causes it to strike a filter fiber
4-65
-------
during the passage of the air stream around the fiber, the particle will be collected.
The mechanism of inerdal collection plays a major pan where high linear velocities
of airflow are present. Increasing the particle size and deeieasing the iiber
diameier (or pore diametei) improves the effect of inertia! collection.
Electrical Forces
II the aerosol particle has an electrical charge, and the filtei liber has a charge of
opposite sign, the particle will be attracted to the tiller media. Many investigalois
feel ihat such a mechanism has a part to play in the filter sampling of air.
Experiments (5) have indicated that both atmospheric panicles and filter media
possess electrical charges and that collection does occur by this means. The
magniiude ot the effect is not well known, but it probablv i-. a detinue contribution
10 the process of collecting panicles smaller than the poie si/e.
Combined Factors
Duiing actual filtration all ot the above-mentioned collection mechanisms are
working together so that it is often difficult to sepaiate one horn another. Fuilliei-
moie, othei mechanisms occur that complicate the situation. All of the methods of
collection that have been discussed postulate that once a panicle touches the filler
material it stays there. This is not entirely so -there is some reenirainment of par
tides resulting from the airflow through the filter picking up and carrying inaienal
that has been previously deposited some distance before the particles are once- more
entrapped by the filter fibers. Increasing the filter lace velocity will increase the
chancc-s of reenirainment.
Collection Efficiency of Filters
In any procedure where an attempt is made to relate a sample activity to an
activity concentration in the environment, the volume ol original sample must be
known. In the sampling of air particulates by filtration, not all particles in the air
are collected. That is to say, the collection mechanisms that have been discussed
are not 100% efficient. It is, therefore, not sufficient just to know what volume of
air was passed through the filter, the fraction of the airborne particles collected
must also be- known. Collection efficiency may be stated in terms of several
parameters. The most common is to determine the percent penetration (percent
passing through the filter) of a certain particle si/.e as a I unction of the linear
velocity (volume rate of flow divided by the filter area) through the filter. Another
way of presenting the efficiency would be to have percent collection of a certain
particle si/.e as a function of linear velocity through the tiller. Although efficiencies
of this type are useful, it would be most difficult to use them to determine the frac-
tion of the atmospheric particulates collected unless the particle si/e distribution
existing in the atmosphere were known.
A further complication arises in the determination ot airborne radioactivity
because, in order to evaluate what fraction of the radioactive particulars have
been sampled, a knowledge of the size distribution of radioactive particulates in the
atmosphere should be known. This difficulty has great significance if an attempt is
4-66
-------
to be made to relate the analysis to health significance, as the radioactive particle
size versus lung retention should be evaluated. In practice many simplifying
assumptions are made. One method is to assume a minimum collection efficiency
for all particle sizes, and further assume that the radioactive particles are all of this
minimum collectable size. It is admitted that these are greatly simplified assump-
tions, but from the standpoint of health significance, they are also on the safe side.
There is one other type of filter collection efficiency that is reported in the
literature. This is a mass collection efficiency based entirely on the percentage of
the mass of the airborne particulates that are collected. This can be grossly
misleading, as the small particles predominate in number, and yet, constitute a
minor fraction of the total paniculate mass. The only time a mass collection
efficiency would represent a total particle collection efficiency would be if all the
particles were of the same size.
Theory of Collection Efficiency
Before continuing the discussion of collection efficiency, it is necessary to discuss
the theory of collection efficiency for the various collection mechanisms previously
mentioned, and also, the effect of various parameters on collection. Figure 4-48
shows a qualitative picture of the theoretical percent penetration versus linear
velocity that might be expected for a given filter and an aerosol with a specific size
particle. By referring to this graph, and the associated notes, the interrelation of
the various mechanisms can be examined.
Interception
As radius increases curve moves
As fiber diameter increases curve moves A
As inter fiber distance increases curve moves
II direct interception
As radius increases curve moves -«
As density increases curve moves
A As fiber diameter increases
'As radius increases
/ curve moves •<
As fiber diameter increases
/ curve moves
As inter fiber distance increases
curve moves
Linear velocity (cm/sec) 1
Figure 4-48. Filtration mechanisms.
4-67
-------
/> iff us to n Efficiency
I'.ie d.llus.on line (I) of Figure 4-48 shows how the dillus.on mechanism is affected
m Inuai velocity. As the velocity increases, the diffusion mechanism decreases. As
pi caously mentioned, this is due to the shortei iransii time ihiough the filter ai
;hr lughei flow tales. Ihe diffusion function always passes ihiough the origin but
the slope vanes with the particle radius, the filler libei diamete., and the distance
beiwecn liliei hbeis.
\s i,ie panicle radius increases, the slope becomes gicaiei. (bus icsultmg in a
•educed dillusion effect lor a given linear velocity This is because larger panicles
a,e no as readily affected by diffusion mechanisms as smallei panicles.
The elfects ol fiber diametei and imerfibcr distance are quite similar. As ciihei
,,l ihese iwo parameters inc.ease, the influence of the dilfus.onal mechanism
mcieases. and the slope of the line decreases. A grealc. libei d.amete. mc.eases i he
(hifusi.m effect, because to. a given linear velocity the panicle spends a longe, i.me
passing bv a filter fiber. Hie effect of interfibei distance is similar as it too alleeis
;he n.msit time oi the panicle by a filter fiber. 1 he la.ge, ihe mterfibe, distance
i lie greate. is the open space in the filler; therefore, lor a given ove.all line-it
velocity, ihe velocity in the open space of the filter is less for larger mte.fibei
distance, and transit time is longer.
Direct Interception Efficiency
As pieviouslv mentioned, direct interception is analgous 10 simple mechanical
Miaimng. As shown in line 11 of Figure 4-18. there would 1101 In- any elfeci ol
changing linear velocity on this collection mechanism.
The effects ol the particle radius and interfibei distance on the mte.ception
mechanism aie quite simple. As the particle radius me .eases, the permit penena
uon decreases as would be expected tor straining. As m.e.libei distance' increases,
,hc pe.com penetration increases tor a given s./.e, again, as would be exported Horn
mechanical straining.
The relation between panicle penetration and I.be, diameter tor mte.ception is
somewhat more' involved. As the fiber diamete. increases in si/e, the an ilowl.nes
aio affected at an increasing distance upstream from ihe liber. This may cause
some ol the flowhnes to dive.ge from a path thai would have Inoughi the pa.i.cle
w.lliin contact distance of .ho fiber. Therefore, as the I,be, diameter increases, the
colled ion efficiency due to interception decreases.
Inertial Effect Efficiency
The inortial collection of particles depends on the panicles leaving their an
flowlines and contacting a filter fiber due to their inertial Jones. For a particle ol a
j.iven size this effect would start to show at some velocity, v... and would reach a
maximum at some greater velocity, v.,,. The fact that the effect increases with
velocity is due to the dependence of inertial forces on both mass and velocity.
Although Figure 4-18. line HI shows a /.ero percent peneiration tor this mechanism
•H v the penetration might be significant, depending on the imerfiber distance. II
a larger imerfiber distance existed, the maximum effect might occu, at some pom.
v},,; in which case, a definite amount of penetration would occur.
4-68
-------
The effect of increasing particle radius in inertia! collection is improved collec-
tion for a given linear velocity. This is related to the fact that the larger the par
tide, the greater is the probability of it coming within a distance from the filter
fiber where it will make contact. Figure 4-57 lists the collection efficiency of some
filter media as a function of particle size.
The effect of increasing particle density also improves inertia! collection. Like
the effect of velocity, this is due to the increasing of menial forces, thus causing
greater deviation from the air flowlines.
As the filter fiber diameter increases, the effect of inertial collection decreases.
With the air flowlines being affected at a greater distance upstream from the fiber,
the change in their trajectory is much more gradual and, therefore, the tendency
of inertial forces to cause a particle to leave a flowline is less.
Overall Efficiency
In filter sampling all of the collection mechanisms are taking place simultaneously
and their effects are algebraically additive. The overall relation between percent
penetration and linear velocity might well look like the dotted line in Figure 4-48.
From this overall efficiency curve, it is seen that there appears to be some
velocity, vp, where a maximum penetration occurs and this has been confirmed in
several investigations (4, 9, 12, 14, 15, 18). It should be pointed out, however, that
if the direct interception effect is dominant enough, the overall efficiency curve
may well have a flat plateau, rather than a single velocity of maximum
penetration.
The mechanism of electrical forces has not been covered in this discussion of the
theory of collection efficiency. The degree to which it will change the overall
efficiency is dependent on many factors. The theoretical aspect of this effect is
beyond the scope of this discussion, but the fact that it may well play an important
part in certain instances is not to be ignored (2, 5, 8).
Experimental Collection Efficiencies
Many investigators have studied the actual collection efficiencies of filter media and
their results are reported in the literature (2, 3, 5, 9, 12, 13, 14, 18, 22). A degree
of caution must be exercised in utilizing this experimental data. A number of dif-
ferent types of particulates have been used, including dioctyl phthalate (DOP)
smoke, atmospheric dust, duraluminum dust, radon daughter products, polystyrene
aerosols, and lead fumes. The experimental results of the various investigators often
appear to be in great disagreement for many of the filter media, and these dif-
ferences should be evaluated in selecting an efficiency value to use.
One additional complication factor should be mentioned in ielation to filter
efficiency. During the time that the sampler is running, the increasing amount of
particulates that accumulate on the filter will cause the collection efficiency to
improve as sampling continues. At the same time, however, the resistance of the
filter to airflow also increases, thus perhaps interfering with the sampling pro-
cedure and decreasing the flowrate.
4-69
-------
Characteristics of Filter Media
A great number of individual types of filter papers are available. In general, they
fall into four main categories: cellulose fiber, glass fiber, mixed fiber and mem-
brane filters. These various categories will be discussed in terms of then general
characteristics, sampling considerations, and analysis consideiations.
Cellulose Fiber Filters
The liher papers in this category are typically called chemical filters. They were
designed for use in "wet" chemistry where liquid-solid separations were desired.
Although not designed with air sampling in mind, a number of these filters have
seen extensive application in this area (see Figure 4-49).
Filter
Whatman 1
4
in
41
4Z
44
541
S&S 604
MSA Type S
Cellulose
Corrugated
Cellulose
MSA BM-2133
IPC 1478
Gelman W-41
Void
size
microns
2 +
4 +
1 -
2
4 +
i
4 +
-
-
24
••^ ^^— •
Fiber
diameter
microns
-
-
-
-
-
Av.17
Thickness
microns
130
180
150
150
180
180
150
100
130
200
100.0
1000
1000
1830
560-760
Weight
per unit
area
mg/cm2
8.7
9 2
10.0
9.5
9.1
10.0
8.0
10.0
8.2
32.7
14.6
Ash
content
0.06
0.06
0.025
0.01
0.01
0.01
0.01
0.025
0.008
0.03
1.3
1.3
1.3
0.12
0.04
Maximum
operating
temp. °C
150
150
150
150
150
150
150
150
150
80
120
120
120
120
Tensile
strength
.67 kg/cm
.41 kg/cm
2.24 kg/cm
0.58 kg/cm
0.18 kg/cm
Flow
resistance
00 ft/min
inH2O
— — — — ^— —
40.5
11.5
38 (Ifm)
54
8.1
46 (28 Ifm)
40 (28 Ifm)
49 (28 Ifm)
~
8.5
6.5
9.1
0.31
J — J
Figure 4-49. Cellulose fiber filters.
Sampling Considerations
Cellulose fiber filters are made of purified cellulose pulp, thus rendering them
ineffective at high operating temperatures. The low ash content of these filters
make them highly suitable to analysis where heat or chemical ashing is a required
preparation. These filters generally have a high affinity for moisture 1 his limita-
tion means that the relative humidity must be controlled while weighing the filters
to ensure there is no error introduced by absorbed moisture. Recent stud.es show
that cellulose filters irreversibly absorb water, thus a control blank is required when
humidity is a problem. Cellulose filters also enhance the artifact formation ot
sulfate and nitrate.
4-70
-------
Specific Filters
Whatman 11 filter paper is the most widely used of the cellulose fiber filters.
Whatman 11 has become the preferred substitute to glass fiber filters for high
volume sampling because of the good collection efficiency and acceptability for fur-
ther chemical analyses. It has also received wide use in all metals analysis including
neutron activation analysis because of the low blanks (see Figure 4-58). This filter
has also found applications as the tape used with paper tape samplers lor the deter-
mination of soiling index.
The MSA type "S" filter is well adapted to high-volume sampling because of ihe
low piessure drop. This filter is difficult to handle because of iis bulkiness and its
variable ash content and organic binder make it unsuitable loi some chemical
analyses.
TFA filteis have also been used for high-volume sampling when extensive
chemical analysis is required. Whatman 41 is preferred ovci this liltei because of
lire highei collection efficiency (see Figure 4-57).
Glass Fiber Filters
Ihese tiltei papeis are made from finely-spun glass fibei b\ combining the- liber
with an organic binder and compressing this material in a paper machine. These
filters have seen increasing use in air sampling.
Sa i up ling Co nsidera t io ns
These tihers have (he ability to withstand high temperatures (up to 5i()°C), thus
making them most attractive for stack sampling. They are funhet typified by high
coiled ion efficiency. In some cases, the organic binder may interfere with subse-
quent analysi-,. so the fillet is flash fired to remove ilie bmdci maietial. This causes
some loss in tensile strength and usually requires that a backing material be used
duimg sampl'tig 1'he glass tillers are nonhydroscopic. thus allowing them to be
used in areas where humidity is high. Being glass also makes them the filler choice
for most con >sive atmospheres. All of the filters in this caiei>oi\ ate quite fragile
aird must be handled with rare.
A nalysis Considerations
Glass fiber tihers. because ->t the high silicate content, are extremely difficult to ash
by chemicals or heat. Therefore, extraction procedure's are performed on these
filters to remove the sample for subsequent chemical analysis. Foi this reason flash
fired glass fillers are major atmospheric sampling fillets.
The pll of the tiller will effect ihe collection of the sample It has been recom-
mended (13) lhai neutral pH (6.5
-------
Specific Filter
1'hr MSA 1106 BH is one of the flash-fired glass fiber filters, and theiefore, free
of 01 game binder. Special mention should be made of the Gehnan Spectro Grade
l'\pe A Glass Fiber Filter. This neutral pH glass fiber filtei has been idled as ulna
puie lor metals analysis to minimi/.e interference with irate metal background
values. (Kadi box of 8" X 10" filters contains assay informal ion on trace metals loi
ihe en. losed I IUMS.) Spectro grade filteis do not absorb detectable am milts of SO.,
horn the atmosphere, thereby minimizing its effect. A suninuuv of the physical
characteristics of glass fiber filters is given in Figure 4-.r>()
Filter
MSA 1106B*
1106BH#
Gelman A#
E*
G
M
H
Whatman
AGF.'Atf
AGF/ B#
AGF/ D
AGF F
Il&V H-93
H-94
S&S 24 '
IT'
Void
size
microns
> j
-> j
>!
>1
Fiber
diameter
microns
0.6
0.5-3
Thickness
microns
180-270
180-460
380
380
810
580
510
340
840
460
380
460-560
380
200
125-180
127
Weight
per unit
area
mg/cm2
6.1
5.8
9.3
10 0
11.6
10.8
12.7
5.3
15.0
5.5
6.3
9.3
8.2
6.05
5.4
Ash
content
- 95
~ 100
99.4
98 1
100
100
100
100
96-99
96-99
98
98
Maximum
operating
temp. °C
540
540
480
480
480
480
480
540
540
540
540
540
480
400
400
Tensile
strength
3.5 Ib/m
1.5 Ib/in
1.29 Ib/m
3.14 Ib/in
.56 Ib/in
73 Ib/in
2.5 Ib, in
2.5 Ib/in
Flow
resistance
100 ft/min
in H20
19.8
19.8
18.9
18 9
3.0
6 1
21 7
2.3@2 Ifm
Benzene
extract/ 100
in2 mg
17 3
0.6
0 6
0.8
3.2
00
5
0.6
^without organic binder
1 with organic binder
Figure 4-50. Glass fiber filter characteristics.
Mixed Fiber Filters
\'o! LOO much can be said of this category except thai they possess the
clidiacu-iistics of the individual fibers composing ihem. I'nc ( hemical analysis of
mixed libei fillers also depends on the individual fibet tonsmiu-niv General
c naiac.e.isiics of a numbet of I he members of this category an- given in
1'igui e 1 -i 1.
4-72
-------
Filter
designation
H&V H-70,
9 mil
H-70, 18 mil
H-64
H-90
H-91
N-15
5-G
MSA glass &
cellulose
Whatman
ACG/A
ACG/B
H&V CWS-6
H&V AEC-1
VM-100
Gelman
VM-1
Composition
Cellulose
asbestos
Cellulose
asbestos
Cellulose
asbestos
Cellulose glass
Cellulose glass
Synthetic fiber
& glass
Synthetic fiber
glass & cotton
Glass & cellulose
Glass & cellulose
Glass & cellulose
Cellulose
asbestos
Cellulose
asbestos
Vinyl metracel
Vinyl metracel
Void
size
microns
> 1
> 1
10.0
5 0
Fiber
diameter
microns
0.1-35
0.1-35
0.1-35
9-35
1.5-35
0.5-15
0.5-15
Thickness
microns
230
460
830-1090
685
710
1270
685
1000
330
990
762
762
Weight
per unit
area
mg/cm*
8.2
15.4
22.7
13.4
13 5
24 9
14.5
5.5
19 5
Ash
content
%
20-25
20-25
15-20
70
80
15
4-6
11%
13%
Maximum
operating
temp. °C
150
150
150
150
150
150
150
120
150
150
Tensile
strength
2.5 Ib/in
4.0 Ib/in
2.0 Ib/in
3.2 Ib/in
3.5 Ib/in
1.0 Ib/in
gauze backed
270 gm/cm
330 gm/cm
Flow
resistance
at 100 ft/min
( — 50 cm/sec)
in H2O
17
26
15
0.4
0.89
9.9
2.0
0.9 (20 Ifm)
2 6 (20 Ifm)
17
13.3
Figure 4-51. Mixed fiber filter characteristics.
Membrane Filters
This filter media consists of dry gels of cellulose esters, usually produced as
cellulose acetate, or cellulose nitrate, polyvinyl chloride, acryloiiiti-lie, and
Teflon"; . The filters are cast on a smooth flat substrate and exposed to a con
trolled atmosphere. The process can control both the internal membrane structure
and pore size (8). Some filters are formed with pores while others aie lormed as
sheets with pores formed later.
Sampling Considerations
Membrane filters are typically very brittle and require careful handling. In air
sampling they should be backed by some support structure to avoid breakage. The
filters ire not too well suited to stack sampling, as they have an operating
temperature range comparable to cellulose fiber filters (see Figure 4-52).
The particle sizes collected by membrane filters have been found to be much
smaller than the pore size; this is thought to be due to electrostatic forces.
A comparison can be made between the filters supplied by two major manufac-
turers, Millipore and Gelman. The Millipore filters have a different texture on the
two sides created in the manufacturing process. Lindeken ei. al. (7) report that
Millipore filters with nominal pore diameters less than one micron have smaller
openings on ihe top side than on the bottom. For Millipores with nominal pore
sizes above one micron, this situation is reversed. It is the top side pore opening
that is reported as the nominal size, and Millipore filters are packed with this side
4-73
-------
up. The one exception to this is Millipore SM, which is packed with its nominal
pore side down. The Gelman filters have approximately the same pore size on both
sides. This consideration becomes important when sampling tor alpha radioactivity
as the small side of the Millipore would be the preferred upstream side in order to
minimize the depth that particles go into the filter. Because of these differences in
the two makes of filters, the Gelman filters have a smaller pressure drop at
equivalent pore diameters. Both filters have appreciable pressure drops, however,
and this may limit the volume of air that can be sampled. Another factor limiting
Hie volume of air that can he sampled is the build-up of a second layer of dust on
the filter surface. This layer has a tendency to slough off. causing loss of part oi
i lie sample.
Filter
Millipore SM
SS
ws
RA
AA
DA
HA
WH
Gelman AM-1
AM-3
AM-4
AM-5
AM-6
S&S AF-600
AF-400
AF-250
AF-150
AF-100
AF-50
AF-30
Gelman GM-1
GM-3
GM-4
GM-6
GM 8
GM-9
GM-10
Pore
size
microns
5 0
3.0
3 0
1.2
0 80
0.67
0.45
0.45
5 0
2.0
0.65
0 65
0 40
7.5
4.0
2 0
0.85
0.70
0.60
0.40
5.0
1.2
0 80
0.45
0.20
0 10
0 05
Index
refraction
1.495
1.495
1.512
1 510
1.510
1 510
1.49
1.49
1.49
1.49
1.49
1.49
1 49
Thickness
microns
170
170
150
150
150
150
150
150
200
200
200
200
200
180-250
180-250
160-210
160- 210
Av 150
Av-135
Av-120
Weight
per unit
area
rag/cm8
3.6
3.8
4 9
4.2
4 7
4.8
4 9
5.7
3.6
6.6
5 8
6.3 8.7
63 87
5673
5.6 -7.3
Av 5.3
Av 4.7
Av 4 2
Ash
content
%
0.0001
0.0001
0.0001
0.01
0.01
0.008
0.008
0 007
0.006
0.005
Maximum
operating
temp. °C
125
125
••
••
Continuous 100
Peak 200
Dry 176
1 7fi
1 /D
176
176
176
176
176
Tensile
strength
100 psi
150 psi
300 psi
350 psi
400 psi
450 psi
.41 kg/cm2
.72 kg/ cm2
Flow resistance
100 ft/min
in H2O
19
38
-~ 100
62
91 (39)*
120
210
— 270
11
33
T\
I j
'Alternate value reported
Table 4-52. Membrane filter media.
Another filter that should be discussed is the Nucleopore membrane filter. These
filters have been reported to have high efficiencies, but even filters with 0.8 /xg
holes are not very efficient (10). The Nucleopore membranes also have the disad-
vantage of an extremely high pressure drop across the filter. The pressure drop is
almost double that of the Millipore membrane filter. The codes used by filter
manufacturers are shown in Figure 4-53.
4-74
-------
Code
I PC
s&s
Whatman
H&V
MSA
Gelman
Millipore
C.W.S.
AEC
Manufacturer
Institute of Paper Chemistry (U.S.A.)
Schleicher and Schuell (Germany)
W. and R. Balston Ltd. (England)
Hollingsworth and Vose (U.S.A.)
Mine Safety Appliances Co. (U.S.A.)
Gelman Instrument Co. (U.S.A.)
Millipore Filter Corp. (U.S.A.)
Chemical Warfare Service (Chemical Corps, U.S.A.
Atomic Energy Commission (U.S.A.)
Figure 4-53. Filter manufacturers code.
Analysis Considerations
Figure \ .") t lisls ihe effects of various chemicals on membrane tilieis. This listing
should be quiie helpful in setting up an analysis puxediue.
Solvent
Hydrocarbons
Pentane
Hexane
Petroleum ether
Benzene
Toluene
Xylene
Halogenated hydrocarbons
Methylene chloride
Ethylene chloride
Carbon tetrachloride
Trichloroethylene
Freon TF
Dichloroethane
Alcohols
Methanol
Ethanol
Propanol
Isopropyl alcohol
Butyl alcohol
Glycerol
Ether alcohols
Carbitol
Methyl cellusolve
Butyl cellosolve
Ketones
Acetone
Methyl ethyl ketone
Methyl isobutyl ketone
Cycohexanone
Diacetone
Effect on filter
None
None
None
None
None
None
None
Shrinks
None
None
None
None
Dissolves
Swells
Distorts
None
None
None
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
Figure 4-54. Membrane filter solubility characteristics. (Continued next page.)
4-75
-------
Solvent
Effect on filter
Esters
Methyl acetate
Ethyl acetate
Propyl acetate
Butyl acetate
Miscellaneous organic
Pyridine
Dimethyl formamide
Unsymmetrical dimethyl hydrazine
Nitro benzene
Ethylene glycol
Varsol1
Cobehn2
Mineral spirits
Turpentine
JP 4
JP-5
Kerosene
Acids
Glacial acetic acid
10% acetic acid
6N Hydrochloric acid
6N Sulfuric acid
6N Nitric acid
Alkalis
6N Sodium hydroxide
6N Ammonium hydroxide
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
Dissolves
None
None
None
None
None
None
None
None
DissoUes
None
None
None
None
Disintegrates
None
'Reg. T. M. of Esso Standard
2Reg. T. M. of Cobehn Corporation
i 1-54. Membrane filter solubility characteristics, continued.
(g/m!)
Item
Dates
7-26
7-27
7-28
7-29
7-30
7-31
8-2
8-3
8-4
Mean
Total particulates
pH 6.5 pH 11
68.9 79.3
49.8 58.3
61.4 78.9
66.7 104.1
104.5 113.7
83.1 94.4
44.5 50.9
43.3 49.6
79.8 94.5
66.89 80.41
Sulfates
pH 6.5 pH 11
6.0 11.1
4.5 7.9
4.4 8.3
5.0 12.2
6.8 9 1
6.7 10.1
4.6 6.5
4.6 7.2
5.8 10.9
5.38 9.26
Nitrates
pH 6.5 pH 11
0.8 3.4
1.3 2.0
0.4 1.3
1.9 4.9
1.9 3.5
1.6 3.6
0.7 2.2
0.3 1.1
1.0 3.6
1.10 2.84
Organics
pH 6.5 pH 11
8.9 5.9
5.2 3.4
3.8 3.1
6.2 6.1
6.9 4.9
5.7 5.2
3.3 3.3
3.7 2.7
5.5 4.4
5.47 4.33
Alkaline filters yielded significantly higher results for total particles, nitrates and sulfates, but not
organics. The difference is ascribable to the adsorption of acid gases.
Figure 4-55. Effect of glass fiber pH on concentration observed with 24-hr, standard high
volume sampling in Anderson, Calif., July-Aug., 1972.
4-76
-------
Filter
None
None
Size, mmh
90 (holder
only)
90 + 47
(holder
only)
No. of
trials
1
1
Flow
rate
cfma
4.3 (max)
3.0
2.5
4.2 (max)
3.0
2.5
Overall
pressure
drop, cm. Hg
25.0
10.3
8.0
26.6
9.0
6.9
Efficiency
by CNC, %b
—
—
—
—
—
—
Membrane filters
0.8 fi Nucleopore
5 ii Nucleopore
3.0 n MF millipore (SS)
3.0 n MF millipore (SS)
5.0 n MF millipore (SM)
5.0 n MF millipore (SM)
8.0 n MF millipore (SC)
Gelman GA-1.5 /* cell.
acetate
1.2 n Silver membrane8
S&S 0.45 n (B6A)
cellulose acetate
47
47
47
90
47
90
47
47
47
47
2
1
1
2
1
1
2
2
2
1
1.5 (max)
3.0
3.4 (max)
3.0
2.5
4. 1 ± 2(max)
3.0
2.5
3.7 (max)
3.0
2.5
4.4 (max)
3.0
2.5
4.0 (max)
3.0
2.5
3. 8 ±0.1
(max)
3.0
2.5
3.6±0.2
3.0
2.5
1.5 (max)
1.0
50.8
11.5
32.2
25.6
19.9
26.3 + 2.1
14.2 + 2.4
10.8+1.8
31.7
22.6
17.6
23.2
11.4
8.6
27.1
16.5
13.1
30 2 + 0.7
20.2 + 0.2
15.6 + 0.2
29.3 + 0.5
20.8 + 2.7
16.4+1.4
53.1
41
72 ± r
<50
f
98
96 + 2
98
96
98
98
97+ 1
97 + 1
98
97 + 2
66
ND
ND
Teflon fiber fill ers
5.0 n Teflon millipore (LS)
5 fi Teflon millipore (LS)
10 n Teflon millipore (LC)
90
47
90
2
1
1
3.2±0.1
(max)
3.0
2.5
2.0
1.7 (max)
4.7 (max)
3.0
2.5
36.6 + 0.6
30
24.8+ 1.6
16.0
55
26
10.8
8.6
93
-
73
Figure 4-56. Initial filtration efficiency and flow behavior of the Lundgren impactor
as a function of the after filter media**'6, continued.
4-77
-------
Filter
Size, mmn
No. of
trials
Flow
rate
c£ma
Overall
pressure
drop, cm. Hg
Efficiency
by CNC, %b
Glass fiber filters
Gelman A
Reeves Angel #900 AF
MSA 1106 BH
S&S #25 acid washed
S&S green ribbon #599
TFA
Whatman 41
47
47
47
47
47
47
47
2
1
1
2
1
2
1
3.9 + 0.3
(max)
3 0
2 5
4.1 (max)
3 0
2 5
4.2 (max)
3.0
2 5
3.6±0.2
(max)
3.0
2 5
3.2 (max)
3.0
2.5
3.7±0.3
(max)
3 0
2.5
3.7 (max)
3.0
2 5
27.9±1.1
17.6±2.5
13. 1± 1.5
23.2
11.4
8.6
26.4
14.8
11.2
29. 9 ±0.7
21.6 + 2.6
16.0 + 2.0
32.9
28.1
20.5
29.6+1.0
20.1 + 2.7
14.8+1.9
31.0
21.2
16.0
95
98
98
98,
99
98
59
50+ 12
60
66
52+14
70
81
a. Determined with built-in flowmeter of Lundgren Impactor.
b. Efficiency of particle removal as measured by a GE portable condensation nuclei counter. The condensation
nuclei counter (CNC) works on much the same principle as a cloud chamber. The particle laden gas is
entered into a chamber. Inside the chamber is a light source. The inside atmosphere is saturated with water
vapor. When the gas is suddenly expanded, the particles act as condensation nuclei, having water droplets
form around them. The droplets are counted by a light scattering technique.
c. Measured at 1.2 1/min with the filter holder attached directly to CNC
d. Face velocity at 3.0 cfm =89 cm/sec for 47 rnm filter and s30 cm/sec for 90 mm filter.
e The impactor stages were neither covered by film nor rotated.
f. Results erratic at this high Ap
g Selas Flotronics
h. The effective size for the filters was 45 mm and 78 mm for 47 and 90 mm filters, respectively
Figure 4-56. Initial filtration efficiency and flow behavior of the Lundgren impactor
as a function of the after filter media*''6.
4-78
------- An error occurred while trying to OCR this image.
-------
One advantage of membrane filters is that they are primarily surface collectors,
consequently, the problem of self-absorption of radiation becomes negligible,
although the difference in the two sides of Millipore filters must be taken into con-
sideration in alpha spectroscopy.
The membrane filters lend themselves very reathlv to particle si/.e analysis by
microscopy. By using an immersion oil with an index of refraction (N,,) equal to
that of the filter (see Figure 4-52), the filter can be made transparent to light, thus
allowing light-transmitted microscopic analysis. Houevei, care mu.st be taken to
ensuie that the N,, of the particle is not the same a-, the N,, of the immersion oil.
The filters can also be readily ashed and leave very little residue. This can be a
definite advantage in some analysis schemes. Most membrane filters are also readily
soluble in many organic solvents, thus allowing removal of paniculate mattei with
little pioblem,
Sample Problems
Problem 1
A filtei it. to be chosen for subsequent analysis by a chemical ashing technique.
The maximum vacuum flow resistance the pump can overcome is about 10 in. of
water when the lace velocity is about 100 ft. mm. When dealing with atmospheric
sampling, the temperature is not a limiting facioi since most filters will operate at
tempeiatu.es over 100°C (212°F). Which filter, based on the available information
and assuming all were at hand, would be the best choice.'
Solution .
From Figures 4 50 and 4-51 it can be seen that glass fiber and mixed fiber fillers
have relatively high ash contents, making them impractical for chemical ashing.
From Figures 4-49 and 4-52, it can be seen that cellulose fiber fillers and mem-
brane filters exceed the requirements of the pumps. From Figure 4-49, it can be
seen ihai Whatman 41, S&S 604, MSA BM-2133. and the IPC 1478 all have ash
contents less than 1% and flow resistances less than 10 in H2O. 1 hese would all be
acceptable and availability would determine which would be used.
Problem 2
Fhe pump capacity from Problem 1 has now been doubled so that 20 in.
H20@100 ft./min. can now be sampled. An efficiency of 100% for all particles
down to 0.3 IJL is desired. Which filter should be used?
Solution
From Figure 4-57 the 5.0 /( MF Millipore (SM) 47 mm & 90 mm, S&S B6A
cellulose acetate, Gelman A 25 mm and MSA 1106 BH all show apparent particle
efficiencies of 100% for all sizes. From Figure 4 50 both the Gelman and MSA
filters show an ash content over 95%. The S&S B6A does not appear on Figures
1 49 through 4-52 but since the cellulose acetate is a mixed fiber filter we can
assume the ash content is too high. From Figure 4-52, the 5.0 ,t Millipore SM has a
low ash eunient (0.0001%) and an acceptable flow resistance (19.8 in. H,O).
4-80
-------
Problem 3
A filter is needed for subsequent analysis by acid extraction. The sampling
atmosphere is very humid. The flow resistance must be kept below 10 in. H2O.
Which filter should be used?
Solution
Cellulose filters are very hygroscopic so they should not be used in this situation.
Membrane filters have extremely high flow resistances, so they, too, should not be
used in this situation. Glass fiber filters are non-hydroscopic and are the filters of
choice in water vapor laden atmospheres. By referring to Figure 4-50, Gelman G or
M filters could be used since they both qualify for the low flow resistance
requirements.
Summary
No single type of filter is the right one for all air sampling problems. In evaluating
a filter one must consider many factors: general filter characteristics, collection
efficiency, background filter impurities, sampling conditions, ease of analysis, self-
absorption, flow resistance, and the purpose for which the sample is being taken.
References
1. Appel, B. R., and Wesolowski, J. J. Selection of Filter Media for Particulate
Sampling with a Lundgren Impactor. State of California Air and Industrial
Hygiene Laboratory Report 125.
2. Chambers, L. Filter Media for Air Sampling. Industrial Hygiene Quarterly 15:
290, 1954.
3. Dams, R.; Rahn, K. A.; and Winchester, J. W. Evaluation of Filter Materials
and Impaction Surfaces for Nondestructive Neutron Activation Analysis of
Aerosols. Environ. Sci. Tech. 6: 441, 1972.
4. Fitzgerald, J., and Detwiler, C. Collection Efficiency of Air-Cleaning and Air-
Sampling Filter Media. AEC Report KAPL-1088.
5. Gillespie, J. The Role of Electric Forces in the Filtration of Aerosols by Fiber
Filters. Journal of Colloid Science 10: 299, 1955.
6. Latorre, P., and Silverman, L. Collection Efficiencies of Filter Papers for
Sampling Lead Fume. AMA Arch, of Industrial Health 11: 243, 1955.
7. Lindeken, C., et al. Surface Collection Efficiency of Large-Pore Membrane
Filters. Health Physics 10: 495, 1964.
8. Lippman, M. Filter Media and Filter Holders for Air Sampling. Air Sampling
Instruments for Evaluation of Atmospheric Contaminants. American Con-
ference of Governmental Industrial Hygienists, Cincinnati, Ohio, 1972.
9. Lockhart, L., et al. Characteristics of Air Filter Media Used for Monitoring
Airborne Radioactivity. NRL Report 6054.
4-81
-------
10. Mueller, P. K.; Twiss, S.; and Sanders, G. Selection of Filter Media: An
annotated Outline. Presentation at the 13th Conference on Methods in Air
Pollution and Industrial Hygiene Studies, Berkeley, October 30-31, 1972.
11. Pate, J., and Tabor, E. Analytical Aspects of the Use of Glass Fiber Filters for
the Collection and Analysis of Atmospheric Particulate Matter. American
Industrial Hygiene Association Journal 2$: 145, 1962.
12. Ramskill, E. and Anderson, W. The Inertial Mechanism in the Mechanical
Filtration of Aerosols. Journal of Colloid Science 6: 416, 1951.
13. Smith, F., and Nelson, A. C. Guidelines for Development of A Quality
Assurance Program—Reference Method for the Determination of Suspended
Particulates in the Atmosphere (High Volume Method). Prepared for Research
Triangle Institute under Contract 68-02-0598, June, 1973.
14. Smith, J.; and Surprenant, N. Properties of Various Filtering Media for
Atmospheric Dust Sampling. Proceedings of the American Society of Testing
Materials 53: 1122.
15. Stern, S., et al. The Aerosol Efficiency and Pressure Drop of a Fibrous Filter
at Reduced Pressures. Journal of Colloid Science 15: 546, I960.
16. Dust Topics, 1, No. 1., Gelman Instrument Company, spring 1964.
17. Parker, R. D., et al. A Two Stage Respirable Aerosol Sampler Using
Nuclepore Filters in Series. Atmospheric Environment 11: 617, 1977.
18. Liu, Benjamin Y. H., and Kuhlmey, G. A. Efficiency of Air Sampling Filter
Media. Presented at the Symposium and Workshop on X-ray Fluorescence
Analysis of Environmental Samples, Chapel Hill, N.C., January 26-28, 1978.
Particle Technology Laboratory Publication No. 293, University of Minnesota,
Minneapolis, Minnesota.
19. Kirsh, A. A.; Stechkina, I. B.; and Fuchs, N. A. Efficiency of Aerosol Filters
Made of Ultrafine Polydisperse Fibres./ Aerosol Science 6: 119, 1975.
20. Pupp, Christian; Lao, R. C.; Murray, J. J.; and Pottie, R. F. Equilibrium
Vapor Concentrations of Some Polycyclic Aromatic Hydrocarbons, As 4O6, and
SeO2 and the Collection Efficiencies of These Air Pollutants. Atmos. Environ.
8: 915, 1974.
21. Caroff, M.; Choudhary, K. R.; and Gentry, J. W. Effect of Pore and Particle
Size Distribution on Efficiencies of Membrane Filters./. Aerosol Science,
4: 93, 1973.
22. Hounam, R. F. The Filtering Efficiency of Selected Papers. Ann. Occup. Hyg.
4: 301, 1962.
4-82
-------
Chapter 5
Gaseous Sampling
Principles of Absorption
Introduction
Absorption of pollutants in various media plays an important role in air pollution
monitoring. It is particularly important in the wet-chemical methods of analysis.
Before the advent of continuous monitoring instrumentation, techniques employing
absorption were the most inexpensive and up-to-date methods available.
Absorption is the process "of transferring one or more gaseous components into a
liquid or solid medium in which they dissolve." Absorption of gaseous pollutants in
solution is frequently utilized in atmospheric sampling because of the numerous
methods available to analyze the resulting solution. These methods include
photometric, conductimetric, and titrimetric techniques. Details of sampling and
analysis of specific gaseous pollutants by absorption are given elsewhere. This
discussion concentrates on a description of the gas-liquid absorption process and
factors affecting collection efficiency. Devices frequently utilized in gas-liquid
absorption and several current applications are also discussed.
Types of Absorption
In gas-liquid absorption the collecting liquid (i.e., the absorbent) may change
either chemically or physically, or both, during the absorption process. In
gas-liquid absorption sampling, two types of absorption have been recognized:
(a) physical absorption and (b) chemical absorption.
A typical chemical absorption process would involve drawing a volume of air
through a solution that reacts with the gaseous contaminant to form a nongaseous
compound. For example, an acid mist is drawn through a volume of sodium
hydroxide. The acid reacts with the base to form a stable salt. Titration of the
unreacted base with standard acid indicates the quantity of pollutant reacted.
Physical Absorption (2, 3)
Physical absorption involves the physical dissolving of the pollutant in a liquid. The
process is usually reversible in that the pollutant exhibits a relatively appreciable
vapor pressure. The solubility of the pollutant in a given absorbent is dependent on
the partial pressure of the pollutant in the atmosphere and the temperature and
purity of the absorbent. An ideal solvent would be relatively nonvolatile, inexpen-
sive, noncorrosive, stable, nonviscous, nonflammable, and nontoxic. In many cases
5-1
-------
distilled water fulfills many of these characteristics and is used as the solvent for
collecting some gases. The suitability of distilled water for several selected gases is
presented in Figure 5-1.
Gas
Nitrogen
Oxygen
Nitric oxide
Carbon dioxide
Hydrogen sulfide
Sulfur dioxide
Volume absorbed
per volume of water*
0.015
0.031
0.047
0.878
2.582
39.374
*Gas volumes reduced to 0°C and 760 mm Hg
Figure 5-1. Solubility of selected gases in distilled water at 20 °C.
The physical absorption process involves collecting the pollutant by solution in
the absorbent. The solution is then analyzed for pollutant concentration by a con-
venient analytical method. In general, low efficiency will be obtained for physical
absorption unless the pollutant is very soluble and the ratio of dissolved gas to
liquid volume is small. For this reason, physical absorption is rarely the only
absorption process involved in collecting gaseous pollutants.
Chemical Absorption (2, 3)
In contrast to physical absorption, chemical absorption is a process that involves a
liquid absorbent that reacts with the pollutant to yield a nonvolatile product. The
solvent selected is one that reacts with the pollutant in an irreversible fashion; for
example, the reactions of ammonia and carbon dioxide gases with acidic and basic
solvents respectively. These reactions produce carbonic acid (H2CO3) and
ammonium hydroxide (NH4OH). The solubilities of these acids and bases are much
greater than gaseous CO2 or NH3. Primary factors affecting the choice of an absor-
bent in chemical absorption are the solubility of the pollutant, reactive properties
of pollutant and absorbent, and the subsequent analytical method to be used. Care
should be taken to avoid an absorbent that will interfere with subsequent chemical
analysis.
A typical process involving chemical absorption is the reaction of SO2 and
aqueous H2O2 to produce sulfuric acid. The concentration of SO2 is determined by
titrating the H2SO4 formed with Ba (C1O4)2. This procedure is currently the
reference method for determining SO2 emissions from stationary sources.
Collection Efficiency (2, 4)
Each absorption sampling device must be assembled from units found to be most
suitable for the specific pollutant involved. It is not necessary to have 100 percent
collection efficiency; however, the efficiency under sampling conditions should be
known and reproducible. In some circumstances a sampling system having a
relatively low collection efficiency (e.g., 60-70 percent) could be used provided that
the desired sensitivity, reproducibility, and accuracy are obtainable.
5-2
-------
There is much information available in the literature concerning optimum flow
rates for specific pollutants and collection efficiencies with respect to the pollutant
and absorbent for many sampling devices. However, much more information is
needed on the variation of collection efficiency with the rate of sampling, concen-
trations of a variety of compounds, and the nature of the collecting medium. For
available information on gas-liquid absorption theory and the mathematical treat-
ment of the variables affecting collection efficiency, the reader is referred to the
literature (Refs. 1-10). In the present discussion only a qualitative description of
the factors affecting collection efficiency has been attempted.
Factors Affecting Collection Efficiency (2, 3, 4)
The variables affecting the collection efficiency of methods that use absorbers for
the collection of gaseous contaminants may be conveniently considered as: (a) those
associated with the absorber such as an acceptable flow rate, bubble size, and
height of the liquid column: (b) the chemical characteristics of the sampling situa-
tion such as the chemical nature and concentration of the pollutant in the air and
the absorbing medium, the chemical nature and concentration of the absorbing
solution, and the reaction rate; and (c) the physical characteristics of the sampling
situation such as temperature, pressure, and pollutant solubility.
Absorber Characteristics
The gas flow rate through the absorber is one of the major factors determining the
collection efficiency of an absorber. Absorption collection efficiency varies inversely
with the flow rate. An increase in the flow rate through the solution will decrease
the probability of adequate gas-liquid contact. In addition, high flow rates increase
the possibility of liquid entrainment in the effluent gas. If varying flow rates are
used in sampling, a collection efficiency versus flow rate curve should be deter-
mined for each absorber and absorber type. All other variables (e.g., temperature,
pollutant and absorbent types, etc.) should be held at the desired values.
The collection efficiency of the absorption process for a gas or vapor by chemical
absorption or physical absorption depends on the probability of successful collisions
of reagent or solvent molecules with gas molecules. For a given concentration of
reagent this probability of collisions will depend on the surface area of the gas
bubbles, on the length of the column of liquid through which the bubbles must
pass, and on the rate at which they rise through the liquid. As the volume of
individual bubbles decreases, the surface area presented to the liquid increases.
Hence smaller bubbles have a greater possibility of gas transfer into the absorbent
phase. For this reason many absorption devices use fritted discs as opposed to
injection-type dispersion tubes to achieve a smaller bubble size. However, due to
possible surface reactions that can take place at the frit, fritted bubblers may not
be appropriate for certain types of sampling (e.g., for ozone). The length of the
column of liquid in the absorber is another prime factor affecting the collection
efficiency. The longer the gas bubble is in contact with the liquid, the more pollu-
tant transferred. However, in many cases this variable cannot be used to its
maximum advantage; for example, when the sampled pollutant has a low concen-
tration in the atmosphere it must be collected in a small absorbent volume so that
it is in the sensitivity range of the subsequent analytical method to be used. Bubble
5-3
-------
rise time is a function of bubble size and absorbent height. A compromise is
usually reached by having the smallest feasible bubble size combined with the
highest absorbent column possible for the particular analysis.
Chemical Characteristics
The best situation, with respect to collection efficiency, is to choose an absorbent
with a very large capacity for absorbing the pollutant without building up
appreciable vapor pressure. This can be accomplished by choosing a chemical
reagent that reacts with the pollutant in an irreversible fashion. For example, the
irreversible reaction that occurs when carbon dioxide is absorbed in a sodium
hydroxide solution to form the carbonate (COr) ion.
The concentration of the absorbing medium to be used is a function of the
expected concentration of the contaminant encountered, and the rate of the par-
ticular chemical reaction being used. An excess of the reactant in the absorbing
solution is preferable to ensure that all the pollutant is collected and that the reac-
tion rate is at a maximum. Ideally the reaction should be instantaneous since the
period of contact between the pollutant and the absorbent is a short one.
Since the rate of reaction is proportional to concentrations of the reacting
substances, other variables being equal, the rate of the process falls off as the reac-
tion proceeds. This phenomenon must be compensated for by increasing the con-
centration of the absorbing liquid, thereby, forcing the reaction to approach com-
pletion rapidly.
Physical Characteristics
The primary physical characteristics affecting collection efficiency are pressure,
temperature, and pollutant solubility in the absorbing medium. In many sampling
situations, these variables are fixed by ambient conditions.
The solubility of the pollutant in the absorbing medium is related to its partial
pressure (by Henry's law), and the partial pressure of the pollutant in turn is
related to its concentration. The net effect considering ideal gas behavior is that an
increase in pollutant concentration in the air will result in an increase in pollutant
solubility in the liquid. Increased pollutant solubility, other variables being equal,
results in a higher collection efficiency.
An increase in temperature enhances chemical reactions but decreases pollutant
solubility in the absorbent. In most cases the net effect is a decrease in collection
efficiency with increasing temperature.
Determination of Collection Efficiency (2, 4)
The method of determining collection efficiency will depend on how the results are
to be used. If the most accurate values are needed, the best available method for
determining collection efficiency should be used. On the other hand, if only
approximate values are needed, a less stringent method for determining collection
efficiency may be satisfactory. In all cases collection efficiency should be defined
with respect to the method of determination.
5-4
-------
The most accurate method of determining the collection efficiency of a par-
ticular absorber is by a trial on a synthetic atmosphere duplicating in every respect
the actual sampling conditions. Calibration techniques consist of both dynamic
dilution and static dilution systems. In dynamic dilution a continuous supply of a
known pollutant concentration is available that can be sampled, while the static
system consists of a container holding a known volume of pollutant of a known
concentration. In both of these calibration procedures the investigator must be
assured that the atmosphere being sampled actually contains the pollutant concen-
tration it is believed to contain.
Another method that may be used for collection efficiency calibration is the
comparison of the technique of interest to a previously calibrated method. In this
technique the conditions of the calibrated method are imposed on the method of
interest. All variables in both methods should be identical, especially with respect
to interferences.
Absorption Devices (2, 11)
A variety of devices have been used for sampling pollutants from the atmosphere.
One of the simplest and most common devices used is an ordinary gas washing
bottle containing the absorbent plus a gas dispersion tube for introduction of the
pollutant into the solution. A typical device of this type is illustrated in Figure 5-2.
Absorbing solution
Figure 5-2. Absorption device adapted from an Erlenmeyer flask.
Gas flows from the unrestricted opening into the absorbent solution. A variety of
absorbers of this type are available. They are usually glass and may be conical or
cylindrical in shape. Typical flow rates through the various devices range from 1 to
5 liters per minute.
5-5
-------
The majority of other absorption devices used in atmospheric sampling fall into
two categories: (a) fritted-glass absorbers and (b) impingers.
Fritted-Glass Absorbers
A great variety of shapes and sizes of these absorbers are being used. A few are
illustrated in Figure 5-3.
Impinger tube
+ frit
Impinger tube
+ frit
Midget
impinger tube
+ frit
Smog
bubbler
Figure 5-3. Typical fritted-glass absorbers.
These units usually provide the most efficient collection of gaseous pollutants. In
addition to the commercially available units, homemade devices may be created
using normal gas dispersion tubes. The fritted part of the dispersion tube is readily
available in the form of a disc or cylinder of various pore size. The coarse and
extra coarse frits provide good pollutant dispersion with a minimum head loss.
The collection efficiency of any one device will depend on the factors previously
mentioned. However, under optimal conditions of flow rate, absorbing medium
and pollutant type, many of the fritted-glass absorbers have a collection efficiency
in excess of 90 percent. Several of their more important characteristics are
presented in Figure 5-4.
5-6
-------
Principle of operation
Simple gas-washing
bottles. Gas flows from
unrestricted opening
into solution. Glass,
conical or cylindrical
shape
Modified gas-washing
bottles
Large bubbler traverses
path extended by
spiral glass insert.
Impingers — designed
principally for collec-
tion of aerosols. Used foi
collection of aerosols
Used for collection of
gases. Restricted
opening. Fritted tubes
available which allow
use as bubbler
Smog bubbler
Devices
Standard
Drechsel
Fleming
Fritted
bubbler
Glass bead
bubbler
Fisher
Milligan
bottle
Greiner-
Friedrichs
Greenburg
Smith
Midget
Fritted
bubbler
Capacity
(ml)
125-500
125-500
100
100-500
100-500
275
100-200
500
100
10-20
Sampling
rate
f/min
1-5
1-5
1-5
1-5
1-5
1-5
1-5
1-5
.1-.5
1-4
Efficiency*
%
90-100
90-100
90-100
95-100
90-100
90-100
90-100
90-100
90-100
95-100
i
Comment
- -
Bubblers are '
large. Reduc-
tion of sampl-
ing rate in-
creases efficien-
cy. Several
J
units in series
raises efficiency
Similar to
above
Difficult to
clean
Fritted tubes
available for
simple gas
washing, items
above. Smaller
bubblers pro-
vide increased
gas-liquid con-
tact
Provides for
longer gas-
liquid contact
smaller bubbles
Similar to
Fisher
Milligan
Cylindrical
shape
Cost
\ 6.00
9.00
4.00
6.00
3.65
7.25
12.00
13.00
28.00
7.50
18.00
J 1
Source
Pyrex
Fisher
A. H.
Thomas
A. H.
Thomas
Self con-
structed
Fisher
Sci.
Glass
Sci.
Glass
Sci.
Glass
Ace
Glass
•Under optimum conditions of flow rate, absorbing medium etc. for a particular pollutant.
Figure 5-4. (11) Absorption sampling devices.
Absorbers that use frits with a pore size of approximately 50 micrometers or less
gradually become clogged with use. They may be cleaned by surging the
appropriate cleaning solution back and forth through the frit and thent rinsing with
distilled water in the same fashion. Various substances may be removed from the
frits by cleaning with the appropriate solvent (e.g., hot hydrochloric acid for dirt,
hot concentrated sulfuric acid containing sodium nitrite for organic matter, etc).
5-7
-------
Impingers
Impingers are often used in sampling for gaseous and vaporous pollutants from the
atmosphere. Two types of impingers are shown in Figure 5-5.
Greenburg-Smith Midget
Figure 5-5. Two types of impingers.
A limited amount of investigation has indicated that the impinger is somewhat
less efficient than the fritted absorber for collecting gaseous pollutants. When
several types of absorbers were operated under optimal conditions, the midget
impingers were found to be less efficient than the fritted-glass absorber. In addi-
tion, the threshold concentration for collection with the midget impinger was found
to be somewhat higher than that for several types of fritted-glass absorbers.
Summary
Gas-liquid absorption is the process by which a gaseous pollutant is dissolved in a
liquid medium. If a chemical reaction occurs between the pollutant (absorbate)
and absorbent, the process is termed chemical absorption, whereas, physical solu-
tion of the pollutant in the absorbent is termed physical absorption.
The collection efficiency of any particular absorption process is a function of the
characteristics of the absorption device, and the chemical and physical properties
of the absorbate-absorbent pair. A collection efficiency should be determined for
each sampling analysis situation by a method that gives the accuracy desired.
5-8
-------
References
1. Roberts, Louise R., and McKee, Herbert C. Evaluation of Absorption
Sampling Devices./. Air Poll. Control Assoc. 9: 51, 1959.
2. Stern, Arthur C. Air Pollution, vol. 1, chpt. 11, pages 392-424. New York:
Academic Press, 1962.
",. Calveit, Seymour, and Workman, Walter. The Efficiency of Small Gas
Absorbers./. Ind. Hygiene 22: 318, 1961.
t. Hochheiser, Seymour Methods of Measuring and Monitoring Atmospheric
Sulfur Dioxide, Environmental Health Series-Air Pollution, No. 999-AP-6.
1964.
5. Gage, J. C. The Efficiency of Absorbers in Industrial Hygiene Air Analysis.
Analyst 85: 196, 1960.
6. Calvert, Seymour, and Workman, Walter. Estimation of Efficiency for Bubble
Type Gas Absorbers. Talanta. 4: 89, 1960.
7. Holland, F. A. Brush-Up Your Absorption Theory. British Chemical Engineer-
ing. 9: 294, 1964.
8. Becker, H. G. Mechanism of Absorption of Moderately Soluble Gases in
Water. Industrial and Engineering Chemistry. 16: 1220, 1924.
9. Halsom, R. T.; Hershey, R. L.; and Keen, R. H. Effect of Velocity and
Temperature on Roles of Absorption. Industrial and Engineering Chemistry
16: 1224, 1924.
10. Dankwerts, P. V. Gas Absorption Accompanied by Chemical Reaction. A. I.
Ch. E.J. 1: 456, 1955.
11. Air Sampling Instruments for Evaluation of Atmospheric Contaminants, 5th
ed. American Conference of Governmental Industrial Hygienists. Chapter B-6.
Cincinnati, Ohio, 1978.
5-9
-------
Principles of Adsorption
Basic Principles
Adsorption is the phenomenon by which gases, liquids, and solutes within liquids
are attracted, concentrated, and retained at a boundary surface. The boundary
surface may be the interface between a gas and liquid, liquid and liquid, gas and
solid, liquid and solid, or solid and solid. Of the various boundary surfaces, the
adsorption mechanism between liquid and solid, and gas and solid have received
the most attention. The former with respect to removal of substances from solution
with a solid adsorbent (e.g., purification), and the latter with respect to removing
gaseous pollutants on solid adsorbents of high surface area (1).
A solid adsorbent has a crystal lattice structure. The atoms at the surface of the
lattice are arranged in a regular sequence, which depends on the particular solid's
crystalline structure. The valence or other attractive forces at the surface of a solid
are unsatisfied or unsaturated because they are not united with other atoms. As a
result of this unbalanced condition, the solid surfaces will tend to satisfy their
residual forces by attracting and retaining gases or other substances with which
thev come in contact. This surface concentration of substance is the adsorption
process. The attracted substance is known as the adsorbate, while the surface
substance is called the adsorbent.
In air pollution work, adsorption techniques are commonly used for collecting a
specific gas or combination of gases. A typical process consists of passing a gas
stream through a container filled with an adsorbent such as activated charcoal,
alumina, or silica gel. The gas is bound to the adsorbent by molecular forces and,
if condensation does not occur, the gas remains physically and chemically
unchanged. Following collection, the gas may be removed from the absorbent for
analysis or ultimate deposition by applying heat, passing inert carrier gases through
the system, or treating chemically.
Adsorption can be distinguished from absorption. In absorption the material is
not only retained on the surface, but it passes through the surface and is
distributed throughout the absorbing medium. The term absorption in many cases
implies a chemical reaction between the absorbing medium (absorbent) and the
collected substance (absorbate). For example, water is absorbed by a sponge and by
anhydrous calcium chloride. However, various gases are ., .uiiacr of
activated carbon. Often when the true process is not known the term sorption is
used. (2, 3)
Types of Adsorption
Investigation of the adsorption of gases on various solid surfaces has revealed that
the operating forces are not the same in all cases. Two types of adsorption have
been recognized: (a) physical or van der Waals' adsorption (physiosorption) and
(b) chemical adsorption (chemisorption).
5-10
-------
Physical Adsorption
In physical adsorption, the attractive forces consist of van der Waals' interactions,
dipole-dipole interactions, and/or electrostatic interactions. These forces are similar
to those causing the condensation of a gas to a liquid. The process is further
characterized by low heats of adsorption, on the order of 2-15 kilocalories per mole
of adsorbate, and by the fact that adsorption equilibrium is reversible and rapidly
established.
Physical adsorption is a commonly occurring process. For example, this is the
type of adsorption that occurs when various gases are adsorbed on charcoal. If the
temperature is low enough, any gas will be physically adsorbed to a limited extent.
The quantity of various gases adsorbed under the same conditions is roughly a
function of the ease of condensation of the gases. The higher the boiling point or
critical temperature* of the gas, the greater is the amount adsorbed. This concept
will be discussed in more detail subsequently.
Chemical Adsorption
In contrast to physical adsorption, chemical adsorption is characterized by high
heats of adsorption, on the order of 20-100 kilocalories per mole of absorbate,
which leads to a much stronger binding of the gas molecules to the surface. Heats
of adsorption are on the same order of magnitudes as chemical reactions and it is
evident that the process involves a combination of gas molecules with the adsorbent
to form a surface compound. This type of adsorption resembles chemical bonding
and is called chemical adsorption, activated adsorption, or chemisorption. For
example, in the adsorption of oxygen on tungsten it has been observed that
tungsten trioxide distills from the tungsten surface at about 1200°K. However, even
at temperatures above 1200°K, oxygen remains on the surface, apparently as
tungsten oxide. Additional examples of chemical adsorption are the adsorption of
carbon dioxide on tungsten, oxygen on silver, gold on platinum, and carbon and
hydrogen on nickel.
A comparison of physical and chemical adsorption can be made by considering
the adsorption of oxygen on charcoal. If oxygen is allowed to reach equilibrium
with the charcoal at 0°C, most of the oxygen may later be removed from the char-
coal by evacuating the system at 0°C with a vacuum pump. However, a small por-
tion of the oxygen cannot be removed from the charcoal no matter how much the
pressure is decreased. If the temperature is now increased, oxygen plus carbon
monoxide and carbon dioxide are released from the charcoal. Thus most of the
oxygen is physically adsorbed and can be easily removed, but a small quantity
undergoes a chemical reaction with the adsorbent and is not readily removed. In
some cases, chemical adsorption may be preceded by physical adsorption, the
chemical adsorption occurring after the adsorbent has received the necessary
activation energy.
*Critical temperature may be defined as that temperature above which it is impossible to liquify
gas no matter how high an external pressure is applied.
5-11
-------
In general, with respect to the adsorbent-adsorbate pairs, chemical adsorption is
more specific in nature than physical adsorption. It is usually a much slower pro-
cess, requiring the displacement or selection of the molecules where the reaction is
to occur. The chemisorption process is enhanced at higher temperatures where
existing energy barriers between the adsorbent and adsorbate are overcome. At low
temperatures, chemical adsorption in some systems may be too slow to reach a
measurable amount. In many cases the adsorption occurring is a combination of
both types. At low temperatures physical adsorption may predominate, whereas at
higher temperatures chemisorption may be more prominent. This situation is true
for the adsorption of hydrogen on nickel. However, because of the non-specificity
of van der Waals' forces, physical adsorption may be occurring but be hidden by
chemisorption. Finally, chemical adsorption is usually limited to the formation of a
single layer of molecules on the adsorbent's surface (monolayer adsorption),
whereas in physical adsorption the adsorbed layer may be several molecules thick
(multilayer adsorption).
In most of the adsorption equipment in air pollution control work, physical
adsorption plays the most prominent part (2, 3, 4).
Variables Affecting Gas Adsorption
The quantity of a particular gas that can be adsorbed by a given amount of
adsorbent will depend on the following factors: (a) concentration of the gas in the
immediate vicinity of the adsorbent; (b) the total surface area of the adsorbent;
(c) the temperature of the system; (d) the presence of other molecules competing
for a site on the adsorbent; (e) the characteristics of the adsorbate such as weight,
electrical polarity, chemical reactivity. Ideal physical adsorption of a gas would be
favored by a high concentration of material to be adsorbed, a. large adsorbing sur-
face, freedom from competing molecules, low temperature, and by aggregation of
the adsorbate into a form that conforms with the pore size of the attracting adsor-
bent (5, 6).
Several of the above listed variables will not be discussed in greater detail.
Adsorption Isotherms
Adsorption processes where physical adsorption rather than chemisorption
represents the final state can be explained in terms of equilibrium measurements.
For a given amount of adsorbent with a given surface area the amount of gas
adsorbed is dependent on the pressure (or concentration) of the gas surrounding
the adsorbent. The higher the pressure or concentration of the gas at a given
temperature, the greater the amount of gas adsorbed. When an adsorbent and gas
are mixed, the amount adsorbed will gradually increase while the concentration of
the adsorbate in the system decreases until the rate of adsorption becomes equal to
the rate of desorption. Thus an equilibrium between the two reactions is estab-
lished. If additional gas is added to the system the amount adsorbed will increase
until equilibrium is again established. Likewise, if the gas concentration is
decreased the adsorbent will lose gas to its surroundings until equilibrium is again
reached.
5-12
-------
The description of the relationship between the quantity of gas adsorbed at
various concentrations or pressures at constant temperature is called an adsorption
isotherm. An adsorption isotherm consists of a plot of the data obtained from
measuring the amount of gas adsorbed (e.g., grams adsorbed per gram of adsor-
bent) at various gas concentration or pressure (e.g., moles per liter or
atmospheres), as the case may require, at equilibrium under a condition of
constant temperature. Adsorption isotherms are useful in that they provide a means
of evaluating: (a) the quantity of gas adsorbed at various gas concentrations;
(b) adsorptive capacities at various gas concentrations; (c) the adsorptive capacity
as a function of concentration and type of gas; and (d) the surface area of a given
amount of adsorbent (1, 2, 3).
Types of Adsorption Isotherms (1, 3)
The graphic plots of adsorption isotherms yield a wide variety of shapes. Six
general types of isotherms have been observed in the adsorption of gases on solids;
these are illustrated in Figure 5-6. In physical adsorption all six isotherms are
encountered, while in chemisorption only type 1 occurs.
Pressure or concentration
Figure 5-6. Gas adsorption isotherms.
Type 1 —This type represents the adsorption of a single layer of gas molecules on
the adsorbent. There is no interaction between the adsorbed molecules.
Type 2-This isotherm begins like type 1 but is modified at high pressure by
multilayer adsorption. There is definite interaction between the layers of
adsorbed gas molecules.
5-13
-------
Type 3 —This type of isotherm is rare. It occurs only when initial adsorption
favors a very few strong sites. The interaction between adsorbed molecules is so
strong that vacant sites next to occupied sites are stronger than any other vacant
sites. In this type of adsorption the number of effective sites increases with
coverage of the adsorbent.
Type 4 & 5 —These two are similar to types 2 and 3 respectively, except that
they continue to exhibit adsorption at high adsorbent coverage.
Type 6— This type resembles type 3 with monolayer adsorption first and then
continued deposition of a multilayer film.
Adsorbate Characteristics
The major adsorbate characteristics affecting the amount of gas adsorbed are: the
ease of liquefaction of the gas, adsorbate size, concentration of the gas, and
presence of other gases.
Gas Liquefaction
The specificity by which certain gases are adsorbed on solid adsorbents is illustrated
in Figure 5-7, where the volumes of different gases adsorbed by one gram of char-
coal at 15°C are tabulated.
Gas
H2
N2
CO
CH4
CO2
HC1
H2S
NH3
C12
SO2
Volume
adsorbed (cc)
4.7
8.0
9.3
16.2
48.0
72.0
99.0
181.0
235.0
380.0
Critical temperature
<°K)
33
126
134
190
304
324
373
406
417
430
* Volumes of gases have been reduced to standard conditions (0°C and 1 atmosphere pressure).
Figure 5-7 (3). Adsorption of gases on one gram of charcoal at 15°C*.
Figure 5-7 indicates that the extent of adsorption parallels the increase in critical
temperature. This correlation suggests that gases which liquify easily (high critical
temperatures) are more readily adsorbed. However, it does not imply that the
adsorbates exist as liquids on the adsorbent's surface. A similar relationship is
obtained with boiling points (3).
Adsorbate Size
The size of the gas molecule to be removed by adsorption is characterized by a
lower and upper range. The lower size limit is imposed on physical adsorption by
the requirement that the pollutant must be higher in molecular weight than the
normal components of air. In general, gases with molecular weights greater than
45 are readily removed by physical adsorption. This size includes most odorous and
5-14
-------
toxic gases of air pollution interest. Gases of interest of lower molecular weight,
such as formaldehyde and ammonia, may be removed by chemical adsorption
methods using appropriately impregnated adsorbents.
For the upper limit the individual particles must be sufficiently small so that
Brownian motion or kinetic velocities will ensure effective contact by collision
between them and the granular adsorbent. Although moderate efficiencies may be
obtained for very fine mists, the upper limit is generally in the range of molecular
size.
Gas Concentration
As seen from the examination of adsorption isotherms, the quantity of gas
adsorbed is a function of the gas concentration or pressure. An increase in concen-
tration or pressure in the vicinity of the adsorbent results in an increase of the total
amount of gas adsorbed.
Presence of Other Gases
Since the presence of additional gas molecules in a particular adsorbent-adsorbate
system causes competition for the limited number of adsorption sites present, the
observed effect is a reduction in the amount of adsorbate removed.
Adsorbent Characteristics
Most of the common adsorbents in use are more or less granular in form and are
supported in a column through which the gas to be sampled is drawn. Common
adsorbents have the capacity to adsorb 8 — 40 percent of their weight. An ideal
adsorbent should be granular and of such size and form that it offers little or no
resistance against flow. It should have a high adsorptive capacity; be inert and
specific; be resistant to breakage, deterioration, and corrosion; be easily activated;
and provide an easy release of adsorbate. Unfortunately, no one adsorbent possesses
all these characteristics, so it becomes a matter of choosing the best adsorbent for a
particular job (5, 7, 8,).
Surface Area
All solids are capable of adsorbing gases to some extent. However, since adsorption
is a surface phenomenon, it is not very pronounced unless the adsorbent possesses a
large surface area for a given mass. For this reason, materials like silica gel and
charcoals obtained from wood, bone, coconut shells, and lignite are very effective
adsorbing agents. Since large surface areas are desirable for extensive adsorption,
this factor is of primary importance in determining the amount of absorbate that
can be held by a unit of adsorbent. Solid adsorbents may vary in surface area from
less than 1 to over 2,000 square meters per gram. Typical approximate surface
areas of several adsorbents are presented in Figure 5-8. The latter two substances
owe their high surface area to their porosity. They are thus capable of taking up
large volumes of various gases.
5-15
-------
Adsorbent
Clay
Asbestos
Chalk
Carbon black
Silica or alumina gel
Activated carbon
Area (mVg)
5-15
10-20
20-30
50-100
200-800
500-2000
Figure 5-8. Typical surface areas of adsorbents.
The extent of adsorption can be further increased by activating the adsorbents
by various methods. For example, wood charcoal is activated by heating to between
350 — 1000°C in a vacuum, in air, in steam, and/or in the presence of other gases
to a point where the adsorption of carbon tetrachloride at 24°C can be increased
from 0.011 gram per gram of charcoal to 1.48 gram. The activation process
involves distilling out various impurities from the adsorbent, thus leading to the
formation of a larger free surface area for adsorption. Occasionally, large surface
areas are produced by the original cellular structure of the plant, as in the case of
coconut shell charcoal. However, the activation process will increase the porosity of
the material and may, under some circumstances, cause it to be less stable as an
adsorbent. For example, if the temperature is raised, the porous structure of the
adsorbent may aggregate into larger units that tend to become smooth and inac-
tive. In many cases the past history of the adsorbent with respect to preparation
and method of activation is just as important as the chemical characteristics in
determining the adsorption capacity (1, 3, 4).
Often the adsorbent will exhibit an inherent preference for the adsorption of cer-
tain gases. This preference is primarily due to such factors as the method of
preparation and activation, and the chemical nature of the adsorbent's surface.
Preparation and activation methods not only may increase total adsorptive capa-
city, but they may also affect the adsorption process with respect to adsorbate size.
Pore Size
The pore size in the more porous adsorbents may vary in diameter from a few to
several hundred angstrom units. This may become a critical factor in selecting an
adsorbent to remove a particular adsorbate. For example, iodine may be adsorbed
on an adsorbent with a pore size of 10 A in diameter, while methylene blue is
excluded by pores having a diameter less than about 15 A (1).
Chemical Nature
The chemical nature of the adsorbent's surface is an additional factor of con-
siderable importance. It is of particular interest in chemical adsorption where a
rapid rate and a large degree of chemical reaction is desirable. In physical adsorp-
tion the nature of the surface is one of the primary factors influencing the strength
of the adsorbent-absorbate attraction. For example, a pure graphite surface
physically adsorbs hydrophobic (i.e., water-hating) compounds to a large extent,
while oxygenated surfaces are generally required to adsorb hydrophyllic (i.e.,
water-loving) compounds appreciably at room temperature (1).
5-16
-------
Typical Adsorbents t
The various adsorbents used in physical adsorption may be classified according to
their degree of polarity. For example, activated carbon, which is commonly known
as a non-polar adsorbent, is largely composed of neutral atoms of a single species
exhibiting little polarity. The non-polar adsorbents are most effective for gross
decontamination of moist air streams containing materials of little polarity (e.g.,
organic molecules).
The majority of the commercially important adsorbents other than carbon
derivatives are simple or complex oxides. Their surfaces consist of heterogeneous
distributions of charge on a molecular scale. They are strongly polar in nature.
These adsorbents show a greater selectivity than do the carbon derivatives and
exhibit a much stronger preference for polar than for non-polar molecules. In
separation of various gases, the polar solvents are more useful than carbon
derivatives. However, they are much less useful for overall decontamination of
moist air streams, since the strongly polar water molecules are preferentially
adsorbed (6).
Carbon
Various forms of carbon serve as efficient adsorbents. It has been shown that the
material from which the carbon is prepared has a demonstrable effect upon the
ability of the carbon to adsorb various gases. Carbon prepared from logwood, for
instance, has approximately twice the capacity for adsorption as carbon from
rosewood. Similarly, coconut shell is about twice as efficient as logwood. Strangely
enough the carbon prepared from harder, denser materials such as peach and
other fruit pits, and coconut shells have the highest adsorptive capacities. Primary
carbon is not nearly as efficient as activated carbon. The adsorbents "activated
charcoal," "active carbon," "adsorbent carbon," and "adsorbent charcoal" may be
activated in a slightly different manner, but the terms are generally considered
synonymous.
Activated carbon has a high adsorptive capacity, a high degree of hardness, high
reliability and other premium qualities. Almost all volatile materials, whether they
are chemicals or mixtures of odor-causing substances, are retained within the
microscopic porous structure to some extent. The only gaseous materials that it will
not adsorb very well are low molecular weight gases such as oxygen, nitrogen, and
carbon monoxide. Activated carbon finds its major application in solvent recovery
and odor removal. It is also employed to a limited extent in the removal and
monitoring of hydrogen sulfide, sulfur dioxide, and other toxic gases. Activated
carbon is perhaps the most widely used adsorbent in air pollution control. The
following substances are some of those that have been shown to be appreciably
adsorbed upon activated carbon:
acetic acid acetone ammonia
benzene acetaldehyde hydrochloric acid
ethyl alcohol mercury vapor nitrous oxide
carbon tetrachloride iodine carbon dioxide
methyl alcohol carbon disulfide noble gases
chloroform diethyl ether PVC
5-17
-------
Molecular Sieve
Molecular sieve adsorbents* are synthetic sodium or calcium alumino-silicate
zeolites of very high porosity. They are another representative of the siliceous
adsorbents.
The structural formula of a typical molecular sieve is
Me,/ JA
where Me represents exchange cations of charge n. The zeolite is precipitated as a
white powder, bonded with clay, and formed into roughly spherical beads of four
to twelve mesh size. The adsorbent is activated with heat to drive off water of
hydration. The resulting product is a crystalline solid of very porous structure.
Again the adsorptive characteristics are dependent on the method of preparation.
Molecular sieves can be made very specific with respect to pore size. This
characteristic gives them the outstanding property of being specific on the basis of
adsorbate size and shape. Molecular sieves show a strong preference for the more
polar molecules. For example, these adsorbents will not adsorb organic molecules
that match their pore size from a moist stream of air. The accompanying water
molecules being adsorbed in preference. Molecular sieves are truly selective
adsorbents because they can separate mixtures on the basis of differences in
molecular size, degree of polarity, and extent of carbon bond saturation. In addi-
tion to their selective properties, molecular sieves possess a high adsorptive capacity
over wide ranges of concentration and temperature. They also are capable of
removing impurities to extremely low concentrations. These adsorbents have been
tested successfully on carbon dioxide, hydrogen sulfide, acetylene, ammonia, and
sulfur dioxide. They show promise for adsorption of compounds of low molecular
weight (9).
Adsorption Losses in Air Sampling
Each adsorption medium used in atmospheric sampling has different limitations
and problems. The problems most frequently encountered are:
• irreversible adsorption
• variable desorption efficiency
• interference by water vapor
Activated carbon is used extensively because of its high affinity for organic
substances. Irreversible adsorption and variable desorption efficiencies are two prin-
cipal problem areas associated with carbon sampling devices. Carbon can also serve
as a potent catalyst creating the possibility of m-situ reactions during sampling.
Other alternative adsorption media that have recently been used extensively in
air sampling are thermally stable, polystyrene divinyl benzene co-polymers (15, 16,
17). These media were used at ambient temperatures to collect volatile organic
compounds. The volatile organics were recovered by thermal desorption followed
"Often referred to as molecular sieve absorbents.
5-18
-------
by gas chromatographic analysis. These polymers are non-polar and have little
affinity for water. Water desorption represents a severe problem in the analysis pro-
cedure. Adsorption sampling devices are being used primarily for the collection of
volatile organic compounds.
Current Applications of Adsorption in Atmospheric Sampling
Carbon, porous polymers such as Porapack Q_, Porapack P, Tenax GC, XAD-
resins, and polyurethane foam have been used extensively in collecting pesticides,
polychlorinated biphenyls, and other organic compounds in ambient air (18, 19,
20). These media can be used in sampling devices, which can be modifications of
the hi-volume sampler (see Figure 5-9). Here polyurethane foam is used to collect
organics, namely PCB's.
Stainless steel throat extension
Polyurethane foam
plug location
Faceplate
Throat extension
Wire
retainer
Exhaust duct
(3 m minimum length)
Figure 5-9. Assembled sampler and shelter with exploded view of the filter holder.
5-19
-------
To pump
Stainless steel tubing
Spring
Stainless steel screen
Chromosorb 102
Stainless steel screen
Stainless steel tubing
Figure 5-10. High speed organic vapor collector.
Figure 5-10 illustrates an adsorbent sampling cartridge and Figure 5-11 shows a
cartridge placed in a thermal desorption system. Figure 5-12 shows dynamic enrich-
ment, which is repeated absorption from many different cartridges onto a single
cartridge to attain enough of the specie of interest for measurement.
Carrier gas
Cap
Teflon body
Insulation
Class tube
Class wool
Tenax absorbent
^Carrier gas and
pollutants—gas chromatograph
Figure 5-11. Desorption of pollutants from a tenax-GC cartridge.
5-20
-------
Flash heater
Tenax
column
Sample inlet
He carrier
Carrier
purifier
Oven
- 180°C
condenser
tube
GC column
direct
GC iniet
Figure 5-12. Dynamic enrichment on adsorption column (experimental set-up).
Summary
The adsorption process is characterized by either physical or chemical forces. In
some cases both types may be involved. Where physical forces predominate the pro-
cess is termed physical adsorption, whereas chemical adsorption describes chemical
action.
Adsorption phenomena may be quantitated by considering such adsorbate-
adsorbent characteristics as gas composition, concentration, and temperature, as
well as adsorbent type, surface area, and pore size.
5-21
-------
References
1. Graham, D. Adsorption Equilibrium, Adsorption, Dialysis, and Ion Exchange.
Chemical Engineering Progress Symposium Series. 24, pp 17-23. American
Institute of Chemical Engineers, New York. 1959.
2. Daniels, F., and Alberty, R. A. Physical Chemistry, dipt. 17, pp 522-526. Lew
York: John Wiley and Sons, Inc., 1955.
3. Maron, S. H., and Prutton, D. F. Principles of Physical Chemistry, chpt. 7,
pp. 214-225. New York: The MacMillan Co., 1958.
4. Brey, W. S., Jr. Principles of Physical Chemistry, chpt. 7, pp. 244-253. New
York: Appleton-Century-Crafts, Inc., 1958.
5. Stern, A. C. Air Pollution, vol. I, chpt. 11. New York: Academic Press, 1962.
6. Stern, A. C. Air Pollution, vol. II, chpt. 33. New York: Academic Press, 1956.
7. Magill, P. L.; Holden, F. R.; and Ackley, C. Air Pollution Handbook,
chpt. 13, p. 83. New York: McGraw-Hill Book Company, Inc., 1956.
8. Air Sampling Instruments, chpt. Al and B6. American Conference of Govern-
mental Industrial Hygienists. Cincinnati.
9. Gresmer, G. J.; Jones, R. A.; and Lautensach, H. Molecular Sieves, Adsorp-
tion, Dialysis, and Ion Exchange. Chemical Engineering Progress Symposium
Series, pp. 45-50. American Institute of Chemical Engineers. New York, 1959.
10. Codudal, M. Determination of Radon in Uranium Mines by Sampling on
Activated Charcoal./ Phys. Radium, 16: 479, 1955.
11. Shleien, B. The Simultaneous Determination of Atmospheric Radon by Filter
Paper and Charcoal Adsorptive Techniques./ Amer. Industrial Hygiene 24:
180-187, March-April 1963.
12. Sell, C. W., and Flygare, J. K., Jr. Iodine Monitoring at the National Reactor
Testing Station. Health Physics 2: 261-268, 1960.
13. McConnon, D. Radioiodine Sampling with Activated Charcoal Cartridges.
AEG Research and Development Report. HW-77126. April 1963.
14. Browning, W. E. Removal of Volatile Fission Products from Gases. Nuclear
Reactor Chemistry. First Conference, Gatlinburg, Tennessee. TID-7610,
October 1960.
15. Dravnieks, A., et al., High-speed Collection of Organic Vapors from the
Atmosphere. Env. Sci. and Tech. 5: 1220, 1971.
16. Williams, F. N., et al., Determination of Trace Contaminants in Air by Con-
centrating on Porous Polymer Bead Anal. Chem. 40: 2232, 1969.
17. Zlatkis, A., et al. Concentration and Analysis of Trace Volatile Organics in
Gases and Biological Fluids with a New Solid Adsorbent. Chromatographia
6:67, 1973.
18. Lewis, R. G.; Brown, A. R.; and Jackson, M. D. Evaluation of Polyurethane
Foam for Sampling of Pesticides, Polychlorinated Biphenyls and
Polychlorinated Naphthalenes in Ambient Air. Anal. Chem. 49: 1668, 1977.
19. Versino, B.; Groot de M.; and Caeiss, F. Air Pollution-Sampling by Adsorp-
tion Columns Chromatographic 7: 302, 1974.
20. Stratton, Charles L., et al., A Method for the Analysis of Polychlorinated
Biphenyls (PCBs) in Air, EPA-600/4-78-048, August, 1978.
5-22
-------
Selection and Performance of Wet Collector Media
Introduction
In the design of sampling trains, the most important component of the entire
system is the collector. The process of pollutant removal is generally accomplished
by absorption, adsorption, etc. The collector may take the form of a bubbler,
impinger, etc. The process to be discussed is that using a wet collector for the col-
lection of gases, vapors, and particulate matter. Some of the more important fac-
tors to consider are:
• Gas flow rate
• Bubble size
• Height of liquid column
• Reaction rate
• Solubility of pollutant
Absorber Design
General Considerations
Solubility of Pollutant
The solubility of a pollutant in a solvent must be considered in determining the
type of absorber to choose. It will also determine the conditions under which the
sample will be taken. The absorption coefficient is one method employed to express
the results of solubility measurements with gases. The absorption coefficient a is
given by:
Where:
V0 = the volume of gas dissolved (ml)
V=the volume of solvent (ml)
p = the partial pressure of the gas (atm.)
Some typical absorption coefficients are given in Figure 5-13.
Solvent
Water
Carbon disulfide
Chloroform
Ethyl alcohol
Acetone
Ethyl ether
Benzene
H,
.017
.031
—
.080
.065
.12
.066
He
.009
—
—
.028
.030
—
.018
N8
.015
.049
.120
.130
.129
.24
.104
0*
.028
—
.205
.143
.207
.415
.163
CO
.025
.076
.177
.177
.198
.38
.153
CO*
.88
.83
3.45
3.0
6.5
5.0
NO
.047
—
—
—
—
—
—
HUS
2.68
—
—
—
—
—
NHS
710
—
—
—
—
—
" ,
(Glasstone, S., Textbook of Physical Chemistry, p. 695, D. Van Nostrand, New York, 1946)
Figure 5-13. Absorption coefficient of gases at 20°C.
5-23
-------
Influence of temperature-"When gases dissolve in a liquid, there is generally a
liberation of heat; it follows, therefore, that an increase of temperature will result
in a decrease of solubility. It is for this reason that gases may be readily expelled
from solution by boiling. By thermodynamic methods, it is possible to show that an
increase in temperature will decrease the solubility of a gas. This effect can be seen
in Figure 5-14.
Gas/temp
OC
30C
Helium
.0094
.0081
Nitrogen
.0235
.0134
Oxygen
.0489
.0261
Carbon dioxide
1.713
.665
(Glasstone, S., Textbook of Physical Chemistry, p. 696).
Figure 5-14. Influence of temperature on solubilities of gases in water.
Influence of pressure-The most important factor influencing the solubility of a
gas is pressure; increasing the pressure of the gas will tend to increase its solubility.
The pressure is expressed by Henry's law, which states that the mass of a gas
dissolved by a given volume of solvent, at constant temperature is proportional to
the pressure of the gas which it is in equilibrium.
(Eq. 5-9) m = kP
Where: m = mass of gas dissolved by unit volume of solvent (g)
p = equilibrium pressure (atm.)
k = constant
Some examples of pressure versus solubility effects are given in Figure 5-15.
Solvent/pressure
100mm
200mm
400mm
700mm
Methyl alcohol
42.5
42.7
43.1
43.3
Acetone
67.2
68.0
69.2
72.8
Methyl acetate
75.8
77.1
77.6
79.0
(Glasstone, S., Textbook of Physical Chemistry, p. 697).
Figure 5-15. Influence of pressure on solubility of CO* in various solvents at - 59°C.
Rate of Reaction
All chemical reactions take place at a definite rate, depending on process condi-
tions The most important factors are: concentration of reactants, temperature,
and presence of a catalyst or inhibitor. Some reactions are so rapid that they
appear to be instantaneous, whereas others are so slow at ordinary temperatures
that no detectable change would be observed in the course of years. Between these
two extremes are many processes taking place with measurable velocities at
temperatures easily accessible in the laboratory.
5-24
-------
Since the rate of a reaction is proportional to the concentration of the reacting
substances it is evident that the rate of the process must fall off as the reaction pro-
ceeds. This phenomenon can, however, be used to advantage by increasing the
concentration of the absorbing liquid, thereby forcing the reaction to approach
completion rapidly.
Collection Efficiency
There are three major factors inherent in the design of a bubbler that can affect
the efficiency of the absorber.
• Flow rate
• Bubble size
• Height of liquid column
Bubble Size
The surface area at the gas-liquid interface is inversely related to the average
volume of the gas bubble. As the volume of individual bubbles decreases, the sur-
face area at the gas-liquid interface increases.
The efficiency of absorption of a gas or vapor by chemical reaction or physical
absorption depends on the probability of successful collisions with molecules of
reagent or solvent at the gas-liquid interface. For a given concentration of reagent
this will depend on the surface area of the gas bubbles, on the length of the col-
umn of liquid through which the bubbles must pass, and the rate at which they
rise through the liquid.
Collection efficiency varies inversely with flow rate and bubble size and varies
directly with the height of the liquid column.
Flowrate
The gas flow rate through an absorber is one of the factors determining the effi-
ciency of an absorber. Figure 5-16 shows clearly that as flow rate increases, for the
absorbers studied, the efficiency decreases. This efficiency versus flow rate curve
should be determined for each absorber and used in any analysis.
Height of Liquid Column
The length of the column of liquid in an absorber is important in determining effi-
ciency. The velocity of rise of bubbles is approximately constant at 24 cm. sec. for
bubble diameters greater than 0.2 cm (7). Since the bubbles rise at approximately
24 cm. sec. they will be in contact with a liquid column 24 cm. long for 1 second.
48 cm. long for 2 seconds, etc. The longer the gas bubble is in contact with the
liquid, the more pollutant is transferred from the gas phase to the liquid phase
until gas-liquid equilibrium is approached.
5-25
-------
£
a
u
u
100
90
80
70
60
50
1 T
i 1 r~
Greenburg-Smith
impinger \
(a)
Midget bubbler
250 ml gas
washing bottle
_L
_L
_L
Ammonia, 1 ppm
Pressure = 26.5" Hg
Temperature = 80°F (approx.)
(a) Maximum obtainable
not advisable due to entrainment
' ' 1 1
0.1 0.2 0.3 0.4 0.5 0.6
Air flow rate, cubic feet/minute
0.7
0.8 0.9
Figure 5-16. Performance curves—commercially available absorbers.
Retention of Gases and Vapors by Solution
The equation defining Raoults' Law is:
(Eq. 5-10) p = Npa
Where: p = partial pressure of gas to be dissolved (atm.)
N= mole fraction of gas
p0 = vapor pressure of gas (atm.)
From this relationship one can calculate the solubility of a gas below its critical
temperature, on the assumption that the solution behaves in an ideal manner. For
example, the critical temperature of ethane is 34 °C. At 25 °C the pure liquid has a
vapor pressure of 42 atmospheres. According to Raoult's Law, therefore, the
solubility of ethane at 25 °C and a pressure of 1 atm. in any solvent in mole frac-
tion can be determined by the following:
N= 2- = — =0.024
po 42
since p is 1 atm., and p0 is 42 atm. The actual solubility in n-hexane at 25°C and
1 atm. pressure is 0.017 mole fraction. This variation is due to n-hexane being a
non-ideal solvent.
5-26
-------
To extend the method for calculating gaseous solubilities to temperatures above
the critical temperature, it is necessary to estimate the hypothetical vapor pressure
of the liquid by a suitable extrapolation. This is best done by using the integrated
form of the Clapeyron-Clausius equation, which is,
(Eq. 5-11)
where Le is the latent heat of vaporization.
If the vapor pressure at any two temperatures is known, the value at any other
temperature may be evaluated on the assumption that the molar heat of vaporation
remains constant. The critical temperature of methane is 95.5°C, and the
hypothetical vapor pressure of the liquid at 20 °C is 310 atm., giving an ideal
solubility at this temperature and a pressure of 1 atm. of 1/310 = 0.0032 mole frac-
tion; this is very close to the solubilities actually found in n-hexane and m-xylene.
Since the solubility in mole fractions of a gas at 1 atm. pressure is equal to l/p,
where p is the vapor pressure of the liquified gas, it is evident that, for ideal solu-
tions, the lower the vapor pressure at the given temperature the greater will be the
solubility of the gas. Gases that are liquified only with difficulty, that is to say,
those having very low boiling points, may be regarded as having high vapor
pressures; such gases will, therefore, have low solubilities. It follows that, in
general, easily liquifiable gases will be the most soluble; this is in agreement with
observation in most cases.
Although the solubility of a gas, in mole fractions, should theoretically be
independent of the nature of the solvent, this is not true in practice because of
departure from ideal behavior. Some data for solutions of gases, showing devia-
tions, are listed in Figure 5-17. The solubilities in water are exceptionally low, since
water is both polar and associated, and also has a very high internal pressure, solu-
tions of gases of the type mentioned in Figure 5-17 would hardly be expected to
behave ideally. Even chlorine and carbon dioxide, which interact with water and
are generally regarded as relatively soluble gases, have solubilities considerably
below the calculated values, because of their low polarity and internal pressure. A
quite different type of behavior is shown by ammonia, which is a highly polar
substance with a high internal pressure. In hydrocarbon solvents, therefore,
solubility is considerably below the ideal value, whereas in alcohol and water the
observed solubility is somewhat greater than that calculated. If allowance could be
made for interaction between ammonia and the solvent, good agreement would be
found. A corollary to the foregoing conclusions is that for a number of gases of
similar polarity and internal pressure (e.g., hydrogen, nitrogen, carbon monoxide,
oxygen) that do not react with the solvent, the ratio of the solubilities in various
solvents should be approximately independent of the nature of the gas. This
generalization is roughly true in practice, and only gases such as carbon dioxide
and ammonia, which are not in the same category, are exceptions.
5-27
-------
Gas
Nitrogen
Carbon dioxide
Oxygen
Argon
Ideal
10
11
16
21
Nitrobenzene
2.6
3.9
Ethyl alcohol
3.3
4.5
6.5
A niline
1.1
1.9
Water
0.13
0.19
0.17
0.17
Figure 5-17. Ideal and observed solubilities at 20°C.
Retention of Gases and Vapor by Chemical Reaction
The usual objective in the selection of an absorbent for scrubbing a gas is to find a
liquid, possibly a solution, with a very large capacity for absorbing the solute
without building up an appreciable equilibrium back pressure. This can be
accomplished readily by choosing a chemical with which the solute reacts irrever-
sibly, as when an aqueous solution of sodium hydroxide is used to absorb carbon
dioxide. There are indeed very few absorptions of a gas in a liquid that are not
accompanied by a chemical reaction to some degree. Thus, when ammonia
dissolves in water, ionization occurs that may be looked upon as a chemical
change. A similar phenomenon, though potentially weaker, occurs when carbon
dioxide dissolves in water. A much stronger and more definite chemical change
takes place when ammonia is dissolved in an acid, or carbon dioxide in a base.
There is no sharp line dividing pure physical absorption from absorption con-
trolled by the rate of a chemical reaction. Most cases fall in the intermediate
range; the rate of absorption being limited both by resistance to diffusion and by
the finite rate of reaction. Simultaneous occurrence of a chemical reaction renders
the mechanism of absorption more complicated. The theory of purely physical
absorption rests on the assumption of the two-film concept. This theory may be
carried over to the case where a simultaneous reaction occurs, however, modifica-
tion in film resistance will become apparent. Thus when carbon dioxide is dissolved
in water, the rate controlling factor is not the migration of the dissolved carbon
dioxide from the liquid surface into the liquid interior, simply because the rate of
solution of the gas in water is small from the very start. On the other hand if
absorption of carbon dioxide in a solution of sodium hydroxide is considered, the
rate of absorption is very rapid and then the rate of migration of the carbonate
into the main body of the liquid becomes rate controlling.
These phenomena are complex, and, although considerable advances have been
made, the situation is still very obscure. Whenever there is a pronounced chemical
reaction occurring simultaneously vtith absorption, there are essentially two effects
that must be considered; these pertain to:
• Modification of capacity (rate) data
• Modification of the driving force.
Capacity coefficients will generally, but not always increase when a chemical
reaction occurs simultaneously with absorption. At present there is no data
available to permit a correlation that will allow for estimation of capacity data.
5-28
-------
As far as driving force is concerned, an increase is usually observed as a conse-
quence of a chemical reaction. In many cases the dissolved gas, once having
reacted with a constitutent in the liquid, offers virtually no resistance to further
absorption. This is the case when carbon dioxide or sulfur dioxide are dissolved in
basic solutions.
Retention of Particulate Matter
The design of the absorber plays a most important part in the retention of par-
ticulate matter by a liquid. A liquid absorber is highly efficient for retaining par-
ticles only when the velocity of the air at the jet approaches that of sound and the
particles impinge with high velocity on a surface in the liquid. The sudden change
in kinetic energy results in the virtually complete trapping of all particles having a
diameter greater than 1 micrometer.
References
1. Roberts, L. R. and McKee, H. C. Evaluation of Absorption Sampling Devices.
Journal Air Pollution Control Assoc. 9: 41-53, 1959.
2. Droege, H. F., and Ping, A. Y. Relative Efficiencies of Various Collection
Devices Used for Source Testing. Presented at the Sixth Conference on
Methods in Air Pollution Studies, Berkeley, California, Jan. 6-7, 1964.
3. Elkins, H. B.; Hobby, A. K.; and Fuller, J. E. The Determination of
Atmospheric Contaminants I. Organic Halogen Compounds. Journal of
Industrial Hygiene and Toxicology, 19: 474-485, 1937.
4. Saltzman, B. E. Preparation and Analysis of Calibrated Low Concentrations of
Sixteen Toxic Gases. Analytical Chemistry. 33: 1100-1112, July 1961.
5. Perry, R. H., and Pigford, R. L. Kinetics of Gas-Liquid Reactions. Industrial
and Engineering Chemistry, 45: 1247-1253, June 1953.
6. Calvert, S., and Workman, W. The Efficiency of Small Gas Absorbers.
Industrial Hygiene Journal, pp. 318-324, August 1961.
7. Calvert, S. and Workman, W. Estimation of Efficiency for Bubbler-Type Gas
Absorbers. Talanta, 4: 89-100, 1960.
8. Gage, J. C. The Efficiency of Absorber in Industrial Hygiene Air Analysis.
Industrial Hygiene Journal 85: 196-203, March 1960.
9. Leva, M. Tower Packings and Packed Tower Designs. Akron: United States
Stoneware Co., 1951.
10. Sherwood, T. K., and Pigford, R. L. Absorption and Extraction. New York:
McGraw-Hill, 1952.
5-29
-------
Principles of Grab Sampling
Introduction
The term "grab sample" suggests two concepts: (a) a sample taken at a particular
time and place within an interval of a few seconds to a minute or two, and (b) a
small representative portion removed from the gross sample with no alteration.
Grab samples are usually collected in one of the following ways:
• Using an evacuated container,
• Purging (displacement of air),
• Displacement of a liquid,
• Inflation of a plastic bag,
• Using a syringe,
• Using an adsorbent cartridge.
Evacuated Containers
Evacuated containers used for grab sampling are of several types. One common type
is a strong glass bulb of 250 — 300 ml capacity (although bulbs with volumes as
large as 1 — 2 liters are sometimes used) (see Figure 5-18).
Scratch
250 to 1000 ml
Figure 5-18. Vacuum tube.
To use this type of container, the bulb is evacuated until almost all the air has
been removed. In the last stages of evacuation, the neck is sealed. Then at the
sampling site, the neck is scratched and broken. Sampling is instantaneous, and
will continue until the internal pressure is equivalent to the external pressure. The
broken end is then sealed with wax and sent to the lab for analysis.
5-30
-------
There are several advantages in the use of this collector: it is simple to use and
no pump or manometer need be taken to the sampling site. However, the tube
must be redrawn, re-evacuated and resealed if it is to be used again. There is also
the danger of breakage.
An evacuated flask fitted with a stopcock or vacuum cap can also be used in this
type of sampling (see Figure 5-19). The flask is evacuated and then sealed by giving
the cap a half turn. When sampling is to occur, the cap is turned to the "open"
position and the air will be drawn into the flask. The cap is closed after sampling
and the flask is returned to the laboratory.
Figure 5-19. Vacuum flask.
During the transport of the evacuated container to the sampling site there is a
possibility of slow leakage through poorly-fitted stopcocks. This would, of course,
completely invalidate the results. The pollutants to be analyzed may also degrade
or react with other compounds in the evacuated flask. This apparatus has the
advantage of being easy to reuse. Such collectors should be placed in a protective
container or wrapped with adhesive tape to reduce hazards of implosion.
If for some reason the containers are not completely evacuated, it may be
necessary to subtract a residual volume from the volume of the flask to determine
the volume of air sampled. Let Vf be the volume of the vessel; after evacuation let
the temperature and residual pressure in the flask be TI and Pt. The flask is
transported to the sampling site and opened; the flask temperature and pressure
now become T2 and P2. The volume of air sampled, Vs, is given by:
(Eq. 5-12) V.= Vf- V,
5-31
-------
Where Vx is the volume occupied by the residual gas. Assuming gas ideality for the
residual gas:
(Eq. 5-13)
Hence: K = F, I 1 -
T,
Pi ?V
P2T,
If the ratio Pi/Pz is small (almost complete evacuation) then the correction can
be neglected and
The presence of the pollutant in the residual gas would further complicate the
matter.
Air Displacement or Purging
Cylindrical tubes with stopcocks at each end are used as collectors (see Figure
5-20). The stopcocks are opened and the tube is thoroughly purged. After
sampling, the tube should be held in place until the stopcocks have been closed
and the aspirating device has been removed.
250 to 300 ml
Figure 5-20. Gas-displacement collector
Metal containers of the same general design have been employed, but they have
been found to react with many samples. Their real advantage lies in the fact that
they are virtually unbreakable.
The sample air is drawn through the container using any of a variety of pumps.
Enough air must be drawn through to completely flush out old unrepresentative air
that may be present.
The necessary volume of air required will vary, but in all cases it will be at least
several times greater than the volume of the container. Theoretically all of the old
air can never be eliminated by pumping. Since this pumping process may take a
relatively long time, it is not strictly an instantaneous sample. If the concentration
of pollutant in the air changes radically during purging, the results will not
necessarily be close to the average over the time interval involved.
5-32
-------
Liquid Displacement
Another technique used in gas sampling is liquid displacement. In this method a
liquid is allowed to drain from the bottom of a container, while an opening at the
top allows the gas to enter and fill the space left by the liquid. Any suitable liquid
that will not dissolve the sample nor react with it can be used. The choice of liquid
will depend upon the material being sampled; some commonly used liquids are
water, brine, mercury, or water saturated with the gas to be sampled.
Containers used are of two basic types: (a) a glass tube with two stopcocks as
used in air displacement (see Figure 5-21) and (b) an aspirator bottle (see
Figure 5-22).
250 to 300 ml
Figure 5-21. Liquid displacement collector.
5-33
-------
Figure 5-22. Aspirator bottle.
In both cases, the liquid is allowed to drain through the lower opening (the rate
can be controlled by adjusting the stopcock) and the gas is drawn in through the
upper stopcock or tube. This method requires a minimum of equipment and no
special training. The container may be calibrated to indicate the volume of gas
sampled.
Inflation
A fourth gas sampling method is the collection of a sample by inflation of a plastic
bag (2). Plastics of various types have been used. The choice of material will
depend upon the gas to be sampled and the storage period.
Some hazards to look out for in "bag sampling" are: wall effects, memory effects
(where previous constituents linger), sample deterioration over time, sample
deterioration due to sunlight in some cases, and the possibility of reactions among
the various gases in the bag.
The deflated plastic bag is placed in a closed air tight box, with only a tube
extending outside the box. An opening in the box itself is connected to a vacuum
source, and the air is pumped out of the box. As the air is removed from the outer
container, the bag will inflate, drawing in the sample. The air may be metered as
it is pumped out of the box, thus indicating the volume of gas sample drawn into
the bag (see Figure 5-23).
Syringes
Syringes may be used in the collection of small gas samples. This technique has
been widely applied in the field of odor measurement.
5-34
-------
Blower
Blower opening
Bag valve
I Plastic bag
Figure 5-23. Inflation sampler.
Grab Sampling Techniques
Grab sampling techniques are preferable to continuous sampling in certain situa-
tions. Some constituents have absorption rates too slow for efficient collection by
absorption. Field conditions (lack of electricity and lab facilities) often necessitate
this type of sampling.
Grab sampling is useful when concentrations vary considerably over a period of
time, and it is necessary to obtain a sample at a specific time. Most grab sampling
techniques utilize a minimum of equipment and require little or no special training
or experience on the part of the operator (5).
Grab sampling has a serious limitation —the sample obtained is generally not
large enough to detect very small quantities of materials except by the most sen-
sitive techniques.
5-35
-------
References
1. Jacobs, M. B. The Analytical Chemistry of Industrial Poisons, Hazards and
Solvents. New York: Interscience Publishers, Inc., 1949.
2. Connor, William D., and Nader, J. S. Air Sampling with Plastic Bags. Amer.
Indus. Hyg. Assoc. J. 25: 291-297, May-June, 1964.
3. Altshuller, A. P.; Wartburg, A. F.; Cohen, I. R.; and Sleva, S. F. Storage of
Gases and Vapors in Plastic Bags. Int. J. Air Wat. Poll, 6: 75-81, 1962.
4. Silverman, Leslie. Industrial Air Sampling and Analysis. Philadelphia:
Industrial Hygiene Foundation, 1947.
5. Devorken, H.; Chass, R. L.; Fudurich, A. P.; and Ranter, C. V. Source
Testing Manual. Los Angeles: Air Pollution Control District, 1963.
5-36
-------
Principles of Freezout Sampling
Introduction
Air pollutants existing as gases can be trapped or removed by the freezout or con-
densation method. Trapping in this discussion refers to the mechanism of sample
collection, and removed implies an air-cleaning mechanism to rid unwanted gas
contaminants from the gas stream. The method has a very high efficiency at
relatively low flow rates. Certain problems are encountered when using the
freezeout method, thus necessitating an appraisal of the method for particular
applications.
Concept
The method consists essentially of drawing air through collection chambers with
progressively lower temperatures. If the temperatures of the chambers are approx-
imately equal to or less than the boiling point (the temperature at which a liquid is
converted to a gas) of the gaseous components of the air passing through it, these
components will exhibit a phase change from the gaseous phase to the liquid
phase. The condensate (liquid phase) is collected in the chamber where the phase
change occurs. The gaseous contaminants to be collected will determine the
temperatures required in the collection chambers. The temperatures of the
chambers can be controlled by using different immersion bath liquids. Con-
taminants with boiling points as low as - 195°C can be collected by this method.
Equipment
The type of freezeout equipment required depends to a large extent on the applica-
tion. The required amount of equipment of a given type depends on whether the
sampling apparatus is a single or multistage unit.
The size of the collection chamber varies according to the immersion bath for
which it was designed. The collection chambers themselves are placed in Dewar
flasks containing the cooling solutions (see Figure 5-24).
Collection chamber
Vacuum
Bath solution
Dewar flask
Figure 5-24. Freezeout unit.
5-37
-------
Figure 5-25 indicates various bath solutions and some sizes of the Dewar flasks
that have been used for each. The volume of the bath solutions and thus the size of
the collecting chamber itself are partially due to factors such as:
• Temperature gradients across the collecting chambers as related to the
criticality of the boiling point of the contaminant being collected;
• The surface area as related to the evaportation rate of the bath solution; and
• The condensation of water vapor in the primary collection chambers, thus
necessitating a larger volume.
Bath solution
Ice + salt
Dry ice & acetone or
methyl-cellosolve
Liquid oxygen
Liquid nitrogen
Temperature
-16°C
-80°C
-183°C
-195°C
Volume of solution
-2 liter
750 ml
100 ml
100 ml
Figure 5-25. Bath solutions.
The level of the solutions in the baths should be kept at 2" to 4" within the top
of the collection chambers in an attempt to maintain a constant temperature
throughout the chamber.
Among the collection chambers utilized, U-shaped and spiral-shaped tubes are
prominent. Large radius bends should be designed into the tubes to facilitate
smooth airflow and to prevent ice accumulation at the bends. Freezout devices can
be classified into two categories, single and multistage units.
Single-Stage Units
A single-stage unit (see Figure 5-24), consists of one collection chamber (glass or
metal) immersed in a bath solution. As has already been mentioned, the
temperature of the bath and consequently the liquid of the bath will depend on the
particular gas to be sampled.
Multistage Units
Multistage units consist of a series of collection chambers. These chambers can be
arranged in either horizontal or vertical trains (see Figure 5-26 and 5-27). In these
trains the temperatures of the baths are progressively lower. This allows for con-
densation of different gases in different chambers.
5-38
-------
Air intake
Dewar flask
Ice salt
Dry ice
and acetone
Dewar
flask
To flow meter
and pump
Liquid
nitrogen
Figure 5-26. Freezeout equipment for atmospheric samples (horizontal sampling train).
Air intake
Insulating boxes
Immersed trap
Condensing traps made
from Dewar blanks
Vacuum pump
Figure 5-27. Freezeout equipment for atmospheric samples (vertical sampling train).
5-39
-------
Efficiency
The collection efficiencies of the previously described systems are not very good. To
efficiently condense gases it is necessary for the gas to come in contact with the
cold surface of the collection chamber. Therefore, the efficiency of collection by
freezeout can be improved by: (a) filling the collection chamber with some type cf
material that will increase the cold surface area and (b) reducing the flow rate.
Packing Materials
To increase the cold surface area within the collection chamber, various materials
such as glass beads, metal packing (1) and activated carbon (2) have been used (see
Figure 5-28). In one application, for collecting benzene and formaldehyde, the
glass beads and metal packing increased the efficiency from 50% to 65% and
80%, respectively (1). The lower collection efficiency of the unpacked train was
due partly to the formation of a fine mist that was not retained by the walls of the
traps.
In another application using activated carbon, a collection efficiency of 100% was
reported for xenon and krypton (2). The activated carbon gave a much larger sur-
face area for the gas to pass over. The use of activated carbon will give the added
advantage of adsorbing gases from the air stream.
Collection chamber
Vacuum
Bath solution
Dewar flask
Figure 5-28. Freezeout unit showing packing material.
5-40
-------
Flow Rate
The flow rate through the train should be such that a sufficient "detention time"
(time allowed for the gas to come into equilibrium with its surrounding
temperature) be available to allow the desired collection efficiency. For an
unpacked train the detention time must be relatively large due to the small cold
surface area. By packing the train with a surface-area-increasing material the cold
surface area will increase and the detention time can become smaller. With a
smaller required detention time the rate of flow through the train can be greater.
Flow rates on the order of 0.1 to 0.2 cfm have been reported for unpacked trains,
while 1 to 2 cfm (2) have been reported for trains packed with activated carbon.
Another factor affecting flow rate is the formation of ice crystals in the bends of
the collection chambers. This will be discussed in another section.
Errors
One possible source of error is that gases soluble in water will be removed to some
extent prior to their removal in a collection chamber. Other errors may be
introduced when electrostatic precipitators, drying towers, etc., are placed ahead of
the freezeout train. Electrostatic precipitators will aid in the removal of par-
ticulates, but they may also alter the gas chemically (1). Adsorption of vapors by a
desiccant placed before the collection chamber has also been reported. This
adsorption might introduce errors in the final results.
Sensitivity
The sensitivity of the freezeout method depends primarily on the gas collected,
volume of air sampled, and how the collected gas is analyzed. Hydrocarbon
samples were analyzed on a mass spectrometer to detect pollutant concentrations of
10~4 ppm from a 1 liter sample. With larger sample volumes, concentrations on the
order of 10"6 ppm have been reported (3).
Applications
The freezeout method has proved useful in sampling gases. The freezeout device
can be used as a collecting train itself, or it can be used in conjunction with other
sample collection devices. Also, freezeout traps are used in the lab to concentrate
trace amounts of pollutants (such as halogenated hydrocarbons) desorbed from
adsorption cartridges in order to increase their detectibilities. This probably is its
chief use today.
Freezeout Train
Trains composed of several collection chambers have the ability of collecting
several gases at the same time. This may aid in the gross analysis of the sample
because the sample will be broken into fractional parts according to the various
boiling points of the gases.
5-41
-------
Probably the main disadvantage of a freezeout train is the plugging of the collec-
tion chambers by ice crystals. Drying towers placed on the inlet side of the train
will help alleviate this problem as well as filter some particulates. When drying
towers are used the flow rate is dependent upon the speed at which the desiccant
will help alleviate this problem as well as filter some particulates. When drying
towers are used the flow rate is dependent upon the speed at which the desiccant
will effectively remove the water moisture from the air. Flow rates of 1 to 2 cfm
have been reported when using a drying tower (2).
Liquid oxygen creates another problem when it is used as a bath solution or
when collected in a collection chamber. When used as a bath solution extreme care
must be employed because of the ability of liquid oxygen to support combustion.
Therefore, a restricted personnel area around the sampler must be maintained.
Oxygen (B.P. = - 183°C) will condense when liquid nitrogen (B.P. = - 196°C) is
used as a bath solution. This is undesirable since it will dilute the collected con-
taminants. If the solution is allowed to warm up after sampling, a portion of the
contaminants may be carried off by the escaping oxygen (1).
Multicollection Train
The freezeout train may be part of a larger train where particulate filters, elec-
trostatic precipitators, activated charcoal cartridges, etc., make up the rest of the
train. The major advantage of such a train would be the removal of particulates
and gases that were not of interest. Probably the main disadvantage of the larger
train is the possibility of altering the chemical composition of the gas of interest.
Summary
Freezeout trains have proved to be an efficient collection device. Collection efficien-
cies of 100% for flow rates up to 2 cfm have been reported for certain con-
taminants. Problems such as water vapor condensation with subsequent plugging of
collecting chambers can be alleviated by using a desiccant on the inlet side of the
train. Collection efficiency improvements such as increasing the cold surface area
can be accomplished by using a packing material. The use of freezeout devices for
"field" operations has its limitations because ot us bulkmess and the problem of
keeping the bath solutions at a constant level.
5-42
-------
References
Cadle, R. D.; Rolston, Myra; and MaGill, P. L. Cold-Surface Collection of
Volatile Atmospheric Contaminants. Analytical Chemistry 23: 475-477, March
1951.
Flygare, J. K., Jr.; Wehmann, George; Harbertson, A. R.; and Sill, C. W. A
Method for the Collection and Identification of Radioactive Xenon and Kryp-
ton. Sixth AEG Air Cleaning Conference, TID-7593, pp. 18-25, 1959.
Shepherd, M.; Rock, S. M.; Howard, R.; and Stormes, S. Isolation Identifica-
tion and Estimation of Gaseous Pollutants of Air. Analytical Chemistry 23:
1431 1440, October 1951.
Johns, vrcd B., Chief, Projects, Southwestern Radiological Health Laboratory.
Personal Communication.
5-43
-------
Chapter 6
Generation of Standard
Test Atmospheres
Introduction
Calibration of atmospheric sampling equipment is very important for air moni-
toring; and it must be done to ensure that the data generated by air monitors
represent the actual concentration of pollutants in air. Many factors may affect the
calibration of sampling or monitoring devices, preventing them from providing a
true measure of the atmospheric contaminant concentrations. The generation oi
standard test atmospheres is essential to the calibration procedures for continuous
air monitoring instrumentation.
There is a need for reliable, accurate methods to generate pollutant gases of
known concentration. For example, an air monitor may be designed to operate at a
certain efficiency, but due to factors such as reagent deterioration, electrical or
electronic component variability, and flow rate changes, data generated by an air
monitor may differ from the true concentration of a pollutant in air. To evaluate
the performance of sampling equipment, known contaminant concentrations must
be introduced; by knowing the input concentration, the output of the monitor can
be determined for accuracy and a function generated showing the relationship
between the input concentration and output instrument response. The purpose of
this section is to discuss the methods of preparing gases of known concentration.
The most important factor in the preparation of these standard atmospheres is
devising a method of preparing gases of known concentration. Moreover, in the
preparation of the standard atmospheres, devising a method of creating accurate
calibration gases at the extremely small concentrations typically found in the
atmosphere is often difficult. Many methods are now available for creating gases in
high concentrations, but •a-tmospheric testing often requires concentrations in the
sub part-per-million range.
Static Systems
Pressurized Systems
Although pressurized tanks of known contaminant concentrations are usually pur-
chased, they may be prepared by the following procedures. Cylinders containing a
pollutant gas are prepared by adding a known volume of pollutant gas and then
pressurizing the cylinder with a diluent gas. The gas is then of known concentra-
tion and can be used for calibration purposes. The range of concentrations that
6-1
-------
can be achieved by this method are typically less than 100 ppm to more than 5000
ppm, depending on the stability of the gaseous pollutant. The concentration of the
mixture can be calculated as follows:
10(JX V. 10h/7,
(Eq. 6-1) cppm= = --
or
(Eq. 6-2) c%=-
Vd+Vc
Where: cppm = concentration of gas mixture, ppm by volume
CM = concentration of gas mixture, percent
l/c — volume of contaminant gas
Vd = volume of diluent gases
p, = partial pressure of contaminant gas
p, = total pressure of the gas mixture
Using the rigid container procedure it is necessary to construct a gas-handling
manifold that interconnects the vacuum source, the calibrated volume, the source
of contaminant gas, and the gas cylinder. Subsequently, the entire system is
e\a< nated; the known calibration volume is flushed and filled at atmospheric
pressure with the contaminant gas and isolated, and connecting lines are again
evacuated. The contaminant gas is then swept by diluent carrier gas into the
cylinder; and the cylinder pressurized with diluent gas to the desired pressure.
Because of compressional heating the cylinder should be allowed to equilibrate at
room temperature before reading the pressure to be used in the concentration
calculations. The concentration of the mixture is calculated as follows:
106x F, xP,,
(Eq. 6-3) €„„„ =
V cvl X I 1
Where: cppm = concentration of mixture, ppm by volume
yc = volume of pure contaminant gas
Vcy! = volume of cylinder
Pfc = barometric pressure at time of filling
P, = final total pressure of cylinder
One factor that must be watched closely when preparing gas mixtures by this
method is the thoroughness of the mixing. When introducing the gases into the
cylinder one at a time, a layering effect may occur and result in incomplete
mixing. This effect can be counteracted by allowing for adequate mixing time
before use.
It should be noted that at room temperature and pressure most gas mixtures
conform closely to the ideal gas law. However, at the higher pressures that are
present in the cylinders, gaseous mixtures can deviate from this law and create er-
rors of up to 20%. This can be corrected by using a quantity called the com-
6-2
-------
pressibility factor (K). The units for the pressure and volume are not important as
long as Pc and P, are in the same units and Vc and Vd are in the same units since
both parts-per-million and percent are unitless quantities.
In commercial practice, compressed gas cylinders of calibration gases are often
prepared using high load mass balances: in this procedure, a precise tare weight
for an evacuated cylinder is obtained, and the cylinder is weighed again following
the addition of the desired trace constituent and after the addition of the diluent
gas (prepurified nitrogen in most cases) under pressure. The mass fraction of the
contaminant gas is converted to a volumetric concentration by application of the
usual formula involving molecular weight and molar volume.
When prepared by the user, a pressure dilution technique of some sort is
generally necessary. Here the volume of the contaminant gas introduced, Vc, the
volume of the cylinder, Vcyt, the evacuation pressure, Pvac, and the final total
pressure in the cylinder P, must be known.
Using this procedure the cylinder is evacuated using a vacuum pump capable of
producing a very low pressure. Depending upon the exact capacity of the pump,
the previous contents and size of the cylinder, and the final pressure to be used, the
evacuation pressure may be ignored in the calculations. For example, if the
cylinder was evacuated to a few torrs and the final pressure was 30 atmospheres the
calculation error in ignoring the evacuation pressure, i.e., assuming it is zero,
would only be about 1 part in 10,000.
As a matter of good practice, cylinders should be continued in the same service
and not interchanged; for example, a cylinder formerly used for SO2 span gas
should not be converted to NO2 service. Further, cylinder materials consistent with
the gases to be contained therein should be used.
Following the evacuation of the cylinder the contaminant gas is introduced using
a syiinge technique, as described in the bag procedure, or using a small rigid con-
tainer of precisely known volume.
Rigid Chambers
Rigid chambers such as the one illustrated in Figure 6-1 (1) are another method of
preparing an analytical gas standard.
The gaseous contaminant is introduced into the vessel, mixed with the diluent
gas, and sampled. The volume of the chamber may vary with the type of applica-
tion but the principle remains the same. Preparation of standard gas mixtures by
this procedure has been largely replaced by more accurate permeation devices and
pressurized cylinders containing a specified pollutant concentration.
6-3
-------
Contaminant inlet
Control valve
Diluent gas
Motor-driven stirrer
Rubber septum (for injection of pollutant)
Control valve
Gas mixture
Container of known volume
Figure 6-1. Rigid chamber used for producing standard gas mixture.
Nonrigid Chambers — Bag Samples
In this procedure, a bag, usually of a flexible nonreactive plastic material, is filled
with a known volume of diluent gas, and a known volume of contaminant gas is
added to the system. The diluent gas should be cleaned of all interfering consti-
tuents and be nonreactive to the desired contaminant. After allowing for complete
mixing in the bag, a sample can be drawn off for calibration purposes. The con-
centration of contaminant in the bag mixture can be calculated by (if the initial
contaminant concentration is 100%):
Where: c = concentration of diluted contaminant, ppm by volume
V c — volume of contaminant gas, \t$
Vd = volume of diluent gas, f
One of the first and most important steps in this preparation is the selection of
the bag. The bag must be of a flexible material and be chemically inert to the
gases it will contain. Chemical inertness is very important; if the contaminant gas
6-4
-------
reacts with the bag, the amount of contaminant will decrease and the actual con-
centration of the bag mixture will be unknown. Teflon, Mylar and copolymer
Tedlar bags have been widely used because they are inert to most materials.
Before any bag is to be used, it should be tested with the contaminant gas to be
sure that no reaction will take place between the bag and the gas. After the bag
has been selected, it should be checked for leaks before the sample is introduced;
particularly the seams of the bag and the area around the valve. The bag should
be flushed and evacuated at least three times to ensure that all unwanted con-
taminants have been removed. The actual filling of the bag must be done under
controlled conditions to guarantee measurement of the volume. Clean air is
pumped into the bag through an accurately calibrated flowmeter. The flow rate of
the diluent gas must be kept as constant as possible throughout, the filling pro-
cedure to obtain an accurate measure of volume. After the flow rate is set, a stop-
watch is used to get an accurate filling time. The product of the flow rate and the
filling time is the total volume of diluent gas added to the system.
The contaminant sample should be introduced into the stream of diluent gas as
the bag is filling. This should be done after the bag has filled about one-quarter of
the desired volume. By introducing the sample at this time, mixing can take place
as the bag fills to its final volume. The sample is introduced into the diluent gas
stream with the use of a syringe. A rubber septum and tee assembly is located in
the filling line for insertion of the syringe. Care must be used when handling the
syringe to ensure that the desired quantity of contaminant is introduced.
The syringe should have a graduated barrel so that the amount of contaminant
entered can be read directly. The plunger of the syringe must be gas tight so no
sample will escape while being injected into the diluent gas stream. A Teflon cap
can be fitted over the plunger to eliminate gas escaping during introduction. The
sample gas is extracted with the syringe from a source of known concentration.
When drawing the sample, the syringe should be filled and evacuated at least six
times with the sample gas to eliminate any air that may have been present in the
needle. The plunger of the syringe should be drawn well past the desired volume
when entering the sample. Just before introducing the sample into the diluent gas
stream, the syringe should be adjusted to the desired volume to eliminate any error
from air that may have diffused into the needle. Care should be taken in handling
the syringe; it should never be held by the barrel or the needle (2). Heat from the
hand of an analyst will cause the gas in the syringe to expand and part of the sam-
ple will be lost from the needle. When introducing the sample, the tip of the nee-
dle should be in the middle of the diluent gas stream to prevent the sample from
being lost on the walls of the filling lines. After depressing the plunger, remove the
needle from the gas stream immediately to ensure that none of the sample will
diffuse out of the needle and into the gas stream, giving a higher resulting concen-
tration. Mixing time can be decreased by kneading the bag for several minutes.
The concentration of the calibrated gas will change during storage. The decay
rate will depend upon the substance being stored, the relative humidity, and the
bag material. Substances such as nitrogen dioxide and ozone will decay faster than
carbon monoxide and hydrocarbons. Figure 6-2 (1) illustrates the rate of decay
6-5
-------
with time of a known concentration of SO2 stored in a bag. The decay rate can be
decreased if the bag is preconditioned. The preconditioning requires that the bag
be flushed several times with the sample gas. The bag should be left overnight at
least once with a sample gas in it as a preconditioning step.
.428 <
.424
.420
S .416
.2 .412
£
S
8 .408
I
O .404
t/5
.400
.396
.392
1 1
\
\o
X
\
1
100 200 300 400 500 600 700 800
Time, mins
Figure 6-2. Rate of decay of SO8 concentration in a bag.
It should be noted that the bag-filling method is only an approximate method of
preparing a known gas mixture concentration. A reference method should be used
to determine the actual concentration of the bag mixture.
Polasek et al. (2) found that sample stability depends clearly on the compound
and type of bag material used. Figure 6-3 gives the results of carbon monoxide
deterioration with time in bags of various materials.
6-6
-------
Bag material
No. of bags tested
Concn of calibration gas used
to fill bags, ppm
O h after filling, av ppm
Av deviation, ppm
Av sqd deviation, ppm2
24 h after filling, av ppm
Av deviation, ppm
Av sqd deviation, ppm2
48 h after filling, av ppm
Av deviation, ppm
Av sqd deviation, ppm2
100 h after filling, av ppm
Av deviation, ppm
Av sqd deviation, ppm2
pvc
10
9.0
8.9
- 0.12
0.056
8.5
- 0.50
0.306
8.4
- 0.63
0.497
7.9
- 1.2
1.5
Tedlar
10
9.0
8.5
- 0.5
0.352
7.5
- 1.5
4.2
6.8
- 2.2
7.7
5.2
- 3.8
17.8
Snout*
5
8.2
8.2
0
0.012
8.0
— 0.16
0.048
8.3
0.10
0.010
8.0
-0.18
0.066
Aluminized
polyester
3
8.2
8f\
.0
— 0.17
0.030
7.9
— 0.30
0.097
8.2
— 0.03
0.010
7.7
— 0.50
0.25
*Consists of layers of polyester, polyvinyl chloride, aluminum, polyamide, and polyethylene.
Figure 6-3. Carbon monoxide sample deterioration with time in bags of various materials.
Dynamic Systems
Permeation Systems
The use of permeation techniques for preparation of standard mixtures is very
useful for some contaminants. The method is based on the theory that a gas con-
fined above its liquified form at a constant temperature will permeate through
some materials at a constant rate. By putting a liquified gas into a Teflon tube, for
example, permeation of the vapor through the tube will take place because of the
concentration gradient that exists between the inner and outer tube walls. By
passing different flows of diluent gas over the tube, gases of varying concentration
can be generated.
The actual concentration of a sample gas can be calculated by Equation 6-5 (1).
(Eq. 6-5)
Where:
(PR)\(M)
'24.46/tf/M-wofeV T°K \/760 mm Hg^
/i-mo/e/V>98°/J\P mm Hg )
c =
f/min
c = concentration, pA/i or ppm by volume
T - temperature of the system, °K
P = pressure of the system, mm Hg
PR = permeation rate, fig/min
Q= total flow rate, liters/min
M- molecular weight of the permeating gas,
24 .46 = molar volume (V) of any gas at 25 °C & 760 mm Hg,
-mole
6-7
-------
Permeation tubes allow for the generation of gases with concentrations in the
sub-part-per-million range.
Permeation tubes are made from a variety of different materials. The material
must allow the diffusion of the contaminant gas through the walls and also be inert
to the diffusing gas. If some reaction took place between the tube material and the
gas, the permeation rate would be affected and might no longer be constant.
Teflon, Mylar, and Saran Wrap are materials often used because of their chemi-
cal inertness and good permeation properties. Before any material is used for
the permeation tube, it should be tested to ensure that no changes will occur in the
material when it comes in contact with liquified gas. To test the material, a piece
should be placed in some of the liquified gas it will contain. The material should
be removed after a few days and checked to see if any changes in the material have
occurred (i.e. brittleness, holes, stickiness, etc.). If there are no apparent changes,
the material is probably suitable for use.
The first step in the construction of the tube is to compress the desired sample
gas to a liquid state. The liquified gas is then put into a tube and the tube ends
are sealed. One method of sealing the tube ends is to force glass beads or stainless
steel balls into the tube ends. To seal properly, these beads should be approxi-
mately one and one-half times the inside diameter of the tubes. Once the tubes
have been prepared, they should be stored for two to three days in order to
equilibrate. Since the permeation rate is extremely dependent upon temperature
and relative humidity, the permeation tubes should be stored at a constant
temperature and zero humidity. After the waiting period has ended, the tubes
should be weighed on an analytical balance and replaced in the storage area. Time
required to weigh the tube should be minimized and kept constant to compensate
for the effects of moisture absorption. High humidity will cause the permeation
tube to absorb moisture thereby increasing the tube weight. This will yield an
erroneously low value for the permeation rate. Absorbed moisture on the tube can
form acids that may cause tube blistering, thus changing the permeation rate. The
tube should be stored and weighed several times to yield enough data to
demonstrate that the permeation rate is constant. The results of these weighings
should be plotted on a graph as weight versus time.
From the slope of the resulting best "fit" line, the permeation rate can be
calculated in micrograms per minute as shown in Figure 6-4. The permeation rate
for a particular material and sample is usually given as nanograms per minute per
centimeter of the tube length.
-------
r
I
4_i
s
a
w
VM
o
4_>
.£!
Slope = rate of permeation
492
I
I
1
a
I
BC
'C
0 100 200 300 400 500 600 700
Time, hours
Figure 6-4. Calibration of two permeation tubes.
Once the tube has been calibrated, it can be used to generate test gases of
known concentration. The permeation tube is placed in a stream of diluent gas.
The gas passes over the tube and the permeated gas is mixed into the gas stream.
I tic desired concentration can be varied by varying the flow rate of the diluent gas
oi by varying the permeation tube length. The diluent gas must be kept at a
constant temperature during the time the calibration gas is being generated to be
sure the permeation rate is constant. The temperature dependence of the permea-
tion rate is illustrated for four gases in Figure 6-5. To accomplish this, the diluent
gas is drawn through a constant temperature chamber before passing over the
tube as in Figure 6-6. Permeation tubes are commercially available from many
sources offering a variety of precalibrated tubes with different permeation rates.
The National Bureau of Standards offers some reference sources. Figure 6-7 lists
permeation rates for a number of compounds through a Teflon film. A number of
configurations other than the original tube design are also commercially available.
Some of these are designed to provide a longer useful life. Figure 6-8 lists some
materials used to construct a permeation tube.
(.-9
-------
a
1
^
a
u
rt
u
i-i
u
Pk
Scale right
(/ig/cm/min) .
10
25 20 15
Temperature, °C
Figure 6-5. Permeation rate vs. temperature for four gases.
Flow meter
or dry test
meter
• Clean dry air
Flow meter
I or
critical orifice
Purified air
or cylinder
nitrogen
Water
pump
Vent
Figure 6-6. Components and flow of a typical permeation system.
6 10
-------
.
• /
I . 1
; IK
, < . ' .
C,,H,(:H<
• , ukucss
Mil >
[1
1 1
•
12
0 ill Li
') )•!.'
' (M .'.
,, , i j
' i ' i J
<' ;2
• :
-i mi.
1 1 , i . i
t) 01 h
0.030
Temperature
(°C)
,-t) (1.5
JO • 0.5
•20.1
29.1
!3.8
Jl.l
L'9.1
.'1.1
'."I 1
; .)
. '). !
JO
93
•1
M
20
Permeation
rate
(ng/cm/min)
213
138
203
396
605
1110
2290
53
119
6.4
22.3
2.8
1.3
0.29
0.065
0.00006
Figure 6-7. Permeation rates of some typical compounds through FEP Teflon.
,,..., ,l,l,v ,
• i U1
a i <
h'l: indc
i , phi .ijl.itc
1 rude name
i.i 1 d'lun
Alatiion
Saran Wrap
N'vlon 6
.\lvlar
Mylar
Diothene
. .ncrials used c> << nstruci a permeation tube.
The performance of a permeation device depends on the polymer films used to
construct these tubes and the pollutant for which a standard concentration is
needed. The important factors to be considered in the use of a permeation device
are temperature, humidity, gas stability, equilibration time, etc. These parameters
have been studied for nitrogen dioxide (3), sulfur dioxide (5), and recently for
numerous halogenated hydrocarbons (5), and recently for permeation tubes con-
structed with FEP (fluorinated ethylene propylene copolymer) Teflon. In the case
of nitrogen dioxide, the permeation device demonstrating the greatest stability and
utility in terms of equilibration time and lifetime is illustrated in Figure 6-9.
6-11
-------
Teflon plug
Permeating area
L
Glass tube •
Stainless steel bands
Teflon tube
Liquid nitrogen dioxide
Figure 6-9. Nitrogen dioxide permeation device.
Expt: mu ntal results indicated that a small permeating area of Teflon was
requin a tor NO, due to the rapid permeation rate of NO2 through the Teflon
wall. 1 ubes constructed as in Figure 6-9 have a useable lifetime of up to 2 years
(3). In contrast, the permeation rate of sulfur dioxide through the same Teflon
wall is considerably slower. Permeation tubes containing this pollutant are con-
structed as shown in Figure 6-10.
Teflon tube
Retaining collar
(indicating tube
I.D.#"33-64")
\
Retaining collar
(indicating tube
contents, "SO2")
Figure 6-10. SO, permeation tube (NBS—standard reference material).
b 12
-------
To determine the rate of permeation in this type of device, the tube may simply
be removed from the permeation system and weighed to the nearest 0.1 mg on an
analytical balance. Generally weighings can be made daily, weekly, or monthly
depending on the gas and type of permeation device. As indicated previously, the
permeation rate can be determined by measuring the slope of the least-squares
error line used to fit the weight vs. time data. A more rapid calibration, however,
can be obtained by using the apparatus shown in Figure 6-11.
Weighing unit
Purge air.
Control unit
\
r r
Constant
temperature
bath ±0. 1C
•^—
1
— i
r
Recorder
Temperature
monitor
Figure 6-11. Gravimetric calibration apparatus.
In addition to rapid calibration, this apparatus has the advantage of continuous
direct read-out of weight change (see Figure 6-12), and the permeation tube is
maintained in the constant temperature bath at all times. The potential problem
of spurious weight increase due to the hygroscopic nature of some pollutants (e.g.,
SO2, NO2) is eliminated with this device for the tube need never have to leave the
constant temperature bath.
fi 13
-------
-0800
-0700
-0600
240
0
-0500
-0400
240 240
mm
12/3/70 SO2 #24 temp. 25°C, range 0-1 Mg F.S.
0300. i I . i I I \ L
20 30 40 50 X.60 70 80 90 100
705
Figure 6-12. Typical stripchart read-out from an in situ gravimetric apparatus.
Dilution Systems
Single Dilution
A simple system for diluting gas entails mixing the gas with the diluent in a mixing
chamber after measuring their flow rates. A single dilution system using rotameters
to measure flow is illustrated in Figure 6-13 (1). The concentration of the test mix-
ture can be calculated with the following formula:
(Eq. 6-6)
=c
"
Where: cu = concentration of undiluted contaminant gas
cd = concentration of diluted test mixture
Qu -flow of undiluted contaminant gas
Q_d =flow of diluent gas
Flow measuring devices other than rotameters are frequently used to increase the
accuracy and precision of the test mixture concentration.
-------
Mixing chamber
Test mixture
Control valve
Component Diluent
gas gas
Figure 6-13. Single dilution system.
Multiple Dilution
Low-level concentrations of pollutants may be obtained by a multiple dilution
system. A system such as that shown in Figure 6-14 for a double dilution avoids
high dilution ratios and the use of low flow rotameters. Other methods for measur-
ing and controlling flow (i.e., orifices, mass flowmeters, etc.) in a multiple dilution
system provide a variety of means for constructing multiple dilution systems. The
test mixture concentration can be calculated with the following equation:
(Eq. 6-7)
Where (see Figure 6-14):
CuQu
ft.
C7~r
tm
Qu -flou} of undiluted contaminant gas
Qdi -flow of first diluent gas
(?m =flou> of first diluent mixture
QdZ—flou] of second diluent gas
cu = concentration of undiluted contaminant gas
c,m = concentration final test mixture
6-15
-------
High-concentration
waste gas
Manometer
Control valves
Qu
Qdi
Flowmeter
I/
Control valve
t
Test mixture
Qd2
Contaminant Diluent
gas gas
Diluent gas
1st 2nd
Figure 6-14. Sketch of a system for making double dilutions.
Ozone Generation
Ozone is unlike most of the gaseous pollutants in that there are no gaseous calibra-
tion standards available. This is due to ozone's instability, which makes it impossi-
ble to produce gas cylinders of standardized ozone concentrations or ozone permea-
tion tubes. The only means available for the calibration of ozone monitors is to
produce stable, known amounts of ozone at the site of calibration.
Ozone is most commonly produced by irradiating oxygen with an ultraviolet light
source in an ozone generator. An ozone generator may utilize a tubular quartz
chamber into which ozone-free air is admitted at a controlled, constant rate. This
incoming air (from a zero air source) is subjected to ultraviolet radiation from a
mercury vapor lamp.
The mercury vapor lamp is housed within an aluminum shield which can be
adjusted to expose a selected portion of the lamp, thereby controlling the amount
of ultraviolet radiation to which the air stream is exposed. In this way varying
amounts of ozone can be produced by simply moving the shield to a number of
positions across the mercury lamp. Using a pressure regulator to assure a stable
flow through the generator plus a constant line source transformer to assure cons-
tant voltage to the UV lamp, the ozone generator can be expected to give stable
ozone concentrations which vary by no more than ±10 parts per billion.
6-16
-------
Varying concentrations of ozone can also be obtained by dilution of a stable
ozone source (from an ozone generator) with varying amounts of zero air (ozone-
free air). One or more of the dilution techniques previously mentioned can be
employed.
Evaporative Systems
Gases may be introduced into a test atmosphere by injection of the gas in the
liquid phase into the moving diluent gas stream. Figures 6-15 through 6-17
exemplify several types of systems for introducing a liquified gas into a system that
will produce an atmosphere of known concentrations.
Careful consideration should be given to the specific pollutant and the type of
injection used. For example, syringe injection of nitrogen dioxide into a bag
through a rubber septum is unadvisable since nitrogen dioxide will react with the
rubber giving rise to spurious concentrations.
The approximate concentration of the test mixture can be calculated with the
following formula.
(Eq. 6-8) c =
Where: c = concentration of test mixture, in ppm
Qj=flou> of liquid contaminant, in mUmin
Q(= density of liquid, in g/ ml
V= molar volume of ideal gas at operating temperature and pressure
in Vmole
M= molecular weight of contaminant, in g/gm-mole
Qd =flou> of diluent gas, in V/min.
106 = conversion to ppm
Teflon needle
Septum
Liquid/gas
Diluent gas
Test mixture
Figure 6-15. Method for injecting liquids and gases into a test atmosphere.
6-17
-------
Diluent gas
Excess liquid
Test mixture
Stainless steel needle
Liquid
Figure 6-16. Method for injecting liquids into a test atmosphere.
Test mixture
t t I
t t
Diluent gas
Liquid
Figure 6-17. Method for injecting liquids into a test atmosphere.
6-18
-------
References
1. Nelson, G. O. Controlled Test Atmospheres. Ann Arbor, ML: Ann Arbor
Science Publishers, Inc., 1972.
2. Polasek, J. C.; Bullin, J. A.; Evaluation of Bag Sequential Bag Sampling
Technique for Ambient Air Analysis. Environ. Set. Technol. 12: 709, 1978.
3. Hughs, E. E., et al., Performance of a Nitrogen Dioxide Permeation Device.
Anal. Chem. 49: 1823, 1977.
4. Scaringelli, F. P.; O'Keeffe, A. E.; Rosenberg, E.; and Bell, J. P. Preparation
of Known Concentrations of Gases and Vapors with Permeation Devices
Calibrated Gravimetrically. Anal. Chem. 42: 871, 1970.
5. Purdue. L. J., and Thompson, R. J. A Rapid, Sensitive Method for Calibra-
tion of Permeation Devices. Anal. Chem. 44: 6, 1972.
6-19
-------
Preparation of Zero Air
Introduction
Zero air can be defined as air that is free of contaminants and interferences for a
particular analysis technique (9). It is important to note that a zero air for one
analysis may not be a zero air for another analysis. For example, an analysis
technique may require a zero air that has only sulfur dioxide and water vapor
removed while another analysis technique may call for only hydrocarbons to be
removed.
Zero gases are used extensively in atmospheric sampling both in laboratory and
field applications. Many continuous monitors require a constant supply of zero air
for parallel path reference cells. Cylinder air can be used, but by applying the
proper adsorption or absorption column on the inlet gas stream, zero gas can be
continuously produced from atmospheric air. This can cut costs and eliminate the
necessity of replenishing zero air cylinders. The contaminants and their typical con-
centrations found in clean, dry ambient air are summarized in Figure 6-18.
Component
Nitrogen
Oxygen
Argon
Carbon dioxide
Neon
Helium
Methane
Krypton
Hydrogen
Nitrous oxide
Xenon
Ozone*
Ammonia
Iodine
Nitrogen dioxide
Sulfur dioxide
Carbon monoxide
Formula
N2
02
Ar
C02
Ne
He
CH4
Kr
H2
N2O
Xe
o,
NH3
I*
NO2
SO2
CO
Content
% by Vol. ppm
78.09
20.94
0.93
0.033
18X1Q-"
5.2X10'4
1.5X10'"
i.oxio-"
0.5X1Q-"
0.5X1Q-"
0.08X10-"
0.03x10-*
0. 01x10'"
0.01x10-"
0. OOlXlO'4
0.0002X10-"
0 to trace
780,900
209,400
9,300
330
18
5.2
1.5
1.0
0.5
0.5
0.08
0.03
0.01
0.01
0.001
0.0002
~
*Ozone content in winter is 0.02 ppm and in summer is 0.03.
Figure 6-18. typical composition of clean, dry air near sea level.
There are many techniques for producing zero air. The purpose of this section is
to discuss the most widely used methods for removing contaminants from a gas
stream to produce a zero air. It should be noted that water vapor is a contaminant
that must be removed from many gas streams. This is especially true when
introducing a gas stream into a permeation tube system where the gas must be free
from water vapor. This ensures proper permeation through the walls of the
permeation tube (11, 12). For this reason, a separate discussion of drying a gas
stream is provided in this section.
6-20
-------
Gaseous Contaminant Removal
* -f YY\ d o~n ^ sf T'^sm
There are many ways of removing a gaseous ™™™™™ir ^ 6.19 lists
including the use of catalytic devices, adsorption, and absorption, g
the more commonly used materials.
Material
1. Activated charcoal
(6x16 mesh)
Purpose
Removes many gases such as
ozone, S02, NO2l higher molecu-
lar weight organic vapors. Does
not remove CO, CO2.
NO oxidizer Converts NO to NO2 for subse-
(a) Chromium trioxide (CrO3) quent removal with TEA or soda
v ' lime1'.
(b) Humidifier
Furnishes water vapor for proper
operation of oxidizer.
Preparation
Commercially activated coconut
hull charcoal 6X 16 mesh." Place
in 2 inch I.D. X 18 inch long sec-
tion of plastic pipe. Use glass
wool plugs to retain charcoal in
place. Plastic pipe caps on each
end of cylinder are drilled and
threaded to accept standard Vi
inch O.D. tube fittings. Con-
tainers of other materials and
similar configuration and volume
may be used.
Soak firebrick or alumina 15-40
mesh in a solution containing
16 g CrO3 in 100 ml of water.
Drain, dry in an oven at
105-115 °C for 30 to 60 minutes
and cool. Spread a thin layer of
the dry pellets in a dish and
place in a desiccator containing
a saturated salt solution which
maintains the relative humidity (RH)
between 50 to 70%. The reddish
color changes to a golden orange
when equilibrated. Place in a ft
inch I.D. X 15 inch glass tube.
Use glass wool plugs to hold
pellets in place. (Caution: pro-
tect eyes and skin when handling
this material. Do not breathe
oxidizer dust.) Before using, pass
air containing 30 to 70% RH
through the oxidizer for about 1
hour at 0.5 f/min to condition.
Discard when more than % ot
the oxidizer bed depth turns
brown.
Pass dilute NO air stream over
water at a fixed temperature
such that the humidity of the air
stream is maintained within
50 ±20% RH
6-21
-------
3. Triethanolamine (TEA)
Removes NO,
4. Desiccant
Remove water
5. Carbon monoxide
Oxidizer (Hopcalite)
Catalytically oxidize CO to CO2
for subsequent removal with
Ascarite or soda lime.
6. CO2 Absorber
Remove CO,, H,O
Soak Firebrick 10-20 mesh in a
20% aqueous solution of TEA.
Drain, spread on a dish and dry
for 30 to 60 min at 95 °C. Pellets
should be free-flowing. Place in
a V2 inch O.D. (7/16 inch
I.D.)x 15 inch long polyethylene
or stainless steel tube with stan-
dard V? inch O.D. tube fittings
on each end. Use glass wool
plugs to hold the pellets in place.
Commercial 6x 16 mesh silica
gel with color indicator.* Place
in 3 inch I.D. X 24 inch long
clear plastic cylinder capped at
both ends. Caps are drilled and
threaded to' accommodate stan-
dard yz inch O.D. tube fittings.
Use glass wool to hold granules
in place. When the change in
color exceeds % of the desiccant
bed depth, regenerate by expos-
ing the silica gel to 120°C
atmosphere overnight.
Commercial mixture of copper
and manganese oxides'. Place
granules in a i/i> inch O.D. (7/16
inch I.D.)x 15 inch long section
of copper pipe with standard i/£
inch O.D. tube fittings on each
end. Use glass wool to hold
granules in place.
a. Soda lime'': commercial
preparation calcium and
sodium hydroxides; 4 to
8 mesh.
b. Ascarite": a commercial
preparation of sodium
hydroxide in an asbestos
matrix, 8-30 mesh.
Place soda lime or ascarite in a 2
inch I.D. X 15 inch long section
of plastic pipe. Plastic caps at
both ends are drilled and
threaded to accept standard i/£
inch O.D. tube fittings.
"Activated coconut hull charcoal, Type PCB 6x 16 mesh, Pittsburgh Activated Charcoal,
Merk & Co., Inc. Pittsburgh, PA 15230
'Silica gel, indicating, 6x 16 mesh, Grace Davison Chemical, Baltimore, MD 21226
'Carbon monoxide purifier, (Hopcalite) Model RAF-BCHDI, Robbins, Aviation, Inc.,
Vernon, CA 90058
"Soda lime, 4-8 mesh, J. T. Baker Chemical Co., Phillipsburg, NJ 08865
" Ascarite, 8-30 mesh, ibid.
Figure 6-19. Materials used in producing zero air up to 30 (/rain.
6-22
-------
Catalytic Devices
Catalytic devices have been used to selectively remove a gas from the sample
stream. This leaves a reference or zero gas that is minus only the pollutant to be
measured. Using this method, the change in output when the catalyst bed is
bypassed will be due only to the pollutant measured. An example of a specialized
catalytic conversion is the UV absorption ozone monitor. Ozone is selectively
removed from the sample using a manganese dioxide catalyst that reduces ozone to
oxygen (1, 2). This ozone-free sample, which still contains all other gases, is the
reference or zero air for the UV ozone monitor and is used as the baseline. The gas
stream then bypasses the catalyst, and the ozone-carrying gas enters the monitor
where the change in output can be attributed specifically to the ozone.
Adsorption
Adsorption is a widely used method for removing contaminants from a gas stream
to form a reference or zero air. Solid adsorbents have the ability to adsorb quan-
tities of gases because they have extremely high surface areas per unit weight, e.g.,
activated carbon has a surface area ranging from 300 to 1400 mVgm (18). A more
detailed discussion of the porous quality of adsorbents may be found in the
literature (3, 8, 18). Figure 6-20 lists the properties of some common adsorbents.
Adsorbent
Activated
carbon
Silica gel
Activated
alumina
Molecular
sieves
Form
Pellets
Beads (G)
Beads (G)
Beads (S)
Beads (G)
Beads (S)
Pellets
Beads (G)
Beads (S)
External
surface
Area, ftVlb
10.5-21.5
15.0-24.0
5.0-16.0
6.0
7.0-18.5
4.0- 8.0
9.0-14.5
32.0
7.5-12.5
Pore
volume
ftVlb
0.010-0.013
0.007
0.006
/
/
/
Reactivation
Temp, °F
200-1000
200-1000
250-450
300-450
350-600
350-1000
300-600
300-600
300-600
Max. gas
flow,
CFH/lb
/
/
75
75
50
50
/
/
/
Sp. heat,
c,
Btu/(lb°R)
0.25
0.25
0.22
0.25
0.22
0.25
0.23
0.23
0.23
Typical
Adsorbates
CH4 through
n-C5H12, CO2, H2S
CH4 through
C^Hio, C2H4
through C4Ha.
H2O, H2S, SO2
H2O, H2S
Oil vapors
See Figs. 6-22
6-2S
6-24
G = Granules
S = Spheroids
Figure 6-20. Typical properties of adsorbents.
Activated carbon and molecular sieves are the most widely used solid adsorbents
for the removal of contaminants. Activated carbon has been used extensively for
the adsorption of many contaminant gases in a sample stream not only in
atmospheric sampling and in laboratory use but also as an industrial adsorbent in
continuous flowing adsorption towers (16). The term "activated" carbon derives
from a method of enhancing the adsorption properties of regular carbon. The car-
bon is "activated" by heating at 900 °C in a reducing atmosphere to increase the
porous nature of the carbon thereby increasing the adsorbency of the carbon (4).
6-23
-------
Activated carbon can be made from many substances including soft coal, fruit pits
nut shells, and coconut shells. Coconut shell carbon is the desired form of activated
carbon because of its porous nature. Figure 6-21 lists the adsorptive capacity of
activated carbon for several gases.
Gas
SO2
CH3C1
NH3
H2S
HC1
C2H2
CO2
CH4
CO
02
N2
H2
Volume adsorbed, cm3
380
277
181
99
72
49
48
16
9
8
8
5
——^——^^^—^^^—^^"
Figure 6-21. Adsorption of gases by carbon (1 gm of adsorbent, temperature 15°C).
Molecular sieves are also widely used as adsorbents because of their porous
nature The molecular sieves are usually made from synthetic zeolite crystals or
metal aluminosilicates. The diameter of the pores or passageways of the molecular
sieve regulates the size of the molecules that may pass through the sieve. Because of
this molecular sieves have been used extensively in the fractionation of organic
gases (hydrocarbons) as packing material in gas chromatographic columns. Various
properties of molecular sieves are summarized in Figures 6-22, 6-23, and 6-^4.
Adsorbed on 4A & 5A
molecular sieve
Water
CO2
CO*
H2S
SO2
NH3
NO2*
02*
CH4*
Methanol
Ethane
Ethylene
Acetylene
Propylene
n-Propanol
Ethylene Oxide
Adsorbed on 5A but not
4A molecular sieve
Propane and higher
n-paraffins to Ci4
Butene and higher n-olefins
n-Butanol and higher
n- alcohols
Cyclopropane
Freon 12
Not adsorbed on 5A or 4A
molecular sieve
Isobutane and all
iso-paraffins
Isopropanol and all iso,
sec and tert-alcohols
Benzene and all aromatics
Cyclohexane and all cyclics
with at least 4 numbered
rings
Carbon Tetrachloride
Sulfur Hexafluoride
Hexachlorobutadiene
Freon 114 and 11
Boron Trifluoride
Molecules larger than 5 A
* Adsorbed below -20°F
Figure 6-22. Molecular sieve adsorption characteristics.
6-24
-------
3A
4A
5A
10X
13X
Nominal pore
diameter, /*
0.003
0.004
0.005
0.008
0.010
Molecules adsorbed*
< 3 A effective diameter
(e.g. H20, NH3)
<4 A diameter (e.g. H2S,
Ethanol, CO2, SO2> C2H4
& C2H6)
< 5 A diameter (e.g.
n-C4H9-OH
n-C4Hio, CaHg to C22H46)
<8 A diameter (e.g. iso-
paraffin & olefins, C6H6)
< 10 A diameter
Remarks
Used for drying
and dehydration
Scavenge water
from solvents and
sat. hydrocarbons (HC).
Separates n-paraffins
from branched &
cyclic HC.
Separates aromatic HC.
Drying, H2S & mercap-
tan removal (gas
sweetening)
*Each type adsorbs listed molecules plus those of preceding types
Figure 6-23. Molecular sieves—Linde type.
Zeolite
4A
5A
B
X
Y
Erionite
Offretite
Mordentite
Amount sorbed @ 25 °C
(g/lOOg of molecular sieve)
HZO
24.5
24.5
20.0
31.5
28.0
11.3
16.6
13.3
n-C6HM
—
12.0
—
16.8
16.5
4.4
8.6
6.1
Cyclohexane
—
—
—
18.5
19.0
6.7
5.3
7.3
Figure 6-24. Effective sorption capacities of molecular sieves.
Acid gases in the atmosphere can prove to be interferences for analytical and
continuous monitoring techniques. An often used adsorbent for these gases is
Ascarite. This strongly basic adsorbent, 91% NaOH, will remove acid gases from
the sample stream. Ascarite has also been used extensively for selectively removing
carbon dioxide from a gas stream (17).
Absorption
Absorption is also a means of removing unwanted gases from the sample stream.
One absorptive method could be much like the sampling of gases through
impingers: the contaminant could be scrubbed from the system using a liquid solu-
tion. Solid absorbents are also used. Levaggi et al. (7) discuss a method of
absorbing nitrogen dioxide from a gas stream by using triethanolamine on
firebrick. This will selectively absorb nitrogen dioxide while allowing the passage of
nitric oxide.
6-25
-------
Water Vapor Removal
As mentioned previously, water vapor removal is very important in the preparation
of calibration gases using permeation tubes. Water vapor removal is also important
when using some catalysts or adsorbents. For example, water vapor must be
removed before passing a gas stream over a bed of Hopcalite, because water vapo
causes Hopcalite to lose its oxidizing properties (19). There are many other situa-
tions where water vapor must be removed. The three most widely used laboratory
and air sampling methods of removing water vapor (drying) from a gas stream are
adsorption, absorption, and condensation.
Adsorption
Adsorption of water vapor on solid desiccants is the most common method of
drying a gas stream. This is because solid desiccants are readily available, easy to
handle and store, can be regenerated, and can be prepared with an indicator
material in them that changes color when the desiccant is spent.
The choice of a drying agent should not be based solely on its drying ability.
Other factors, such as stability, temperature dependency, ability to perform in high
humidity situations, emission of gases through reaction with the moisture, etc.
should be taken into account.
The three most widely used drying adsorbents are silica gel, calcium sulfate, and
anhydrous magnesium perchlorate. Efficiency and capacities of these and other
solid desiccants are listed in Figures 6-25 and 6-26.
Material
CuSO4 (anhy)
CaCl2 (gran.)
CaCl2 (tech. anhy.)
ZnCl2 (sticks)
Ba(ClO4)2 (anhy.)
NaOH (sticks)
CaCl2 (anhy.)
Mg(C104)2.3H20
Silica gel
KOH (sticks)
A12O3
CaSO4 (anhy.)
CaO
Mg(C104)2 (anhy.)
BaO
Volume of air
per hr. per ml.
desiccant, ml
36 to 50
66 to 165
115 to 150
120 to 335
26 to 36
75 to 170
75 to 240
65 to 160
65 to 135
55 to 65
65 to 135
75 to 150
60 to 90
95 to 130
64 to 66
Total vol. of air
per ml. of
desiccant, liters
0.45 to 0.7
6.1 to 24.2
4.0 to 5.8
0.8 to 2.1
2.3 to 3.7
2.3 to 8.9
1.2 to 7.8
4.0 to 7.2
6.5 to 7.7
3.2 to 7.2
6.5 to 7.7
1.2 to 18.5
7.6 to 10.1
6.4 to 13.2
10.6 to 25.0
Residual water
vapor per liter of
air, mg
2.8 (2.7 to 2.9)
1.5 (1.4 to 1.6)
1.25 (1.23-1.27)
0.98 (0.94-1.02)
0.82 (0.76-0.88)
0.80 (0.78-0.83)
0.36 (0.33-0.38)
0.031 (0.028-0.033)
0.03 (0.02-0.04)
0.014 (0.010-0.017)
0.005 (0.004-0.009)
0.005 (0.004-0.006)
0.003 (0.003-0.004)
0.002 (0.002-0.003)
0.00065 (0.006-0.0008)
Figure 6-25. Comparative efficiency of various drying agents.
6-26
-------
Desiccant
Sodium hydroxide?
Anhydrous barium
perchlorate
Calcium oxide
Magnesium oxide
Potassium hydroxides
MekohbiteS
Anhydrous magnes-
ium perchlorated
Anhydroned> f
Barium oxide
Activated alumina
Phosphorus
pentoxideS
Molecular sieve 5Af
Indicating anhydrous
magnesium
perchlorated
Anhydrous lithium
perchlorateS
Anhydrous calcium
chlorideJ
Drieritef
Silica gel
Ascaritef
Calcium chlorides
Anhydrous calcium
chlorides
Anhydrocelf
Initial composition
NaOH«0.03H,O
Ba (C1O4)2
CaO
MgO
KOH«O.52H2O
68.7% NaOH
Mg(ClO4)2.0.12H20
Mg(ClO4)2.1.48H2O
96.2% BaO
A120,
P205
Calcium aluminum
silicate
88% Mg (C1O4)2 and
0.86% KMn04
LiC104
CaCl2.O.18H2O
CaSO4.O.O2H2O
—
91.O%NaOH
CaCl2.O.28H2O
CaCl2
CaSO4.O.21H2O
Regeneration requirements
Drying time
(hr)
—
16
6
6
—
—
48e
—
—
6 to 8
—
—
48e
12e, 12
16e
1 to 2
12
—
—
16e
1 to 2
Temperature
(°C)
—
127
500, 900
800
—
—
245e
240
1000
175,400
—
—
240e
70e, 110
127e
200 to 225
118 to 127
—
200e
245e
200 to 225
Average
efficiency11
(mg/liter)
0.513
0.599
0.656
0.753
0.939
1.378
0.0002
0.0015
0.0028
0.0029
0.0035
0.0039
0.0044
0.013
0.067
0.067
0.070
0.093
0.099
0.137
0.207
Relative
capacity0
(liters)
178
28
51
22
18.4
68
1168
1157
244
263
566
215
435
267
33
232
317
44
57
31
683
aNitrogen at an average flow rate of 225 ml/min was passed through a drying train consisting of
three Swartz drying tubes (14 mm i.d. by 150 mm deep) maintained at 25°C.
hThe average ampunt of water remaining in the nitrogen after it was dried to equilibrium.
1 The average maximum volume of nitrogen dried at the specified efficiency for a given volume of
desiccant.
''Hygroscopic.
''Dried in a vacuum.
'Trade name.
^Deliquescent.
Figure 6-26. Comparative efficiencies and capacities of various solid desiccants in drying a
stream of nitrogen.3
Silica gel is easy to handle, and it can readily and indefinitely be regenerated at
temperatures near 120 °C (4, 10). Attempting to regenerate silica gel above 260 °C
will cause loss of the water vapor extractive properties. More often, a series of
adsorbents are used for drying a gas stream. A popular series method for water
vapor removal is silica gel followed by molecular sieve. The silica gel, which can be
regenerated easily, removes a major portion of the water vapor; then the molecular
sieve, which is a more efficient drying agent, removes most of the remaining water
vapor.
6-27
-------
Calcium sulfate also has excellent regenerative capabilities (1-2 hours at 200 °C);
however, unlike silica gel it will gradually lose its drying properties because of the
destruction and reformation of the dehydration elements. Calcium sulfate is also
able to operate at a constant efficiency over a wide range of temperatures. Drierite
and Anhydrocel are trade names for commercially available calcium sulfate mix-
tures, and, as with silica gel, these adsorbents are available in indicating and non-
indicating forms. Anhydrous magnesium perchlorate has the best drying efficiency
of the compounds named, but it has a certain drawback: explosive compounds may
be formed if the regeneration step occurs in the presence of organic vapors. For
this reason, hydrocarbons must be removed from the gas stream before regenera-
tion (10). Anhydrous magnfesium perchlorate is also deliquescent, i.e., it will melt
when removing moisture from the air.
Absorption
Absorption is another method of drying a gas stream. Absorption, usually with
liquid desiccants, is not as efficient as with solid desiccants, but it has a higher
drying capacity because the liquid can be constantly recirculating. This process
with liquid desiccants takes place in much the same way as in a scrubbing tower:
the gas comes into contact with the liquid and the water vapor is absorbed. Strong
acids and bases are good liquid desiccants, but they will emit corrosive vapors.
Condensation
Drying gases by condensation (cooling) is an excellent method for some purposes;
all that is required is that the gas be cooled below its dewpoint, thereby removing
the water vapor from the gas stream (5). The process is quite simple; the sample
gas enters a vessel and is cooled. When the gas has been cooled below the dew-
point, the water vapor condenses on the inner walls of the vessel and is removed
from the gas stream, e.g., a solution of dry ice and acetone at a temperature of
- 78.5°C will remove all but 0.01 mg/liter, and a cooling bath of liquid nitrogen
at a temperature of - 196°C will remove all but 1 X lO^ing of water vapor/liter of
air (10). Figure 6-27 lists various cold bath solutions and their temperatures.
6-28
-------
Coolant
Ice and water3
Ice and NaCl
Carbon tetrachloride slusha'b
Chlorobenzene slusha'b
Chloroform slusha'b
Dry ice and acetonea
Dry ice and cellosolve3
Dry ice and isopropanol3
Ethyl acetate slush3 ib
Toluene slushb
Carbon disulfide slusha'b
Methyl cyclohexane slusha>b
n —Pentane slushb
Liquid air
Isopentane slushb
Liquid oxygen
Liquid nitrogen
Temperature
0
- 21
- 22.9
- 45.2
- 63.5
- 78.5
- 78.5
- 78.5
- 83.6
- 95
-111.6
-126.3
-130
-147
-160.5
-183
-196
aAdequate for secondary temperature standard.
'The slushes may be prepared by placing solvent in a Dewar vessel and adding small increments of
liquid nitrogen with rapid stirring until the consistency of a thick milkshake is obtained.
Figure 6-27. Summary of cold bath solutions.
Mechanical refrigeration devices especially designed for water removal are com-
mercially available.
Summary
Molecular sieves and activated carbon are used extensively for removal of gaseous
contaminants. These solid adsorbents have a porous quality that gives them an
extremely high surface area per unit weight, thus increasing their adsorptive
capacities.
There are adsorbents and absorbents that will selectively remove one gas from an
air stream and leave the others. These include Hopcalite for selectively removing
CO from an air stream, manganese dioxide for removing ozone from an air stream,
Ascarite for removing CO2 from an air stream and triethanolamine on firebrick for
selectively removing NO2 from an air or NO stream.
For the removal of water vapor, there are basically three methods used: conden-
sation, absorption and adsorption. Condensation is the most efficient means of
drying a gas stream, but may be awkward to use. Absorption, using liquid desic-
cants, has the greatest capacity for drying a gas stream because the liquid can be
recirculated continuously. This method, too, may be awkward for field work.
Adsorption of water vapor using solid adsorbents is used extensively because of the
ease of handling and storage and the advantage of the regenerative properties.
6-29
-------
Sample Problems
Problem 1
A sample gas stream containing sulfur dioxide is to be scrubbed with an adsorbent
while the sample collection is performed. The SO2 concentration is known to be
approximately 10 ppm (10 \dlL The adsorbent to be used is activated carbon. The
sampling rate is 200 f/min to be maintained for 24 hours. How much activated
carbon would be needed to remove all of the SO2 for the entire length of the
sampling period?
Solution
The total amount of sampled air can be calculated 200 f/miri X 60 min/hour X 24
hour = 288,000 liters.
The total amount of SO2 that must be scrubbed from the sample is calculated:
288,000 ! of Air X 10 \d of SO2/£ of Air = 2,880,000 \d of SO2 to be removed. For
the purposes of this example 1 m£= 1cm3 (this will actually add very little error).
2.88 X 106 \d SO2 X 10~3 ml/nlx 1 cm3/m£= 2880 cm3 of SO2 to be removed.
From Figure 6-21, activated carbon will adsorb 380 cm3 of SO2 per gram of
adsorbent (assuming adsorption takes place at 15°C).
The total carbon that is necessary can be calculated:
2,880 cm3 SO2
—— = 7.6 g Carbon
380 cm3 SO2/1 g Carbon
7.6 grams of carbon would be needed to effectively remove the SO2 from the gas
sample. This is an approximate amount.
Problem 2
A sample stream contains approximately 2% (by weight) water vapor, which must
be removed. Molecular Sieve Type 4A has been chosen as the drying agent.
Sampling is to be performed at 2.5 l/min for 8 hours. How much 4A molecular
sieve is needed to dry the air for the length of the sample period? (Assume adsorp-
tion takes place at 25°C).
Solution
The total amount of air sampled:
2.5 f/min X 60 min/hour X 8 hours = 1200 liters.
The density of air at 25 °C and 1 atm (13) is 1.1844 mg/m£
The weight of the air sampled:
= 1200 liters x 1000 mf/f/ X 1.1844 mg/rm°= 1,421,300 mg of air sampled.
The weight of water vapor to be removed:
1,421,300 mgX 0.001 g/mgX 0.02 = 28.43 g.
From Figure 6-24, 4A molecular sieves will remove 24.5 g of H2O per 100 g of sieve.
The amount of sieve necessary to fully dry the air stream (this is an approximate
value):
28.43 g H2O
= 116 g of Molecular sieve 4A
24.5 g H2O/100 g sieve
6-30
-------
Summary
Standard test atmospheres are very important as calibration sources for atmospheric
monitors. The bag-filling method is best suited for "batch" calibration purposes
where only a small amount of calibrated gas is needed at one time. For example, a
series of known CO concentrations can be made very quickly with this technique.
The cylinder method, permeation tube method, and ozone generators are best used
where a constant flow of calibration gas is needed. Some continuous atmospheric
monitors and all manual sampling trains require this type of calibration technique
because of the finite time required to obtain an adequate sample.
References
1. Bowman, L. D., and Horak, R. F. A Continuous Ultraviolet Absorption Ozone
Photometer, pp. 102-108. Dasibi Corporation, ISA AID 72430, 1972.
L. Bryan, R. J., and Cherniak, I. A Comparison Study of Various Types of Ozone
and Oxidant Detectors Which Are Used for Atmospheric Air Sampling.
Presented at the 57th Annual Meeting of the Air Pollution Control Associa-
lion, Houston, Texas, June, 1964.
3. Hei.^h, C. K. Molecular Sieves. New York: Reinhold Publishing Co., 1961.
4. Hesketh, H. E. Understanding and Controlling Air Pollution. Ann Arbor, ML;
Ann Arbor Science Publishers, 1972.
5. Jo. prison, R , editor, Fan Engineering. Buffalo, NY: Buffalo Forge Company,
1970.
6. Landolt, G R. Method for Rapid Determination of Adsorption Properties of
Molecular Sieves. Analytical Chemistry 43: 613-615, April 1971.
7. Levaggi, D. A.; Siu, W.; Feldstein, M.; and Kothny, L., Quantitative Separa-
tion of Nitric Oxide From Nitrogen Dioxide at Atmospheric Concentration
Ranges. Environ. Sci. Tech. 6: 250, 1972.
o. Mantell, C. L. Adsorption. New York: McGraw-Hill Book Company, 1951.
9. Mueller, P. K., et al. A Guide For The Evaluation of Atmospheric Analyzers.
:alif. Dept. of Health, Air and Industrial Hygiene Laboratory, for EPA under
mtract 68-02-0214, June 1973.
10. Nelson, G. O. Controlled Test Atmospheres. Ann Arbor, MI; Ann Arbor
Science Publishers, 1972.
11. Saltzman, B. E.; Burg, W. R.; and Ramaswamy, G. Performance of Permea-
tion Tubes as Standard Gases. Environ. Sci. & Tech. 5: 1121, 1971.
1 Scaringelli, F. P.; Rosenberg, E.; and Rehme, K. A. Comparison of Permea-
i ion Devices and Nitrite Ion as Standards For the Colorimetric Determination
of Nitrogen Dioxide. Environ. Sci. Tech. 4: 924, 1970.
Li. Sheeny, J. P.; Achinger, W. C.; and Simon, R. A. Handbook of Air Pollution.
Public Health Service Publication, No. 999-AP-44, 1968.
14. >mith, G. F. Dehydration Using Anhydrous Magnesium Perchlorate. Colum-
bus, Ohio: G. Frederick Smith Chemical Co.
15. Stern, A. C., Air Pollution, volume II, New York: Academic Press, 1968.
6-31
-------
16. Treybal, R. E. Mass-Transfer Operations. New York: McGraw-Hill Book Co.,
1968.
17. Methods of Air Sampling and Analysis. Intersociety Committee, American
Public Health Association, New York, 1972.
18. 'Basic Concepts of Adsorption on Activated Carbon". Pittsburgh: Pittsburgh
Activated Carbon Co.
19. Mine Safety Appliance Corp. (MSA), Bulletin No. 0102-2.
6-32
-------
Chapter 7
Standard Methods
for Criteria Pollutants
Introduction
At the present time, there are seven pollutants for which the Environmental Protec-
tion Agency has promulgated primary and secondary ambient air quality stan-
dards. These pollutants are listed in Figure 7-1(1) together with the averaging time
and reference measurement methods. These primary or health related and secon-
dary or welfare related pollutant standards are contained in Title 40 Part 50 of the
Code of Federal Regulations (40 CFR 50) and were set forth by the authority
granted in Section 109 of the Clean Air Act (1977) (2). The legal methods for
monitoring ambient atmospheres for these seven pollutants are contained in the
appendices (Appendix A through G) of Part 50. Calibration procedures are also
prescribed.
Section 109 of the Clean Air Act as amended in 1977 requires the Environmental
Protection Agency to evaluate the criteria for which standards have been pro-
mulgated at five-year intervals and to issue any new standards as may be
appropriate. The issuance of reference methods designed to monitor these criteria
pollutants has a legal basis in Section 301 of the Clean Air Act which states that
regulations may be promulgated by the Administrator which are necessary to carry
out the provisions of the Act. In order to evaluate and ascertain the status of air
quality with regard to the criteria pollutants, uniform analytical methods are used
to insure consistency and accuracy in the data generated. Since these methods will
be required for extensive monitoring (for state implementation plans, prevention of
significant deterioration, etc.) other factors considered are cost, availability, and
degree of sophistication.
Reference Method, Equivalent Method, and Measurement Principle
By definition, a reference method is a "method of sampling and analyzing the
ambient air for an air pollutant that is specified as a reference method in the
appendices" of Title 40 Part 50 of the Code of Federal Regulations or a method
that has been designated as a reference method in accordance with Part 53 of
Title 40b. An equivalent method is a method of "sampling and analyzing the am-
bient air that has been designated as an equivalent method in accordance with
Part 53 of 40 CFR." Two types of methods may be considered for equivalent
method determination:
1. manual method
2. automated method
A reference method may be either a manual method or an automated as well.
7-1
-------
If a manual method is to be considered for equivalent designation, it must
demonstrate a consistent relationship to the reference method when both methods
are used to measure pollutant concentrations in a real atmosphere. For automated
methods, each instrument must meet the performance specifications listed in
Figure 7-2(5).
Pollutant
Sulfur dioxide
Suspended
paniculate matter
Carbon monoxide
Ozone
Hydrocarbons
(corrected for
methane)
Nitrogen
dioxide
Lead
Averaging
time
Annual arithmetic
mean
24 hr
3 hr
Annual geometric
mean
24 hr
8 hr
1 hr
1 hr
3 hr
Annual arithmetic
mean
Calendar quarter
Primaryb>c
standards
80 fig/m3
(0.03 ppm)
365 jtg/m3
(0.14 ppm)
—
75 /tg/ms
260 fig/m3
10 mg/m3
(9 ppm)
40 mg/m3
(35 ppm)
0.12 ppm
160 ng/m3 (f)
(0.24 ppm)
100 jig/m3
(0.05 ppm)
1.5 /*g/m3
Secondary*1
standards
1300 jtg/m3
(0.5 ppm)
60 ftg/m3 (f)
150 /ig/m3
Same as primary
standard
Same as primary
standard
Same as primary
standard
Same as primary
standard
Same as primary
standard
Reference6
method
Pararosaniline
method
High volume
sampling method
Nondispersive infra-
red spectroscopy
Gas-phase
chemiluminescent
method with ethylene
Flame ionization
detection using
gas chromatography
Gas-phase
chemiluminescence
with ozone
High volume
sampling, atomic
absorption analysis
Environmental Protection Agency, Federal Register 40 CFR 50 p. 4-6 (July 1, 1979).
bNational standards other than those based on annual arithmetic means, annual geometric means,
or quarterly arithmetic means are not to be exceeded more than once per year.
cNational Primary Standards: The levels of air quality necessary, with an adequate margin of
safety, to protect the public health.
dNational Secondary Standards: The levels of air quality necessary to protect the public welfare
from any known or anticipated adverse effects of a pollutant.
eReference method as described by EPA. An "equivalent method" means any method of sampling
and analysis which can be demonstrated to have a "consistent relationship to the reference
method."
fFor use as a guideline in assessing implementation plans.
Figure 7-1. National ambient air quality standards11.
7-2
-------
Performance parameter
2 Noise
3 Lower detectable limit
5 Zero drift 12 & 24 hr
6 Span drift 24 hr
20% of upper range limit . . .
80% of upper range limit . . .
20% of upper range limit . . .
80% of upper range limit . . .
Units
Parts per
million
do
do
do
do
do
Percent
do
Minutes
do
do
Parts per
million
do
Sulfur
dioxide
0-0.5
.005
.01
±.02
.06
±.02
±20.0
±5.0
20.
15
15
.01
.015
Ozone
0-0.5
.005
.01
±.02
.06
±.02
±20.0
±5.0
20.
15
15
.01
.01
Carbon
monoxide
0-50
.50
1.0
±1.0
1.5
±1.0
±10.0
±2.5
10
5.
5
.5
.5
Nitrogen
dioxide
0-0.5
.005
.01
±0.02
.04
±.02
±20.0
±5.0
20
15
15
.02
.03
Definitions
and test
procedures
Sec.53.23(a).
Sec.53.23(b).
Sec.53.23(c).
Sec.53.23(d).
Sec.53.23(e).
Do.
Do
Do
Do
Do
Figure 7-2. Performance specifications for automated methods. For definitions of the various
performance parameters see Reference 5 of this chapter.
The distinction between automated reference and equivalent methods is based
upon the measurement principle that an instrument employs. For example, the
measurement principle for the automated method for the detection of the oxides of
nitrogen is chemiluminescence with ozone. Any instrument, therefore, which meets
the performance specification for automated methods and uses chemiluminescence
with ozone for detection is a reference method. Instruments using other measure-
ment principles would be designated equivalent methods provided they meet the
performance specifications. This section is designed to provide a brief overview of
the reference methods as they are described in the Federal Register, discuss poten-
tial problem areas and give some quality assurance considerations from a practical
standpoint.
Sulfur Dioxide
The reference method for the measurement of sulfur dioxide in ambient
atmospheres is a manual wet-chemical method —the Pararosaniline Method. Sulfur
dioxide is bubbled through a solution of potassium tetrachlorosulfitomercurate
(TCM) which forms a dichlorosulfitomercurate complex. This complex forms an
intensely colored solution upon addition of pararosaniline and formaldehyde. The
concentration of sulfur dioxide can be determined spectrophotometrically by
measuring the absorbance of the colored solution.
7-3
-------
The potential problems associated with interferences are minimized by the pro-
cedures listed in Figure 7-3.
Interferent
Ozone
heavy metals
oxides of Nitrogen
Corrective procedure
time delay
addition of EDTA and
Phosphoric Acid
addition of Sulfamic Acid
Figure 7-3. Pararosaniline interferences.
Other precautions to be considered relate to the sampling train (Figure 7-4) and
sampling conditions. The diameter of the impinger should be checked "such that a
No. 79 jewelers drill will pass through but a No. 78 jewelers drill will not(6)." The
temperature instability of the dichlorosulfitomercurate complex during sampling
also poses potential problems. This can be minimized by using a temperature con-
trolled sampling apparatus (7). Collected samples should be kept at a temperature
less than 5°C. Since this method involves a laboratory analysis of the sample after
collection in the field, care should be taken in handling the sample once it has
been collected to avoid contamination.
Hypodermic needle
To impinger
To air pump
Optional critical orifice flow control
To air pump
Impinger
Figure 7-4. Manual SO, sampling train for 30 minute to 1 hour sampling.
7-4
-------
Nitrogen Dioxide
The reference method for the determination of nitrogen dioxide involves the gas
phase chemiluminescence reaction of nitric oxide with ozone (Eq. 7-1). A more
complete discussion of chemiluminescence is given in Chapter 8.
(Eq. 7-1) NO+O3—NO2 + hi> (light)
Nitrogen dioxide is measured indirectly by measuring total oxides of nitrogen
(\O> = NO2 + NO) and electronically subtracting nitric oxide concentration yielding
a nitrogen dioxide determination. A measure of the total oxides of nitrogen is
obtained by passing the sample across a catalytic converter which reduces nitrogen
dioxide to nitric oxide (Eq. 7-2).
(Eq. 7-2) NO2 + NO + catalytic converter-^NO
The calibration of the NO and NOX channels of the instrument is accomplished
by diluting a pressurized tank of WO. The calibration of the NO2 channel is
accomplished with a permeation device or the gas phase titration of an NO stan-
dard with ozone. The gas phase titration involves the following reaction (Eq. 7-3).
(Eq. 7-3) NO + 03-N02
This produces a known amount of nitrogen dioxide which allows one to calibrate
the NO2 channel of the monitor. The other calibration procedure specified in
Appendix F, 40 CFR 50 uses a dynamic dilution system in combination with a
permeation device to produce a known amount of NO2. The use of permeation
devices and dynamic calibration procedures are discussed in Chapter 6 of this
manual.
The problem areas associated with this method are interferents such as
peroxyacetyl nitrate and other nitrogen containing compounds (8). In periods of
high photochemical activity, corrections for these interferences may be necessary.
A recent smog chamber study (9) indicated that the presence of high levels of
halocarbons gave a positive interference in a reference method NO2 analyzer.
The chemical composition of the atmosphere plays an important role in deter-
mining the validity of the nitrogen dioxide measurements obtained using a
reference method analyzer. For most ambient air measurements, however,
interferents such as the ones mentioned previously are minimal. The frequency of
calibration and other maintenance (e.g., replacement of charcoal ozone filter,
check of converter efficiency, etc.) are important quality assurance considerations.
Carbon Monoxide
The reference method used to measure ambient levels of carbon monoxide is non-
dispersive infrared spectrometry. A complete discussion of the principle of opera-
tion of an NDIR spectrometer is contained in Chapter 8. In brief, a signal is pro-
7-5
-------
duced as a result of the infrared absorption of carbon monoxide which can be
related to the absolute concentration by calibrating the instrument as specified in
Appendix C of 40 CFR 50.
The problem areas associated with this method are the broad band absorption of
carbon dioxide and water vapor. These two compounds' absorption result in an
interference. The technique of a "negative filter" NDIR analyzer (discussed in
detail in Chapter 8) corrects for these problems to a limited extent.
Total Suspended Particulate Matter
The High Volume Method is the reference method for the determination of total
suspended particulates in ambient air. In this method, air is drawn through a glass
fiber filter and the amount of particulate matter is determined gravimetrically.
The glass fiber filter is used to collect the particulate matter in a sampler as
illustrated in Figure 7-5. A rotameter, attached to the sampler motor near the
exhaust port, is calibrated against actual air-flow with an orifice calibration unit
(Figure 7-6). This orifice calibration unit is first calibrated against a standard roots-
meter"'. A detailed description of the calibration procedure is in Chapter 4.
Figure 7-5. Assembled sampler and shelter.
7-6
-------
Orifice
Resistance plates
Figure 7-6. Orifice calibration unit.
The problems with this method are the non-uniformity of sampling flow rate,
artifact formation of sulfates and nitrates, and the wind direction sensitivity of the
apparatus. These problems are discussed more fully in the High Volume section of
Chapter 4.
Ozone
The measurement principle for the measurement of ozone is the gas-phase
chemiluminescence resulting from the reaction of ozone with ethylene. The
unstable nature of ozone requires that ozone be generated in-situ in order to
calibrate ozone monitors. The NBKI wet-chemical procedure which was previously
used to measure ozone concentration from ozone generators demonstrated some
"inherent shortcomings" (10). UV photometry was found the most satisfactory for
the measurement of ozone concentrations generated in-situ and has been
designated the calibration procedure. An 18-month phase-in period using a Boric
Acid Potassium Iodide procedure began February 8, 1979 (10) to permit air
monitoring agencies to acquire and train personnel to operate UV photometers.
Figure 7-7 (10) illustrates the BAKI calibration system option and a schematic of
the UV photometry calibration system is shown in Figure 7-8.
Nonmethane Hydrocarbons
The reference method for the determination of nonmethane hydrocarbons is gas
chromatography with a flame ionization detector (FID). Hydrocarbons are
separated from water, carbon dioxide, and methane by a stripper column and a
measurement of "total hydrocarbon" is obtained.
7-7
-------
Of all the gaseous reference methods, this method is the least reliable. Since
different hydrocarbons have different responses to a FID, "total hydrocarbon"
measurements are subject to question. The "calibration gas" is simply not specified
in the method and, therefore, any hydrocarbon or mixture of hydrocarbons may be
used.
Research by some instrument companies has involved converting all hydrocar-
bons to carbon dioxide or methane and then performing a CO2 or CH4 measure-
ment (11). These instruments in principle could operate similar to water "total
organic carbon" analyzers.
o
Zero air
Flow
controller
Flowmeter
^
O9 generator
Output manifold
Vent
Extra outlets capped
when not in use
\ \
To inlet of To inlet of analyzer
KI sampling train under calibration
Figure 7-7. Schematic diagram of a typical BAKI calibration system.
7-8
-------
Output manifold
Zero air
Extra outlets capped
when not in use
To inlet of analyzer
under calibration
Vent
Figure 7-8. Schematic diagram of a typical UV photometric calibration system.
Lead
The reference method for lead consists of measuring the lead content of particulate
matter collected by the total suspended particulate reference method's high volume
sampling procedure. After sample collection, lead is acid extracted from the par-
ticulate matter of a %" X 8" strip of the high volume filter. Finally, the lead con-
tent of the resulting solution is determined by atomic absorption spectrometry (12).
Potential problems exist in both the sampling and analysis portions of the
method. In addition to the sampling problems associated with the reference
method for total suspended particulates, which were discussed in Chapter 4, lead is
non-uniformly distributed across the filter when sampling near heavily traveled
roadways. The problem can be alleviated by analyzing a larger portion of the
filter. Chemical and light scattering interferences may be encountered during the
atomic absorption spectrophotometric analysis. Chemical interferences can be over-
come by using the method of standard additions. Light scattering interferences can
be corrected by using a dual channel atomic absorption spectrophotometer
equipped with a continuum light source, by using a non-absorbing wavelength that
is near the lead analytical wavelength, or by using a chelating agent. Furthermore,
for accurate particulate lead analyses it is necessary that the variation of lead con-
tent from filter to filter within a given batch of blank filters be small (12).
7-9
-------
References
1. Coloff, S. C.; Cooke, M.; Drago, R. J.; and Sleva, S. F. Ambient Air Monitor-
ing of Gaseous Pollutants, American Laboratory, 1973.
2. The Clean Air Act as amended August 1977.
3. Environmental Protection Agency, Fed. Reg. 42 1271-1289, December 14,
1977.
4. Title 40 Part 50.1, Code of Federal Regulations, pp. 4-6, July 1, 1979.
5. Title 40 Part 53.3, Code of Federal Regulations, pp. 979-986, July 1, 1979.
6. Title 40 Part 50, Code of Federal Regulations, Appendix A, pp. 6-11, July 1, 1979.
7. Research Appliance Company Applications Note, Rt. 8, Gibsonia, PA, 1978.
8. Winer, A. M.; Peters, J. W.; Smith, J. P.; and Pitts, J. N. Response of Com-
mercial Chemiluminescence NO-NO2 Analyzers to Other Nitrogen-Containing
Compounds. Environ. Sci. Technol. 8: 1118, 1974.
9. Joshi, S. B., and Bufalini, J. J. Halocarbon Interferences in
Chemiluminescence Measurements of NOX. Environ. Sci. Technol. 12: 5, 1978.
10. Environmental Protection Agency, Fed. Reg. 43, 121 36962, June 1978.
11. McElroy, F. Personal communication.
12. Federal Register, vol. 43, no. 194, pp. 46258-46261, October 5, 1978.
7-10
-------
Chapter 8
Continuous Air Monitoring
Instrumentation
Introduction
Instrumental methods are assuming an increasingly prominent role in monitoring
gaseous pollutants in air. The advantages of instrumental methods, together with
advances in the associated technology, have made older wet-chemical methods
largely obsolete. Real-time data output, greater sensitivity in meeting the
requirements of specific applications, in-situ measurement, and the ability to input
directly into computer data systems represent major advantages of instrumental
techniques.
The methods that are used for the measurement of oxides of nitrogen exemplify
the advantages of an instrumental method. The wet-chemical techniques involve
analyzing a grab sample (phenoldisulfonic acid method) for source samples and a
24-hour sample (TGS-ANSA or Sodium Arsenite method) for ambient samples.
These techniques, while reliable and relatively inexpensive, require considerable
sample handling and lack real-time data output. Moreover, for ambient air
monitoring, a short term (less than 24 hours) averaging time requirement will be
promulgated by the Environmental Protection Agency. This requirement will
reduce the applicability of the wet-chemical methods since they were developed
specifically for 24-hour samples. The development of a continuous instrumental
method that measures the chemiluminescence produced by the reaction of nitric
oxide and ozone has provided a reliable real-time method for the measurement of
oxides of nitrogen. This method has been applied to both source and ambient
in-situ monitoring.
Optical techniques using chemiluminescence and other spectroscopic properties
of molecules provide the basis for many continuous air monitoring instruments.
These instruments utilize characteristics of a gaseous pollutant that are relatively
specific for that molecule. The ultraviolet absorption of ozone and infrared absorp-
tion of carbon monoxide are examples of the spectroscopic properties measured in
instrumental methods. The ability of a particular pollutant to oxidize halogens is
also used in continuous monitors. This section deals with the advantages, disadvan-
tages, principles of operation, and current applications of instruments using
physical and chemical properties of air pollutants.
8-1
-------
Coulometric Instruments
Principles of Operation
Coulometric analytical techniques use the electrical charge generated by oxidation-
reduction reactions occurring in an electrolytic cell to measure gaseous pollutant
concentrations. The air sample containing the pollutant flows through the cell and
oxidizes (or reduces) part of the solution at one electrode. A balancing reaction
occurs at the other electrode causing a small current to flow across the cell. The
concentration can be determined by Faraday's Law, which states that one gram-
equivalent of a material is oxidized or reduced by one Faraday of electricity. By
measuring the current across the cell, the concentration of the sample may be
determined since the quantity of electricity (Q, coulombs) is g;iven as the integral of
current (i, amperes) over the time interval (t, seconds) (1):
(Eq. 8-1)
Where: m - mass in grams of the species consumed or produced during
electrolysis
M — gram molecular, weight
z = number of Faradays (equivalents) of electricity required per
gram mole (i.e., the number of electrons appearing in the
equation for the net reaction of interest)
F = proportionality constant: 96,487 coulombs /mole
This type of Coulometric technique measures the amount (i.e., coulombs) of elec-
tricity directly produced as the result of a reaction of a pollutant at the electrode.
In current instrumentation, however, pollutant concentration is determined
indirectly by measuring the current required to maintain a constant halogen con-
centration. These halogens are typically bromine (Br2) or iodine (I2). The sample
gas passes through an electrolytic cell and oxidizes the halogen to halide (for exam-
ple, Br2 + pollutant — Br~), thereby reducing the halogen concentration. The cur-
rent required to maintain the electrochemical balance is directly proportional to
the pollutant concentration. Since current and not charge is measured, these
analyzers are properly termed amperometric analyzers.
Current Applications
Coulometric or amperometric techniques have been most widely applied to
analyzing ambient sulfur dioxide. A constant iodine concentration can be
generated by application of a constant current to the electrodes in the detector cell.
Upon entering the cell, sulfur dioxide undergoes hydrolysis, and the hydrolysis pro-
duct reduces the steady state concentration of iodine in the electrolytic cell:
(Eq. 8-2) S02 + I2 + 2H2 O - SO4= + 2I~ + 4/T
8-2
-------
Since the iodine concentration has been reduced, the cell loses the capacity to
carry the full applied anodic current to the cathode. As a result, current flows
through the reference electrode until the electrochemical balance in the cell has
been restored. Figure 8-1 illustrates the flow of current and the reactions occuring
at each electrode (2).
Detector
cell*
Constant
current
source
I
Platinum anode (21 ~ —1« + 2e - )
I
Reference electrode
Platinum
' -cathode
1
Amplifier,
^y v_x
2e- -2I-)
*Gas phase sample is continuously introduced
into detector cell
Figure 8-1. Coulometric titration of iodine.
Instrumentation also employing this type of coulometric titration measures the
reduction of bromine rather than iodine. Typically, bromine concentration is
maintained at a constant level by a set of indicator-reference electrodes and the
electronic circuits. As sulfur dioxide is introduced into the detector cells, the
bromine (Br2) concentration is reduced by the reaction (8-3):
(Eq. 8-3)
SO2 + Br2
This reaction disturbs the Br2/Br ratio, and this change is sensed by a basic
amplifier. The current required to regenerate the bromine (Br2) through the
generator-auxiliary electrodes is proportional to the sulfur dioxide concentration.
Figure 8-2 illustrates the four electrodes and circuit employed in this type of
instrument (3).
8-3
-------
Air
Basic amplifier
Titration amplifier Current amplifier
Figure 8-2. Coulometric titration of bromine.
Recently, Lindqvist (4) described a sulfur dioxide monitor based on the anodic
oxidation of sulfur dioxide in a galvanic cell. In this type of instrument, sulfur
dioxide is oxidized directly by the electrode, and the resulting current can be
related to the sulfur dioxide in ambient air. The design of this detector is unique
in that current generated by interferences is eliminated by a differential measuring
technique. The cell contains two anodes.
One anode is the reference electrode, and the other the measuring electrode. By
measuring the potential difference (AV) across the cell section when the air flow is
equal in both anode sections, the current generated by interferences will be
eliminated, and the current will be due solely to the presence of sulfur dioxide.
The oxidation of sulfur dioxide at the measuring anode in reaction 8-4 can be
related to concentration by Faraday's Law.
(Eq. 8-4)
Advantages
As Lindqvist has noted, one of the principle advantages of the coulometric tech-
nique for ambient SO2 monitoring is that these instruments require no compressed
gases such as instruments using gas chromatography and flame photometric
methods (4). They can, therefore, be used in aircraft and other sites where the use
of compressed hydrogen might pose a potential explosion hazard. The major
advantage, however, is that coulometric analyzers require minimal maintenance.
The halide contained in the cell is regenerated and volatilization of the halogen is
8-4
-------
not a problem (5). The electrolyte solution that is lost by evaporation can be
replaced by condensation from air, or a thermistor (a device for temperature
measurement) can be used to automatically maintain the reservoir at a certain
level.
Potential Interferents and Disadvantages
The potential interferents in the sulfur dioxide (SO2) coulometric instruments
described here are compounds that are able to oxidize halogens, reduce halogens,
or complex with either. Sulfur compounds such as hydrogen sulfide, mercaptans,
organic sulfides, and organic disulfides are the most notable interferents. Nitric
oxide (NO), ozone (O3), nitrogen dioxide (NO2), chlorine (C12), ethane (CH3CH3),
among others are also interferents but are much less significant than sulfur com-
pounds. Chemical filters to remove these interferents are usually incorporated into
coulometric sulfur dioxide analyzers.
Second Derivative Spectroscopic Instruments
Principles of Operation
In direct absorption spectroscopy, pollutant gases absorb energy at a specific
wavelength (X) and the absorption is related to concentration by Beer's Law
(Equation 8-5):
(Eq. 8-5) 7(X) = /0(X)e-«Wc'
Where: I0 = incident intensity at X
I— observed intensity at X
c = concentration of absorbing gas
?= path length through the gas
a —absorption coefficient of the gas at X
e — natural Iqg function
Second derivative instruments, on the other hand, relate the slope and curvature
characteristics of absorption bands to the concentration of gaseous pollutants.
These instruments, rather than produce a second derivative spectrum, produce a
signal from which a second derivative voltage output can be extracted. This can be
accomplished by using phase lock amplifier circuits. The second derivative voltage
output produced by the instrument is directly related to concentration by
Equation 8-6 (8).
(Eq. 8-6)
Signal from constant, term linear with gas
derivative independent concentration
spectrometer of intensity with fixed
pathlength
8-5
-------
The method for producing a second derivative signal consists of modulating with
time essentially monochromatic light with the amplitude of light approximately
equal to the band width being detected. This light is passed through the sample
gas and strikes a photomultiplier tube. The (dzl/d\z)/l second derivative portion of
the signal is extracted by the electronics — "signal analysis" — and can be related to
the concentration of the pollutant. Figure 8-3 (7) illustrates the arrangement of the
optics employed to produce a second derivative signal.
Sample cell
Photomultiplier tube
Sample inlet
Sample exhaust
0.25—m
Monochromator
Signal
analysis
electronics
Wobbler
Wavelength
drive
programmer
Ultraviolet
source
Figure 8-3. Schematic of optics employed in a second derivative spectrometer.
Since second derivative instruments, like direct absorption spectrometers, do not
provide absolute measurements, they must be calibrated with various concentra-
tions of the pollutant being measured.
Current Applications
Second derivative spectrometers are being used to measure both source and
ambient pollutants. They have been applied to the measurement of nitrogen diox-
ide, sulfur dioxide, and ammonia. Another interesting application has been the
determination of trace nitrogen dioxide in high purity argon (6).
8-6
-------
Advantages
Second derivative spectrometers have the advantage of being highly specific for
trace analysis of complex mixtures of atmospheric pollutants (8). Broad band
absorption by interferent gases or other light scattered by paniculate matter is
minimized by these instruments. Also, they have been adapted to in-situ stack
sampling and unattended ambient monitoring, and the instrumentation employed
is relatively simple for single component analysis. Furthermore, no support gases or
reactive solutions are required.
Disadvantages
As a practical consideration, the high cost of these instruments in comparison to
other instrumental and wet-chemical techniques is the major disadvantage. The
mathematical basis for the instrument is complex, and the lack of trained per-
sonnel to install, and maintain these instruments are further disadvantages. It
should be noted, however, that this technique has only recently been applied to
process control, source and ambient monitoring. Its utility, therefore, remains to
be determined.
Flame Photometric Instruments
Principles of Operation
Instrumentation employing flame photometric detection measures the emission
intensity of a selected wavelength while the pollutant is introduced into a hydrogen-
rich flame. Figure 8-4 illustrates the principal components of a flame photometric
detector (FPD).
Air pump
t
Optical window
Light filter
Secondary flame •
Primary flame •
±
Power supply
PMT
Amplifier and
read-out
Sample gas
Figure 8-4. Flame photometric detector.
8-7
-------
The sample gas containing sulfur pollutants is introduced into the hydrogen-rich
flame producing a S2 species that reacts with the available hydrogen and hydroxyl
radicals to produce a high energy species S2*. In an excellent review by Farwell
et al. (9), the principal reactions involved in the formation of the high energy S2
from the sulfur-containing pollutant are discussed. They may be briefly sum-
marized in reactions 8-7 through 8-14.
(Eq. 8-7) Sulfur compound at S atoms or indirectly by
(Eq. 8-8) Sulfur compound ea H2S
(Eq. 8-9) H2S + •H-HS + H2
(Eq. 8-10) HS '!f-S + H2
In the cool outer cone of the flame there results in the formation by excited S*
molecules by the following set of reactions:
(Eq. 8-11) »H+»H+ S2—*H2 + S2
where the recombination of atomic hydrogen furnishes the excitation energy for
S2 —S* or by the reaction
(Eq. 8-12) «H+ •OH+S2-~S?+H20
where the formation of H2O supplied the excitation energy for S2 —S*.
The mechanism for the production of strong luminescent emission with the con-
current conversion of the high energy S* to the lower energy state S2 may be
summarized:
(Eq. 8-13) S,+ »H+»H~&
or S + S + Af — S*+M (where M is a third body)
(Eq. 8-14) S*-~S2 + hv
The observed luminescent emission (hv) is related to the concentration of the
sulfur pollutant by Equation 8-15.
(Eq. 8-15) Is2 = Io[S]n
Where Is = observed intensity of the molecular emission due to the S2 species
[S] = concentration of sulfur atoms
I0 = constant under given experimental conditions
n - constant (usually assumed to be 2) under given experimental
conditions
8-8
-------
It should be noted that the concentration of sulfur pollutant is not linearly pro-
portional to the observed intensity. Most instruments, however, use linearizing elec-
tronic circuitry to yield a linear voltage output. Moreover, instruments using FPD
must be calibrated with various concentrations of the particular pollutant gas since
this measurement is not absolute and different sulfur gases have different responses.
Current Applications
The flame photometric detector is most frequently used in the detection of ambient
sulfur dioxide. This instrumentation uses a scrubber to remove potential
interferents (usually sulfur compounds other than SO2) and the signal is converted
to a linear ppm SO2 output. The flame photometric detector in combination with
gas chromotography has been used to detect mixtures of sulfur-containing
pollutants (SO2, H2S, CH3SH, and others). Stevens, et al. (10), for example,
described a gas chromatographic system using flame photometric detection to
separate and quantitate various sulfur pollutants.
Potential Interferents
Instrumentation designed to monitor for a specific sulfur compound, such as SO2,
must contain scrubbers that remove all other sulfur-containing compounds. This is
a consequence of the fact that the flame photometric detector does not
discriminate between different sulfur-containing compounds. Moreover, as Farwell
notes (9), the presence of other organic compounds in the sample gas (as well as
variable carbon dioxide) may result in a change in the response of the FPD. This
change in response may be attributed to "the inactivation of the excited S* species
by their combination or collision with organic compounds and/or the organic
degradation products." Ambient relative humidity may represent another potential
interferent.
Advantages
The principal advantages of FPD analyzers are low maintenance, high sensitivity,
fast response, and selectivity for sulfur compounds. There are no solutions, and the
only reagents are hydrogen and air for the flame. These analyzers have the
potential for unattended operation. They are particularly useful at most ambient
air monitoring sites since interference from non-sulfur species is essentially non-
existent (10).
Fluorescence Instruments
Fluorescence emission can be differentiated from other types of luminescence by
the type of excitation energy. Chemiluminescence, for example, uses a chemical
reaction, x-ray fluorescence uses x-rays, and fluorescence uses ultraviolet-visible
light. Fluorescence can be distinguished from phosphorescence mechanistically and
8-9
-------
empirically by observing the lifetime of the excited states (~ 10~8 second for
fluorescence; several seconds for phosphorescence). The energy level diagram in
Figure 8-5 illustrates the mechanism involved in fluorescence emission (11).
Vibrational relaxation
(10-12 second) v
i
Abso
So
ion ^
JA* I T
n
i
?<
0
i
i<
A '
k i i
I
s,
3n (excita
18 second
y
k T
' ' 1
Y /
t j! !
don) . | I
1 !
I ii1
Fluorescence ^
10 ~8 second D
|j!!
-------
Advantages
These instruments use no consumable gases, as does the FPD, and the inherent
stability enables the instrument to operate for long unattended periods. The
instruments are linear over a wide range (0.5-5000 ppm), responsive, and insen-
sitive to temperature and flow variations.
Potential Interferents and Disadvantages
The interferents in "UV fluorescence" instruments are related primarily to the
quenching effects of O2) N2, CO2, and water vapor. For ambient monitoring, the
level of O2, N2, and, for the most part CO2 and water vapor, are constant, and
thus the quenching effect is fairly constant. In flue gases, however, the level of
CO2, water vapor, and O2 vary considerably. Consequently, the SO2 values may be
spurious. Jahnke et al. (13) have obtained approximate values for the quenching
coefficients of these gases providing corrections for different concentrations of CO2,
N2, and O2. Hydrocarbons may also interfere but all "equivalent" analyzers now
come equipped with a hydrocarbon scrubber or "cutter" to eliminate this potential
problem.
Chemiluminescence Instruments
Principles of Operation
Chemiluminescence instruments in air pollution monitoring measure the emission
of light produced by the reaction of a gaseous pollutant and a reagent gas. The
process involved in homogeneous gas-phase Chemiluminescence is summarized in
Reactions 8-16 and 8-17.
(R8-16) A+B~C* + D
(R8-17) C*~C+hv
The reaction of pollutant A with excess reagent gas B produces a chemi-excited
species C*, which reverts to a lower energy state by emission of light (hv). The light
thus produced passes through an optical filter to isolate a given spectral region and
strikes a photomultiplier tube (PMT). The PMT signal is amplified and is directly
proportional to the concentration of pollutant (14).
Instrumentation employing this type of Chemiluminescence is compound or
pollutant specific and should be distinguished from the flame Chemiluminescence
detector (flame photometric detector) described previously, which is element
specific. The flame photometric detectors are used almost exclusively for the
measurement of sulfur compounds. The Chemiluminescence instruments discussed
here use a reagent gas rather than a hydrogen-rich flame to produce a chemi-
excited species.
8-11
-------
Figure 8-6 illustrates the basic components and gas flow of a compound-specific
chemiluminescence detector. Sample gas containing the pollutant enters the reac-
tion chamber and reacts with the reagent gas.
Sample gas
Reagent gas
Amplifier and
read-out
Pump
Figure 8-6. Compound specific chemiluminescence detector.
The intensity of the light produced is related to the concentration of pollutant
and reagent gas (Equation 8-16).
(Eq. 8-16) I=k[A][B]
Where: I—intensity of light emitted
k — rate coefficient (a constant)
[A] — concentration of pollutant
[B] = concentration of reagent gas
Since the reagent gas is usually present in large excess, the concentration is
assumed to be constant and Equation 8-17 can be used to relate the intensity to the
concentration of pollutant.
(Eq. 8-17) I=c[A]
Where: I'= intensity of light emitted
c = constant under given experimental cQnditions
[A] = concentration of pollutant
8-12
-------
Current Applications
Chemiluminescence techniques are routinely used for the measurement of ozone
and oxides of nitrogen. The chemiluminescence resulting from an ethylene/ozone
reaction was first reported by Nederbragt et al. (15) in 1965 and forms the basis
for present ozone monitors using homegenous gas-phase chemiluminescence. This
reaction produces a continuum at 435 nm presumably due to an excited aldehyde
linkage (16). After extensive field testing and evaluation of prototype instruments,
ozone/ethylene chemiluminescence was designated as the reference principle for
ozone detection by the Environmental Protection Agency (17).
The reaction of ozone (O3) with nitric oxide (NO) produces an emission spec-
trum that is shown in Figure 8-7 (15).
U
s
o1
III"
Photomultiplier
Reaction
100
50
I
_g
c
o
400 800 1200 1600 2000
Wavelength, nm
2400
2800
3200
Figure 8-7. Emission spectrum of ozone/nitric oxide.
This spectrum results from the chemiluminescence of ozone and nitric oxide
shown in Reactions 8-18 and 8-19.
(Eq. 8-18)
(Eq. 8-19)
8-13
-------
Nitric oxide is measured directly. Nitrogen dioxide measurements can be made
by catalytic conversion of nitrogen dioxide to nitric oxide with carbon-based and
other metal converters resulting in a total oxide of nitrogen concentration
(NOX = NO + NO2). The difference between total oxides of nitrogen (NO2 + NO)
and nitric oxide is a measure of nitrogen dioxide concentration.
Potential Interferents
For a positive interference, a species must react with the reagent to yield
chemiluminescence in the spectral region of the pollutant emission. In nitrogen
oxide detectors a filter that absorbs emissions below 600 nm is placed between the
reaction chamber and photocathode. This eliminates emissions from the reaction of
ozone with olefins that could produce a positive interference. Ambient instruments
also respond to other nitrogen pollutants such as peroxyacetyl nitrate, ammonia,
and other organic nitrates (18). These nitrogen compounds are converted to NO in
the converter. Quenching effects are a source of negative interference. In ambient
air, molecular nitrogen and oxygen are the predominate quenching agents; but,
since these agents do not vary significantly in concentration the quenching effect is
generally constant. In source measurements, carbon dioxide and water vapor are
additional quenching agents posing a potential negative interference.
Advantages
The advantages of gas-phase chemiluminescence are that emissions are relatively
specific for the pollutant being monitored. Gas-phase reactions in instruments using
continuous flow are insensitive to changes in surface properties in the reaction
chamber (19). Furthermore, a linear response in the range of 4 ppb to 100 ppm
can be obtained.
Ultraviolet Photometric Instruments
Principle of Operation
The visible and ultraviolet absorption of molecules are associated with electronic
energy level transitions. For example, TT—TT* or a—a* transitions could result in
ultraviolet absorption. Instruments making use of the ultraviolet absorption of
pollutants measure the absorption at a specific wavelength that can be related to
the concentration of the pollutant by Beer's Law (Equation 8-5).
Current Applications
The ultraviolet absorption of ozone is shown in Figure 8-8. This characteristic of
ozone will be used to calibrate ozone monitors using chemiluminescent detection by
the Environmental Protection Agency after evaluation of UV photometry, a wet
8-14
-------
1000
a
V
o
'{L
o
0.1
1800 2000
2200 2400 2600
Wavelength A
2800 3000 3200
Figure 8-8. Ultraviolet absorption of ozone.
chemical method, and gas-phase titration. The procedure using a UV calibration
photometer was shown to "have the best performance —low variability, high
accuracy, and minimum operator involvement (20)." Figure 8-9 shows the gas flow
and components of an instrument using UV photometry for the detection of ozone
(21). The calibration of ambient ozone instruments (chemiluminescence), con-
tinuous monitors for source NOX, SO2, H2S, and other gases, and the use of a UV
photometer as an analyzer itself are but a few of its current uses.
Advantages
The UV photometry method for calibration of ozone monitors eliminated a wet
chemical method —"the NBKI procedure," which had demonstrated "some
inherent short-comings (20)." In source applications, instruments measure the dif-
ference in absorption of radiation (in the UV-visible region) by the stack gas at two
different wavelengths. This provides continuous measurement of gaseous pollutants
(NOX, SO2, etc.) in stacks. These instruments have numerous advantages over wet-
chemical techniques, which require considerable sample handling and lack con-
tinuous data output.
8-15
-------
Ozone sample line
Valve
De-ozoner
UV source I j I Source detector/monitor
n
Reference sample line
-a
Flow orifice
•JT>
Sample cell
Pump
D
UV detector
Figure 8-9. Flow and components of ozone detector using UV photometry.
NDIR Instruments
Principle of Operation
Nondispersive infrared (NDIR) analyzers have been developed to monitor SO2,
NOX, CO, CO2, and other gases that absorb in the infrared, including hydrocar-
bons. They are termed nondispersive analyzers because they use no prism or
grating to disperse the infrared radiation. A narrow band filter that absorbs all fre-
quencies except the one of interest is used. Figure 8-10 illustrates the essential com-
ponents and gas flow of a typical NDIR instrument.
Light passes through the sample and reference cell and then through a narrow
bandpass filter, which limits the wavelengths of light entering the detector cells to
the area of the electromagnetic spectrum at which the pollutant molecule strongly
absorbs.
The reference cell is filled with a non-absorbing inert gas such as nitrogen or
argon. If no pollutant gas is present in the sample, the amount of radiation
reaching the reference half and the sample half of the detector will be equal;
therefore, the temperature (and resulting pressure) of both halves of the detector
will be equal. Thus, with no pollutant in the sample side, the diaphragm shown in
Figure 8-11 is in a null (a zero distention) position.
As pollutant gas enters the sample cell, the energy reaching the sample half of
the detector will be less, due to the absorption of energy by the pollutant. This
produces a distention of the diaphragm because the temperature (and resulting
pressure) is lowered on the sample side of the detector cell, (e.g., the pressure is
directly related to the temperature of a gas at constant volume). The distention of
8-16
-------
the diaphragm produces a capacitance change in the detector that is converted to a
voltage output by the instrument. The optical chopper simply creates an alter-
nating current (AC) signal that can be easily amplified.
This technique, while seemingly specific for CO, suffers from interferences due
to the broad band absorption of carbon dioxide, water vapor, and hydrocarbons.
Figure 8-11 shows the absorption of carbon monoxide and the broad band
interferences.
Infrared source
Chopper.
Reference cell
UJ
J*
^=
•
~5>
r*^-
W
'-*---
Detector
Recorder signal
Sample in
Sample out
Diaphragm
distended
Component of interest
Other molecules
Figure 8-10. Components and gas flow of a NDIR instrument.
8-17
-------
>*
•w
"w
a
•M
a
Carbon
monoxide •
(pollutant)
Interi'erents
(COj, water vapor, HC)
Wavelength
Figure 8-11. Overlap of carbon monoxide and broad band absorption curves.
If broad band interferences were present, the NDIR depicted in Figure 8-11
would yield a higher concentration than was actually present.
An NDIR instrument has been designed that minimizes these broad band
interferences. In this type of instrument, two detector cells are placed in series as
shown in Figure 8-12.
Sensor
Beam chopper
Sample in
Motor
Infrared source
Diaphragm
Reference cell _ _ _
Rear
Band pass Front measuring
filter measuring chamber
chamber
Detector
Figure 8-12. "Negative filter" analyzer.
8-18
-------
As light enters the sample cell, the gas absorbs energy in the CO spectral region.
The carbon monoxide in the front detector cell absorbs the reduced energy from
the sample cell, largely from the center frequencies of the CO band (see Figure
8 ISa) The rear measuring chamber absorbs the remaining energy (see Figure
8-13b) The longer cell path of the rear measuring chamber allows for more
absorption in the rear chamber. The resulting distention in the diaphragm, then, is
a ratio of the CO absorption in the front and rear measuring chambers. 11 the
interferences are largely linear across the CO absorption band, this ratio will
accurately reflect the CO concentration; that is, the absorbances in the front and
rear chambers wil icrease by the same amount due to interferences, but the ratio
will be constant regardless of -'
-------
Current Applications and Advantages
NDIR analyzers in ambient monitoring are used principally for the measurement of
carbon monoxide. This technique has been designated as the reference method for
the continuous measurement of carbon monoxide in air by the Environmental Pro-
tection Agency. This technique is a relatively inexpensive, reliable method that
does not require highly trained personnel for its operation.
Potential Interferents and Disadvantages
Since NDIR spectroscopy is most frequently used in the measurement of carbon
monoxide, the two most prominent interferents are carbon dioxide and water
vapor. "Negative filter" (22) analyzers discussed previously or more recent analyzers
using a palladium catalyst and a flowing reference cell all minimize spectral
interferences from CO2 and water vapor (23). The normal hydrocarbon levels in
the atmosphere are too small to cause possible interference in NDIR analyzers.
References
1. Shults, W. D. Coulometric Methods, p. 459-492. In Standard Methods of
Chemical Analysts, vol. 3, F. J. Welcher, ed. Princeton, NJ: D. Van Nostrand
Co., 1966.
2. Beckman Instruments, Inc.. Model 906A Sulfur Dioxide Analyzer, p. 43.
Beckman Instruments, Inc. Fullerton, CA 1972.
3. Philips Electronics Industries, Inc., Philips PW9700 SO{ Analyzer, Philips
Electronics Industries, Inc., Mahwah, NJ 15, 1975.
4. Lindqvist, Finn Galvanic Detection of Sulfur Dioxide in Ambient Air at Trace
Levels by Anodic Oxidation. /. Air Pollut. Control Assoc. 28: 138, 1978.
5. Hollowell, Craig D.: Gee, Glenn Y.; and McLaughlin, Ralph D. Current
Instrumentation for Continuous Monitoring for SO2. Anal. Chem. 45: 63A,
1973.
6. Hager, R. N. Second Derivative Spectroscopy as Applied to the Measurement
of Trace Nitrogen Dioxide. Presented at Eastern Analytical Symposium and
Society for Applied Spectroscopy National Meeting, New York, NY, 1973.
7. Staudner, R. Derivative Spectroscopy. Proc. Analyt. Div. Chem. Soc. 212: July
1976.
8. Hager, R. N. Derivative Spectroscopy with Emphasis on Trace Gas Analysis.
Anal. Chem. 45: 1131A, 1973.
9. Farwell, S. O., and Rasmussen, R. A. Limitations of the FPD and ECD in
Atmospheric Analysis: A Review. /. Chromatog. Sci. 14: 224, 1976.
10. Stevens, R. K., and O'Keefe, A. E. Modern Aspects of Air Pollution Monitor-
ing. Anal. Chem. 42: 143A, 1970.
11. Karasek, F. W. Fluorescence Spectrometry. Research/Development, January
28, 1976.
8-20
-------
12. Okabe, H.; Splitstone, P. L.; and Ball, J. J. Ambient and Source SO2 Detector
Based on a Fluorescence Method./. Air Pollut. Control Assoc. 28: 514, 1973.
13. Jahnke, J. A.; Cheney, J. L.; and Homolya, J. B. Quenching Effects in SO2
Fluorescence Monitoring Instruments. Environ. Set. Technol. 10: 1246, 1976.
14. Fontijn, A. Chemiluminescence Techniques in Air Pollution Monitoring. In
Modern Fluorescence Spectroscopy, vol. 1 ed. (Milton Birnbaum), pp. 159-192.
15. Nederbragt, G. W.; van der Horst, A.; and van Duyn, J. Ozone Determination
Near an Accelerator. Nature 206: 87, 1965.
16. Finlayson, B. J.; Pitts, J. N.; and Atkinson, R. Low-Pressure Gas-Phase Ozone-
Olefin Reactions. Chemiluminescence, Kinetics, and Mechanisms. /. Amer.
Chem. Soc. 96: 5356, 1974.
17. Environmental Protection Agency, Fed. Reg. 36 (228), 22384-97, Nov. 25,
1971.
18. Winer, A. M.; Peters, J. W.; Smith, P. J.; and Pitts, J. N. Response of Com-
mercial Chemiluminescence NO-NO2 Analyzers to Other Nitrogen-Containing
Compounds. Environ. Set. Technol. 6: 348, 1972.
19. Fontijn, A.; Sabadell, A. J.; and Ronco, R. J. Homogenous
Chemiluminescence Measurement of Nitric Oxide with Ozone. Anal. Chem.
42: 575-579, 1976.
20. Environmental Protection Agency, Fed. Reg., 43, (121), 26962-86.
21. Bowman, L. D., and Horak, R. F. A Continuous Ultraviolet Absorption Ozone
Photometer. Presented at the 18th Annual Symposium of the Analysis Instru-
ment Division of the Instrument Society of America, May 3-5, 1972.
22. Jahnke, J. A., and Aldina, G. J. Continuous Source Monitoring Systems Hand-
book. Prepared for ERIC under EPA contract #68-03-2561, 1978.
23. Houben, W. P. Continuous Monitoring of Carbon Monoxide in the
Atmosphere by an Improved NDIR Method. Beckman Instruments, Inc.,
Fullerton, California, 1978.
24. Green, A. S. The Middle Ultraviolet, Its Science and Technology, New York:
Wiley, 1966.
8-21
-------
Chapter 9
Design of Surveillance Networks
Elements of Surveillance Networks
All ambient air surveillance networks are composed of several major elements that
can be divided into two subsystems: a sensor subsystem and a data subsystem. The
data subsystem can be further divided into a data recording and transmission
system, and a data processing system. This chapter will describe the elements of
each subsystem, along with the basic considerations necessary for overall design. It
should be remembered throughout this process that designing an ambient air
surveillance network is a complex process requiring some experience. There are no
easy procedures to follow that will eliminate all potential problems. This chapter
will introduce you to a basic approach and will provide you with sufficient infor-
mation to design basic air monitoring networks.
Defining Network Uses and Objectives
Before beginning considerations for system components for your surveillance net-
work, the ultimate uses of the data must first be identified. The design of an effec-
tive network must establish these uses and employ them throughout the design
process. For example, several of the major uses for sulfur dioxide (SO2) data are:
1. Judging attainment of SO2 NAAQS;
2. Evaluating progress in achieving/maintaining the NAAQS or State/local
standards;
3. Developing or revising State Implementation Plans (SIPs) to attain/maintain
NAAQS; evaluating control strategies;
4. Reviewing new sources;
5. Establishing baseline air quality levels for preventing significant deterioration
(PSD) and for air quality maintenance planning (AQMP);
6. Developing or revising national SO2 control policies, e.g., new source perfor-
mance standards (NSPS), tall stacks, supplementary control systems (SCS);
7. Providing data for model development and validation;
8. Providing data to implement the provisions of the Energy Supply and
Environmental Coordination Act (ESECA) of 1974;
9. Supporting enforcement actions;
10. Documenting episodes and initiating episode controls;
11. Documenting population exposure and health research;
12. Providing information to:
a. public —air pollution indices; and
b. city/regional planners, air quality policy/decision makers— for activities
related to programs such as air quality maintenance planning (AQMP),
prevention of significant deterioration (PSD), and the preparation of
environmental impact statements. (1)
9-1
-------
These uses are broad and program oriented and, therefore, serve to delineate the
direction of the whole program. The second step in the design process is to
generate a list of specific objectives from the above design criteria. To continue the
example of SO2, some of the specific objectives and the associated data uses are as
follows (data use numbers refer to above list of SO2 data uses).
1. Determination of peak concentrations in urban areas
• major data uses: 1, 2, and 3.
• other data uses: 8, 9, and 12.
2. Determination of the impact of individual point sources in multi-source urban
settings
• major data uses: 3, 4, 6, 8, and 9.
• other data uses: 12.
3. Determination of the impact of isolated point sources
• major data uses: 3, 4, 6, 8, and 9.
• other data uses: 5 and 12.
4. Assessment of interregional SO2 transport
• major data uses: 2, 3, 5, and 12.
5. Determination of base concentrations in areas of projected growth
• major data uses: 5 and 12.
6. Initiation of emergency episode abatement actions
• major data uses: 10 and 12.
7. Determination of population exposure
• major data uses: 11 and 12.
8. Assessment of background concentrations in rural areas
• major data uses: 5 and 12.
• other data uses: 2 and 3. (2)
Having identified specific objectives, there are some additional general considera-
tions that must be discussed before beginning the actual design of the three sub-
systems. It should be remembered, however, that these objectives should be the
basis for the total design of the surveillance network. They should be prioritized on
the basis of several points, including legal requirements, topography, population
data, source characteristics, administrative and political concerns, and of course,
cost. Some of these topics will be covered in the section following on specific siting
procedures. By organizing these objectives by priorities, it should be possible to
minimi/.e expense while still meeting the identified needs.
Sensor Subsystem
Having established the why of our monitoring network, we can now begin to choose
a sensor system. The first step in choosing an appropriate sensor is to identify what
is to be monitored. If we are monitoring for more than one pollutant, it may be
necessary to have more than one type of shelter, probe, and monitoring location.
For example, measurements made in a street canyon may be valid for determining
maximum CO concentrations, but probably are not valid for regional SO2
measurements. This problem centers around the spatial scale of representativeness
of a station.
9-2
-------
It is important to be aware of the distinction between the spatial scale desired to
be represented by a measurement, and the spatial scale actually represented by
that measurement. The spatial scales and the corresponding dimensions
characterizing each are:
• Microscale. Ambient air volumes with dimensions ranging from meters up
to about 100 meters are associated with this scale.
• Middle Scale. This scale represents dimensions of the order from about 100
meters to 0.5 kilometers and characterizes areas up to several city blocks
in size.
• Neighborhood Scale. Neighborhood scale measurements would characterize
conditions over areas with dimensions in the 0.5 to 4.0 kilometer range.
• Urban Scale. Urban scale measurements would be made to represent condi-
tions over areas with dimensions on the order of 4 to 50 kilometers.
• Regional Scale. Conditions over areas with dimensions of as much as hun-
dreds of kilometers would be represented by regional-scale measurements.
These measurements would be applicable mainly to large homogeneous areas,
particularly sparsely populated areas.
• National and Global Scales. These measurement scales represent concentra-
tions characterizing the nation and the globe as a whole (3).
These dimensions are approximate but provide useful divisions for defining the
final site. Within each scale, there may be more than one site required to define
the area: for pollutants that have large, individual sources; for areas with steep
concentration gradients (concentrations changing rapidly over relatively short
distances); for reactive pollutants. The concentration gradients would be deter-
mined from previous sampling or dispersion modeling.
The choice of "what" will be monitored should be based on the needs, uses, and
objectives that we defined. At this stage, we should prioritize these needs based on
the individual considerations of each case. Each case will represent a unique situa-
tion. Some common considerations include:
• Pollutants with legal standards;
• Available resources, material, and personnel;
• Political factors;
• Problem pollutants - expected hazards.
The importance given each of these factors will vary with each case.
Monitoring Methods
The completion of this task leads to the next step in designing the sensor sub-
system—how are we going to monitor? If a criteria pollutant is being monitored,
one of the EPA Federal reference methods should be considered. These are
included in the appendix of this manual. Also available from EPA* is the "List of
Designated Reference and Equivalent Methods," which provides a current listing of
all the methods for measuring ambient concentrations of specified air pollutants
*U.S. Environmental Protection Agency
Office of Research and Development
Environmental Monitoring and Support Laboratory
Research Triangle Park, NC 27711
9-3
-------
'hat have been designated "reference or equivalent" methods in accordance with
1-0 CFR 53. There are cases where you will be required to use a reference or
equivalc-iu method such as when monitoring for State Implementation Plans (SIPs)
or most monitoring required under the Prevention of Significant Deterioration
(P'.D) regulations (40 CFR§51.24). If use of the EPA Federal reference or
equivalent methods is not required, it is important to ascertain if the method of
sampling is specified by a law or statute. If the data is to be used in legal pro-
ceedings, this fact may be of considerable importance. Other sources of analytical
methods exist including the NIOSH Manual of Analytical Methods and the Annual
\STM Standards series.
If using an automated or instrumental method, maintenance and calibration are
important. It can be difficult to estimate time and money costs for these items, but
there are sources of information available. For maintenance, the manufacturer
should be able to provide some estimates. It has been noted that these estimates
can be somewhat optimistic. Another source may be others who are or who have
used the instrument. Experience can be an accurate, if time consuming, instructor.
There may even exist previous in-house maintenance records that will provide
valuable information.
For calibration requirements, again the manufacturer should provide some
estimates. However, there may be legal requirements such as those specified in the
appendixes to 40 CFR 58 detailing minimum requirements for PSD or SIP
monitoring. Calibration guidance is also contained in "Quality Assurance Hand-
book for Air Pollution Measurement Systems, Volumes I and II,
[EPA-600/9-76-005 (Vol. I) and EPA-600/4-77-027a (Vol. II)]. The time
requirements for calibration and the frequency of audits may vary as in-house
records establishing performance are developed. For example, with a new instru-
ment, calibration may be weekly for the first 3 months; then biweekly with weekly
audits; and finally monthly with biweekly audits as performance parameters are
documented for that particular instrument. Of course, detailed records of
maintenance and calibration must be maintained.
Siting
Having established "why," "how," and "what," the next step in designing the sensor
system is "where." This topic involves properly siting the sensor to collect represen-
tative data. Since 1975, the EPA has issued several documents detailing a pro-
cedure for selecting the proper site for a pollutant for different purposes. By
following the procedure outlined in these documents, it is possible to logically select
a proper site for a particular purpose and to clearly document the procedure
followed. This allows for a subsequent audit or review of the site selection process
to react to changing conditions without having to duplicate earlier work. The
documents available are:
"Selecting Sites for Carbon Monoxide Monitoring."
September 1975. EPA-450/3-75-077.
"Optimum Site Exposure Criteria for SO2 Monitoring."
April 1977. EPA-450/3-77-013.
9-4
-------
"Selecting Sites for Monitoring Total Suspended Participates."
June 1977. EPA-450/3-77-018.
"Site Selection for the Monitoring of Photochemical Air Pollutants."
April 1978. EPA-450/3-78-013.
All of these siting documents follow the same approach. First the needs or uses of
the data are identified, then these uses are coordinated with specific program
objectives. After these steps, a flow chart is developed for each siting objective.
It should also be noted that each siting case may be unique and the "proper site"
is not necessarily an absolute, fixed item. Given the same background information
and a similar siting objective, two "experts" may arrive at different site locations.
Often there is a lack of needed information, and even under the best of conditions,
siting sensors involves the abstract correlation of large amounts of chemical,
meteorological, geographical, and political information. There are no "instant
experts" or short cuts to actual experience. However, by following a set of logical
procedures, it is possible to arrive at a site location that will adequately meet the
program objectives and allow for concise documentation of the selection process.
This is important especially in view of the proposed revision to the ambient
monitoring regulations.
Section 40 CFR§58.20 requires an annual review of each surveillance network.
These changes to the EPA monitoring regulations were promulgated May 10, 1979.
In addition'to organizing and revising almost all of the EPA ambient monitoring
regulations into one section of the Code of Federal Regulations, there is also siting
and network design guidance contained in Appendix D of these regulations. The
guidance in Appendix D is consistent, both in approach and content, with that
contained in the individual siting documents. The documents contain more detail
and greater depth than is possible in the regulations.
The location of the sensor is only part of the answer to the where of monitoring.
Besides locating the sensor, it is also important to obtain a correct probe location.
The location of the probe depends on the chemistry of the specie sampled and, of
course, the monitoring objectives. Guidance on probe location is contained in the
siting documents and also in Appendix E of 40 CFR§58. By adopting a uniform
probe location for each pollutant, comparison of the data from different projects
and times is simplified. As mentioned earlier foi sensor location, the probe location
may vary for each pollutant and each objective. The final choice would represent
an optimal compromise between needs and resources.
As part of our network design system, we have specified the why, how, what, and
where of our sensor package and network. There are two remaining systems of the
data subsystem to define: the recording and transmission system, and the data pro-
cessing system. Both of these systems are concerned with the collection and pro-
cessing of the data or signals produced by the sensor systems, and there is some
overlapping. It is important to spend time on properly planning each of these, as a
great deal of resources have been expended needlessly in these areas. We will work
with each separately.
9-5
-------
Data Subsystem
Data Recording
The first part of this subsystem concerns data recording. We may have an electrical
signal from one or more instruments from one or more locations, or a set of
analytical results from manual methods. In either case, the data must be recorded.
If we have an electrical output from an analyzer, several choices are available. We
can manually read a meter and record data, use a strip chart recorder, record on
punch cards or paper tape, or record on machine-readable magnetic tape or discs.
When designing data systems, the major problem is usually optimizing man and
machine interactions. In some cases you may have excess manpower and little
automation, while the opposite may be true in other cases. Therefore, the first step
in this process is to identify the needs. Is data required or produced continuously,
are all outputs from analyzers matched, is there a mix of manual and automated
sensor systems, and finally is the data for immediate or future use? If the data is to
be used for a daily Pollutant Standards Index or to operate a Supplemental Con-
trol System, there may be a need for automated recording and processing.
However, if the data is to be used for annual trend reports, monthly strip charts
may be the best choice.
With the needs identified, resources available should be assessed. This assessment
allows a realistic data system to be designed. The goal here again should be to
meet the program requirements while minimizing costs. Too often overly elaborate
systems have been designed with no consideration of needs. Needs must be
prioritized, separating necessary items from desirable, but less essential com-
ponents. The total mix of components in a system must also be accounted for in
the design. If grounding and interface problems exist, it may be advantageous to
solve these separately and use one data logger for each station rather than separate
strip charts or other separate recorders. Unfortunately, no hard and fast rules exist
to assist in making these decisions. Advice should be sought from competent
systems engineers or other personnel familiar with the equipment involved.
Data Transmission
The same situation exists for the transmission part of this subsystem. It is impor-
tant not to overdesign the system and thus waste resources. In reviewing the data
needs, speed of utilization must be determined. If it arrives quickly, but is not
going to be processed for a month, a simpler system may be more cost effective.
The best rule to follow is to make it as simple, cheap, and uncomplicated as is
feasible. The choice between on-site pickup of data versus telemetered data will be
based on data needs, resources available, and the existing network. Signal condi-
tioning must be designed by a competent electrician to avoid later problems.
Data processing comprises the final part of the Data System. The data processing
system involves the selection of a format, validation, analysis, storage, and
retrieval. A format is simply a systematic listing of the data recorded. The format
chosen should be clearly defined and well documented. This prevents the problem
of reconciling incompatible formats later, a process that can be extremely expen-
9-6
-------
sive.
ai^. If the data is to be submitted to another office, identical formats would be
desirable. If no standard format exists, a logical, easily understood format should
be designed. The EPA has available a computer program, the Comprehensive Data
Handling System, which processes aerometric data and puts it into acceptable EPA
(SAROAD) format. This software program is offered free to governmental agencies
and for a nominal copying charge (<$100) to others from:
National Air Data Branch
OAQPS/MDAD (MD-14)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
Data Validation
After choosing a format to use for the data that provides a copy, validation can
begin. The validation should be performed by an air pollution professional with
intimate knowledge of the sensors, recording system, transmission system, as well as
any peculiarities of the air pollution problem in the area. This validation should be
performed at the lowest level possible to permit the maximum possible recovery of
valid data. For example, if there is a two week period in a month's data that has
values three times the "normal" values, several possibilities exist. The data may
have been analyzed by a substitute while the usual person was on vacation and a
factor of three may have been dropped from the denominator of an equation; the
instrument may have malfunctioned; meteorological conditions (poor dispersion,
stable atmosphere, low mixing depth) may have resulted in increased concentra-
tions. Obviously, in the first case the data can be corrected by including the factor
and recomputing; in the second case the data probably will have to be invalidated
unless an accurate correction factor can be determined; and in the third case, the
data is correct. Only an air pollution professional would recognize the difference,
and the closer that person is to the network, the less expensive this process will be.
Computer validation is possible to identify outliers, or abnormal values, but
manual checking is needed to perform final validation.
Data Analysis
After the data has been formated and validated, the final analysis can be per-
formed. Depending on the nature of the data and the final data use, a variety of
statistical procedures exist. These techniques and their use with aerometric data are
complex and a professional statistician should be consulted. The type of analysis
may be dictated by the standard (requiring arithmetic or geometric means), or the
data (long term trends represented by log-normal distributions). Advice from a
statistician can help prevent erroneous conclusions.
Storage of the data and/or the samples should prevent physical deterioration,
provide a logical system, and should be secure to provide a legal chain of custody.
If using thermally sensitive chart paper, storage in a warehouse subject to elevated
daytime temperatures would prove problematic and soon the records would be
uniformly gray. A logical system provides continuity and, along with a secure
storage area and sign-off sheets, helps establish a legal chain of custody for records.
9-7
-------
The storage system should not, however, prevent easy access to the data by the
proper personnel. Retrieval procedures should not be overly complex, yet must pro-
vide a systematic procedure ensuring data security. In automated systems, provi-
sions can be made to restrict access with confidential identification codes or
restricted data discs.
After designing the sensor and data subsystems, one final item requiring alloca-
tion of resources exists. This is to incorporate a comprehensive quality control pro-
gram to assure the data produced is correct. Guidance on these programs is con-
tained in the Quality Assurance Handbooks mentioned earlier and in Appendixes
A and B of 40 CFR§58. Without adequate quality control of each subsystem, the
question of ultimate data validity may be impossible to answer.
References
1. Ball, R. J., and Anderson G. E. Optimum Site Exposure Criteria for SO2
Monitoring. EPA-40/3-77-013, US EPA, RTF, NC, 1977.
9-8
-------
Chapter 10
Statistical Techniques Employed
in Atmospheric Sampling
Introduction
Proper use of statistics and statistical techniques is necessary for evaluating air
pollution data. This chapter serves as an introduction to statistics and statistical
concepts. Topics to be discussed include: (a) Measures of Central Tendency;
(b) Measures of Dispersion; (c) Distribution Curves; and (d) Data Plotting.
Although the above topics are not simple, they can be understood and used by
nonstatisticians. If a detailed statistical analysis of data is required, it is recom-
mended that an experienced statistician be consulted.
Data Plotting
Data is usually unmanageable in the form in which it is collected. In this section
we shall consider the graphical techniques of summarizing such data so that the
meaningful information can be extracted from it. Basically there are two kinds of
variables to which we assign data: continuous variables and discrete variables. A
continuous variable is one that can assume any value in some interval of values.
Examples of continuous variables are weight, volume, length, time and
temperature. Most air pollution data are taken from continuous variables. Discrete
variables, on the other hand, are those variables whose possible values are integers.
Therefore, they involve counting rather than measuring. Examples of discrete
variables are the number of sample stations, number of people in a room, and
number of times a control standard is violated. Since any measuring device is of
limited accuracy, measurements in real life are actually discrete in nature rather
than continuous, but this should not keep us from regarding such variables as con-
tinuous. When a weight is recorded as 165 pounds, it is assumed that the actual
weight is somewhere between 164.5 and 165.5 pounds.
Graphical Analysis
Frequency Tables
Let us consider the set of data in Figure 10-1 which represents SO2 levels for a
given hour for 25 days. As a first step in summarizing the data from Figure 10-1,
we would form a frequency table as shown in Figure 10-2.
10-1
-------
Days
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
SO2 Concentration (ppb)*
53
72
59
45
44
85
77
56
157
83
120
81
35
63
48
180
94
110
51
47
55
43
28
38
26
* ppb = parts per billion collected SO2 levels
Figure 10-1. SO* levels.
Class Interval (ppb)
25- 40
40- 55
55- 70
70- 85
85-100
100-115
115-130
130-145
145-160
160-175
175-190
Frequency of Occurrence
4
7
4
4
2
1
1
0
1
0
1
Figure 10-2. Frequency table.
10-2
-------
In constructing the frequency table it can be seen that we have divided the data
into 11 class intervals with each interval being 15 units in length. The choice of
dividing the data into 11 intervals was purely arbitrary. However, in dealing with
data it is a rule of thumb to choose the length of the class interval such that 8-15
intervals will include all of the data under consideration. Deriving the frequency
column involves nothing more than counting the number of values in each interval.
From observation of the frequency table, we can now see the data taking form.
The values appear to be clustered between 25 and 85 ppb. In fact, nearly 80 per-
cent are in this interval.
The Frequency Polygon
As a further step we can graph the information in the frequency table. One way of
doing this would be to plot the frequency midpoint of the class interval. The solid
line connecting the points of Figure 10-3 forms a frequency polygon.
r 4
32.5 62.5 92.5 122.5 152.5 182.5
Figure 10-3. Pollution concentration (mid-point of class interval) frequency polygon.
The Histogram
Another method of graphing the information would be by constructing a histogram
as shown in Figure 10-4. The histogram is a two-dimensional graph in which the
length of the class interval is taken into consideration. The histogram can be a very
useful tool in statistics especially if we convert the given frequency scale to a
relative scale so that the sum of all the ordinates equals one. This is shown in
Figure 10-5. Thus, each ordinate value is derived by dividing the original value by
the number of observations in the sample, in this case 25. The advantage in con-
structing a histogram like the one in Figure 10-5 is that we can read probabilities
from it, if we can assume a scale on the abscissa such that a given value will fall in
10-3
------- An error occurred while trying to OCR this image.
------- An error occurred while trying to OCR this image.
-------
The cumulative frequency table gives the number of observations less than a given
value. Probabilities can be read from the cumulative frequency curve. For exam-
ple, to find the probability that a value will be less than 85 we read up to the curve
at the point X = 85 and across to the value 0.76 on the Y axis.
Distribution of Data
When we draw a histogram for a set of data we are representing the distribution of
the data. Different sets of data will vary in relation to one another and, conse-
quently, their histograms will look different. Basically, there are three
characteristics that will distinguish the distributions of different sets of data. These
are central location, dispersion, and skewness. These are characterized in Figure
10-8. Curves A and B have the same central location, but B is more dispersed.
However, both A and B are symmetrical and are, therefore, said not to be skewed.
Curve C is skewed to the right as well as having a different central location than A
and B. Mathematical measures of central location and dispersion will be discussed
later.
-4
-2
024
Standard unit scale
Figure 10-8. Relative frequency distribution showing: Curve A and B both centrally located;
Curve B being more disperse than Curve A and the skewness of Curve C.
Transformation of Data
In most statistical work, it is required to have data that closely approximate a
particular symmetrical curve called the normal curve. Both curves A and B in
Figure 10-8 are examples of normal curves. In dealing with skewed curves, such as
C in the same figure, it is desirable to transform the data in some way so that a
symmetrical curve resembling the normal curve results. Referring to the frequency
table (Figure 10-2) and histogram (Figure 10-4) of the data used earlier, it can be
seen that for this set of data the distribution is skewed in the opposite direction as
curve C above.
10-6
-------
The Logarithmic Transformation
One of the most successful ways of deriving a symmetrical distribution from a
skewed distribution is by expressing the original data in terms of logarithms. The
logarithms of the original data are given in Figure 10-9.
Day
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Pollutant Cone. X
53
72
59
45
44
85
77
56
157
83
120
81
35
63
45
180
94
110
51
47
55
43
28
58
26
LogioX
1.724
1.857
1.771
1.653
1.644
1.929
1.887
1.748
2.196
1.919
2.079
1.909
1.544
1.799
1.681
2.255
1.973
2.041
1.708
1.672
1.740
1.634
1.447
1.580
1.415
Figure 10-9. Logarithmic transformation.
Arbitrarily dividing the logarithmic data into 9 class intervals each of 0.1 unit in
length we can write the logarithmic frequency table in Figure 10-10. As can be
seen a frequency plot versus the logarithmic scale would more closely approximate
a symmetrical curve than would the arithmetic plot.
Class interval
1.4-1.5
1.5-1.6
1.6-1.7
1.7-1.8
1.8-1.9
1.9-2.0
2.0-2.1
2.1-2.2
2.2-2.3
Tally
11
11
1111
1111 1
11
1111
11
1
1
Frequency
2
2
5
6
2
4
2
1
1
Cumulative
frequency
2
4
9
15
17
21
23
24
25
Figure 10-10. Logarithmic frequency table.
10-7
-------
Probability Graph Paper
Probability graph paper is used in the analysis of cumulative frequency curves; for
example, the graph paper can be used as a rough test of whether the arithmetic or
the logarithmic scale best approximates a normal distribution. The scale,
arithmetic or logarithmic, on which the cumulative frequency distribution of the
data is more nearly a straight line is the one providing the better approximation to
a normal distribution. Plotting the cumulative distribution curve of the data above
on the two scales, it can be seen that the logarithmic scale yields the better fit
(Figure 10-11).
These probability plots can also be used to get estimates of the mean and stan-
dard deviation of the data. The estimate of the mean, as will be shown later, is the
50th percentile point, and the estimation of the standard deviation is the distance
from the 50th percentile to the 16th percentile.
O Log plot
A Arithmetic plot
H
Cfl
o1
i
o
M^
O
10
20 30 40 50 60 70 80
Cumulative frequency
90 95
Figure 10-11. Normalized data plot vs. non-transformed data.
10-8
-------
Least Squares Linear Regression
If the relationship between two variables is significant, a linear regression line or
line of "best fit" may be drawn to represent the data. Algebraically a straight line
has the following form
(Eq. 10-1)
Where:
= mx
y - variable plotted on the ordinate
x = variable plotted on abscissa
b=y intercept
m = slope = change in y/ change in x
Linear regression minimizes the vertical distance between all data points and the
straight line (see Figure 10-12).
This distance minimized
for all points
Figure 10-12. Linear regression curve.
The constants m and b for the "best fit" line can be determined using the fol-
lowing equations:
(Eq. 10-2)
(Eq. 10-3)
Where:
m =
n
Lx*
(Lx)*
n
b=y — mx
n = number of observations
"y — Ey/n; x — la: //
10-9
-------
Example Problem
The following data was collected during a calibration of a chemiluminescent NO,
analyzer.
X = Concentration NO, (ppm)
Y = Instrument response (volts)
0.05
1.20
0.10
2.15
0.20
3.90
0.30
6.20
0.45
9.80
Values for m and b for the "best fit" line can be calculated from Ex, I/y, Ex2,
Lxy, n, y, and x.
Solution:
Ex = 0.05 -I- 0.10 + 0.20 + 0.30 + 0.45 =1.1
E}>=1.20 f 2.15 + 3.90 + 6.20 + 9.80 = 23.25
Ex2 - (0.05)2 + (0.10)2 + (0.20)2 + (0.30)2 + (0.45)2 = .345
Exy = (0.05)(1.20) + (0.10X2.15) + (0.20)(3.90) + (O.SO)(6.20) + (0.45)(9.80) - 7.33
n= 5
1.1
_ Ex
x = —
n
-=0.22
E-y 23.25
Y = —- =
y n 5
1.65
7.33-
(1.1X23.25)
m =
0.345 -
(l-l)2
2.22
0.103
-=21.6
6 = 4.65-(21.6)(0.22)= -0.102
The equation for this calibration curve would be
y = 21.6*-0.102
Volts = 21.6 (ppm)-0.102
Rearranging
volts + 0.102
10-10
-------
Measures of Central Tendency
Arithmetic Average or Mean
The most basic way of summarizing data is the computation of a central value.
The most commonly used central value statistic is the arithmetic average or the
mean. This statistic is particularly useful when applied to a set of data having a
fairly symmetrical distribution. The mean is an efficient statistic in that it sum-
marizes all the data in the set, because each piece of data is taken into account in
its computation. The formula for computing the mean is
(Eq. 10-4) 2 X,
x=
n n n
Where: X = arithmetic mean
X, = any individual measurement
n - total number of observations
X^Xz.Xs, = measurements 1, 2, and 3 respectively
The arithmetic mean is not a perfect measure of the true central value of a given
data set. Arithmetic means overemphasize the importance of one or two extreme
data points. Many measurements of a normally distributed data set will have an
arithmetic mean that closely approximates the true central value.
Median
When a distribution of data is asymmetrical, such as that of Figure 10-13, it is
sometimes desirable to compute a different measure of central value. This second
measure, known as the median, is simply the middle value of a distribution, that
is, the quantity above which half the data lie and below which the other half of the
data lie. If n data are listed in their order of magnitude, the median is the
\(N+ l)/2]th value. If the number of data is even, then the numerical value of the
median is the value midway between the two data nearest the middle. The median,
being a positional value, is less influenced by extreme values in a distribution than
the mean. The median alone is not a good measure of central tendency.
Example:
Find the median of 22, 10, 15, 8, 13, 18.
Solution: The data must first be arranged in order of magnitude such as:
8, 10, 13, 15, 18, 22
Now the median is 14, or the value half way between 13 and 15, since this data
set has an even number of measurements.
10-11
-------
2
rt
u
u
, Median
Mean
Magnitude of data
Figure 10-13. Example of nonsymmetrical distribution of data (median vs. mean).
Geometric Mean
Another measure of central tendency used in more specialized applications is the
geometric mean (Xg). The geometric mean can be calculated using the following
equation:
(Eq. 10-5) Xg =
If scientific calculators are not available, a formula that more readily lends itself
to a four function calculator is
Logi0Xg= — £ LogloX,
n
The geometric mean is most often used for data whose causes behave exponen-
nally rather than linearly, such as in the growth of bacteria, measurements that are
i.uios, or lognormal distributions.
In a distribution shaped like that of Figure 10-13, the geometric mean, like the
,, •dian, will yield a value closer to the main cluster of values than will the mean.
1 lie arithmetic mean is always higher than the geometric mean.
10-12
-------
Example
Calculate the geometric mean of 3.0, 2.5, 2.2, 3.4, 3.2
Solution:
= ^/(3.0)(2.5)(2.2)(3.4)(3.2) = 2.
or
LogloXg = —(.477 + .398 + .342 + .531 + .505)
5
Log10X,= 0.4506
Measures of Dispersion
It is, of course, quite possible for two separate distributions of data to have equal
means and yet be considerably different. This fact is depicted in Figure 10-14.
Distribution A has data that are bunched relatively much closer to the mean than
the data of distribution B. Distribution B is said to have greater variability or
dispersion than A. Thus, it is apparent that, instead of drawing a diagram each
time one wishes to compare distributions, a statistic that measures variability is
desirable.
o
V
Distribution A
Distribution B
Quantity variations
Figure 10-14. Dispersion characteristic curves.
10-13
-------
Standard Deviation
The most commonly used measure of dispersion, or variability, of sets of data is the
standard deviation. Its defining formula is given by the expression:
(Eq. 10-6)
5=
Where: s = the standard deviation (always positive)
X, = a datum value
X-the mean of the data sample
n = the number of observations
The expression (X-X) shows that the deviation of each piece of data from the
mean is taken into account by the standard deviation.
Although the defining formula for the standard deviation gives insight into its
meaning, this algebraically equivalent formula makes computation much easier:
nEX,2-(EX,)2
n(n — \)
When using this formula, it is only necessary to array the data in a column, con-
struct another column of squared data, and find the sums for both columns. (Note
that EX2 does not equal (EX)2).
Suppose one wishes to compute the mean and standard deviation of 220 data
points. After the computation, it is found that one of the data points has been
omitted. Unless the omitted value exactly equals the obtained mean, the mean of
the complete set of data will be different from the one obtained; therefore, the
deviations of the data from the mean will all be changed, and will necessitate
untold additional laborious computation. But with the computational formula for
5, it is only necessary to add the omitted value to the X column and its square to
the X2 column and plug the corrected values into the computational formula. The
computational formula is also handy for desk calculators, as they are designed to
keep running totals of numbers and their squares simultaneously. Also, no errors
inherent in the computation are introduced, as when one obtains a mean such as
3.333.
The Range
Despite the existence of a computational formula for the standard deviation, the
easiest way to obtain a measure of dispersion of a set of data is to find the dif-
ference between the maximum and the minimum values in the set. This simple
measure is appropriately termed the range. Obviously the range does not make full
use of the information contained in the data, since only two of the data points are
taken into account. The range is, however, reasonably efficient compared to s
when the number of data is 10 or less. Thus the range is an extremely handy
measure of variability for small samples.
10-14
-------
Standard Geometric Deviation
Dispersion of skewed data such as lognormal distributions is better measured by the
standard geometric deviation. The standard geometric deviation is very similar to
the standard deviation. The dispersion in the log of the measurements is measured
by the geometric standard deviation instead of the dispersion of the measurements.
The log calculation normalizes the data to better approximate a normal distribu-
tion. The formula for calculating the standard geometric deviation is:
(Eq.10-7)
The following formula is mathematically identical, yet it is much easier to use in
calculation:
;- (LlogX)*'
*-i n
sg = antilog
n-1
Distribution Curves
Many types of distribution curves exist; binomial, t, chi, F, normal, and lognormal
are just a few of the existing distributions. However, in air pollution measurements,
the normal and lognormal are the most commonly occuring ones. Thus only these
two will be discussed.
The Normal Distribution
One reason the normal distribution is so important is that a number of natural
phenomena are normally distributed or closely approximate it. In fact many
experiments when repeated a large number of times will approach the normal
distribution curve. In its pure form, the normal curve is a continuous, symmetrical,
smooth curve shaped like the one shown in Figure 10-15. Naturally, a finite
distribution of discrete data can only approximate this curve. The normal curve
has the following definite relations to the descriptive measures of a distribution.
The Mean and Median
The normal distribution curve is symmetrical; therefore, both the mean and the
median are always to be found in the middle of the curve. Recall that in general
the mean and median of a nonsymmetrical distribution do not coincide.
The Range
The normal curve ranges along the x axis from minus infinity to plus infinity.
Therefore, the range of a normal distribution is infinite.
10-15
-------
u
V
a1
I Mean
Measurement
Figure 10-15. Gaussian distribution curve "normal curve".
The Standard Deviation
The standard deviation (s) becomes a most meaningful measure when related to the
normal curve. A total of 68.2 percent of the area lying under a normal curve is
included by the part from one standard deviation below to one standard deviation
above the mean. A total of 95.4% lies - 2 to +2 standard deviations from the
mean (see Figure 10-16). By using tables found in statistics texts and handbooks,
one can determine the area lying under any part of the normal curve.
-2s
Figure 10-16. Characteristics of the Gaussian distribution.
10-16
-------
These areas under the normal distribution curve can be given probability inter-
pretations. For example, if an experiment yields a nearly normal distribution with
mean equal to 30 and standard deviation of 10, we can expect about 68% of a
large number of experimental results to range from 20 to 40, so that the prob-
ability of any particular experimental result's having a value between 20 and 40 is
about 0.68.
Applyi.ig the propeities of the normal curve to the testing of data readings one
can determine whether a change in the conditions being measured is shown or
whether only chance fluctuations in the readings are represented. For a well-
established set of criterion data, a frequently used set of control limits is plus and
minus three standard deviations. That is, a special investigation of data readings
trying these limits can be used to determine whether the conditions under which
the original data were taken have changed. Since the limits of three standard
deviations on either side of the mean include 99.7 percent of the area under the
normal curve, it is very unlikely that a reading outside these limits is due to the
conditions producing the criterion set of data. The purpose of this technique is to
separate the purely chance fluctuations from the other causes of variation. For
example, if a long series of observations of an environmental measurement yield a
mean of 50 and a standard deviation of 10, then control limits will be set up as 50
plus and minus 30, in other words, plus and minus three standard deviations, or
from 20 to 80. So, a value of 80 would suggest that the underlying conditions have
changed, and that a large number of similar observations at this time would yield a
distribution of results with a mean different (larger) than 50.
This process of determining whether a value represents a significant change is
closely related to the use of control charts. Often in setting up control limits it is
necessary to divide the available data into subgroups and calculate the mean and
standard deviations of each of these groups, making careful note of the conditions
prevailing under each subgroup. In collecting data to establish control limits, the
time sequence can be important, for the data in a time series may reveal some non-
random patterns. It is apparent that as much information as possible should be
gathered about the causes a,nd conditions in effect during the period of obtaining a
criterion set of data. Generally the conditions during this period should be "nor-
mal" or as much in control as possible.
In the situation where one takes readings of some environmental quantity, the
appearance of data beyond the control limits might suggest the starting of a new
data grouping to further ascertain whether the underlying environmental variable
has changed.
It should be kept in mind that the limits of plus and minus three standard devia-
tions are traditional rather than absolute. They have been found through
experience to be very useful in many control situations, but each experimenter
must decide what limits would be most suitable for a given purpose by determining
what levels of probability would be needed to quantify acceptance and rejection
bounds.
10-17
-------
Lognormal Distributions
Lognonnal distributions can best be demonstrated by means of an example.
If hourly sulfur dioxide concentrations are plotted against frequency of occur-
rence, as in Figure 10-15, a skewed distribution would exist similar to the one in
Figure 10-17. Such a curve indicates that many concentrations are close to zero and
that lew arc very high. Unlike temperature, or wind direction, sulfur dioxide con-
ci lit atious are blocked on the left because values less than zero do not exist.
Because numerous aids exist for normal distributions it is desirable to normalize
tins : /pc of distribution. By plotting the log of hourly SO2 concentrations against
i Ue irequency of occurrence, a "bell-shaped" curve similar to Figure 10-15 is
obtained. By making this simple normalizing feature, all existing normal distribu-
tion tables can be used to make probability interpretations.
Hourly SO2 concentration
Figure 10-17. Frequency vs. concentration of SO2.
10-18
-------
References & Extra Reading
Quality Control Practices in Processing Air Pollution Samples. APTD 1132,
U.S. Environmental Protection Agency, March, 1973.
Bennett, C. A., and Franklin, N. L. Statistical Analysis in Chemistry and the
Chemical Industry. New York: Wiley, 1954.
Dixon, W. J., and Massey, F. J., Jr. Introduction to Statistical Analysts. New
York: McGraw-Hill, 1957.
Spiegel, Murray R. Statistics, Shaum's Outline series. New York: McGraw-Hill
Book Co., 1961.
Hayslett, H. T. Statistics Made Simple—A Comprehensive Course Jor SelJ-
Study and Review. New York: Doubleday & Company, Inc., 1968.
Neville, A. M., and Kennedy, J. B. Basic Statistical Methods for Engineers and
Scientists. Scranton, PA: International Textbook Company, 1964.
Cadle, R. D. The Measurement of Airborne Particulates, pp. 13-43. New
York: Wiley-Interscience, 1975.
10 19
-------
Appendixes
Appendix 1. Theory and Calibration Procedures for a Rotameter Al- 1
Nomenclature A1" £
Description of a Rotameter Al - 2
Development of Flow Equations Al - 2
Common Practices in the Use of a Rotameter for Gas Flow Measurement. . Al - 6
Appendix 2. Federal Register Reference Methods • • • A2-a
A —Reference Method for the Determination of Sulfur Dioxide in the
Atmosphere (Pararosaniline Method) A2- 1
B —Reference Method for the Determination of Suspended Particulates
in the Atmosphere (High Volume Method) A2- 5
C —Measurement Principle and Calibration Procedure for the
Continuous Measurement of Carbon Monoxide in the Atmosphere
(Non-Dispersive Infrared Spectrometry) A2' 9
D —Measurement Principle and Calibration Procedure for the
Measurement of Ozone in the Atmosphere A2-11
E —Reference Method for Determination of Hydrocarbons Corrected
for Methane A2-15
F—Measurement Principle and Calibration Procedure for the
Measurement of Nitrogen Dioxide in the Atmosphere (Gas Phase
Chemiluminescence) A2-17
G —Reference Method for the Determination of. Lead in Suspended
Particulate Matter Collected from Ambient Air A2-23
Appendix 3. Conversion Factors and Useful Information A3- 1
-------
Appendix 1
Theory and Calibration Procedures
for the Use of a Rotameter
Nomenclature
Af = cross sectional area of the float
Am = annular area between the circumference of the float and the inside circum-
ference of the meter tube at that position
C = drag coefficient
Cm = length which is characteristic of the physical system under study (used to
calculate Reynold's Number)
d = length which is characteristic of the physical system under study
D/= diameter of the float
D,= diameter of the tube at the float position
g = local acceleration due to gravity
gc = dimensional constant
m/= mass of the float
Mm = molecular weight of the metered gas
M,, M2l Ms. . .etc. = value of molecular weight of the metered gas at conditions
1, 2, 3. . .etc.
NRe = Reynold's Number
NRe/Cm = dimensionless factor defined by equation 14
Pm = absolute pressure of the metered gas
PI, PZ, PS .etc. = values of absolute pressure at conditions 1, 2, 3. . .etc.
Q^, = volumetric flow rate through the meter at conditions of pressure (Pm), tem-
perature (Tm), and molecular weight (Mm)
R = universal gas constant
Tm = absolute temperature of the metered gas
TI, T2, Ts. . .etc. = values of absolute temperature at conditions 1, 2, 3. . .etc.
v = average gas velocity through the annual area of the meter
V/= volume of the float
/i = viscosity of flowing fluid (used to calculate Reynold's Number)
/im = viscosity of the metered gas
Q = density of flowing fluid (used to calculate Reynold's Number)
Qf— density of the float
Qm = density of the metered gas
AM
-------
Description of a Rotameter
The rotameter (Figure A-l) is a variable area meter which consists of a vertical,
tapered, transparent tube containing a float. The float moves upward as the fluid
flow increases. A variable ring or annulus is created between the outer diameter of
the float and the inner wall of the tube. As the float moves upward in the tube,
the area of the annulus increases. The float will continue to move upward uiiiil a
pressure drop across the float which is unique for each rotameter is reached. This
pressure drop across the float is constant regardless of the flow ..i,<>. Graduations
are etched on the side of the tube so that an instantaneous reading may be
observed.
r
Float
Annulus
Float
Section A-A
B
t_
t
O
t
Y
A
Tapered tube
Annulus —
Float
Section B-B
B
Figure A-l. Rotameter
Development of Flow Equations
General Flow Rate Eqttations
A free body diagram of the forces acting upon the rotameter float is shown in
Figure A2. The weight of the float is equal to the force of gravity acting on the
float. The buoyant force is equal to the weight of the gas that is displaced by the
float. The drag force is equal to the frictional forces acting between the float and
the moving gas stream. Mathematically these forces are as follows:
Al-2
-------
Buoyant force
Drag force
Weight of float
Figure A-2. Forces acting upon a rotameter float.
Drag force —
Weight of float =
VfQfg
Buoyant force =
Where: As- cross sectional area of the float
C = drag coefficient
g = local acceleration due to gravity
:_; = dimensional constant
v = average gas velocity through the annular area of the meter
Vf —volume of the float
Qf- density of the float
Qm = density of the metered gas
When the forces acting in an upward direction exactly equal the force acting in
a downward direction the float will remain stationary in the tube. Equating these
forces yields:
CAfQmvz VfQmg _
2g;gc go
Cancelling like terms (gc) and rearranging yields:
Al-3
-------
Solving for v and factoring out V, and g from the first two terms yields:
"21
(Eq. A-l)
CAfQm
The area of the float is equal to irD//4 where Df is the diameter of the float.
Substituting 7TD/V4 for As in Equation A-l yields:
(Eq. A-2)
v —
CirD/Q,
Let Cm equal (8/CTr)1^ where Cm is called a meter coefficient and is dependent
upon the drag coefficient. Substituting Cm for (8/CTr)1^ in Equation A-2 yields:
(Eq. A-3)
g(e/~
D?Q,
Because the drag coefficient C is dependent upon Reynold's Number, Cm must
also be a function of Reynold's Number. Because the density of the gas flowing in
the rotameter is very small compared to the density of the float, it can be ignored
in the (Q(- Qm) term. Modifying the (Q,- Qm) term in Equation A-3 yields:
(Eq. A-4)
v = C»
The volumetric flow rate (Qm) through the rotameter is equal to the product of
the velocity (v) and the annular area of the meter (Am). Substituting QJAm for v
in Equation A-4 yields:
" r'ge/~T/2
Rearranging terms and removing D/ from the radical yields:
CmAm \l gQfV/2
(Eq. A-5)
Df
The density of the float Qf is equal to the mass of the float (mf) divided by the
volume of the float. Substituting ms/Vf for Qf in Equation A-5 and cancelling the
F/'s yields:
(Eq. A-6)
Df
gmf
Al-4
-------
The density of the gas mixture passing through the meter (Qm) is equal to
PmMjRTm where Pm is the absolute pressure at the meter, Mm is the apparent
nTolecular weight of the gas mixture passing through the meter, R is the universal
gas constant, and Tm is the absolute temperature of the gas mixture. Substituting
PmMjRTm for Qm in Equation A-6 yields the general flow rate equation for a
rotameter:
CmA
Computation of Reynold's Number
Reynold's Number is defined as vdQ/n where v is the velocity flow, d is a length
which is characteristic of the physical system under study, Q is the density of the
Rowing fluid, and n is the viscosity of the flowing fluid. When calculating
Reynold's Number for a gas flowing through a rotameter, the length characteristic
of the physical system (d) is the difference between the tube diameter (D,) and the
diameter of the float (D/). Therefore, Reynold's Number may be calculated by
using the following equation:
v(D,-Df)Q
(Eq. A-8) NR* =
The average velocity of flow through the rotameter is given by Q_JAm where Q.m
is the volumetric flow rate through the meter and Am is the annular area between
the inside circumference of the tube at the float position.
Substituting Qm/Am for v in Equation A-8 yields:
Q_m(D,-Df)Q
(Eq. A-9) NRe=-
Am\t.
The density of the flowing fluid Q is equal to PmMjRTm where Pm is the
absolute pressure of the metered gas, Mm is the apparent molecular weight of the
metered gas, R is the universal gas constant, and Tm is the absolute temperature of
the metered gas.
Substituting PmMm/RTm for Q in Equation A-9 yields:
Qm(Dt-Df)PmMm
(E<*-A-10) "--—
Adding the subscript m to the viscosity term ^ in Equation A-10 to denote the
viscosity of the metered gas yields the following equation which is used to calculate
Reynold's Number for gas flow in a rotameter.
Q_m(L
(Eq. A-ll) N*. =
Al-5
-------
Common Practices in the Use of a Rotameter
for Gas Flow Measurement
It can be seen from Equation A-7 that the volumetric flow rate through a
rotameter can be calculated when such physical characteristics as the diameter and
the mass of the float and the annular area of the meter at each tube reading are
known providing measurements are made of the temperature, pressure and
molecular weight of the metered gas. Before these calculations of the volumetric
flow rate can be made, data must be known about the meter coefficient Cm. The
meter coefficient being a function of Reynold's Number is ultimately a function of
the conditions at which the meter is being used. To obtain data on the meter coef-
ficient, the meter must be calibrated. However, because of the ease involved in
using calibration curves, common practice is to use calibration curves to determine
volumetric flow rates instead of calculating the flow rates from raw data.
Procedures for the Calibration of a Rotameter
A common arrangement of equipment for calibrating a rotameter is shown in
Figure A-3.
Rotameter
Metering valve
777 /
Gas mover
Test meter I Manometer
Figure A-3. Test setup for calibrating a rotameter.
Flow through the calibration train is controlled by the metering valve. At various
settings of the rotameter float, measurements are made of the flow rate through
the train and of the pressure and temperature of the gas stream at the rotameter.
The temperature of the gas stream is usually assumed to be the same as the
temperature of the ambient air. If the test meter significantly affects the pressure
or temperature of the gas stream, measurements should also be made of the actual
pressure and temperature at the test meter. A typical rotameter calibration curve is
illustrated in Figure A-4.
Al-6
-------
Tube reading
At some temperature and pressure
Volumetric flow rate
Figure A-4. Rotameter calibration curve.
To make the calibration curve useful, the temperature and pressure of the
olumetric flow rate must be specified.
A Universal Calibration Curve
The normal arrangement of the components in a sampling train is shown in
Figure A-5. Since the meter is usually installed downstream from the pollutant col-
lector, it can be expected to operate under widely varying conditions of pressure,
temperature, and molecular weight. This requires a different calibration curve for
each condition of pressure, temperature, and molecular weight. This can be
facilitated by drawing a family of calibration curves, which would bracket the
anticipated range of pressures, temperatures and molecular weights, as shown in
Figure A-6.
Pollutant
collector
Meter
Metering
valve
Gas
mover
Figure A-5. Arrangement of sampling components.
Al-7
-------
Tube reading
Volumetric flow rate
Figure A-6. Family of rotameter calibration curves.
Operation of a rotameter under extreme sampling conditions, particularly
extreme temperatures, complicates the calibration setup. It is difficult if not
impossible for most laboratories to be able to calibrate flow metering devices at
high temperatures or unusual gas mixtures (especially where toxic gases are
involved). For these reasons, it is desirable to develop a calibration curve which is
independent of the actual expected sampling conditions. As previously mentioned,
the flow through a rotameter is dependent upon the value of Cm, the meter coeffi-
cient (see Equation A-7), which is a function of the Reynold's Number for the flow
in the rotameter. Therefore, to be independent of the sampling conditions, the
calibration curve must be in terms of Cm and NRe.
Development of a Universal Calibration Curve
Solving Equation A-7 for Cm gives the following relationship:
,/ PmMm XV*
\gnifRT m/
(Eq. A-12)
Dividing Equation A-ll by Equation A-12 yields:
Q_m(Dt-Df)PmMn
Am \gmfRT
Al-8
-------
Cancelling the like terms Qm and Am yields:
AT,
Re
D,( PJVf,
\gmfR7
Simplifying:
N
R.
PmM
Combining the like terms Pm, Mm, R and Tm yields:
(Eq. A-13)
NRe
(D,-Df)
Simplifying the (Dt-Df) and D/ relationship in Equation A-13 yields a dimen-
sionless factor which has no limitations on either Reynold's Number or the meter
coefficient Cm.
N
Re
(Eq. A-14)
A plot of the dimensionless factor NRe/Cm defined by Equation A-14 versus the
meter coefficient Cm as calculated from Equation A-12 on regular graph paper will
yield a universal calibration curve which is independent of the sampling conditions.
Such a plot is illustrated in Figure A-7.
Figure A-7. A universal calibration curve for a rotameter.
Al-9
-------
Use of the Universal Calibration Curve for a Rotameter
To Determine an Existing Flow Rate
To determine an existing flow rate, measurements must be made of the gas
temperature and pressure as well as the float position. Data from the manufacturer
of the rotameter will yield information on the diameter of the tube at the various
float positions and on the diameter and mass of the float. The apparent molecular
weight of the gas being metered can be calculated if the composition of the gas
stream is known. The viscosity of the gas stream can be determined if the
temperature of the gas stream is known (see Perry's Chemical Engineer's Hand-
book). From this data the NRe/Cm factor (see Equation A-14) can be calculated.
The universal calibration curve is then entered at the calculated value of NRe/Cm
and the corresponding Cm is noted. Q,m is then calculated from Equation A-7.
To Establish a Required Sampling Rate
To establish a required sampling rate, estimates are made of the metered gas
pressure (Pm), the metered gas temperature (Tm), the apparent molecular weight of
the metered gas (Mm), and the area of the meter (Am) which will exist at the
desired sampling rate. Using these estimated values, the meter coefficient Cm is
calculated (see Equation A-12) for the desired sampling rate Qm. The universal
calibration curve (see Figure A-7) is entered at this value of Cm and the corres-
ponding NRe/Cm factor is noted. [(£>,/£>,)- 1] is solved by using the following equa-
tion which is a rearrangement of Equation A-14:
(Eq. A-15)
The float position can be determined from the value [(Dt/Df)- I]. For some
rotameters the value of [(Dt/Df)- 1] is the tube reading divided by 100. If the area
of the meter corresponding to this float position is not equal to the original
estimated value for the meter arear the new value of area is used as an estimate
and the entire procedure is repeated until the estimated area and the calculated
area are equal. Then upon setting the float position at this tube reading, Tm, Pm,
and Mm are noted. If they are different from the original estimates, the procedure
is repeated using the observed values of Tm, Pm, and Mm as estimates. Experience
will aid in selecting original estimates that are nearly accurate so that the required
sampling rate may be set fairly rapidly.
To Predict Calibration Curves
The above techniques are very cumbersome to apply in the field and as a result the
universal calibration curve should not be used in such a manner.
The real utility of the universal calibration curve is that it can be used to predict
calibration curves at any set of conditions. This results in a great reduction in
laboratory work in that the rotameter need only be calibrated once and not every
time the conditions at which the meter is operated change.
Al-10
-------
cm
li st step in predicting calibration curves from the universal calibration
LU1V, A ., rotameter is to ascertain the anticipated meter operating range for the
samph.ig application of concern. Once this operating range is established, an
arbi.rary selection of a point on the universal calibration curve is made (see point
in Figure A-8) The coordinates of point (a), NRe/Cm (point b) and Cm (point c) are
determined. Values of T,, A, and M, and the value of NRe/Cm are used to
calculate a value for D,/Df- I by means of the following equation:
(Eq. A-15)
NK.\/
Cm I \gmfPmMn
Figure A-8. Predicting calibration curves from the universal calibration curve.
The area of the meter (Am) is calculated from this value of [(/>,/£>,)-!] and is
used along with the assumed values of T,, P,, and M, and the value of Cm from
the universal calibration curve to calculate a volumetric flow rate by means of the
following equation:
(Eq. A-7)
D,
PmMm
This procedure is repeated until enough points are available to plot a normal
calibration curve. The entire procedure is repeated using new values for
temperature, pressure, and molecular weight until a family of calibration curves is
plotted Of course this family of curves should bracket the anticipated meter
operating conditions for the sampling application of concern. The volumetric flow
rate (£m) is plotted versus either the area of the meter (Am) or the tube reading
which corresponds to the meter area.
AMI
-------
Field operation is greatly simplified if the tube reading is used. A typical family
of calibration curves is shown in Figure A-9.
m
or
Tube
reading
Figure A-9. Calibration curves predicted from universal calibration curve.
Notice that these curves are similar to the calibration curves illustrated in Figure
A-6. The difference between them is the manner in which they were obtained. The
curves of Figure A-6 were obtained by an actual laboratory calibration run for
each set of conditions illustrated, whereas the curves of Figure A-9 were obtained by
mathematical manipulation of data from only one calibration run. This can, of
course, save considerable laboratory time. In addition, it may not be possible to
ascertain, in the laboratory, calibration data at extreme conditions, particularly at
high temperatures.
Al-12
-------
Appendix 2
Federal Register Reference Methods
A2-a
-------
APPENDIX A-REFERENCE METHOD FOR THE
DETERMINATION OF SULFUR DIOXIDE IN THE
ATMOSPHERE (PARAROSANILINE METHOD)
1 Principle and Applicability. 1.1 Sulfur
dioxide is absorbed from air in a solution of
potassium tetrachloromercurate (TCM). A
dichlorosulfitomercurate complex, which re-
sists oxidation by the oxygen in the air, is
formed (1, 2). Once formed, this complex is
stable to strong oxidants (e.g., ozone, oxides
of nitrogen). The complex is reacted with
pararosaniline and formaldehyde to form
intensely colored pararosaniline methyl sul-
fonic acid (3). The absorbance of the solu-
tion is measured spectrophotometrically.
1.2 The method is applicable to the mea-
surement of sulfur dioxide in ambient air
using sampling periods up to 24 hours.
2 Range and SensitivUy. 2.1 Concentra-
tions of sulfur dioxide in the range of 25 to
1,050 MK/m.3 (0.01 to 0.40 p.p.m.) can be
measured under the conditions given. One
can measure concentrations below 28 n?./
m.3 by sampling larger volumes of air out
only if the absorption efficiency of the par-
ticular system is first determined. Higher
concentrations can be analyzed by using
smaller gas samples, a larger collection
volume, or a suitable aliquot of the collected
sample. Beer's Law is followed through the
working range from 0.03 to 1.0 absorbance
units (0.8 to 27 Mg- of sulfite ion in 25 ml.
final solution computed as SO,).
2 2 The lower limit of detection of sulfur
dioxide in 10 ml. TCM is 0.75 ^g. (based on
twice the standard deviation) representing a
concentration of 25 ^g./m3SO, (0.01 p.p.m.)
in an air sample of 30 liters.
3. Interferences. 3.1 The effects of the
principal known interferences have been
minimized or eliminated. Interferences by
oxides of nitrogen are eliminated by sulfa-
mic acid (4, 5), ozone by time-delay (6), and
heavy metals by EDTA (ethylene-diamine-
tetraacetic acid, disodium salt) and phos-
phoric acid (4, 6). At least 60 >ig. Pe (III), 10
Mg. Mn (II), and 10 ng. Cr (III) in 10 ml. ab-
sorbing reagent can be tolerated in the pro-
cedure. No significant interference was
found with 10 nS- Cu (II) and 22 /ig. V (V).
4. Precision, Accuracy, and Stability. 4.1
Relative standard deviation at the 95 per-
cent confidence level is 4.6 percent for the
analytical procedure using standard sam-
ples. (5)
4.2 After sample collection the solutions
are relatively stable. At 22° C. losses of
sulfur dioxide occur at the rate of 1 percent
per day. When samples are stored at 5' C.
for 30'days, no detectable losses of sulfur
dioxide occur. The presence of EDTA en-
hances the stability of SO, in solution, and
the rate of decay is independent of the con-
centration of SO,. (7)
5. Apparatus.
5.1 Sampling.
5.1.1 Absorber Absorbers normally used
in air pollution sampling are acceptable for
concentrations above 25 fig./m.3 (0.01
p.p.m.). An all-glass midget impinger, as
shown in Figure Al, is recommended for 30-
minute and 1-hour samples.
For 24-hour sampling, assemble an absorb-
er from the following parts:
Polypropylene 2-port tube closures, spe-
cial manufacture (available from Bel-Art
Products, Pequannock, N.J.).
Glass impingers, 6 mm. tubing, 6 inches
long, one end drawn to small diameter such
Appendix A
that No. 79 jewelers drill will pass through,
but No. 78 jewelers drill will not. (Other end
fire polished.)
Polypropylene tubes, 164 by 32 mm. (Nal-
gene or equal).
5.1.2 Pump. Capable of maintaining an
air pressure differential greater than 0.7 at-
mosphere at the desired flow rate.
5.1.3 Air Flowmeter or Critical On/ice. A
calibrated rotameter or critical orifice capa-
ble of measuring air flow within ±2 percent.
For 30-minute sampling, a 22-gauge hypo-
dermic needle 1 inch long may be used as a
critical orifice to give a flow of about 1 liter/
minute. For 1-hour sampling, a 23-gauge hy-
podermic needle five-eights of an inch long
may be used as a critical orifice to give a
flow of about 0.5 liter/minute. For 24-hour
sampling, a 27-gauge hypodermic needle
three-eighths of an inch long may be used
to give a flow of about 0.2 liter/minute. Use
a membrane filter to protect the needle
(Figure Ala).
5.2 Analysis.
5.2.1 Spectrophotometer. Suitable for mea-
surement of absorbance at 548 nm. with an
effective spectral band width of less than 15
nm. Reagent blank problems may occur
with spectrophotometers having greater
spectral band width. The wavelength cali-
bration of the instrument should be veri-
fied. If transmittance is measured, this can
be converted to absorbance:
6. Reagents.
6.1 Sampling.
6.1.1 Distilled water. Must be free from
oxidants.
6.1.2 Absorbing Reagent 10.04 M Potas-
sium Tetrachloromercurate (TCAf)]. Dis-
solve 10.86 g. mercuric chloride, 0.066 g.
EDTA (ethylenediaminetetraacetic acid, di-
sodium salt), and 6.0 g. potassium chloride
in water and bring to mark in a 1,000-ml.
volumetric flask. (Caution: Highly poison-
ous. If spilled on skin, flush off with water
immediately). The pH of this reagent
should be approximately 4.0, but it has been
shown that there is no appreciable differ-
ence in collection efficiency over the range
of pH 5 to pH 3. (7). The absorbing reagent
is normally stable for 6 months. If a precipi-
tate forms, discard the reagent.
6.2 Analysis.
6.2.1 Sulfamic Acid (0.6 percent). Dis-
solve 0.6 g. sulfamic acid in 100 ml. distilled
water. Prepare fresh daily.
6.2.2 Formaldehyde (0.2 percent). Dilute 5
ml. formaldehyde solution (36-38 percent)
to 1,000 ml. with distilled water. Prepare
daily.
6.2.3 Stock Iodine Solution (0.1 TV). Place
12.7 g. iodine in a 250-ml. beaker; add 40 g.
potassium iodide and 25 ml. water. Stir until
all is dissolved, ther dilute to 1,000 ml. with
distilled water.
6.2.4 Iodine Solution (0.01 N). Prepare
approximately 0.01 N iodine solution by di-
luting 50 ml. of stock solution to 500 ml.
with distilled water.
6.2.5 Starch Indicator Solution. Tri-
turate 0.4 g. soluble starch and 0.002 g. mer-
curic iodide (preservative) with a little
water, and add the paste slowly to 200 ml.
boiling water. Continue boiling until the so-
lution is clear; cool, and transfer to a gk.ss-
stoppered bottle.
6.2.6 Stock Sodium Thiosulfate Solution
(0.1 N). Prepare a stock solution by dissolv-
ing 25 g. sodium thiosulfate (Na&O, 5H,O)
in 1,000 ml freshly boiled, cooled, distilled
water and add 0.1 g. sodium carbonate to
the solution. Allow the solution to stand 1
day before standardizing. To standardize,
accurately weigh, to the nearest 0.1 mg., 1.5
g. primary standard potassium iodate dried
at 180° C. and dilute to volume in a 500-ml.
volumetric flask. To a 500-ml. iodine flask,
pipet 50 ml. of iodate solution. Add 2 g. po-
tassium iodide and 10 ml. of 1 N hydrochlo-
ric acid. Stopper the flask. After 5 minutes,
titrate with stock thiosulfate solution to a
pale yellow. Add 5 ml. starch indicator solu-
tion and continue the titration until the
blue color disappears. Calculate the normal-
ity of the stock solution:
N = (W/M)x2.80
N = Normality of stock thiosulfate solution.
M=Volume of thiosulfate required, ml.
W=Weight of potassium iodate, grams.
2.80=C103(conversion of g. tomg.JxO.l
(fraction iodate used)]/35.67 (equivalent
weight of potassium iodate)
6.2.7 Sodium Thiosulfate Titrant (0.01
N). Dilute 100 ml. of the stock thiosulfate
solution to 1,000 ml. with freshly boiled dis-
tilled water.
Normality = Normality of stock
solution • 0.100
6.2.8 Standardized Sulfite Solution for
Preparation of Working Sulfite-TCM Solu-
tion. Dissolve 0.3 g. sodium metabisulfite
(Na,S,O5) or 0.40 g. sodium sulfite (NaJSO3)
in 500 ml of recently boiled, cooled, distilled
water. (Sulfite solution is unstable, it is
therefore important to use water of the
highest purity to minimize this instability )
This solution contains the equivalent of 320
to 400 jig./ml. of SO, The actual concentra-
tion of the solution is determined by adding
excess iodine and back-titrating with stand-
ard sodium thiosulfate solution. To back-ti-
trate, pipet 50 ml of the 0.01 N iodine into
each of two 500-ml. iodine flasks (A and B).
To flask A (blank) add 25 ml. distilled water,
and to flask B (sample) pipet 25 ml. sulfite
solution. Stopper the flasks and allow to
react for 5 minutes. Prepare the working
sulfite-TCM Solution (6.2.9) at the same
time iodine solution is added to the flasks.
By means of a buret containing standard-
ized 0 01 N thiosulfate, titrate each flask in
turn to a pale yellow. Then add 5 ml. starch
solution and continue the titration until the
blue color disappears.
6 2.9 Working Sulfite-TCM Solution.
Pipet accurately 2 ml. of the standard solu-
tion into a 100 ml volumetric flask and bring
to mark with 0.04 M TCM. Calculate the
concentration of sulfur dioxide in the work-
ing solution.
HgSO,/ml.-«A-B> (N) (32.000V25) ( > 0.02)
A = Volume thiosulfate for blank, ml.
B = Volume thiosulfate for sample, ml.
N = Normality of thiosulfate titrant.
32.000 = Milliequivalent wt. of SO,, ,ug.
25 = Volume standard sulfite solution, ml.
0.02 = Dilution factor.
A2-1
-------
This solut in is stable for 30 days if kept at
5 C. (refrigerator). If not kept at 5 C.. pre-
pare daily.
6.2.10 Punfied Pararosamhne Stock So-
lution (0.2 percent nominal).
6.2.10.1 Dye Specifications. The pararo-
saniline dye must meet the following per-
formance specifications: (1) The dye must
have a wavelength of maximum absorbance
at 540 nrn. when assayed in a buffered solu-
tion of 0 1 M sodium acetate-acetic acid; (2)
the absorbance of the reagent blank, which
is temperature-sensitive (0.015 absorbance
unit/'C), should not exceed 0.170 absor-
bance unit at 22° C with a 1-Tn. optical
path length, when the blank is prepared ac-
cording l,o the prescribed analytical proce-
dure and to the specified concentration of
the dye; (3) the calibration curve (Section
8.2.1) should have a slope of 0.030±0.002 ab-
sorbance units/^g. SO, at this path length
when the dye is pure and the sulfite solu-
tion is properly standardized.
6.2.10.2 Preparation of Stock Solution. A
specially purified (99-100 percent pure) so-
lution of pararosaniline, which meets the
above specifications, is commercially availa-
ble in the required 0 20 percent concentra-
tion (Harleco*). Alternatively, the dye may
be purified, a stock solution prepared and
then assayed according to the procedure of
Scaringe'ili, et al. (4)
6.2.11 Pararosamhne Reagent To a 250-
ml. volumetric ilask, add 20 ml. stock parar-
osaniline solution. Add an additional 0.2 ml.
stock solution for each percent the stock
assays below 100 percent. Then add 25 ml. 3
M phosphoric acid and dilute to volume
with distilled water. This reagent is stable
for at least 9 months.
7. Procedure.
7.1 Sampling. Procedures are described
for short-term (30 minutes and 1 hour) and
for long-term (24 hours) sampling. One can
select different combinations of sampling
rate and time to meet special needs. Sample
volumes should be adjusted, so that linear-
ity is maintained between absorbance and
concentration over the dynamic range.
7.1.1 JO-Minute and 1-Hour Samplings.
Insert a midget impinger into the sampling
system. Figure Al. Add 10 ml. TCM solution
to the irnpinger Collect sample at 1 liter/
minute for 30 minutes, or at 0.5 liter/minute
for 1 hour, using either a rotameter, as
shown in Figure Al. or a critical onficr, as
shown in Figure Ala. to control flow. Shield
the absorbing reagent from direct sunlight
during and after sampling by covering the
impinge: with aluminum foil, to prevent de-
terioration. Determine the volume of air
sampled by multiplying the flow rate by the
time in minutes and record the atmospheric
pressure and temperature. Remove and
stopper the impinger. If the sample must be
stored for more than a day before analysis,
keep it at 5° C. in a refrigerator (see 4.2).
7.1.2 24-Hour Sampling. Place 50 ml.
TCM solution in a large absorber and collect
the sample at 0.2 liter/minute for 24 hours
from midnight to midnight. Make sure no
entrainment of solution results with the im-
pinger. During collection and storage pro-
tect from direct sunlight. Determine the
total air volume by multiplying the air flow
rate by the time in minutes. The correction
of 24-hour measurements for temperature
and pressure is extremely difficult and is
not ordinarily done. However, the accuracy
of the measurement will be improved if
meaningful corrections can be applied. If
storage is necessary, refrigerate at 5' C. (see
4.2).
7.2 Analysis.
7.2.1 Sample Preparation. After collec-
tion, if a precipitate is observed in the
sample, remove it by centrilugation.
7.2.1.1 30-Mmute and 1-Hour Samples,
Transfer the sample quantitatively to a 2b-
ml. volumetric flask, use about 5 m! dis-
tilled water for rinsing. Delay analyses for
20 minutes to allow any ozone to decom-
pose.
7.2 1 2 24-Hour Sample. Dilute the entire
sample to 50 ml with absorbing solution
Pipet 5 ml of the sample into a 25-ml volu-
metric flask for chemical analyses Bring
volume to 10 ml with absorbing reagent
Delay analyses for 20 minutes to allow any
ozone to decompose
7.2.2 Determination. For each set oi de-
terminations prepare a reagent blank by
adding 10 ml. unexposed TCM solution to a
25-ml. volumetric flask Prepare a control
solution by adding 2 ml. of working sulfite-
TCM solution and 8 ml. TCM solution to a
25-ml. volumetric flask. To each flask con-
taining either sample, control solution, or
reagent blank, add 1 ml. 0.6 percent sulfa-
mic acid and allow to react 10 minutes to de-
stroy the nitrite from oxides of nitrogen.
Accurately pipet in 2 ml. 0.2 percent formal-
dehyde solution, then 5 ml pararosaniline
solution Start a laboratory timer that has
been set for 30 minutes. Bring all flasks to
volume with freshly boiled and cooled dis-
tilled water and mix thoroughly. After 30
minutes and before 60 minutes, determine
the absorbances of the sample (denote as
A), reagent blank (denote as A0) and the
control solution at 548 nm. using 1-cm. opti-
cal path length cells Use distilled water, not
the reagent blank, as the reference (NOTE'
This is important because of the color sensi-
tivity of the reagent blank to temperature
changes which can be induced in the cell
compartment of a spectrophotometer.i Do
not allow the colored solution to stand m
the absorbance cells, because a film of dye
may be deposited. Clean cells with alcohol
after use. If the temperature of the determi-
nations does not differ by more than 2" C.
from the calibration temperature (8.2), the
reagent blank should be within 0.03 absor-
bance unit of the y-intercept of the calibra-
tion curve (8.2). If the reagent blank differs
by more than 0 03 absorbance unit from
that found in the calibration curve, prepare
a new curve.
7.2.3 Absorbance Range If the absor-
bance of the sample solution ranges be-
tween 1.0 and 2.0, the sample can be diluted
1.1 with a portion of the reagent blank and
read within a few minutes. Solutions with
higher absorbance can be diluted up to six-
fold with the reagent blank in order to
obtain onscale readings within 10 percent of
the true absorbance value.
8. Cadbratton and Efficiencies.
8.1 Flowmeters and Hypodermic Needle.
Calibrate flowmeters and hypodermic
needle (8> against a calibrated wet test
meter
8.2 Calibration Curves
•Hartmen-Leddon, 60th and Woodland
Avenue, Philadelphia, PA 19143.
8.2.1 l-'Gcedure with Sulfite Solution Ac-
curately pipet graduated amounts of the
working sulfite-TCM solution (62.9) (such
as 0. 0 5, 1, 2, 3, and 4 ml.) into a series of 25-
ml. volumetric flasks. Add sufficient TCM
solution to each flask to bring the volume to
approximately 10 ml. Then add the remain-
ing reagents as described in " 2 2 For maxi-
mum precision use a const; nt-temperature
batn The temperature of calibration must
be maintained within ~ 1 C. ..nd in the
range of 20' to 30' C. The temperature of
calibration and the temperature of analysis
must be within 2 degrees. Plot the absor-
bs n re against the total concentration in fig.
SO; for the corresponding solution. The
total fig. SO, in solution equals the concen-
tration of the standard (Section 6.2.9) in us
SO./ml times the ml. sulfite solution added
(fig SO,-fig/I SO,xml added). A linear re-
lationship should bt; obtained, and the y-m-
tercept should be within 0.03 absorbance
unit of the zero standard absorbance. For
maximum precision determine the line of
best fit using regression analysis by the
method of least squares Determine the
slope of the line of best fit, calculate its re-
ciprocal and denote as B,. B, is the calibra-
tion factor. (See Section 6.2.10.1 for specifi-
cations on the slope of the calibration
curve). This calibration factor can be used
for calculating results provided there are no
radical changes in temperature or pH At
least one control sample containing a known
concentration of SO, for each series of de-
terminations, is recommended to insure the
reliability of this factor.
8.2 2 Procedure with SO, Permeation
Tubes.
8 2.2.1 General Considerations. Atmos-
pheres containing accurately known
amounts of sulfur dioxide at levels of inter-
est can be prepared using permeation tubes.
In the systems for generating these atmos-
pheres, the permeation tube emits SO, gas
at a known, low, constant rate, provided the
temperature of the tube is held constant
dtO.r C.) and provided the tube has been
accjrately calibrated at the temperature of
use. The SO, gas permeating from the tube
is carried by a low flow of inert gas to a
mixing chamber where it is accurately dilut-
ed with SO,-free air to the level of interest
and the sample taken. These systems are
shown schematically in Figures A2 and A3
and have been described in detail by
O'Keeffe and Ortman (9), Scaringelli, Frey.
and Saltzman (10). and Scaringelli,
O'Keeffe, Rosenberg, and Bell (11).
8.2.2.2 Preparation of Standard Atmos-
pheres. Permeation tubes may be prepared
or purchased. Scaringelli, O'Keeffe, Rosen-
berg, and Bell (11) give detailed, explicit di-
rections for permeation tube calibration.
Tubes with a certified permeation rate are
available from the National Bureau of
Standards. Tube permeation rates from 0.2
to 0.4 fig./minute, inert gas flows of about
50 ml./minute, and dilution air flow rates
from 1.1 to 15 liters/minute conveniently
give standard atmospheres containing de-
sired levels of SO, (25 to 390 fig./m.'; 0.01 to
0.15 p p.m. SO,). The concentration of SO,
in any standard atmosphere can be calculat-
ed as follows:
C = (Pxl01)/(Ra + R1)
A2-2
-------
Where:
C = Concentration of SO,, ns./m. * at refer-
ence conditions.
P=Tube permeation rate, jig./minute.
Rd = Flov rate of dilution air, liter/minute
at reference conditions.
P., = Flow rate of inert gas, liter/minute at
reference conditions.
8.2.2.3 Sampling and Preparation of
Calibration Curve. Prepare a series (usually
six) of standard atmospheres containing SO,
levels from 25 to 390 pg. SO,/m.5. Sample
each atmosphere using similar apparatus
and taking exactly the same air volume as
will be done in atmospheric sampling. De-
termine absorbances as directed in 7.2. Plot
the concentration of SO, in jig./m.3 (x-axis)
against A'A. values (y-axis), draw the
straight line of best fit and determine the
slope. Alternatively, regression analysis by
the method of least squares may be used to
calculate the slope. Calculate the reciprocal
of the slope and denote as B,.
8.3 Sampling Efficiency. Collection effi-
ciency is above 98 percent; efficiency may
fall off, however, at concentrations below 25
Hg./m.3. (12, 13)
9. Calculations.
9.1 Conversion of Volume. Convert the
volume of air sampled to the volume at ref-
erence conditions of 25° C. and 760 mm. Hg.
(On 24-hour samples, this may not be possi-
ble.)
VR=Vx(P/760)x (298/1 + 273)
VR = Volume of air at 25' C. and 760 mm. Hg,
liters.
V = Volume of air sampled, liters.
P=Barometric pressure, mm. Hg.
t=Temperature of air sample, "C.
9.2 Sulfur Dioxide Concentration.
9.2.1 When sulfite solutions are used to
prepare calibration curves, compute the
concentration of sulfur dioxide in the
sample:
lig. SO./m.^CUA-A.XlO'HB,))/ VR]xD
A=Sample absorbance.
Ao=Reagent blank absorbance.
10 3=Conversion of liters to cubic meters.
VH = The sample corrected to 25° C. and 760
mm. Hg, liters.
B, = Calibration factor, ^g./absorbance unit.
D = Dilution factor. For 30-minute and 1-
hour samples, D=l. For 24-hour sam-
ples. D=10.
9.2.2 When SO, gas standard atmos-
pheres are used to prepare calibration
curves, compute the sulfur dioxide in the
sample by the following formula:
SO,, ng./m.3 = (A-Ao)xB,
A ^Sample absorbance.
Ao=Reagent blank absorbance.
B. = (See 8.2.2.3).
9.2.3 Conversion of p.g./m.' to p.p.m. = lt
desired, the concentration of sulfur dioxide
may be calculated as p.p.m. SO, at reference
conditions as follows:
p.p.m. SO,=jig. SO./m.^x3.82x10"'
10. References.
(1) West, P. W., and Gaeke, G. C., "Fixation
of Sulfur Dioxide as Sulfitomercurate
III and Subsequent Colorimetric Deter-
mination", Anal. Chem. 28, 1816 (1956).
(2) Ephraims, P., "Inorganic Chemistry," p.
562, Edited by P.C.L. Thome and E. R.
Roberts, 5th Edition, Interscience.
(1948).
(3> Lyles. G. R., Dowling, F. B., and Blan-
chard, V. J., "Quantitative Determina-
tion of Formaldehyde in Parts Per Hun-
dred Million Concentration Level", J.
Air Poll. Cant. Assoc. 15, 106 (1965).
(4) Scaringelli, F. P., Saltzman, B. E., and
Frey, S. A., "Spectrophotometric Deter-
mination of Atmospheric Sulfur Diox-
ide", Anal Chem. 39, 1709 (1967).
(5) Pate, J. B., Ammons, B. E., Swanson, G.
A., Lodge, J. P., Jr., "Nitrite Interfer-
ence in Spectrophotometric Determina-
tion of Atmospheric Sulfur Dioxide",
Anal. Chem. 37, 942 (1965).
(6) Zurio, N. and Griff ini, A. M., "Measure-
ment of the SO, Content of Air in the
Presence of Oxides of Nitrogen and
Heavy Metals", Med. Lavoro, 53, 330
(1962).
(7) Scaringelli, F. P., Elfers, L., Norris, D.,
and Hochheiser, S., "Enhanced Stability
of Sulfur Dioxide in Solution", Anal
Chem. 42, 1818 (1970).
(8) Lodge, J. P. Jr., Pate, J. B., Ammons, B.
E. and Swanson, G. A., "Use of Hypoder-
mic Needles as Critical Orifices in Air
Sampling," J. Air Poll. Cont, Assoc. 16,
197(1966).
(9) n'Keeffe, A. E., and Ortman, G. C., "Pri-
mary Standards for Trace Gas Analy-
sis", Anal. Chem. 38, 760 (1966).
(10) Scaringelli, F. P., Frey, S. A., and Saltz-
man, B. E., "Evaluation of Teflon Per-
meation Tubes for Use with Sulfur
Dioxide", Amer. Ind. Hygiene Assoc. J.
28, 260(1967).
Ill) Scaringelli, F. P.. O'Keeffe, A. E.,
Rosenberg, E., and Bell, J. P., "Prepara-
tion of Known Concentrations of Gases
and Vapors with Permeation Devices
Calibrated Gravimetrically", Anal.
Chem. 42, 871 '1970).
(12) Urone, P., Evans, J. B., and Noyes, C.
M., "Tracer Techniques in Sulfur Diox-
ide Colorimetric and Conductiometric
Methods", Anal Chem. 37, 1104 (1965).
(13) Bostrom, C. E., "The Absorption of
Sulfur Dioxide at Low Concentrations
(p.p.m.) Studied by an Isotopic Tracer
Method", Intern. J. Air Water Poll. 9, 33
(1965).
HYPODERMIC
NEEDLE
Fi0ur«A1. Sampling imn
A2-3
------- An error occurred while trying to OCR this image.
-------
Appendix B
APPENDIX B—REFERENCE METHOD FOR THE
DETERMINATION OF SUSPENDED PARTICU-
LATES IN THE ATMOSPHERE (HIGH VOLUME
METHOD)
1. Principle and Applicability.
1.1 Air is drawn into a covered housing
and through a filter by means of a high-
flow-rate blower at a flow rate (1.13 to 1.70
m.Vmin.; 40 to 60 ft. Vmin.) that allows sus-
pended particles having diameters of less
than 100 ^m. (Stokes equivalent diameter)
to pass to the filter surface. (1) Particles
within the size range of 100 to O.lnm. diame-
ter are ordinarily collected on glass fiber fil-
ters. The mass concentration of suspended
particulates in the ambient air (^g./m.3) is
computed by measuring the mass of collect-
ed particulates and the volume of air sam-
pled.
1.2 This method is applicable to measure-
ment of the mass concentration of suspend-
ed particulates in ambient air. The size of
the sample collected is usually adequate for
other analyses.
2. Range and Sensitivity.
2.1 When the sampler is operated at an
average flow rate of 1.70 m.Vmin. (60 ft.3/
min.) for 24 hours, an adequate sample will
be obtained even in an atmosphere having
concentrations of suspended particulates as
low as 1 M8-/m.3. If particulate levels are un-
usually high, a satisfactory sample may be
obtained in 6 to 8 hours or less. For determi-
nation of average concentrations of sus-
pended particulates in ambient air, a stand-
ard sampling period of 24 hours is recom-
mended.
2.2 Weights are determined to the near-
est milligram, airflow rates are determined
to the nearest 0.03 m.Vmin. (1.0 ft.Vmin.J.
times are determined to the nearest 2 min-
utes, and mass concentrations are reported
to the nearest microgram per cubic meter.
3. Interferences.
3.1 Particulate matter that is oily, such
as photochemical smog or wood smoke, may
block the filter and cause a rapid drop in
airflow at a nonuniform rate. Dense fog or
high humidity can cause the filter to
become too wet and severely reduce the air-
flow through the filter.
3.2 Glass-fiber filters are comparatively
insensitive to changes in relative humidity,
but collected particulates can be hygroscop-
ic. (2)
4. Precision, Accuracy, and Stability.
4.1 Based upon collaborative testing, the
relative standard deviation (coefficient of
variation) for single analyst variation (re-
peatability of the method) is 3.0 percent.
The corresponding value for multilabora-
tory variation (reproducibihty of the
method) is 3.7 percent. (3)
4.2 The accuracy with which the sampler
measures the true average concentration de-
pends upon the constancy of the airflow
rate through the sampler. The airflow rate
is affected by the concentration and the
nature of the dust in the atmosphere. Under
these conditions the error in the measured
average concentration may be in excess of
±50 percent of the true average concentra-
tion, depending on the amount of reduction
of airflow rate and on the variation of the
mass concentration of dust with time during
the 24-hour sampling period. (4)
5. Apparatus.
5.1 Sampling.
5.1.1 Sampler. The sampler consists of
three units: (1) The faceplate and gasket,
(2) the filter adapter assembly, and (3) the
motor unit. Figure Bl shows an exploded
view of these parts, their relationship to
each other, and how they are assembled.
The sampler must be capable of passing en-
vironmental air through a 406.5 cm.2 (63
in.2) portion of a clean 20.3 by 25.4 cm. (8-
by 10-in.) glass-fiber filter at a rate of at
least 1.70 m. Vmin. (60 ft. Vmin.). The motor
must be capable of continuous operation for
24-hour periods with input voltages ranging
from 110 to 120 volts, 50-60 cycles alternat-
ing current and must have third-wire safety
ground. The housing for the motor unit
may be of any convenient construction so
long as the unit remains airtight and leak-
free. The life of the sampler motor can be
extended by lowering the voltage by about
10 percent with a small "buck or boost"
transformer between the sampler and power
outlet.
5.1.2 Sampler Shelter. It is important
that the sampler be properly installed in a
suitable shelter. The shelter is subjected to
extremes of temperature, humidity, and all
types of air pollutants. For these reasons
the materials of the shelter must be chosen
carefully. Properly painted exterior ply-
wood or heavy gauge aluminum serve • ell.
The sampler must be mounted vertically in
the shelter so that the glass-fiber filter is
parallel with the ground. The shelter must
be provided with a roof so that the filter is
protected from precipitation and debris.
The internal arrangement and configura-
tion of a suitable shelter with a gable roof
are shown in Figure B2. The clearance area
between the main housing and the roof at
its closest point should be 580.5±193.5 cm.2
(90 ±30 in. 2).The main housing should be
rectangular, with dimensions of about 29 by
36cm. (ll'/z by 14 in.).
5.1.3 Rotameter. Marked in arbitrary
units, frequently 0 to 70, and capable of
being calibrated. Other devices of at least
comparable accuracy may be used.
5.1.4 Onfice Calibration Unit. Consisting
of a metal tube 7.6 cm. (3 in.) ID and 15.9
cm. <6V, in.) long with a static pressure tap
5.1 cm. (2 in.) from one end. See Figure B3.
The tube end nearest the pressure tap is
flanged to about 10.8 cm. (4'/4 in.) OD with a
male thread of the same size as the inlet
end of the high-volume air sampler. A single
metal plate 9.2 cm. (3% in.) in diameter and
0.24 cm. (%a in.) thick with a central orifice
2.9 cm. (I'/s in.) in diameter is held in place
at the air inlet end with a female threaded
ring. The other end of the tube is flanged to
hold a loose female threaded coupling,
which screws onto the inlet of the sampler.
An 18-hole metal plate, an integral part of
the unit, is positioned between the orifice
and sampler to simulate the resistance of a
clean glass-fiber filter. An orifice calibration
unit is shown in Figure B3.
5.1.5 Differential Manometer. Capable of
measuring to at least 40 cm. (16 in.) of
water.
5.1.6 Positive Displacement Meter. Cali-
brated in cubic meters or cubic feet, to be
used as a primary standard.
5.1.7 Barometer. Capable of measuring
atmospheric pressure to the nearest mm.
5.2 Analysis.
5.2.1 Filter Conditioning Environment.
Balance room or desiccator maintained at
15° to 35° C. and less than 50 percent rela-
tive humidity.
5.2.2 Analytical Balance. Equipped with
a weighing chamber designed to handle un-
folded 20.3 by 25.4 cm. (8- by 10-in.) filters
and having a sensitivity of 0.1 mg.
5.2.3 Light Source. Frequently a table of
the type used to view X-ray films.
5.2.4 Numbering Device. Capable of
printing identification numbers on the fil-
ters.
6 Reagents.
6.1 Filter Media. Glass-fiber filters
having a collection efficiency of at least 99
percent for particles of 0.3 ^m. diameter, as
measured by the DOP test, are suitable for
the quantitative measurement of concentra-
tions of suspended particulates. (5) al-
though some other medium, such as paper.
may be desirable for some analyses. If a
more detailed analysis is contemplated, care
must be exercised to use filters that contain
low background concentrations of the pol-
lutant being investigated. Careful quality
control is required to determine background
values of these pollutants.
7 Procedure.
7.1 Sampling.
7 1.1 Filter Preparation. Expose each
filter to the light source and inspect for pin-
holes, particles, or other imperfections. Fil-
ters with visible imperfections should not be
used. A small brush is useful for removing
particles. Equilibrate the filters in the filter
conditioning environment for 24 hours.
Weigh the filters to the nearest milligram;
record tare weight and filter identification
number. Do not bend or fold the filter
before collection of the sample.
7.1.2 Sample Collection. Open the shel-
ter, loosen the wing nuts, and remove the
faceplate from the filter holder. Install a
numbered, preweighed, glass-fiber filter in
position (rough side up), replace the face-
plate without disturbing the filter, and
fasten securely. Undertightening will allow
air leakage, overtightening will damage the
sponge-rubber faceplate gasket. A very light
application of talcum powder may be used
on the sponge-rubber faceplate gasket to
prevent the filter from sticking. During in-
clement weather the sampler may be re-
moved to a protected area for filter change.
Close the roof of the shelter, run the sam-
pler for about 5 minutes, connect the rota-
meter to the nipple on the back of the sam-
pler, and read the rotameter ball with rota-
meter in a vertical position. Estimate to the
nearest whole number. If the ball is fluctu-
ating rapidly, tip the rotameter and slowly
straighten it until the ball gives a constant
reading. Disconnect the rotameter from the
nipple; record the initial rotameter reading
and the starting time and date on the filter
folder. (The rotameter should never be con-
nected to the sampler except when the flow
is being measured.) Sample for 24 hours
from midnight to midnight and take a final
rotameter reading. Record the final rota-
meter reading and ending time and date on
the filter folder. Remove the faceplate as
described above and carefully remove the
filter from the holder, touching only the
outer edges. Fold the filter lengthwise so
that only surfaces with collected particu-
lates are in contact, and place in a manila
folder. Record on the folder the filter
number, location, and any other factors,
such as meteorological conditions or razing
of nearby buildings, that might a.'fect the
results. If the sample is defective, void it at
this time. In order to obtain a valid sample,
A2-5
-------
the high-voi 'trie sampler must be operated
with the same rotameter and tubing that
were used during its calibration.
7.2 Analysis. Equilibrate the exposed fil-
ters for 24 hours in the filter conditioning
environment, then reweigh. After they are
weighed, the filters may be saved for de-
tailed chemical analysis.
7.3 Maintenance.
1 3 1 Sampler Motor. Replace brushes
before they are worn to the point where
motor damage can occur.
7.3.2 Faceplate Gasket, Replace when the
margins of samples are no longer sharp. The
gasket may be sealed to the faceplate with
rubber cement or double-sided adhesive
tape.
7.3.3 Rotameter. Clean as required, using
alcohol.
8. Calibration.
8.1 Purpose. Since only a small portion of
the total air sampled passes through the ro-
tameter during measurement, the rotameter
must be calibrated against actual airflow
with the orifice calibration unit. Before the
orifice calibration unit can be used to cali-
brate the rotameter, the orifice calibration
unit itself must be calibrated against the
positive displacement primary standard.
8.1.1 Onfice Calibration Unit. Altai ;:
the orifice calibration unit to the intake end
of the positive displacement primary stand-
ard and attach a high-volume motor blower
unit to the exhaust end of the primary
standard. Connect one end of a differential
manometer to the differential pressure tap
of the orifice calibration unit and leave the
other end open to the atmosphere. Operate
the high-volume motor blower unit so that a
series of different, but constant, airflows
'usually six) are obtained for definite time
periods. Record the reading on the differen-
tial manometer at each airflow. The differ-
ent constant airflows are obtained by plac-
ing a series of loadplates, one at a time, be-
tween the calibration unit and the primary
standard. Placing the orifice before the inlet
reduces the pressure at the inlet of the pri-
mary standard below atmospheric; there-
fore, a correction must be made for the in-
crease in volume caused by this decreased
inlet pressure. Attach one end of a second
differential manometer to an inlet pressure
tap of the primary standard and leave the
other open to the atmosphere. During each
of the constant airflow measurements made
above, measure the true inlet pressure of
the primary standard with this second dif-
ferential manometer. Measure atmospheric
pressure and temperature. Correct the
measured air volume to true air volume as
directed in 9.1.1. then obtain true airflow
rate, Q, as directed in 9.1.3. Plot the differ-
ential manometer readings of the orifice
unit versus Q.
8.1.2 High-Volume Sampler. Assemble a
high-volume sampler with a clean filter in
place and tun for at least 5 minutes. Attach
a rotameter, read the ball, adjust so that
the ball reads 65, and seal the adjusting
mechanism so that it cannot be changed
easily. Shut off motor, remove the filter,
and attach the orifice calibration unit in its
place. Operate the high-volume sampler at a
series of different, but constant, airflows
(usually six). Record the reading of the dif-
ferential manometer on the orifice calibra-
tion unit, and record the readings of the ro-
tameter at each flow. Measure atmospheric
pressure and temperature. Convert the dif-
ferential manometer reading to m. Vmin., Q,
then plot rotameter reading versus Q.
8.1.3 Correction for Differences in Pres-
sure or Temperature. See Addendum B.
9. Calculations.
9.1 Calibration of Onfice.
9.1.1 True Air Volume. Calculate the air
volume measured by the positive displace-
ment primary standard.
V. = ((P.-Pm)/P.)(VM)
Va-True air volume ?t atmospheric pres-
sure, m.3
P.^Barometric pressure, mm. Hg.
Pm = Pressure drop at inlet of primary stand-
ard, mm. Hg.
VM = Volume measured by primary standard.
m.'
9.1.2 Conversion Factors.
Inches Hg x25.4 = mm. Hg.
Inches water x 73.48 < 10'3= inches Hg.
Cubic feet air x 0.0284 = cubic meters air.
9.1.3 True Airflow Rate.
Q = V./T
Q = Flow rate, m. Vmin
T = Time of flow, min.
9.2 Sample Volume.
9.2.1 Volume Conversion. Convert the
initial and final rotameter readings to true
airflow rate, Q, using calibration curve of
8.1.2.
9.2.2 Calculate volume of air sampled
V = (Q,Q,/2)xT
V^Air volume sampled, m.3
Q, = Initial airflow rate, m. Vmin.
Q,=Final airflow rate, m. Vmin.
T = Sampling time, mm.
9.3 Calculate mass concentration of sus-
pended particulates
S.P. = [(W,-W,)xl06]/V
S.P. = Mass concentration of suspended par-
ticulates, jig/m.3
W, = Initial weight of filter, g.
W,^Final weight of filter, g.
V = Air volume sampled, m.'
10 6=Conversion of g. to ^g.
10. References.
<;> Robson. C. D.. and Foster. K. E., "Evalu-
ation of Air Paniculate Sampling Equip-
ment". Am. Ind. Hyg Assoc. J. 24, 404
(1962).
(2) Tierney, G. P.. and Conner, W. D., "Hy-
groscopic Effects on Weight Determina-
tions of Particulates Collected on Glass-
Fiber Filters". Am. Ind. Hyg. Assoc. J.
28, 363 (1967).
(J) Unpublished data based on a collabora-
tive test involving 12 participants, con-
ducted under the direction of the Meth-
ods Standardization Services Section of
the National Air Pollution Control Ad-
ministration, October, 1970.
(4) Harrison, W. K., Nader, J. S., and
Fugman, F S., "Constant Flow Regula-
tors for High-Volume Air Sampler", Am.
Ind. Hyg. Assoc. J. 21, 114-120 (1960).
(5) Pate, J. B., and Tabor, E. C., "Analytical
Aspects of the Use of Glass-Fiber Filters
for the Collection and Analysis of At-
mospheric Paniculate Matter", Am. Ind.
Hyg. Assoc. J. 23, 144-150 (1961).
ADDENDA
A. Alternative Equipment.
A modification of the high-volume sam-
pler incorporating a method for recorjing
the actual airflow, over the entire sampling
period has been described, and is acceptable
for measuring the concentration of suspend-
ed particulates (Henderson. J. S., Eighth
Conference on Methods in Air Pollution and
Industrial Hygiene Studies. 1967, Oakland,
Calif.). This modification consists of an ex-
haust orifice meter assembly connected
through a transducer to a system for con-
tinuously recording airflow on a circular
chart. The volume of air sampled is calculat-
ed by the following equation:
V = QxT
Q = Average sampling rate, m. Vmin.
T=Sampling time, minutes.
The average sampling rate, Q, is determined
from the recorder chart by estimation if the
flow rate does not vary more than 0.11 m.V
min. (4 ft. Vmin.) during the sampling
period. If the flow rate does vary more than
0.11 m. ' (4 ft. Vmin.) during the sampling
period, read the flow rate from the chart at
2-hour intervals and take the average.
B. Pressure and Temperature Corrections.
If the pressure or temperature during
high-volume sampler calibration is substan-
tially different from the pressure or tem-
perature during orifice calibration, a correc-
tion of the flow rate, Q, may be required. If
the pressures differ by no more than 15 per-
cent and the temperatures differ by no
more than 100 percent ("O, the error in the
uncorrected flow rate will be no more than
15 percent. If necessary, obtain the correct-
ed flow rate as directed below. This correc-
tion applies only to orifice meters having a
constant orifice coefficient. The coefficient
for the calibrating orifice described in 5.1.4
has been shown experimentally to be con-
stant over the normal operating range of
the high-volume sampler (0.6 to 2.2 m.V
min., 20 to 78 ft. Vmin.). Calculate corrected
flow rate:
Q., = Corrected flow rate, m. Vmin.
Q, = F'low rate during high-volume sampler
calibration (Section 8.1.2), m.Vmin.
T, = Absolute temperature during orifice
unit calibration (Section 8.1.1), °K or 'R.
P, = Barometric pressure during orifice unit
calibration (Section 8.1.1), mm. Hg.
Tj = Absolute temperature during high-
volume sampler calibration (Section
8.1.2), "Kor °R.
P, = Barometric pressure during high-
volume sampler calibration (Section
8.1.2), mm Hg.
A2-6
------- An error occurred while trying to OCR this image.
-------
ORIFICE
RESISTANCE PLATES
Figure 63. Orifice calibration unit.
A2-8
-------
Appendix C
APPENDIX C—MEASUREMENT PRINCIPLE AND
CALIBRATION PROCEDURE FOR THE CONTINU-
OUS MEASUREMENT OF CARBON MONOXIDE
IN THE ATMOSPHERE (NON-DISPERSIVE IN-
FRARED SPECTROMETRY)
1. Principle and applicability.
1.1 This principle is based on the absorp-
tion of infrared radiation by carbon monox-
ide in a non-dispersive photometer. Both
beams pass into matched cells, each contain-
ing a selective detector and CO. The CO in
the cells absorb infrared radiation only at
its characteristic frequencies and the detec-
tor is sensitive to those frequencies. With a
non-absorbing gas in the reference cell, and
with no CO in the sample cell, the signals
from both detectors are balanced electroni-
cally. Any CO introduced into the sample
cell will absorb radiation, which reduces the
temperature and pressure in the detector
cell and displaces a diaphram. This displace-
ment is detected electronically and ampli-
fied to provide an output signal,
1.2 An analyzer based on this principle
will be considered a reference method only
if it has been designated as a reference
method in accordance with Part 53 of this
chapter.
2.—6. [Reserved]
7. Procedure.
7.1 Calibrate the instrument as described
in 8.1. All gases (sample, zero, calibration,
and span) must be introduced into the
entire analyzer system. Figure Cl shows a
typical flow diagram. For specific operating
instructions, refer to the manufacturer's
manual.
8. Calibration.
8.1 Calibration Curve. Determine the lin-
earity of the detector response at the oper-
ating flow rate and temperature. Prepare a
calibration curve and check the curve fur-
nished with the instrument. Introduce zero
gas and set the zero control to indicate a re-
corder reading of zero. Introduce span gas
and adjust the span control to indicate the
proper value on the recorder scale (e.g. on
0-58 mg./m.' scale, set the 46 mg./m.'
standard at 80 percent of the recorder
chart). Recheck zero and span until adjust-
ments are no longer necessary. Introduce in-
termediate calibration gases and plot the
values obtained. If a smooth curve is not ob-
tained, calibration gases may need replace-
ment.
9. Calculations.
9.1 Determine the concentrations direct-
ly from the calibration curve. No calcula-
tions are necessary.
9.2 Carbon monoxide concentrations in
mg./m.3 are converted to p.p.m. as follows:
p.p.m.CO = mg. CO/m.nv 0.873
10. Bibliography.
The Intech NDIR-CO Analyzer by Frank
McEl^oy. Presented at the llth Methods
Conference in Air Pollution, University of
California, Berkeley, Calif.. April 1. 1970.
Jacobs. M. B. et al., J.A.P.C.A. 9, No. 2,
110-114. August 1959.
MSA LIRA Infrared Gas and Liquid Ana-
lyzer Instruction Book, Mine Safety Appli-
ances Co., Pittsburgh, Pa.
Beckman Instruction 1635B, Models 215A,
315A and 415A Infrared Analyzers, Beck-
man Instrument Company, Fullerton, Calif.
Continuous CO Monitoring System, Model
A 5611. Intertech Corp., Princeton, N.J.
Bendix-UNOR Infrared Gas Analyzers.
Ronceverte, W. Va.
[36 FR 22384, Nov. 25, 1971, as amended at
40 FR 7043, Feb. 18. 1975)
A2-9
-------
Appendix D
APPENDIX D—MEASUREMENT PRINCIPLE AND
CALIBRATION PROCEDURE FOR THE MEASURE-
MENT OF OZONE IN THE ATMOSPHERE
MEASUREMENT PRINCIPLE
1. Ambient air and ethylene are delivered
simultaneously to a mixing zone where the
ozone in the air reacts with the ethylene to
emit light, which is detected by a photomul-
tiplier tube. The resulting photocurrent is
amplified and is either read directly or dis-
played on a recorder.
2. An analyzer based on this principle will
be considered a reference method only if it
has been designated as a reference method
in accordance with Part 53 of this chapter
and calibrated as follows:
CALIBRATION PROCEDURE
1. Principle. The calibration procedure is
based on the photometric assay of ozone
(O,) concentrations in a dynamic flow
system. The concentration of d in an ab-
sorption cell is determined from a measure-
ment of the amount of 254 nm light ab-
sorbed by the sample. This determination
requires knowledge of (1) the absorption co-
efficient (a) of O3 at 254 nm, (2) the optical
path length (I) through the sample, (3) the
transmittance of the sample at a wave-
length of 254 nm, and (4) the temperature
(T) and pressure (P) of the sample. The
transmittance is defined as the ratio I/I,,
where I is the intensity of light which
passes through the cell and is sensed by the
detector when the cell contains an d
sample, and !„ is the intensity of light which
passes through the cell and is sensed by the
detector when the cell contains zero air. It is
assumed that all conditions of the system,
except for the contents of the absorption
cell, are identical during measurement of I
and I0. The quantities defined above are re-
lated by the Beer-Lambert absorption law.
Transmittance
-ad
where:
a=absorption coefficient of O, at 254
nm = 308±4 atnr1 cm'1 at 0°C and 760
torr." » »«•*«•»
c = Oi concentration in atmospheres
/=optical path length in cm
In practice, a stable O, generator is used
to produce O, concentrations over the re-
quired range. Each O, concentration is de-
termined from the measurement of the
transmittance (I/I0) of the sample at 254 nm
with a photometer of path length I and cal-
culated from the equation.
c(atm) « - -I
or,
c(ppn)
(In I/I0)
(2a)
(2b)
The calculated O, concentrations must be
corrected for O, losses which may occur in
the photometer and for the temperature
and pressure of the sample.
2. Applicability. This procedure is applica-
ble to the calibration of ambient air O3 ana-
lyzers, either directly or by means of a
transfer standard certified by this proce-
dure. Transfer standards must meet the re-
quirements and specifications set forth in
Reference 8.
3. Apparatus. A complete UV calibration
system consists of an ozone generator, an
output port or manifold, a photometer, an
appropriate source of zero air, and other
components as necessary. The configuration
must provide a stable ozone concentration
at the system output and allow the photom-
eter to accurately assay the output concen-
tration to the precision specified for the
photometer (3.1). Figure 1 shows a common-
ly used configuration and serves to illus-
trate the calibration procedure which fol-
lows. Other configurations may require ap-
propriate variations in the procedural steps.
All connections between components in the
calibration system downstream of the d
generator should be of glass, Teflon, or
other relatively inert materials. Additional
information regarding the assembly of a UV
photometric calibration apparatus is given
in Reference 9. For certification of transfer
standards which provide their own source of
O,, the transfer standard may replace the
O, generator and possibly other components
shown in Figure 1; see Reference 8 for guid-
ance.
3.1 UV photometer. The photometer con-
sists of a low-pressure mercury discharge
lamp, (optional) collimation optics, an ab-
sorption cell, a detector, and signal-process-
ing electronics, as illustrated in Figure 1. It
must be capable of measuring the transmit-
tance, I/I, at a wavelength of 254 nm with
sufficient precision such that the standard
deviation of the concentration measure-
ments does not exceed the greater of 0.005
ppm or 3% of the concentration. Because
the low-pressure mercury lamp radiates at
several wavelengths, the photometer must
incorporate suitable means to assure that
no O, is generated in the cell by the lamp,
and that at least 99.5% of the radiation
sensed by the detector is 254 nm radiation.
(This can be readily achieved by prudent se-
lection of optical filter and detector re-
sponse characteristics.) The length of the
light path through the absorption cell must
be known with an accuracy of at least
99.5%. In addition, the cell and associated
plumbing must be designed to minimize loss
of O> from contact with cell walls and gas
handling components. See Reference 9 for
additional information.
3.2 Air flow controllers. Devices capable
of regulating air flows as necessary to meet
the output stability and photometer preci-
sion requirements.
3.3 Ozone generator. Device capable of
generating stable levels of Os over the re-
quired concentration range.
3.4 Output manifold. The output mani-
fold should be constructed of glass, Teflon,
or other relatively inert material, and
should be of sufficient diameter to insure a
negligible pressure drop at the photometer
connection and other output ports. The
system must have a vent designed to insure
atmospheric pressure in the manifold and to
prevent ambient air from entering the
manifold.
3.5 Two-way valve. Manual or automatic
valve, or other means to switch the photom-
eter flow between zero air and the O> con-
centration.
3.6 Temperature indicator. Accurate to
±\'C.
3.7 Barometer or pressure indicator. Ac-
curate to ±2 torr.
4. Reagents.
4.1 Zero air. The zero air must be free of
contaminants which would cause a detect-
able response from the 03 analyzer, anJ it
should be free of NO. C,H, and other spe-
cies which react with O,. A procedure for
generating suitable zero air is given in Ref-
erence 9. As shown in Figure 1, the zero air
supplied to the photometer cell for the 1,
reference measurement must be derived
from the same source as the zero air used
for generation of the ozone concentration to
be assayed (I measurement). When using
the photometer to certify a transfer stand-
ard having its own source of ozone, see Ref-
erence 8 for guidance on meeting this re-
quirement.
5. Procedure.
5.1 General operation. The calibration
photometer must be dedicated exclusively
to use as a calibration standard. It should
always be used with clean, filtered calibra-
tion gases, and never used for ambient a ir
sampling. Consideration should be given to
locating the calibration photometer in a
clean laboratory where it can be stationary,
protected from physical shock, operated by
a responsible analyst, and used as a common
standard for all field calibrations via trans-
fer standards.
5.2 Preparation. Proper operation of the
photometer is of critical importance to the
accuracy of this procedure. The following
steps will help to verify proper operation.
The steps are not necessarily required prior
to each use of the photometer. Upon initial
operation of the photometer, these steps
should be carried out frequently, with all
quantitative results or indications recorded
in a chronological record either in tabular
form or plotted on a graphical chart. As the
performance and stability record of the
photometer is established, the frequency of
these steps may be reduced consistent with
the documented stability of the photometer.
5.2.1 Instruction manual: Carry out all
set up and adjustment procedures or checks
as described in the operation or instruction
manual associated with the photometer.
5.2.2 System check: Check the photom-
eter system for integrity, leaks, cleanliness,
proper flowrates, etc. Service or replace fil-
ters and zero air scrubbers or other consum-
able materials, as necessary.
5.2.3 Linearity: Verify that the photom-
eter manufacturer has adequately estab-
lished that the linearity error of the pho-
tometer is less than 3%, or test the linearity
by dilution as follows: Generate and assay
an O3 concentration near the upper range
limit of the system (0.5 or 1.0 ppm), then ac-
curately dilute that concentration with zero
air and reassay it. Repeat at several differ-
ent dilution ratios. Compare the assay of
the original concentration with the assay of
the diluted concentration divided by the di-
lution ratio, as follows
E =
x 100%
(3)
A2-11
-------
where:
E = linearity error, percent
A, = assay of the original concentration
Ajr assay of the diluted concentration
R = dilution ratio=flow of original concen-
tration divided by the total flow
The linearity error must be less than 5%.
Since the accuracy of the measured flow-
rates will affect the linearity error as meas-
ured this way, the test is not necessarily
conclusive. Additional information on veri-
fying linearity is contained in Reference 9.
5.2.4 Intercomparison: When possible,
the photometer should be occasionally in-
tercompared, either directly or via transfer
standards, with calibration photometers
used by other agencies or laboratories.
5.2.5 Ozone losses: Some portion of the
O, may be lost upon contact with the pho-
tometer cell walls and gas handling compo-
nents. The magnitude of this loss must be
determined and used to correct the calculat-
ed O, concentration. This loss must not
exceed 5%. Some guidelines for quantita-
tively determining this loss are discussed in
Reference 9.
5.3 Assay of O, concentrations.
5.3.1 Allow the photometer system to
warm up and stabilizer.
5.3.2 Verify that the flowrate through
the photometer absorption cell, P allows the
cell to be flushed in a reasonably short
period of time (2 liter/min is a typical flow).
The precision of the measurements is in-
versely related to the time required for
flushing, since the photometer drift error
increases with time.
5.3.3 Insure that the flowrate into the
output manifold is at least 1 liter/min great-
er than the total flowrate required by the
photometer and any other flow demand
connected to the manifold.
5.3.4 Insure that the flowrate of zero air,
F,, is at least 1 liter/min greater than the
flowrate required by the photometer.
5.3.5 With zero air flowing in the output
manifold, actuate the two-way valve to
allow the photometer to sample first the
manifold zero air, then P,. The two photom-
eter readings must be equal (I = I0).
NOTE: In some commercially available
photometers, the operation of the two-way
valve and various other operations in sec-
tion 5.3 may be carried out automatically by
the photometer.
5.3.6 Adjust the O, generator to produce
an Oj concentration as needed.
5.3.7 Actuate the two-way valve to allow
the photometer to sample zero air until the
absorption cell is thoroughly flushed and
record the stable measured value of I0.
5.3.8 Actuate the two-way valve to allow
the photometer to sample the ozone concen-
tration until the absorption cell is thor-
oughly flushed and record the stable meas-
ured value of I.
5.3.9 Record the temperature and pres-
sure of the sample in the photometer ab-
sorption cell. (See Reference 9 for guid-
ance.)
5.3.10 Calculate the O. concentration
from equation 4. An average of several de-
terminations will provide better precision.
where:
COj]ol,T=Oj concentration, ppm
a=absorption coefficient of O, at 254
nm=308 atnr1 cnr' at 0"C and 760 torr
1=optical path length, cm
T=sample temperature, K
P=sample pressure, torr
L=correction factor for O, losses from
5.2.5=(l-fraction O, lost).
NOTE: Some commercial photometers may
automatically evaluate all or part of equa-
tion 4. It is the operator's responsibility to
verify that all of the information required
for equation 4 is obtained, either automati-
cally by the photometer or manually. For
'automatic" photometers which evaluate
the first term of equation 4 based on a
linear approximation, a manual correction
may be required, particularly at higher O,
levels. See the photometer instruction
manual and Reference 9 for guidance.
5.3.11 Obtain additional O, concentration
standards as necessary by repeating steps
5.3.6 to 5.3.10 or by Option 1.
5.4 Certification of transfer standards. A
transfer standard is certified by relating the
output of the transfer standard to one or
more ozone standards as determined accord-
ing to section 5.3. The exact procedure
varies depending on the nature and design
of the transfer standard. Consult Reference
8 for guidance.
5.5 Calibration of ozone analyzers.
Ozone analyzers are calibrated as follows,
using ozone standards obtained directly ac-
cording to section 5.3 or by means of a certi-
fied transfer standard.
5.5.1 Allow sufficient time for the Oj ana-
lyzer and the photometer or transfer stand-
ard to warmup and stabilize.
5.5.2 Allow the O, analyzer to sample
zero air until a stable response is obtained
and adjust the O, analyzer's zero control.
Offsetting the analyzer's zero adjustment to
+ 5% of scale is recommended to facilitate
observing negative zero drift. Record the
stable zero air response as "Z".
5.5.3 Generate an O3 concentration
standard of approximately 80% of the de-
sired upper range limit (URL) of the O3 ana-
lyzer. Allow the O, analyzer to sample this
O, concentration standard until a stable re-
sponse is obtained.
5.5.4 Adjust the O, analyzer's span con-
trol to obtain a convenient recorder re-
sponse as indicated below:
ard, record the O, and the corresponding
analyzer response.
5.5.6 Plot the O, analyzer responses
versus the corresponding O, concentrations
and draw the O3 analyzer's calibration curve
or calculate the appropriate response factor.
5.5.7 Option 1: The various O, concentra-
tions required in steps 5.3.11 ar 1 5.5.5 may
be obtained by dilution of the Ga concentra-
tion generated in steps 5.3.6 and 5.5.3. With
this option, accurate flow measurements are
required. The dynamic calibration system
may be modified as shown in Figure 2 to
allow for dilution air to be metered in down-
stream of the O3 generator. A mixing cham-
ber between the O, generator and the
output manifold is also required. The flow-
rate through the O, generator (F0) and the
dilution air flowrate (FD) are measured with
a reliable flow or volume standard traceable
to NBS. Each O3 concentration generated by
dilution is calculated from:
recorder response U scale) • ( u^— « 100) * Z
(5)
(4 in})
o
, >
(77,)
,760,
-
10
T
where:
URL= upper range limit of the Oa analyz-
er, ppm
Z=recorder response with zero air, % scale
Record the O3 concentration and the cor-
responding analyzer response. If substantial
adjustment of the span control is necessary.
recheck the zero and span adjustments by
repeating steps 5.5.2 to 5.5.4.
5.5.5 Generate several other O, concen-
tration standards (at least 5 others are rec-
ommended) over the scale range of the O,
analyzer by adjusting the O, source or by
Option 1. For each Oj concentration stand-
Co,]
3JOUT
[0,]
'3JOUT
(6)
where:
[O]]'0LT=diluted Q3 concentration, ppm
F«== flowrate through the O, generator,
liter/min
Fp=diluent air flowrate, liter/min
REFERENCES
1. E.C.Y. Inn and Y. Tanaka, "Absorption
coefficient of Ozone in the Ultraviolet and
Visible Regions", J. Opt. Soc. Am., 43, 870
(1953).
2. A. G. Hearn, "Absorption of Ozone in
the Ultraviolet and Visible Regions of the
Spectrum". Proc. Phys. Soc. (London), 78,
932 (1961).
3. W. B. DeMore and O. Raper, "Hartley
Band Extinction Coefficients of Ozone in
the Gas Phase and in Liquid Nitrogen,
Carbon Monoxide, and Argon", /. Phys.
Chem.. 68, 412 (1964).
4. M. Griggs, "Absorption Coefficients of
Ozone in the Ultraviolet and Visible Re-
gions". J. Chem. Phys., 49, 857 (1968).
5. K. H. Becker, U. Schurath, and H. Seitz,
"Ozone Olefin Reactions in the Gas Phase.
1. Rate Constants and Activation Energies",
Intl Jour, of Chem. Kinetics, VI, 725 (1974).
6. M. A. A. Clyne and J. A. Coxom, "Kinet-
ic Studies of Oxy-halogen Radical Systems",
Proc. Roy. Soc., A303, 207 (1968).
7. J, W. Simons, R. J. Paur, H. A. Webster,
and E. J. Bair, "Ozone Ultraviolet Photoly-
sis. VI. The Ultraviolet Spectrum", J. Chem.
Phys., 59, 1203 (1973).
8. "Transfer Standards for Calibration of
Ambient Air Monitoring Analyzers for
Ozone', EPA Publication available from
EPA, Department E (MD-77), Research Tri-
angle Park, N.C. 27711.
9. "Technical Assistance Document for
the Calibration of Ambient Ozone Moni-
tors". EPA Publication available from EPA,
Department E (MD-77). Research Triangle
Park, N.C. 27711.
A2-12
------- An error occurred while trying to OCR this image.
------- An error occurred while trying to OCR this image.
-------
Appendix E
APPENDIX E—REFERENCE METHOD FOR DETER-
MINATION OF HYDROCARBONS CORRECTED
FOR METHANE
1. Principle and Applicability.
1.1 Measured volumes of air are delivered
semicontinuously (4 to 12 times per hour) to
a hydrogen flame ionization detector to
measure its total hydrocarbon (THC) con-
tent. An aliquot of the same air sample is in-
troduced into a stripper column which re-
moves water, carbon dioxide, and hydrocar-
bons other than methane. Methane and
carbon monoxide are passed quantitatively
to a gas chromatographic column where
they are separated. The methane is eluted
first, and is passed unchanged through a
catalytic reduction tube into the flame ion-
ization detector. The carbon monoxide is
eluted into the catalytic reduction tube
where it is reduced to methane before pass-
ing through the flame ionization detector.
Between analyses the stripper column is
backflushed to prepare it for subsequent
analysis Hydrocarbon concentrations cor-
rected for methane are determined by sub-
tracting the methane value from the total
hydrocarbon value.
Two modes of operation are possible: (DA
complete chromatographic analysis showing
the continuous output from the detector for
each sample injection; (2) The system is pro-
gramed for automatic zero and span to dis-
play selected band widths of the chromato-
gram. The peak height is then used as the
measure of the concentration. The former
operation is referred to as the chromatogra-
phic or spectro mode and the latter as the
barographic or "normal" mode depending
on the make of analyzer.
1.2 The method is applicable to the semi-
continuous measurement of hydrocarbons
corrected for methane in ambient air. The
carbon monoxide measurement, which is si-
multaneously obtained in this method, is
not required in making measurements of
hydrocarbons corrected for methane and
will not be dealt with here.
2. Range and Sensitivity.
2.1 Instruments are available with var-
ious range combinations. For atmospheric
analysis the THC range is 0-13.1 mg./m.3
(0-20 p.p.m.) carbon (as CH.) and the meth-
ane range is 0-6.55 mg./m.3 (0-10 p.p.m.).
For special applications, lower ranges are
available and in these applications the
range for THC is 0-1.31 mg./m.3 (0-2 p.p.m.)
carbon (as CH.) and for methane the range
is 0-1.31 mg./m.3 (0-2 p.p.m.).
2.2 For the higher, atmospheric analysis
ranges the sensitivity for THC is 0.065 mg./
m.1 (0.1 p.p.m.) carbon (as CH,) and for
methane the sensitivity is 0.033 mg./m.1
(0.05 p.p.m.). For the lower, special analysis
ranges the sensitivity is 0.016 mg./m. ' (0.025
p.p.m.) for each gas.
3. Interferences.
3.1 No interference in the methane mea-
surement has been observed. The THC mea-
surement typically includes all or a portion
of what is generally classified as the air
peak interference. This effect is minimized
by proper plumbing arrangements or is ne-
gated electronically.
4. Precision, Accuracy, and Stability.
4.1 Precision determined with calibration
gases is ±0.5 percent of full scale in the
higher, atmospheric analysis ranges.
4.2 Accuracy is dependent on instrument
linearity and absolute concentration of the
calibration gases. An accuracy of 1 percent
of full scale in the higher, atmospheric anal-
ysis ranges and 2 percent of full scale in the
lower, special analysis ranges can be ob-
tained.
4.3 Variations in ambient room tempera-
ture can cause changes in performance
characteristics. This is due to shifts in oven
temperature, flow rates, and pressure with
ambient temperature change. The instru-
ment should meet performance specifica-
tions with room temperature changes of ±3°
C. Baseline drift is automatically corrected
in the barographic mode.
5. Apparatus. 5.1 Commercially Available
THC, CH,, and CO Analyzer. Instruments
should be installed on location and demon-
strated, preferably by the manufacturer, or
his representative, to meet or exceed manu-
facturer's specifications and those described
in this method.
5.2 Sample Introduction System. Pump,
flow control valves, automatic switching
valves, and flowmeter.
5.3 Filter (In-line). A binder-free, glass-
fiber filter with a porosity of 3 to 5 microns
should be immediately downstream from
the sample pump.
5.4 Stripper or PrecoZumn. Located out-
side of the oven at ambient temperature.
The column should be repacked or replaced
after the equivalent of 2 months of continu-
ous operation.
5.5 Oven. For containing the analytical
column and catalytic converter. The oven
should be capable of maintaining an elevat-
ed temperature constant within ±0.5° C.
The specific temperature varies with instru-
ment manufacturer.
6. Reagents.
6.1 Combustion Gas. Air containing less
than 1.3 mg./m.3 (2 p.p.m.) hydrocarbon as
methane.
6.2 Fuel. Hydrogen or a mixture of hy-
drogen and inert gas containing less than
0,065 mg./m.3 (0.1 p.p.m.) hydrocarbons as
methane.
6.3 Carner Gas. Helium, nitrogen, air or
hydrogen containing less than 0.005 mg./
m. 1 (0.1 p.p.m.) hydrocarbons as methane.
6.4 Zero Gas. Air containing less than
0.065 mg./m.' (0.1 p.p.m.) total hydrocar-
bons as methane.
6.5 Calibration Gases. Gases needed for
linearity checks (peak heights) are deter-
mined by the ranges used. Calibration gases
corresponding to 10, 20, 40, and 80 percent
of full scale are needed. Gases must be pro-
vided with certification or guaranteed anal-
ysis. Methane is used for both the total hy-
drocarbon measurement and methane mea-
surement.
6.6 Span Gas. The calibration gas corre-
sponding to 80 percent of full scale is used
to span the instrument.
7. Procedure,
7.1 Calibrate the instrument as described
in 8.1. Introduce sample into the system
under the same conditions of pressure and
flow rates as are used in calibration. (The
pump is bypassed only when pressurized cyl-
inder gases are used.) Figure El shows a
typical flow diagram; for specific operating
instructions refer to manufacturer's
manual.
8. Calibration.
8.1 Catibration Curve. Determine the lin-
earity of the system for THC and methane
in the barographic mode by introducing
zero gas and adjusting the respective zero-
ing controls to indicate a recorder reading
of zero. Introduce the span gas and adjust
the span control to indicate the proper
value on the recorder scale. Recheck zero
and span until adjustments are no longer
necessary. Introduce intermediate calibra-
tion gases and plot the values obtained. If a
smooth curve is not obtained, calibration
gases may need replacement.
9. Catenation.
9.1 Determine concentrations of total hy-
drocarbons (as CH.) and CH,, directly from
the calibration curves. No calculations are
necessary.
9.2 Determine concentration of hydro-
carbons corrected for methane by subtract-
ing the methane concentration from the
total hydrocarbon concentration.
9.3 Conversion between p.p.m. and mg./
m.3 values for total hydrocarbons (as CH,)
methane and hydrocarbons corrected for
methane are made as follows:
p.p.m. carbon (as CH.)=Cmg. carbon (as
CH,)/m.3]xl.53
10. Bibliography.
Fee, G.. "Multi-Parameter Air Quality
Analyzer", ISA Proceedings AID/CHEM-
PID Symposium, Houston, Texas, April 19-
21, 1911.
Villalobos, R., and Chapman, R. L., "A
Gas Chromatographic Method for Automat-
ic Monitoring of Pollutants in Ambient Air",
ibid.
Stevens, R. K., "The Automated Gas
Chromatograph as an Air Pollutant Moni-
tor", 1970 Conference on Environmental
Toxicology, U.S. Air Force, Wright-Patter-
son Air Force Base, Dayton, Ohio.
Stevens, R. K., and O'Keeffe, A. E.. Anal
Chem. 42, 143A(1970).
Schuck, E. A., Altshuller, A. P., Barth, D.
S. and Morgan, G. B., "Relationship of Hy-
drocarbons to Oxidants in Ambient Atmos-
pheres", J. Air Poll. Cant. Assoc. 20, 297-302
(1970).
Stevens, R. K., O'Keeffe, A. E., and
Ortman, G. C., "A Gas Chromatographic
Approach to the Semi-Continuous Monitor-
ing oi Atmospheric Carbon Monoxide and
Methane", Proceed.ngs of llth Conference
on Methods in Air Pollution on Industrial
Hygiene Studies, Be.-keley, Calif., March 30-
April 1, 1970.
Swinnerton, J. W., Linnenbom, V. J. and
Check, C. H.. Environ. Sci. Technol 3, 836
(1969).
Williams, I. G., Advances in Chromato-
grapfiy, Giddings, J. C., and Keller, R. A.,
editors, Marcell Dekker, N.Y. (1968), pp.
173-182.
Altshuller, A. P.. Kopcznski, S. L., Lonne-
man, W. A., Becker, T. L. and Slater, R., En-
viron. Sci. Technol. 1, 899 (1967).
Altshuller, A. P.. Cohen, I. R., and Purcell,
T. C., Con. J. Chem., 44, 2973 (1966).
DuBois, L., Zdrojewski, A., and Monkman,
J. L., J. Air Poll. Cont. Assoc. 16, 135 (1966).
Ortman, G. C., Anal. Chem. 38, 644-646
(1966).
Porter, K., and Volman, D. H., AnaL
Chem. 34, 748-749 (1962).
Crum, W. M., Proceedings, National Anal-
ysis Instrumentation Symposium ISA, 1962.
Schwink, A., Hochenberg, H., and Forder-
reuther, M., Brennstoff-Chemie 72, No. 9,
295 (1961).
Instruction Manual for Air Quality Chro-
matograph Model 6800, Beckman Instru-
ment Co., Fullerton, Calif.
Instruction Manual, Bendix Corp., Ronce-
verte. W. Va.
A2-15
-------
Instruction Manual, Byron Instrument
Co., Raleigh, N.C.
MSA Instruction Manual for GC Process
Analyzer for Total Hydrocarbon, Methane
and Carbon Monoxide, Pittsburgh, Pa.
Monsanto Enviro-Chem System for Total
Hydrocarbons, Methane and Carbon Mon-
oxide Instruction Manual, Dayton, Ohio.
Union Carbide Instruction Manual for
Model 3020 Gas Chromatograph for CO-
CH.-T/1, White Plains, N.Y.
Instruction Manual for 350 P Analyzer,
Tracor Inc., Austin, Tex.
ADDENDA
A. Suggested Performance Specifications
for Atmospheric Analyzers for Hydrocarbons
Corrected for Methane:
Range (minimum)
Output (minimum)
Minimum detectable sensitivity
Zero drift (maximum) .
Span drift (maximum)
Precision (minimum)
Operational period (minimum)
Operating temperature range
(minimum)
Operating humidity range (mini-
mum)
Linearity (maximum)
0 3 mg /m 1 (0-5 p p m )
THC
0-3 mg /m ' (0-5 p p m )
CH.
0-10 mv full scale
0 1 p p m THC
0 1 p p m CH,
Not to exceed 1 percent/
24 hours.
Not to exceed 1 percent/
24 hours.
±05 percent
3 days
5-40° C
10-100 percent
1 percent of full scale
B. Suggested Definitions of Performance
Specifications:
Range—The minimum and maximum mea-
surement limits.
Output—Electrical signal which is propor-
tional to the measurement; intended for
connection to readout or data processing
devices. Usually expressed as millivolts or
milliamps full scale at a given impedence.
Full Scale—The maximum measuring limit
for a given range.
Minimum Detectable Sensitivity—The
smallest amount of input concentration
that can be detected as the concentration
approaches zero.
Accuracy—The degree of agreement be-
tween a measured value and the true
value; usually expressed at ± percent of
full scale.
Lag Time—The time interval from a step
change in input concentration at the in-
strument inlet to the first corresponding
change in the instrument output.
Time to 90 Percent Response—The time in-
terval from a step change in the input
concentration at the instrument inlet to a
reading of 90 percent of the ultimate re-
corded concentration.
Rise Time (90 percent)—The interval be-
tween initial response time and time to 90
percent response after a step decrease in
the inlet concentration.
Zero Drift—The change in instrument
output over a stated time period, usually
24 hours, of unadjusted continuous oper-
ation, when the input concentration is
zero; usually expressed as percent full
scale.
Span Drift—The change in instrument
output over a stated time period, usually
24 hours, of unadjusted continuous oper-
ation, when the input concentration is a
stated upscale value; usually expressed as
percent full scale.
Precision—The degree of agreement be-
tween repeated measurements of the same
concentration. It is expressed as the aver-
age deviation of the single -esults from
the mean.
Operational Period—The period ol time
over which the instrument can be expect-
ed to operate unattended within specifica-
tions.
Noise—Spontaneous deviations from a mean
output not caused by input concentration
changes.
Interference—An undesired positive or neg-
ative output caused by a substance other
than the one being measured.
Interference Equivalent—The portion of in-
dicated input concentration due to the
presence of an interferent.
Operating Temperature Range—The range
of ambient temperatures over which the
instrument will meet all performance
specifications.
Operating Humidity Range—The range of
ambient relative humidity over which the
instrument will meet all performance
specifications.
Linearity—The maximum deviation between
an actual instrument reading and the
reading predicted by a straight line drawn
bttu, een upper and lower calibration
points.
A2 16
-------
Appendix F
APPENDIX P—MEASUREMENT PRINCIPLE AND
CALIBRATION PROCEDURE FOR THE MEASURE-
MENT or NITROGEN DIOXIDE IN THE ATMOS-
PHERE (GAS PHASE CHEMILUMINESCENCE)
Principle and Applicability
1. Atmospheric concentrations of nitrogen
dioxide (NO.) are measured indirectly by
photometrically measuring the light intensi-
ty, at wavelengths greater than 600 nano-
meters, resulting from the chemilumtnes-
cent reaction of nitric oxide (NO) with
ozone (O,). (1,2,3) NO, is first quantitatively
reduced to NO(4,5,6) by means of a convert-
er. NO, which commonly exists in ambient
air together with NO,, passes through the
converter unchanged causing a resultant
total NO, concentration equal to NO+NO,.
A sample of the input air is also measured
without having passed through the convert-
ed. This latter NO measurement is subtract-
ed from the former measurement
(NO+NO,) to yield the final NO, measure
ment. The NO and NO-tNO, measurements
may be made concurrently with dual sys-
tems, or cyclically with the same system
provided the cycle time does not exceed 1
minute.
2. Sampling considerations.
2.1 Chemiluminescence NO/NO./NO,
analyzers will respond to other nitrogen
containing compounds, such as peroxyacetyl
nitrate (PAN), which might be reduced to
NO in the thermal converter. (7) Atmos-
pheric concentrations of these potential in-
terferences are generally low relative to NO,
and valid NO, measurements may be ob-
tained. In certain geographical areas, where
the concentration of these potential inter-
ferences is known or suspected to be high
relative to NO,, the use of an equivalent
method for the measurement of NO, is rec-
ommended.
2.2 The use of integrating flasks on the
sample inlet line of Chemiluminescence NO/
NO./NO, analyzers is optional and left to
couraged. The sample residence time be-
tween the sampling point and the analyzer
should be kept to a minimum to avoid erro-
neous NO, measurements resulting from the
reaction of ambient levels of NO and O, in
the sampling system.
2.3 The use of paniculate filters on the
sample inlet line of Chemiluminescence NO/
NO./NO, analyzers is optional and left to
the discretion of the user or the manufac-
turer. Use of the filter should depend on the
analyzer's susceptibility to interference.
malfunction, or damage due to particulates.
Users are cautioned that particulate matter
concentrated on a filter may cause errone-
ous NO, measurements and therefore filters
should be changed frequently.
3. An analyzer based on this principle will
be considered a reference method only if it
has been designated as a reference method
in accordance with Part 53 of this chapter.
Calibration
1. Alternative A—Gas phase titration
(GPT) of an NO standard with O,.
Major equipment required: Stable O, gen-
erator. Chemiluminescence NO/NO./NO,
analyzer with strip chart recorder(s). NO
concentration standard.
1.1 Principle. This calibration technique
is based upon the rapid gas phase reaction
between NO and O, to produce stoichiome-
tric quantities of NO, in accordance with
the following equation: (.8)
(1)
The quantitative nature of this reaction is
such that when the NO concentration is
known, the concentration of NO, can be de-
termined. Ozone is added to excess NO in a
dynamic calibration system, and the NO
channel of the Chemiluminescence NO/
NO./NO, analyzer is used as an indicator of
changes in NO concentration. Upon the ad-
dition of O,, the decrease in NO concentra-
tion observed on the calibrated NO channel
is equivalent to the concentration of NO,
produced. The amount of NO, generated
may be varied by adding variable amounts
of O, from a stable uncalibrated O, gener-
ator. (9)
1.2 Apparatus. Figure 1, a schematic of a
typical GPT apparatus, shows the suggested
configuration of the components listed
below. All connections between components
in the calibration system downstream from
the O, generator should be of glass, Teflon*
or other non-reactive material.
1.2.1 Air now controllers. Devices capable
of maintaining constant air flows within
±2% of the required flowrate.
1.2.2 NO flow controller. A device capa-
ble of maintaining constant NO flows within
±2% of the required flowrate. Component
parts in contact wiih the NO should be of a
non-reactive material.
1.2.3 Air flowmeters. Calibrated flow-
meters capable of measuring and monitor-
ing air flowrates with an accuracy of ±2%
of the measured flowrate.
1.2.4 NO flowmeter. A calibrated flow-
meter capable of measuring and monitoring
NO flowrates with an accuracy of ±2% of
the measured flowrate. (Rotameters have
been reported to operate unreliably when
measuring low NO flows and are not recom-
mended.)
1.2.5 Pressure regulator for standard NO
cylinder. This regulator must have a non-
reactive diaphragm and internal parts and a
suitable delivery pressure.
1.2.6 Ozone generator. The generator
must be capable of geneiating sufficient and
stable levels of O, for reaction with NO to
generate NO, concentrations in the range
required. Ozone generators of the electric
discharge type may produce NO and NO,
and are not recommended.
1.2.7 Valve. A valve may be used as
shown in Figure 1 to divert the NO flow
when zero air is required at the manifold.
The valve should he constructed of glass.
Teflon^ or other nonreactive material.
1.2.8 'Reaction chamber. A chamber, con-
structed of glass, Teflon", or other nonreac-
tive material, for the quantitative reaction
of O, with excess NO. The chamber should
be of sufficient volume (VHi) such that the
residence time (ta) meets the requirements
specified in 1.4. For practical reasons, U
should be less than 2 minutes.
1.2.9 Mixing chamber. A chamber con-
structed of glass, Teflon*, or other nonreac-
tive material and designed to provide thor-
ough mixing of reaction products and di-
luent air. The residence time is not critical
when the dynamic parameter specification
given in 1.4 is met.
1.2.10 Output manifold. The output
manifold should be constructed of glass, Tef•
Ion*, or other non-reactive material and
should be of sufficient diameter to insure an
insignificant pressure drop at the analyzer
connection. The system must have a vent
designed to insure atmospheric pressure at
the manifold and to prevent ambient air
from entering the manifold.
1.3 Reagents.
1.3,1 NO concentration standard. Cylin-
der containing 50 to 100 ppm NO in N, with
less than 1 ppm NO,. The cylinder must be
traceable to a National Bureau of Standards
NO in N, Standard Reference Material
(SRM 1683 or SRM 1684) or NO, Standard
Reference Material (SRM 1629). Procedures
for certifying the NO cylinder (working
standard) against an NBS traceable NO or
NO, standard and for determining the
amount of NO, impurity are given in refer-
ence 13. The cylinder should be recertified
on a regular basis as determined by the
local quality control program.
1.3.2 Zero air. Air, free of contaminants
which will cause a detectable response on
the NO/NO./NO, analyzer or which* might
react with either NO, O,, or NO, in the gas
phase titration. A procedure for generating
zero air Is given in reference 13.
1.4 Dynamic parameter specification.
1.4.1 The O, generator air flowrate (F0)
and NO flowrate (FN<)) (see Figure 1) must
be adjusted such that the following rela-
tionship holds:
2.75 ppm-minutes
VH
„ .
ro + F so
2 minute-
(2)
(4)
where:
P« = dynamic parameter specification, de-
termined empirically, to insure com-
plete reaction of the available O,,
ppm-minute
[NO1«( =NO concentration in the reaction
chamber, ppm
„ = residence time of the reactant gases in
the reaction chamber, minute
[NO]STD=concentration of the undiluted
NO standard, ppm
FNO = NO flowrate, scmVmiri
F0=O, generator air flowrate, scm Vmin
VK< = volume of the reaction chamber,
scm3
1.4.2 The flow conditions to be used in
the GPT system are determined by the fol-
lowing procedure:
(a) Determine FT, the total flow required
at the output manifold (/> = analyzer
demand plus 10 to 50% excess).
(b) Establish [NOloir as the highest NO
concentration (ppm) which will be required
at the output manifold. [NO]0iT should be
approximately equivalent to 90% of the
upper range limit (URL) of the NO, concen-
tration range to be covered.
(c) Determine FN() as
[ATP]..,
(d) Select a convenient or available reac-
tion chamber volume. Initially, a trial VM
may be selected to be in the range of ap-
proximately 200 to 500 scm \
(e) Compute FO as
A2-17
-------
1>75
(f) Compute £„ as
(0)
(7)
Verify that U < 2 minutes. If not, select a
reaction chamber with a smaller VH< •
Compute the diluent air flowrate as
(8)
where:
/V diluent air flowrate, scm Vmin
(h) If F,, turns out to be impractical for
the desired system, select a reaction cham-
ber having a different VM and recompute F,,
and F,,.
NOTE. A dynamic parameter lower than
2.75 ppm-minutes may be used if it can be
determined empirically that quantitative re-
action of O, with NO occurs. A procedure
for making this determination as well as a
more detailed discussion of the above re-
quirements and other related considerations
is given in reference 13.
1.5 Procedure.
1.5.1 Assemble a dynamic calibration
system such as the one shown in Figure 1.
1.52 Insure that all flowmeters are cali-
brated under the conditions of use against a
reliable standard such as a soap-bubble
meter or wet-test meter. All volumetric
flowrates should be corrected to 25° C and
760 rnm Hg. A discussion on the calibration
of flowmeters is given in reference 13.
1.5.3 Precautions must be taken to
remove O, and other contaminants from the
NO pressure regulator and delivery system
prior to the start of calibration to avoid any
conversion of the standard NO to NO,. Fail-
ure to do so can cause significant errors in
calibration. This problem may be minimized
by (1) carefully evacuating the regulator,
when possible, after the regulator has been
connected to the cylinder and before open-
ing the cylinder valve; (2) thoroughly flush-
ing the regulator and delivery system with
NO after opening the cylinder valve; (3) not
removing the regulator from the cylinder
between calibrations unless absolutely nec-
essary. Further discussion of these proce-
dures is given in reference 13.
1.5.4 Select the operating range of the
NO/NO,/NO, analyzer to be calibrated In
order to obtain maximum precision and ac-
curacy for NO, calibration, all three chan-
nels of the analyzer should be set to the
same range. If operation of the NO and NO,
channels on higher ranges is desired, subse-
quent recalibration of the NO and NO,
channels on the higher ranges is recom-
mended.
NOTE: Some analyzer designs may require
identical ranges fer NO, NO,, and NO:
during operation of the analyzer.
1.5.5 Connect the recorder output
cable(s) of the NO/NO,/NO, analyzer to the
input terminals of the strip chart
recorder(s). All adjustments to the analyzer
should be performed based on the appropri-
ate strip chart readings. References to ana-
lyzer responses in the procedures given
below refer to recorder responses.
1.5.6 Determine the GPT flow conditions
required to meet the dynamic parameter
specification as indicated in 1.4.
1.5.7 Adjust the diluent air and O, gener-
ator air flows to obtain the flows deter-
mined in 1.4.2. The total air flow must
exceed the total demand of the analyzer(s)
connected to the output manifold to insure
that no ambient air is pulled into the mani-
fold vent. Allow the analyzer to sample zero
air until stable NO, NO,, and NO, responses
are obtained. After the responses have sta-
bilized, adjust the analyzer zero control(s).
NOTE: Some analyzers may have separate
zero controls for NO, NO,, and NO,. Other
analyzers may have separate zero controls
only for NO and NO,, while still others may
have only one zero control common to all
three channels.
Offsetting the analyzer zero adjustments
to + 5 percent of scale is recommended to fa-
cilitate observing negative zero drift. Record
the stable zero air responses as ZNO, ZNOX,
and ZNO,.
1.5.8 Preparation of NO and NO, calibra-
tion curves.
1.5.8.1 Adjustment of NO span control.
Adjust the NO flow from the standard NO
cylinder to generate an NO concentration of
approximately 80 percent of the upper
range limit (URL) of the NO range. This
exact NO concentration is calculated from:
the amount of NO, impurity in the standard
NO cylinder are given in reference 13. The
exact NO, concentn.tion is calculated from:
(9)
where:
[NO],»T=diluted NO concentration at the
output manifold, ppm
Sample this NO concentration until the NO
and NO, responses have stabilized. Adjust
the NO span control to obtain a recorder re-
sponse as indicated below:
recorder response (percent scale)
(.0)
where:
URL=nominal upper range limit of the
NO channel, ppm
NOTE: Some analyzers may have separate
span controls for NO, NO,, and NO,. Other
analyzers may have separate span controls
only for NO and NO,, while still others may
have only one span control common to all
three channels. When only one span control
is available, the span adjustment is made on
the NO channel of the analyzer.
If substantial adjustment of the NO span
control is necessary, it may be necessary to
recheck the zero and span adjustments by
repeating steps 1.5.7 and 1.5.8.1. Record the
NO concentration and the analyzer's NO re-
sponse.
1.5.8.2 Adjustment of NO, span control.
When adjusting the analyzer's NO, span
control, the presence of any NO, impurity in
the standard NO cylinder must be taken
into account. Procedures for determining
where:
[NOJoir-diluted NO, concentration at
the output manifold, ppm
[NOJ,Mp = concentration of NO, impurity
in the standard NO cylinder, ppm
Adjust the NO* span control to obtain a re-
corder response as indicated below:
recorder response (% scale)
NOTE: If the analyzer has only one span
control, the span adjustment is made on the
NO channel and no further adjustment is
made here for NO,
If substantial adjustment of the NO, span
control is necessary, it may be necessary to
recheck the zero and span adjustments by
repeating steps 1.5.7 and 1.5.8.2. Record the
NO, concentration and the analyzer's NO,
response.
1.5.8.3 Generate several additional con-
centrations (at least five evenly spaced
points across the remaining scale are sug-
gested to verify linearity) by decreasing FNO
or Increasing FD. For each concentration
generated, calculate the exact NO and NO,
concentrations using equations (9) and (11)
respectively. Record the analyzer's NO and
NO, responses for each concentration. Plot
the analyzer responses versus the respective
calculated NO and NO, concentrations and
draw or calculate the NO and NO, calibra-
tion curves. For subsequent calibrations
where linearity can be assumed, these
curves may be checked with a two-point cali-
bration consisting of a zero air point and
NO and NO, concentrations of approximate-
ly 80% of the URL.
1.5.9 Preparation of NO2 calibration
curve.
1.5.9.1 Assuming the NO, zero has been
properly adjusted while sampling zero air in
step 1.5.7, adjust F0 and FD as determined in
1.4.2. Adjust F,,,, to generate an NO concen-
tration near 90% of the URL of the NO
range Sample this NO concentration until
the NO and NO, responses have stabilized.
Using the NO calibration curve obtained in
1.5.8, measure and record the NO concentra-
tion as [NO]om. Using the NO, calibration
curve obtained in 1.5.8, measure and record
the NO, concentration as [NOJort«.
1.5.9.2 Adjust the O, generator to gener-
ate sufficient O, to produce a decrease in
the NO concentration equivalent to approxi-
mately 80% of the URL of the NO, range.
NOTE: If the analyzer has only one or two
span controls, the span adjustments are
made on the~ NO channel or NO and NO,
channels and no further adjustment is made
here for NO,.
If substantial adjustment of the NO, span
control is necessary, it may be necessary to
A2-18
-------
recheck the zero and span adjustments by
repeating steps 1.5.1 and 1.5.9.3. Record the
NO, concentration and the corresponding
analyzer NOt and NO. responses.
gen dioxide are generated by means of a
permeation device. (10) The permeation
device emits NOi at a known constant rate
provided the temperature of the device Is
recorder response (% scale) <
1.5.9.4 Maintaining the same FN0. FO. and
FD as in 1.5.9.1, adjust the ozone generator
to obtain several other concentrations of
NO, over the NO, range (at least five evenly
spaced points across the remaining scale are
suggested). Calculate each NO, concentra-
tion using equation (13) and record the cor-
responding analyzer NO, and NO, re-
sponses. Plot the analyzer's NO, responses
versus the corresponding calculated NO,
concentrations and draw or calculate the
NO, calibration curve.
1.5.10 Determination of converter effi-
ciency.
1.5.10.1 For each NO, concentration gen-
erated during the preparation of the NO,
calibration curve (see 1.5.9) calculate the
concentration of NO, converted from:
INO.lctmv=ENO, W< tNO, W CNO.J™.)
(15)
where:
[NO,]coNv=concentration of NO, convert-
ed, ppm
[NO.Jon,=original NO, concentration prior
to addition of O,, ppm
[NOJrem=NO, concentration remaining
after addition of O-. com
The decrease must not exceed 90% of the
NO concentration determined in step 1.5.9.1.
After the analyzer responses have stabi-
lized, record the resultant NO and NO. con-
centrations as CNO]nm and [NO.]nn,.
1.5.9.3 Calculate the resulting NO, con-
centration from:
lNO,]ooT=[NOJ.,,«-[N'0],.m
where:
[NO,]0irt= diluted NO, concentration at
the output manifold, ppm
[NO].n,=original NO concentration, prior
to addition of O,, ppm
[NO]rem=NO concentration remaining
after addition of O,. ppm
Adjust the NO, span control to obtain a re-
corder response as Indicated below:
NOTE: Supplemental information on cali-
bration and other procedures in this
method are given in reference 13.
Plot [NOJcoNv (y-axis) versus [NO,]OUT
-------
perature mui be adjusted and controlled to
within +0 TC or less of the calibration tem-
perature as monitored with the temperature
measuring device
244 Precautions must be taken to
remove O, and other contaminants from the
NO pressure regulator and delivery system
prior to the start of calibration to avoid any
conversion of the standard NO to NO,. Fail-
ure to do so can cause significant errors in
calibration This problem may be minimized
b.\ (1) Carefully evacuating the regulator,
v\hen possible, after the regulator has been
conm rtod to the cylinder and before open-
ing the cylinder valve; (2) Thoroughly flush-
ing the regulator and delivery system with
NO after opening tt> olmder valve, (3) Not
returning the regulator from the cylinder
between calibrations unless absolutely nec-
essary Further discussion of these proce-
dures is given in reference 13.
245 Select the operating range of the
NO/NO,NO, analyzer to be calibrated. In
order to obtain maximum precision and ac-
curacy for NO, calibration, all three chan-
nels of the analyzer should be set to the
same range If operation of the NO and NO,
channels on higher ranges is desired, subse-
quent recalibration of the NO and NO,
channels on the higher ranges is recom-
mended.
NOTE —Some analyzer designs may require
identical ranges for NO, NO,, and NO,
during operation of the analyzer.
2.4.8 Connect the recorder output
cablets) of the NO/NO,/NO, analyzer to the
input terminals of the strip chart
recorder(s). All adjustments to the analyzer
should be performed based on the appropri-
ate strip chart readings. References to ana-
lyzer responses in the procedures given
below refer to recorder responses.
2.4.7 Switch the valve to vent the flow
from the permeation device and adjust the
diluent air flowrate, F,,, to provide zero air
at the output manifold. The total air flow
must exceed the total demand of the
analyzer(s) connected to the output mani-
fold to insure that no ambient air is pulled
into the manifold vent. Allow the analyzer
to sample zero air until stable NO, NO,, and
NO, responses are obtained. After the re-
sponses have stabilized, adjust the analyzer
zero control(s).
NOTE —Some analyzers may have separate
zero controls for NO, NO,, and NO, Other
analyzers may have separate zero controls
only for NO and NO,, while still others ma>
have only one z.eru common control to all
three channels
Offsetting the analyzer zero adjustments to
-r5% of scale is recommended to facilitate
observing negative zero drift. Record the
stable zero air responses as ZN0. ZN = concent-ation of the undiluted
NO standard, ppm
F,, = diluent air flowrate, scm Vmin
Sample this NO concentration until the NO
and NOX responses have stabilized. Adjust
the NO span control to obtain a recorder re-
sponse as indicated below:
recorder response (% scale)
(17)
•where:
URL = nominal upper range limit of the
NO channel, ppm
NOTE.—Some analyzers may have separate
span controls for NO, NO,, and NO,. Other
analyzers may have separate span controls
only for NO and NO,, while still others may
have only one span control common to all
three channels. When only one span control
is available, the span adjustment Is made on
the NO channel of the analyzer.
If substantial adjustment of the NO span
control is necessary, it may be necessary to
recheck the zero and span adjustments by
repeating steps 2.4.7 and 2.4.8.1. Record the
NO concentration and the analyzer's NO re-
sponse.
2.4.8.2 Adjustment of NO, span control.
When adjusting the analyzer's NO, span
control, the presence of any NO, impurity in
the standard NO cylinder must be taken
into account. Procedures for determining
the amount of NO, impurity in the standard
NO cylinder are given in reference 13. The
exact NO, concentration is calculated from:
([NO18TD+ !N021,MP '
respectively. Record the analyzer's NO and
NO, responses for each concentration. Plot
the analyzer responses versus the respective
calculated NO and NO, concentrations and
draw or calculate the NO and NO, calibra-
tion curves. For subsequent calibrations
where linearity can be assumed, these
curves may be checked with a two-point cali-
bration consisting of a zero poi .t and NO
and NO, concentrations of approximately 80
percent of the URI..
2.4.9 Preparation of NO, calibration
curve.
2.4.9 1 Remove the NO flow. Assuming
the NOs zero has been properly adjusted
while sampling zero air in step 2.4.7, switch
the valve to provide NO, at the output
manifold.
2.4.9.2 Adjust Fi, to generate an NO, con-
centration of approximately 80 percent of
the URL of the NO, range. The total air
flow must exceed the demand of the
analyzer(s) under calibration. The actual
concentration of NO, is calculated from:
RXK
[N'Olc
(10)
(1M
where:
[NO,]0iT = diluted NO, cencentration at
the output manifold, ppm
[NO,],MP^concentration of NO, impurity
in the standard NO cylinder, ppm
Adjust the NO, span control to obtain a con-
• enient recorder response as indicated
below:
recorder response (% scale)
(19)
NOTE: If the analyzer has only one span
control, the span adjustment is made on the
NO channel and no further adjustment is
made here for NO,.
If substantial adjustment of the NO, span
control is necessary, it may be necessary to
recheck the zero and span adjustments by
repeating steps 2.4.7 and 2.4.8.2. Record the
NO, concentration and the analyzer's NO,
response.
2.4.8.3 Generate several additional con-
centrations (at least five evenly spaced
points across the remaining scale are sug-
gested to verify linearity) by decreasing FNO
or increasing F,,. For each concentration
generated, calculate the exact NO and NO,
concentrations using equations (16) and (18)
(20)
where:
[NO,]0,T-diluted NO, concentration at
the output manifold, ppm
R = permeation rate, ^g/min
ff=0.532jil NO,/ng NO, (at 25°C and 760
mm Hg)
Ff = air flowrate across permeation device,
scm Vmin
f,, = diluent air flowrate, scm Vmin
Sample this NO, concentration until the
NO, and NO, responses have stabilized.
Adjust the NO, span control to obtain a re-
corder response as indicated below:
recorder response (percent scale)
= /[NTOjWTxloo\+z (21)
\ Li RL I
NOTE. If the analyzer has only one or two
span controls, the span adjustments are
made on the NO channel or NO and NO,
channels and no further adjustment is made
here for NO,.
If substantial adjustment of the NOS span
control is necessary it may be necessary to
recheck the zero and span adjustments by-
repeating steps 2.4.7 and 2.4.9.2. Record the
NO, concentration and the analyzer's NO,
response. Using the NO, calibration curve
obtained in step 2.4.8, measure and record
the NO, concentration as [NO,]M-
2.4.9.3 Adjust F,, to obtain several other
concentrations of NO, over the NO, range
(at least five evenly spaced points across the
remaining scale are suggested). Calculate
each NO, concentration using equation (20)
and record the corresponding analyzer NO,
and NO, responses. Plot the analyzer's NO,
responses versus the corresponding calculat-
ed NO, concentrations and draw or calculate
the NO, calibration curve.
2.4.10 Determination of converter effi-
ciency.
2.4.10.1 Plot [NO,]M (y-axis) versus
[NO.]0iT (x-axis) and draw or calculate the
converter efficiency curve. The slope of the
curve times 100 is the average converter ef-
ficiency, E(. The average converter efficien-
A2-20
-------
cy must be greater than 96 percent; if it is
less than 96 percent, replace or service the
converter,
NOTE: Supplemental information on cali-
bration and other procedures in this
method are given in reference 13.
3. Frequency of calibration. The frequen-
cy of calibration, as well as the number of
points necessary to establish the calibration
curve and the frequency of other perform-
ance checks, will vary from one analyzer to
another. The user's quality control program
should provide guidelines for initial estab-
lishment of these variables and for subse-
quent alteration as operational experience is
accumulated. Manufacturers of analyzers
should include in their instruction/oper-
ation manuals information and guidance as
to these variables and on other matters of
operation, calibration, and quality control.
REFERENCES
1. A. Fontijn, A. J. Sabadell, and R. J.
Ronco, "Homogeneous Chemiluminescent
Measurement of Nitric Oxide with Ozone,"
Anal. Chem., 42, 575 (1970).
2. D. H Stedman, E. E. Daby, P. Stuhl,
and H. Niki, "Analysis of Ozone and Nitric
Oxide by a Chermluminiscent Method in
Laboratory and Atmospheric Studies of
Photochemical Smog," J. An Poll. Control
Assoc., 22, 260 (1972).
3. B. E. Martin, J. A. Hodgeson. and R. K.
Stevens, "Detection of Nitric Oxide Chemi-
luminescence at Atmospheric Pressure,"
Presented at 164th National ACS Meeting,
New York City, August 1972.
4 J. A. Hodgeson, K. A. Rehme, B. E.
Martin, and R. K. Stevens, "Measurements
for Atmospheric Oxides of Nitrogen and
Ammonia by Chemiluminescence," Present-
ed at 1972 APCA Meeting, Miami, Florida,
June 1972.
5. R. K. Stevens and J. A. Hodgeson, "Ap-
plications of Chemiluminescence Reactions
to the Measurement of Air Pollutants,"
Anal. Chem., 45, 443A <1973).
6. L. P. Breitenbach and M. Shelef, "De-
velopment of a Method for the Analysis of
NO, and NH, by NO-Measuring Instru-
ments," J. Air Poll. Control Assoc., 23, 128
(1973).
7. A M. Winer, J. W. Peters, J. P. Smith,
and J. N. Pitts. Jr., "Response of Commer-
cial Chemiluminescent NO-NO, Analyzers
to Other Nitrogen-Containing Compounds,"
Environ. Sci. Technol., 8, 1118 (1974).
8. K. A. Rehme, B. E. Martin, and J A.
Hodgeson, Tentative Method for the Cali-
bration of Nitric Oxide, Nitrogen Dioxide,
and Ozone Analyzers by Gas Phase Titra-
tion," EPA-R2-73-246, March 1974.
9. J. A. Hodgeson, R. K. Stevens, and B. E.
Martin, "A Stable Ozone Source Applicable
as a Secondary Standard for Calibration of
Atmospheric Monitors," ISA Transactions,
11, 161 (1972).
10. A. E. O'Keeffe and G. C. Ortman. "Pri-
mary Standards for Trace Gas Analysis,"
Anal. Chem., 38, 760 (1966).
11. F. P. Scanngelli, A. E. O'Keeffe, E.
Rosenberg, and J. P. Bell, "Preparation of
Known Concentrations of Gases and Vapors
with Permeation Devices Calibrated Gravi-
metrically," Anal. Chem.. 42, 871 (1970).
12. H. L. Rook, E. E. Hughes, R. S. Fuerst,
and J. H. Margeson, "Operation Character-
istics of NO, Permeation Devices," Present-
ed at 167th National ACS Meeting, Los An-
geles, California. April 1974.
13. E. C. Ellis, "Technical Assistance Doc-
ument for the Chemiluminescence Measure-
ment of Nitrogen Dioxide," EPA-E600/4-
75-003 (Available in draft form from the
United States Environmental Protection
Agency, Department E (MD-76), Environ-
mental Monitoring and Support Laboratory,
Research Triangle Park, North Carolina
27711).
'o ^
"l
FLOW
CONTROLISR
FlOWMfTER
V»IV1
(
\
[IIRAOUIlSTSCAPffO
*M!II NOT IN US!
OUTfUT
MANKOIO
TOINUTOt ANALV!t«
uNomcAinnATion
! o' a typical GPT calibration ivltti
UNOIHCAU»»Tlt«
F>gurt ? Schematic dijflum ol a tvpKal e«lib*llio*> apoafllul uy"9 i" *^D^ w—Mt'O" fl«vic*
(Sec. 4. Pub. L. 91-604, 84 Stat. 1678 (42 U.S.C. 1857c-4»
[41 FR 52688. DM. 1.1978)
A2 21
-------
Appendix G
APPENDIX G—REFERENCE METHOD FOR THE
DETtRMrNATION OF LEAD IN SUSPENDED PAR-
TICTJIATC MATTER COLLECTED FROM AMBI-
ENT AIR
1. Principle and applicability.
1.1 Ambient air suspended participate
matter is collected on a glass-liber filter for
24 hours using a high volume air sampler.
1.2 Lead in the particulate matter is solu-
bilized by extraction with nitric acid
(HNO,), facilitated by heat or by a mixture
of HNO, and hydrochloric acid (HC1) facili-
tated by ultrasonication.
1.3 The lead content of the sample is
analyzed by atomic absorption spectrometry
using an air-acetylene flame, the 283.3 or
217.0 nm lead absorption line, and the opti-
mum instrumental conditions recommended
by the manufacturer.
1.4 The ultrasonication extraction with
HNOa/HCl will extract metals other than
lead from ambient particulate matter.
2. Range, sensitivity, and lower detectable
Hmit. The values given below are typical of
the methods capabilities. Absolute values
will vary for individual situations depending
on the type of instrument used, the lead
line, and operating conditions.
2.1 Range. The typical range of the
method is 0.07 to 7.5 >ig Pb/m3 assuming an
upper linear range of analysis of 15 jig/ml
and an air volume of 2,400 m3.
2.2 Sensitivity. Typical sensitivities for a
1 percent change in absorption (0.0044 ab-
sorbance units) are 0.2 and 0.5 >ig Pb/ml for
the 217.0 and 283.3 nm lines, respectively.
2.3 Lower detectable limit (LDL). A typi-
cal LDL is 0.07 fig Pb/m3. The above value
was calculated by doubling the between-lab-
oratory standard deviation obtained for the
lowest measurable lead concentration in a
collaborative test of the method.US) An air
volume of 2,400 m1 was assumed.
3. Interferences. Two types of interfer-
ences are possible: chemical and light scat-
tering.
3.1 Chemical. Reports on the absence (1,
2, 3, 4, 5) of chemical interferences far
outweigh those reporting their presence, (ff)
therefore, no correction for chemical Inter-
ferences is given here. If the analyst sus-
pects that the sample matrix is causing a
chemical interference, the interference can
be verified and corrected for by carrying out
the analysis with and without the method
of standard additions.* 7)
3.2 Light scattering, Nonatomic absorp-
tion or light scattering, produced by high
concentrations of dissolved solids in the
sample, can produce a significant interfer-
ence, especially at low lead concentrations.
12) The interference is greater at the 217.0
nm line than at the 283.3 nm line. No inter-
ference was observed using the 283.3 nm
line with a similar method.U)
Light scattering interferences can, ho\ ev-
er, be corrected for instrumentally. Since
the dissolved solids can vary depending on
the origin of the sample, the correction may
be necessary, especially when using the
217.0 nm line. Dual beam instruments with
a continuum source give the most accurate
correction. A less accurate correction can be
obtained by using a nonabsorbing lead line
that is near the lead analytical line. Infor-
mation on use of these correction tech-
niques can be obtained from instrument
manufacturers' manuals.
If instrumental correction is not feasible,
the interference can be eliminated by use of
the ammonium pyrrolidinecarbodithioate-
methylisobutyl ketone, chelation-solvent ex-
traction technique of sample preparation.(S)
4. Precision and bias.
4.1 The high-vok.me sampling procedure
used to collect ambient air particulate
matter has a between-laboratory relative
standard deviation of 3.7 percent over the
range 80 to 125 (ig/m3.<9) The combined ex-
traction-analysis procedure has an average
within-laboratory relative standard devi-
ation of 5 to 6 percent over the range 1.5 to
15 jig Pb/ml, and an average between labo-
ratory relative standard deviation of 7 to 9
percent over the same range. These values
include use of either extraction procedure.
4.2 Single laboratory experiments and
collaborative testing indicate that there Is
no significant difference in lead recovery be-
tween the hot and ultrasonic extraction pro-
cedures.(75)
5. Apparatus.
5.1 Sampling.
5.1.1 High-volume sampler. Use and cali-
brate the sampler as described in reference
10.
5.2 Analysis.
5.2.1 Atomic absorption spectrophoto-
meter. Equipped with lead hollow cathode
or electrodeless discharge lamp.
5.2.1.1 Acetylene. The grade recommend-
ed by the instrument manufacturer should
be used. Change cylinder when pressure
drops below 50-100 psig.
5.2.1.2 Air. Filtered to remove particu-
late, oil, and water.
5.2.2 Glassware. Class A borosilicate
glassware should be used throughout the
analysis.
5.2.2.1 Beakers. 30 and 150 ml. graduated,
Pyrex.
5.2.2.2 Volumetric flasks. 100-ml.
5.2.2.3 Pipettes. To deliver 50, 30, 15, 8, 4,
2.1 ml.
5.2.2.4 Cleaning. All glassware should be
scrupulously cleaned. The following proce-
dure is suggested. Wash with laboratory de-
tergent, rinse, soak for 4 hours in 20 percent
(w/w) HNO,. rinse 3 times with distilled-
deionized water, and dry in a dust free
manner.
5.2.3 Hot plate.
5.2.4. Ultrasonication water bath, un-
heated. Commercially available laboratory
ultrasonic cleaning baths of 450 watts or
higher "cleaning power," i.e., actual ultra-
sonic power output to the bath have been
found satisfactory.
5.2.5 Template. To aid in sectioning the
glass-fiber filter. See figure 1 for dimen-
sions.
5.2.6 Pizza cutter. Thin wheel. Thickness
1mm.
5.2.7 Watch glass.
5.2.8 Polyethylene bottles. For storage of
samples. Linear polyethylene gives better
storage stability than other polyethylenes
and is preferred.
5.2.9 Parafilm "M".' American Can Co.,
Marathon Products, Nennah, Wis., or equiv-
alent.
6. Reagents.
6.1 Sampling.
6.1.1 Glass fiber filters. The specifica-
tions given below are intended to aid the
user in obtaining high quality filters with
reproducible properties. These specifica-
tions have been met by EPA contractors.
6.1.1.1 Lead content. The absolute lead
content of filters is not critical, but low
values are, of course, desirable. EPA typical-
ly obtains filters with a lead content of 75
ng/filter.
It is important that the variation in lead
content from filter to filter, within a given
batch, be small.
6.1.1.2 Testing.
6.1.1.2.1 For large batches of filters
(>500 filters) select at random » to 30 fil-
ters from a given batch. For small batches (-
500 filters) a lesser number of fibers may be
taken. Cut one •>«": 8" stiii from each filte
anywhere in the filter. Analyze all strips,
separately, according to the directions in
sections 7 and 8.
6.1.1.2.2 Calculate the total lead in each
filter as
F.= ug Pb/ml x
100 mi
strip
12 strips
~ filter
where:
Fb=Amount of lead per 72 square inches of
filter, ng.
6.1.1.2.3 Calculate the mean, Fb, of the
values and the relative standard deviation
(standard deviation/mean x ICO). If the rel-
ative standard deviation is high enough so
that, in the analysts opinion, subtraction of
Fb, (section 10.3) may result in a significant
error in the >ig Pb/m3- the batch should be
rejected.
6.1.1.2.4 For acceptable batches, use the
value of Fb to correct all lead analyses (sec-
tion 10.3) of particulate matter collected
using that batch of filters. If the analyses
are below the LDL (section 2.3) no correc-
tion is necessary.
6.2 Analysis.
6.2.1 Concentrated (15.6 M) HNO,. ACS
reagent grade HNO, and commercially avail-
able redistilled HNO, has found to have suf-
ficiently low lead concentrations.
6.2.2 Concentrated (11.7 Af) HC1. ACS
reagent grade.
6.2.3 Distilled-deionized water. (D.I.
water).
6.2.4 3 M HNO,. This solution is used in
the hot extraction procedure. To prepare,
add 192 ml of concentrated HNO, to D.I.
water in a 1 / volumetric flask. Shake well,
cool, and dilute to volume with D.I. water.
Caution: Nitric acid fumes are toxic. Pre-
pare in a well ventilated fume hood.
6.2.5 0.45 M HNO, This solution is used
as the matrix for calibration standards
when using the hot extraction procedure.
To prepare, add 29 ml of concentrated
HNO3 to D.I. water in a 1 / volumetric flask.
Shake well, cool, and dilute to volume with
D.I. water.
6.2.6 2.6 M HNO, + 0 to 0.9 M HC1. This
solution is used in the ultrasonic extraction
procedure. The concentration of HC1 can be
varied from 0 to 0.9 M. Directions are given
for preparation of a 2.6 M HNO,+ 0.9 Af HC1
solution. Place 167 ml of concentrated HNO,
into a 1 I volumetric flask and add 77 ml of
concentrated HC1. Stir 4 to 6 hours, dilute
to nearly 1 I with D.I. water, cool to room
temperature, and dilute to 1 I.
•Mention of commercial products does
not imply endorsement by the U.S. Environ-
mental Protection Agency.
A2-23
-------
6.i 7 0.40 U HNO, + X M HC1. This solu-
tion is used as the matrix for calibration
standards when using the ultrasonic extrac-
tion procedure. To prepare, add 26 ml of
concentrated HNO,. plus the ml of HC1 re-
quired, to a 1 I volumetric flask. Dilute to
nearly 1 I with D.I. water, cool to room tem-
perature, and dilute to 1 I. The amount of
HC1 required can be determined from the
following equation:
77 ml x 0.15_x.
where:
y = ml of concentrated HC1 required.
x = molanty of HC1 in 6.2.6.
0 35 = dilution factor in 7.2.2.
6.2.5 Lead nitrate. Pb(NO3),. ACS reagent
grade, purity 99.0 percent. Heat for 4 hours
at 120' C and cool in a desiccator
6.3 Calibration standards.
6.3.1 Master standard, 1000 ^g Pb/ml in
HNO, Dissolve 1.598 g of PWNO,), in 0.45 M
HNO, contained in a 1 / volumetric flask
and dilute to volume with 0.45 J/HNO,.
6.3.2 Master standard, 1000 fig Pb/ml in
HNO,/HC1. Prepare as in 6.3.1 except use
the HNOs/HCl solution in 6.2.7.
Store standards in a polyethylene bottle.
Commercially available certified lead stand-
ard solutions may also be used.
7. Procedure.
7.1 Sampling. Collect samples for 24
hours using the procedure described in ref-
erence 10 with glass-fiber filters meeting the
specifications in 6.1.1. Transport collected
sample? to the laboratory taking care to
minimize contamination and loss of
sample. 116).
7.2 Sample preparation.
7.2.1 Hot extraction procedure.
7.2 1.1 Cut a %" x 8" strip from the ex-
posed filter using a template and a pizza
cutter as described in Figures 1 and 2. Other
cutting procedures may be used.
Lead in ambient particulate matter col-
lected on gla;is fiber filters has been shown
to be uniformly distributed across the
filter.1 1 " Another study " has shown that
when sampling near a roadway, strip posi-
tion contributes significantly to the overall
variability associated with lead analyses.
Therefore, when sampling near a roadway,
additional strips should be analyzed to mini-
mize this variability.
1 2.1.2 Fold the strip in half twice and
place in a 150-ml beaker. Add 15 ml of 3 M
HNO, to cover the sample. The acid should
completely cover the sample. Cover the
beaker with a watch glass.
7.2.1.3 Place beaker on the hot-plate,
contained in a fume hood, and boil gently
for 30 min. Do not let the sample evaporate
to dryness. Caution: Nitric acid fumes are
toxic,
7.2,1.4 Remove beaker from hot plate
and cool to near room temperature.
7.2.1.5 Quantitatively transfer the
sample as follows:
7.2.1.5.1 Rinse watch glass and sides of
beaker with D.I. water.
7.2.1.5.2 Decant extract and rinsings into
a 100-ml volumetric flask.
7.2.1.5.3 Add D.I. water to 40 ml mark on
bea.ts.er, cover with watch glass, and set aside
for a minimum of 30 minutes. This is a criti-
cal step and cannot be omitted since it
allows the HNO, trapped In the filter to dif-
fuse into the rinse water.
7.2.1.5.4 Decant the water from the filter
into the volumetric flask.
7.2.1.5.5 Rinse filter and beaker twice
with D.I. water and add rinsings to volumet-
ric flask until total volume is 80 to 85 ml.
7.2.1.5.6 Stopper flask and shake vigor-
ously. Set aside for approximately 5 minutes
or until foam has dissipated.
7.2.1.5.7 Bring solution to volume with
D.I water Mix thoroughly.
7.2.1.5 3 Allow solution to settle for one
hod! oefore proceeding with analysis.
7.2.1.S.9 If sample is to be stored for sub-
sequent analysis, transfer to a linear poly-
ethylene bottle.
7.2.2 Ultrasonic extraction procedure
7.2.2.1 Cut a %"x8" strip from, the ex-
posed filter as described in section 7.2.1.1.
7.2.2.2 Fold the strip in half twice and
place in a 30 ml beaker. Add 15 ml of the
HNOj/HCl solution in 6.2.6. The acid should
completely cover the sample. Cover the
beaker with parafilm.
The parafilm should be placed over the
beaker such that none of the parafilm is in
contact with water in the ultrasonic bath.
Otherwise, rinsing of the parafilm (section
7.2.2.4.1) may contaminate the sample.
7.2.2.3 Place the beaker in the ultrasoni-
cation bath and operate for 30 minutes.
7.2.2.4 Quantitatively transfer the
sample as follows:
7,a.2.4.1 Rinse parafilm and sides of
beafcer with D.I. water
7.2.2.4.2 Decant extract and rinsings into
a 100 ml volumetric fls.sk.
7.2.2.4.3 Add 20 ml D.I. water to cover
the filter strip, cover with parafilm, and set
aside for a minimum of 30 minutes. This is a
critical step and cannot be omitted. The
sample is then processed as in sections
7.2.1.5.4 through 7.2.1.5.8.
NOTE: Samples prepared by the hot ex-
traction procedure are now in 0.45 M HNO,.
Samples prepared by the ultrasonication
procedure are in 0.40 M HNO, + X M HC1.
8. Analysis.
8.1 Set the wavelength of the monochro-
mator at 283.3 or 217.0 nm. Set or align
other instrumental operating conditions as
recommended by the manufacturer.
8.2 The sample can be analyzed directly
from the volumetric flask, or an appropriate
amount of sample decanted into a sample
analysis tube. In either case, care should be
taken not to disturb the settled solids.
8.3 Aspirate samples, calibration stand-
ards and blanks (section 9.2) into the name
and record the equilibrium absorbance.
8.4 Determine the lead concentration in
jig Pb/ml, from the calibration curve, sec-
tion 9.3.
8.5 Samples that exceed the linear cali-
bration range should be diluted with acid of
the same concentration as the calibration
standards and reanalyzed.
9. Calibration.
9.1 Working standard, 20 ng Pb/ml. Pre-
pared by diluting 2.0 ml of the master
standard (6.3.1 if the hot acid extraction
was used or 6.3.2 if the ultrasonic extraction
procedure was used) to 100 ml with acid of
the same concentration as used in preparing
the master standard.
9.2 Calibration standards. Prepare daily
by diluting the working standard, with the
same acid matrix, as indicated below. Other
lead concentrations may be used.
Volume of 20 ftgl
ml working
standard, ml
0
1 0
20
20
40
80
150
300
500
1000
Final volume, ml
100
200
200
100
100
100
100
100
100
100
Concentration fig
Pb/ml
0
01
02
04
08
1 6
30
60
100
200
9.3 Preparation of calibration curve.
Since the working range of analysis will
vary depending on which lead line is used
and the type of instrument, no one set of
instructions for preparation of a calibration
curve can be given. Select standards (plus
the reagent blank), in the same acid concen-
tration as the samples, to cover the linear
absorption range indicated by the instru-
ment manufacturer. Measure the absor-
bance of the blank and standards as in sec-
tion 8.0 Repeat until good agreement is ob-
tained between replicates. Plot absorbance
(y-axis) versus concentration in fig Pb/ml
(x-axis). Draw (or compute) a straight line
through the linear portion of the curve. Do
not force the calibration curve through
zero. Other calibration procedures may be
used.
To determine stability of the calibration
curve, remeasure—alternately—one of the
following calibration standards for every
10th sample analyzed: Concentration £ l)tg
Pb/ml; concentration & 10 >ig Pb/ml. If
either standard deviates by more than 5 per-
cent from the value predicted by the cali-
bration curve, reca.ibrate and repeat the
previous 10 analyses.
10. Calculation.
10.1 Measured air volume. Calculate the
measured air volume at Standard Tempera-
ture and Pressure as described in Reference
10.
10.2 Lead concentration. Calculate lead
concentration in the air sample.
(ftg Pb/ml X 100 ml/strip X 12 strips/filter)-Fb
c, _
where:
C=Concentration, >ig Pb/sm3.
^g Pb/ml = Lead concentration determined
from section 8.
100 ml/strip=Total sample volume.
12 strips =Total useable filter area, 8" x 9".
Exposed area of one strip, %" x 7".
Filter=Total area of one strip, %" x 8".
Fb=Lead concentration of blank filter, jig,
from section 6.1.1.2.3.
VSTP=Air volume from 10.2.
11. Quality control
%" x 8" glass fiber filter strips containing
80 to 2000 jig Pb/strip (as lead salts) and
blank strips with zero Pb content should be
used to determine if the method—as being
used—has any bias. Quality control charts
should be established to monitor differences
between measured and true values. The fre-
quency of such checks will depend on the
local quality control program.
A2-24
-------
To minimize the possibility of generating
unreliable data, the user should follow prac-
tices established for assuring the quality of
air pollution data, (13) and take part in
EPA's semiannual audit program for lead
analyses.
12. Trouble shooting.
1. During extraction of lead by the hot ex-
traction procedure, it is important to keep
the sample covered so that corrosion prod-
ucts—formed on fume hood surfaces which
may contain lead—are not deposited in the
extract.
2. The sample acid concentration should
minimize corrosion of the nebulizer. Howev-
er, different nebulizers may require lower
acid concentrations. Lower concentrations
can be used provided samples and standards
have the same acid concentration.
3. Ashing of particulate samples has been
found, by EPA and contractor laboratories,
to be unnecessary in lead analyses by atomic
absorption. Therefore, this step was omitted
from the method.
4. Filtration of extracted samples, to
remove particulate matter, was specifically
excluded from sample preparation, because
some analysts have observed losses of lead
due to filtration.
5. If suspended solids should clog the ne-
bulizer during analysis of samples, centri-
fuge the sample to remove the solids.
13. Authority.
(Sees. 109 and 301(a), Clean Air Act as
amended, (42 U.S.C. 7409, 7601(a)).)
14. References.
1. Scott, D. R. et al. "Atomic Absorption
and Optical Emission Analysis of NASN At-
mospheric Particulate Samples for Lead."
Envir. Sci. and Tech., 10, 877-880 (1976).
2. Skogerboe, R. K. et al. "Monitoring for
Lead in the Environment." pp. 57-66, De-
partment, of Chemistry, Colorado State Uni-
versity. Port Collins, Colo. 80523. Submitted
to National Science Foundation for publica-
tions. 1976.
3. Zdrojewski, A. et al. "The Accurate
Measurement of Lead in Airborne Particu-
lates." Inter. J. Environ. Anal. Chew.., 2, 63-
77 (1972).
4. Slavin, W., "Atomic Absorption Spec-
troscopy." Published by Interscience Com-
pany, New York, N.Y. (1968).
5. Kirkbright, G. F., and Sargent, M.,
"Atomic Absorption and Fluorescence Spec-
troscopy." Published by Academic Press,
New York, N.Y. 1974.
6. Burnham, C. D. et al., "Determination
of Lead in Airborne Particulates in Chicago
and Cook County, 111. by Atomic Absorption
Spectroscopy." Envir. Sci. and Tech., 3, 472-
475 (1969).
7. "Proposed Recommended Practices for
Atomic Absorption Spectrometry." ASTM
Book of Standards, part 30, pp. 1596-1608
(July 1973).
8. Koirttyohann, S. R. and Wen, J. W.,
"Critical Study of the APCD-MIBK Extrac-
tion System for Atomic Absorption." Anal
Chem.,45, 1986-1989(1973).
9. Collaborative Study of Reference
Method for the Determination of Suspended
Particulates in the Atmosphere (High
Volume Method). Obtainable from National
Technical Information Service, Department
of Commerce, Port Royal Road, Springfield,
Va. 22151, as PB-205-891.
10. "Reference Method for the Determina-
tion ol Suspended Particulates in the At-
mosphere (High Volume Method)." Code of
Federal Regulations, Title 40, Part 50, Ap-
pendix B, pp. 12-16 (July 1, 1975).
11. Dubois, L., et al., "The Metal Content
of Urban Air." JAPCA, 16, 77-78 (1966).
12. EPA Report No. 600/4-77-034, June
1977, "Los Angeles Catalyst Study Sympo-
sium." Page 223.
13. Quality Assurance Handbook for Air
Pollution Measurement System. Volume 1—
Principles. EPA-60C/9-76-005, March 1976.
14. Thompson, R. J. et al., "Analysis of Se-
lected Elements in Atmospheric Particulate
Matter by Atomic Absorption." Atomic Ab-
sorption Newsletter, 9, No. 3, May-June
1970.
15. To be published. EPA, QAB, EMSL,
RTP, N.C. 27711
16. Quality Assurance Handbook for Air
Pollution Measurement Systems. Volume
II—Ambient Air Specific Methods. EPA-600/
4-77/027a, May 1977.
MAKIUFIlf FOlOtd
FILTEA FNOMSTICKMC TO F1MTIC
—_^ RIGID PLASTIC
Figure 1
A2-25
-------
19mm (%")
STRIPS FOR
OTHER ANALYSES
V» I" STRIP FOR
LEAD ANALYSIS
Figure 2
(Sees. 109, 301(a) of the Clean Air Act, as amended (42 U.S.C. 7409, 7601(a»
[43 PR 46258, Oct. 5, 1978, as amended at 44 FR 37915, June 29, 1979]
A2-26
-------
Appendix 3
Conversion Factors
and Useful Information
International Metric System —Le Systeme International d'Unites (SI Units)
Base Units of the International Metric System (SI)
Quantity
Length
Mass
Time
Temperature
Electric current
Luminous intensity
Amount of substance
Name of the Unit
meter
kilogram
second
Kelvin
ampere
candela
mole
Symbol
m
kg
s
K
A
cd
mol
Recommended decimal multiples and submultiples and the corresponding prefixes
and names.
Pressure
Factor
1012
10"
106
103
102
10
10-'
io-2
10-'
10'6
10'9
10'12
10~15
io-18
Prefix
tera
g'ga
mega
kilo
hecto
dec a
deci
centi
milli
micro
nano
pico
fern to
atto
Symbol
T
G
M
k
h
da
d
c
m
/JL
n
P
f
a
Meaning
One trillion times
One billion times
One million times
One thousand times
One hundred times
Ten times
One tenth of
One hundredth of
One thousandth of
One millionth of
One billionth of
One trillionth of
One quadrillionth of
One quintillionth of
FromS^o
mm Hg
in Hg
InH2O
ftH20
atm
Ib/in*
Kg/cm2
mmHg
1
25.40
1.868
22.42
760
51.71
735.6
in Hg
0.03937
1
0.07355
0.8826
29.92
2.036
28.96
InH.O
0.5353
13.60
1
12
406.8
27.67
393.7
ftH.O
0.04460
1.133
0.08333
1
33.90
2.307
32.81
atm
0.00132
0.03342
0.00246
0.02950
1
0.06805
0.9678
lb/in!
0.01934
0.4912
0.03613
0.4335
14.70
1
14.22
Kg/cm8
0.00136
0.03453
0.00254
0.03048
1.033
0.07031
1
A3-1
-------
Volume
Froin\ro
cm3
liter
m3
in3
ft3
cm'
1
1000
1 x ID'6
16.39
2.83X10'4
liter
0.001
1
1000
0.01639
28.32
m9
1 x ID'6
0.001
1
1.64x10-"
0.02832
in3
0.06102
61.02
6.10X10'4
1
1728
ft3
3.53X10"5
0.03532
35.31
5. 79 XlO-4
1
Temperature
°C = 5/9 (°F-32)
°K=°C + 273.2
°F=9/5 °C + 32
°R=°F + 459.7
Conversion factors —flow
\. Desired
Give>
-------
Conversion factors — ppm vs. /*g/ms
x\ Desired
\xiinit
Given\.
unit \^
/ig/m3
mg/m3
Parts per million by volume — ppm
0,
5.10x10-*
—
NO,
5. 32X10-*
—
SO,
3.83XKT4
—
HZS
7.19XKT4
—
CO
-
0.875
HC
as methane
—
1.53
\. Desired
Given\i«»t
unit ^\
ppm
/tg/m"
0,
1960
NOj
1880
SO*
2610
HjS
1390
mg/m3
CO
1.14
HC
0.654
To convert a value from a given unit to a desired unit, multiply the given value by the factor
opposite the given units and beneath the desired unit.
A3-3
------- An error occurred while trying to OCR this image.
------- An error occurred while trying to OCR this image.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
450/2-80-004
TITLE AND SUBTIJLE
APTI Course 435
Atmospheric Sampling
Student Manual
5. REPORT DATE
September 1980
6. PERFORMING ORGANIZATION CODE
I. RECIPIENT'S ACCESSION-NO.
AUTHOR(S)
M. L. Wilson, D. F. Elias, R. C. Jordan,
_B. M. Ray, K. C. Joerger, o. G. Durham
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
Northrop Services, Inc.
P.O. Box 12313
Research Triangle Park, N.C. 27709
10. PROGRAM ELEMENT NO.
B 18A2C
11. CONTRACT/GRANT NO.
2. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
U.S. Environmental Protection Agency
Manpower and Technical Information Branch
Air Pollution Training Institute
Research Triangle Park, N. C. 27711
Student Manual
14. SPONSORING AGENCY CODE
EPA-OANR-OAQPS
5. SUPPLEMENTARY NOTES
Project Officer for this manual is R. E. Townsend, EPA-ERC, MD-17, RTP, NC 27711
6. ABSTRACT
This manual is used in conjunction with Course #435, "Atmospheric Sampling", as
designed and presented by the EPA Air Pollution Training Institute (APTI). The
manual supplements the course lecture material, presenting detailed discussions
in an introductory manner on the following topics:
Basic Gas Properties and Mathematical Manipulations
Air Measuring Instruments
Particulate Sampling
Gaseous Sampling
Generation of Controlled Test Atmospheres
Standard Methods for Criteria Pollutants
Continuous Air Monitoring Instruments
Design of Surveillance Networks
Atmospheric Sampling Statistical Techniques
This Student Manual is designed to be used in conjunction with the Instructor's
Guide (EPA 450/2-80-006) and the Laboratory Manual (EPA 450/2-80-005) for APTI
Course 435.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Training
Air Pollution
Atmospheric Sampling
Training Course
Student Manual
13B
51
68A
18. DISTRIBUTION STATEMENT
Unlimited. Avail-
able from: National Technical Info. Ser.
5285 Port Royal Rd.
Springfield, VA 22161 —
1-1 (9-73^
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
365
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-
A3-6
------- |