NATIONAL TRENDS IN TRACE METALS
              IN AMBIENT AIR
                  1965 - 1974
       U.S. ENVIRONMENTAL PROTECTION AGENCY
           Office of Air and Waste Managensent
         Office of Air Quality Planning and Standards
        Research Triangle Park, North Carolina 27711

-------
                       EPA-450/1-77-003
NATIONAL TRENDS
 IN TRACE  METALS
  IN  AMBIENT  AIR
       1965 -  1974
              by

   Robert B.Faoro and Thomas B. McMullen

    Monitoring and Data Analysis Division
      Monitoring and Reports Branch
 U.S. ENVIRONMENTAL PROTECTION AGENCY
     Office of Air and Waste Management
  Office of Air Quality Planning and Standards
    Monitoring and Data Analysis Division
  Research Triangle Park, North Carolina 27711


           February 1977

-------
The Office of Air and Waste Management of the Environmental Protection Agency would like to thank the
Environmental Monitoring  and Support Laboratory,  RTF,  for providing  air quality  data from the
National Air Surveillance Network.
   This report has been reviewed by the Monitoring and Data Analysis Division, Office of Air Quality
 Planning and Standards, Office of Air and Waste Management, Environmental Protection Agency, and
 approved for publication. Mention of trade names or commercial products does not constitute endorse-
 ment or recommendation for use.  Copies are available  free of charge to Federal employees, current
 contractors and grantees, and nonprofit organizations - as supplies permit - from the Office of Library
 Services, Environmental Protection Agency, Research Triangle Park, North Carolina 27711; or copies
 may be purchased from the Superintendent of Documents, U.S. Government Printing Office, Washington,
 D.C. 20460.
                                Publication No. EPA-450/1-77-003
                                              11

-------
                             CONTENTS

Section                                                              Page

LIST OF FIGURES 	 iv
LIST OF TABLES 	 iv
1. INTRODUCTION AND OVERVIEW 	  1
2. BACKGROUND 	  3
  2.1  Ambient Monitoring Data  	  3
  2.2  Major Metal Emission Sources 	  4
3. TRENDS IN METALS  	  7
  3.1  Urban Trends  	  7
  3.2  Nonurban Trends 	 10
4. POSSIBLE CAUSES FOR METAL TRENDS OBSERVED	11
  4.1  Fuel-Combustion-Related Metals—Lead, Vanadium, Nickel, and Titanium  .... 11
      4.1.1 Lead  	 11
      4.1.2 Vanadium and Nickel  	12
      4.1.3 Titanium	 17
  4.2  Industry-Related  Metals—Cadmium, Chromium, Copper, Iron, and Manganese. 17
5. SUMMARY AND CONCLUSIONS  	19
6. ACKNOWLEDGMENTS 	21
7. REFERENCES  	23
APPENDIX  	25
TECHNICAL REPORT  DATA 	28
                                  111

-------
                           LIST OF  FIGURES

Figure                                                                      Page

  1       Trends in the 50th Percentile at Urban Sites of Annual Averages for
          Metals Associated with Primarily Fuel Combustion Sources  	  8

  2       Trends in 50th Percentile of Annual Averages for Metals Associated
          with Metal Industry Sources at Urban Sites  	  9

  3       Seasonal Patterns and Trends in Quarterly Average Urban Lead
          Concentrations	12

  4       Nationwide Trends in Regular and Premium Gasoline Sales and
          Lead Content, 1960-1974  	 13

  5       Five Geographical Areas of the Country Used in Urban Metal Summaries  14

  6       Regional Trends in the 90th Percentile of the Annual Averages for
          Vanadium 	 15

  7       Seasonal Variation in Quarterly Averages at Urban Sites in the Northeast  16
                            LIST OF TABLES

Table                                                                       Page

  1       Three Highest Emission Categories for Metals Studied 	  5

  2       Percent Change in 50th and 90th Composite Statistics for Urban Sites  	10

  3       Trends in Urban Metal Concentrations and Their Possible Causes  	 19
                                       IV

-------
                  NATIONAL TRENDS
                   IN  TRACE METALS
                     IN  AMBIENT AIR
                           1965 -  1974
              1.  INTRODUCTION AND OVERVIEW
   Airborne concentrations of metals are a concern for three reasons:  They are generally
associated with particles in the respirable size range, many possess known toxic properties,1
and they can act as catalysts in atmospheric reactions such as the conversion of sulfur dioxide
to sulfates. This report examines trends over the past 10 years in ambient concentrations
for 11 metals:  beryllium (Be), cadmium (Cd), chromium (Cr), copper (Cu), cobalt (Co), iron
(Fe), lead (Pb), manganese (Mn), nickel (Ni), titanium (Ti), and vanadium (V). The trends
are derived from samples collected from 92 urban and 16 nonurban hi-vol stations in the
National Air Surveillance Network (NASN).  All samples were analyzed at the central NASN
laboratory, now part of EPA's Environmental Monitoring and Support Laboratory. Because
of limitations inherent in these data, the results are intended primarily as a qualitative
description of general patterns rather than as a precise quantitative analysis.

   This report consists of four major sections: (1) a background section describing the data
base and providing general information on emissions, (2) a section presenting the observed
trends, (3) a section discussing the possible reasons for the trends, and (4) a concluding or
summary section.  The basic information is highlighted in three  tables summarizing emis-
sions, trends, and possible causes for the observed  trends. For purposes of presentation,
the trace metals have been grouped into two broad categories: the metals related to fuel
combustion—beryllium,  lead, nickel,  titanium, and vanadium—and those  related to
industry—cadmium, chromium, cobalt, copper, iron, and manganese.

   The major findings of this investigation are as follows:
   •   In general, ambient metal concentrations have declined in most urban areas with the
   exception of copper, titanium, and possibly chromium.

-------
•  The downward  trend in lead concentrations is due to the lower lead content of
gasolines sold in recent years.
•  The decline in vanadium and nickel concentrations, particularly in the Northeast
sector of the United States, results from desulfurization of petroleum, which also removes
these impurities.
•  The absence of trends in copper may be at least partly due to copper contamination
of the  sample from the commutator of the hi-vol itself.
•  The increase in titanium concentrations over the 10-year period may be caused by the
increase in coal consumed by electric utilities.

-------
                             2.  BACKGROUND

2.1 AMBIENT MONITORING DATA

   Hi-vol filter samples from the NASN have been routinely analyzed for certain metals
going back to the early 1960's. The years 1965-74 are covered in this report. Earlier NASN
metals data are of such a sketchy nature that they are of no practical importance for trends
purposes.  As they become available, data for subsequent years will be analyzed in future
reports.

   The NASN network consists of individual monitoring stations located in urban and
nonurban areas throughout the Nation. The urban sites are usually located in the center-city
business area, while the nonurban sites are either in Federal or State parks. For the metal
analysis, individual  24-hour samples taken throughout  most of the  10-year period on a
biweekly schedule were combined by quarter and then analyzed to obtain quarterly com-
posite measurements at each site. The laboratory methodology, lower discrimination level
(LDL), and other characteristics of the 1970-74 data are described in an EPA report.2 For
most metals the LDL values were generally higher in the first 5 years (1965-69) than the most
recent 5 years (1970-74) by at least an order of magnitude.  Composite values less than these
limits were given the value of one half of these limits—for example, a beryllium urban annual
average in which every quarterly composite value is less than the 1965-69 LDL of 0.0002 micro-
grams per cubic meter (/xg/m3) will be given a value of 0.0001 Mg/m3 for computational pur-
poses.  The 92 urban and 16 nonurban sites qualifying for trend analysis  are given  in the
Appendix to this report, with a brief description of the sites and how they were selected.

   The data on airborne concentrations of metals reported here, by the very nature of their
collection and analysis, provide useful measures of relative rather than absolute concentra-
tions across time, between seasons, or between geographical regions of the country. Through-
out the 10-year period, basically the same sampling and analysis procedures were used on all
data.  However, there  are some possibly important differences in the manner the samples
were analyzed:

   (1)  Since 1970, the LDL for most metals have been reduced substantially. Consequently,
the procedure is  more sensitive for  detection of background metal concentrations  in the
blank or unexposed filters.  In the earlier years, of course, blank metal concentrations may
not have been detected, and thus no adjustment  was made in the final  measured value
because of the much larger LDL used. The effect of this refinement in the analytical techni-
que on the reported values is not known.

   (2)  Either all or part of the data for 1965 may involve samples where the muffle furnace
was used for ashing as opposed to the much lower temperature ashing procedure used in the
other years. An analysis of metal concentration before and after this change concluded that
it does not apparently  affect the more volatile metals such as cadmium and lead, which are
the most likely to be affected. Hence, these data are included in the  analysis.

   (3)  Samples for the first 5 years were analyzed using an emission spectrometer, while
those  for the 5 most recent  years were generated with an emission spectrograph.

   (4)  The data were analyzed at different times and even places since the laboratory was
moved in 1970 from Cincinnati, Ohio, to Research Triangle Park, N.C.
                                         3

-------
   (5) The NASN metals data for the most recent 5 years are thought to be more precise and
internally consistent than the earlier data since these data were analyzed together as a group;
furthermore, the newer data are the result of two separate determinations, using the same
composite extract, rather than one as was done prior to 1970.

   The high data recorded in 1969 for most of the metals cannot be explained. The high
values were most pronounced in vanadium and cadmium. This finding was discussed with
appropriate personnel in the Office of Research and Development, and they could offer no
explanation.3 Removal of the 1969 data would not alter our assessment of overall trends.

2.2  MAJOR METAL EMISSION SOURCES

   The three highest emission categories have been determined for 9 of the 11 metals studied
(Table 1). The rankings are based on nationwide estimates of emissions for 1970—the most
recent year available.4"7 No estimates are available for cobalt or iron.

   Of the metals studied,  iron is the most abundant in the earth's crust and in the ambient
air.  It exists as an impurity in fuels, has wide use, and, therefore, has a pervasive potential
for emissions.  A significant portion of the iron collected likely comes from its corrosion
and incineration.

   Beryllium, lead, nickel, titanium, and vanadium are most likely to be present in urban
areas because their emissions are associated with combustion of gasoline, coal, or oil. They
are present as impurities  or, in the case of lead, as an additive in gasoline.

   Other metals (Cd, Co, Cr, Cu, and Mn) are more closely associated with their use in making
steel or other alloys or their fabrication into end products. Higher concentrations of these
metals would be most likely in urban or nonurban areas near steel plants, smelting opera-
tions, or other plants associated with the metals industry.  Although many of the NASN sites
are in areas  not directly  associated with major metal industries, several  steel-producing
centers, such as Pittsburgh, Pa.; East Chicago, Ind.; Ashland, Ky.; and Bethlehem, Pa., are
represented  in our set of  trend sites.

-------
                   Table 1. THREE HIGHEST EMISSION CATEGORIES FOR METALS STUDIED
Metal
Fuel-combustion-
related metals
Beryllium
Lead
Nickel
Titanium
Vanadium
Industry-related
metals
Cadmium
Chromium
Copper
Cobalt
Iron
Manganese
1970 U.S.
emission
estimates-^
(tons)

170
230,000
7,300
88,000a
20,000

2,200
t
17,000
14,000b
Unknown
Unknown
18,000
Highest

Coal combustion
Combustion of
leaded gasoline
Oil combustion
Coal combustion
Oil combustion

Incineration of
plated metal
Ferrochromium
production
Metallurgical
processing


Manganese alloys
processing
%of
total

88
93
83
83
90

46
68
64


57
2nd
highest

Oil combustion
Secondary lead
smelting
Stainless steel
reprocessing
Use as pigment
Coal combustion

Metallurgical
processing
Refractory
production
Iron and steel
production


Case iron
reprocessing
%of
total

6
2
7
5
9

43
10
20


17
3rd
highest

Metallurgical
processing
Solid waste
disposal
Mining and
processing
Pigment
production
Metallurgical
processing

Incineration
of radiators
Coal
combustion
Coal
combustion


Coal
combustion
%of
total

3
1
4
5
1

6
9
7


11
aAs Ti02.
b1969 emissions.

-------
                         3.  TRENDS IN METALS
   In the appraisal of trends, the 50th percentile (the median) and the 90th percentile of
the annual averages were chosen as the statistics to best describe the change in metals con-
centrations over time.  This was done to minimize the influence of individual extreme values
and to simplify the characterization of trends for metals having large amounts of data below
the detection limit. Also, these statistics can portray different aspects of the yearly average
distribution—the 50th percentile, the typical, and the 90th, the high concentration site.
Rigorous  statistical techniques  such as  time  series or regression analysis were not used
because this degree of sophistication was neither necessary nor advisable in light of the
uncertainties present in  the data. Instead, our presentation will rely heavily on graphical
displays of the  two percentile values and a subjective interpretation of these patterns.

3.1 URBAN TRENDS

   Figures 1 and 2 graphically present the urban metal concentrations (1965-74) for the two
broad emission categories (combustion  and industry) in terms of the 50th percentile of
annual averages. The 90th percentile plots are not shown since they provided very similar
results.  Beryllium  and cobalt are omitted because both the 50th percentile and 90th per-
centile values fall below  the lower detection limit for  all 10 years.

   Trends in these metals are mixed over the 10-year period. For example, lead, iron, copper,
and chromium  do not show a discernible trend over the 10-year period, although both lead
and iron show  a downward trend over the past 5 or so years. Cadmium, manganese, nickel,
and vanadium all show downward trends over  the 10-year period and, with the exception of
cadmium, over the most  recent 5 years as well. Titanium shows an increasing trend for the
first 5 years and a stable pattern over the last 5  years. These patterns describe the trends for
the group or urban sites  as a whole and characterize the vast majority of individual station
trends.

   The percent of changes for the 11 metals and total suspended particulates were examined
for the 10-year period and for the most recent 5-year period (Table 2). These figures are based
on a comparison of averages for the 1965-67 versus 1972-1974 periods for 10-year interval and
1970-71 versus  1973-74 for the 5-year interval. In the 10-year trends, only titanium shows
positive changes in both the 50th and 90th percentile statistics. Chromium shows no change in
the median  value and a 17 percent decrease in  the 90th percentile. Copper and lead changes
lie mostly in the stable interval (-10 to +10 percent); the remainder (with the exception of
beryllium and  cobalt for which no trend can be determined) appear to have declined in
ambient concentrations over this time period. Copper's stability appears to be an artifact
caused by a contamination problem for the hi-vol sampler.

   Trends for the most recent 5-year period (1970-74) are similar with the exception of lead,
which now shows a modest decline, and titanium and chromium, which exhibit more stable
patterns. The median and 90th percentile TSP concentrations have declined 18 percent and
17 percent, respectively, over the 10-year period for these groups of sites. However, TSP con-
centrations  have changed much  less (-4 percent) over the most recent 5-year period in the
typical or median concentration range.  TSP concentrations at the high end of the concentra-

-------
tion distribution, as characterized by the 90th percentile, continue to show a modest decline
(-13 percent).

   The urban trend in the ratio of metal to TSP was considered for iron, lead, and vanadium.
Trends in the fraction paralleled the raw concentration trends almost exactly. This indicates
that it is not just the declining trend in TSP that is causing the change in metal concentra-
tions, but it is a change over and above the expected from just collecting less TSP.
     0.1
  _E
  O1
  a.
  2"
  o
    0.01
   0.001
                                                                       LEAD
                  (Lead values should be multiplied by 10 in order to be in M9/m3 units.)
                                                                       TITANIUM
                                  *•-.
                                                     \
                                                      \
                                                                 \
                                                                  \
                                                                        NICKEL
                                                          , - - A -  -A VANADIUM
                  65    66
                             67
                                   68
                                        69     70
                                         YEAR
                                                    71    72
73     74
Figure 1. Trends in the 50th percentile at urban sites of annual averages for metals associated with
primarily fuel combustion sources.  (A indicates value below lower discrimination limit.)

-------
   0.1
E
~5i
a.
    .01
   .003
   002
  .0008
   0004
                                                   IRON3
                                                             (IRON VALUES SHOULD BE MULTIPLIED BY
                                                              10 IN ORDER TO BE IN w/m3 UNITS)
                                 .
                            COPPERb

                                (TREND PATTERN IS PROBABLY INFLUENCED BY
                                INTERNALLY GENERATED COPPER FROM THE
                                SAMPLING DEVICE.)
\   / MANGANESE


   \
                                                           CHROMIUM
                                           A
                                                     CADMIUM
                   65     66     67     68     69     70     71     72     73     74    YEAR
      Figure 2. Trends in 50th percentile of annual averages for metals associated with metal
      industry sources at urban sites. (A indicates value below lower discrimination limit.)

-------
           Table 2.  PERCENT CHANGE IN 50TH AND 90TH COMPOSITE STATISTICS
                                 FOR URBAN SITES

Metal
Fuel-combustion-
related metals
Beryllium
Lead
Nickel
Titanium
Vanadium
Industry-related
metals
Cadmium
Chromium
Copper
Cobalt
Iron
Manganese
TSP
1965-1974 1970-1974
1 965-67 vs. 1 972-74 1 970-7 1 vs. 1 973-74
% change
50
percentile

Unknown3
+ 5
40
+106
/_ \ b

/ \b
(no change)"
10
Unknown3
-29
-50
-18
90
percentile

Unknown3
5
-32
+22
-43

-57
17
+ 3
Unknown3
-40
-60
17
50
percentile

Unknown3
-23
-40
0
<->b

Unknown3
0
+ 11
Unknown3
-35
50
4
90
" percentile

Unknown3
10
-43
7
-59

-44
- 12
+ 8
Unknown3
-54
- 50
13
           3Change cannot be determined because concentrations are below the lower
            discrimination limit.

           "Only the sign of the change is given when one of the two intervals used
            in the comparison is less than the LDL.
3.2 NONURBAN TRENDS


   Data from the nonurban stations were considered separately. Because of the lower levels
and the smaller number of stations (16), detailed discussion is not warranted, and the con-
clusions must be treated as tenuous.  However,  data from the nonurban  sites do indicate
the following:


   •  Chromium, copper,  lead, and  titanium increased fairly steadily  over the 10-year
      period


   •  Iron, manganese, nickel, and vanadium showed declines, especially during the 1970-
      74 period


   •  Beryllium, cadmium, and cobalt levels were below the LDL of the method.


                                         10

-------
 4.  POSSIBLE CAUSES FOR METAL TRENDS OBSERVED

4.1  FUEL-COMBUSTION-RELATED METALS-LEAD, VANADIUM, NICKEL,
     AND TITANIUM

   Of the four metals associated principally on a nationwide scale with fuel combustion
sources, nickel and vanadium (which are both associated with oil combustion) show a sub-
stantial decline in concentration over the 10-year period, while titanium increased early in
this  period. Lead levels experienced an initial increase and then a decline.

4.1.1 Lead

   The national composite 50th percentile of lead concentrations (Figure 1) increased from
 1965 until 1971 and then declined from about 1.1 Mg/m3 to 0.84 jug/m3—about a 24 percent
decrease, with most of the decline occurring between 1972 and 1973. This general trend
pattern is consistent at most sites studied. A recent report8 of extensive lead data at urban
locations in Los Angeles, Houston,  Tulsa, and Chicago has shown a decreasing trend in
particulate lead at each of these locations. Another site in DuPont, Wash., showed a decline in
lead concentrations, while a distinctly rural site a Starke, Fla., showed a stable pattern. NASN
lead concentrations are about 30 percent higher in the Far West (where the California sites
predominate) and Northeast sectors than in other three geographical areas of the country.
This geographical difference in lead loadings must be due to the greater amount of gasoline
consumed, particularly in the southern California and New York City areas. These results,
however, should be used with caution owing to the relatively small number of stations in each
area and the possible area differences  in the  impact of vehicular and other lead emissions
at these monitoring sites. The NASN site located in downtown Los Angeles experienced the
highest concentrations—averaging between 4 and 5 /ig/m3 until 1971 when the concentra-
tions decreased to about 2 ;ug/m3 in  1974.  Nationwide, lead levels tend to peak during the
winter months at most of the 92 urban sites, especially the California sites even though winter
lead content levels in gasoline9 are generally lower in California and the rest of the country
as well (Figure 3). This seasonal pattern in California has been at least partly explained there
by the poorer overall dispersion of the atmosphere during the winter months.10

   Since about the 1970 model year, automobiles have been built with lower compression
engines—ones requiring lower octane gasoline and thus gasoline with lower lead content. As
a result of this change, practically all new cars built since 1970 are able to use regular gasoline
instead of the more leaded premium fuels. The result of the engine modification can be
clearly seen in the lower lead content in gasoline9 (both in regular and premium grades)
after 1969 (Figure 4).  Subsequently, sales of regular gas increased and sales of premium
gasoline  decreased. Lead contents in gasoline will continue to drop in the future because
of the  increasing use of no-lead gasolines in  new cars equipped with catalytic converters.
These factors, coupled  with the consumption of a modest amount of low  lead or no-lead
gasolines introduced at about this same time, result in the modest (10 to 20 percent) decrease
in ambient lead concentrations over this period.  These changes are more  than enough to
offset a general increase in gasoline consumed from 1970-74. However, this increase in gaso-
line consumed is probably not felt as greatly at these predominately center-city areas since
they are  generally already at or near vehicular saturation.  There may even have been  a
reduction in vehicle miles traveled in downtown areas during this time because of car pool-
ing, improved mass transit systems, and the loss of business activity to suburban shopping
centers.
                                         11

-------
       4.0
        3.0
   •a
   V)
   z
   o
   EC
   I-
   o
   o
   a
        2.0
        1.0
                 i       i      r
                                                                90th percentile
                  65
66
67
68
 69    70
YEAR
71
72
73
74
     Figure 3.  Seasonal patterns and trends in quarterly average urban lead concentrations.
4.1.2 Vanadium and Nickel

   The 50th percentile of vanadium annual averages for 1971-74 (Figure 1) falls in the LDL
region and are assumed at the LDL value (0.003 Mg/m3), even though there are slight varia-
tions for these years below this limit.  The abrupt rise in vanadium concentration in 1969 is
puzzling and may suggest a positive bias in these data since most of the sites record their
highest annual average for this year. However, the nonurban sites fail to show this unusual
pattern in 1969.

   Figure 5 shows five broad geographical areas of the United States. On the basis of 90th
percentiles of the annual averages, vanadium concentrations in the Northeast are much
higher over the entire 10-year period than the other regions (Figure 6). Once again, the 1969
value appears peculiar for most of the regional summaries, although the abrupt peak does
not show up in the Northeast.  Concentrations in the Northeast decreased 74 percent from
1969 to  1974 (0.35 Mg/m3 to 0.09 /ug/m3), with most of this drop occurring between 1971 and
1972. The changes in vanadium concentrations  in the South are caused mainly by two or
three stations showing relatively high readings in the 1972-74 period. This does not appear
to be a  pattern characteristic of sites in the region. Similar regional trends and regional
gradients in concentrations are observed also for nickel. Nickel shows a fairly steady decline
over the 10-year period.  Both vanadium and nickel show a pronounced and regular high
winter-low summer seasonal variation in the 50th and  90th percentile in the Northeast
(Figure 7). This is attributed to space heating emissions and poorer atmospheric dispersion
in the winter.
                                         12

-------
       o
       co
            80
            70
            60
            50
            40
            30
       o
            20
            10
                   REGULAR GAS SALES (%}
                                                          \
                            PREMIUM GAS SALES (%)
           60     62     64     66     68     70


                                YEAR
                                                         72     74
      
-------
HAWAII
                                            39 MILLION POPULATION
                                            19 sites
44 MILLION POPULATION
14  sites
64 MILLION POPULATION
                                                                                             29  sites
                                                                                               32 MILLION POPULATION

                                                                                                  15  sites
                   29 MILLION POPULATION
                 15  sites
        Figure 5.  Five geographical areas of the country used in urban metal summaries.

-------
  0.10
<
cr
   .01
   .001
                         A
                        i  \
                       1
                   V
                                                               NORTH EAST (29 SITES)
                                                             •* SOUTH (16 SITES)
                                   WEST (15 SITES)
       /   \\
      /      »\
                                                           4
\    '  A

 \//\
T,      /

 \>~-
                                                             NORTH CENTRAL (14 SITES)
                                       \
                                             \
                                                                MIDWEST (19 SITES)3
               65
                    66    67    68    69
                                         70    71
                                                    72    73
                                                                   YEAR
    Figure 6. Regional trends in the 90th percentile of the annual averages for vanadium.

    (Aindicates value below lower discrimination limit.)

    31971-74 90th percentile below lower discrimination limit  0.003
                                      15

-------
    0.800
E    0.600 -
     0.400 -
o
u
     0.200 -
                                                                     72      73     74
     0.150
3:   0.100

2
O

<
CC
o
u
     0.050
        0
                                                                     I        I
                                                                     NICKEL       _
                  65     66     67      68     69     70      71      72      73      74
                                             YEAR
       Figure 7. Seasonal variation in quarterly averages at urban sites in the northeast.
                                           16

-------
   This decline in vanadium and nickel concentrations, particularly in the Northeast, is
attributed to the sulfur regulations that went into effect during this time.11  Removal of
sulfur from residual oil also indirectly removes vanadium and nickel, which are present in
very high concentrations ( 200 ppm for vanadium and about 20 ppm for nickel) in South
American petroleum.12  South American or Venezuelian  petroleum is imported almost
exclusively in the New York and northeastern seaboard areas.  As an example, in New York
City and the neighboring counties of Westchester, Suffolk, Rockland, and Nassau, sulfur
regulations have reduced the maximum allowable sulfur in residual oil from about 1 percent
in the 1968-1969 period to about 0.3 percent in 1973—a decrease of 70 percent. This percent
change agrees very well with the  percent change in vanadium concentrations (70 to 80 per-
cent) over this same time at the New York City NASN site  and with the percent change in
vanadium in the composite 90th  percentile for the Northeast  region.

4.1.3 Titanium

   Titanium concentrations have been increasing over the 10-year period (Figure 1), with
65 of 92  sites  showing this pattern.  Titanium concentrations are highest for  the western
stations, where the 50th percentile shows an increase from 0.045 /ig/m3 in 1965 to 0.058 jug/m3
in 1974. The apparent upward trend may be the result of the 50 percent increase in consump-
tion13 of coal in electrical generating plants; coal combustion is the principal source of tita-
nium, accounting for 83 percent  of total TiO2 emissions (Table 1).

4.2  INDUSTRY-RELATED METALS-CADMIUM, CHROMIUM, COPPER, IRON,
     AND MANGANESE

   Copper and chromium show fairly stable patterns on the basis of 50th percentiles (Figure
2).  Copper's very steady concentration pattern with time is rather surprising in light of the
declining trends over the 10-year period shown by iron, manganese, and cadmium, other
metals of this emission category group. It has been hypothesized, that there is persuasive
supporting evidence, that copper contamination from wearing of the commutator on the
motor of the hi-vol itself masks any possible trend.14 Highest concentrations of iron, manga-
nese, and cadmium are generally found in the North Central States and the Northeast where
steel production is concentrated. The East Chicago, Ind.,  site recorded  the highest iron
concentration of any site studied, reaching its peak in 1969 of  almost 10 /ug/m3; it then fell
off to just over 4 jug/m3 in 1974.

   The decline in urban iron concentrations since 1968 has occurred at practically every site
including  sites as different as Bethlehem,  Pa.,  with a steel plant, and  Durham, N.C., without
one.  The trend must be partially due to decreased iron emissions from other sources—for
example, burning oil or gas instead of coal or improving incineration of refuse  or other im-
proved waste burning or waste removal practices.  Declines in some steel producing areas,
may be also due to particulate controls at steel plants, although particulate controls in the
steel industry as a whole have not been implemented extensively as yet. The downward
trends of some other metals, such as cadmium and manganese, are more difficult to attribute
to specific causes; it must be assumed that they were brought about by general particulate
controls in processes emitting these elements. This apparent decrease occurred even though
domestic production and consumption of cadmium and manganese stayed at about the same
level throughout the period of interest.
                                         17

-------
                  5.  SUMMARY AND CONCLUSIONS

   In general, metal concentrations have declined in most urban areas with the exception
of copper, titanium, and possibly chromium. Nickel, vanadium, iron, manganese, and cad-
mium show declines since the late 1960's, while lead concentrations began to drop in 1972.
In contrast, titanium concentrations in recent  years have remained at the same concentra-
tions measured in the late  1960's.  Chromium  and especially copper were fairly constant
throughout the 10-year period.

   Table 3 summarizes metal trends and possible causes for these trends.  Particularly in
the cases of cadmium and manganese, the  causes for the observed concentration patterns
are more uncertain; however, they are given here to indicate plausible explanations. A cause
for the inconsistent chromium pattern is not known. Trends in beryllium and cobalt could
not be determined  because of the very low concentrations. Trends in some of the metals
studied have been correlated with known changes in emissions of these substances.  Nation-
wide lead averages have declined because of the  lower lead content of gasolines sold in recent
years, primarily due to the introduction of lower compression engines around 1970.  Vana-
dium, and nickel have dropped, particularly in the  Northeast because the desulfurization
of petroleum also removes these impurities.  Titanium may have increased due to the rise
in coal consumed by electric utilities.  The absence of trends in copper may be at least par-
tially explained by  a contamination problem from the commutator of the hi-vol sampler.
Decreasing iron, manganese, and cadmium  concentrations are probably related to reduced
particulate emissions from steel plants and related industries and from improved incinera-
tion and waste burning practices.

                     Table 3. TRENDS IN URBAN METAL CONCENTRATIONS
                               AND THEIR POSSIBLE CAUSES
                      Metal
                  Fuel-combustion-
                  related metals
                    Beryllium
                    Lead

                    Nickel
                    Titanium

                    Vanadium

                  Industry-related
                  metals
                    Cadmium
                    Chromium
                    Cobalt
                    Copper

                    Iron


                    Manganese
                                   Observed trends
   Unknown
Down last 5 years

    Down
     Up

    Down
    Down


   No trend
   Unknown
   No trend

    Down


    Down
                                                        Possible causes
 Lower lead content in gasolines
        after 1969
 Reduction of Ni in residual oils
   Increasing use of coal in
      electric utilities
 Reduction of V in residual oils
  Controls in metal industry
  and improved incineration
        practices
        Unknown
  Contamination from hi-vol
       commutator
 Improved incineration or waste
burning practices, fuel switching,
   controls in steel industry
  Controls in metals industry
                                          19

-------
                      6.   ACKNOWLEDGMENTS

   The authors wish to express their appreciation to Dr. Richard J. Thompson and his staff
of EPA's Environmental Monitoring and Support Laboratory for doing the chemical anal-
yses for the metals data used in this report.  The authors also acknowledge the many helpful
suggestions and comments provided by: Gerald G. Akland, Alan Hoffman, Justice Manning,
Gary McCutchen, and Jacob Summers, all of EPA. A special thanks goes to Mrs. Joan Bivins
for editing and typing the manuscript.
                                      21

-------
                              7.  REFERENCES

 1. Lead (also Manganese, Chromium, and Vanadium).  Committee on the Biologic Effects
   of Atmospheric Pollutants, Division of Medical Sciences, National Research Council,
   National Academy of Sciences, Washington, D.C. 1971, 1972, 1973, 1974.

 2. Air Quality Data for Metals, 1970-1974, from the National Air Surveillance Networks.
   U.S. Environmental Protection Agency, Office of Research and Development, Environ-
   mental Monitoring and Support Laboratory, Research Triangle Park, N.C. 27711.
   July 1976.

 3. Akland, Gerald.  U.S.  Environmental Protection Agency, Environmental Monitoring
   and Support Laboratory, Research Triangle Park, N.C. 27711. Personal communications,
   August 1976.

 4. Data file of Nationwide Emissions, 1970.  U.S. Environmental Protection Agency, Office
   of Air Programs, Applied Technology Division, National Air Data Center.  July 1972.

 5. National Emissions Inventory of Sources and Emissions of Chromium, 1970. GCACorp.,
   GCA Technology Division, Bedford, Mass.  01730.  Prepared for U.S. Environmental
   Protection Agency, Office of Air and Waste Management, Office of Air Quality Planning
   and Standards, Research Triangle Park, N.C. 27711.  May 1973.

 6. National Inventory of Sources and Emissions: Copper -1969. W.E. Davis and Associates,
   9726 Sagamore  Rd.,  Leawood, Kans.  Prepared for U.S.  Environmental Protection
   Agency, Office of Air and Waste Management, Office of Air Quality Planning and Stand-
   ards, Research Triangle Park, N.C. 27711. April 1972.

 7. National Emissions Inventory of Sources and Emissions of Titanium.  GCA Corp. GCA
   Technology Division, Bedford, Mass. 01730.  Prepared for U.S. Environmental Protec-
   tion Agency, Office of Air  and Waste Management, Office of Air Quality Planning and
   Standards, Research Triangle Park,  N.C. 27711.  May 1973.

 8. Trends in Air Lead Concentration.  Petroleum Laboratory, E.I. DuPont De Nemours
   and Co., Inc., Wilmington, Del. 19898. July 26,  1976.

 9. Motor Gasolines, U.S. Energy Research and Development Administration, Bartlesville
   Energy Research Center, Bartlesville, Okla.

10. California Air Quality Data, July-September 1973, Vol. V, No. 3. California Air Resources
   Board, Technical Services  Division.

11. Monitoring and  Air Quality  Trends Report,  1972.  U.S. Environmental Protection
   Agency, Office of Air and Waste Management, Office of Air Quality Planning and Stand-
   ards, Monitoring and Data Analysis Division,  Research Triangle Park, N.C. 27711.
   December  1973.

12. Vanadium.  Committee on the Biologic Effects of Atmospheric Pollutants,  Division of
   Medical Sciences, National Research Council, National Academy of Sciences, Washing-
   ton, D.C.  1974.

                                        23

-------
13. Mineral Yearbook - Minerals Fuels, 1965-1974. U.S. Department of the Interior, Washing-
   ton, D.C.

14. King, Robert B., and John Toma.  Copper Emissions from a High-Volume Air Sampler.
   NASA Technical Memorandum, Lewis Research Center, Cleveland, Ohio 44135. March
   1975.

15. Directory of Air Monitoring Sites, 1972.  U.S. Environmental Protection Agency, Office
   of Air Quality Planning and Standards, Monitoring and Data Analysis Division, Research
   Triangle Park, N.C. 27711.  September 1973.
                                        24

-------
                                  APPENDIX
   Only NASN sites meeting a minimum amount and distribution of data in the period 1965-
1974 were included in the trend analysis. In order to be included in the trend sample, a site
had to meet the following criteria:  (1) at least 2 years in the period 1965-69 had to be repre-
sented with at least three valid quarters and (2) at least 3 years in the period 1970-74 had to be
represented with at least three valid quarters,  with two of these years being 1970 or 1971
and 1973 or 1974. A valid quarter, which refers to the number of measurements submitted
and not to the validity of the measurement, has at a minimum 5 days of data spread through-
out the quarter.  If one of the months in the quarter does not have any samples, then the
other 2 months must have at least two samples apiece. Missing years of data were replaced
by either interpolating from data on both sides of the gap or by simply using the preceeding
or following annual average to replace missing annual average values at either end of the data
record.  A comparison  of the results for sites having a complete year—that is, four valid
quarters of data with that set having at least three quarters—did not reveal any important
bias in the results.

   The sites satisfying  these  basic requirements together with a brief description  of the
site15 are given below:
            City name

Gadsden, Alabama
Huntsville, Alabama
Mobile, Alabama
Montgomery, Alabama
Anchorage, Alaska
Phoenix, Arizona
Tucson, Arizona
Little Rock, Arkansas
West Memphis, Arkansas
Glendale, California
Long Beach, California
Los Angeles, California
Oakland, California
Sacramento, California
San Bernardino, California
San Diego, California
San Francisco, California
Hartford, Connecticut
New Haven, Connecticut
Waterbury, Connecticut
Newark, Delaware
Jacksonville, Florida
Tampa, Florida
 EPA
Region

  IV
  IV
  IV
  IV
  X
  IX
  IX
  VI
  VI
  IX
  IX
  IX
  IX
  IX
  IX
  IX
  IX
   I
   I
   I
  III
  IV
  IV
    Site description

Center city-industrial
Center city-commercial
*
Center city-commercial
Center city-commercial
*
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-industrial
Center city-industrial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-industrial
*
Center city-commercial
Center city-commercial
                                           25

-------
            City name
Atlanta, Georgia
Honolulu, Hawaii
Boise City, Idaho
Joliet, Illinois
Springfield, Illinois
East Chicago, Indiana
Terre Haute, Indiana
Cedar Rapids, Iowa
Davenport, Iowa
Des Moines, Iowa
Topeka, Kansas
Wichita, Kansas
Ashland, Kentucky
Covington, Kentucky
New Orleans, Louisiana
Shreveport, Louisiana
Baltimore, Maryland
Worchester, Massachusetts
Flint, Michigan
Grand Rapids,  Michigan
Trenton, Michigan
Duluth, Minnesota
St. Paul, Minnesota
St. Louis, Missouri
Omaha, Nebraska
Reno, Nevada
Concord, New Hampshire
Bayonne, New Jersey
Glassboro, New Jersey
Newark, New Jersey
Trenton, New Jersey
Albuquerque, New Mexico
New York City, New York
Charlotte, North Carolina
Durham, North Carolina
Bismarck, North Dakota
Akron, Ohio
Youngstown, Ohio
Tulsa, Oklahoma
Allentown, Pennsylvania
Altoona, Pennsylvania
Bethlehem, Pennsylvania
Harrisburg, Pennsylvania
Hazelton, Pennsylvania
 EPA
Region
  IV
  IX
   X
   V
   V
   V
   V
  VII
  VII
  VII
  VII
  VII
  IV
  IV
  VI
  VI
  III
   I
   V
   V
   V
   V
   V
  VII
  VII
  IX
   I
   II
   II
   II
   II
  VI
   II
  IV
  IV
 VIII
   V
   V
  VI
  III
  III
  III
  III
  III
    Site description
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Suburban-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
*
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
*
Suburban-commercial
Center city-residential
Suburban-commercial
Suburban-residential
Center city-commercial
Center city-commercial
Rural-commercial
Center city-residential
Center city-commercial
Center city-commercial
Center city-residential
Rural-commercial
Center city-industrial
Center city-residential
Center city-commercial
*
Suburban-commercial
Center city-commercial
Center city-residential
                                          26

-------
            City name
Pittsburgh, Pennsylvania
Reading, Pennsylvania
Scranton, Pennsylvania
Wilkes-Barre, Pennsylvania
York, Pennsylvania
Providence, Rhode Island
Columbia, South Carolina
Greenville, South Carolina
Memphis, Tennessee
Nashville, Tennessee
Dallas, Texas
Houston, Texas
San Antonio, Texas
Ogden, Utah
Burlington, Vermont
Hampton, Virginia
Norfolk, Virginia
Portsmouth, Virginia
Roanoke, Virginia
Seattle, Washington
Charleston, West Virginia
Kenosha, Wisconsin
Madison, Wisconsin
Milwaukee, Wisconsin
Casper, Wyoming
Grand Canyon Nat. Park, Arizona
Montgomery Co., Arkansas
Mesa Verde Nat. Park, Colorado
Butte Co., Idaho
Acadia Nat. Park, Maine
Jackson Co., Mississippi
Thomas Co., Nebraska
White Pine Co., Nevada
Coos Co., New Hampshire
Jefferson Co., New York
Cherokee Co., Oklahoma
Curry Co., Oregon
Black Hills Nat. For., South Dakota
Matagorda Co., Texas
Orange Co., Vermont
Yellowstone Nat. Park, Wyoming
""Information not available from National Aerometric Data Bank.
EPA
Region
III
III
III
III
III
I
IV
IV
IV
IV
VI
VI
VI
VIII
I
III
III
III
III
X
III
V
V
V
VIII
IX
VI
VIII
X
I
IV
VII
IX
I
II
VI
X
VIII
VI
I
VIII

Site description
Center city-commercial
*
Center city-commercial
Center city-commercial
*
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
*
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Center city-commercial
Remote
Remote
Remote
Remote
Remote
Remote
Remote
Remote
Rural
Remote
*
Remote
Remote
Remote
Remote
Rural
                                         27

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
     FPA-d.Wl-77-003
                                                           3. RECIPIENT'S ACCESSI ON-NO.
4. TITLE AND SUBTITLE
  National Trends
  1965-1974
      in Trace  Metals in Ambient Air  -
5. REPORT DATE
 February  1977
                                               6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
  Robert B.
  Thomas B.
                                                           8. PERFORMING ORGANIZATION REPORT NO.
Faoro
McMullen
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  U.S. Environmental  Protection Agency
  Office of Air  and  Waste Management
  Office of Air  Quality Planning and Standards
  Research Triangle  Park, North Carolina  27711
                                               10. PROGRAM ELEMENT NO.

                                                  2AF643
                                               11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
                                                            13. TYPE OF REPORT AND PERIOD COVERED

                                                              Special	
                                                            14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
       This report  examines trends over  the past 10 years  (1965-74) in ambient  con-
  centrations for  11  metals:   beryllium,  cadmium, chromium,  copper, cobalt,  iron,
  lead, manganese,  nickel, titanium, and  vanadium.  The trends  are derived from
  samples collected from 92 urban and  16  nonurban hi-vol stations in the National
  Air Surveillance  Network.  For purposes of presentation,  the  trace metals  were
  grouped into  two  broad categories:   the metals related to  fuel  combustion  and
  those related to  industry.   Possible reasons for the trends observed are given
  for most of the  metals studied.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS
                                                             c. cos AT I Field/Group
  Air Pollution Trends
  Beryl 1ium
  Cadmium
  Chromium
  Copper
  Cobalt
  Iron
                  Titanium
                  Vanadium
                  National Air
                    Surveillance
                    Network
                  Manganese
                  Nickel
      R1BUTION STATEMENT
  Release Unlimited
                                              19. SECURITY CLASS (This Report)
                                                Unclassified
                                                             21. NO. OF PAGES

                                                               41
                                              20. SECURITY CLASS (Thispage)

                                                Unclassified
                                                                         22. PRICE
EPA Form 2220-1 (9-73)
                                             28

-------