GUIDELINES
Air Quality
Surveillance
Networks
U. S. ENVIRONMENTAL PROTECTION AGENCY
-------
GUIDELINES: AIR QUALITY
SURVEILLANCE NETWORKS
ENVIRONMENTAL PROTECTION AGENCY
Office of Air Programs
Research Triangle Park, North Carolina
May 1971
For sale by the Superintendent of Documents, U.S. Government-Printing Office, Washington, B.C. 20402 - Price 20 cents
-------
The AP series of reports is issued by the Office of Air Pro-
grams, Environmental Protection Agency, to report the
results of scientific and engineering studies, and information
of general interest in the field of air pollution. Information
reported in this series includes coverage of Air Program
intramural activities and of cooperative studies conducted in
conjunction with state and local agencies, research institutes,
and industrial organizations,, Copies of AP reports are avail-
able free of charge to Federal employees, current contractors
and grantees, and nonprofit organizations - as supplies per-
mit - from the Office of Technical Information and Publica-
tions, Office of Air Programs, Environmental Protection
Agency, P. O. Box 12055, Research Triangle Park, North
Carolina 27709. Other requestors may purchase copies from
the Superintendent of Documents, Washington, D. C. 20402.
Office of Air Programs Publication No. AP-98
11
-------
CONTENTS
INTRODUCTION 1
OBJECTIVES OF REGIONAL SURVEILLANCE .... 2
DESIGN OF AN AIR QUALITY SURVEILLANCE
NETWORK 3
Informa.tion Required for Network Design 4
Network Size 4
Station Location 9
Sampling Frequency 12
Sampling Site Characteristics 12
Methodology and Instrumentation 14
LABORATORY OPERATIONS 16
DATA ACQUISITION AND ANALYSIS . 16
iii
-------
GUIDELINES: AIR QUALITY
SURVEILLANCE NETWORKS
INTRODUCTION
Air Quality Control Regions have been designated for the
entire United States. For these regions an inherent part of
the control effort is the development of an air quality surveil-
lance program. In some areas, existing networks will be
modified or expanded; in others, new surveillance programs
must be developed. Although past efforts were concerned
primarily with sulfur dioxide and particulates, regional sur-
veillance programs will need to be expanded to include other
pollutants such as carbon monoxide, nitrogen dioxide, non-
methane hydrocarbons, and oxidants.
The guidelines presented here will assist State and local
agencies in setting up air quality surveillance programs. The
development of an air quality surveillance program includes
determining the number andlocation of sampling sites, select-
ing appropriate instrumentation, and establishing a data infor-
mation system. Experience and technical judgment are essen-
tial for determining the number and location of sampling sites
because adequate mathematical models or other methods have
not been formulated.
The development and implementation of the system must
by necessity involve a trade-off between what is considered
desirable from a strictly technical point of view and what is
feasible with the available resources. An ideal network will
in almost all instances require more resources than are avail-
able. In light of this, the design discussed in this paper
-------
centers on a minimally adequate surveillance network - a net-
work less than ideal, yet capable of meeting the major sur-
veillance requirements. The basic difference between a
minimally adequate surveillance network and the ideal is
that the minimal network has fewer and perhaps less sophis-
ticated instruments. Designers of the network should attempt
to maximize the effectiveness of the minimally adequate net-
work through careful selection of sampling sites, scheduling
of variable sampling frequencies, and possible use of mechan-
ical (integrated) as well as automatic (continuous) samplers.
Because of limited resources, some air quality control
regions may be required to build up to an adequate surveil-
lance network over a period of time. The surveillance net-
work established under conditions of limited resources would
be the starting point upon which a future network could be
built. In such cases a schedule for expansion should be com-
piled early to allow for systematic buildup and to guide the
allocation of resources.
This report deals with four major aspects of regional
surveillance: (1) objectives of surveillance, (Z) design of a
minimally adequate surveillance network, (3) laboratory re-
quirements, and (4) data acquisition and analysis.
OBJECTIVES OF REGIONAL SURVEILLANCE
Regional air quality surveillance networks are inherent
parts of the implementation plans currently required for sul-
fur oxides, particulates, carbon monoxide, hydrocarbons,
oxidants, and nitrogen oxides. Generally, surveillance net-
works for all of these pollutants must be established in a
region. Although each pollutant requires separate analysis,
the collection of samples can be generalized into two groups:
(1) a particulate network, which is the source of information
for suspended particulates, and (2) a gas network, which
consists of sampling devices for CO, SC>2, NO, NO2) non-
methane and total hydrocarbons, and oxidants. The need for
surveillance for each pollutant will depend on the amount of
pollution present within the region. For example, whereas
one region may require extensive surveillance, of, say, oxi-
dants, the relative absence of this pollutant in another region
may preclude such an extensive surveillance effort.
Air quality surveillance within a region must provide
information to be used as a basis for the following actions:
2 'GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
-------
1. To judge compliance with and/or progress made toward
meeting ambient air quality standards.
2. To activate emergency control procedures to prevent
air pollution episodes.
3. To observe pollution trends throughout the region in-
cluding the nonurban areas. (Information on the non-
urban areas is needed to evaluate whether air quality
in the cleaner portions of a region is deteriorating
significantly and to gain knowledge about background
levels.)
4. To provide a data base for application in evaluation of
effects; urban, land use, and transportation planning;
development and evaluation of abatement strategies;
and development and validation of diffusion models.
DESIGN OF AN AIR QUALITY SURVEILLANCE NETWORK
An air quality surveillance program is composed of three
distinct but interrelated elements: sampling networks, labo-
ratory support, and data acquisition and analysis. With auto-
matic (continuous) instrumentation the need for routine labora-
tory support is greatly reduced, but a problem of data trans-
mission, validation, and reduction is introduced. Network
design entails such considerations as the number and type of
stations needed, their locations, frequency of sampling, and
duration of collection period for each sample. The kind of net-
work specified for a given region will also determine the re-
quirements for laboratory sampling and analysis procedures,
laboratory support, and data acquisition and analysis systems.
The two general types of networks required for regional
air quality surveillance are (1) a particulate network and (2)
a gaseous network.
The particulate network should be composed primarily of
high-volume samplers (Hi-Vols) and tape samplers. The Hi-
Vols are used to collect total suspended particulates (TSP),
which may be subsequently fractionated into trace elements and
compounds. Tape samplers provide an indication of suspended
particulate loading over periods of less than 24 hours, pri-
marily for use during air pollution episodes. The total par-
ticulate network design, i. e. , number and location of Hi-Vol
stations, will most likely be determined by the sampling re-
quirements for TSP. The extent to which the Hi-Vol samples;
are analyzed for a particular constituent depends upon local
circumstances. It is important to note that practical field
Design of an Air Quality Surveillance Network 3
-------
techniques are being perfected for measuring the respirable
size fractions of TSP.
The gas network may be composed of a mixture of mech-
anized and automated sampling devices. For some pollutants,
such as sulfur dioxide, nitrogen oxides, and oxidants, both
types of samplers may be used. "Where pollution levels are
substantial, theurbancore network will typicallybe composed
of automatic sampling stations, with the mechanical stations
being relegated to areas of lower concentrations between
widely spaced automated stations.
Information Required for Network Design
Knowledge of the existing pollution levels and patterns
within the region is essential in network design. The areas
of highest pollution levels must be defined, together with geo-
graphical and temporal variations in the ambient levels. Iso-
pleth maps of ambient concentrations derived from past sam-
pling efforts and/or from diffusion modeling are the best tools
for determining the number of stations needed and for suggest-
ing the station locations. Additionally, information on mete-
orology, topography, population distribution, present and
projected land uses, and pollution sources is extremely help-
ful in network design. In fact, where isopleth maps are not
available, such information, which can usually be obtained
readily from organizations such as the Bureau of Census,
National Weather Service, and local planning agencies, pro-
vides the basis for initial design.
In the absence of isopleth maps, information on emission
densities and/or land use can be used together with wind-rose
data to pinpoint areas of expected higher concentrations. To-
pographical maps provide additional information on wind flow
and pollution dispersion characteristics. Maps of population
distribution are essential in locating key stations for moni-
toring during episodes.
In some cases adequate information for network design
will not be available; the resulting networks will need to be
modified as more information and experience are obtained.
Network Size
The number of sampling stations required depends pri-
marily on the existing pollution levels, their variability, and
the size of the region. The number of sampling stations must
GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
-------
be adequate to allow definition of the area or areas where am-
bient concentrations may be expected to exceed air quality
standards. Information on air quality in other areas, in-
cluding the nonurban portions of the region, should also be
gathered.
A first approximation of the number of stations required
in a region may be obtained from Figure 1, in which the num-
ber of stations is shown as a function of total population. The
curves in Figure 1 show a spread suggesting a minimum and
a maximum number of stations for each population class de-
pending on the extent and degree of pollution. For example,
a region of 1 million inhabitants -with a severe SO2 problem
may require up to 25 samplers, whereas one of similar size
with a minimal SC>2 problem would require only 10 samplers.
Although the curves should provide good estimates for appli-
cation to population- or motor-vehicle-related pollutants such
as CO, HC, NOX, and oxidant (ozone), they do not necessarily
apply as well to SC>2 and particulate matter. For the latter
pollutants, industrial complexity and fuel-use patterns in the
region strongly influence the pollution levels and thus affect
network size regardless of the population.
Surveillance of SO2 and NQx requires an additional de-
cision concerning the mixture of mechanical and automatic
samplers. A first approximation can again be obtained from
Figure 1. The curve for mechanical samplers provides the
estimate for the total number of SC>2 samplers needed. The
difference between the estimates for mechanical and auto-
matic samplers is the actual number of mechanical samplers
required.
Figure 1 is intended only as a general guide to network
size. The curves are based on a qualitative evaluation of
cities of different population classes in terms of their existing
networks, pollution patterns, and geographic distribution of
sources. The relationship between population and network
size was derived from such investigations combined with ex-
perience and knowledge. In general, population is a good
index to network size and such data are easily obtainable.
There are, however, certain situations, such as the relative
absence of SO? pollution in the western portion of the United
States, in which these curves are not applicable. In these
cases, more-specific information on sources and emissions
is essential before network size can be determined.
Design of an Air Surveillance Network
-------
J
o
6
Q_
a.
u,uuu,uuu
5,000,000
2,000,000
1,000,000
500,000
200,000
100,000
50,000
20,000
innrm
= 1 " /I X 1
= J^'/^ <
'/ / ' /x
- // ' / /
ill if /
-All i ! /
-ft'' /^ '"'
• 1 * * •
-ti / / /
n ; i i
-n ' '
"1 ; / /
1 / . •
II // /
iJ /•' 7
H •'• /
B/ -7'
14; '
/ y ...
J/
i • _
I 1 1 1
T 1 /' \ \ L' — •
¥ ^ ~
/' —
/
/
—
Z
^~
—
—
—
MFPHAKIirAI -INTFRRATINR —
•^ inL^nnii iunL iiii^unniinu
(TSP,S02) —
AUTOMATIC- CONTINUOUS
(S02, CO, HC, NOX, OXIDANTS)
1 1 1 1 1
0 5 10 15 20 25 30 35 40 45 50
NUMBER OF STATIONS
Figure 1. Region population versus number of stations.
GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
-------
Another approach for approximating the number of stations
required within a region incorporates information on existing
levels of pollution as a function of the area of the region.
Where isopleth maps are available, they should be used for
estimating network size. At the present time, use of the
mathematical approach is limited to the design of surveil-
lance networks for suspended particulates and sulfur dioxide.
The equation for estimating network size relates the
number of stations to the degree of pollution and the land area
of the region. It is based on the fact that more stations are
needed in zones where ambient air pollutant concentrations
may be expected to exceed the ambient air quality standards
and that the concentrations influence the number of stations.
The equation considers distinct areas: the area, X, where
the pollution levels are higher than the ambient air quality
standard; the area, Y, where pollution levels are above back-
ground but lower than the standard; and the area, Z, where
existing concentrations are at background levels. In these
calculations all air quality data are expressed in terms of
annual averages. The total number of samplers, N, required
for the entire region is obtained by summing the estimated
numbers of samplers for each of the three subareas.
N = Nx + Ny + Nz (1)
The subareas are described as follows:
Nx = 0.0965 X (la)
Cs - Cb
Ny = 0.0096 Y (Ib)
cs
Nz = 0.0004 Z (lc)
Where:
Cm = Value of maximum isopleth* (with a contour interval
of 10), M-g/m3
*Table 1 provides estimated background values for total sus-
pended particulates and SOz for use when isopleth information
is not available.
Design of an Air Quality Surveillance Network
-------
Table 1. TSP AND SOz VALUES FOR NONURBAN
BACKGROUND TERM IN EQUATION 1
Total suspended
particulate
Sulfur dioxide
Proximatea
45
20
Intermediate
40
10
R emote c
20
5
Proximate values based upon NASN stations in the following
states: Connecticut, Delaware, District of Columbia, Mary-
land, Massachusetts, New Jersey, New York, Pennsylvania,
Rhode Island.
Intermediate values for all other states.
"Remote values based upon NASN stations in the following
states: Colorado, Idaho, Michigan, Minnesota, Montana,
Nebraska, Nevada, New Hampshire, New Mexico, North
Dakota, Utah, Wisconsin, Wyoming.
Cs = Ambient air quality standard, |j,g/m3
C"b = Value of the minimum isopleth (again with a con-
tour interval of 10), fig/m.3
X = Area wherein concentrations are higher than ambi-
ent air quality standard, km^
Y = Area wherein concentrations are above background
but less than ambient standard, km^
Z = Area wherein concentrations are at background
levels, km^
Use of these equations requires the division of the region
into three zones onthebasis of isointensity lines representing
the ambient air quality standard and the background value ap-
propriate for the region. The land areas of each zone are
determined from the isopleth map, as are the maximum and
minimum concentrations in the region. Note once more that
concentration values used in the equation are annual averages.
The above equations should not be used to estimate the numbers
of background stations where regions encompass excessively
large unpopulated areas. No more than four or five background
stations should be necessary in any region.
GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
-------
The equation for determining number of stations was
derived from an in-depth analysis of the relationship between
pollution levels and patterns, geometric distribution of
sources, meteorology, and land area in the National Capital
Interstate Air Quality Control Region. The equation was
verified by application to several other cities in the United
States with various population and pollutant distributions. As
mentioned earlier, it is applicable only to SO2 and TSP net -
works. The same general approach, with different constants,
can be applied to determine the size of networks for the
measurement of other pollutants. These constants will be
developed when more data become available.
Station Location
Selecting the locations of stations and samplers involves
decisions regarding (1) distribution of samplers within the
region and (2) specific site selection for each station. The
first decision requires consideration of surveillance objec-
tives, overall pollution patterns, and the needs for govern-
mental jurisdictional coverage. Selection of the particular
site is based upon representativeness of the area and other
practical aspects such as housing the samplers, electric
power, and security from vandalism.
The information required for selecting sampler location
is essentially the same as that for determining network size,
i.e., isopleth maps, population density maps, source loca-
tions. Following are suggested guidelines:
1. The priority area is the zone of highest pollutant con-
centration within the region. One or more stations are
to be located in this area.
2. Close attention should be given to densely populated
areas within the region, especially when they are in
the vicinity of heavy pollution.
3. For assessing the quality of air entering the region,
stations must also be situated on the periphery of the
region. Meteorological factors such as frequencies of
wind direction are of primary importance in locating
these stations.
4. For determining the effects of future development on
the environment, sampling should be undertaken in
areas of projected growth.
Design of an Air Quality Surveillance Network
-------
5. A major objective of surveillance is evaluation of the
progress made in attaining the desired air quality. For
this purpose, sampling stations should be strategically
situated to facilitate evaluation of the implemented
control tactics.
6. The National Air Surveillance Network (NASN) sampling
sites within a region should be included as an integral
part of the proposed network. It is expected the appro-
priate agency will take over and continue to operate
this station(s) at the existing site to provide continuity
of air quality at a given location.
7. Some information of air quality should be available to
represent all portions of the region.
The air quality surveillance network should consist of
stations that are situated primarily to document the highest
pollution levels in the region, to measure population exposure,
to measure the pollution generated by specific classes of
sources, and to record the nonurbanlevels of pollution. Many
stations will be capable of meeting more than one of these
criteria; e.g., a station located in a densely populated area
besides measuring population exposure could also document
the changes in pollutant concentrations resulting from new
control strategy employed in the area.
Although the sampler locations depend on many factors,
some idea of sampler distribution may be obtained from Tables
2 and 3, which show sampler location as a function of network
size. Table 2 summarizes distribution of mechanical samplers,
such as Hi-Vols; Table 3 shows distribution of automatic
samplers. With respect to locations shown in Tables 2 and 3
it will be necessary to consider wind patterns, source loca-
tions, and distribution of emissions in selecting approximate
locations for these sites. For example, stations in the highly
populated area should be so situated that they can accurately
assess the pollution impact under different meteorological
conditions. Although both types of stations follow the same
general pattern, the tendency is for wider distribution of
mechanical sampling stations.
It is the intent of these guidelines to suggest that a simple
network be developed to measure the concentration of as
many pollutants as possible. It is likely that common sites,
although not necessarily ideal for each pollutant, maybe
10 GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
-------
Table 2. DISTRIBUTION OF MECHANICAL, (INTEGRATED)
SAMPLING STATIONSa
Total number
of stations
1
2
3
4
5
10
15
20
25
30
Number of stations in:
Center city/
industrial
1
1
2
2
2
5
8
12
14
17
Residential
zones
-
1
1
2
2
3
5
6
8
10
Nonurban
-
-
-
-
1
2
2
2
3
3
Includes Hi-Vol sampler and
collecting devices. ,
NO2, and oxidant (ozone)
Table 3. DISTRIBUTION OF AUTOMATIC. (CONTINUOUS)
SAMPLING STATIONS3"
Number of stations in;
Total number
of stations
1
2
3
4
5
6
10
15
Center city/
industrial
1
1
2
2-3
3
4
6
10
Residential
z one s
-
1
1
1-2
2
2
4
5
Nonurbanb
-
_
_
_
_
_
_
-
Includes SO2, CO, HC, NO, NO2, and oxidant (ozone).
Where ozone damage has been identified in nonurban areas,
surveillance may be necessary.
selected to provide adequate coverage for the pollutants of
concern. Each pollutant, however, should be considered
individually during the design phase to pinpoint pockets of high
pollution that otherwise might be overlooked.
Design of an Air Quality Surveillance Network
11
-------
The final task in determining sampler placement is to find
a specific location with the proper facilities for operating the
sampler. Availability of space and power, accessibility,
security, and representativeness of the site determine the
precise location.
Sampling Frequency
The sampling frequencies for mechanical samplers and
the averaging times for automatic samplers are dictated by
the ambient air quality standards. If, for instance, standards
are set in terms of days, hours, or minutes, then the sampling
frequencies must be in the same averaging time.
Although standards for TSP and SO2 are prescribed in
terms of annual averages and maximum daily concentrations,
it is impractical to operate the entire network on a daily basis.
Adequate coverage maybe maintained with intermittent sampl-
ing at frequencies calculated statistically for desired levels
of precision. Suggested sampling frequencies are presented
in Table 4, which relates frequency of sampling to the degree
of pollution, ranging from every third day sampling in the
highly polluted zones to once every sixth day in the nonurban
zones. Twenty-four-hour sampling should be from midnight
to midnight to represent calendar days and to permit direct
utilization of the sampling data with standard daily meteoro-
logical summaries.
Sampling Site Characteristics
The preceding sections gave guidelines for the general
distribution of sampling stations within a region. The selec-
tion of a particular site for a single sampler or a complex
station is equally important. It is essential that the sampler
be situated to yield data representative of the location without
undue influence from the immediate surroundings. No defin-
itive information is available concerning how air quality
measurements are affected by the nearness of buildings ,
height from ground, and the like. There are, however, gen-
eral guidelines that should be considered in site selection:
1. Uniformity in height above ground level is desirable
for the entire network within the region. Some excep-
tions may include canyons, high-rise apartments, and
sites for special-purpose samplers.
2. Constraints to airflow from any direction should be
avoided by placing inlet probes at least 3 meters from
12 GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
-------
-------
buildings or other obstructions. Inlet probes should be
placed to avoid influence of convection currents.
3. The surrounding area should be free from stacks,
chimneys, or other local emission points.
4. An elevation of 3 to 6 meters is suggested as the most
suitable for representative sampling, especially in
residential areas. Placement above 3 meters prevents
most reintrainment of particulates, as well as the direct
influence of automobile exhaust.
Methodology and Instrumentation
Information on types of instrumentation, its use, specii-
ficity, and associated costs is summarized in Table 5.
Development of instruments and techniques for sampling
and analysis is progressing rapidly. Numerous instruments
and techniques are now available for sampling and analysis.
Standard methods are under development by the Air Pollution
Control Office and the Intersociety Committee. * Until such
standard methods are prescribed, the preferred methods of
sampling and analyses are those for which a large body of data
is available. Recommended reference methods are described
in the Federal Register (Ap*il 30, 1971) with the National
Ambient Air Quality Standards for SO2, CO, NOX, oxidant,
hydrocarbons, and particulate matter. Acceptable methods
for sampling and analysis are those that have been compared
with standard or reference methods and have proved compar-
able in collection efficiency and in specificity. Whatever
method is chosen, caution should be exercised to purchase
instruments that have been thoroughly field tested.
Continuous production of valid data requires that instru-
ments be properly maintained. This includes calibration
prior to installation and on a routine basis thereafter. Most
modern continuous instruments provide for automatic d:pa.amiic
*Member societies are: Air Pollution Control Association
(APCA), American Council of Governmental Industrial
Hygienists (ACGIH), American Industrial Hygiene Associa-
tion (AIHA), American Public Health Association (APHA),
American Society of Mechanical Engineers (ASME), Ameri-
can Society for Testing and Materials (ASTM), and Associa-
tion of Official Analytical Chemists (AOAC).
14 GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
-------
Table 5. CLASSIFICATION OF AIR POLLUTION SAMPLING TECHNIQUES
I
I
CD
Z
-------
calibration at least daily. Dynamic calibration either by per-
meation devices or gaseous mixtures is preferred. Static
calibration is subject to stoichiometric factors and does not
take into account collection efficiency of the continuous
sampling instrument.
LABORATORY OPERATIONS
Support of the surveillance networks will require labora-
tory operations of varying levels of complexity. The require-
ments for laboratory support, in terms of size and complexity,
will be dictated by the pollutants of interest in the region, size
of the networks, and the degree of pollution. The laboratory
should be equipped for analyses of samples for at least TSP ,
SOz, NO2, and oxidant and should provide for calibration of
all collecting and measuring devices and preparation of
reagents.
Some regions will require laboratory capability for
analyses of trace elements, fluorides, and other pollutants.
Inlarge laboratories with requirements for a large number of
analyses, automated laboratory procedures should be con-
sidered.
DATA ACQUISITION AND ANALYSIS
The design of a network must be accompanied by design
of a system of data acquisition and analysis that considers the
flow of data and their use.
To insure uniformity of data across the country and also
to assist State and local agencies in data handling and analyses,
the Environmental Protection Agency is expanding the National
Aerometric Data Information Service (NADIS). This service
encompasses the operation of a National Aerometric Data
Bank, the systematic gathering of all aerometric data, and
the provisionf or data dissemination in the form of summaries
and special analyses. State and local agencies will routinely
submit all air quality data collected to the National Aero-
metric Data Bank quarterly. All data should be expressed in
the SAROAD format. A SAROAD Users Manual gives detailed
instructions for coding data in the SAROAD format.
16 GUIDELINES: AIR QUALITY SURVEILLANCE NETWORKS
------- |