-------
2.1.2 Factors Influencing Network Design
The factors that are typically involved in estimating an adequate network
size are of course also the factors involved in designing an ultimate
network configuration as well, primarily climatological and topographic
factors. These factors are typically cited as meaningful in network
design, but it is frequently difficult to make practical use of them.
This is because they are significant primarily in the extremes, as noted
below, rather than in the broad middle range prevalent throughout most of
the country.
2.1.2.1 Meteorology and Climatology - The meteorological factors that
have the greatest effects on ambient pollution concentrations are the
horizontal wind (speed and direction, and the vertical distribution of
both) and the vertical mixing structure (stability, mixing heights). At
most locations, however, these parameters vary significantly over time
scales in hours and distance scales in tens of meters. Thus, while they
are of significance in a number of air pollution areas, they are not of
much help in the design of networks, which depends on longer-term average
parameters.
Dilution climatology is defined as the long-term average combination of
those meteorotogical conditions that affect the interchange and disper-
sion of pollutants over relatively large areas and long time intervals.
These factors, the frequency, persistence, and height variations of wind
speed and direction, of stable (inversion) layers of air, and of mixing
heights, collectively provide a measure of the dilution climatology of
an area. Dilution climatology accounts for the effects of large scale
topographic features, such as large bodies of water and mountain ranges,
that exert their influence at that scale. The relative frequency of
recurrence of short-term phenomena such as stagnation episodes is also
considered. Small scale obstructions such as hills and buildings are
classified as localized influences and are not considered in dilution
climatology. Atmospheric areas possessing similar dilution climatologies
II-5
-------
have been defined on a geographic basis for the contiguous United States.
They are illustrated in Figure II-l and described in Table II-2; interim
definitions for areas outside the contiguous United States in which
AQCR's have been designated are also included in Table II-2. Figure II-2
presents isopleths of mean annual solar radiation which, in conjunction
with the dispersion characteristics of the various atmospheric areas,
relates to the potential for formation of photochemical pollutants.
As was noted above, these climatological factors are of primary signifi-
cance in the extremes. The Great Plains Area has frequent high winds
which, coupled with the nature of the fuel use patterns, reduces concern
with S09; however, because of increased fugitive dust entrainment, par-
ticulate problems require increased concern. Considering north-to-south
variations in solar radiation, it is apparent that the Southwest and the
Gulf Coast will have an accordingly greater concern with photochemical
oxidant levels.
2.1.2.2 Topography - The dispersion patterns in some sectors of an Air
Quality Control Region can be significantly altered by local topographi-
cal factors. The most significant with respect to their influence on a
monitoring network are:
Valley Effects - Valleys tend to channel the wind flow
along their axis, restrict horizontal dispersion, in-
crease the tendency for inversions to form, and may
cause aerodynamic downwash from stacks not extending
above the valley walls. "Air quality discontinuities
between valley-ridge sectors often exist. Thus, val-
leys almost always need monitors in ;excess of the
requirement for level terrain.
II-6
-------
Cfl
0)
tfl
4-1
CO
H
C
W
3
O
60
H
4J
C
O
M-l
O
0)
V-i
a.
en
O
B
4J
60
II-7
-------
Table II-2. ATMOSPHERIC AREAS OF THE UNITED STATES
Atmospheric area
Extent of area
Meteorological And topographical
characteristtca
California-Oregon coastal area
Washington coastal area
tocky Mountain area
Great Plains arta
Great Lakes Northeast area
Appalachian area
Extends about 20 to SO nlles Inland fro
the Pacific Ocean.
Extend* about 20 to 30 nlles Inland from
the Puget Sound region, frora which
the eastern boundary extends south-
westward to the vicinity of Lon^view
on the Columbia River and then west-
ward to the coast.
Extends eastward froo the California
Oregon and Washington coastal areas to
terminate as a north-south oriented
eastern boundary, essentially core-
pond ing to the 3.000 to -.,000 toot oean
sea level contour interval which in gen-
eral defines the eastern oost extension
of the major mountain ranges. This
eastern boundary stretches troa the
Canadian border in Montana south-
vard through extreoe eastern Colorado,
eastern N
-------
Table II-2 (continued). ATMOSPHERIC AREAS OF THE UNITED STATES
Atmospheric ar«a
Extent of area
Meteorological and topographical
characteristic*
ttd-AtlantU araa
South rlorlds-Carlbbesn *raa
Hawaiian-Pacific ara«
Alaska* feclCia Knltl** av*a
Alaskan Baring Maritime araa
Alaskan Arctic Maritlma araa
Alaskan Continental araa
Encompasses the Atlantic coastal plain
from extreme southwestern Connecticut,
Including the New York City and Long
Island region, southward to the South
Carolina border at the coastline, and
extends Inland to the Appalachian araa.
Extends south from the DaytQns Beach
Cedar Key line to Include the aoothern
half of Florida. Puerto Rico, and ths.
Virgin Islands.
Includes sll of the islands making up
the State of Hawaii, and the territories
of Guam snd American Samoa.
Bound*4 by tVi* Unltsd Statst-Csnsd* bor-
der to th« southeast) the Chugateh Moun-
tain Range to the north, and the Aleutian
Range to the northwest. As such this araa
Includes the Alexander Archipelago, the
coastal regions of the Gulf of Alaska,
Kodlsk Island, the Alaskan Peninsula, and
the Aleutian Islands.
Bounded by the southwestern and western
slopes of mountain'ranges and the ridge
line of the Sevard Peninsula. As such,
the area Includes the coastal plateaus and
valleys of the southwest and western main-
land, the southern half of the Seward
Peninsula, and offshore island*.
Bounded by the western slopes of mountain*
from the Scward Peninsula northward to tha
Brooks Mountain Range then eastward to
United States-Canadian border. As such,
this area includes the northern half of tha
Seward Peninsula, the coastal regions to
tha north, and the tundra region between
the Brooks Range and the Arctic Ocean.
Bounded by tha inland portion of tha
Alaska-Canadian border to the east and
tha previously described Atmospheric Area
boundaries to the north, south, and west.
Shallow mixing depths, less frequent low-
level stability and higher wind speeds
are features of the dilution climate that
distinguish this cosatsl art* fron thosa
adjacent.
The climate of this ares Is predominantly
tropical-marine In nature* Atmospheric
stagnation Is practically nonexistent;
there 1* a small frequency of low-level
stability; snd relatively good' vertlcsl
mixing prevails.
Relatively good ventilation; occasional
surface-based nocturnal Inversion* In In-
land areas; persistent period* of stag-
nation are rare.
Undor th* influence o! Pacific Matitlma
weather patterns; relatively good ventila-
tion associated with frequent storms; oc*
caatonal strong nocturnal inversions may
persist throughout the daytime during the
winter season; persistence of such condl*
tlons Is not marked, however, because of
the frequency of atormlness.
Under the Influence of Bering Maritime
weather conditions. Air Pollution cli-
matology varies from that of the Pacific
Maritime area because of less frequent
storm activity and the resultant poten-
tial of greater persistence of surface-
based inversions. In spite of difference*,
persistent stagnations are not frequent.
Under the influence of two, seasonally-
oriented weather conditions; continental
during tha winter months when the ocean,
la frozen; maritime during the warmer
months when the ocean is partially free
of Ice. Relatively high wind speeds pro-
vide good ventilation; the lack of solar
radiation in the winter snd cold maritime
winds during summer days result in the
highest annual frequency of daytime
surface-based Inversions of any of the
areas discussed here.
Under the influence of continental
weather conditions; sheltered from mari-
time Influence by medium-to-high mountain
ranges on all sides; has the highest an-
nual frequency of nighttime, surface-
based inversions of any of the adjacent
areas; low wind speed during the winter,
combined with extremely persistent ground-
level inversions, gives this araa tha most
restrictive pollution climatology of any
AoMspharlc Araa*
II-9
-------
c
60
C
G
O
H
H
13
CO
.H
O
c
rt
(U
CM
I
11-10
-------
Shoreline Effects - Airflow along shorelines undergoes
frequent changes brought about by the changes in rela-
tive temperature of the air and water. Discontinuities
and convergence zones in the dispersion patterns occur
which indicate need for monitoring beyond required
minimums.
Hilly and Mountainous Terrain Effects - Complexities
introduced by hills and mountains include disrupted
airflow patterns, intersection of their interface by
elevated plumes, induced mechanical turbulence and more
frequent inversions in low-lying protected areas.
Hilly and mountainous terrain usually increase the need
for monitors.
In general, these concerns are greatest in the case of SC^ and particu-
lates which are often dispersed from major point sources. They are of
lesser importance for automotive pollutants such as CO, or secondary
pollutants like CL and N0?.
2.1.3 General Patterns of Basic Networks
The overall configuration of a basic fixed network is primarily a function
of the purpose of the monitoring and the typical spatial distribution of
the pollutant under consideration. It is important to initially design a
separate network for each pollutant under consideration, and only then to
consider whether and to what extent the networks may be combined, with
sensors at common sites. The following sections consider each pollutant
in turn, discussing the configuration of networks as they typically exist
and suggesting changes as appropriate.
11-11
-------
2.1.3.1 Sulfur Dioxide - The general configuration of a typical sulfur
dioxide network is one of roughly uniform distribution over the built-up
or populated portion of a Region, usually with a decreasing density in
the areas farther from the urban center. One or more of the sites is
usually in the area of anticipated maximum levels, to monitor for the
attainment and maintenance of NAAQS, while the others serve to monitor
the exposure in neighborhoods (residential, downtown, commercial, etc.).
The primary goals of SO,, monitoring are all relatively well-served by
such a population-oriented network with a typical site-to-site distance
of at least 2 to 4 kilometers. Typically regional S02 networks consist
of a mixture of continuous instrumentation and bubbler sites, and this
is considered appropriate; an acceptable distribution between the two
types is presented in Table II-3. The use of continuous instrumentation
at more sites than indicated in Table II-3 at all sites, is acceptable
(if somewhat expensive); the use of less continuous and more bubbler sites
is not recommended.
Table II-3. DISTRIBUTION OF CONTINUOUS AND
BUBBLER S02 INSTRUMENTATION
Number of S02 Sensors
Total
1
2
3
4
5
6
10
15
20
25
30
35
40
Continuous
0
0
0
1
1
2
3
5
7
10
13
15
17
Bubbler
1
2
3
3
4
4
7
10
13
15
17
20
23
11-12
-------
This overall assessment of adequacy of SCL networks is based on the
use of emission inventories to develop S02 emission density patterns
with special consideration given to any major industrial process S02 .
source that might cause significant deviations from a relatively smooth
geographic distribution. In those few Regions with no significant S02
emissions, relatively less dense networks are adequate.
It there are significant industrial sources, or concentrations of smaller
sources, the network should include additional sites to monitor exposures
in any adjacent residential areas. In this context, significant S02
source is intended to refer to such as refineries, smelters, etc., that
have numerous emission points. Major fuel-burning sources, such as power
plants, which have only a very few elevated emission points, should be
considered in the context of the discussions in Supplement B. A third
situation requiring a significant deviation in the density of the network
is that of unusual topography. Major topographic features, such as hills
and valleys, that destroy the smooth uniformity of air quality patterns,
require additional monitoring to define the discontinuity.
In general, current SCL state monitoring efforts are typically adequate
in comparison to the monitoring efforts directed at other pollutants. The
primary need in the near future will be for some reallocation of monitors
in the form of increased density in designated Air Quality Maintenance
Areas for S02 and around major isolated point sources, from the urban
core area or the CBD.
2.1.3.2 Suspended Particulate Matter - The general pattern of particu-
late networks is usually similar to that for S0«, in many cases consist-
ing of the same sites. This is reasonable, since the two pollutants both
have a widespread multitude of small sources, frequently the same sources.
There are, however, differences in the nature of the two pollutants that
may lead to some differences in the network configuration. Since entrain-
ment from the ground and other "fugitive dust" sources can be important
11-13
-------
for participates, the issues of actual siting become of greater importance
than with S0?. In the past, a suspended particulate network that was
largely coincident with the S0~ configuration was generally considered
adequate. However, as the traditionally important particulate sources
(industrial processes and fuel combustion, small coal-fired boilers) have
been eliminated or controlled, other types of sources (re-entrained urban
dust, rural fugitive emissions) have become of jajor concern. Hence, a
reallocation of monitors to neighborhood and rural sites, as opposed to indus-
trial peak sites, will be needed to understand these new problems and
develop appropriate control tactics.
2.1.3.3 Carbon Monoxide - In contrast to the case with S02 and particu-
lates, the general configuration of a typical CO monitoring network is
neither well-defined nor adequate. In most cases, CO monitoring is con-
ducted at only three or four sites in an urban AQCR. Because the measured
CO levels are very sensitive to the exact placement of the inlet probe,
the possibility of biased information resulting from this scarcity of
sites is greatly increased.
Designing a CO monitoring network is, thus, quite complicated in com-
parison to other pollutants. This is because of the nature of the NAAQS
for CO, and the differing circumstances in which they are typically vio-
lated. As there is no long-term (annual or seasonal) standard for CO,
the objective of determining trends and patterns is of a good bit less
importance, and the objective of monitoring attainment and maintenance of
NAAQS is more complicated. The issue is further complicated by the dif-
fering circumstances under which the 1-hour and 8-hour standard are typi-
cally violated, which is determined by the interaction of the strong
daily cycle in CO source strength with the seasonal and daily cycles of
atmospheric mixing potential. The 1-hour NAAQS for CO is typically vio-
lated under circumstances of maximum traffic during the morning rush hour,
often on mornings when a nocturnal radiation inversion has persisted
until the time of the rush hours. Because they depend on having heavy
11-14
-------
traffic for a short period, the peak 1-hour levels typically occur near
points of major traffic volumes. In contrast, the highest 8-hour CO levels
tend to occur in the evening and overnight, and may well occur quite apart
from short-term traffic peaks. This is due primarily to differential cooling,
down slope drainage and a general reduction in mixing height, commonly occuring
in the evenings and early morning.
It is recommended that the overall CO network configuration should involve
sites of four types which are discussed in detail and prioritized in Supplement f.
These types are:
Street Canyon
- Peak
- Average
Neighborhood
- Peak
- Average
Corridor
Background
As discussed in Supplement A, there is generally little likelihood of
totally defining an area's CO air quality patterns with a monitoring net-
work, because the variation in CO levels is so dramatic over such short
distances that the number of monitoring sites required would be totally
prohibitive. Rather, it is considered appropriate to monitor a few care-
fully selected neighborhood and street canyon sites. These should be
selected to typify population exposures under a variety of conditions, so
that one can develop from these a relationship adequate to project the
impact in other similar areas.
2.1.3.4 Photochemical Oxidants/Ozone - The typical configuration used
for oxidant or ozone monitoring has been too often only one site in the
urban center of an AQCR, and frequently the precursor pollutants are mon-
itored at the same site. Because oxidant, as a secondary pollutant, is
not closely related to any geographic source pattern, oxidant levels have
been presumed to be relatively uniform over large areas, and one downtown
11-15
-------
sampling site was not considered too grossly inadequate. This is not neces-
necessarily the wisest practice, however, due to the scavenging effect of
freshly generated NO from mobile sources. Figure II-3 presents the typical
diurnal afctern experienced at such a combined site. The maximum oxidant
levels are not coincident in time with the peak levels of the precursors
and hence are not likely to be coincident in space either. This leads to
the recommendation that peak sites be located 15 to 25 km from the center
of the city in at least two general directions. These two general areas
should be selected based on wind directions during the ozone season. This
season varies, according to local climatology, from May to September in the
North to April to October in the South. Generally, ozone levels above the
NAAQS are not found when daytime ambient temperatures are below 15°C (60°F).
Consideration may well be given to reduced operation of isolated 03 monitors
during the winter months.
In addition to these peak sites, several neighborhood sites may be necessary
for monitoring population exposures in residential, commercial, and downtown
areas, depending on the population and size of the Region. For purposes of
determining possible transport of ozone into the region, it may be necessary
to have sites in remote areas upwind.
2.1.3.5 Nitrogen Dioxide - Nitrogen dioxide has a dual role in air pollu-
tion, so that two different sets of network needs must be considered. There
is an NAAQS for N02, so that peak and neighborhood population exposure must
be monitored. Because of the lag time indicated in Figure II-3, the peak
N02 exposure will not necessarily be at major traffic points of high NO
emissions. However, the timing of the peak can vary significantly through
the year (Figure II-4), so that it does not provide a very rigorous
guide for placing sites. In general, in areas where levels exceed the
standards, a population-oriented network involving both bubbler and con-
tinuous monitoring should be done in peak areas, while the intermittent
monitoring should be at neighborhood and background sites. The peak sites
should be located similar to the peak ozone sites, except that they should
be only 10 to 15 km from the center city.
11-16
-------
E
CX
8
CD
1
cvi o
cc
o
E
6
O
O
(6
60
g
03
Cfl
O
r-4
S
0)
J2
O
O
D- TO
T)
M-l
O 0)
00
O bO
-« C!
O -i-l
§ S
0 §
i-H -H
CQ 4-1
>> o
H 4-1
01
I
60
r-l
C\J
IO
ci
o
o
IO
q
o
11-17
-------
0.10
0.05 -
0.10
0.05 >
0.101'r i iiriiiriir-
0.05 -
E
a
a
Z
O
- 0.05 -
-------
2.1.3.6 Nonmethane Hydrocarbons and Nitric Oxide - The existing monitoring
of nonmethane hydrocarbon and nitric oxide is typically a very limited
effort with one or a few continuous sites in an area. Since there are no
NAAQS, population-oriented monitoring is not necessary and in most areas
is not conducted.
However, data on both NO and nonmethane hydrocarbons are required along
with N0» and oxidant data to provide research and planning information with
respect to photochemical oxidant reduction. Single sites at the urban
center are clearly not adequate for this purpose, as they do not permit any
resolution of spatial distribution and transport-reaction time questions.
It is recommended (although not required) that hydrocarbon and paired NO
X
sensors be located in the CBD of the urban core area when reliable instru-
ments for measuring non-methane hydrocarbons become available.
2.1.3.7 Meteorological Sensors - In addition to data on pollutant concen-
trations, it is necessary to have available some source of meteorological
data for use in dispersion modeling and other data analysis efforts utiliz-
ing the monitoring network data. Hie data should include wind speed, wind
direction, and vertical stability information, although most networks
include only wind speed and direction, since vertical temperature param-
eters are difficult to monitor in urban areas.
Wind data may often be adequately supplied by the National Weather Service
or by commercial consultants. In other cases, however, the National
Weather Service airport site may be too remote, or the data otherwise less
than adequate, and wind speed and direction sensors should be included in
the air quality monitoring network. Such sensors, if included, should be
placed at sites where several continuous instruments are housed together,
in order to obtain the greatest use of the data for modeling and research
purposes. An adequate number of meteorological sensors would probably be
on the order of one-half or less of the number of such stations.
11-19
-------
Information on vertical stability can usually be adequately obtained in-
directly by utilizing inferred relationships between wind conditions, time
of day, insolation and vertical stability classes. Observations of temper-
ature at several heights near the surface are very useful to infer stability
for short-term modeling and air quality forecasting, but extensive measure-
ment of vertical parameters is usually only done on a research basis.
2.1.3.8 Combined Sites - As has been noted above, it has been common prac-
tice to consider the configuration of an entire network covering all pol-
lutants, as a whole rather than on a strict pollutant-by-pollutant basis.
This is, of course, done as a matter of economy, both of cost and manpower,
it generally being more economical to have as many sensors collected at
one site as possible.
It is considered appropriate to combine instruments to a certain extent.
However, it is not appropriate to routinely house all instruments for all
pollutants together as has often been common practice, except for back-
ground sites.
The peak and neighborhood type sites for total suspended particulate and
SOp may very reasonably be combined. As was noted, it is specifically
recommended that in the case of research and planning sites, hydrocarbon
and oxides of nitrogen sensors be collected together together into sta-
tions, which may also reasonably include hi-vols and 862 sensors.
However, as also noted above, the locations of peak levels of the various
pollutants are in most cases not at the same location within the area.
Most prominent example of this is carbon monoxide. Although it is obvi-
ously convenient to have all the continuous sensors together, it is
extremely rare to find a site large enough for a full monitoring station
that is also in an appropriate location for peak CO monitoring and,
indeed, sites suitable for CO monitoring are not necessarily suitable for
11-20
-------
other pollutants, depending on the purpose. Hence it is not prudent to
presume that CO sensors can be located with the others, although if pos-
sible of course it should be done.
2.1.4 Additional Guidance
Other recent and current EPA contract efforts relevant to the issues of
network design, optimization, and evaluation include:
Guidelines for Air Quality Maintenance Planning and Analysis,
Volume 11; "Air Quality Monitoring and Data Analysis"
Subject Matter; This document provides states with
planning information and guidance for the preparation
and implementation of a monitoring system which is
compatible with the goal of air quality maintenance
and the need for the development of Air Quality Main-
tenance Plans.
Status; The guideline document (also identified as
EPA-450/4-74-012 and OAQPS #1.2-030) has been completed
by the GGA Corporation, September 1974.
EPA Project Officer: Alan J. Hoffman, MRB, HDAD, OAQPS.
Collection and Integration of Operational Characteristics of
Existing Pollutant Monitoring Networks
Subject Matter; This study deals with the analysis of
operational data gathered from five superior air and
water monitoring networks to identify the most efficient
and economical methodology by which a monitoring network
can satisfy its responsibilities and optimize the cost-
effectiveness of daily operations. The goal of the pro-
ject is the development of manuals that would furnish
the desired techniques for evaluating operations, and to
provide methodologies by which the efficiency and/or
cost-effectiveness of all operations could be readily
considered along with the effects of alternative actions
where the evaluation indicates that improvement is needed,
while remaining within budgetary constraints and meeting
network objectives.
Status; The project is being carried out by URS Research
Corp. and is expected to be completed by November 1975.
EPA Project Officers: Edward A. Schuck and
Leslie Dunn, MSA, NERC-LV
11-21
-------
2.2 INSTRUMENT SITING AND PROBE EXPOSURE
After the general location of a sampling site is selected, based on con-
sideration of the Region-wide configuration, it is necessary to select a
site for the sensor or station, and then within the confines of that
choice to determine the precise location of the inlet probe in the case
of gaseous pollutants.
2.2.1 Site Selection
The selection of a precise site, once a general area has been
selected, is primarily a question of availability, accessibility, secu-
rity, and the potential effect of surrounding structures. The issues of
accessibility and security are the ongoing concerns of the daily opera-
tion of a network, and there is little additional guidance to be offered.
The issues of ground-level versus rooftop sites might be considered a
site-selection problem, as availability is one of the primary reasons for
seeking rooftop sites; however, the impact of the choice is more in the
nature of a probe placement issue, and it is so considered here.
Sulfur dioxide is considered to be rather well mixed near the ground, at
least at receptors not overly affected by specific point sources. There-
fore, either ground or roof-top sampling is adequate, and the choice can
be made on the basis of site availability. However, care must be taken to
ensure that rooftops are 'clean,' i.e., free from space heating vents,
laboratory hood vents, and the like, that may have S02 emissions. Once above
the effect of reentrainment from the ground, it is generally considered
that TSP is also fairly well mixed for the next few hundred feet above the
ground. Hence rooftop sampling has traditionally been recommended in order
to avoid influence of possible reentrainment effect, and rooftops up to
several stories high have been used, particularly at center city sites. If
the reentrainment is to be considered, however, perhaps as part of the popu-
lation exposure, then a site that permits ground level (2 to 3 meters)
sampling is required. If such a site is not attainable, an alternate
arrangement such as a portable sampler should be considered. This is a
11-22
-------
clear instance where the purpose of the monitoring needs to be very
precisely stated to determine the appropriate siting action.
The obvious case where station siting depends on the purpose of the
monitoring is with CO, where a station may be either a street canyon,
neighborhood, corridor, or background station. In contrast to the case
with SC>2, the horizontal distribution of CO across an urban area consists
of so many alternating areas of peak and valley levels, one at each
street or major traffic center, that one must consider site locations
for CO primarily in probe placement terms, in scale of plus or minus a
meter or two. Hence a peak station site needs to be essentially adja-
cent to the street in question and needs to permit nose-level sampling,
while a neighborhood site must be located at least 35 meters from the
nearest street. This setback will limit the influence of the nearest street
to about 1 ppm and make the reading more representative of the general
community in which the monitor is located. The strong dependence of carbon
monoxide concentration upon distance from the nearest roadway has been
illustrated in a number of studies. ' Generally it was found that the
concentrations experienced by pedestrians exceeded those measured at a
typical air monitoring site, while concentrations at randomly selected
locations throughout the survey grid were less than those at the site.
More specifically, the data in one study indicated that average concentra-
tions determined by the monitor would be reduced to near the urban back-
ground level by moving the monitor approximately 200 feet farther back
o
from the street. Figure II-5 indicates how the CO levels at the various
stations in Los Angeles are closely related to the slant distance from
the street, despite presumably different traffic volumes in the various
locales. It is also known that for peak CO sampling within street canyons,
the side of the street which is opposite the side facing the rooftop-level
winds will experience the higher concentrations (see Figure II-6). Hence
in any location with a significantly prevailing wind direction, even the
choice of the side of the street becomes a relevant siting question.
11-23
-------
Z
UJ
o
o
s
UJ
X
«o
o
v£> CO
ON
r-l 60
CD -H
4J O
4-1
14-1 -H
o c
00 M
CO -,-1
)-i CB
0)
> 4J
LO
i
H
00
H
11-24
-------
BACKGROUND
CO CONCENTRATION
PRIMARY RECEPTOR
VORTEX
TRAFFIC
LANE
-W-
Figure II-6. Schematic of cross-street circulation in street canyon'
11-25
-------
2.2.2 Probe Placement
Within the several meters scale involved in a typical monitoring site,
there is general latitude in the precise placement of the inlet probe.
For the gaseous pollutants (excepting CO), this is an issue primarily
involving security from vandalism, the avoidance of any restrictions
to the air flow, such as from the station itself, trees, etc., and any
undue influence from a minor local source, such as a stack located on
the roof of a building where the air inlet is located. These require-
ments are generally taken to indicate a height above the ground of 3
to 15 meters, and either a vertical clearance above the roof of 1 to
2 meters or, in a different configuration, a horizontal clearance
beyond the supporting structure of at least 2 meters.
In the case of particulates, the hi-vol represents a special situation.
Historically, the NASN hi-vols have been on rooftops, sometimes 8 to
10 stories high. This avoids reentrained surface dust, and the atten-
dant variability, and in so doing provides a smoother, more reliable
record for trend purposes. However, it can be argued that elevating
the sampler in this way makes the resulting data an inaccurate reflec-
tion of true population exposure. Table II-4 provides, as an example,
a comparison of 5 months' data from the CAMP Station in Philadelphia
and the Franklin Institute site operated by the Philadelphia Department
of Public Health. The CAMP Station hi-vol, at 11 feet, reads consis-
tently higher than the City Station, at the same location, which is at
about 50 feet. It is probably also true, though perhaps less thoroughly
demonstrated, that the distance of the hi-vol from a nearby street is of
importance. Since streets, walkways, and other such areas are a source
of reentrained particulate matter, it is probable that placing a hi-vol
on a one-story roof, for instance, is not the same as placing it in an
open area or on a trailer, even if the height above the ground is the
same.
11-26
-------
The pollutants NMHC, NO, N02 and Oo are tied together in a precursor -
secondary product relationship and should therefore be considered as
an integrated system in site selection. While hydrocarbons are emitted
in much the same pattern as CO, an elevated site in the CBD is more appro-
priate than a ground level one. This is to limit the influence of any
single street and provide a more representative measurement of the CBD
as a whole. NO and N0£ should be monitored at this location to provide
information on ratios of NMHC to NO and N02-
As photochemically produced secondary pollutants, N02 and 03 are con-
sidered to be well mixed vertically and of relatively uniform concen-
tration over a large area. Therefore, either rooftop or ground level
sampling are adequate and the prime concern is the location of a
favorable distance downwind of the CBD to locate the zone of maximum
concentration. Under normal wind speeds, this zone thought to be 10-
15 km for N0£ and 15-25 km for oxidants. Special precaution should be
taken not to locate 03 sites within 100 meters of major traffic
arteries or large parking areas due to the scavenging effect of NO
emissions.
II-26B
-------
Table II-4. COMPARISON OF HI-VOL DATA AT TWO DIFFERENT
HEIGHTS - FRANKLIN INSTITUTE, PHILADELPHIA
Date
Nov. 1, 1974
Nov. 13, 1974
Nov. 19, 1974
Nov. 25, 1974
Dec. 1, 1974
Dec. 7, 1974
Dec. 13, 1974
Dec. 19, 1974
Dec. 25, 1974
Dec. 31, 1974
Jan. 1, 1975
Jan. 6, 1975
Jan. 8, 1975
Jan. 12, 1975
Jan. 18, 1975
Jan. 24, 1975
Jan. 30, 1975
Feb. 5, 1975
Feb. 23, 1975
Mar. 1, 1975
Mar. 7, 1975
Mar. 13, 1975
Mar. 19, 1975
Mar. 25, 1975
Mar. 31, 1975
Geometric mean
City
199
48
122
69
73
113
174
94
101
54
31
88
107
42
54
153
44
45
64
80
119
84
56
84
46
76
CAMP
264
76
187
116
117
154
237
133
143
82
52
190
136
71
76
228
81
76
138
254
206
122
92
128
94
125
Ratio
1.33
1.58
1.53
1.68
1.60
1.36
1.36
1.41
1.42
1.52
1.68
2.16
1.27
1.69
1.41
1.49
1.84
1.69-
2.16
3.18
1.73
1.45
1.64
1.52
2.15
1.64
11-27
-------
It is recognized that the effect of height seen in Table 2, and other
similar concerns, indicate that many of the hi-vol networks and sam-
pling sites that have been used in the past are generally not as com-
parable with each other as is the case with other pollutants. Ulti-
mately, the need for a greater degree of homogeneity will likely require
that adjustments be made in the way hi-vols are typically placed. How-
ever, because of the large number of sites involved, and the length of
historical record at many of them, such an adjustment would be an issue
of major concern and significance. Since a large amount of good quanti-
tative information on the topic is not currently available, it is con-
sidered inappropriate to make major recommendations at the present. The
effect of height, etc., can be taken into consideration in interpreting
hi-vol data, and it is recommended that this be consistently done.
Several study programs are underway that will provide much better infor-
mation on these questions in the near future, and the guidance material
will then be revised as appropriate. It is expected that guidance in this
area will be in the form of a supplement to this document, similar to the
CO Supplement, and will be issued in early 1976.
However, the issue of probe placement is the most serious concern in the
case of CO. Even within the scale of a typical monitoring site, CO
levels can vary dramatically. As indicated in Figure II-7, CO levels
can change with vertical height at a rate more than 1/2 ppm per meter.
Figure II-8 illustrates the sizable changes possible with short hori-
zontal changes in the vicinity of a typical peak site location. Thus,
while there is no single "right" position for a CO probe, it is obious
that some major degree of standardization is needed to ensure uniformity.
The currently recommended positions are discussed in Supplement A. These
positions were selected not only to standardize probe and station
locations, but also to provide a reasonable measure of population exposure
in the breathing zone.
A summary of the current recommendations concerning station siting and
probe placement is presented Table II-5.
11-28
-------
40
30
20
UJ
X
10
LEEWARD SIDE
OF STREET
WINDWARD SIDE
OF STREET
i I I I I I I I I V i i i i I
0
10 15
CO,ppm
20
s
Figure II-7. The vertical distribution of CO concentration on a
street with traffic volume of 1,500 vehicles/hour''
11-29
-------
C
o
H
4J
CJ
QJ
W
t-l
OJ
CO
01
C
60
H
0)
0)
(U
CO 03
G
D, rfl
a)
B
00
I
3
60
H
11-30
-------
2.2.3 Additional Guidance
Other recent and current EPA contract studies relevant to the issues
of site selection and probe placement include:
"Selecting Sites for Carbon Monoxide Monitoring"
Subject Matter; This report presents procedures and
criteria for selecting appropriate locations for CO
monitoring stations which fulfill specific monitoring
objectives. Procedures are given for selecting loca-
tions that will provide CO measurements representative
of downtown street canyon areas, urban neighborhoods,
and larger interurban regions. Specific recommenda-
tions are given for inlet heights, distance from major
and minor roadways and placement re ative to urban
areas. Ihe rationale behind each specific recommenda-
tion is also given.
Status; The first draft report prepared for EPA by
Stanford Research Institute is being reviewed. A
final report is expected by the end of September 1975
EPA Project Officer; Neil J. Berg, Jr., MRB, MDAD, OAQPS
11-31
-------
*f
2
9
t*
S
-1
u
a
M a
5 2
8 °
4*^
.. M
| S
8
-> u
o. e
« M
I 8
to "o
S
tl
J
.c M
oo oo
H
15
"8
§
H O
0 4J
U CO
r-( O
1) 0
01 0)
4J 4J
i-l -i-l
S
V
4J
4J
C
CO
w
3
i-l
O
to
M CM CM EM
A A A A
CM CM CM
e o o i
W 4J 4J
i-l r-4 r-4
m m m m
,_! iH ,-4 t-l
o o o o
U 4J 4J 4J
o en m CM
0
CD 0)
O, -S »
to « 4J 4J
Eco co 0
4J 4J a ai
TO C O -rl
to 4J to
§,-10) to M
CO U rH 00
M 0 ft, 3 (2
IH -rl oi O, X O
MM- O 4J i-l
O O (U P« 1-4 00
01 4-* T3 M v-1 OJ
C M C 3 « 0 Pti
^ -H CO CO O 3
e j: o o- C
M " P* to 1-4
0) - X X i-l M J3
4J t-H 4J OJ tO W 4J
"° 1 c o "f -D 0°"
4-» £ <1) i-J C C CO C/D
C T3 4J O TO 0)
^ C « J4 M
O O C '"' "^ (00) O
o.^) o d o> c 'w
« -ft (l. C )-i C
§D (fl O « i
2 *
i 1
CMCMCM CMCMCMCMlSl
AAA AAAAA
-
CM CM CM CM CM CM
I 1 O O O O O O
4J ^i [j | > [i [_f
inmm mmmmm
i-li-lf-l i-4r-4i-lt-4r-l
ooo o o o o o
4J4J.4J 4J 4J4J 4J4J
CMCMC*> roPicotnco
O
PQ
r i ***
s °
, TJ *J
(U U i-l - tl
"5 111 « e
^ C 4J -rl 0
» M 4J 4J
"4J o a O to
O u 0) 1
§0 -o g -o
rf Tl 5
CO1" CO -rl *J
rl T3 <4-l OJ MO.
S 10 14 01
CBM (3 o
Sou > M
Oh x 0 w
"O I4J p tl ^1
o 1-1 js
3B ""i t» o 4B * »2 «J
M a a m 1-1 c B a
cNoih B 3 M ra > o O o
OO Kjinnaio N M M
C/3C/3 CMflc04J(4ta OSO O
oo 5 c co tu Si
MM COMttg) M h Li M
oo >ri4JM-id Ur-iv omo o
I4-IM-I4J i-l CD *4-4 r-l IH 1*. (
C m B .* B -ri 0
noi Vr-4 MI-IOK BOB a
ran » O « MA ouig to
o> « O a w a> h _
0101 Mctti-l 0)E3 9!Om ci
e e a i-i . M u 1 c s^g g
coco «i «j " o -rj o o a a a
cocom CAOKM CACA
"S "8 "8
.§ "§ ^ "§ ^ "§
M 3 MS M 3
O O O O O O
n M «^ M ,O L|
j: oo js oo ja op
oc j: ji oo ^ J^ ooj?
rlUCO -rlUCO-rlO
oitntu aicooicuco
SSEQp* 2PQPL|?5PQ
tol
II
L)
a
3 ' a
a C -a
rl ' U -rl
> . NX
t «| O O
a SB
11-32
-------
P*J
23
w
8
PM
S
§
PM
1
g
M
H
M
CO
B
3
en
CO
g
H
*
S
O
g
£.0
CO
X)
I
H
4J
0
o
in
B
3
^
^
&
,tl
° -H S
O*^ jj
^ i ^ i *^
rtt 0)
* ^ C Tj o
H c 5
<<<. *'
o ay a) 01 f| Jj «, ,-j r-H
i 1 r 1 fH i-H ^i ^1 J __J
a P. tx tx*-; g co tH *::!
§§§ §-2 8 § 2«2«
»c/ic/>c/i<|iri>4~ " ni n|< B
a) _c -.-i M> ai r! " coco
v &0r-i^;r;;11- a)r-i VrH
W-rJMO a> o aj co iH 41 a)
MZU«(1»O. * "^
06
gai!
T3,
C ,0 -H
S 2 b K
1 5 3 5 £^
£ 2 ^} O 0 00 -H '
,2 § S H -H 0 X
Ss it) )-i M o
H C > *J 4-) «H
0 O .C -H rt Q
8
-------
"Determine Optimum Site Exposure Criteria for TSP Monitoring"
Subject Matter; The purpose of this contract is to develop
specific optimum site exposure criteria for TSP monitoring
which could be applied generally. Criteria will be developed
for a limited number of different types of sites, each of
which achieves some specific monitoring objective or set of
objectives. This will be accomplished by the following:
(1) conduct a literature search on the nature and purpose
of ambient TSP monitoring; (2) determine a specific set of
objectives to be achieved by ambient TSP monitoring and the
relative importance of each objective; (3) delineate repre-
sentative types of monitoring sites which achieve one or
more of the monitoring objectives; (4) for each representa-
tive type of site, determine optimum exposure criteria which
could be applied uniformly to that type of site; and (5) for
each type of site, determine the relative effects of various
TSP sources both nearby emitters and those further away.
Status; Stanford Research Institute has been chosen to
perform this study.
The expected completion date is March 1, 1976.
EPA Project Officer; Neil J. Berg, Jr., MRB, MDAD, OAQPS.
"Determine Optimum Site Exposure Criteria for SO^ Monitoring"
Subject Matter: This project is similar to the one for TSP
monitoring described above.
Status: The Center for Environment and Man has been chosen
to perform this study. Estimated completion date is February 1, 1976.
EPA Project Officer; Neil J. Berg, Jr., MRB, MDAD, OAQPS.
Monitoring and Data Analysis Division.
"Study of the Feasibility of Determining Optimum Site Exposure
Criteria for 0,_, NO,,, and Hydrocarbon Monitoring"
y~- L
Subject Matter: The purpose of this contract is to in-
vestigate the feasibility of determining optimum site
exposure criteria for Ox, N02 and hydrocarbon monitoring
which could be applied generally. This will be accom-
plished by the following: (1) conduct a literature search
on the nature and purposes of Ox, N02 and HC monitoring;
(2) determine a specific set of objectives to be achieved
by Ox, N02 and HC monitoring, and the relative importance
11-34
-------
of each (if there is a lack of data which precludes this
determination of monitoring objectives, fully document
this data void and suggest means to obtain the necessary
information); (3) delineate representative types of moni-
toring sites which achieve one or more of the monitoring
objectives; and (4) prepare a final report summarizing as-
sumptions, findings, and conclusions of this study.
Status; Stanford Research Institute has been chosen to
perform this study. Estimated completion date is December 1,
1975.
EPA Project Officer; Neil J. Berg, Jr., MRB, MDAD, OAQPS
"Development of a Study Plan to Determine the Air Quality Grad-
ients at Air Monitoring Sites"
Subject Matter; The purpose of this contract is to develop
a study plan to define the area for which a point samplers
data may be "representative." The study plan should address
various pollutants, differing monitoring objectives, and
site exposure criteria in determining the three-dimensional
air quality gradients around monitoring sites. The plan
should define limits of "representativeness" as well.
Status: Rockwell International Air Monitoring Center has
been chosen to perform this study. The expected completion
date is December 1, 1975
EPA Project Officer; Alan J. Hoffman, MRB, MDAD, OAQPS
2.3 NETWORK OPERATION
In addition to defining the configuration of the network and actually
siting the monitors, the process of monitoring network design also in-
cludes the selection of appropriate instrumentation and the definition
of various procedures for the operation of the network.
2.3.1 Monitoring Equipment Selection
The selection of monitoring instruments for use in a network is an im-
portant aspect of overall network planning. EPA has established a set
of procedures for establishing whether monitoring methods are reference
11-35
-------
methods or equivalents, and thus acceptable for meeting SIP requirements.
This was published as a regulation in 40 CFR 53 on February 18, 1975.
The burden of proof of whether an analyzer is a reference method or
equivalent falls upon the manufacturer. Many analyzers currently in
use are no longer manufactured per se (that is the specific make and
model). Since the vendor will have no incentive to test these analyzers
for reference method or equivalency, EPA will in most cases make the
necessary tests.
2.3.1.1 Reference Method Determination - For S0? and TSP, the measure-
ment principle specified is a manual method. (Pararosaniline for SCL,
hi-vol for TSP) thus, there is only one reference method for S02 and TSP
since the method consists of a series of mechanical steps or chemical
operations to be performed. For CO, ozone and N0«, only the measurement
principle and calibration procedure has been specified. Any analyzer
utilizing the specified measurement principle and calibration procedure
and which meets the performance specifications in 40 CFR 53 will be
designated as a reference method. Thus, as an example, there could be
as many reference methods for CO as there are different models of
NDIR analyzers.
2.3.1.2 Equivalency Determination - In general, equivalency to a refer-
ence method is determined by passing the tests for demonstrating a
consistent relationship to a reference method and by meeting performance
specifications. If the candidate equivalent method is a manual method
only the consistent relationship need by established. If the method is
an automated method then both the consistent relationship and performance
specification tests must be passed in order to be designated as an
equivalent method.
At the present time, reference methods exist only for SO and TSP which
are described in 40 CFR 50. They are the high volume procedure for TSP
and the pararosaniline (sulfamic acid) procedure for S0~. Any other
manual methods for these pollutants are unacceptable.
11-36
-------
For CO and ozone, EPA is awaiting data from manufacturers before designat-
ing any reference methods. For the present time, any instrument utiliz-
ing the NDIR measurement principle for CO and the chemiluminescent prin-
ciple for ozone will be acceptable. It is possible that instruments
utilizing the NDIR principle for CO or the chemiluminescent for 0,. will
be unacceptable. This situation could occur if the manufacturer fails
(or if EPA tests the analyzer and it fails) to pass the performance
specifications tests.
For NO-, no reference measurement principle or methods exists since the
Jacobs-Hochheiser (J-H) technique was rescinded. The chemiluminescent
measurement principle will be proposed very soon to replace the J-H tech-
nique. The triethanolamine guiacol sulfite orifice method (TGS) and the
sodium arsenite orifice (ARS) method will be tested for equivalency as
soon as a reference method is designated.
Unacceptable manual methods should be changed to the reference method
or equivalent within 6 months. Unacceptable analyzers (automated methods)
should be changed to a reference method or equivalent as soon as prac-
ticable but no later than 5-years (February 1980).
Automated analyzers not utilizing the reference measurement principle
and calibration procedure and which fail equivalency tests should be
replaced with a reference method as soon as practicable but no later
than 5-years (February 1980).
2.3.2 Operating Procedures
There are at least two types of operational decisions that affect the
design of the network in the sense that they affect the type of data
produced. In the case of intermittent sampling, the frequency of opera-
tion is such a decision, and in the case of continuous monitoring, the
selection of the instrument operation range is also. Note that the many
11-37
-------
other operating procedures associated with the quality assurance aspects
of a monitoring network are not considered here, although they are none-
theless of major importance also.
2.3.2.1 Intermittent Sampling Frequency - The entire point of sampling
intermittently, such as the every-6-days schedule used for hi-vols and
bubblers, is to provide some measure of air quality knowledge at a cost
less than that associated with more frequent sampling. Such a program
necessarily introduces some uncertainty into the statements that can be
made based on the resulting data. However, standard statistical pro-
cedures are available to provide estimates of this uncertainty, and to
indicate how to adjust the sampling frequency to provide an appropriate
degree of uncertainty.
Figure II-9 indicates how the range of uncertainty with respect to the
NAAQS varies with the sampling frequency. At 61 samples per year, an
3
annual mean TSP level of 75 may be 8 ug/m higher or lower (95 percent
confidence limits); if the sampling frequency is tripled, the uncer-
3
tainty drops to + 3 (ig/m . Thus, with sites having levels near the
standard, greater sampling frequency may be needed to precisely define
compliance, while at sites with levels well above or well below the
standard, less frequent sampling may be adequate. It should also be
noted that the uncertainty increases or decreases linearly with the
value of the standard geometric deviation; the 1.6 used to calculate
Figure II-9 is a typical value.
Table II-6 presents similar information relevant to the 24-hour standards;
specifically the table presents the probability, for each of three sam-
pling frequencies, that at least 2 of the days over the 24-hour standard
will be detected; i.e., that the site will be considered in violation of
the standard. Note that, again, if the site is only marginally in viola-
tion, quite frequent sampling is needed to detect this, while a site a
large number of excursions is almost assured of being so identified
11-38
-------
SiET
&&:
ABOVE THE STANDARD
iH
iffi
*
|:|^
JO
X)
1 1
IMPJMillil
i= g ' i
^ Li- i
S '
i£ j
u_ 1 «
0 == |
XvXOX"X*X*X'X*X*X*X*X'X*X*X*X>X*X*X"X*X*X"'
1
1
1
1
1
1
1
1
1
1
1
1
1
1 i
co r r^. «j
14T>
CO
CO
oo
r- «
CM
CM
, «
^H
^-«
->
3
ce
-
a:
UJ
o_
oo
Q_
co
,-1
n)
3
cr
4J O
CO -H
4J
!>*» cO
Vi -i-l
0 Hi
H ^
V4
& O
H
"ZL
&
0»
4J
C
H
01
O
0)
*o
H
C
o
o
4-1
a
o
^
01
ex
cu
£>
H
M-l
1
4-1
01
H
!Z3
cu
rt
CO
en
cfl
CO
cu
H
o
Q)
3
cr
cu
j*^
m
60
C!
H
rH
B*
CO
cn
CO
O
H
CO
>
£uj/3rf 'NO!iVaiN33NOO dSl
I
H
H
CU
3
H
11-39
-------
Table II-6. PROBABILITY OF SELECTING TWO OR MORE DAYS WHEN SITE
EXCEEDS STANDARD
Actual number of
excursions
2
4
6
8
10
12
14
16
18
20
22
24
26
Sampling frequency, days/year
61/365
0.03
0.13
0.26
0.40
0.52
0.62
0.71
0.78
0.83
0.87
0.91
0.93
0.95
122/365
0.11
0.41
0.65
0.81
0.90
0.95
0.97
0.98
0.99
0.99
0.99
0.99
0.99
183/365
0.25
0.69
0.89
0.96
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
2.3.2.2 Instrument Operating Range - Since continuous instruments can
usually be adjusted electronically to operate in various concentration
ranges, the selection of such range is a necessary decision in the pro-
cess of network design. Generally, the decision is simply to utilize
the smallest range that will encompass the maximum expectable levels,
and this is usually adequate. However, in the case where very high
levels (usually S09) from a major source are received at a site that
normally experiences low background levels, the use of a single range may
not be possible. If the range is set too low, accurate documentation of
the peaks is lost offscale, while if it is chosen high, there will not be
adequate precision in the data concerning the low levels. There is no
way to resolve this with a single instrument. In such a case, one or the
other orientation (population or source) must be selected as primary, and
the instrument site coded 01 or 02 as appropriate. A good solution would
11-40
-------
be to have a continuous instrument adjusted to measure the peak levels,
and a bubbler for long term population exposure.
11-41
-------
SECTION III
REFERENCES
1. Code of Federal Regulations. Title 40. Section 51.17.
2. Federal Register. Volume 33, Number 10. January 16, 1968.
3. McCormick, Robert A. Air Pollution Climatology. Chapter 9 in
Stern A.C., Air Pollution. 2nd Edition. Academic Press. 1968.
4. Lynn, David A. Air Pollution-Threat and Response (in press)
Addison-Wesley. Reading, Mass. 1976.
5. CAMP in Washington, B.C. 1962-1963. Publication Number AP-23.
Department of Health, Education and Welfare. 1966.
6. Kinosian, John R. and Dean Simeroth. The Distribution of Carbon
Monoxide and Oxidant Concentrations in Urban Areas. California
Air Resources Board. 1973.
7. Johnson, W. B. et al. Field Study for Initial Evaluation of
an Urban Diffusion Model for Carbon Monoxide (APRAC-la). Contract
CAPA-3-68 (1-69). Stanford Research Institute. 1971.
8. Ott, Wayne R. An Urban Survey Technique for Measuring the Spatial
Variation of Carbon Monoxide Concentrations in Cities. Depart-
ment of Civil Engineering. Stanford University. 1971.
9. Guidelines for the Interpretation of Air Quality Standards. OAQPS
Guideline Numbers 1.2 - 008. 1974.
III-l
-------
-------