-------
TABLE 9-2. CORRELATION/REGRESSION ANALYSIS BETWEEN AIRPORT
EXTINCTION AND COPPER SMELTER SOV EMISSIONS
y\
OJ
fa
Data set
Tucson (1950-75)
Tucson (1959-75)
Phoenix (1959-75)
Wins low (1948-73)
Prescott (1948-75)
Prescott (1948-69)
Correlation
coefficient
0.91
0.88
0.81
0.68
0.70
0.70
Regression coefficient
extinction/emissions,
(10* mfVdOOO TPD)
0.035
0.038
0.041
0.047
0.031
0.039
t- statistic
(t * 1.7 for 95% confidence)
(t s 2.5 for 99% confidence)
11.1
7.2
5.4
4.5
5.0
4.4
-------
?
u>
o
' 250 TONS/DAY SO
URBAN SITE
NONURBAN SITE
100 200
I I
SCALE, milM
MESA VERDE
7%l NATIONAL PARK
PHOENIX 1-62%
X
67%)TUCSON
Figure 9-15. Seasonally adjusted changes in sulfate during the copper strike are compared with the
geographical distribution of smelter SOX emissions.
Source: Trijonisand Yuan (1978a).
-------
(£>
CO
cn
• - 250 TONS/DAY SO2
(J URBAN AIRPORTS
NONUR^AN AIRPORTS
0
L
100
I
SCALE. milM
200
_J
Figure 9-16. Seasonally adjusted percent changes in visibility during the copper strike are compared
with the geographical distribution of smelter SOX emissions.
Source: Trijonisand Yuan (1978a).
-------
Visibility improved at almost all locations during the strike, with the largest improve-
ments occurring near and downwind (north) of the copper smelters in southeast Arizona and near
the copper smelters in Nevada and Utah. The nine locations showing statistically significant
improvements are all within 150 miles of a copper smelter Attributing the improvement in visi-
bility entirely to the drop in sulfate levels yields an estimated extinction efficiency of 3 9
~6 ~1 3
x 10 m /(ug/m ), in agreement with data in Table 9-1.
Altshuller (1973) has noted an increase over the past decade in sulfate concentration at
nonurban sites in the Eastern United States, which is not inconsistent with the decreasing
trend in nonurban site median visibilities noted by Trijonis et al. (1978b). Unfortunately,
the historical record of sulfate concentrations extends back only to the mid-I9601s. Within
the Eastern United States, over 90 percent of the S0x emissions are associated with the com-
bustion of coal and oil. One apparent conclusion is that visibility reduction is currently
due in largp part to increases in sulfate aerosols, which are formed primarily from coal
combustion-related SO* emissions; examination of the trends and changing spatial distributions
of coal use should be comparable with the change in the light extinction coefficient.
Air pollutants emitted over the Eastern United States result mainly from the combustion
of fossil fuels—coal, oil products, and gas. The great spatial and seasonal variability of
haziness (inverse of visibility) prompted Husar et al. (1979) to examine the patterns of coal
consumption in the Eastern United States over the past few decades. For comparison with coal
consumption estimates, visibility data are expressed in terms of a light extinction coeffi-
cient, b ., via the Koschmieder formula: b . = 3.92/V.
Figure 9-17 illustrates the striking similarity between summertime average haziness and
coal use within the Eastern United States over the past three decades.
Further support for the relationship between coal consumption and haziness is given in
Figure 9-18.
As shown in Figure 9-18, in 1951 the haziness was most pronounced in the wintertime, when
the coal consumption was the highest. By 1974, there was a shift toward a summer peak, coin-
cident with the increasing summer use of coal. Such coincident behavior alone cannot estab-
lish cause-effect relationships. It is nevertheless instructive to examine the more detailed
spatial and temporal patterns of coal use and haziness (specifically, extinction).
Since 1940, the trend in coal consumption has been more pronounced in the summer than in
the winter (Figure 9-18; U.S. Bureau of Mines, 1933-74); since 1960 summer coal use has grown
by about 5.8 percent per year compared with 2.8 percent per year for winter coal demand. The
monthly coal combustion peaked in the winter in the early 1950's, but the seasonal pattern had
shifted to a summer peak by 1974 (Figure 9-19). f
The corresponding regional trends of haziness in the Eastern United States (Figure 9-20;
Husar et al., 1979) exhibit changes similar to those of coal combustion.
In the Ohio River Valley region, the winter (quarter 1) average extinction (bext, km )
decreased slightly, whereas the spring (quarter 2) average increased. The summer (quarter 3)
SOX9A/A 9-37 12-23-80
-------
176
150
125
110°
I
1 75
i
50
25
I I I I I I I I I
I I I I I I I I I I I Ins
4.0
3.5
3.0
2.5 E
o
2.0 i
1.5
1.0
1940 1950 1960 1970 1980 1990 2000
YEAR
Figure 9-17. Compared here are summer trends of U.S. coal
consumption and Eastern United States extinction coefficient.
Source: Adapted from Husar and Patterson (1980).
I I I I I I I I I I I
rrT"!
ELECTRIC UTILITIES
iimiiiiiiikiii
JFMAMJJASOND JFMAMJ JASOND
g
i
Figure 9-18. In the 1950's the seasonal coal consumption peaked in
the winter primarily because of increased residential and railroad use.
By 1974, the seasonal pattern of coal usage was determined by the
winter and summer peak of utility coal usage. The shift away from a
winter peak toward a summer peak in coal consumption is consistent
with a shift in extinction coefficient from a winter peak to a summer
peak in Dayton, OH, for 1948-52.
Source: U.S. Bureau of Mines Yearbooks 1933-74.
9-38
-------
2
O
t-
cc
O
I
c/3
H
O
175
150 -
125 -
100 -
175
1940
1980
" ' ' i ' ' ' ] ' • i
WINTER
1940
1980
Figure 9-19. In 1974, the U.S. winter coal consumption was well below, while the summer consumption
was above, the 1943 peak. Since 1960 the average growth rate of summer consumption was 5.8 percent
per year, while the winter consumption increased at only 2.8 percent per year.
Source: U.S. Bureau of Mines, Minerals Yearbooks 1933-1974.
9-39
-------
OHIO RIVER
OUPBTE* ?
0.1
o L.
194050 60 70 80 90 194050 60 70 80 90
NE. MEGALOPOLIS
1940 50 60 70 80 90 1940 50 60 70 80 90
EASTERN SUNBELT
1940 50 60 70 80 90 1940 50 60 70 80 90
YEAR
1940 50 60 70 80 90 1940 50 60 70 80 90
SMOKY MOUNTAINS
1940 50 60 70 80 90 1940 50 60 70 80 90
0 L_ ..._.....,........., J ' .,_ '
1940 50 60 70 80 90 1940 50 60 70 80 90
YEAR
Figure 9-20. Trends in the light extinction coefficient (bgxt) in the Eastern United States are shown by
region and by quarters; 1 (winter), 2 (spring), 3 (summer), 4 (fall).
Source: Husar et al. (1979).
9-40
-------
extinction increased from roughly 0.25 in the 1950's (a visibility of 10 miles) to abo'.t 0.-
in the 1970's (a visibility of less than 6 miles). Fall (quarter 4) extinction remain
essentially unchanged. The summer average increased from about 2 to 3, corresponding to *
reduction in visibility from 12 to 8 miles (20 to 13 km). The Northeast megalopolis regie-
shows a general decline in haziness during quarters 1 and 4, whereas quarters 2 and 3 display
a slight increase from b = 2.7 (9 miles, 15 km) to 3 (8 miles, 13 km) x lo'^if1. The Smo*y
Mountain region displays a strong increase in the average summer quarter extinction coeffici-
ent from about 1.6 to 4, corresponding to visibility deterioration from 15 to 6 miles. Small-
er but still pronounced increases are noted for quarters 2 and 4. Evidently the Smoky Moun-
tains have become appreciably "smokier" over the past 20 summers. The Eastern Sunbelt region
has an increased haziness for all quarters, most pronounced being the summer quarter, with an
increased from 2 (12 miles) to 3.5 (7 miles). In the Midwest, extinction during the first
quarter fluctuated slightly with no discernible trend. The spring and fall quarters have
increased appreciably, but summer values have nearly doutfled, from 1.5 to 3 (16 to 8 miles).
The spatial shifts of Eastern United States haziness are displayed in greater detail in
Figure 9-21 (Husar et al., 1979).
9-5.1 Natural Versus Manmade Causes
Vision in the natural, unpolluted atmosphere is restricted by blue sky scattering, (air
molecule light scattering is often termed Rayleigh scattering) by curvature of the earth's
surface, and by suspended liquid or solid natural aerosols. Important sources of natural
aerosols include water (fog, rain, snow), wind-blown dust, forest fires, volcanos, sea spray,
vegetative emissions, and decomposition processes. The particle free atmosphere scatters
light and limits visual range to about 200 miles at sea level.
Dark objects, such as distant mountains, when viewed in daytime through a particle-free
atmosphere, appear bluish because blue light is scattered preferentially into the line of
sight. Bright snow-covered mountain tops or clouds on the horizon can appear yellow to pink
because the atmosphere scatters more of the blue light from bright "targets" out of the line
of sight leaving the longer wavelength colors. The actual visual range in the particle-free
atmosphere is also limited by the earth's curvature. Thus, Rayleigh scattering is seldom the
limiting factor in the detection of the most distant objects, i.e., the visual range.
Rayleigh scattering is, however, important in reduction of visual texture and in bluish
coloration of distant dark visual targets. Moreover, air scattering is solely responsible for
the blue color of the non-horizon sky.
Fog is a naturally occurring phenomenon which can reduce the visual range to nearly zero.
It is characterized by high liquid water content, typically over 1000 ug/m , dispersed in
droplets with a mean diameter of several micrometers or more. In "natural" fogs all colors
are scattered and absorbed about equally, so the atmosphere appears white (Husar, et al.,
1979).
SOX9A/A 9-41 12-23-80
-------
194852
196064
1970-74
0.277 A
\
cn t
EXTINCTION
COEFFICIENT, km"1
VISIBILITY, miles
>0.36
< 6.6
0.3-0.36
6.6-8
0.24-0.30
810
0.18-0.24
10-13.3
< 0.18
>13.3
Figure 9-21. The spatial distribution of 5-year average extinction coefficients shows the substantial
increases of third-quarter extinction coefficients in the Carolines, Ohio River Valley, and Tennessee-
Kentucky area. In the summers of 1948-52, a 1000-km size multistate region centered around Atlanta,
GA, had visibility greater than 15 miles; visibility has declined to less than 8 miles by the 1970's. The
spatial trend of winter (first quarter) visibility shows improvements in the Northeast megalopolis
region and some worsening in the Sunbelt region. Both spring and fall quarters exhibit moderate but
detectable increases over the entire eastern United States.
Source: Husar et al. (1979).
9-42
-------
The historical frequency of occurrence of fogs in the continental United States reveals
considerable geographic variability (Figure 9-22). Coastal areas experience the highest fre-
quency. Most inland portions of the United States west of the Applachians can expect fe«er
than 20 days of fog per year, with less than five days of fog annually in the arid West.
With the exception of coastal and mountainous regions, fogs are rare during the surfer
months. Fogs tend to be localized events of, at most, a few hours duration, commonly dj^i-c
the early morning hours. On an hourly basis, fogs exist less than one percent of the time
(Conway, 1963). Thus, the overall contribution of fog to the degradation of visual air
quality is small, and it is an insignificant cause of reduced visibility during the daylight
hours.
Thunderstorms, other rainfall, and snow can also reduce visibility. East of Nevada, most
of the U.S. experiences from 30 to 50 days each year with thunderstorm activity. Such storms
are most common on summer afternoons. Since thunderstorms are usually intense but brief, they
also contribute to visibility reduction less than one percent of the time on an annual basis.
Snow is a major natural impediment to visibility. It is an important factor in man,
regions of the North and in some mountainous areas, where blowing snow occurs from 1 to 12
percent of winter hours (Conway, 1963). During the winter months snowstorms may account for
most of the hours of reduced visibility, and certainly may dominate the epidoses of extremely
low visibility in winter months.
The natural contribution of fog, thunderstorms, snow, and other forms of precipitation
can thus cause severe degradation of visual air quality. With few exceptions, however, these
intense but infrequent events do not dominate the average visual range within the continental
U.S.; typically only a small percentage of the hours involve storms or fog.
In the arid West the contribution of wind-blown dust to degradation of visual air quality
is an important problem. Because human activities which disturb natural soil surfaces add
significantly to wind-blown dust, dust storms are only partially natural.
The suspension of particles from the surface is determined by cohesiveness of the
particles to the underlying material, the force of the surface wind, and the topography of the
surface layer. The ideal situation leading to suspension of surface material is a dry,
crumpling or disturbed crust in flat terrain without vegetation. Agitation of such surfaces
by strong winds and turbulence can transform a pristine arid atmosphere into a dust storm with
severely reduced visibility. Suspended crustal material in a dust storm usually consists of
coarse solid particles with volume mean diameters of tens of micrometers (urn) or more.
Patterson et al. (1976) found that the optically important fugitive dust particles include
those up to 40 urn in diameter. Orgill and Sehmel (1976) have analyzed the frequency of
occurrence of dust storms in the continental United States in great detail, based on National
Weather Service observations of wind-blown dust and sand associated with visibility of seven
miles or less. The peak hours for dust are noon to eight p.m., during the period of maximum
SOX9A/A 9-43 12-23-SC
-------
Figure 9-22. Average annual number of days with occurrence of dense fog. Coastal
and mountainous regions are most susceptible to fog.
Source: Conway (1963).
Figure 9-23. Annual percent frequency of occurrence of wind-blown dust when
prevailing visibility was 7 miles or less, 1940-1970. Dust is a visibility problem
in the Southern Great Plains and Western desert regions.
Source: Adapted from Orgill and Sehmel (1976).
9-44
-------
thermal turbulence. Forested, coastal and mountainous regions have few, if any, episodes.
The Pacific coast has high (>0.1 percent) incidence of dust only in the San Joaquin Valley *'~
the Los Angeles Basin. Western desert areas in Eastern Washington, Western Nevada, Utah, Ne«
Mexico and Arizona are also prone to dust. The highest dust frequency is in the Southern
Great Plains, where' wind-blown dust is a serious problem up to 3 percent of the time (Figure
9-23).
9.5.2 Summary—The impact of sulfates in reducing visual air quality in individual plumes and
on the regional scale of 1000 km is no longer a matter of dispute. Considerable evidence from
chemical mass balance methods indicates that sulfates, which make up approximately 50 percent
of the fine-aerosol mass, contribute considerably more to visibility degradation than do other
chemical species. Finally, the 30-year record of the spatial and temporal trends of coal
combustion and visibility suggests that the increasing emissions of SO since the 1950's have
been associated with similar increases in haziness. This association is not as well defined
<
as that for sulfate concentrations and visibly decreases.
9.6 SOLAR RADIATION
Incoming solar radiation is composed of the direct beam and the diffuse skylight arising
from the light-scattering atmosphere (Figure 9-24; Gates, 1966). The relative contribution of
the skylight is least at noon and greatest at sunrise and sunset. At sea level, and for a
clean atmosphere, skylight contributes at least 10 percent of the total radiation.
Aerosol layers in the atmosphere scatter and absorb solar radiation (Figure 9-25). Some
of the scattered radiation is directed upwards and lost to space; some is directed downwards
to the earth's surface. Most of the solar radiation eventually reaches the surface, but its
spectral and directional composition, that is, the "quality" of the solar radiation, may be
changed by atmospheric haze (Weiss et al., 1978). A small fraction of the scattered radiation
may also be lost back to space, in which case the amount of energy reaching the surface is
reduced, contributing to the cooling at the earth's surface. A fraction of the radiation may
also be absorbed by aerosols, further reducing the amount of radiation reaching the surface
but at the same time heating the aerosol layer itself from the earth's surface.
It should be noted that because aerosols are not uniformly distributed in the atmosphere,
their effects are spatially nonhomogeneous. First, the horizontal spatial scale encompassing
aerosol source, transport, and removal in the lower troposphere is variable but often about
1000 km. The vertical spatial scale of noninfluenced aerosols is also quite variable, but
often the particles are concentrated in a layer from 500 to 2000 m deep at the earth's sur-
face. Hence, the aerosol effects should be concentrated in the lowest layers of the atmos-
phere, especially in industrial regions. %
Global-scale effects might also occur. If the effects in industrial regions are strong
enough, then the heat balance of the entire earth could be influenced. On the other hand.
effects from long-lived aerosol, for example, in the stratosphere, might lead to direct physi-
cal effects on a global scale.
SOX9A/A 9-45 12-23-80
-------
I
GLOBAL RADIATION
DIRECT SOLAR
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
WAVELENGTH, jmm
Figure 9-24. Solar radiation intensity spectrum at sea level in
cloudless sky peaks in the visible window, 0.4-0.7 jum wavelength
range, shows that in clean remote locations, direct solar radiation
contributes 90 percent and the skylight 10 percent of the incident
radiation on a horizontal surface. The airmass, m, is a measure of the
amount of air the direct solar beam has to pass through.
Source: Gates (1966).
9-46
-------
Figure 9-25. Extinction of direct solar radiation by aerosols is
depicted.
9-47
-------
9,6,1 Spectral and Directional Quality of Solar Radiation
The spectral quality of solar radiation on a clear day and on a hazy day in Texas is shown
in Figure 9-26 (McCree and Keener, 1974). On the hazy day, the direct solar radiation is
reduced to about one-half of that on a clear day, but most of the energy has reappeared as
diffuse skylight. The net effect is that there is an overall loss of up to about 10 to 20
percent of the radiation reaching the surface.
If we take the typical backscattered fraction for haze aerosols to be 10 percent of scat-
tering, and the absorption to be also about 10 percent, as suggested by the data of Weiss
(1978), then we can estimate the amount of energy lost from the surface, the amount lost to
space, and the amount absorbed by the atmosphere. On a day with half of the direct beam
transmitted, we conclude that 10 percent of the other half, or 5 percent, is lost to space,
and the other 5 percent results in atmospheric heating. Together, these phenomena lead to a
loss of 10 percent of the radiation. Although it is not possible to calculate accurately the
influence this loss might have on surface temperature, rate of thawing of frozen ground,
growing season, or other climatological measures, it is highly probable that this loss cools
the ground and heats the hazy lower layers of the atmosphere. In turn, if this occurs, it
must increase atmospheric stability, decrease convective mixing, and therefore increase the
rate at which pollutants accumulate.
No detailed and routine measurements of the quality of solar radiation are available for
the United States. However, the total solar energy reaching the surface is monitored routine-
ly at many meteorological observation sites in the United States and worldwide. Unfortunate-
ly, the large variability of such data does not allow manmade aerosol effects to be distin-
guished from other natural causes.
A data base that gives some information on the quality of solar radiation is the U.S.
turbidity network operated at about 40 stations in the country since 1961 (Flowers et a!.,
1969). If there are no clouds between the observer and the sun, the intensity of direct solar
radiation for a given solar elevation depends on the variable amount of dust, haze, and water
vapor in the atmosphere. The extinction produced by these constituents is called "atmospheric
turbidity." The seasonal pattern of atmospheric turbidity in the United States at 29 sites
for 1961-66 is shown in Figure 9-27 (Flowers et al., 1969). At all sites, the highest turbi-
dity occurs in the summertime and the lowest occurs in the winter, which is consistent with
the haziness pattern obtained from visibility observations (Section 9.4.2). However, the
turbidity of the atmosphere in the United States has a strong spatial dependence. ' In the
Southwestern States with an annual turbidity coefficient of about 0.06, the incoming direct
solar radiation is attenuated by only 13 percent (mostly scattered) compared with Midwestern
State values of about 20 percent. The highest turbidity coefficients were observed in the
Eastern United States, where summer values of 0.2 and winter values of 0.1 were typical.
This means that in the summertime about 35 percent of the direct solar beam is diverted to
SOX9A/A 9-48 12-23-80
-------
1.0 -
0.3
0.4 0.5
WAVELENGTH,
0.6
Figure 9-26. On a cloudless but hazy day in Texas, the direct solar
radiation intensity was measured to be half that on a clean day, but
most of the lost direct radiation has reappeared as skylight. However,
there is about 20 percent of the solar radiation missing on the hazy
day, some absorbed, and some backscattered to space.
Source: McCree and Keener (1974).
9-49
-------
Figure 9-27. To interpret these 1961-1966 monthly average turbidity data in terms of aersol effects on
transmission of direct sunlight use the expression I/I0 = 10'B, where B is turbidity and l/lo is the fraction
transmitted.
Source: Flowers etal. (1969).
9-50
-------
skylight, backscattered to space, or absorbed. This means that about 3.5 to 7 percent is back-
scattered to space and another 3.5 to 7 percent is absorbed into the atmosphere.
Since the first report of Flowers et al. (1969), the turbidity data have been reported
yearly by the World Meteorological Organization (WMO, 1977). Comparison of the seasonal
turbidity pattern for 1961-66 and 1972-75 is shown in Figure 9-28. Since the mid-I9601s there
has been a further increase in Eastern U.S. turbidity, particularly in the summer season.
Currently the. summer average turbidity in the region including Memphis, TN, Oak Ridge, TN,
Greensboro, NC, and Baltimore, MD, is about 0.3. This corresponds to a 50 percent attenuation
of the direct solar beam on an average summer day. During hazy episodes, turbidity coeffi-
cients of 0.6 to 1.0 are often reported, resulting in a condition in which 75 to 90 percent of
the solar radiation is removed from the direct beam, 7.5 to 18 percent is lost to space, and
7.5 to 18 percent is lost as atmospheric heating. One of the consequences of such hazy atmos-
phere is the disappearance of shadow contrasts. It is stropgly suspected but has not yet been
proved that there are effects on agricultural productivity.
The spatial distribution and trends of regional-scale turbidity in the Eastern United
States are consistent with the observed pattern of haziness obtained through visibility obser-
vations. Both the turbidity and visibility reduction by haze in the Eastern United States can
be attributed primarily to manmade fine particles (Section 9.4.3). Bolin and Charlson (1976)
suggest that many of these radiative effects are due to sulfates and conclude that the magni-
tude of effects is comparable to that summarized here.
9.6.2 Total Solar Radiation: Local to Regional Scale
Changes in the total radiant energy have been observed within urban areas. Early measure-
ments in central city locations, primarily in Europe, showed levels typically 10 to 20 percent
below surrounding rural areas. Robinson (1962) discussed some observations made in London and
in Vienna. In London the deficit was considerably reduced after the implementation of a clean
air act. Measurements on 47 days in autumn 1973 in the Los Angeles area are summarized in
Table 9-3 (Peterson and Flowers, 1977). In the St. Louis area, however, smaller urban-rural
differences were observed. On 12 cloudless days in summer 1972, the average solar and UV
fluxes at an urban site were only 3 and 8'percent, respectively, below those at a rural site
about 50 km from the city. The difference between the St. Louis and Los Angeles and European
urban areas appears to involve both decreased urban and increased rural attenuation, and it
may be that neither the city of St. Louis nor its surroundings over a wide area modify solar
radiation in a manner typical of other locations.
Recently, Angel! and Korshover (1975) analyzed the solar radiation duration data (hours
of sunshine) for the eastern half of the United States. Data for the 1950-70 time period were
obtained with on/off detectors, and these data are believed to be more reliable for long-tern
trend analysis than data from recorders of solar radiation intensity. Angel 1 and Korshover
noted some marked trends: in the Southeast and South Central United States, the solar radia-
SOX9A/A 9-51 12-23-80
-------
IU
5
E
I" I'l "I'l' I'l
MEMPHIS, TN
TURBIDITY TREND
1961-66 --»-•
1972-75 ——
OAK RIDGE, TN
I.I.I. I.I.I .1.1,1,1.1.1
GREENSBORO, NC
JFMAMJJASOND
MONTH
BALTIMORE, MD
i.l.I.I.I.1. 1.1.1 1.1.
Figure 9-28. Seasonal turbidity patterns for 1961-66 and 1972-75 are shown for selected regions in the
Eastern United States.
Source: Flowers et al. (1969).
9-52
-------
TABLE 9-3. SOME SOLAR RADIATION MEASUREMENTS IN
THE LOS ANGELES AREA3
Measurement
Minimum
Average
Maximum
Total
4
11
20
UV
15
29
44
Values for the daily average percentage decrease of total
and UV solar radiation between El Monte (urban) and
Mt. Disappointment (rural).
Source: Peterson and Flowers (1977)..
tion duration has decreased by about 4 to 6 percent; however, the North Central is increasing
(Figure 9-29; Angel 1 and Korshover, 1975). Although the authors do not attribute these trends
to any specific cause, it should be noted that there has been an increase in haziness within
that time period in areas with decreased solar radiation. It is conceivable, therefore, that
increased haziness causes sunshine-duration detectors to delay the turn-on time in the morning
and advance the turn-off time. It should also be stressed, however, that changes in the solar
radiation duration may be caused by other natural or manmade phenomena.
9.6.3 Radiative Climate: Global Scale
The attenuation of solar radiation from scattering and absorption by particles in the
atmosphere is probably an important factor in climatic change. The effect could arise from
redirecting the photon energy, from effects on cloud structure (perturbations of the colloidal
system), or from effects on optical properties, such as increased photon retention from
absorption on soot particles. A report by Hobbs et al. (1974) argues that aerosols are most
likely the principal agents by which pollutants may affect weather and climate, by influencing
the structure and distribution of clouds. On a global scale, they noted that the effects of
manmade pollutants on climate are still a matter of debate. Others, e.g., Twomey (1974), have
pointed up a direct connection between pollution and the number of drops in a cloud and, hence,
an influence on optical thickness and reflectance of the clouds (cloud albedo) and in doing
so, affect climate.
The importance of particulate matter on climate may be overshadowed by that of carbon
dioxide. A doubling of the concentration of carbon dioxide could result in a 1.5 to 3 degree
centigrade of warming the lower atmosphere, according to various predictions viewed by the
U.S. Department of Energy (U.S. Department of Energy, 1978). One set of calculations fre-
quently cited in the scientific literature predicts a 2 to 3 degree centigrade rise in the
SOX9A/A 9-53 12-Z3-80
-------
i
s
c
u.
Z
O
<
UJ
Q
as
20
18
16
14
12
10
8
6
4
2
0
-2
-4
1<
-
—
_ —
—
—
•» ^
— —
SOUTH CENTRAL
NE\\
^L..^..---""
*** \S 1 1 1 1 1 1 1 1 1 1 1 1 ii i 1 r 1 1 1 1 1 1 1 1 1 1 1 1
>40 50 60 70 80 90 20
YEAR
00
Figure 9-29. Analysis of the hours of solar radiation since the 1950's
shows a decrease of summer solar radiation over the Eastern United
States. There may be several causes for this trend, including an
increase of cloudiness; some of the change may also be due to haze.
Source: Angell and Korshover (1975).
9-54
-------
average temperature of the lower atmosphere in the middle latitudes with the doubling of the
C02 content of the air and a 3 to 4 fold greater temperature increase in the polar regions
(Manabe and Whetherald, 1975). Current model estimates suggest that the earth should ha;e
experienced a few tenths of a degree of warming since the late 1880s due to the increase of
C02 concentrations from about 290 to the current 335 ppm. However, it appears that natural
variations currently are large enough to mask the expected Co2 effect on temperature (U.S.
Department of Energy, 1980 and Stuivert, 1978).
Effects of volcanic emissions (see Chapter 4) on weather have also been suggested.
Although "spectacular" sunsets have been associated with major eruptions, for example Krakatoa
and Katmai, the effects of scattering of solar radiation from volcanic dust are not clearly
understood. Whether a universal greenhouse effect is created which can result in cooler
weather is still at issue. To date, surface temperature and rainfall changes are not conclu-
sively related to volcanic events. The large variability of weather and self-preserving
aspects of the atmospheric system tend to obscure all but the most dramatic changes in
climate.
On local scales associated with urban and industrial areas, any significant attenuation
of radiation by air pollution can, in addition to other we 11-recognized factors, result in
changes in local weather (e.g., Landsberg, 1970). It is possible that local- and regional-
scale changes in solar radiation caused by human activity may ultimately influence the heat
and water vapor contents of the atmosphere on very large scales, but solar radiation and
aerosol levels measured at stations remote from pollutant sources have not as yet displayed
any trend that can be related to human causes (Fischer, 1967; Ellis and Pueschel, 1971; Hodge
et al.f 1972).
Unfortunately, there is little agreement about whether the net effect of increased air-
borne particulate concentrations is the warming or cooling of the earth as a whole. Most
models-can predict either an increase or a decrease in the effective albedo of the earth under
cloudless, skies, depending on which combination of surface albedo, sun angle, particle size
distribution, and particle refractive index is assumed. The effects of clouds are very impor-
tant, and the contributions from infrared radiation must be considered in order to obtain a
t *
complete energy budget (Wesely and Lipschutz, 1976).
9.7 CLOUDINESS AND PRECIPITATION
The global cloud cover plays a vital role in the earth's radiative budget in reflecting
energy back,to space, in absorbing both solar and longwave (terrestrial) radiation, and in
emitting its energy downward and outward into space. Changes in cloud cover, therefore, alter
the global heat balance. Cloud- and precipitation-forming^processes may be divided into two
broad classes: (1) macrophysical processes, which affect the rise and descent of air currents
and the amount of water vapor available for condensation; and (2) microphysical processes,
which affect the inature of cloud particles formed during condensation. The role of atmos-
SOX9A/A 9-55 12-23-80
-------
pheric aerosols, primarily those that are strongly hygroscopic, is to influence the micro-
physics of cloud formation.
On a global or even regional scale, the very small amounts of moisture that man adds by
land practices or combustion of fossil fuels are negligible in comparison with global evapora-
tion. On a regional scale, only one form of increasing cloudiness suggests itself: the
formation of aircraft contrails (Machta and Telegadas, 1974). Aircraft contrail formation
results mainly from the injection of water vapor rather than of aerosols.
In urban areas, inadvertent changes of cloudiness as well as the quantity of precipita-
tion have been well established. Such urban impacts also include the frequencies of thunder-
storms and hail as well as total amounts of rain. In a classical study, Changnon (1968) has
reported a notable increase in days of precipitation, thunderstorms, and hail occurring since
1925 at La Porte, IN. Since La Porte is 30 miles east of the Chicago urban-industrial com-
plex, he proposed that the increased precipitation results from inadvertent manmade modifi-
cations. Figure 9-30 (Changnon, 1968) shows the 5-year running totals of days with smoke and
haze restricted visibility in Chicago. This measure of atmospheric pollution has a temporal
distribution after 1930 rather similar to the La Porte precipitation curve. A noticeable
increase in smoke-haze days began in 1935 and became more marked after 1940, when the La Porte
precipitation curve began its sharp increase.
A recent report by Clark (1979) critically reexamined this anomoly historically and tech-
nically through the use of a budgetary hydrologic model. Clark proposed several explanations
of the anomoly. If the anomoly was real, it was, at most, a microscale phenomenon), and its
disappearance by the mid-I9601s arose from the dissipation of the precipitation-producing
mechanism. He noted that although increases in local precipitation have been reported down-
wind of urban areas, no records exist of any alteration equalling the magnitude of the La Port
precipitation variation relative to surrounding sites. Additionally, the 1929 to 1963 period
coincides roughly with the tenure of the sole observer. The accuracy of the guage was also at
issue. Clark concluded that errors by the observer and/or guage most likely explain the
anomaly.
As part of project Metromex, studies by the Illinois State Water Survey suggest increases
of about 30 percent in rain and 200 percent in thunderstorms and hail at single gauging sta-
tions downwind of St. Louis, with increases of about 10 percent over a two-county area. Here
again, the physical causes of the maxima are not well understood, but they do appear to be
associated with perturbations of the planetary boundary layer and enhanced cloudiness, possi-
bly resulting from the addition of aerosols. It is regrettable that the complex interactions
of cloud- and precipitation-forming processes obscure the specific role of manmade aerosols.
The incorporation of particles into rain and fog droplets can change the "quality" of
precipitation by changing its chemical composition. The most important impact on precipita-
tion water quality is probably that of "acid rain," discussed in more detail in chapter 8.
SOX9A/A 9-56 12-23-80
-------
300
CHICAGO!
SMOKE-I -
HAZE DAYS
OBSERVER j
CHANGES AT /
LA PORTE
1910 1920 1930 1940 1950 1960
I
ENDING YEAR OF 5-YEAR PERIOD
Figure 9-30. Numbers of smoke/haze days are plotted per 5 years at
Chicago, with values plotted at end of 5-year period.
Source: Changnon (1968).
9-57
-------
9.S SUW1ARY
Traditionally, visibility has been defined in terms of the distance from an object that
is necessary to produce a minimum detectable contrast between that object and its background.
Although visibility is often defined in terms of this "visual range," with its reasonably pre-
cise definition, visibility includes more than being able to see or not see a target. Visibility
eludes seeing targets at shorter distances than the visual range and being able to appreciate
the details of the target, including colors. Visibility impairment can manifest itself in two
ways: (1) a layer of haze (or a plume), which is visible because it has a visual discontin-
uity between itself and its background or (2) a uniform haze which reduces atmospheric clarity.
The type and degree of impairment are determined by the distribution, concentrations, and
characteristics of atmospheric particles and gases, which scatter and absorb light traveling
through the atmosphere. Scattering and absorption seem to determine extinction.
The consequences of visibility degradation range from the deterioration or loss of the
aesthetic perception of scenic vistas to annoyances or outright hazards to automobile and
aircraft traffic.
On a regional scale, the extinction of light is generally dominated by particle
scattering. In urban areas, absorption by particles becomes important. Extinction by
particles is usually dominated by particles of diameter 0.1 to 2 urn ("fine" particles).
Extinction due to scattering is closely proportional to the fine-particle mass concentration,
_r _1 O
with extinction/mass ratios in the range of 3-5 x 10 m /(pg/m ).
Current knowledge indicates that fine aerosol is composed of varying amounts of sulfate,
ammonium, and nitrate ions, elemental carbon, organic carbon compounds, water, and smaller
amounts of soil dust, lead compounds, and trace species. Sulfate often predominates the fine
mass and visibility impairment, while elemental carbon is often the primary visibility-
impairing specie in urban areas. Significant variations can occur at different times and sites.
Our knowledge of the roles of several possibly important species is hindered by the lack of
sufficient good data. The 30-year record of the spatial and temporal trends of coal combus-
tion and visibility suggest that the increasing emissions of SO since the 1950s have been
J\
associated with similar increases in haziness. Nevertheless, the relationship between SOg
emissions and resulting sulfate concentrations is not as well defined as the relationships
between sulfate concentrations and visibility reduction.
The currently available visibility monitoring methods measure different aspects of visi-
bility impairment. Generally, contrast type measurements (such as photography, telephotq-
metry, and human eye observations) relate well to the perception of visual air quality, while
extinction or scattering measurements (such as transmissometry and nephelometry) relate to the
cause of visibility degradation. Each of the above measurement methods can be used to
approximate visual range. No single method has been proven totally effective in measuring
light absorption. The methods used thus far include determining the difference between ex-
tinction and scattering, several filtering methods, and a refractive index method.
SOX9A/A 9-58 12-23-80
-------
Pollutants released to the atmosphere alter the environment in ways other than visibility
reduction. They may lead to slow and subtle changes in the nature of the atmosphere ana,
possibly, in climate. For example, a fraction of the solar radiation may be absorbed by
aerosols, further reducing the amount of radiation reaching the earth's surface and, at the
same time, heating the aerosol layer itself. On a hazy day, the direct solar radiation is
reduced to about one-half of that on a clear day, but most of the energy reappears as diffuse
skylight. However, there is an overall loss of up to about 10 to 20 percent of the radiation
reaching the surface.
If there are no clouds between the observer and the sun, the intensity of direct solar
radiation for a given solar elevation depends on the variable amount of dust, haze, and water
vapor in the atmosphere. The extinction produced by these constituents is called "atmospheric
turbidity." during hazy episodes, turbidity coefficients of 0.6 to 1.0 are often reported,
resulting in a condition in which 75 to 90 percent of the* solar radiation is removed from the
direct beam, 7.5 to 18 percent is lost to space 7.5 to 18 percent is lost as atmospheric
heating. One of the consequences of such a hazy atmosphere is the disappearance of shadow
contrast.
The attenuation of solar radiation from scattering and absorption by particles in the
atmosphere is probably an important factor in climatic change on all scales. On local scales
associated with urban and industrial areas, any significant attenuation of radiation by air
pollution can, in addition to other well-recognized factors, result in changes in local weather.
It is possible that local- and regional-scale changes in solar radiation caused by human
activity may ultimately influence the heat and water vapor contents of the atmosphere on very
large scales, but solar radiation and aerosol levels measured at stations remote from pol-
lutant sources have not as yet displayed any trend that can be related to human causes.
Cloud- and precipitation-forming processes may be divided into two broad classes: macro-
physical and raicrophysical processes. Macrophysical processes involve the rise and descent of
air masses and the amount of water vapor available for condensation. Atmospheric aerosols,
primarily those that are strongly hygroscopic, influence the microphysics of cloud formation.
The incorporation of particles into rain and fog droplets can change the "quality" of pre-
cipitation by changing its chemical composition. However, the complex interactions of cloud-
and precipitation-forming processes obscure the specific role of manmade aerosols.
SOX9A/A 9-59 12-23-80
-------
9.9 REFERENCES
Allen, J., R. B. Husar, and E. S. Macias. Aerosol size and the shape determination Using a
laser light scattering spectrometer. In: Aerosol Measurement, a Workshop, University of
Florida, Gainesville, Florida, March 24-26, 1976. D. A. Lundgren, M. Lippmann, F. S.
Harris, Jr., W. E. Clark, W. H. Marlow, and M. D. Durham, eds., University Presses of
Florida, Gainesville, FL, 1979. pp.312-320.
Altshuller, A. P. Atmospheric sulfur dioxide and sulfate, distribution of concentration at
urban and nonurban sites in the United States. Environ. Sci. Technol. 7:709-713, 1973.
Angell, J. K., and J. Korshover. Variation in sunshine duration over the contiguous United
States between 1950 and 1972. J. Appl. Meteorol. 14: 1174-1181, 1975.
Appel, B. R., S. M. Wall, Y. Tokiwa, and M. Haik. Interference effects in sampling parti-
culate nitrate in ambient air. Atmos. Environ. 13:319-325, 1979.
Barnes, R. A., and D. 0. Lee. Visibility in London and the long distance transport of
atmospheric sulfur. Atmos. Environ. 12:791-794, 1978.
Bergstrom, R. W. Extinction and absorption coefficients of the atmospheric aerosol as a
function of particle size. Beitr. Phys. Atmos. 4§:223-224, 1973.
Bohn, P. An approach to the problem of estimating demand for public good. Swed. J. Econ. 73:
1971.
Bolin, B., and R. J. Charlson. On the role of the tropospheric sulfur cycle in the shortwave
radiative climate of the earth. Ambio Ei:47-54, 1976.
Bradford, D. Benefit-cost analysis and demand curves for public goods. Kyklos 23: 1970.
Brookshire, D. S. Issues in valuing visibility: an overview. In: Proceedings of the Work-
shop in Visibility Values. D. Fox, R. J. Loomis and T. C. Green, technical coordinators,
U.S. Department of Agriculture, Fort Collins, CO, 1979. pp. 130-138.
Brookshire, D., B. C. Ives, and W. D. Schulze. The valuation of aesthetic preferences. J.
Environ. Manag. 3:325-346, 1976.
Cass, G. R. On the relationship between sulfate air quality and visibility with examples in
Los Angeles. Atmos. Environ. 13:1069-1084, 1979.
Chandrasekhar, S. Radiative Transfer. Dover Publishers, New York, NY, 1960.
Changnon, S. A., Jr. The La Porte weather anomaly—fact or fiction? Bull. Am. Meteorol. Soc.
49:4-11, 1968.
Charlson, R. J., .A. H. Vanderpol, D. S. Covert, A. P. Waggoner, and N. C. Ahlquist.
H9SOA/(NH..)9SO, Background aerosol: optical detection in St. Louis region. Atmos.
Environ. 8:1257-1268, 1974.
Charlson, R. J., A. P. Waggoner, J. F. Thielke. Visibility Protection for Class I Areas: the
Technical Basis. Council on Environmental Quality, Washington, DC, August 1978.
Charlson, R. J., D. S. Covert, T. V. Larson, and A. P. Waggoner. Chemical properties of
tropospheric sulfur aerosols. In: Sulfur in the Atmosphere, Proceedings of the Inter-
national Symposium, United Nations, Dubrovnik, Yugoslavia, September 7-14, 1977. Atmos.
Environ. 12:39-53, 1978.
SOX9/E 9-60 12-20-80
-------
Char!son, 1978.
Charlson, R. J., D. S. Covert, Y. Tokiwa, and P. K. Mueller. Multiwavelength nephelometer
measurements in Los Angeles, Smog Aerosol, III. Comparison to light extinction by NCL
J. Coll. Int. Sci. 39:260-265, 1972. 2
Clark, R. R. A hydrologic reanalysis of the La Port anomoly. Bull. Am. Meteorol. Soc. 60:
1979. —
Conway, H. M., ed. The Weather Handbook. Conway Research, Inc., Atlanta, GA, 1963.
Cornsweet, T. N. Visual Perception. Academic Press, Inc., New York, NY, 1970.
Covert, D. S., A. P. Waggoner, R. E. Weiss, N. C. Ahlquist, and R. J. Charlson. Atmospheric
aerosols, humidity, and visibility. In: The Character and Origins of Smog Aerosols.
Adv. Environ. Sci. Technol. 9:574, 1980.
Dave, J. V. Scattering of electromagnetic radiation by a large, absorbing sphere. IBM J. Res.
Develop. 13:302-313, 1969.
Davis, R. K. Recreational planning as an economic problem. Nat. Resour. J. 3: 1963.
Duntley, S. Q., A. R. Boileau, and R. W. Preisendorfer. Image transmission by the troposphere
I. J. Opt. Soc. Am. 47:499, 1957.
Eggleton, A. E. J., and D. H. F. Atkins. The chemical composition of atmospheric aerosols on
Teesside and its relation to visibility. Atmos. Environ. 3:355-372, 1969.
Eiden, R. Determination of the complex index of refraction of spherical aerosol particles.
Appl. Opt. 10:749-754, 1971.
Ellestad, T. G. Aerosol composition of urban plumes passing over a rural monitoring site.
Ann. New York Acad. Sci. 338:202-218, 1980.
Ellestad, T. G., and R. E. Speer. Application of a telephotometer to visibility measurements
in the eastern United States. Presented at Symposium on Plumes and Visibility, Grand
Canyon, AZ, November 1980.
Ellis, A. T., and R. F. Pueschel. Absence of air pollution trends at Mouana Loa. Science
(Washington, D.C.) 172:845-846, 1971.
Ensor, 1972. '
Ensqr, D. S., R. J. Charlson, N. C. Ahlquist, K. T. Whitby, R. B. Husar, and B. Y. H. Liu.
Milltiwavelength Nephelometer Measurements in Los Angeles Smog Aerosol. I. Comparison of
calculated and measured light scattering. J. Coll. Int. Sci. 39:242-251, 1972.
Ensor, D., and A. P. Waggoner. Angular truncation error in the integrating nephelometer.
1 Atmos. Environ. 4:481-487, 1970.
Faxvog, F. R. Optical scattering per unit mass of single particles. Appl. Opt. 14:269-270,
1975.
Faxvog, F. R., and D. M. Roessler.' Carbon aerosol visibility vs. particle size distribution.
. Appl. Opt. 17:2612-2616, 1978.
Fischer, K. Mass absorption coefficients of natural aerosols in the 0.4-2.4 mm wavelength
interval. Beitr. Physik de Atmosphiire 46, 1973.
i
S6X9/E 9-61 12-20-80
-------
Fischer, W. H. Some atmospheric turbidity measurements in Antarctica. J. Appl. Meteorol.
6:958-959, 1967.
Fischer, 1975.
Flowers, E. C., R. A. McCormick, and K. R. Kurfis. Atmospheric turbidity over the United
States, 1961-1966. J. Appl. Meteorol. 8:955-962, 1969.
Fowler, B. W., and C. C. Sung. Scattering of an electromagnetic wave from dielectric bodies
of irregular shape. J. Opt. Soc. Am. 69:756-761, 1979.
Fox, D., R. J. Loomis, and T. C. Green. Proceedings of the workshop in Visibility Values, Ft.
Collins, CO, January 18-February 1, 1979. USDA Forest Service, Washington, DC.
Garland, J. A. Condensation on ammonium sulfate particles and its effect on visibility. Atm.
Environ. 3:347-354, 1969.
Gates, D. M. Spectral distribution of solar radiation at the Earth's surface. Science
(Washington, D.C.) 151:523-529, 1966.
Green, D. M., and F. W. Campbell. Optical and retinal factors affecting visual resolution.
J. Physio!. 181:576-593, 1965.
Groblicki, P. J., G. T. Wolff, and R. J. Countess. Visibility Reducing Species in the Denver
Brown Cloud. Part One: Relationships Between Extinction and Chemical Composition.
Presented at the Symposium on Plumes and Visibility Measurements and Model Components,
Grand Canyon, AZ, November 10-14, 1980.
Grosjean, D. et al. Concentration, size, distribution and modes of formulation of particulate
nitrate, sulfate and ammonium compounds in the eastern part of the Los Angeles air basin.
Presented at the Annual Meeting of the Air Pollution Control Association, 1976. Paper
No. 76-20.3.
Hall, J. S., and L. A. Riley. Basic spectrophotometric measures of air quality over long paths.
In: Radiative Transfer and Thermal Control, vol. 49: Progress in Astronautics and
Aeronautics. A. M. Smith, ed., American Institute of Aeronautics and Astronautics, New
York, NY, 1976. pp. 205-212.
Hansen, A. D. A., H. Rosen, R. L. Dod, and T. Novakov. Optical attenuation as a tracer for
the primary component of the carbonaceous aerosol. In: Proceedings: Carbonaceous
Particles in the Atmosphere, Lawrence Berkeley Laboratory and National Science Foundation,
Berkeley, California, March 20-22, 1978. T. Novakov, ed., LBL-9037, University of
California, Berkeley, CA, June 1979.
Heisler, S. L., R. C. Henry, and J. G. Watson. The sources of the Denver Haze in November and
December of 1978. Paper No. 80-58.6, Annual Meeting of the Air Pollution Control Associa-
tion, 1980.
Henry, R. C. Psychophysics and Visibility Values. Proceedings of the Workshop on Visibility
Values, D. Fox, R. J. Loomis, and T. C. Greene (technical coordinators). Fort Collins,
Colorado. U.S. Department of Agriculture, 1979. jj
Henry, R. C. The application of the linear system theory of visual acuity to visibility
reduction by aerosols. Atraos. Environ. 11:697-701, 1977.
Higgins, I. T. T., et al. Airborne Particles, Washington, DC: National Academy of Sciences,
1977.
SOX9/E 9-62 12-20-80
-------
Hobbs, P. V., H. Harrison, and E. Robinson. Atmospheric effects of pollutants. Science
(Washington, DC) 183:909, 1974.
Hodge, P. W., N. Laulainen, and R. J. Charlson. Astronomy and air pollution. Science
(Washington, DC) 178:1123-1124, 1972.
Hodkinson, J. R. Calculations of colour and visibility in urban atmospheres polluted by
gaseous NOg. Air and Wat. Pollut. Int. J. 10:137-144, 1966.
Horvath, H., and K. E. Noll. The relationship between atmospheric light scattering co-
efficient and visibility. Atmos. Environ. 3:543-550, 1969.
Husar, R. B., and D. E. Patterson. Regional scale air pollution: sources and effects.
Annals of the NY Academy of Sciences 338:399-417, 1980.
Husar, R. B., D. E. Patterson, J. M. Holloway, W. E. Wilson, and T. G. Ellestad. Trends of
eastern U.S. haziness since 1948. J_n: Proceedings of the Fourth Symposium on Atmospheric
Turbulence, Diffusion, and Air Pollution, American Meteorological Society, Reno, NE, 1979.
pp. 249-256.
Koschmieder, H. Theoric der horizontalen Sichtweite. Beitr. Phys. Freien Atm. 12:33-53,
171-181, 1924. ~~
Landsberg, H. E. Man-made climatic changes. Science (Washington, DC) 170:1265-1274, 1970.
Latimer, D. A., R. W. Bergstrom, S. R. Hayes, M. K. Liu, J. H. Seinfeld, G. Z. Whitten, M. A.
Wojcik, and M. J. Hi 1Iyer. The Development of Mathematical Models for the Prediction of
Anthropogenic Visibility Impairment. EPA-450/3/78-110a, U.S. Environmental Protection
Agency, Research Triangle Park, NC, September 1978.
Leaderer, B. P., D. M. Bernstein, J. M. Daisey, M. T. Kleinman, T. J. Kneip, E. 0. Knutson, M.
Lippmann, P. J. Lioy, K. A. Rahney, D. Sinclair, R. L. Tanner, G. T. Wolff. Summary
R. J. Charlson. Absorption coefficient of atmospheric aerosol: a
method for measurement. App. Opt. 12:1356-1363, 1973.
Lindberg, J. D., and L. S. Lande. Measurement of the absorption coefficient of atmospheric
dust. Appl. Opt. 13:1923-1927, 1974.
Machta, L., and K. Telegadas. Inadvertent Large-scale Weather Modification. In: Weather and
Climate Modification. W. N. Hess, ed., John Wiley &, Sons, New York, NY, 1974. pp.
687-725.
Macias et al., 1975. !
Macias, E. S., and R. B. Husar. A review of atmospheric particulate mass measurement via the
beta attenuation technique. In: Fine Particles, B. Y. H. Liu, ed., Academic Press, New
York, NY, 1976.
Macias, E. S., D. L. Blumenthal, J. A. Anderson, and B. K. Cantrell. Characterization of
visibility. Reducing aerosols in the southwestern United States; interim report on Project
VISTTA. MRI Report 78 IR 15-85, Midwest Research Institute, Kansas City, MO, 1978.
Malm, W. Visibility: A Physical Perspective. In: Proceedings of the Workshop in Visibility
Values, D. Fox, R. J. Loomis, and T. C. Greene (technical coordinators). Fort Collins,
Colorado, U.S. Department of Agriculture, 1979. pp. 56-68.
Malm, W. C. Considerations in the measurement of visibility. J. Air Pollut. Control Assoc.
29:1042-1052, 1979.
SOX9/E ,9-63 12-20-80
-------
Malm, W. C., E. G. Walther, K. E. O'Dell, and M. Kleine. Visibility in the southwestern
United States from Summer 1978 through Spring 1979. In: Proceedings Symposium on
Visibility, Grand Canyon, No. 3, 1980c.
Malm, W. C., K. K. Kelly, and J. V. Molenar. Human perception of visual air quality (regional
haze). In: Proceedings Symposium on Visibility, Grand Canyon, N. 33, 1980b.
Malm, W. C., K. K. Leiker, and J. V. Molenar. Human perception of visual air quality. J. Air
Pollut. Control Assoc. 30:122-131, 1980a.
Manabe, F., and R. Whetherald. The effects of doubling the C02 concentration on the climate
of a general circulation model. J. Atraos. Sci. 32:315, 1975.
Marians, M., and J. Trijonis. Empirical Studies of the Relationship Between Emissions and
Visibility in the Southwest. Prepared under grant 802015 by Technology Service Corp.
U.S. Environmental Protection Agency, Cincinnati, OH, 1979.
McCree, K. J., and M. E. Keener. Effect of atmospheric turbidity on the photosynthetic rates
of leaves. Agric. Meteorol. 13:349-357, 1974.
Middleton, W. E. K. Vision Through the Atmosphere. University of Toronto Press, Toronto,
Canada, 1952.
Mie, G. Contributions of optics to cloudy media, especially colloidal metal works. Am. Phys.
25:377-445, 1908.
Mulu, et al., 1980a.
Mungnai, A., and W. J. Wiscombe. Scattering of radiation by moderately nonspherical particles.
J. Atm. Sci. 37:1291-1307, 1980.
National Park Service. My Eyes Need a Good Stretching, Seven Authorities Speak Out on Visi-
bility — Clean Air and Unique Natural Areas. U.S. Department of the Interior, Washing-
ton, DC, 1978.
Orgill, M. M., and G. A. Sehmel. Frequency and diurnal variation of dust storms in the con-
tiguous United States. Atmos. Environ. 10:813-825, 1976.
Orr, C., F. K. Hurd, and W. J. Corbett. Aerosol size and relative humidity. J. Colloid Int.
Sci. 13:472-482, 1958.
Patel, A. S. Spatial resolution by the human visual system. The effect of mean retinal
illuminance. J. Opt. Soc. Am. 56:689-694, 1966.
Patterson, E. M., and D. A. Gillette. Measurements of visibility vs. mass-concentration for
airborne soil particles. Atmos. Environ. 11:193-196, 1977.
Patterson, R. K., and J. Wagman. Mass and composition of an urban aerosol as a function of
particle size for several visibility levels. J. Aerosol Sci. 8:269-279, 1977.
Penndorf, R. Tables of the refractive index for standard air and the RayTeigh scattering co-
efficient for the spectral region between 0.2 and 20.0 m and their application to atmos-
pheric optics. J. Opt. Soc. Am. 47:176-182, 1957. •
Peterson, J. T., and E. C. Flowers. Interactions between air pollution and solar radiation.
Sol. Energy 19:23-32, 1977.
SOX9/E 9-64 12-20-80
-------
Pierson, W. R., and P. B. Russell. Aerosol carbon in the Denver area in November 1973 ,>-
Environ. 3:347-354, 1969.
Pierson, W. R., W. W. Brachaczek, T. J. Truex, J. W. Butler, and T. J. Korniski. Ambient
sulfate measurements on Alleghany Mountains and the question of atmospheric sulfate in
the northeastern United States. Ann. New York Acad. Sci. 338:145-173, 1980.
Pinnick, R. G., D. E. Carroll, and D. J. Hofmann. Polarized light scattered from monodisperse
randomly oriented nonspherical aerosol particles: measurements. Appl. Opt. 15-384-393
1976. —
Pitchford, M. L., R. E. Flocchini, R. G. Draftz, T. A. Cahill, L. L. Ashbaugh, R. A. Eldred.
Silicon in submicron particles in Southwest. Atmos. Environ., in press, 1980.
Pueschel, R. , and D. L. Wollman. On the nature of atmospheric background aerosol. Presented
at the 14th Conference on Agricultural and Forest Meteorology, Minneapolis, MN, April
1978.
Rabinoff, R., and B. Herman. Effect of aerosol size distribution on the accuracy of the in-
tegrating nephelometer. J. Appl. Meteorol. 12:184-186', 1973.
Randall, A., B. Ives, C. Eastman. Bidding games for valuation of aesthethic environmental
improvements. J. Environ. Econ. Manag. 1:132-149, 1974.
Robinson, G. D. Absorption of solar radiation by atmospheric aerosol, as revealed by measure-
ments at the ground. Arch. Meteorol. Geophys. Bioclimatol. Ser. B 12:19, 1962.
Robinson, G. D. Inadvertent Weather Modification Workshop. Final Report to the MAS; The Center
for the Environment and Man (CEM) Report 4215-604, 1977.
Rodhe, H., C. Persson, and 0. Akesson. An investigation into regional transport of soot and
sulfate aerosols. Atmos. Environ. 6:675-693, 1972.
Roessler, D. M., and F. R. Faxvog. Optical properties of agglomerated acetylene smoke parti-
cles at 0.5145 mm and 10.6 mm wavelengths. J. Opt. Soc. Am. 70:230-235, 1980.
Rosen, H., A. D. A. Hansen, R. L. Dod, and T. Novakov. Soot in urban atmospheres: deter-
mination by an optical absorption technique. Science (Washington, DC) 208:741-744, 1980.
Rowe, R. D., R. C. d'Arge, and D. S. Brookshire. An experiment on the economic value of visi-
bility. J. Environ. Econ. Manag. 7:1-19, 1980.
Samuels, H. J., S. A. Twiss, E. W. Wong, H.1 Wong-Woo, P. K. Mueller, J. A. Maga, and 6. S. Taylor.
Visibility, Light Scattering and Mass Concentration of Particulate Matter. Report of the
California Tri-City Aerosol Sampling Project. State of California Air, Resources Board,
Sacramento, CA, July 1973.
Schade, 0. H. Optical and photoelectric analog of the eye. J. Opt. Soc. Am. 46:721-739, 1956.
Schober, H. A. W..'and R. Hilz. Contrast sensitivity of the human eye for square-wave grating.
J. Opt. Soc. Am. 55:1086-1091, 1965.
Spicer, C. W., and P. M. Schumacher. Interferences in sampling atmospheric particulate nitrate.
' « Aim. Environ. 11:873-876, 1977.
Spicer, C. W., and P. M. Schumacher. Particulate nitrate: laboratory and field studies of
major sampling interferences. Atm. Environ. 13:543-552, 1979.
&X9/E '-SJ 9-65 12-20-80
-------
Stevens, R. K., T. G. Dzubay, G. Russworm, and D. Rickel. Sampling and analysis of atmos-
pheric sulfates and related species. Atm. Environ. 12:55-68, 1978.
Stevens, R. K., T. G. Dzubay, R. W. Shaw, W. A. McClenny, C. W. Lewis, and W. E. Wilson.
Characterization of aerosol in the Great Smoky Mountains. Accepted for publication in
Environ. Sci. Tech., 1980.
Stuiverts, M. Atmospheric carbon dioxide and carbon reservoir changes. Science (Washington,
DC) 199:253-258, 1980.
Tang, I. N., H. R. Munkelwitz, and J. G. Davis. Aerosol growth studies, IV. Phase trans-
formation of mixed salt aerosols in a moist atmosphere. J. Aerosol Sci. 9:505-511, 1978.
Tanner, R. L., R. Garber, W. Marlow, B. P. Leaderer.and M. A. Leyko. Chemical composition of
sulfate as a function of particle size in New York summer aerosol. Ann. New York Acad.
Sci. 322:99-113, 1979.
Tombach, I. H., and M. W. Chan. Physical, chemical, and radiological characterization of
background particulate matter in northeastern Utah. Presented at the Annual Meeting of
the Air Pollution Control Association, 1977. Paper No. 77-48.6.
Tombach, I., and D. Allard. Intercomparison of visibility measurement methods. J. Air Pollut.
Control Assoc. 30:134-142, 1980.
Trijonis, J., and R. Shapland. Existing Visibility Levels in the U.S.: Isopleth Maps of Visi-
bility in Suburban/Nonurban Areas During 1974-1976. EPA450/5-79-010, U.S. Environmental
Protection Agency, Research Triangle Park, NC, 1979.
Trijonis, J., K. Yuan, and R. B. Husar. Visibility in the Southwest: An Exploration of the
Historical Data Base. EPA-600/3/78/039, U.S. Environmental Protection Agency, Research
Triangle Park, NC, April 1978a.
Trijonis, J., K. Yuan, and R. B. Husar. Visibility in the Northeast: Long Term Visibility
Trends and Visibility/Pollutant Relationships. EPA-600/3-78-075, U.S. Environmental
Protection Agency, Research Triangle Park, NC, August 1978b.
Truex, T. J., and J. E. Anderson. Mass monitoring of carbonaceous aerosols with a spectrophone.
Atm. Environ. 13:507-509, 1979.
Twomey, S. Pollution and the planetary albedo. Atmos. Environ. 8:1251, 1974.
U.S. Bureau of Mines. Minerals Yearbook, Annual Publication, 1933-74. U.S. Department of the
Interior, Washington, D.C.
U.S. Department of Energy. Carbon Dioxide Effects, Research, and Assessment Program. A com-
prehensive plan for carbon dioxide effects, research, and assessment, Part I: The Global
Carbon Cycle and Climatic Effects of Increasing Carbon Dioxide. DOE/EV-0094, U.S. Depart-
ment of Energy, Washington, DC, August 1980.
U.S. Department of Energy. Office of Carbon Dioxide Effects, Research, and Assessment.! A(Com-
prehensive Plan for Carbon Dioxide Effects, Research, and Assessment. U.S. Department of
Energy, Washington, DC, 1978.
U.S. Environmental Protection Agency. EPA Report to Congress: Protecting Visibility. EPA-450/
5-79-008, U. S. Environmental Protection Agency, Research Triangle Park, NC, 1979.
Van Nes, F. L., and M. A. Bouman. Spatial modulation transfer in the human eye. J. Opt. Soc.
Am. 57:401-406, 1967.
SOX9/E 9-66 12-20-80
-------
Waggoner, A. P., A. J. Vanderpol, R. J. Charlson, S. Larsen, L. Granat, and C. Tragardh.
Sulphate-light scattering ratio as an index of the role of sulphur in troposphen>
optics. Nature (London) 261:120-122, 1976.
Waggoner, A. P., and R. E. Weiss. Comparisons of fine particle mass concentration and light
scattering extinction in ambient aerosol. Atmos. Environ. 14:623-626, 1980.
Waggoner, A. P., and R. J. Charlson. Measurement of aerosol optical parameters. _In: Fine
Particles, B. Y. H. Liu, ed., Academic Press, New York, 1976.
Waggoner, A. P., et al. Optical absorption by atmospheric aerosols. Appl. Opt. 12:896, 1973.
Waggoner, 1980.
Waggoner, A. P., R. E. Weiss, N. C. Ahlquist, D. S. Covert, and R. J. Charlson. Optical Charac-
teristics of Atmospheric Aerosols. Presented at: Symposium on Plumes and Visibility
Measurements and Model Components, Grand Canyon, Arizona, November 10-14, 1980.
Weiss, R. E. Ph. D. Dissertation, 1978.
t
Weiss, R. E. et al. Studies of the optical, physical, and chemical properties of light-absorb-
ing aerosols. In: Proceedings of the Conference on Carbonaceous Particles in the
Atmosphere, 1978. p. 257.
Weiss, R. E., A. P. Waggoner, R. J. Charlson, and N. C. Ahlquist. Sulfate aerosol: its geo-
graphical extent in the midwestern and southern United States. Science 195:979-980, 1977.
Wesely, M. L., and R. C. Lipschutz. An experimental study of the effects of aerosols on diffuse
and direct solar radiation received during the summer near Chicago. Atmos. Environ.
10:981-987, 1976.
White, W. H., and D. E. Patterson. Nitrogen dioxide, particles, and the color of smoke
plumes. Presented at Symposium on Plumes and Visibility, Grand Canyon, AZ, November 1980.
White, W. H., and P. T. Roberts. On the nature and origins of visibility reducing aerosols in
the Los Angeles Air Basin. Atmos. Environ. 11:803-812, 1977.
Wiscombe, W. J. Improved Mie scattering algorithms. App. Opt. 19:1505-1509, 1980.
WMO/EPA/NOAA/UNEP. Global Monitoring of the Environment for Selected Atmospheric Constituents
1975. Environmental Data Service National Climatic Center, Asheville, NC, 1977.
Wolff, G. T., P. J. Groblicki, S. H. Cadle, and R. J. Countess. Particulate carbon at various
locations in the United States. Presented at the General Motors Symposium on Particulate
Carbon, Warren, MI, October 1980.
SOX9/E 9-67 12-20-80
-------
10. EFFECTS ON MATERIALS
10.1 INTRODUCTION
Beginning with associative observations in the 17th century and continuing through modern
analytical investigation, scientists have gathered evidence that air pollutants damage man-made
and natural materials. Pollutant-related damage may lead to the need for measures such as
increased maintenance, use of protective coatings, replacement of materials, or the control of
emissions. The decision to pursue one course rather than another is guided, ultimately, by
weighing the damage against the costs of prevention. Neither damage nor prevention cost is
necessarily measured in monetary units, especially with regard to works of art or items of
historical significance. However, for better or worse, society tends to assign monetary units
to value gained or lost. Thus, the objective of this report is to present and evaluate studies
useful for estimating the dollar cost of material damage caused by sulfur oxides and particulate
matter.
The relation between pollutant emissions and economic damage is depicted in Figure 10-1.
As shown in the figure, one may (1) proceed from ambient pollutant levels to economic damage
estimates directly or (2) follow a route involving estimate of damage based on physical damage
functions. The latter route, which we all call the damage function approach, has been the
method of choice in the past. More recent studies have employed the first route. The two
alternative routes have an element in common. This element is estimation of willingness to
pay-
Economic damage (benefit) as a result of increased (decreased) pollutant concentrations
can be estimated by willingness to pay approaches. All willingness to pay approaches attempt
to estimate the aggregate monetary values that all affected individuals assign to the effects
of a change in pollutant concentration. Willingness to pay approaches can be divided into
three classes: damage function approaches, non-market approaches, and indirect market
approaches. Damage function approaches utilize the relationship of pollutant exposure to
physical damage as a first step. The second step links the physical damage to a dollar
estimate of willingness to pay. With proper consideration of substitution possibilities this
damage function approach can be a good estimate of willingness to pay. Non-market approaches
generally utilize surveys which attempt to ascertain the monetary values assigned to the
effects in question by all individuals. Indirect market approaches use information about the
demand for marketed goods to estimate the willingness to pay for non-marketed environmental
attributes which are closely related to the marketed good (e.g., property value studies which
estimate the willingness to pay for a change in the level of pollutant concentration through
the analysis of the price of residential property) (Freeman, 1979). Each of the three groups
of approaches have different data requirements. The physical damage function approaches have
been the most widely used in the past and will therefore receive the most extensive treatment
in this chapter. Non-market and indirect market studies will be discussed later in Section
10.5.
SOX10A/A 10-1 12-15-80
-------
LEGISLATION
EXECUTIVE
ACTS
CONTROL
MEASURES
POLICY
DECISIONS
POLLUTANT
EMISSIONS
PHYSICAL
AND
CHEMICAL
INTER-
ACTIONS
COSTS OF
CONTROL
ECONOMIC
DAMAGE
FUNCTIONS
PHYSICAL
DAMAGE
FUNCTIONS
PHYSICAL
DAMAGE
AMBIENT
LEVELS
PROCESSES AND ACTIONS
RESULTANT PRODUCTS
Figure 10-1. Relationship among emissions, air quality, damages and benefits, and policy decisions.
Source: Hershaft (1976)
•
10-2
-------
In the damage function approach, before economic damage is estimated, physical damage
must be determined. This determination begins by calculating the exposure of certain
materials in specific locations. Exposure is estimated from ambient air levels over specific
time intervals. Depending on the pollutants and the kinds of material studied, both average
short-term and long-term exposure data may be needed. These data are used to develop a
physical damage function, which is a mathematical expression linking exposure to damage. The
damage function is expressed in terms appropriate to the interaction of the pollutant and
material. For example, the corrosion of metal might be expressed in units of thickness lost,
while the deterioration of paint could be stated in units of reflectance or thickness lost.
The pollutant level may be expressed in any concentration unit. Since material damage usually
develops over time, average long-term pollutant levels are often used in calculating damage.
Damage studies often encompass a yearly time scale, which facilitates economic comparisons and
allows averaging of conditions in changeable climates.
A major problem in establishing reliable dama'ge function has been to separate the
influence of the target pollutant from that of meteorological parameters and other air
pollutants. For the corrosion of metals, relative humidity is the most important variable.
Rainfall, time of wetness, sunlight, and wind speed and direction are available for many
locations and, if relevant, can be factored into damage calculations. The influence of
atmospheric components should be considered in the context of the material being studied. For
example, in studies of SO -related steel corrosion, even low atmospheric chloride levels may
have a significant impact on the rate of damage.
Before an economic value is placed on the incremental damage caused by pollution, a
critical damage level must be determined. This level represents the point at which the
service life or functional utility of the material has ended or is severely impaired. When
this point is reached, replacement or repair is necessary, and cost is incurred. For example,
if a typical coat of paint is 60 urn thick, the critical damage level at which repainting is
necessary may be around 50 urn. By comparing the amount of surface erosion in a clean environ-
ment with that in an area with a specific pollutant, the apparent damage from the pollutant
can be calculated and used in assessing physical damage. The valuation is determined through
economic damage functions, (1) which may be developed from physical damage functions; (2)
which may relate economic damage directly to pollutant level by comparing the life-time costs
of a given material in different ambient environments; (3) which may combine both approaches.
In the first case, physical damage functions are coupled to the use and value of the material.
Included in the estimate are exposure, replacement, protection, and other data. None of these
approaches account for irreplaceable items, such as works of art, where the only measurable
cost is that of preservation. The social losses of artifacts cannot be quantified
economically.
As discussed above in the context of Figure 10-1, most estimates of economic daaage to
materials have followed the damage function approach. This approach is discussed here to
illustrate how damage functions may be used as an integral component of cost estimation. The
SOX10A/A 10-3 12-15-80
-------
chapter is organized in several parts. Presented first are laboratory studies. These con-
trolled experiments investigate variables that influence damage to a material at various
pollutant concentrations. The results of these studies are used to develop damage functions,
which include only those variables that affect material damage.
Field experiments are performed to test the validity of extrapolating laboratory results
to ambient conditions. Parameters measured in field experiments are those found or suspected
to be significant factors in laboratory experiments. For example, if relative humidity were
part of the damage function derived from laboratory results, it would be carefully measured in
the field experiment.
Results from field experiments are then compared with laboratory-derived damage
functions. The comparisons may show similarities or differences. A field study may measure a
parameter not measured in a controlled experiment but which correlates well with damage. In
such a case, a laboratory experiment may be designed to determine whether or not a causal
relationship between parameter and damage exists. When a damage function has been validated
by field and laboratory experimental results, it may be used as a basis for determining total
damage to materials exposed to an air pollutant.
In the damage function approach to cost estimation described above, the physical damage
function is one of several components of an economic damage function. Other necessary com-
ponents to this approach include a mathematical expression of surface area of materials
exposed and cost factors associated with units of physical damage. Problems arising from this
approach have been (1) valid physical damage functions do not exist for all pollutants and all
materials; (2) estimates of the amount and type of material exposed have usually been based on
some surrogate such as production figures modified by service life data (field surveys of
exposed material in place have not been performed); (3) cost factors may or may not accurately
reflect cost associated with pollution as opposed to other causes. This latter consideration
is especially a problem in estimates of soiling damage, where socioeconomic factors are
heavily involved.
Past estimates of costs associated with particulate matter, sulfur oxides, or other
pollutants must be considered in light of the above discussion. The last section of this
chapter discusses cost estimates and their limitations for decision-making purposes.
10.2 SULFUR OXIDES
10.2.1 Corrosion of Exposed Metals
Sulfur oxides in the environment accelerate the corrosion of metals. Several factors
other than concentration of S02 are important. These are discussed first. ;
10.2.1.1 Physical and Chemical Considerations—The atmospheric corrosion of most metals is a
diffusion-controlled electrochemical process. For electrochemical action to take place, the
following are necessary: (1) an electromotive force between points on the metal surface; (2>
a mechanism for charge transfer between the electronic conductors; and (3) a conduction path
between the cathode and anode reaction centers. Measurements of the rate of SOg-accelerated
rusting of iron vary greatly from site to site, despite careful monitoring of pollutant
SOX10A/A 10-4 12-15-80
-------
concentrations, a fact that has often puzzled researchers. Several factors might oe
responsible for inconsistent results. These include: (1) the deposition rate of gaseous or
dissolved S02 and particles; (2) the variability in the electrochemical actions that cause
corrosion; (3) the influence of rust on the rate of subsequent corrosion; and (4; t-.e
interaction between the pollutant effects and "wetness time" often indicated by relati/e
humidity (RH) on surface electrolyte concentrations.
According to Nriagu (1978), once corrosion has begun, the progress of the reaction is
largely controlled by the sulfate ions formed by oxidation of the adsorbed SO-. The actual
mechanism for the oxidation of S02 (and its hydrated products) at metal -water interfaces is
little understood. Barton (1973) proposed the following reaction:
or
S02 + 02 + 2e
302 + 4e
The electron is provided by the oxidation of the metal (M):
M -» Mn+ + ne
Duncan and Spedding (1974), using an electrophoretic method, found that the rates of sulfate
formation on iron and zinc surfaces were similar; the pseudo-first-order half-life was
determined to be about 24 hours. Other workers (Karraker 1963; Yoshihara et al., 1964)
reported higher oxidation rates (half-life, 10 to 100 minutes) in bulk solutions using Fe3+
catalysts as summarized in Nriagu (1978). Nriagu also noted that rust on iron and steel is
first restricted to localized sites or "nests" and then spreads across the entire exposed
surface. At an SO, concentration of about 260 ug/m , corrosion products were obvious on iron
53
surfaces after 6 to 8 weeks, whereas at SOp concentrations of 4 x 10 ug/m , they could be
seen after only a few hours.
2
Barton (1976) showed that the critical SO, flux for corrosion was 6 to 10 ug/m per year
2
for steel and 18 to 20 pg/m per year for zinc and copper. The formation of rust drastically
increases the adsorption rate for S02. For example, at an S02 concentration of >0.001 percent
and relative humidity (RH) of >96 percent, virtually all of the S09 that comes into contact
i f-
with the rusting surface is adsorbed or consumed in reaction.
10.2.1.1.1 Relative Humidity and Corrosion Rate. According to Schwarz (1972), the corrosion
rate of a metal should increase by 20 percent for each increase of 1 percent in the relative
humidity (RH) above the critical RH value. It is evident that RH has a considerable influence
on the corrosion rate. This influence has been established in laboratory trials by Barton and
Bartonova (1969) and Sydberger and Ericsson (1976). It is apparent from Figure 10-2 (Haynie
and Upham, 1974) that the corrosion rate of steel increases with increasing RH as well as with
increasing S02 concentration.
The climate of an area is usually characterized by average RH rather than RH
distribution. Since average RH is calculated from the distribution, there should be an
empirical relationship between average RH and the fraction of time some "critical huaidity
SOX10A/A 10-5 1^15-80
-------
I
100
90
80
70
E
uT 60
2 50
g
CO
S 40
ec
6 30
o
t/i 20
10
0
I
SOj CONCENTRATION. jUfl/m'
_L
J_
I
I
0 10 20 30 40 50 60 70 80 90 100
AVERAGE RELATIVE HUMIDITY, %
Figure 10-2. Steel corrosion behavior is shown as a function of aver-
age relative humidity at three average concentration levels of sulfur
dioxide.
Source: Haynie and Upham (1974).
10-6
-------
value" (minimum concentration of water vapor required for corrosion to proceed) is exceeded
(Mansfeld and Kenkel, 1976; Sereda, 1974). The fraction of time that the surface is wet must
be zero when the average RH is zero and unity when the average RH is 100 percent. The
simplest single-constant first-order curve that can be fitted to observed data and that meets
these conditions is described by the equation f = [(1 - k)RH]/(100 - kRH), where f is the
fraction of time the RH exceeds the critical value, RH is the average RH, and k is an
empirical constant less than unity (Haynie, 1980).
Ten quarter-year periods of RH data from St. Louis International Airport were analyzed
and fitted by the least-squares method to the above equation. The fraction of time the RH
exceeded 90 percent gave a value of 0.86 for k. This fraction and the data points are plotted
in Figure 10-3 (Haynie, 1980).
When the temperature of a metal is below the ambient dewpoint, condensation of water on
the metal surface will take place. The metal temperature at which condensation occurs will
vary with heat transfer between ground and metal and between air and metal. Condensation will
occur when the RH adjacent to the surface exceeds a value in equilibrium with the vapor
pressure of a saturated solution on the surface. The solution may contain corrosion products,
other hygroscopic contaminants, or both. Temperature, wind, and sunshine then become factors
in establishing corrosion rates, since they determine whether there will be sufficient dew
condensation.
Haynie (1980) reported on the relationship between diffusion theory and thermodynamics
for the observed effects of five variables: pollution level, RH, temperature, wind velocity,
and surface geometry. He observed that metals must be wet to corrode electrochemically.
Surfaces are wet from condensation much more often than from precipitation.
10.2.1.1.2 Influence of Rainfall on Corrosion. Steel surfaces shielded from the leaching
effect of rain may corrode at a higher rate than those exposed to rain. The sulfate content
of rust has been identified as a dominant factor in corrosion and is found at higher concen-
trations on surfaces sheltered from rain than on exposed surfaces because soluble sulfate is
leached from the rust. However, sulfur deposition during rainfall must also be considered.
Haagenrud and Ottar (1975) noted that the rate of corrosion of unalloyed steel and zinc
correlated with the amount of sulfur (S02 and sulfate) in air and in precipitation.
As Kucera's (1976) review of this problem indicates, the mode of deposition complicates
the analysis of acidic precipitation's effects. For example, in an area where dry deposits of
hydrogen and sulfate ions exceed deposits in wet precipitation, flat steel plates corrode more
rapidly on their undersides than on their upper surfaces, suggesting that rainfall has more of
a washing effect than a corrosive action. However, in other areas where wet and dry
deposition were about equal, the upper sides of the plates corroded more quickly, suggesting
that the corrosive effect of the rainfall predominated. Other variables, including amount and
frequency of precipitation, its pH level, humidity, and temperature, also determine the iapact
of acidic precipitation (Kucera, 1976). For a more thorough discussion of the effects of
acidic precipitation on corrosion, see Chapter 7.
SOX10A/A 10-7 12-15-80
-------
£
3.
Ill"
C
o
55
O
DC
C
1
Q
D
100
90
80
70
60
50
40
30
20
10
I
I
I
0 100 200 300 400
AVERAGE SULFUR DIOXIDE CONCENTRATION,/u 9/m3
Figure 10-3. Steel corrosion behavior is shown as a function of aver-
age sulfur dioxide concentration and average relative humidity (RH).
Source: Haynie and Upham (1974).
10-8
-------
10.2.1.1.3 Influence of Temperature on Corrosion. From chemical reaction kinetics, one might
expect corrosion rates to follow the Arrhenius law according to which the logarithm of a reac-
tion rate is inversely proportional to the absolute temperature. However, this does not
appear to be a major factor controlling corrosion rates. In most cases, as discussed later,
the reaction rate is controlled by diffusion, either in the corrosion product film or in the
environment. In either case, the rate of diffusion is relatively insensitive to changes in
ambient temperature. A decrease in temperature raises the RH while decreasing diffusivity;
thus, the normal temperature range will most likely not observably affect the overall
corrosion rate. Even at -20°C metal corrosion is not halted (Sereda, 1974).
Guttman (1968) and Haynie and Upham (1974), using statistical techniques of multiple
linear regression and nonlinear curve fitting, did not obtain from their data a significant
correlation between corrosion and temperature. However, increased temperature decreases the
solubility of oxygen in the electrolyte. Moreover, temperature decreases can cause
condensation, the increased quantity of condensed droplets can affect the corrosion rate, and
the rate at which a surface dries depends on both temperature and air movement.
Guttman and Sereda (1968) made continuous measurements of sulfur oxides, time of wetness,
and temperature in their outdoor exposure tests. The corrosion rate increased markedly with
temperature, which may be attributed to a speeding up of the electrochemical process. The
effects are not simply thermodynamic since other factors involved include evaporation of
surface electrolyte and decrease in the solubility of oxygen and gaseous pollutants. Barton
(1976) did find that the effect of increased temperature was more pronounced when the rust
contained little water and sulfate. Haynie et al. (1976) found temperature to be a
significant variable in chamber studies of weathering steel.
10.2.1.1.4 Hygroscopicity of Metal Sulfates. The sulfate in rust stimulates further
corrosion by a mechanism that is related to the critical relative humidity at which an
electrolyte film is formed. The hygroscopicity of iron sulfates in the rust lowers the
critical RH for corrosion; however, sulfates are not the most deliquescent salts. For
example, chloride and nitrate salts which have higher hygroscopicity than sulfates make
corrosion possible at lower humidities.
Surfaces contaminated by sea salt (mostly sodium chloride) can be expected to be wet when
the RH exceeds 75 percent. In contrast, calcium chloride keeps surfaces wet at an RH as low
as 30 percent. A saturated solution of zinc sulfate at 20°C is in equilibrium at 90 percent
RH. Thus, zinc corroded by sulfur dioxide is expected to be wet when the RH exceeds 90
percent (Haynie, 1980).
Some pollutants that react chemically with various materials and with other materials are
not consumed and produce catalytic effects. The following four mechanisms act to increase
corrosion rates without consuming the pollutants: (1) hygroscopic materials increase the
amount of time that a corroding surface is wet; (2) conductivities of solutions and corrosion
product films increase; (3) pollutants form soluble intermediate reaction products or
activated complexes that destroy the protective nature of films; and (4) lowering the pH
usually increases the solubility of corrosion products (Haynie, 1980).
SOX10A/A 1°-9
-------
100
90
80
70
60
50
u.
O 40
u.
O
ui
5
I I
I 1 I I I I
t- 30
.20
10
0
0 10 20 30 40 50 60 70 80 90 100
AVERAGE RELATIVE HUMIDITY, %
Figure 10-4. Empirical relationship between average relative humid-
ity and fraction of time relative humidity exceeded 90 percent is
shown for data from St. Louis International Airport
Source: Haynie (1980).
10-10
-------
10.2.1.1.5 Electronic Conductivity of Rust. Barton (1976) postulated that sulfate ions
influence the anodic dissolution of iron as a function of their concentration at the
steel-rust interface. The corrosion rate of the rust layer is based in part on the high
electronic conductivity of rust, which allows the reduction of oxygen to occur within the rust
layer. The rate is also influenced by the porosity of rust, which permits rapid diffusion of
oxygen to the cathode.
In the presence of SO-, ferrous sulfate is formed before insoluble rust develops. The
amount of S02 required is small; each SO- molecule can generate 20 to 30 molecules of rust.
Once ferrous sulfate is formed, rusting can continue even though SO- is no longer present in
gaseous form.
10.2.1.1.6 Cathodic Reduction of Rust. Evans (1972) suggests that oxidative hydrolysis of
ferrous sulfate occurs slowly, and would be important only in the initial stage of corrosion.
He proposes that there is a rate-controlling cathodic process. Thus, the corrosion products
in the ferric state would be converted to magnetite (Fe^) by a reaction involving the reduc-
tion of ferric oxyhydroxide (FeOOH):
Fe2"1" + SFeOOH + 2e~ -» 3Fe304 + 4H20
10.2.1.1.7 Corrosion-Protective Properties of Sulfate in Rust. The rust layer on steel is
somewhat protective against further corrosion, though far less so than the corrosion layer on
zinc and copper. A limiting factor in rust's protection of steel is the content of soluble
compounds in rust.
Rust samples investigated by Chandler and Kilcullen (1968) and by Stanners (1970)
contained 2 to 2.5 percent soluble sulfate and 3 to 6 percent total sulfate. The outer rust
layer contained a small amount (0.04 to 0.2 percent) of soluble sulfate, compared with 2
percent in the inner rust layer. The concentration of insoluble sulfate was fairly uniform
throughout the rust layers.
The emphasis on the composition of the rust layer has led to studies of the
corrosion-protective properties of rust as a function of exposure history (Nriagu, 1978;
Sydberger, 1976). Steel samples initially exposed to low concentrations of sulfur oxides and
then moved to sites of higher sulfur oxide concentrations corroded at a slower rate than did
samples continuously exposed to the higher concentrations. Exposure tests started in summer
showed slower corrosion rates during the first years of exposure than did those started in
wi nter.
• The- long-term corrosion rate of steel appears to depend on changes in the composition and
structure of the rust layer. During the initiation period, which varies in length with the
S02 concentration and other accelerating factors, the rate of corrosion increases with time
(Barton, 1976). Because it is porous and nonadherent, the rust initially formed offers no
protection; in fact, it may accelerate corrosion by retaining hygroscopic sulfates and chlo-
rides, thus producing a microenvironment with a high moisture content (most often reported in
terms of percent RH). After the initiation stage, the corrosion rate decreases as the protec-
tive properties of the rust layer improve. Satake and Moroishi (1974) relate this slowing
SOX10A/A 10-11
-------
down to a decrease in the porosity of the rust layer. During a third and final stage, a
constant rate of corrosion is attained and the amount of sulfate in rust is proportional to
atmospheric sulfur oxide concentrations.
Where there is no appreciable deposition of sulfur or chloride compounds, the corrosion
of steel is low, even with high atmospheric humidity and temperature (Haynie and Upham, 1974).
The quantitative determination and subsequent interpretation of corrosion rates becomes
difficult if it is not known how long the metal has had a surface layer of electrolyte.
Variations in the "wet states" will occur with RH, temperature, rain, dew, fog, evaporation,
and wind. The surface electrolyte layer may form on a metal surface as a result of rain, dew,
or adsorption of water from the atmosphere. Capillary condensation in rust can be related to
the minimum atmospheric moisture content that allows corrosion to occur (i.e., critical RH).
Centers of capillary condensation of moisture on metals can occur in cracks, on dust particles
on the metal surface, and in the pores of the rust (Tomashov 1966).
10.2.1.2 Effects of Sulfur Oxide Concentrations on the Corrosion of Exposed Metals--Most of
the laboratory studies reviewed in this section have measured corrosion rates related to
exposure to sulfur dioxide alone or in combination with other compounds. In field exposure
studies, where sulfur oxides almost invariably occur in combination with other airborne
pollutants, an attempt is made to assign separate values to sulfur oxides and to describe
pollutant interactive effects on corrosion. The discussion here unavoidably overlaps somewhat
with a later section on particles, since sulfur dioxide contributes to formation of secondary
sulfates. Here, the emphasis is on the direct role of sulfur oxides in the corrosion process
(e.g., the oxidation of SOp with moisture on a metal surface). In the later section, sulfates
are discussed mainly in terms of their indirect role (e.g., their ability to increase wetness
time of a metal surface).
10.2.1.2.1 Ferrous Metals. Ferrous metal products and structures are exposed widely to
ambient pollutant levels. Rusting of these metals is the best documented form of metallic
corrosion affected by sulfur oxides. This subsection reviews studies of rusting rates of
ferrous metals, such as iron, steel, and steel alloys.
A number of investigators reported data during the period 1959-1968 which showed that the
addition of 0.05 to 0.5 percent copper to steel results in improved corrosion resistance
(Larrabee, 1959; Larrabee and Coburn, 1961; Brouns and Kail a, 1965; and Schwenk and Ternes,
1968). In an extensive series of exposure tests (Guttman and Sereda, 1968), the American
Society for Testing and Materials (ASTM) corroborated previous investigations, showing that
adding copper to steel improves its resistance to corrosion.
Stainless steels contain more than 12 percent chromium and are widely-used in outdoor
exposures; they are specified for use in many industrial processes involving corrosive liquids
that rapidly attack ordinary steels. The high corrosion resistance of stainless steels that
incorporate chromium, molybdenum, and nickel is attributed to the protective properties of the
oxide film formed on these alloys. In heavily polluted atmospheres, however, this film is not
completely protective. Particles in settled dust, including sulfates and chlorides, can
\
SOX10A/A 10-12 12-15-80\
-------
promote rupture of the oxide film and cause pitting corrosion, which may be influenced by the
surface finish (see section 10.3.1). Smoothly polished or electropolished surfaces are less
likely to retain solid deposits originating from airborne particles (Larrabee, 1959).
The lowest alloyed stainless steels have little corrosion resistance. In particular, #13
Cr steel suffers pitting attack in industrial atmospheres. Evgang and Rockel (1975) report
that the austenitic steels of 18 percent Cr and 8 percent Ni are reasonably resistant in urban
atmospheres but have shown slight rusting in industrial areas. The rusting rate is decreased
when the steel surface is cleaned of atmospheric deposits.
10.2.1.2.2 Laboratory and Field Studies Emphasizing Ferrous Metals. It is useful to consider
laboratory and field studies of corrosion effects separately because cause and effect is much
clearer in laboratory experiments; field studies are often beset by confounding variables.
Laboratory studies develop from controlled experiments, including as many variables as are
thought likely to influence damage to the material being studied. Data on materials damage
result from exposure to various concentrations of the air pollutant being studied. Analysis
of the data resulting from such laboratory studies is used to develop a mathematical
expression of the relationship between the concentration of a pollutant and damage to
materials. Such a mathematical expression is called a damage function, that is, the
quantitative expression of a relationship between exposure to specific pollutants and the type
and extent of damage to a target population. Factors that are shown to be significant
influences in producing damage are included in the damage function.
Information derived from laboratory studies is used to design field experiments, which
are performed to test the possibility of extrapolating laboratory results to ambient
conditions. The parameters that are measured in field studies are those found or suspected to
be important factors in the laboratory experiments. The results of laboratory experiments do
not easily translate to field situations, however, since ambient air pollution levels and
other influencing environmental factors vary widely both in time and space. Temporal
variables include fluctuations in temperature, wind moisture content, insulation, rainfall,
and its chemical characteristics (e.g., acidic rain), and atmospheric pollutant
concentrations. Spatial factors include differences such as aspect, altitude, electromagnetic
fields, and indigenous microorganisms. Initial conditions of the material being studied must
also be considered.
The results of field studies are compared with the laboratory-derived damage function.
In some cases the results are comparable, and the laboratory damage function is validated. In
other cases the data analysis may result in a markedly different damage function, with more or
fewer variables.
10.2.1.2.2.1 Laboratory Studies. Spence and Haynie (1974) described the design of a
laboratory experiment to identify the effects of environmental pollutants on various naterials
including ferrous metals. The environmental system consisted of five exposure chambers to
control temperature and humidity and chill racks to simulate the formation of dew. Gaseous
pollutants included those usually monitored in field exposures: sulfur dioxide, nitrogen
SOX10A/A
10-13 12-15-80
-------
dioxide, and ozone. Experiments were statistically designed for analysis of variance, and a
system was selected to study the interactive effects of pollutants and other variables. The
effect of particles was not included in the design. The chambers were equipped with a xenon
arc light to simulate sunlight. The system was designed to maintain air contact with the
various materials at preselected temperatures, relative humidities, flow rates, and pollutant
concentrations. A dew-light cycle was used; it produced faster deterioration than did condi-
tions of constant humidity and temperature.
Haynie et al. (1976) exposed weathering steel in the chamber study described above and
measured concentrations of sulfur dioxide, nitrogen dioxide, and ozone in various combinations
and at two levels of pollutant concentration. Ozone was of interest since an earlier field
experiment (see following Field-Studies section) had indicated that the presence of oxidants
inhibited metallic corrosion. The corrosion rate was measured by loss in weight of the
weathering steel. Six panels each were exposed to 16 polluted-air and 4 clean-air conditions,
and measurements were taken at 250, 500, and 1000 hours of exposure. The weight losses were
converted to equivalent thickness loss values. As expected, corrosion was most severe at high
SOy concentrations and high humidity. Ozone neither inhibited nor accelerated corrosion. The
authors concluded that some other oxidant or unmeasured factor that was covariant with ozone
caused the inhibition effect. However, if the data from the sites with high oxidant con-
centrations in the field exposure experiments were excluded, the damage function from the
laboratory study was an excellent predictor of the field results. The coefficient of
determination for the field data using the laboratory function was 0.986. The following
physical damage functions were developed by Haynie et al. (1976) to relate S00 concentrations
2
to weathering steel and galvanized steel corrosion (R = 0.91).
corrosion = [5.64 ^ + e <55'44 ' 31,150/RT)^ ^
corrosion = (0.0187 SO, * e 41'85 ' 23,240/RT} t
€m W
where:
corrosion is expressed as \im
3
SQy is expressed at ug/m
t is time of wetness in years
R = gas constant (1.98 cal/gm mol/°K)
T = temperature (°K)
Sydberger and Ericsson (1976) studied the corrosion of mild steel at 1, 10, and 100 ppm
(2620, 26,200, and 262,000 ug/m3) S02 across the range of critical humidities (80 to 96
percent RH). The flow rate of the $Q2 atmosphere'was varied, and some samples were sprayed
with water to simulate rain or condensation. The chemical composition of the corrosion
products was studied by X-ray diffraction, infrared spectrometry, and electron spectroscopy
for chemical analyses (ESCA) techniques. The flow rates of the S02 atmospheres markedly
SOX10A/A 10-14 -12-15-80
-------
influenced the corrosion rates. It appears that corrosion rates are related not only to the
S02 concentration in the atmosphere, but also to the supply of S02 per unit surface area and
time. Spraying the samples with distilled water at intervals substantially increased
corrosion.
Sydberger and Ericsson's (1976) analysis of the corrosion product (rust layer) was based
on the concepts of Schwarz (1972) and Barton (1976) that sulfate is the primary corrosion
stimulant in rust formation. Anodic activity is maintained by the concentration of ferrous
sulfate in the electrolyte. An S02 supply of 4 ug/cm2/hr at the lowest humidity initiated
corrosion at a low rate. A rise above 50 percent RH increased corrosion markedly. Of
particular interest was the finding that a variation of the flow rate at 1 ppm (2620 ug/ra3)
S02 with 96 percent RH gave significant differences in corrosion rates. This study of the
effect of rust on corrosion showed that even at high humidity and high sulfate content, the
corrosion rate decreased to a low level when the S02 concentration was low.
10.2.1.2.2.2 Field Studies. For outdoor exposures? the primary rate-controlling factor
in the delivery of pollutants to a surface is eddy diffusion. This flux is not constant and
is a function of the horizontal wind velocity gradient away from the surface. The transport
of a pollutant to a surface is usually expressed as a "deposition velocity" (u), defined as
the flux to the surface divided by the ambient pollution level at some specific measuring
height. Reported deposition velocities for gaseous pollutants have usually been within an
order of magnitude of 1 cm/sec. These values are consistent with calculated estimates based
on an analogy with momentum flux and measured wind velocity profiles (Sydberger, 1976).
The amount of S02 reaching a steel surface depends on wind direction, wind velocity, and
the orientation of the surface to the emission source. The concept that S02 deposition varies
with flow direction and velocity suggests that data on concentration alone cannot be used to
determine the supply of S02 to metal surfaces; therefore surface adsorption methods like the
lead candle method provide valuable information in relating supply of sulfur oxides to metal
surfaces (Sydberger, 1976). Upham's (1967) work indicated, however, that corrosion of mild
steel at seven Chicago sites increased with time and with increasing mean S02 concentration
(Figure 10-5).
Haynie and Upham (1971) continuously monitored urban pollutants including S02, nitrogen
dioxide, and ozone (oxidants) to determine whether previously unconsidered variables might
affect steel corrosion. Other factors considered were temperature, RH, and total suspended
particulate matter (TSP). Their 5-year program, begun in 1963, involved sites in Chicago,
Cincinnati, New Orleans, Philadelphia, San Francisco, Washington, Detroit, and Los Angeles.
They studied three types of steel expected to show different levels of resistance to
atmospheric corrosion: (1) a plain carbon steel containing some copper (0.1 percent copper);
(2) a copper-bearing steel (0.22 percent copper); and (3) a low-alloy weathering steel (0.4
percent copper with 0.058 percent phosphorus). The exposure periods were 4, 8, 16, and 32
months. The same steels were exposed at rural sites as a control. The rural sites proved to
have higher than expected corrosion rates; however, S02 concentrations were not measured at
SOX10A/A
10-15 12-15-80
-------
18.0
o>
2 16.0
ui
<
9"
§
X
u
5
oc
o
u
14.0
12.0
10.0
8.0
6.0
4.0
2.0
0
3 (months)
I
I
I
I
I
_L
I
I
0.02 0.04 0.06 0.08 0.10
(50) (100) (ISO) (210) (260)
0.12 0.14 0.16 0.18
(310) (360) (415) (470)
MEAN SO2 CONCENTRATION, ppm (
Figure 10-5. Relationship between corrosion of mild steel and cor-
responding mean SC>2 concentration is shown for seven Chicago sites.
(Corrosion is expressed as weight loss of panel.)
Source: Upham (1967).
10-16
-------
these sites. Multiple regression analysis established significant correlations between
average S02 concentrations and corrosion of all three steels at the urban sites.
A first consideration was climatic conditions. For metallic corrosion to occur, a
certain level of wetness on the metal surface is required to provide an electrolyte film in
which electrochemical action can take place. The average humidities recorded by Haynie and
Upham were high enough to produce this critical wetness threshold much of the time.
Temperature was a statistically insignificant variable, although a later chamber study (see
previous section) did find temperature to be a significant variable. Average RH was also
insignificant because the range was only between 62 and 77 percent (Haynie and Upham, 1971).
Inspection of the monitored SO^ and oxidant concentrations revealed wide variations from
site to site. Multiple-regression analysis showed that high concentrations of oxidants
correlated with lowered metallic corrosion rates.
The resulting physical damage functions for corrosion of carbon steel, copper-bearing
steel, and weathering steel developed by Haynie and Upham'(1971) follow:
0.00161 SO
Carbon steel: Y = 9.013 e
Copper-bearing 0.00171 SO
steel:
Weathering
steel:
Y = 8.341 e
(4.768t)
0.7512 - 0.00582 OX
(4.351t)
0.8151 0.00642 OX
0.0045 SO,
Y = 8.876 e
(3.389t)
0.6695 - 0.00544 OX
where:
Y = corrosion in urn
SOp and OX (ozone) are expressed as ug/m
t = time of exposure, years
The correlation coefficient for each of these equations was 0.91. A more recent laboratory
investigation by Haynie et al. (1976), however, (see preceding section) has shown that ozone
is not a significant corrosion controlling variable.
Mansfeld (1980) made observations at nine test sites in and around St. Louis for 30
months beginning in October 1974 as part of the Environmental Protection Agency's Regional Air
Pollution Study to determine the effect of airborne pollutants on galvanized steel, weathering
steel, stressed aluminum, marble, and house paint—essentially the same materials examined in
the chamber study reported by Haynie et al. (1976). During 1975 and 1976, atmospheric
corrosion monitors (ACM) of the type described by Mansfeld and Kenkel (1976) were installed at
four sites to measure time of wetness. Each ACM consists of a copper-zinc or copper-steel
couple that registers current flow when an electrolytic path forms between the two plates as a
result of deposition of water from the air, dew, or rain on corrosion products. The ACM
measures the time that the panel is wet enough for the electrochemical mechanism of corrosion
to occur. Parameters measured in Mansfeld1s study included the following: wind speed, wind
SOX10A/A
10-17
1^15-80
-------
direction, temperature, ozone concentration, total hydrocarbon concentration, total sulfur and
NO concentration, hydrogen sulfide concentration, sulfur dioxide concentration, RH, sulfate
concentration, nitrate concentration, total suspended particle concentration, and time of
wetness.
Mansfeld determined weight losses for galvanized steel, weathering steel, house paint,
and marble; he removed aluminum tension samples after failure. Mansfeld1s data show that
damage to a particular material does not necessarily occur at the same corrosion rate at each
site. Preliminary statistical analysis of the results failed to show significant correlation
between corrosivity and pollutant concentrations. There was substantial error in the
measurement of RH, an extremely important corrosion variable. Relative humidity was,
therefore, not included in the regression analyses. Concentrations of SO, measured by
3
Mansfeld (1980) were generally an order of magnitude lower than the 130 ppb (340 ug/m )
concentrations reported by Upham (1967) at urban sites in St. Louis.
In another study (1974), Haynie and Upham exposed enameling steel containing 0.019
percent carbon and 0.028 percent copper at 57 sites in the National Air Surveillance Network
(NASN). They measured corrosion by weight loss and quantitatively -determined other
pollutants, including gaseous sulfur dioxide, total suspended particles, and the amount of
sulfate and nitrate in the particles. For a discussion of the measurement methodology for
NASN, see Chapter 3. Sulfur dioxide levels ranged from 9 to 374 ug/m , total suspended
particles from 11 to 182 ug/m , and RH from 29 to 76 percent. The average temperature
remained within a fairly narrow range and was considered constant. The temperature, the
quantity of total particles, and the presence of nitrate in the particles did not
significantly affect the corrosion rate of steel. The concentration of sulfur dioxide was
significant only when sulfate was not included in the regression analysis. At each site,
sulfate content of the particles and sulfur dioxide concentration were closely related. On
the basis of this study, Haynie and Upham derived the following empirical expression to obtain
the best relationship between corrosion of enameling steel and atmospheric SOp content:
Corrosion = 325tV0-00275 S02 ' 163'2/RH>
where: corrosion is expressed as urn
o
S0~ is expressed as ug/m
RH = relative humidity (percent)
t = time
Average RH is a substitute variable for the fraction of time the steel is wet.
Considerable effort has gone into isolating environmental variables that predict
long-term corrosion rates. Empirical expressions for corrosion of various steels exposed to
the atmosphere (see Table 10-1) have been developed by Oma et al. (1965), Chandler and
Kilcullen (1968), and Haynie and Upham (1974). These equations may be used to relate
reduction in S02 and sulfates to reduction in corrosion of metals, serving as a basis for a
benefit appraisal.
SOX10A/A 10-18 12-15-80
-------
TABLE 10-1. SOME EMPIRICAL EXPRESSIONS FOR CORROSION
OF EXPOSED FERROALLOYS
Exposure Study
Material Site Duration Empirical Equation
Steel A Sheffield 1 year y = 0.51 + O.Olx
England
Parameter Units
y = corrosion rate
in mils/year 3
x = S02 in ug/nr
Source
Chandler and
Kilcullen
(1968)
Note/Comments
Authors stated that S02 and smoke
has a major influence on the corrosion
rate of steel and accounted for about
50% of the variations found at the
different sites. Other factors, such
as time of wetness, were found equally
important in determining the corrosion
rate of steel.
95% confidence limit + 0.75 roils for
any point on regression.
Steel B Sheffield 1 year
England
y = 0.82 + O.OOSx
y = corrosion rate
in mils/year ,
x = S02 in ug/m
Chandler and
Kilcullen
(1968)
95% confidence limit + 0.5 mils for
any point in regression.
Enameling
steel
NASN
sites
1-2
years
cor =183.5 /I [e0-
Enameling
steel
NASN
sites
1-2
years
faO.00275 SO,-
cor = 325 J~t ie '
cor = depth of Haynie and
corrosion, um Upman
sul = average level (1974)
of sulfate in suspended
particulate, pg/m
RH = average relative
humidity, percent
t = time in years
parameter units Haynie and
same as for previous Upham
equation except (1974)
std. dev. 85.0; 95% conf. limits 13.4
lower, 353 upper
std. dev. 0.00451; 95% conf. limits
0.0552 lower, 0.0732 upper
std. dev. 32.8; 95% conf. limits 97.6
Sul and SO, relationship defined as
sul = 8.9 * 0.0429 SO,
± 2S = 8.2 *
-------
Matsushima et al. (1974), in studies of low-alloy weathering steels, considered the
impact of the washing action of rain, the ease with which water would drain off the surface,
and the drying effect of sunlight to determine the effect of these variables on the retention
of particles that influence the electrolytic corrosion mechanism and the time of wetness. The
authors hypothesized that the geometry of unpainted weathering steels may not favor the
development of a protective oxide film of rust. The model structure used in the exposure
trials contained horizontal and inclined roofs, vertical wall panels, and window frames. Two
sites were chosen: an industrial location and a residential site in the Kawasaki area, which
has a cold, dry winter and a hot, humid summer.
The results showed that the successful use of weathering steel is related not only to the
severity of pollution but also to the specific interplay between shelter and the uniform
washing action of rain. Thus, for areas in which the structural factors are unfavorable, the
optimal rust film is slow to form and may deteriorate. Rust films develop and are then
destroyed, and the surface never develops a protective film. Generally, boldly exposed
surfaces such as horizontal or inclined roofs show the least corrosion.
10.2.1.2.3 Comparison of Ferrous and Nonferrous Metals. Sydberger and Vannenberg (1972)
examined the influence of RH and rust on the adsorption of sulfur dioxide on metal surfaces,
o
using radioactive sulfur. The concentration of S02 was 0.1 ppm (262 ug/m ), and RH varied
between 50 and 98 percent. Polished and preexposed samples of iron, zinc, copper, and
aluminum were compared for their adsorption properties. Iron, zinc, and copper were pre-
exposed to S02 concentrations (100 ppm or 2.6 x 105 ug/m3) at 98 percent RH and 22°C for 3 hr.
The aluminum samples were preexposed for 30 hr. The principal corrosion product identified by
X-ray diffractometry was hydrated metal sulfate. Adsorbed S02 was measured at 30-min
intervals with a Geiger counter. The corrosion rate at 90 percent RH was initially high for
zinc and copper but quite low for aluminum. Adsorption of S02 on preexposed iron samples was
high. At 80 percent RH, almost all of the S02 was adsorbed. The high adsorption rate is
perhaps explained by the rapid oxidation of adsorbed sulfur dioxide caused by the catalytic
effect of the rust. The initial rate of adsorption on polished iron below 80 percent RH is
related to the absence of corrosion; however, at increased humidity corrosion is initiated and
the adsorption rate increases.
Of particular significance is the observation that S02 adsorption can take place at
humidities below the critical humidity (Sydberger and Vannenberg, 1972). This finding
suggests that SO- will be adsorbed on a rusty iron surface during periods of low humidity and
will affect the corrosion rate when humidity rises. The critical humidities for nonferrous
metallic surfaces, as summarized from Nriagu (1978) and National Academy of Sciences (HAS)
(1977), appear in Table 10-2. The corrosion products of copper and aluminum have an extremely
low adsorption capacity below 90 percent RH, confirming the lower sensitivity of these metals
to corrosion by S02 (see Figure 10-6, from Sydberger and Vannenberg, 1972).
Aluminum is generally considered to be corrosion resistant. It is quite resistant to
sulfur oxides when RH is less than 50 percent. At higher humidities aluminum can corrode
SOX10A/A 10-20 12-15-80
-------
E
o 2
V)
CD
E
O
I I
I I
34567
EXPOSURE TIME, hr
COPPER
ALUMINUM
10
Figure 10-6. Adsorption of sulfur dioxide on polished metal surfaces
is shown at 90 percent relative humidity.
Source: Adapted from Sydberger and Vannenberg (1972).
10-21
-------
TABLE 10-2. CRITICAL HUMIDITIES FOR VARIOUS METALS
Metal Critical humidity, percent RH
Alumi num 75-80
Brass 60-65
Copper 65-70
Nickel 65-70
Zinc 70-75
rapidly in the presence of very high SO- concentrations forming a hydrated aluminum sulfate
[A1_(SO«)3 • ISHpO] surface deposit. At low concentrations of acid sulfate particles it forms
a protective film of aluminum oxide. However, when the film becomes contaminated with dirt
and soot particles, there is a change in surface appearance characterized by mottling and
pitting. Simpson and Horrobin (1970) reported that aluminum undergoing long exposure in
industrial areas displayed white areas of crystalline corrosion products. Aluminum surfaces
exposed for periods of more than 5 years to a sulfate concentration of 0.14 ppm (550 ug/m )
had pits as deep as 14 mils (0.36 mm). This is, of course, much higher than typical ambient
concentrations (see Chapter 5).
Fink et al. (1971) summarized measured corrosion rates and depth of pitting of aluminum
3 3
surfaces in rural, mild industrial (30 ug/m or 0.01 ppm SOp), normal industrial (370 ug/m or
0.14 ppm SO- and 80% RH), and severe industrial areas. Their overall conclusion was that,
although some loss of thickness occurred in the first 2-years, structures composed of aluminum
and its alloys are resistant to air pollutants.
In another study, Haynie (1976) found that stressed aluminum specimens exposed to SO, at
3
concentrations of 79 (0.03 ppm) and 1310 ug/m (0.5 ppm) lost approximately 8.6 and 27.6
percent of their bending strength, respectively. He also noted that 7005-T53 high-strength
aluminum alloy tubing, which contains very little copper, is susceptible to stress-corrosion
cracking in industrial environments.
Abe et al. (1971) exposed copper and copper alloys for 2 years in marine, rural, highly
industrial, and urban areas in which there was great variation in pollutant and salt content.
Analyses of the surface deposits showed the presence of basic sulfate, sulfide, and chloride
of copper. The green patina on copper was analyzed and'determined to be basic copper sulfate
in urban areas and basic copper chloride in seacoast areas. These surfaces were protective
against further corrosion.
The formation of these basic copper salts, according to Simpson and Horrobin (1970), can
take as long as 5 or more years and will vary with the concentration of sulfate or chloride
particles, the humidity, and the temperature. They reported the corrosion rate of copper to
SOX10A/A 10-22 12-15-80
-------
be 0.9 to 2.2 pro/year in industrial atmospheres, compared with 0.1 to 0.6 urn/year in rural
areas.
The high corrosion resistance of nickel and copper compared with unalloyed steel is
attributed by Sydberger (1976) to the ability of these metals to form a layer of insoluble
basic sulfate that protects the metal surface. Such layer formation does not occur on steel.
10.2.2 Protective Coatings
Susceptible materials are generally coated for protection against the effects of
exposure. The coatings provide either sacrificial protection or barrier protection. In
galvanization, zinc is applied to ferrous metal for sacrificial protection. Thus, while the
galvanized surface may suffer corrosion damage, it helps to prevent rusting of steel products
such as gutters, cables, wire fencing, and building accessories. Barrier protection is
provided by varnishes, lacquers, and paints by sealing the underlying surface material against
intrusion by moisture.
10.2.2.1 Zinc-Coated Materials—Zinc is generally exposed-as a protective coating for steel
products since zinc coating is fairly resistant to atmospheric corrosion. Zinc is anodic with
respect to steel; when zinc and steel are in contact with an electrolyte, the electrolytic
cell provides current to protect the steel from corrosion with some oxidation of the zinc.
Guttman (1968) carried out a long-term exposure of zinc panels with measurement of the
atmospheric factors. He found that zinc is corroded by SO^ and that time of wetness and
concentration of SOp are the major factors that determine the rate of corrosion.
Fleetwood (1975) conducted 5-year exposure studies of zinc and iron in a number of
locations ranging from dry tropical to industrial. He estimated the service life of
galvanized steel to be 15 to 20 years in an industrial area containing pollutants and 300
years in a dry tropical unpolluted area. Kucera (1976) noted strong correlations between the
corrosion rate and (1) the adsorption of S02 on zinc surfaces and (2) the concentrations of
so2.
Haynie and Upham (1970) exposed zinc panels in eight cities, continuously monitoring S02
concentration and collecting meteorological data, including temperature and RH, from the
nearest weather stations. They developed the following empirical equation, which correlates
corrosion rate with average sulfur dioxide concentration (for the study range of 10 to 479
ug/m^) and RH:
Y = 0.001028 (RH - 48.8) S02,
where
Y = zinc corrosion rate (um/yr),
RH = average annual RH in percent, and
S0? = average S02 concentration (ug/m )-
The regression intercept indicated that no corrosion would occur below an average relative
humidity of 48.8 percent. This expression gave a reasonably good linear fit with the
experimental corrosion results obtained by Haynie and Upham for S02 concentration and RH.
JSGX10A/A 10-23 12-15-80
-------
Based on the St. Louis study results (Haynie, 1980), the corrosion of small specimens of
galvanized steel follows the relationship:
Cz = 2.32 tw + 0.0134 v °'781 - S02 • tw>
C = corrosion in urn
t = time-of-wetness in years,
v = wind velocity in m/s, and
S02 = ug/ra
A theoretical damage coefficient (for purposes of the chapter, damage = any measurable
adverse effect) for a pollutant can be calculated from the stoichiometry of a reaction and the
deposition velocity. For the reaction between SOp and zinc to form zinc sulfate, the
coefficient is 0.045, when the zinc corrosion rate is expressed fn micrometers per year, SO^
in micrograms per cubic meter, and the deposition velocity (u) in centimeters per second. For
a small zinc or galvanized steel sheet specimen, the material damage coefficient for S07 is
0 78
calculated to be 0.0123 v " which agrees well with the above determined empirical
o
coefficient. At a wind velocity of 4 m/sec, the value is 0.0363 (um/yr)/ (pg/m ). For the
same conditions, a similar calculation for marble yields a coefficient of 0.136
3
(um/yr)/(ug/m ).
Haynie (1980) restudied the results of six exposure investigations to relate the,
corrosion of zinc and galvanized steel to the concentration of sulfur dioxide. Each
investigation was different and the data were evaluated differently; thus, no direct
comparison of the results as they were published was possible. However, in reevaluating the
data from each study, Haynie used techniques that permitted the comparison of the various data
sets.
The experimental regression coefficients obtained from all of these studies are compared
in Table 10-3. The SO^ coefficient for the chamber study is low, whereas the analogous
coefficients for the Community Air Monitoring Program (CAMP) (Haynie and Upham, 1970) and
Interstate Surveillance Program (ISP, Cavender et al., 1971) studies are high and agree with
each other. The remaining three S02 coefficients are generally in good agreement. The
average of the time-of-wetness coefficients are within a range of + 0.75 from a mean of 1.73
um/yr.
The specified thickness of galvanized coating varies with intended use. Furthermore, the
thickness of a particular coating varies considerably from one point to another. Bird's
measurements revealed that 5 percent of coating thickness measurements varied from the mean by
more than 46 percent (Bird, 1977). Haynie confirmed this variability with 475 thickness
measurements on a single galvanized steel sheet. The life of a coating is generally
proportional to its thickness; thus, rusting of the substrate steel will occur first at the
thinnest spots and last at the thickest spots. The American Society for Testing and Materials
(ASTM) as reported in Haynie (1980) has observed rusting at thin spots on galvanized steel
wire, fencing, and sheet exposed to various types of atmospheres over many years. Some of
their exposures were started in 1916 and were continued until the test could reveal no
SOX10A/A 10-24 12-15-80
-------
TABLE 10-3. EXPERIMENTAL REGRESSION COEFFICIENTS WITH
ESTIMATED STANDARD DEVIATIONS FOR SMALL ZINC AND
GALVANIZED STEEL SPECIMENS OBTAINED FROM SIX
EXPOSURE SITES
Study
Time-of-wetness
coefficient,
Mm/yr
S0« coefficient,
(ufn/yr)/(Mg/m )
Number of
data sets
CAMP (Haynie and
Upham, 1970)
ISP (Cavender et al.,
1971)
Guttman, 1968
Guttman and Sereda,
1.15 + 0.60
1.05 + 0.96
1.79
0.081 + 0.005
0.073 + 0.007
0.024
37
173
>400
1968
Chamber study (Haynie
et al., 1976)
St. Louis (Mansfeld,
1980)
2.47 + 0.86
1.53 + 0.39
2.36 + 0.13
0.027 + 0.008
0.018 + 0.002
0.022 + 0.004
136
96
153
additional information. In the case of sheet, the product was completely rusted and showed
perforations. In general, the amount of corrosion at each site varied linearly with time.
Corrosion rates at each site were calculated on the basis of time to first rust and time to
complete rust for various zinc thicknesses, assuming +40 percent thickness variability.
Originally, State College, Pa., was selected by ASTM as a control site representing a
"clean" rural environment; however, the corrosion was higher there than at five other rural
locations including the rural-marine environment of Santa Cruz, Calif., where high relative
humidities are expected to accelerate corrosion. ASTM made no pollution measurements but
recognized the effects of "industrial" and "severe industrial" environments.
Zinc corroded nearly twice as fast on wire and fencing as it did on sheet, a finding that
is consistent with the theoretically predicted effects of surface configuration on S02
deposition velocity. One would expect a greater deposition velocity onto fencing than onto
sheet material. ASTM noted that fencing corrodes less near the ground than it does near the
top because wind velocity increases with height, with a resultant increase in deposition
velocity and similar variation of S02 concentration. Another factor may be the decrease in
concentration near the ground, since both soil and vegetation are sinks for S02 (see Chapters
6 and 8).
OX10A/A
10-25
12-15-80
-------
The average corrosion rates in the Haynie (1980) study shown in Table 10-4 correspond to
actual corrosion rates that are two to four times greater when the substrate is wet.
Theoretical calculations indicate that the average SO, levels at the Pittsburgh site over the
3
long period of exposure were between 350 and 700 ug/m . The average at the Altoona site could
3
have been as high as 1000 pg/m .
From the relationships between theoretical and experimental studies, Haynie (1980)
concluded the following:
1. Both short-term laboratory evidence and long-term exposure results for galvanized
steel are consistent with theoretical considerations.
2. Damage functions for some materials can be calculated from theoretical relationships
that consider factors controlling time of wetness and pollutant fluxes.
TABLE 10-4. CORROSION RATES OF ZINC ON GALVANIZED STEEL PRODUCTS EXPOSED TO
VARIOUS ENVIRONMENTS PRIOR TO 1954
Mean corrosion rate and estimated standard
deviation, urn/year
Site
Sheet
Wire and fencing
Altoona, PA
Pittsburgh, PA
Sandy Hook, NJ
Bridgeport, CT
Lafayette, IN
Ithaca, NY
State College, PA
Ames, IA
College Station, TN
Santa Cruz, CA
Manhattan, KS
Davis, CA
7.57 + 0.54
5.63 + 0.34
2.74 + 0.30
1.27 + 0.29
10.86 + 1.02
4.37 + 0.45
4.25 + 0.44
2.94 + 0.34
2.68 + 0.42
2.48 + 0.24
1.68 + 0.19
1.22 + 0.43
0.83 + 0.26
0.79 + 0.27
0.76 + 0.42
Source: Haynie (1980)
SOX10A/A
10-26
12-15-80
-------
3. Wind speed and material geometry should be considered in evaluating
atmospheric corrosion effects.
Marker et al. (1980) examined the variables controlling the corrosion of zinc by SO,, and
sulfuric acid. They used an aerosol flow reactor. Under steady-state conditions, they made
the following measurements:
Environmental Measurements:
o Percent RH and temperature (at two points).
o Average flow velocity (Pitot tube).
o Flow velocity profile (recorded when a steady state had been estab-
lished).
Aerosol Measurements:
o Aerosol size distribution and number concentration determined at
intervals during test by TSI 3050 analyzer.
o Two total-mass filter samples collected.
o Total-deposition sample collected on aluminum foil throughout each
experiment.
o TEM deposition grid samples collected continuously.
o X-ray photoelectron spectroscopy samples (both zinc plate and
aluminum foil) collected continuously during experiment.
Corrosion rate measurements were recorded continuously by an atmospheric corrosion monitor
(ACM), which had been pretreated with either 0.1 N sulfuric acid or ammonium sulfate. The ACM
consists of galvanic cells of alternating plates of copper and zinc separated by sheets of
plastic insulators. Experimental conditions were selected from the following ranges:
Temperature, °C 12-20
Relative humidity, percent 65-100
Mean flow velocity, m/sec 0.5-8
Sulfur dioxide concentration, ppb (volume) 46-216
Sulfate aerosol mass concentration, mg/m 1.2
Aerosol size distribution, urn diameter 0.1-1.0
The factors controlling the rate of corrosion in the Marker et al. study (1980) were RH,
pollutant flux, and chemical form of the P9llutant. Corrosion occurred only at RH high enough
(more than 60 percent) to wet the surface; temperature did not appear to be a controlling
factor within the range 12-20°C. The results indicate that on initial exposure S02-induced
corrosion of zinc proceeds at a rate approximately a factor of two greater than that for the
equivalent amount of deposited sulfuric acid aerosol.
The investigators noted deposition velocities of 0.07 cm/sec for 0.1-1.0 pm sulfate
aerosols and 0.93 cm/sec for S02 at a friction velocity of 35 cm/sec. These factors indicate
that the effects of S02 will dominate the effects of H2S04 in most urban areas.
10.2.2.2 Paint Technology and Mechanisms of Damage—Cotnpared with other environmental impacts
such as sun and precipitation, paint damage due to air pollutants is considered less
important. There are at present no standard ASTM procedures for evaluating the effect of S02>
30X10A/A 10-27 12-15*80
-------
nitrogen dioxide, and/or ozone on paints. Degradation by ultraviolet light has received the
major emphasis; outdoor test stations have been located where SCL levels are low.
The erosion of paint can be measured by loss of thickness of the paint layer, which can
result from the chemical action of SOp and the action of light and ozone. Film erosion rates
are used by paint manufacturers to determine the fail point for their formulations.
In the formulation of paints, the ratio of pigments to film formers is of importance in
the overall properties of gloss, hardness, and permeability to water. If the amount of film
former is too low, soiling is increased and the paint may lose the film flexibility needed for
durability and become brittle. The permeability of paints to water has been reported by Hay
and Schurr (1971). High-permeability films are desirable for surfaces that must allow water
to pass through, such as wooden exterior walls behind poorly ventilated kitchens.
Low-permeability coatings are needed to protect surfaces that corrode when repeatedly
moistened. The low permeability of chlorinated rubber is advantageous for use on concrete, a
use this coating shares with styrene-aerylie.
Paint films permeable to water are also susceptible to penetration by S02 and sulfate
aerosols. The absorption of S02 was observed by Holbrow (1962), who found sulfites and
sulfates in paint, and by Walsh et al. (1977), who used radioactive S02 to determine rates and
saturation values for S02 absorption.
Concentrations of SOp encountered in fogs or near industrial sites can increase the
drying and hardening times of certain kinds of paints. Holbrow (1962) found that the drying
time of linseed, tung, and certain castor oil paint films increased by 50 to 100 percent on
3
exposure to 2620 to 5240 ug/m (1 to 2 ppm) SOp. The touch-dry and hard-dry times of alkyl
and oleoresinous paints with titanium dioxide pigments were also reported to increase substan-
tially; however, the exposure time of the wet films was not reported. Analysis of the dried
films indicated that S02 had chemically reacted with the drying oils, altering the
oxidation-polymerization process. No studies have been reported on the effects of SOp on the
drying of latex paints.
Holbrow (1962) also studied the effects of sulfur dioxide on dried paint film. In these
experiments, paint films-were allowed to dry, were refrigerated, and then were exposed for 15
rain to an atmosphere containing 1.2 percent S02- The paint films with condensed moisture were
finally placed in an accelerated-weathering chamber. For all paints except a pentaerythritol
alkyd paint, the gloss decreased significantly after 1 day in the accelerated-weathering
chamber. Without the accelerated weathering, the actions of sulfur dioxide and moisture on
the paint films produced only a slight reduction in gloss.. Holbrow concluded that the sulfur
dioxide had sensitized the film, permitting water to be absorbed during the weathering cycle.
T
Bluing of lead chromate pigment (green) paints has been observed during the early life of
the film. Holbrow (1962) reproduced this effecttin the laboratory by exposing the film to
sulfur dioxide and moisture and then to warmth and moisture. The bluing was probably caused
by conversion of the lead chromate pigment to lead sulfate. Holbrow/did not attempt to
SX10AA/A 10-28 - 12-15-80
-------
correlate moisture, temperature, and pollutant concentration, or to obtain dose-response aata.
Although very high levels of sulfur dioxide were used, this experimentation does indicate trat
condensation and moisture evaporation are critical in concentrating the pollutant on the
surface of exterior paint films; under these conditions, deterioration of the film occurs.
Svoboda et al. (1973) compared pigmented and unpigmented paint film for S02 permeability
and found that the rate of penetration of SCL into a paint film was related to the pigment
content. Zinc oxide and titanium dioxide pigments caused a 50 to 70 percent decrease in the
rate of penetration of SOp into the paint film.
Spence et al. (1975) carried out a chamber study of the effects of gaseous pollutants on
four classes of paints formulated for exterior exposure: oil-base house paint, vinyl-acrylic
latex house paint, and vinyl and acrylic coil coatings for metals. The house paints were
applied to aluminum panels by spraying. The coil coating panels were cut from commercially
painted stock. The oil-base paint film was 58 urn thick; the acrylic latex, 45 urn; the vinyl
coil coating, 27 urn; and the acrylic coil coating, 20 urn'. The exposure chambers controlled
temperature, humidity, SCL, nitrogen dioxide, and ozone. Each exposure chamber had a xenon
arc lamp to provide ultraviolet radiation. A dew/light cycle was included; light exposure
time was followed by a dark period wherein coolant circulated through racks holding the
specimens, thereby forming dew on the panels. Each dew/light cycle lasted 40 min and
consisted of 20 min of darkness with formation of dew, followed by 20 min under the xenon arc.
The total exposure time was 1000 hr. Damage was measured after 200-hr, 500-hr, and 1000-hr
intervals by loss of weight and by loss of film thickness. In evaluating the data, loss of
weight was converted to equivalent loss of film thickness.
Visual examination of the panels coated with oil-base house paint revealed that all
exposure conditions caused considerable damage. The erosion rate varied from 28.3 to 79.1
urn/year, with an average of 60 urn/year. The investigators concluded that S02 and RH markedly
affected the rate of erosion of oil-base house paint. The presence of nitrogen dioxide
increased the weight of the paint film. A multiple linear regression on S02 concentration and
RH yielded the following relation:
E = 14.3 + 0.0151 S02 + 0.388 RH,
where
E = erosion rate in urn/year,
o
S02 = concentration of S02 in ug/m , and
RH = RH in percent.
The authors reported the 95 percent tolerance limits on 99 percent of the calculated rates to
be ± 44 urn/year.
The formation of blisters on acrylic latex house paint was noted at the high S02 levels.
The blisters resulted from severe pitting and buildup of aluminum corrosion products on the
substrate. The paint acted as a membrane retaining moisture under the surface and excluding
SX10AA/A
10-29 12-15-80
-------
oxygen which would passivate the aluminum. The vinyl coating and the acrylic coating are
resistant to SO^. The visual appearance of the vinyl coil coating showed no damage. The
average erosion rate was low, 3.29 urn/year. The average erosion rate for a clean air exposure
was 1.29 urn/year. The acrylic coil coating showed an average erosion rate of 0.57 urn/year.
A study of the effects of air pollutants on paint was conducted by Campbell et al.
(1974). The paints studied included oil and acrylic latex house paints, a coil coating,
automotive refinish, and an alkyd industrial maintenance coating. These coatings were exposed
3 3
to clean air, S02 at 262 and 2620 ug/m , and ozone at 196 and 1960 ug/m (i.e., equivalent to
0.1 and 1.0 ppm of each pollutant). Other controlled study variables included light,
temperature, and RH. In addition, one-half of the coatings were shaded during the laboratory
exposures. Similar panels (half facing north) were exposed at field sites in Leeds, North
Dakota; Valparaiso, Indiana; Research Center, Chicago, Illinois; and Los Angeles, California.
The laboratory exposure chamber operated on a 2-hour light-dew cycle (i.e., 1 hour of
xenon light at 70 percent RH and a temperature of 66°C followed by 1 hour of darkness at 100
percent RH and a temperature of 49°C). Coating erosion rates were calculated after exposure
periods of 400, 700, and 1000 hours. Estimated erosion rates and statistical characteri-
zations of the results are summarized in Table 10-5. Erosion rates at 03 or S0£ concentra-
tions of 0.1 ppm were not significantly different from values for clean air exposures due to
high variability of the data. The erosion rates on the shaded specimens were significantly
less than the unshaded panel results shown in Table 10-5; panels facing north were also less
eroded. At 1 ppm pollutant concentrations erosion rates were significantly greater than
controls with oil-base house paint experiencing the largest erosion rate increases, latex and
coil coatings moderate increases, and the industrial maintenance coating and automotive
refinish the smallest increases (Yocom and Grappone 1976; Yocom and Upham 1977; and Campbell
et al., 1974). Coatings that contained extender pigments, particularly calcium carbonate,
showed the greatest erosion rates from the SO^ exposures. Results of field exposures also
support these conclusions (Campbell et al., 1974).
10.2.3 Fabrics
Fibers that suffer destructive action upon exposure to acids derived from S02 include (1)
cellulesic fibers such as cotton and its close relative viscose rayon, a regenerated
cellulose, and cellulose acetate; and (2) polyamide fibers such as nylon 6 and 66. Polyester,
acrylic, and polypropylene fibers are not damaged directly by SO^- However, sulfur dioxide
concentrations can be a source of absorbed acid which can accelerate the fading of dyes and
i
result in fabric deterioration through chemical reactions. The possibility of higher acid
content due to oxidation of sulfur dioxide to sulfur trioxide must be considered (Salvin,
1963).
Brysson et al. (1967) exposed cotton fabrics at 12 different environmental sites in St.,
Louis., MO. and Chicago, IL, for a period of up to one year (1963-64). The seven sites in the
St. Louis metropolitan area represented industrial, urban, suburban, and rural-suburban
SX10AA/A 10-30 12-15-1
-------
JABLE_ig-5._ PAINT. EROSION RAJES AND MEST.PROBABILIIY.DATA
TABLE 10-5a. PAINT EROSION RATES AND PROBABILITY DATA (T-TEST)
FOR CONTROLLED ENVIRONMENTAL LABORATORY EXPOSURES
Mean erosion rate (mil loss X 10 /hour with 95 percent
confidence limits) for unshaded panels and
percent probability that differences exit in parentheses
Type of paint
House paint
oil
latex
Coil coating
Automotive refinish
Industrial maintenance
Clean air
control
20.1 + 7.2
3.5 + 1.5
11.9 + 2.3
1.8 + 0.8
18^6 + 5.1
SO,
(1.0 p£m)
141.0 + 19.0
(99)
11.1 + 1.0
(99)
34.1 +4.7
(99)
3.1 + 2.6
(75)
22.4 + 7.0
(66)
°3
(1.0 ppm)
44.7 + 10.5
(99)
- 8.5 + 5.9
(93)
14.9 +2.5
(94)
5.1 + 1.3
(99)
28.1 + 14.0
(85)
TABLE 10-5b. PAINT EROSION RATES AND PROBABILITY DATA (T-TEST)
FOR FIELD EXPOSURES
Mean erosion rate (mil loss X 10 /month with 95 percent
confidence limits) for panels facing south and
percent probability that differences exist in parentheses
Type of paint
House paint
oil
latex
Coil coating
» *»
Automotive refinish
Industrial maintenance
Rural
(clean air)
4.3 + 7.5
1.8 + 0.5
2.1 + 0.8
0.9 + 1.1
3.6 + 1.6
Suburban
14.8 + 4.9
(9973)
3.0 + 0.7
, (9972)
10.0 +1.9
(9979)
2.3 + 0.7
(9776)
8.2 + 4.2
(9773)
Urban
(SQ7 dominant,
-60 pgAr)
14.2 + 4.9
(9871)
3.8 + 0.3
(9778)
9.5 + 0.8
(9979)
1.6 + 0.4
(8672)
6.6 + 3.9
(9172)
Urban
(oxidant dominant
~40 ug/m )
21.0 + 6.2
(99.2)
6.5 + 5.6
(9473)
8.8 + 1.7
(9979)
1.7 + 0.4
(9176)
7.8 + 2.4
(99.7)
Source: Adapted from Yocom and Upham, 1977.
SX10AA/A
10-31
12-15-80
-------
environments. Four of the Chicago sites represented downtown, industrial, commercial, and
rural environments. An additional site was 20 miles southwest of Chicago at Argonne National
Laboratory.
Suspended particulate matter in the St. Louis area was measured using periodic 24-hour
high volume air samplers. Sulfation values were determined by the lead peroxide candle
method. Monthly dustfall measurements were also used. At the Chicago sites, a high volume
air sampler measured 24-hour total particulate matter 3 days per week and SOp was monitored by
bubbler type absorbers for 24 hours twice a week.
Two fabric types were exposed in this study, a desized and scoured cotton print cloth and
scoured cotton army duck. Two samples were untreated, two dyed, and one resin treated. Study
results indicate that there is a significant relationship between air pollution and both
strength degradation and degree of fabric soiling. As shown in Figure 10-7, high pollutant
2 3
levels (mean sulfation 5 mg SO-/100 cm /day and/or SOp concentrations of 0.2 ppm or 520 ug/m )
reduced effective fabric strength by one-sixth when compared with low pollution sites (0.5 mg
2 3
SO-/100 cm /day and/or 0.02 ppm or 60 ug/m SOp concentrations). The relationship between
suspended particulate matter and fabric strength degradation was not as good as that for SOp.
No correlation between dustfall and strength degradation/effective life was demonstrated and
biological deterioration did not appear to be a major factor in this study (Brysson et al.
1967).
In a review of the Brysson et al. (1967) study, Upham and Salvin (1975) report a
correlation coefficient of 0.95 was obtained for breaking strength versus sulfation for cotton
duck cloth. The correlation coefficient for the thinner cotton print was 0.96. Of the
pollutants measured, SOp was most responsible for causing fabric damage (Upham and Salvin,
1975).
Zeronian (1970) carried out laboratory exposures in which cotton and rayon fabrics were
exposed for seven days to clean air with and without 250 mg/m (0.1 ppm) sulfur dioxide. Both
controlled environments included continuous exposure to artificial light (xenon arc) and a
water spray turned on for 18 minutes every two hours. Loss in strength for all fabrics
exposed to clean air averaged 13 percent, while the fabrics exposed to sulfur dioxide averaged
21 percent. Zeronian et al. (1971) also exposed fabrics made from manmade fibers—nylon,
polyester, and modacrylic~to controlled conditions similar to the cotton exposures, except
o
that the sulfur dioxide level was 486 mg/m (0.2 ppm). They found that only the nylon fabrics
were affected, losing 80 percent of their strength when exposed to sulfur dioxide and only 40
i
percent when exposed in clean air. i
In a study designed to determine effects of air pollution on dye fading on fabrics, 67
dye-fabric combinations were tested. The test samples were exposed in the dark at eleven
sites representing climatic regimes and urban/rural conditions. High temperatures and
humidities appeared to increase fading in the presence of air pollutants. Urban sites
\
SX10AA/A 10-32 12-15-
-------
5.0
4.0
E
o
§
o"
in
r 3.0
2.0
1.0
DUCK CLOTH: 12 months EXPOSURE
CORRELATION COEFFICIENT 0.95
SIGNIFICANT AT 0.5% LEVEL
O -DUCK CLOTH
A-PRINT CLOTH
PRINT CLOTH: 5 monlta EXPOSURE '
CORRELATION COEFFICIENT 0.96
SIGNIFICANT AT 0.5% LEVEL
10
20
30
40
50
BREAKING STRENGTH RETAINED, percent
Figure 10-7a. Relationship between retained breaking strength of
cotton fabrics and corresponding mean su If at ion rate measured at
selected sites in St. Louis area.
Source: Bryssonet al. (1967).
0.25
0.20
< 0.15
£C
I
o 0.10
0.05
I T I I
• PRINT CLOTH, UNTREATED
Q PRINT CLOTH. DYED BLUE
O PRINT CLOTH. DYED ORANGE
A PRINT CLOTH. RESIN TESTED -
20 40 60 80
BREAKING STRENGTH RETAINED, percent
100
Figure 10-7b. Relationship between retained breaking strength of
cotton print cloth samples and mean SO2 concentration for 5-
month exposure at three Chicago sites.
Source: Brysson et al. (1967).
10-33
-------
produced significantly higher fading than rural sites. Fading was highest at the site which
had the highest S02 and TSP concentrations (Beloin, 1972).
10.2.4 Building Materials
The deterioration of inorganic building materials occurs initially through surface
weathering. Moisture and salts are considered to be the most important factors in building
material damage. Many researchers believe that the mechanism of damage from air pollution
involves the formation of salts from reactions in the material. Subsequently, these surface
salts dissolve in moist air and are washed away by rainfall. The components of inorganic
building materials can react with S02 and sulfates (Luckat, 1972; Winkler, 1975; Arnold et al.
1976). Other researchers believe that the role of air pollutants in stone and concrete damage
has frequently been overestimated (Riederer, 1974; Niesel, 1979). These authors report that
inorganic building material damage is dominantly associated with RH >65 percent and
freeze/thaw weathering. Some researchers indicate that microorganisms must also be considered
in order to quantify damage to building materials due to ambient pollutant concentrations
(Winkler, 1966; Riederer, 1974; Krumbein and Lange, 1978; Eckhardt, 1978; Hansen, 1980).
Sulfur chemoautotrophs are well known for the damage they can cause to inorganic materials.
These microorganisms (e.g., Thiobacillus) convert reduced forms of sulfur to sulfuric acid
(Anderson, 1978). However, the relative importance of biological, chemical, and physical
mechanisms have not been systematically investigated. Thus, damage functions definitely
quantifying the relationship of pollutant concentrations to stone and concrete deterioration
are not available in the literature. Air pollution damage to glass is also not presently
quantifiable (Newton, 1974).
10.2.4.1 , Stone—Niesel (1979) has recently completed a literature review concerning the
weathering of building stone in atmospheres containing sulfur oxides, which includes
references from 1700 to the present. In brief, he reports that weathering of porous building
stone containing lime is generally characterized by accumulation of calcium sulfate dihydrate
in the near surface region. The effect of atmospheric pollutants on the rate of weathering is
believed dominantly controlled by the permeability and moisture content of the stone.
Migrating moisture serves primarily as a transport medium. SOg is sorbed and thus can be
translocated internally while being oxidized to sulfates. Reacting components of the building
stone are thus leached, the more soluble compounds inward and the less soluble toward the
surface, often forming a surface crust.
Sengupta and DeGast (1972) also report that S02 sorption causes physical changes in stone
involving changes in porosity and water retention characteristics. Removal of( calcium
carbonate changes the physical nature of the stone surface. The hard, nonporous layer that
forms as a result of alternate freezing and thawing may blister, exfoliate, and separate from
the surface. If the stone contains some substances that are unaffected by S02, the surface
can deteriorate unevenly. The conversion of calcium carbonate into calcium sulfate results in
a type of efflorescence termed crystallization spall ing.
SX10AA/A 10-34
-------
Acidic precipitation also contributes to the weathering process of building ston<=r,
according to Gauri (1979). He reported that marble that is directly exposed to rainfall
undergoes nearly continuous erosion as the acid dissolves the calcium carbonate, allowing
calcite granules to break away and wash off.
10.2.4.2 Cement and Concrete—Portland cement, the major active constituent of concrete, is
manufactured by the high-temperature reaction of a mixture of limestone, alumina, silicates,
and iron salts found in clay. Cement, the binding agent in concrete, is an alkaline material
that reacts with SOp and thus also suffers erosion and spalling effects; it can be protected
by paint. Concrete is also subject to damage by sodium sulfate. Sulfate-resistant cement can
be prepared by reducing the calcium aluminate content. Concrete can also be protected by
nonporous paint.
The chemical action of SO- or sulfates on cement or concrete can be described as a dual
nature mechanism. Calcium hydroxide in cement and concrete can be converted to calcium
sulfate, which reacts to form calcium sulfate aluminum hydrate (ettringite), with a
substantial increase in volume. Cement for dams and culverts requires special formulation for
sulfate resistance when exposed to sulfate concentrations > 200 ppm in water (Nriagu, 1978).
Litvin (1968) examined concrete samples containing Portland cement and marble aggregate
with sand at an industrial site in Buffington, Indiana. Some changes were noted in the marble
aggregate, but a more observable change was found in the cement portion. Sealants were
evaluated as protective coatings; their 'use was accompanied in some cases by surface
efflorescence.
10.2.5 Electrical Equipment and Components
Bobbins (1970) and ITT Electro-Physics Labs (1971) studied the damaging effects of S02
and particles on electronic components and estimated the cost of this damage. The report by
ITT Electro-Physics Labs considered damage to 11 categories of electronic components for which
a literature survey indicated that sulfur dioxide pollution would be mainly responsible.
However, information gained directly from manufacturers indicated that particles were the
major factor in degradation and failure of electronic components and equipment. (See Section
10.3.1) Reduction of S02 and particulate concentrations would have little effect on costs for
the prevention of corrosion; low concentrations of pollutants would still require essentially
the same protective measures. Corrosion-resistant metals are used even in environments where
air .pollution is minimal, since their cost is far outweighed by the expense of equipment
failure.
10.2.6 Paper
Modern papers are manufactured from cellulose. On exposure to acids, paper is hydrolyzed
and loses strength. Spedding et al. (1971), in work with radioactive labeling techniques,
determined that S02 is readily absorbed by paper and oxidized to sulfuric acid by the metallic
impurities in the paper. The reaction may also involve the lignins in the paper, resulting in
the formation of lignosulfonic acids. Walsh et al. (1977) showed that S02 is rapidly absorbed
SX10AA/A 10-35
-------
b\ uncoated wallpaper and less rapidly absorbed by vinyl-coated paper. Although most paper is
used in objects with a short service life, the preservation of documents has been of concern
in museums and archives. Coating paper with polymers impervious to gases is an established
process. The preservation of books in libraries is of major concern. The extent of the
problem is emphasized by the estimate that 50 percent of the books printed between 1900 and
1940 are in need of conservation. The New York Public Library conserves books by
microfilming, lamination, and electrostatic reproduction (Kingery, 1960). The library spent
$900,000 between 1952 and 1967 to microfilm books that had deteriorated (Waddell, 1974).
10.2.7 Leather
Leather has a high capacity for absorbing SO,,. Spedding et al. (1971) reported that the
rate of SO- diffusion to the leather surface is the controlling factor in S00 uptake. The
formation, in the presence of water, of sulfuric acid is followed by hydrolysis of the protein
(collagen) of which leather is principally composed. This weakening of leather causes
cracking and ultimately results in reduction of the leather to a red-brown powder (Spedding et
al., 1971; Yocom and Grappone, 1976).
The destruction of leather by absorption of S02 has long been known and was described in
detail by Prenderleith (1946). The buildup of sulfuric acid in aged leathers correlates with
deterioration, which can be reduced by inactivating the sulfate ion and by pH buffering.
Deterioration of leather is important in bookbinding and in leather upholstery; the use of
artificial leathers has reduced damage costs.
10.2.8 Elastomers and Plastics
The deterioration of natural rubber and synthetic elastomers under weathering conditions
has been studied extensively. Heat, light, oxygen, certain metallic ions, and particularly
ozone cause deterioration; but there is no mention in the literature of S02 damage to rubber.
In fact, rubber is used as an acid-resistant coating. The problem of determining ambient air
pollution effects on rubber is complicated by the presence of ozone, which attacks the double
bonds in both natural rubber and the butadiene-styrene and butadiene-aery!onitrile synthetics.
Haynie et al. (1976) conducted a chamber study on rubber to determine the effects of
ozone, S00, and nitrogen dioxide under controlled conditions of temperature, humidity, and
3
light. Exposures were made at concentrations of 0.1 and 1.0 ppm for each pollutant (in ug/m ;
262 and 2620 for S02, 196 and 1960 for Og and 188 and 1880 for N02). As expected, ozone was
responsible for accelerated cracking of the rubber. Sulfur dioxide did not have any effect.
Verdu (1974) presented a theoretical study of the effect of air pollutants on the
weathering of plastics. He attributed a direct deteriorating effect on plastics to ozone and
suggested that air pollutants such as sulfur dioxide may form active compounds through
photochemical reactions leading to oxidation chain reactions. In light-exposure trials, S02
increased the rate of degradation of polystyrene.
SX10AA/A 10-36 12-15-8J
-------
10.2.9 Works of Art
Although works of art are composed of materials already discussed in earlier sectior -,*
this chapter, they are briefly treated here as a separate category because the cost of tnz
materials involved does not represent the cost of the item.
The deteriorating effects of S02 and particles are well known to museum conservators
whose function is to preserve and restore works of art. The rate of pollutant-related
deterioration has increased markedly in the last 50 years. The damage is striking in Europe,
where ancient buildings, paintings, frescoes, stained glass windows, bronze sculptures, and
marble statuary have suffered deterioration.
Newton (1974) has investigated the cause of deterioration of medieval stained glass
windows. He found that the main cause of decay is the leaching of potassium ions from the
silicate glass by condensed water. Another cause is S02, which produces opaque white crusts
containing CaS04'H20 and syngenate (Ca-CaS04-K2S04-H20). The poor durability of medieval glass
is due to its high content of alkaline earths such as lime and magnesia.
Riederer (1974) conducted a study of the corrosion of bronze sculpture by air pollutants.
Sulfates were found in the corroded surfaces.
The dome of the cathedral in Cologne, located in a highly polluted urban area, has
suffered serious erosion of its sandstone due to the reaction of sulfur acids with calcium
carbonate to form calcium sulfate, which is leached out by rain (Luckat, 1976). Decay and
deterioration of the Taj Mahal in India has also been attributed to sulfur dioxide and other
pollutants emitted from a nearly petrochemical complex (Gajendragakar, 1977).
Damage occurring in Venice, Florence, Rome, Athens, London, and Cologne has been
attributed to the effect of S02 from industrial areas in these cities (Yocum and Upham, 1977).
The United States is also concerned about the deterioration of public buildings and monuments.
The National Bureau of Standards was asked by the National Park Service to investigate methods
for preservation of stone after erosion was noted in the facade of the Lincoln Memorial in
Washington, DC (SIeater, 1977).
SI eater (1977) investigated damage to stone from the action of S02, salt, sodium sulfate,
and light. Conservation materials including epoxy resins, fluorosilicates, and silicone
resins were evaluated. The conservation methods recommended to the National Park Service
varied with the exposure conditions.
The damage to the Acropolis caused by S02 and S03 has resulted in a massive
interdisciplinary effort by the Greek government to protect the ancient buildings from further
deterioration (Yocom, 1979).
10.2.10 Review of Damage Functions Relating SOo to Material Damage
Even the most reliable damage functions must be used with caution. Those listed in Table
10-6 were selected on the basis of their treatment of independent variables and their
inclusion in major literature reviews. Time-of-wetness (often expressed as RH above a
critical value) is the most important variable in these damage functions.
SX10AA/A
10-37 12-15-80
-------
TABLE 10-6. SELECTED PHYSICAL DAMAGE FUNCTIONS RELATED TO S02 EXPOSURE
Material
Reference
Dose-Response relationships
o
CO
00
Zinc
Haynie and Upham, 1970 Y = 0.001028 (RH - 48.8) S0£
Galvanized steel Haynie, 1980
corr = 2.32 tw + 0.0134v0'781S00t
W c W
Oil-base house paint Spence et al., 1975 Y = 14.3 + 0.0151 SOy + 0.388 RH
Enameling steel Haynie and Upham, 1974 corr = 325 t^ e(0<00275 S02 " 163-2/RH>
Weathering steel Haynie et al., 1976 corr = [5.64 VSOj + e(55'44 " 31>150/RT)-j ^~
corr = corrosion, urn
Y = corrosion/erosion rate, um/yr
SOp ug/m oK\
R - gas constant (1.98 cal/gm mol/ '
RH = percent average annual relative humidity
t = time-of-wetness in years
v = wind velocity in m/s
T = °K
t = time of exposure, years
0.92
Galvanized steel Haynie et al., 1976 corr = (0.0187 S09 + e 41'85 " 23»240/RT)t 0.91
£» W
0.61
0.91
-------
Functions for zinc or galvanized steel appear to show the best fit, followed by t-.e
functions for oil-based house paint. The field studies by Haynie and Upham (1970) and Haynie
(1980) and chamber study by Haynie et al. (1976) incorporated critical variables and pro/ided
relatively reliable damage functions for galvanized steel. The functions selected for
weathering and enameling steel and for oil-based paint also utilized these critical
environmental variables.
When these functions are used to estimate damage, other factors must be considered, such
as the amount of exposed (uncoated) metal, the percentage of buildings with oil-based (not
Latex) paint, and temperature and humidity variables (sites in the arid southwest compared
with sites in the relatively humid northeast).
10.3 PARTICULATE MATTER
A report by the National Academy of Sciences (1977) on airborne particles notes that
deposition of dust and soot on building materials not only significantly reduces the esthetic
appeal of structures, but also, either alone or in conce'rt with other environmental factors,
results in direct chemical attack. Because of the paucity of data (see Chapter 5) regarding
total suspended particulate matter (TSP) size distribution and composition, it is difficult to
determine the specific types of particles and chemical constituents that have damaged or
soiled a particular structure. As will be discussed in the following section, chemical
composition of particulate matter is highly important to its corrosiveness; particle size may
be important. Since information on both composition and size is spotty and incomplete, it is
difficult to develop physical damage functions associating either TSP or any particular size
function with materials damage or soiling. Research is under way to address the role of
particle size in soiling of paint.
10.3.1 Corrosion and Erosion
Early studies indicated that suspended particulate matter played a significant role in
metal corrosion. Sanyal and Singhania, writing in 1956, termed the influence of suspended
particulate matter "profound." They ascribed the corrosive effects of particles to (1)
electrolytic, hygroscopic, and/or acidic properties and (2) their ability to sorb corrosive
gases, e.g., sulfur dioxide. It has been pointed out that it is quite difficult to predict
corrosion rates separately for sulfur dioxide and particulate matter since they frequently
coexist at high levels (Chandler and Kilcullen, 1968). Other field studies have established
no conclusive correlation between total suspended particulate matter and corrosion (Mansfeld,
1980; Haynie and Upham, 1974; and Upham, 1967), though further analysis of recent data is
still under way.
At an RH exceeding 60 percent, air polluted with S02 and PM results in a much more rapid
corrosion rate than air polluted with S02 alone (Yocom and Grappone, 1976; Johnson et al.,
1977). Kottori (1980) observed that zinc and galvanized steel corrosion rates appeared to be
related to the sulfate content of TSP. Chloride content of dust also may contribute to
accelerated corrosion of steels (Gibbons, 1970; Bresle, 1976).
SX10AA/A 10-39 12-15-80
-------
Barton (1958) reported that dustfalls contribute to the initial stages of corrosion but
that their influence becomes less important as a layer of rust forms. Two classes of
particulate matter appear to be definitely associated with corrosion: hygroscopic salts
(including those of natural origin such as sodium chloride) and acid smut.
A review of atmospheric factors affecting metal corrosion provides evidence of a
relationship between salinity and corrosion (Guttman and Sereda, 1968).
Corrosion of metals can be accelerated by deposition of particles due to their hygroscopic
nature. The influence of hygroscopic substances on metal corrosion rates has been previously
discussed in section 10.2.1.1.4. As also discussed in section 10.2, particles can disrupt the
protective oxide films formed on metal surfaces such as nickel, copper, aluminum and stainless
steel resulting in pitting (Russell, 1976; NAS, 1977). Russell (1976) noted that airborne
particles often play an important role on the attack by sulfur oxides on electrical contact
surfaces by acting as points for the concentration of active ionic species.
Acid smut is highly corrosive, sticky material formed in and emitted mainly from furnaces
burning liquid fuels containing sulfur, notably in power plants (Ireland, 1968). This
material would not usually be considered suspended particulate matter, as it occurs as
agglomerates of carbon, ash, and sulfuric acid up to 0.5 cm or more in diameter which fall
close to the source (Rowden, 1968; Potter, 1971). According to one report, acid smut mainly
settles out within 400 meters of the source under conditions of light wind (Hoshizawa and
Koyata, 1970).
Japanese investigators analyzed a large (>10 drums [sic]) sample of acid smut and found
the sulfuric acid content to be 30 percent (Oyama et al., 1974). Damage to painted surfaces,
automotive finishes, and even agricultural crops can be substantial. As noted in a review of
residual oil firing problems, "public reaction can be quite severe" (Exley, 1970). A report
on the status of "public nuisances" in the electric power industry of Japan reported progress
in determining the cause of acid smut and in developing preventive techniques (Overseas Public
Nuisance Study Mission, 1965).
Finishes on automobiles parked near industrial sites have often been severely damaged.
Staining and even pitting of auto finishes have been traced to iron particles from nearby
industrial operations. Cars parked near brick buildings being demolished have been damaged by
alkali mortar dust during humid weather. Repainting of damaged auto finishes was often
required because color changes were not reversible by washing or polishing (Fochtman and
Langer, 1957).
Parker (1955) reported that large numbers of black specks collected on freshly painted
buildings in industrial areas. The exterior surfaces of these buildings became distinctly
soiled and required cleaning or repainting in 2 or 3 years, depending on the particulate
concentration in the air. When particulate matter became embedded in the paint film, the
coating was both esthetically and physically damaged. Embedding of particles provides
nucleation sites at which other pollutants can concentrate. Cowling and Roberts (1954)
SX10AA/A 10-40 12-15-8d|
-------
suggest that particles promote the chemical deterioration of paint by acting as wicks to
transfer the S02 corrosive solution to the underlying surface.
10.3.2 Soiling and Discoloration
Soiling is the accumulation of particulate matter on the surface of a material. As
defined by Pfafflin and Ziegler (1976), soiling arises from the deposition of particles less
than 10 urn-by impingement on surfaces to thus mingle with settled dust. Soiling produces a
change in reflection from opaque materials and reduces light transmission through transparent
materials (Beloin and Haynie, 1975; MAS, 1977). Soiling due to airborne particles from
manmade sources results in increased cleaning costs for building and other materials and
reduction in the useful life of fabrics.
10.3.2.1 Building Materials—Under high wind conditions, large particles entrained in the
windstream actually result in a slow erosion of surfaces similiar to sandblasting. Particles
also fill surface pores of many sandstones causing'them to become uniformly darkened.
Particles can contribute to chemical decay of marble, limestone and dolomite stone work, and
concrete structures if they carry acids and soluble salts (NAS, 1977). Dose-response
relationships were developed for suspended particulate matter and various building materials
by Beloin and Haynie (1975).
10.3.2.2 Fabrics—Although particulate matter obviously soils fabrics, researchers have noted
that it is only damaging when the particles are highly abrasive and the fabrics are frequently
flexed. Curtains hanging in open windows, serving as filters in polluted areas, provide a
good example. Weakened as a result of such exposure, curtains often split in parallel lines
along the folds. The more tightly woven the cloth, the more resistant it is to soiling (NAS,
1977).
Because of soiling, fabrics must be washed more often. Excessive washing may reduce
fabric strength, leading to a poorer appearance and concomitantly to shortened life
expectancy. However, sunlight, water vapor, SO , NO and ozone concentrations are believed to
/\ y\
affect the service life of fabrics more significantly. Insolation decoloration is considered
to be the most important service life reduction factor (NAS, 1977). Effects of elevated
sulfur oxide concentrations have been previously discussed.
10.3.2.3 Household and Industrial Paints—Exterior paints can be soiled by liquids and by
solid particles composed of soot, tarry acids, and various other constituents. Beloin and
Haynie (1975) have compared the rates of soiling by different levels of TSP on six different
building materials over a 2-year period. The mean annual PM concentrations at the five study
sites ranged from 60 ug/m3 for a rural residential location to 250 ug/m3 for an industrial
residential environment. The exposed materials included painted cedar siding, concrete block,
brick, limestone, asphalt shingles, and window glass. As determined by reflectance
measurements, the degree of soiling of painted surfaces was directly proportional to the
square root of the PM dose, accounting for 74 to 90 percent of the measured variability. The
results of regression for soiling of building materials as a function of TSP dose is included
in Table 10-7. As one example, the linear regression results of soiling of acrylic emulsion
SX10AA/A 1Q-41
-------
TABLE 10-7. RESULTS OF REGRESSION FOR SOILING OF BUILDING MATERIALS AS A FUNCTION OF TSP DOSE
o
ro
Material
Oil base paint
Tint base paint
Sheltered acrylic
emulsion paint
Acrylic emulsion
paint
Shingles
Concrete
Coated limestone
Uncoated limestone
Coated red brick
Uncoated red brick
Coated yellow brick
Uncoated yellow brick
Glass
400
400
400
720
48
160
80
80
80
80
80
80
45
89.43
86.13
91.54
90.79
43.50
41.75
44.57
46.99
12.95
14.88
45.05
43.21
0.2806
-0.2768
-0.2618
-0.593
-0.4131
-0.199
-0.0458
+0.0779
-0.0503
-0.0296
-0.0374
-0.1133
-0.1133
+0.0314
0.0641
0.0571
0.1156
0.0497
0.5771
0.1338
0.2464
0.1500
0.0223
0.0331
0.5337
0. 2740
0.008077
0.000069
0.000061
0.000123
0.000026
0.000258
0.000080
0.000164
0.000089
0.000013
0.000020
0.000317
0.000168
0.000007
7.6510
6.8265
13.8143
8.3791
7.6992
7.5011
6.9046
4.2035
0.6255
0.9274
14. 9533
7.6773
0.6851
0.745
0.738
0.880
0.902
0.769
0.143
0.347
0.266
0.459
0.477
0.342
0.503
0.340
A,
B,
Note: Equation used in this regression analysis was reflectance = B(TSP x months of exposure/5 + A.
N, Number of data sets (dependent upon the number of controlled variables in the factorial experiment).
Intercept of linear regression.
Slope of linear regression.
Estimated variance of intercept.
Estimated variance of slope.
Residual variance (error).
R*", Correlation index (fraction of variability accounted for by regression).
Source: Abstracted from Beloin and Haynie, 1975.
-------
house paint by exposure to suspended particulate matter is shown below: (640 samples, R2 =
0.896)
ln(92.5 Y) = -0.311 + 0.345 ln(SP) + 0.612 ln(t)
Where: 92.5 = Initial reflectance value (80 samples)
Y = Measured percent reflectance (Photvolt Model 625)
SP = Average TSP concentration (ug/m3)
t = Exposure time (months)
Based on this equation, Figure 10-8 was prepared to summarize soiling of acrylic emulsion
house paint as a function of exposure time and particulate concentrations. Although it is
recognized that socio-economic factors control ability and motivation to maintain clean
surfaces (Beloin and Haynie 1975), estimations of repainting frequencies can be formulated.
Assuming an individual responds to a defined change in reflectance of house paint by
repainting, that person will repaint a house twice as5 often in an environment with a TSP
concentration of 260 pg/m in comparison to 75 ug/m . Specifically, at a 35 percent change
this houseowner would repaint every 4 years when TSP is 75 ug/m and every 2 years when 260
ug/m .
10.4 SUMMARY, PHYSICAL EFFECTS OF SULFUR OXIDES AND PARTICULATE MATTER ON MATERIALS
Reviewing the facts presented in Sections 10.2 and 10.3, it should be clear that the best
data base for association of sulfur oxides to materials effects is corrosion of metal and that
the parallel case for particulate matter is soiling. Of the damage functions developed for
corrosion of metals by sulfur oxides, the one for zinc appears to show the best fit. Slightly
less well established are relationships for sulfur oxide exposure and corrosion of other
metals. There is evidence that particulate matter aggravates corrosion, especially when
coexistent with sulfur oxides. This is most likely due to the hygroscopic salt content of the
particulate matter. However, no mathematical expression of the relative contribution of
particulate matter to corrosion of metal has been established.
Both particulate matter and sulfur oxides have been shown to damage paint. Damage
functions for erosion of various paint by sulfur oxides and for soiling of paint by
parti cul ate matter have been developed. The varying properties of the several types of paint
employed for exposed materials make it difficult to construct mathematical expressions for
effect of particulate matter and/or S02 on all paints.
Building materials also are eroded by sulfur oxides and soiled by particulate matter.
That damage is attributable to both pollutants is established; quantitative general
relationships between ambient concentration and effect are, however, lacking. For erosion of
building materials, particularly stone and concrete, the contribution of sulfur oxides
relative to other agents is not clear. As for soiling, though some damage functions have been
developed, the lack of understanding of the role of particle size and composition makes it
difficult to generalize about soiling effects of all suspended particulate matter on all
building materials.
SX10AA/A
-------
45
40
Ul
-------
Concerning effects of particulate matter and/or sulfur oxides on other materials, the
data base is even less well established. There is evidence to show that damage occurs to
fabrics, leather, paper, glass, and plastic, and to works of art composed of one or many
materials. This evidence is mainly qualitative and spotty. For these materials, there are
insufficient data to develop reliable estimates of effects attributable to specific ambient
pollutant concentrations.
10.5 ECONOMIC DAMAGE OF AIR POLLUTION TO MATERIALS—SULFUR OXIDES AND PARTICULATE MATTER
10.5.1 Introduction
The task of estimating costs of air pollution effects on materials involves treatment of
many variables, some of which are difficult to quantify and relate to other factors bearing on
costs. Several approaches to the problem have been reported. These are: (1) the physical
damage function approach, (2) the non-market approach, and (3) the indirect market approach.
The goal of all of the approaches is to estimate the total monetary damage caused by air
pollution, in this instance, by particulate matter and/or sulfur oxides. Monetary damage, as
defined by Maler and Wyzga (1976), is "the total compensating variation of a change that is
detrimental to the environment." Monetary damage comprises two subsets of loss: financial
loss is "that part of the monetary damage which is realized by changes in expenditure, while
amenity loss is the difference between monetary damage and financial loss."
There are several types of financial losses resulting from damage to materials and
soiling: (1) reduced service life of a material, (2) decreased utility of a material, (3)
necessity for use of suitable substitute materials, (4) losses due to an inferior substitute,
(5) protection of suspectible materials, and (6) additional required maintenance, including
cleaning. The major losses of amenity, as defined by Maler and Wyzga, are "those of enduring
and suffering soiled, damaged, or inferior products and materials because of pollution," in
this case particulate matter and/or sulfur oxides. In addition, amenity losses are suffered
when pollution damage repair or maintenance procedures results in inconvenience or other
delays in normal operations. Some of these losses, such as effects on monuments and works of
art, are especially difficult to quantify (Maler and Wyzga, 1976).
In calculating monetary damage, the approach selected will vary depending upon whether
financial losses or losses of amenity are to be emphasized, the type of damage being
considered, and the availability of information. Regarding pollutant effects on materials,
the literature has been dominated by calculations of financial loss based on the physical
damage function approach. Sources of error in financial loss functions using this approach
include the following: (1) present techniques do not reflect the recent introduction of
resistant materials which last longer and require less maintenance; (2) estimates assume that
galvanized steels are left uncoated; (3) many materials may wear out before they are
significantly damaged by pollutants (Glass, 1978).
One area not dominated by the physical damage approach has been the estimation of
monetary damage associated with soiling. In these studies, loss of amenity has been
considered, since socioeconomic variables are heavily involved. The approaches eaployed
SX10AA/A 10-45 12-15-80
-------
reflect the shift in emphasis. Non-market and indirect market approaches have been used to
relate changes in air quality to changes in the amount of money people are willing to pay for
improvement in air quality (reduction in air pollution). Documentation of this relationship
may be sought through survey of affected individuals or through development of relationships
between environmental quality and available data on price differentials. A major source of
error in these approaches is that they demand that all factors other than air quality that
affect cost must be accounted for. It is also very difficult to distinguish among the effects
of many different air pollutants.
Recently, the nonmarket and indirect market approaches have been increasingly employed to
estimate both financial loss and loss of amenity. To date, such studies have not been found
more suitable for regulatory decision-making guidance than the body of literature resulting
from employment of the more traditional (physical damage function) approach. All approaches
are limited by the difficulty in quantification of the human response to damage based upon the
ability and the incentive to pay additional costs. (Yocom and Grappone, 1976).
10.5.2 Economic Damage to Materials
10.5.2.1 Metals—Realistic estimation of the economic damage to metals attributable to sulfur
oxides and particles must take into account several factors, including avoidance costs, such
as the costs of specific protective treatment. For metals, these costs include the use of
anticorrosive primers, the practice of sandblasting before painting, and the use of paints
that are resistant to acid.
A recent report published by the U.S. Department of Commerce (Bennett et al., 1978)
examined the cost of corrosion in the United States in 1975. Unfortunately, the developed
damage costs were not pollutant-specific and were not associated with ambient concentrations
of pollutants. A total annual metallic corrosion cost of $82 billion was estimated with a
model which incorporated a broad range of cost items (e.g., materials, labor, energy, and
technical capabilities). About 40 percent of this cost, or $33 billion, was considered
avoidable. Within this avoidable cost is, of course, the cost of air pollution, a portion of
which is in turn the cost of metallic corrosion resulting from particulate matter and/or
sulfur oxides. The figure of $33 billion is thus only useful as an upper limit for present
purposes.
Fink et al. (1971) estimated that corrosion of external metal structures caused by air
pollution costs $1.45 billion annually in the United States, as shown in Table 10-8. As is
the case for the extensive Bennett et al. (1978) Department of Commerce report, these studies
also were not specific to single pollutants nor were the damage costs associated directlyjwith
ambient pollutant concentrations. Furthermore, in some cases material damage resulting from
causes other than air pollution was included in cost estimates (e.g., the Fink et al. study
included corrosion inside pipes of industrial systems). On the other hand, Haynie (1974)
noted that within Fink's estimate of metal corrosion costs, damage to structural systems
primarily constructed of galvanized steel accounted for more than 90 percent of the cost.
Haynie reasoned that, based on Fink's data, the accelerated corrosion of zinc by sulfur
dioxide accounts for more than 90 percent of corrosion caused by air pollutants.
SX10AA/A 10-46 12-15-80
-------
TABLE 10-8. SUMMATION OF ANNUAL EXTRA LOSSES DUE TO CORROSION DAMAGE BY AIR POLLUTION
TO EXTERNAL METAL STRUCTURES FOR 1970
Steel system or structure
Basis for calculation
Annual
loss in $1000
Steel storage tanks
Highway and rail bridges
Power transformers
Street lighting fixtures
Outdoor metal work
Pole-line hardware
Chain-link fencing
Galvanized wire and rope
Transmission towers
Maintenance
Maintenance
Maintenance
Maintenance
Maintenance
Replacement
Maintenance and replacement
Replacement
Maintenance
$ 46,310
30,400
7,450
11,910
914,015
161,000
165,800
111,800
1,480
$1,450,165
Source: Fink et al. (1971).
-------
The economic damage from corrosion can also be estimated by determining the annual cost
of industrial paints used for corrosion control. This cost is estimated by totaling the costs
of surface preparation, paint, and labor. Coatings applied at this combined cost would
protect metal structures against corrosion for about 10 years, whereas "ordinary" coatings of
paint would provide protection for only 2 years. Painting of structural steel in bridges was
investigated by Moore and O'Leary; (1975) the practice involves sandblasting the steel to
produce a rust-free surface and to remove mill scale. Without such surface preparation, water
is immediately absorbed and sets up a corrosion system, rusting occurs, and the paint surface
deteriorates in 2 to 3 years. The metal surface is protected by primer that inhibits rust
formation, and the primer coat is covered with two coats of SQp-resistant paint, such as vinyl
resin, which is substantially more expensive than household paint. Estimates have been
developed by Banov (1973), Michelson and Tourin (1967), and others to reflect these
preparation and painting costs for protection of metal structures. However, sound fractional
allocation of these costs to sulfur oxide material damage is not available in the literature.
Damage from pollutants during the manufacture of electrical components must be
controlled even in clean environments. To prevent such problems, parts are fabricated in
"clean rooms" with filtered air. If equipment malfunctions in the plant or in service,
additional maintenance costs are' incurred for cleaning, repair, or replacement of defective
equipment. ITT (1971) estimated $15.5 million per year in added costs for clean rooms and
maintenance. Robbins (1970) conducted a survey of the effects of S0? and particles on
electrical contacts such as in switches, relays, connectors, and computers. To reduce
corrosion, contacts are electroplated with corrosion-resistant metals such as gold, platinum,
palladium, and silver. Less expensive metals are susceptible to corrosion failure, mostly
from the action of SOp and HpS. Robbins estimated that 15 percent of the gold and platinum
used in the United States for electrical contacts in 1970 was for the specific purpose of
combating SO^ corrosion, with the remainder going for protection against other environmental
pollutants. However, protection against all deleterious environmental conditions is routinely
provided in clean environments, due to high costs of replacement associated with electrical
system downtime. Thus, costs for protection of electrical hardware cannot be allocated
directly to ambient sulfur oxide and/or particulate matter concentrations.
10.5.2.2 Paints—Spence and Haynie (1972) presented a survey and economic assessment of the
deterioration of exterior paints ("trade paints") caused by air pollution. Included in this
category were both oil-base paints and latex paints containing polyvinyl acetate-acrylic as
the binder. The total annual economic damage to exterior household paints was estimated at
$540 million (1972 prices), including paint loss and a labor factor of three times the cost of
/
the paint. Unfortunately, damage functions were not available for use in developing this
estimate. Furthermore, the estimated damage cost reflects the impacts of all pollutants; no
specific attribution of costs to particulate matter or sulfur oxides is possible.
SX10AA/A 10-48 12-15-80
-------
Salmon (1970) estimated that the annual cost of soiling of household paint would be S35
billion if surfaces were maintained as clean as they are in a clean environment in this
qualitative study. The annual cost of deterioration damage to paints was estimated to be SI.2
billion.
Michelson and Tourin (1967) investigated the frequency of house repainting as a function
of suspended particulate concentration. Questionnaires were sent to residents of three
suburbs of Washington, DC (Suitland, Rockville, and Fairfax) and two cities in the upper Ohio
Valley (Steubenville and Uniontown). Data were compiled from the questionnaires to show
maintenance intervals for exterior repainting in each of the five communities, but paint types
were not reported. In Steubenville, where the mean annual particulate concentration was 235
ug/m , repainting occurred about every year. In Fairfax, where the mean annual particulate
concentration was 60 ug/m , repainting occurred every 4 years. Thus, maintenance frequency
increased as particulate concentration increased. The results of this investigation suggest
that a significant relationship exists between frequency of repainting and particulate
concentration. However, to establish a more definite correlation, additional maintenance data
are needed, particularly for cities with mean annual particulate concentrations greater than
2
150 M9/ra • Any correlation of frequency of repainting with concentration of particulate
matter must take into account the fact that other pollutants are usually present in high
concentration where particle counts are high. Socioeconomic factors also must be considered.
Booz, Allen and Hamilton (1970), in a study conducted for EPA, reported on painting
maintenance frequencies in several zones of the Philadelphia metropolitan area with different
population characteristics, climates, and types of industry. Socioeconomic factors were
delineated by pollution zone; however, paint types were not reported. The percentage of
households with incomes of less than $6000 increased with pollution level, a finding that may
partially explain why there was no statistically significant correlation between painting
frequency and particulate level.
10.5.2.3 Economic Cost of Soiling—Studies on soiling have in some approaches been limited to
household cleaning costs. Other studies employ differences in property values or willingness
to pay as the basis for cost estimates. In the study by Booz, Allen and Hamilton (1970), $5
billion for annual household cleaning costs was attributed to particulate pollution. This
amount did not include laundering, dry cleaning, and personal care items (i.e., face and
hair). The report has been criticized because it omits some sources of damage that might be
responsible for extra maintenance. There is doubt as to which cleaning and maintenance
operations are really sensitive to air particulate levels. Also, the labor of the homemaker
was not added to the cost of the cleaning materials, and the number of urban homes surveyed
was small.
Watson and Jaksch (1978) estimated physical soiling and frequency-of-cleaning functions
using the study of Beloin and Haynie (1975) to establish the relationship between level of
particulate matter and soiling rate and the work of Esmen (1973) to estimate various levels of
SX10AA/A
10-49 12-15-80
-------
particulate soiling for rural, suburban, and industrial locations. Watson and Jaksch (1978)
reported that the frequency of the low-cost cleaning operations depended on the level of
airborne particles but that the frequency of professional house painting was unaffected by
variation in airborne particle levels.
Watson and Jaksch (1978) concluded that 537 to 3816 million dollars (1971 dollars) can be
saved per year in the United States in the form of net welfare benefits if the secondary TSP
standard of 60 mg/m can be met. Watson and Jaksch (1978) utilized data from a review report
by Waddell (1974). Waddell utilized studies that were not specific to particulate matter
pollution and do not contain dependable material exposure allocations (e.g., Gillette, 1975).
Further, Watson and Jaksch (1978) determined frequency-of-cleaning functions from the Booz,
Allen and Hamilton study (1970), which surveyed 1090 households in Philadelphia during 1969.
The Beaver Report (1954) suggested an annual total for damage by all forms of air
pollution in Great Britain of 152 million pounds sterling in direct costs, of which 25 million
was for laundry, 30 million for painting and decorating, and 20 million for cleaning and
depreciation of buildings other than houses; thus, about half the total cost of pollution was
attributed to soiling.
Michelson and Tourin (1967) compared the costs of air pollution in the highly polluted
area of Steubenville, OH, with those in the relatively clean Uniontown, PA, area. The per
capita costs for inside and outside maintenance of houses (painting and cleaning), laundry,
dry cleaning, and personal care (hair and facial) were $84 higher in Steubenville than in
Uniontown. Cost figures were based on data obtained from questionnaires, but the results have
been questioned on the grounds that socioeconomic factors influenced responses and that there
were insufficient statistically reliable data.
Narayan and Lancaster (1973) conducted a questionnaire survey in a rural area and a
polluted area in New South Wales, Australia, to determine the difference in cost of household
upkeep. The cost of maintaining a house in the polluted Mayfield area was about $90 per year
higher than in the relatively unpolluted Rotar area. This cost differential was attributed to
the higher levels of air pollution and airborne particulate matter in Mayfield; however, the
accuracy of the cost data was considered questionable, since the attitude of the respondents
could have introduced a biased point of view.
Waddell (1974), in his review of the economic damage of air pollution, noted that the
Michelson and Tourin (1967) and Booz, Allen and Hamilton (1970) studies considered principally
the costs of household cleaning and maintenance. He questioned the validity of extrapolating
the Michelson and Tourin value of $84 per capita because of insufficient information and
socioeconomic problems in obtaining defensible data. He further asserted that "intuitively,
other than what is implicitly measured in property value differentials, it is difficult to
conclude that there are not significant soiling-related costs." In particular, he stated
that soiling costs should include laundering, dry cleaning, hair and facial care, washing of
automobiles, and costs of cleaning commercial establishments and public structures. Waddell
(1974) concluded that there are insufficient data to assess these soiling costs.
I
SX10AA/A 10-50 12-15-^0
-------
Waddell (1974) also reviewed the concept of property value as an estimator, noting f,e
assumption that the inconvenience of living in a polluted area with soiling and odors leads to
lower property values. He cited the Jacksch and Stoevener (1970) study in Toledo, Oregon,
using dustfall measurements as the variable. Their hypothesis was that air pollution costs,
though not quantitatively known, were reflected in the value of property. They found that
reduction in property values from increasing air pollution was greater in the higher priced,
newer sections of town than in low-cost housing.
Most property value approaches assume a direct and equal relationship between property
values and air quality. In other words, "if two sites are similar in all respects except air
quality, the difference in their values represents the market's willingness to pay for
reduction in air pollution dosages. That is, all air pollution damages will be registered on
differential site values (Crocker, 1970). However, using areas with similar socioeconomlc
characteristics Gillies and Youston (1976) found no -statistical relationship between air
pollution and property values.
In a mere recent study of paired neighborhoods in the Los Angeles area employing both
differences in property value and a willingness to pay survey, Brookshire et al. (1979a,b) found
that the marginal willingness to pay for an approximate 30 percent improvement in air quality
calculated from property values and the survey results ($950 million and $650 million
respectively) were within a factor of 1.5. The study was, however, not pollutant-specific,
and the authors noted that this approach to benefit analysis needs further refinement before
the results "can even cautiously be applied to environmental decision making." They also
cautioned that the study may well not be generalizeable to other situations.
The economic effects of soiling were examined by Liu and Yu (1976) in a Midwest Research
Institute project undertaken for the EPA's Corvallis Environmental Research Laboratory. The
objectives were to generate physical and economic damage functions, by receptor, for SOg and
suspended particles and to establish cost/benefit relationships. The study included the
effects of air pollution damage on health, household soiling, materials, and vegetation.
Liu and Yu (1976) used the data of Booz, Allen and Hamilton (1970) on soiling and
maintenance operations to identify and quantify soiling damage. They surveyed costs of nine
professional cleaning companies in Kansas City, and developed a methodology for use on a
nationwide basis for estimating damage attributable to S02 and particulate matter. They
calculated the gross cost of cleaning and the net (extra) cost from increases in suspended
particles using a formula they derived. For these calculations, they chose a suspended
particulate level of 45 ug/m3 as the level necessitating extra cleaning.
Liu and Yu (1976) derived net and gross household cleaning costs for 65 large cities with
populations greater than 500,000 and for 83 medium-size cities of 200,000 to 500,000. The
extra costs for Chicago, New York, and Los Angeles in 1970 were $516 million, $418 million,
and $388 million, respectively. Per capita costs for household maintenance activities ranged
from $5 per person in San Antonio, Texas, to $104 per person in Cleveland, Ohio. The total
net soilage cost for the urban areas was estimated as $5.033 billion.
SX10AA/A
10-51 12-15-80
-------
10.5.2.4 Combined Studies—Salmon (1970) was one of the first to provide an estimate of
material damage associated with ambient SOp concentrations. He calculated economic loss by
determining the values of materials exposed to pollution and then multiplying by the estimated
difference in useful lifetime between clean-rural and polluted-urban areas. The value of
exposed materials was derived by multiplying annual production by a product lifetime estimate
and then applying a labor-factor estimate and an exposure-factor estimate. Salmon cited
economic damage from sulfur oxides in the United States to the following materials, listed in
decreasing order of the extent of damage: metals, cotton, finishes and other coatings,
building stone, paints, paper, and leather. Paint, zinc, and cement/concrete accounted for 70
percent ($2.647 billion) of the estimated annual economic loss of all major materials ($3.8
billion). Deterioration of paints alone was estimated at $1.2 billion. Salmon noted that it
would cost $35 billion a year to keep household painted surfaces as clean as they would be in
an unpolluted environment. Carbon steel, which has virtually no resistance to SQy and
sulfate, accounted for $54 million of the total loss.
Gillette (1975) reported significant reductions in economic damage to materials from
sulfur oxides attributable to improvements in existing air quality levels throughout the
United States. Comparing annual S02 concentrations from more than 200 monitoring sites with
the estimated inventory of materials exposed in the proximity of these sites, he estimated
that nationwide material damage decreased from more than $900 million in 1968 to less than
$100 million in 1972. These estimates were derived by carefully distinguishing between
physical and economic losses and by attributing current estimates of losses to current
exposure levels.
The distinction between physical and economic damage to materials was based upon the
concept of normal or economic useful life of materials. Whereas physical deterioration to
materials may occur at relatively low exposure levels, economic losses will occur only if the
material requires early replacement or increased maintenance before its normal or economically
useful life is spent. Given the prevailing ambient concentrations observed, Gillette reported
that most materials were not adversely affected economically except for metallic products that
were subjected to corrosion or paint damage. While material losses were much greater during
the early 1960s, the losses in more recent years are substantially lower and reflect the
considerable improvement in air quality (Gillette, 1975).
Waddell assumed that the Gillette study is more defensible than the Fink study, that it
does not significantly overlap the Spence-Haynie study, and that it includes damage to
electrical contacts and components. If the materials evaluated in these studies (zinc,
paints, rubber,%carbon and alloy steel, fibers, cement and concrete, plastics, building brick,
paper, leather, wood, and building stone) plus those believed by Fink to be relatively
unaffected by air pollution (aluminum, copper, stainless steel, and lead) are subtracted from
the Salmon study, a total remainder from the Salmon study of $0.4 billion is obtained. Using
these data and the assumptions outlined above, Waddell derived an estimate of 1970 air
SX10AA/A 10-52 12-15-
-------
TABLE 10-9. ESTIMATES OF TOTAL COSTS FROM AIR POLLUTION
DAMAGE TO MATERIALS IN 1970
Source of
information
Mueller-Stickney
Gillette
Material
category
Elastomers
General materials
Principal
pollutants
Ozone
S00
Estimated cost,
$ billion
0.5
0.4
Salvin
Spence-Haynie
Salmon
Total
but principally
metal corrosion
Textiles and dyes
Paints
Remainder of
important materials
not accounted for
above
Particulates, S0?
Various, but
probably
mostly S02
0.2
0.7
0.4
2.2
Source: Waddell (1974).•
pollution damage costs of $2.2 billion. As shown in Table 10-9, $1.5 billion was directly
attributed to PM and SO .
Yocum and Grappone (1976) also estimated that the economic costs of air pollution damage
to materials for the United States was about $2.2 billion in 1970. They estimated the
relative contribution to this total cost from SOX at 41 percent (about $0.9 billion);
particulate matter, 27 percent (about $0.6 billion); ozone, 26 percent; and NOX> 6 percent.
10.5.3 Summary of Economic Damage of Air Pollution to Materials
The damaging and soiling of materials by airborne pollutants have an economic impact, but
this impact is difficult to quantify. The accuracy of economic damage functions is limited by
several factors. One of the problems has been to separate costs related to sulfur oxides and
particles from those related to other pollutants, as well as those related to normal
maintenance. Cost studies typically involve broad assumptions about the kinds of materials
that are exposed in a given area and then require complex statistical analysis to account for
a selected number of variables. Attitudes regarding maintenance may vary culturally, further
confounding the problem of quantifying economic impact.
Studies have used various approaches to determine pollutant-related costs for extra
cleaning, early replacement, more frequent painting, and protective coating of susceptible
materials, as well as other indicators "of the adverse economic effects of pollutants. Ho
SX10AA/A
10-53
12-15-80
-------
study has produced completely satisfactory results, and estimates of cost vary widely (see
selected examples in Figure 10-9). Studies of household maintenance costs in polluted and
unpolluted areas have shown annual cost differentials of $100 or more, but the results of
these various studies are difficult to compare because of differences in approach,
methodology, and selection of variables. Damage functions indicate that reductions in
pollutants will decrease physical and therefore economic damage, but the data base and
methodology for attribution of costs to sulfur oxides and particulate matter are incomplete at
this time.
10.6 SUMMARY AND CONCLUSIONS, EFFECTS ON MATERIALS
The nature and extent of damage to materials by sulfur oxides and particulate matter have
been investigated by field and laboratory studies. Both physical and economic damage
functions have been developed for specific damage/effect parameters associated with exposure
to these pollutants. To date, only a few of these functions are relatively reliable in
determining damage, while none have been generally accepted for estimating costs.
The best documented and most significant damage from sulfur oxides and particulate matter
is the acceleration of metal corrosion, the erosion and soiling of paint and soiling of
buildings and other structures. Erosion of building materials and stone due to sulfur oxides
is also established, but the importance of sulfur oxides relative to other agents is not
clear. Although evidence of damage to fibers (cotton and nylon), paper, leather and
electrical components has been reported, reliable damage estimates have not.
Relatively accurate physical damage functions have been calculated for the effects of
sulfur dioxide on the corrosion of galvanized steel. Determination of variables such as time
of wetness and surface configuration affect the applicability of the functions. Similar, but
less accurate, functions have also been developed for estimating erosion rates of oil-based
paints from exposure to sulfur dioxide. The least reliable of the "significant" damage
functions are those for soiling from particulate matter. The poorly understood deposition
rates and poorly characterized chemical and physical properties make general application of
the functions difficult, if not impossible.
The limitations of these and other physical damage functions hinder accurate estimates of
total material damage and soiling. Coupled with these limitations is the lack of material
exposure estimates. These problems presently preclude complete and accurate estimates of the
costs of damage based on a physical damage function approach. Other approaches to cost
estimation with lesser data requirements have been attempted, but the studies using these
approaches to date have been inadequate for decision-making guidance.
SX10AA/A 10-54 12-15-^0
-------
POLLUTANTS
MOST
POLLUTANTS
SOX AND TSP
TSP
SOX
1.45
FINK
(1971)
METALS
0.54
SPENCE AND HAYNIE
(1972)
1.2
SALMON
H970)
PAINT
2.2
WADDELL (1974) AND
YOCOM AND GRAPPONE
(1976)
1J5
WADDELL (1974) AND
YOCOM AND GRAPPONE
(1976)
5.0
LIU AND YU (1976) AND
BOOZ-ALLEN (1970)
0.54 - 3.8
WATSON AND JACKSCH
(1978)
0.1-0.9 3.8
GILLETTE SALMON
(1975) (1970)
MOST MATERIALS
MATERIAL TYPE
Figure 10-9. Summary of economic damage estimates of air pollution to materials.
(in billions of dollars per year, early 1970's)
10-55
-------
10.6 REFERENCES
Abe, H., Y. Ishii, and H. Kato. Evaluation of atmospheric factors by analyses of corrosion
products and surface deposits on copper plates. Rail. Tech. Res. Inst. 12:170-174, 1971.
Anderson, J. W. Sulphur in Biology. Studies in Biology No. 101, University Park Press,
Baltimore, MD, 1978. pp. 17-19.
Arnold, L., D. B. Honeyborne, and C. A. Price. Conservation of natural stone. Chem. Ind.
(London) 8:345-347, 1976.
Banov, A. Paintings and Coatings. Structures Publishing Co., Farmington, MI, 1973.
Barton, J. R. Protection Against Atmospheric Corrosion, Theories and Methods. John Wiley &
Sons, New York, NY, 1976.
Barton, K. The influence of dust on atmospheric corrosion of metals. Werkst. Korros.
8/9:547-549, 1958.
Beaver, H., Chairman. Committee on Air Pollution: Report. Her Majesty's Stationery Office,
London, England, 1954.
Beloin, N. J. Fading of dyed fabrics .by air pollution. Text. Chem. Color. 4:77-82, 1972.
-.•
Beloin, N. J., and F. H. Haynie. Soiling of building materials. J. Air Pollut. Control
Assoc. 25:399-403, 1975.
Bennett, L. H., J. Kruger, R. L. Parker, E. Passaglia, C. Reimann, A. W. Ruff, and H.
Yakowitz. Economic Effects of Metallic Corrosion in the United States. Part I: A
Report to the Congress by the National Bureau of Standards. NBS Special Publication
511-1, U. S. Department of Commerce, National Bureau of Standards, Washington, DC, May
1978.
Bird, C. E. Materials Protection and Performance, 16, April 1977.
Booz, Allen and Hamilton, Inc. Study to Determine Residential Soiling Costs of Particulate
Air Pollution. APTD-0715, U.S. Department of Health, Education and Welfare, National Air
Pollution Control Administration, Raleigh, NC, October 1970.
Brauns, E., and U. Kalla. Korrosion 17:81, 1965.
Bresle, A. The corrosion of steel and the dangerous chlorides. Met. Finish. 74:23-25, 30,
1976. ~
Brookshire, D. S., T. D. Crocker, and R. C. D'Arge. Methods Development for Assessing Air
Pollution Control Benefits. Vol. V. EPA-600/5-79-001e, U.S. Environmental Protection
Agency, Washington, DC, February, 1979a.
Brookshire, D. S., R. C. D'Arge, W. D. Schulze. Methods Development for Assessing Air
Pollution Control Benefits. Vol. II: Experiments on Valuing Non-Market Goods: fy case
Study of Alternative Benefit Measures of Air Pollution Control in the South Coast Air
Basin of Southern California. EPA-600/5-79-001b, U.S. Environmental Protection Agency,
Washington, DC, 1979b.
Brysson, R. S., B. J. Trask, J. B. Upham, and S. G. Booras. The effects of air pollution on
exposed cotton fabrics. J. Air Pollut. Control Assoc. 17:294-298, 1967.
Campbell, G. G., G. G. Shurr, D. E. Slawikowski, and J. W. Spence. Assessing air pollution
damage to coatings. J. Paint Techno!. 46:59-71, 1974.
i
SOX10A/D 10-56 12-15-86
-------
Campbell, G. C. , G. G. Schurr, and D. E. Slawlkowski . A Study to Evaluate Techniques -..*
Assessing Air Pollution Damage to Paints. EPA-R3-73-040, U.S. Environmental Protect! c-
Agency, Research Triangle Park, NC, May 1972.
Cavender, J. H. , W. M. Cox, M. Georgevitch, N. A. Huey, G. A. Jutze, and C. E. Zimmer.
Interstate Surveillance Project: Measurement of Air Pollution Using Static Monitors.
APTD-0666, U.S. Environmental Protection Agency, Research Triangle Park, NC, May 1971.
Chandler, K. A., and M. B. Kilcullen. Survey of corrosion and atmospheric pollution in and
around Sheffield. Br. Corros. J. 3:80-84, 1968.
Cowling, J. E. , and M. E. Roberts. Paints, varnishes, enamels, and lacquers. In:
Deterioration of Materials-Causes and Preventive Techniques. G. A. Greathouse and C. J.
Wessel, eds. , Reinhold Publishing Corp., New York, NY, 1954. pp. 596-645.
Crocker, T. D. Urban Air Pollution Damage Functions: Theory and Measurement. California
University, Riverside, CA, 1970.
Duncan, J. R. , and D. J. Spedding. The mode of initial reaction of SO, at a metal surface.
Corros. Sci. 14:241-249, 1974. *
Eckhardt, F. G. W. Microorganisms and weathering of a sandstone monument. In: Environmental
Biogeochemistry and Geochemistry Volume 2: The Terrestrial Environment, Proceedings of
the Third International Symposium, Ministry of Sciences and Arts of Niedersachsen,
Deutsche Forschungsgemeinschaft, Wolfenburetell , Germany, 1977. W. E. Krumbein, ed. , Ann
Arbor Science Publishers, Inc., Ann Arbor, MI, 1978. pp. 675-686.
Esmen, N. A. A direct measurement method for dustfall. J. Air Pollut. Control. Assoc.
23:34-36, 1973.
Evans, U.R. Mechanism of rusting under different conditions. Br. Corros. J. 7:10-14, 1972.
Ergang, R. , and M. B. Rockel. Corrosion resistance of stainless steels in the
atmosphere — evaluation of the results of weathering tests of up to 10 years duration.
Werkst. Korros. 26:36, 1975.
Exley, L. M. A practical review of residual oil firing problems and solutions. Combustion
41:16-23, 1970.
Fink, F. W. , F. H. Buttner, and W. K. Boyd. Technical-Economic Evaluation of Air-Pollution
Corrosion Costs on Metals in the U.S. APTD-0654. U.S. Environmental Protection Agency,
Research Triangle Park, NC, February 1971.
Fleetwood, M. J. Corrosion— resisting metals reviez-zinc coatings. Corros. Prev. Control
22:16-19, 1975.
Fochtman, E. G. , and G. Langer. Automobile paint damaged by airborne iron particles. J. Air
Pollut. Control Assoc. 6:243-247, 1957.
Freeman, A. M. , III. The Benefits of Environmental Improvement, Theory and Practice. Johns
Hopkins University Press, Baltimore, MD, 1979.
Gaj'endr'agadkar, S. K. Effects of sulphur dioxide on materials. Chem. Age India 28:673-677,
1977.
Gauri 'K. L. Effect of acid rain on structures. In: Acid Rain, Proceedings of the National
Convention Workshop Session, American Society of Civtl Engineers, Boston, Massachusetts,
April 2-6, 1979- Preprint 3598, American Society of Civil Engineers, Boston, HA, 1979.
pp. 55-75.
SOX10A/D
10-57 ^ 12-15-80
-------
Gibbons, E. V. The Corrosion Behaviour of the Major Architectural and Structural Metals In
Canadian Atmospheres. Summary of Ten-Year Results of Group I. NRCC 11630, National
Research Council of Canada, Ottawa, Canada, October 1970.
Gillette, D. G. S02 and material damage. J. Air Pollut. Control Assoc. 25:1238-1243, 1975.
Gillies, K. A., and D. J. Youston. Social effects of atmospheric emissions. Presented at the
69th Annual Meeting, Air Pollution Control Association, Portland, Oregon, June 27-July 1,
1976. Paper No. 76-19.4.
Glass, N. Environmental Effects of Increased Coal Utilization: Ecological Effects of Gaseous
Emissions from Coal Combustion. EPA 600/7-78-108, U.S. Environmental Protection Agency,
Corvallis, OR, June 1978.
Guttman, H. Effects of atmospheric factors on the corrosion of rolled zinc. In: Metal
Corrosion in the Atmosphere, A symposium Presented at the Seventieth Annual Meeting,
American Society for Testing and Materials, Boston, Massachusetts, June 25-39, 1967.
ASTM Special Technical Publication No. 435, American Society for Testing and Materials,
Philadelphia, PA, 1968. pp. 223-239.
Guttman, H., and P. J. Sereda. Measurement of atmospheric factors affecting the corrosion of
metals. In: Metal Corrosion in the Atmosphere, A Symposium Presented at the Seventieth
Annual Meeting, American Society for Testing and Materials, Boston, Massachusetts, June
25-30, 1967. ASTM Special Technical Publication No. 435, American Society for Testing
and Materials, Philadelphia, PA, 1968. pp. 326-359.
Haagenrud, S., and B. Ottar. Long range transport of air pollution and corrosion effects.
In: Proceedings of the Seventh Scandinavian Corrosion Congress, Trondheim, Norway, 1975.
pp. 102-115.
Hansen, J. Ailing Treasures. Sci. 80 (Eighty) 1:58-61, 1980.
Harker, A. B., F. B. Mansfeld, D. R. Strauss, and D. D. Landis. Mechanism of S02 and H^SO.
Aerosol Zinc Corrosion. EPA-600/3-80-018, U.S. Environmental Protection Agency, Research
Triangle Park, NC, January 1980.
Hay, T. K., and G. G. Schurr. Moisture diffusion phenomena in practical paint systems. J.
Paint Technol. 43:63-72, 1971.
Haynie, F. H. The economics of clean air in perspective. Mater. Prot. Perf. 13:33-38, 1974.
Haynie, F. H. Air pollution effects on stress induced intergranular corrosion of 7005-T53
aluminum alloy. In: Stress Corrosion - New Approaches. ASTM Special Technical
Publication 610, American Society for Testing and Materials, Philadelphia, PA, 1976.
pp. 32-43.
Haynie, F. H. Theoretical air pollution and climate effects on materials confirmed by zinc
corrosion data. In: Durability of Building Materials and Components. P. J. Sereda and
G. G. Levitan, eds., ASTM Special Technical Publication 691, American Society for; Testing
and Materials, Philadelphia, PA, 1980. pp. 157-175. ,
Haynie, F. H., J. W. Spence, and J. B. Up ham. Effects of Gaseous Pollutants on Materials—A
Chamber Study. EPA-600/3-76-015, U.S. Environmental Protection Agency, Research TJriangle
Park, NC, February 1976. ••;'-.
Haynie, F. H., J. W. Spence, and J. B. Upham. Technical note: effects of air pollutants on
weathering steel and galvanized steel: a chamber study. Mater. Perform. 15:48, 1976.
SOX10A/D 10-58 12-15-*
-------
Haynie, F. H., and 0. B. Upham. Correlation between corrosion behavior of steel and
atmospheric pollution data. ^n: Corrosion in Natural Environments. ASTM Special
Technical Publication 558, American Society for Testing and Materials, Philadelphia, PA,
1974. pp. 33-51.
Haynie, F. H. , and J. B. Upham. Effects of atmospheric pollutants on the corrosion behavior
of steels. Mater. Prot. Perform. 10:18-21, 1971.
Haynie, F. H., and J. B. Upham. Effects of atmospheric sulfur dioxide on the corrosion of
zinc. Mater. Prot. Perform. 9:35-40, 1970.
Hershaft, A. Air pollution damage functions. Environ. Sci. Technol. 10:992-995, 1976.
Hershaft, A., A. M. Freeman, T. D. Crocker, and S. B. Stevens. Critical review of estimating
benefits of air and water pollution control. EPA-600/5-78-014, Washington, DC, June,
1978.
Holbrow, G. L. Atmospheric pollution: its measurement and some effects on paint. J. Oil
Color Chem. Assoc. 45:701-718, 1962.
Hoshizawa, K., and K. Koyata. Formation and their prevention of acid smuts at oil-fired
burner. Nenyro Kyokaishi 49:656-665, 1970.
ITT Electro-Physics Laboratories, Inc. A Survey and Economic Assessment of the Effect of Air
Pollutants on Electrical Components. Volume 1, Sections 1 through 9. APTD-0797, U.S.
Environmental Protection Agency, Research Triangle Park, NC, August 1971.
Ireland, F. E. Pollution by Oxides of Sulfur. Chem. Eng. (London) (221): CE 261-CE262, 1968.
Jacksch, J. A., and H. H. Stoevener. Effects of Air Pollution on Residential Property Values
in Toledo, Oregon. Special Report 304, Oregon State University, Agricultural Experiment
Station, Corvallis, OR, September 1970.
Johnson, J. B., P. Elliott, M. A. Winterbottom, and G. C. Wood. Short-term atmospheric
corrosion of mild steel at two weather and pollution monitored sites. Corrosion Sci.
17:691-700, 1977.
Karraker, D. G. The kinetics of the reaction between sulfurous acid and ferric iron. J.
Phys. Chem. 67:871-874, 1963.
Kottori, M. S. The corrosion rates of metals exposed to pulp-mill and smelter emissions.
Presented -at the 73rd Annual Meeting, Air Pollution Control Association, Montreal,
Quebec, June 22-27, 1980. Paper No 80-67.4.
Krumbein, W. E., and C. Lange. Decay of plaster, paintings and wall material of the interior
of buildings via microbial activity. In: Environmental Biochemistry and Geochemistry,
Volume 2: The Terrestrial Environment, Proceedings of the Third International Symposium,
Minister of Sciences and Arts of Niedersachsen and Deutsche Forschungsemeinschaft,
Wolfbuettell, Germany, 1977. W. E. Krumbein, ed., Ann Arbor Science Publishers, Inc.,
Ann Arbor, MI, 1978. pp. 687-697.
Kingery, R. E. The.extent of the paper problem in large research collections and the
comparative costs of available solutions. In: Summary of a Conference,
Permanent/Durable American Library Association and Virginia State Library, Washington,
DC September 16, I960. Virginia State Library Publications No. 16, Virginia State
Library, Richmond, VA, 1960. pp. 36-41.
SOX10A/D 1Q-59 12-15-80
-------
Kucera, V. Effects of sulfur dioxide and acid precipitation on metals and anti-rust painted
steel. Ambio 5:243-248, 1976.
Larrabee, C. P. Mechanisms by which ferrous metals corrode in the atmosphere. Corrosion
(Houston) 15:526, 1959.
Larrabee, C. P., and S. K. Coburn. The atmospheric corrosion of steels as influenced by
changes in chemical composition. In: First International Congress on Metallic
Corrosion, London, April 10-15, 1961. Butterworth and Co. (Publishers) Ltd., London,
England, 1962. pp. 276-285.
Liu, B., and E. S. Yu. Physical and Economic Damage Functions for Air Pollutants by Receptor.
EPA 600/5-76-011, U. S. Environmental Protection Agency, Corvallis, OR, September 1976.
Litvin, A. Clear coatings for exposed architectural concrete. J. PCA Res. Dev. Lab.
10:49-57, 1968.
Luckat, S. Air pollutant effects on dome at Cologne. Koeln Domblatt 38:39, 1976.
Luckat, S. Investigations concerning the protection against air pollutants of objects of
natural stone. Staub Reinhalt Luft 32: May 1972.
Maler, K. G., and R. E. Wyzga. Economic Measurement of Environmental Damage: A Technical
Handbook. Organization for Economic Co-operation, Paris, France, 1976.
Mansfeld, F. B. Regional Air Pollution Study. Effects of Airborne Sulfur Pollutants on
Materials. EPA-600/4-80-007, U.S. Environmental Protection Agency, Research Triangle
Park, NC, January 1980.
Mansfeld, F., and J. V. Kenkel. Electrochemical monitoring of atmospheric corrosion
phenomena. Corros. Sci. 16:111-122, 1976.
Matsushima, I., Y. Ishizu, T. Ueno, M. Kanazashi, and K. Horikawa. Effect of structural and
environmental factors on the practical use of low-alloy weathering steel. Boshoku
Gijutsu 23:177-182, 1974.
Michelson, I., and B. Tourin. Report on Study of Validity of Extension of Economic Effects of
Air Pollution Damage from Upper Ohio River Valley to Washington, D.C. Area.
Environmental Health and Safety Research Association, August 1967.
Moore, J. C., and J. R. O'Leary. Evaluation of Structural Steel Coatings in Relation to
Industrial Atmospheric Conditions. West Virginia Department of Highways, Research
Project 23, West Virginia Department of Highways, Charleston, WV, January 1975.
Narayanan, R. L., and B. W. Lancaster. Household maintenance costs and particulate air
pollution. Clean Air (Melbourne) 7:10-13, 1973.
National Academy of Sciences. Airborne Particles. University Park Press, Baltimore, MD,
1979.
Newton, R. G. Cathedral chemistry—conserving the stained glass. Chem. Br. 10:89-91, 1974.
Niesel, K. The weathering of building materials in atmospheres containing sulfur oxides-a
literature discussion. Fortsch. Mineral. 57:68-124, 1979.
Nriagu, J. 0. Deteriorative effects of sulfur pollution on materials. In: Sulfur, in the
Environment. Part II: Ecological Impacts. John Wiley & Sons, Inc., New York, NY, 1978.
pp. 1-59.
SOX10A/D 10-60 12-15-8C
-------
Oma, K., T. Sugano, and Y. Hirai. Studies on atmospheric corrosion of steels related tc
meteorological factors in Japan. II. Multiple correlation of meteorological *rA
atmospheric pollution substances on corrosion of steel. Boshoku Gijutsu 14:16-19, 1965.
Overseas Public Nuisance Study Mission. Present Status of Public Nuisance in Electric Power
Industry of Japan. Central Power Council, Tokyo, Japan, September, 1965.
Oyama, M., A. Moriyama, H. Miyazato, and H. Sakugawa. Examples on the acid-smut fallout.
Okinawa-ken Kogai Eisei Kenkysho-ho 8:32-33, 1974.
Parker A. The destructive effects of air pollution on materials. In: Proceedings of the
22nd Annual Conference, National Smoke Abatement Society, Bournemouth, England, September
28, 1955. National Smoke Abatement Society, Brighton, England, 1955. pp. 1-15.
Potter, P. Chimney corrosion and acid smut emission. N.Z. J. Dairy Sci. Technol. 6:11-14,
1971.
Phafflin and Ziegler, 1976.
Prenderleith, H. J. The Preservation of Leather Bookbindings. The British Museum, London,
England, 1957.
Riederer, J. Pollution damage to works of art. In: New Concepts in Air Pollution Research.
Experientia Suppl. (20):73-85. 1974.
Robbins, R. C. Inquiry into the Economic Effects of Air Pollution on Electrical Contacts.
Stanford Research Institute, Menlo Park, CA, April 1970.
Rowden E. Acid corrosion and smut emission by kiln exhaust gases. Trans. Br. Ceram. Soc.
68:227-229, 1968.
Russell, C. A. How environmental pollutants diminish contact reliability. Insul. Circuits
22:43-44, 1976.
Salmon, R. L. Systems Analysis of the Effects of Air Pollution on Materials. APTD-0943, U.S.
Department of Health, Education and Welfare, National Air Pollution Control
Administration, Raleigh, NC, January 1970.
Salvin, V. S. Effect of air pollutants on dyed fabrics. J. Air Pollut. Control Assoc.
13:416-422, 1963.
Sanyal, B., and G. K. Singhania. Atmospheric corrosion of metals: Part I. J. Sci. Ind. Res.
Sect. B. 15:248-255, 1956.
Satake, -J., and T. Moroishi. Iji: Proceedings of the Fifth International Congress on Metal
Corrosion, 1974. National Association of Corrosion Engineers, Houston, TX. pp. 744-749.
Schwarz, H. On the effect on magnetite on atmospheric rust and on rust under a coat of paint.
Werkst. Korros. 23:648-663, 1972.
Schwenk W and H. Ternes. Rate of corrosion of plain carbon and low-alloy structural
steels.' Stahl Eisen 88:318-321, 1968.
Senqupta M and A A de Gast. Environmental deterioration and evaluation for dimension
stone. "Can. Min. Metall. Bull. 65:54-58, 1972.
Sereda P L Weather factors affecting corrosion pf metals. In: Corrosion in Natural
Environments. ASTM Special Technical Publication 558, American Society for Testing and
Materials, Pittsburg, PA, 1974. pp. 7-22.
SOX10A/D
10-61 12-15-80
-------
Simpson, J. W., and P. J. Horrobin, eds. The Weathering and Performance of Building
Materials. Wiley-Interscience, New York, NY, 1970. p. 286.
Sleater, G. A. Stone Preservatives: Methods of Laboratory Testing and Preliminary
Performance Criteria. NBS Technical Note 941, U.S. Department of Commerce, National
Bureau of Standards, Washington, DC, 1977.
Spedding, D. J., P. P. Rowlands, and J. E. Taylor. Sorption of sulphur dioxide by indoor
surfaces: leather. J. Appl. Chem. Biotechnol. 21:68-70, 1971.
Spence, J. W., F. H. Haynie, and J. B. Upham. Effects of gaseous pollutants on paints: A
chamber study. J. Paint Technol. 47:57-63, 1975.
Spence, J. W., and F. H. Haynie. Design of a laboratory experiment to identify the effects of
environmental pollutants on materials. In: Corrosion in Natural Environments. STP 558.
American Society for Testing and Materials, Philadelphia, PA, 1974. pp. 279-291.
Spence, J. W., and F. H. Haynie. Paint Technology and Air Pollution: A Survey and Economic
Assessment. AP-103, U.S. Environmental Protection Agency, Research Triangle Park, NC,
February 1972.
Stanners, J. F. Use of environmental data in atmospheric corrosion studies. Br. Corros. J.
5:117-121, 1970.
Svoboda, M., B. Knapek, and H. Kilcoua. The effect of pigments on the penetration of sulfur
dioxide into coatings. J. Oil Colour Chem. Assoc. 5_6:172174, 1973.
Sydberger, T. Influence of Sulphur Pollution on the Atmospheric Corrosion of Steel.
University of Gothenburg, Department of Inorganic Chemistry, Gothenburg, Sweden, 1976.
pp. 3-8.
Sydberger, T., and R. Ericsson. Laboratory testing of the atmospheric corrosion of steel.
Werkst. Korros. 28:154-158, ,1976.
Sydberger, T., and N. G. Vannerberg. The influence of the relative humidity and corrosion
products on the adsorption of sulfur dioxide on metal surfaces. Corros. Sci. 12:775-784,
1972.
Tomashov, N. D. The Science of Corrosion. The MacMillan Co., New York, NY, 1966.
Upham, J. B. Atmospheric corrosion studies in two metropolitan areas. J. Air Pollut. Control
Assoc. 17:398-402, 1967.
Upham, J. B., and V. S. Salvin. Effects of Air Pollutants on Textile Fibers and Dyes.
EPA-650/3-74-008, U.S. Environmental Protection Agency, Research Triangle Park, NC,
February 1975.
Vannerberg, N. G., and T. Sydberger. Reaction between SO, and wet metal surfaces. Corros.
Sci. 10:43-49, 1970. * ,
Verdu. Effect of air pollutants on aging of plastic materials. Trib. CEBEDEAU 27:360-370,
1974. ~~
Waddell, T. E. The Economic Damages of Air Pollution. EPA-600/5-74-012, U.S. Environmental
Protection Agency, Washington, DC, May 1974.
Walsh, M., A. Black, A. Morgan, and G. Crashaw. Sorption of SO, by indoor surfaces including
carpets, wallpaper, and paint. Atmos. Environ. 11:1107-1111, 1977.
SOX10A/D 10-62 12-15-80
-------
Watson, W. D. , Jr., and J. A. Jaksch. Household cleaning costs and air pollution. Presented
at the 71st Annual Meeting, Air Pollution Control Association, Houston, Texas, June
25-30, 1978. Paper No. 78-52.3.
Winkler, E. M. Stone: Properties, Durability in Man's Environment. Second revised edition,
Springer-Verlag, New York, NY, 1975.
Winkler, E. M. Important agents of weathering for building and monumental stone. Eng. Geol.
(Amsterdam) 1:381-400, 1966.
Yocom, J. E. and Grappone. Effects of Power Plant Emissions on Materials. EPRI EC-139,
Electric Power Research Institute, Palo Alto, CA, July 1976.
Yocom, J. E. Air pollution damage to buildings on the Acropolis. J. Air Pollut. Control
Assoc. 29:333-338, 1979.
Yocom, J. E., and J. B. Upham. Effects on economic materials and structures. In: Air
Pollution. Third Edition. Volume II: The Effects of Air Pollution. A. C. Stern, ed.,
Academic Press, Inc., New York, NY, 1977. pp. 65-116.
Yoshihara, K., T. C. Huang, H. Ebihora, and H. Shibata. Chemical effects of the 35Cl(n,p)35 S
reaction. Radiochim. Acta 3:185-191, 1964.
Zeronian, S. H. Reactions of cellulosic fabrics to air contaminated with sulfur dioxide.
Text. Res. J. 40:695-698, 1970.
Zeronian, S. H., K. W. Alger, and S. T. Omaye. Reactions of fabrics made from synthetic
fibers to air contaminated with nitrogen dioxide, ozone, or sulfur dioxide. In:
Proceedings of the Second International Clean Air Congress, International Union of Air
Pollution, Washington, DC, December 6-11, 1970. H. M. Englund and W. T. Beery, eds.,
Academic Press, New York, NY, 1971. pp. 468-476.
SOX10A/D 10~63
-------
GLOSSARY FOR VOLUME IV
Adsorption: Adhesion of a thin layer of molecules to a liquid or solid surface.
Aerosol: Solid particles or liquid droplets which are dispersed or suspended in air.
Albedo: That fraction of the total light incident on a reflecting surface, especially a
celestial body, which is reflected back in all directions.
Anode: Positive electrode.
Austenite: Gamma iron with carbon in solution.
Austenitic steel: An alloy whose structure is typically that of austenite at room
temperature.
Colloidal system: An intimate mixture of two substances, one of which, called the dispersed
phase (or colloid), is uniformly distributed in a finely divided state through the second
substance, called the dispersion medium (or dispersing medium); the dispersion medium or
dispersed phase may be a gas, liquid, or solid. Also known as colloidal dispersion;
colloidal suspension.
Convective mixing: Atmospheric motions that are predominantly vertical, resulting in vertical
transport and mixing of atmospheric properties.
Crystalization spall ing: The reduction of stone surfaces worn away by a crystal line-forming
process.
Ecosystem: The interacting system of a biological community and its environment.
Efflorescence (mineral): A whitish powder, consisting of one or several minerals produced as
an encrustation on the surface of a rock in an arid region. Also known as bloom.
Electrophoretic coating: A surface coating on a metal deposited by electric discharge of
particles from a colloidal solution.
Exfoliate: Flake away or peel off in scales.
Ferrous (metals): Relating to or containing iron.
Hygroscopic: Pertaining to a marked ability to accelerate the condensation of water vapor.
Insolation decoloration: Removal or fading of color due to exposure to the sun.
Isopleth: A line on a map or chart connecting points of equal value.
Iterative bidding: A research technique for subjectively establishing value.
Mesoscale: Of or relating to meteorological phenomena from 1 to 100 kilometers in horizontal
extent.
Nephelometer- A type of instrument that measures at more than one angle, the scattering
function of particles suspended in a medium; information obtained may be used to
determine the size of suspended particles and the visual range through the medium.
Nuclei (plural of nucleus): A central point, group, or mass about which gathering,
concentration, or accretion takes place.
SOXFM/G
-------
Participate matter (participates, particles, etc.): Fine liquid or solid particles within an
aerosol such as dust, smoke, mist, fumes or smog.
Passivate (metal): To render passive; to reduce the reactivity of a chemically active metal
surface by electrochemical polarization or by immersion in a passivating solution.
Rayleigh scattering: Coherent scattering in which the intensity of the light of wavelength g,
scattered in any direction making an angle with the incident direction, is directly
proportional to 1 + cos r and inversely proportional to g .
Sorb: To take up and hold by absorption or adsorption.
Sorbed: Past tense of "sorb."
Sorption: The process of being sorbed.
Spatial: Relating to, occupying, or of the nature of space.
Spectral: Relating to the electromagnetic spectrum.
Spectroscopy: The branch of physics concerned with the production, measurement, and
interpretation of electromagnetic spectra arising from either emission or absorption of
radiant energy by various substances.
Stoichiometry: The numerical relationship of elements and compounds as reactants and products
in chemical reactions.
Sulfur dioxide (^O^): Colorless gas with pungent odor released primarily from burning of
fossil fuels, 'such as coal, containing sulfur.
Transmissometer: An instrument for measuring the extinction coefficient of the atmosphere and
for the determination of visual range. Also known as hazemeter or transmittance meter.
Troposphere: That portion of the atmosphere in which temperature decreases rapidly with
altitude, clouds form, and mixing of air masses by convection takes place. Generally
extends to about 7 to 10 miles above the earth's surface.
SOXFM/G G-2 12-17-80
-------