Calculating Air Quality and Its Control

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Journal of the Air Pollution Control Association

ISSN: 0002-2470 (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/uawm16

Calculating Air Quality and Its Control

Charles E. Zimmer A.M. & Ralph I. Larsen

To cite this article: Charles E. Zimmer A.M. & Ralph I. Larsen (1965) Calculating Air
Quality and Its Control, Journal of the Air Pollution Control Association, 15:12, 565-572, DOI:
10.1080/00022470.1965.10468424

To link to this article: https://doi.org/10.1080/00022470.1965.10468424

Published online: 16 Mar 2012.

Submit your article to this journal

Article views: 6727

View related articles

Citing articles: 3 View citing articles

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=uawm20
CHARLES E. ZIMMER, A.M.,
Calculating Air Quality and
Chief, Planning and Data Analysis
Unit, Air Quality Section, Laboratory
of Engineering and Physical Sciences
Its Control
RALPH I. LARSEN, Ph.D., Air quality is shown as a function of averaging times offiveminutes to one year for carbon
Assistant Chief, Field Studies Branch, monoxide, hydrocarbons, nitric oxide, nitrogen dioxide, nitrogen oxides, oxidant, and
Division of Air Pollution, Robert A. sulfur dioxide in Chicago, Cincinnati, Los Angeles, New Orleans, Philadelphia, San
Taft Sanitary Engineering Center, Francisco, and Washington, D. C. Concentrations are approximately lognormally
Public Health Service, U.S. Depart- distributed for all pollutants in all cities for all averaging times. Maximum concentra-
ment of Health, Education, and Wel- tion is inversely proportional to averaging time to an exponent. The exponent is a
fare function of the standard geometric deviation. General air quality and control parameters
are derived and shown for one example, nitrogen oxides in Washington, D. C. These
values are compared to one air quality standard.

Mir pollution control officials and time constant.1 The second analysis half of the five-minute values are avail-
researchers measuring air quality have a can be obtained either by averaging able for an averaging period, concentra-
wide variety of air sampling devices at continuous data over a specified time tions are calculated by summing these
their disposal. Such devices range in period or by using integrating samplers values and dividing by their number.
complexity from relatively simple and such as the high-volume sampler,5 the If half or less of the values are available,
inexpensive dustfall collectors and lead AISI sampler, various bubblers, or the no value is calculated for that period.
peroxide candles to the more sophisti- dustfall jar. Cumulative frequency distributions of
cated, and consequently expensive, concentrations of gaseous pollutants for
instruments that provide continuous Concentration Versus Averaging selected averaging times are given in
measurements.1 Each device provides Time and Frequency detail for one example, nitrogen oxides
some indication of air quality; however, in Washington, D. C. (Table I), and
each device also imposes its own restric- Since January 1962 the Public Health in summary for carbon monoxide (Table
tions and limitations upon the resultant Service has operated a Continuous Air II). Maximum concentrations for vari-
data. The choice of air sampling in- Monitoring Program (CAMP), in which ous averaging times are given for carbon
struments to be used in a particular air pollutant concentrations are punched monoxide, hydrocarbons, nitric oxide,
investigation depends upon the objec- automatically into computer paper tape nitrogen dioxide, nitrogen oxides (NO +
tives of the study and the availability of every five minutes. For analysis of NO2), total oxidants, and sulfur dioxide
funds. these data, averaging times are selected in Chicago, Cincinnati, Los Angeles,
Air quality data can be analyzed in that begin at the first minute of a pollu- New Orleans, Philadelphia, San Fran-
two fundamental ways: (1) by the tion season (usually spring, taken here cisco, and Washington, D. C. (Table
frequency and length of time that a as beginning March 1; but when data III). Table I indicates that the mean
certain concentration is exceeded2"4; are scarce, later periods are used). The concentration of nitrogen oxides in
and {2) by concentration averaged over eight-hour-average values then conclude Washington, D. C , during the period
a specified time interval. The first at every 0800, 1600, and 2400 hour time from March 1, 1962, to March 1, 1963,
analysis requires continuous data meas- thereafter. Other averaging-time values was 0.06 ppm. The maximum five-
ured by an instrument with a fairly short are calculated similarly. If more than minute concentration during this period

Table I—Concentration vs. Averaging Time and Frequency for


Nitrogen Oxides (ppm) in Washington, D. C , March 1 , 1962 to March 1, 1963
Aver-
Per- p r r . ^6nt oi J. line if ra firm
lbrtltlOIJL I S J C J X C I
Time Mean Max Min cent 0.001 0.01 0 .1 1 10 20 50 60 70 80 90 99 99.9 99.99 99.999
30 40
5 min 0.06 0.79 0.00 70 0.79 0.77 0 . 53 0.35 0.13 0.09 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.00 0. 00 0.00 0.00
10 0.06 0.79 0.00 70 0.75 0 . 53 0.35 0.13 0.09 0.07 0.06 0.05 0.04 0.03 0.02 0.02 0.00 0. 00 0.00
15 0.06 0.78 0.00 70 0.74 0 . 53 0.35 0.13 0.09 0.07 0.06 0.05 0.04 0.03 0.02 0.02 0.00 0. 00 0.00
30 0.06 0.78 0.00 69 0.69 0 . 53 0.35 0.13 0.09 0.07 0.06 0.05 0.04 0.03 0.02 0.02 0.00 0. 00 0.00
1 hour 0.06 0.74 0.00 70 0 . 51 0.34 0.13 0.09 0.07 0.06 0.05 0.04 0.03 0.02 0.02 0.00 0. 00
2 0.06 0.64 0.00 68 0 . 53 0.33 0.13 0.09 0.07 0.06 0.05 0.04 0.03 0.02 0.02 0.00 0. 00
4 0.06 0.51 0.00 71 0. 49 0.32 0.12 0.09 0.07 0.06 0.05 0.04 0.03 0.03 0.02 0.01 0. 00
8 0.06 0.43 0.00 73 0. 40 0.27 0.12 0.08 0.07 0.06 0.05 0.04 0.03 0.03 0.02 0.01 0. 00
12 0.06 0.35 0.00 77 0.26 0.12 0.09 0.07 0.06 0.05 0.04 0.04 0.03 0.02 0.01
1 day 0.06 0.29 0.01 73 0.22 0.12 0.08 0.07 0.06 0.05 0.05 0.04 0.03 0.03 0.02
2 0.06 0.25 0.02 66 0.20 0.10 0.08 0.07 0.06 0.05 0.05 0.04 0.04 0.03 0.02
4 0.06 0.18 0.03 69 0.10 0.08 0.07 0.06 0.05 0.04 0.04 0.04 0.03
7 0.06 0.13 0.03 73 0.10 0.07 0.07 0.06 0.06 0.04 0.04 0.04 0.03
14 0.06 0.13 0.03 77 0.11 0.08 0.07 0.06 0.06 0.05 0.04 0.04 0.03
1 month 0.06 0.11 0.04 83 0.10 0.08 0.07 0.06 0.06 0.04 0.04 0.04 0.04
2 0.05 0.08 0.04 67 0.04 0.04 0.04 0.04 0.04
3 0.06 0.08 0.04 75 0.05 0.04 0.04 0.04 0.00
6 0.07 0.09 0.05 100 0.05
1 year 0.06 0.06 0.06 100

December 1965 / Volume 15, No. 12 565


Table II—Carbon Monoxide Concentration (ppm) at Camp Sites vs. Averaging Time and Frequency

%
-Period- Averaging Data —Percent of Time Concentration Is Exceeded
City End Time Mean Max Min Avail 0.001 0.01 0.1 1 10 30 50 70 90
Chicago 9/1/62 9/1/63 5 minutes 8 50 0 49 50 47 35 25 15 10
1 hour 8 36 0 49 26 21 14 10
8 hours 8 22 1 50 22 18 13 10
1 day 8 19 3 50 15 12 9
1 week 8 13 5 50 10 9
1 month 8 10 7 58 9 7 7
1 season 8 9 7 50 7 7
Los Angeles 3/1/62 3/1/63 5 minutes 11 67 0 74 66 41 34 25 16 12 10
1 hour 11 36 1 74 32 24 16 12 10
8 hours 11 28 5 76 25 21 15 12 10
1 day 11 22 6 78 19 14 11 10 10
1 week 11 15 8 79 12 12 11 10
1 month 11 13 9 75 12 11 11 9
1 season 11 12 10 75 12 10 0
Philadelphia 9/1/62 9/1/63 5 minutes 13 52 0 16 52 52 48 32 19 15 12 9
1 hour 13 47 4 16 46 32 19 15 12 9
8 hours 13 36 4 16 35 30 18 15 12 9
1 day 13 25 6 16 23 18 14 13 9
1 week 13 18 8 13 17 13 12 11
1 month 14 14 13 17 13 13 13 0
1 season

was 0.79 ppm. The minimum concen-


Table III—Maximum Concentration of Various Pollutants for Various tration recorded was zero, an indication
Averaging Times at Camp Sites, 1962 that the concentration was less than
the minimum detectable by the analyzer.
Averaging Maximum Concentration One percent of the five-minute values
City Time CO IIYC NO NO 2 N O . (ppm;
OXI so3 exceeded 0.35 ppm. Ninety percent
Chicago 5 minutes 50 19 0.66 0.26 0.78 0.16 1.94 exceeded 0.01 ppm. The median con-
1 hour 36 • 12 0.61 0.22 0.69 0.11 1.69 centration (50 percentile) was 0.05 ppm.
S 8 hours 22 8 0.41 0.16 0.54 0.05 0.87 Similarly, the one-day-average concen-
1 day 19 6 0.36 0.12 0.43 0.02 0.71 tration maximum was 0.29 ppm and the
1 week 13 4 0.25 0.08 0.31 0.01 0.47
1 month 10 4 0.17 0.06 0.21 0.01 0.33 one-day-average concentration exceeded
1 season 9 4 0.13 0.06 0.16 0.01 0.28 0.22 on one percent of the days.
1 year (mean) 8 3 0.11 0.04 0.15 0.00 0.14 The cumulative frequency distribu-
Cincinnati 5 minutes 25 0.60 0.30 0.69 0.20 0.68
1 hour 17 0.58 0.25 0.68 0.14 0.48 tion (including the minimum and maxi-
8 hours 11 0.38 0.14 0.50 • 0 .10 0.23 mum values) for various averaging times
1 day 10 0.29 0.09 0.37 0.06 0.11.., can be plotted manually (Figs. 1 and 2)
1 week 6 0.07 0.07 0.09 0.04 0.07 or by computer (Fig. 3). The computer
1 month 5 0.06 0.04 0.09 0.03 0.06
1 season 4 0.05 0.03 0.07 0.03 0.05 plots appear as if typed by a pica type-
1 year (mean) 3 0.03 0.03 0.06 0.01 0.04 writer (six characters per vertical inch
Los Angeles 5 minutes 67 2.11 0.59 2.16 0.53- 0.22 and 10 characters per horizontal inch).
1 hour 36 0.80 0.54 1.07 0.49 0.19 The M values on Fig. 3 indicate maxi-
8 hours 28 0.56 0.37 0.71 0.31 0.09 mum or minimum concentrations. The
1 day 22 0.38 0.20 0.54 0.16 0.07
1 week % 15 0.23 0.13 0.35 0.07 0.03 top 1 is the 0.1 percent frequency.
1 month 13 0.18 0.10 0.28 0.06 0.03 The top 0 is the one percentile. The
1 season 12 0.15 0.10 0.25 0.02 lower 1 and digits 3 to 9 are the 10 to
1 year (mean) 11 0.08 0.06 0.13 0.04 0.02
New Orleans 5 minutes 0.42 0.14 0.49
90 percentiles. The low 0 is the 99
0.18 0.31 percentile. The H values are a plot of
1 hour 0.36 0.13 0.43 0.17 0.10
8 hours 0.15 0.07 0.20 0.11 0.04 the hour-average cumulative distribu-
1 day 0.08 0.05 0.12 0.08 0.03 tion of frequency on a logarithmic
1 week 0.04 0.04 0.07 0.06 0.02 (vertical scale) by normal probability
1 month 0.03 0.03 0.05 0.04 0.02
1 season 0.02 0.02 0.04 0.03 0.02 (lower scale) basis, also shown as a hand
1 year (mean) 0.02 0.02 0.04 0.02 0.01 plot (Fig. 2). Averaging times for Fig.
Philadelphia 5 minutes 52 15 2.63 0.36 2.84 0.12 1.25 3 are on the top scale.
1 hour 47 13 2.37 0.32 2.59 0.10 1.03
. Lt;>' 8 hours 36 7 1.35 0.19 1.49 0.08 0.48 Individual plots, like Fig. 3, of maxi-
L• 1 day 25 5 0.76 0.13 0.89 0.05 0.30 mum concentration and the one-hour-
1 week 18 4 0.19 0.07 0.28 0.02 0.19 average cumulative frequency distribu-
1 month 14 3 0.08 0.05 0.13 0.02 0.10 tion have been printed by the computer.
1 season 3 0.08 0.04 0.12 0.01 0.09
1 year (mean) 3 0.05 0.04 0.09 These have been combined into Fig. 4
13 0.01 0.08 to show the overall data trend. The
San Francisco 5 minutes 17 1.39 0.50 1.47 0.32 0.16
1 hour 12 0.65 0.26 0.72 0.26 0.11 plots indicate two trends:
8 hours 8 0.44 0.16 0.52 0.16 0.07
1 day 6 0.23 0.09 0.27 0.09 0.05 (1) The plot of maximum concentra-
1 week 4 0.15 0.07 0.20 0.05 0.02 tion versus averaging time
1 month 3 0.11 0.06 0.16 0.03 0.01 approximates a straight line on
1 season 3 0.11 0.05 0.16 0.02 0.01
1 year (mean) 2 0.07 0.03 0.10 0.02 0.01 logarithmic paper, indicating that
Washington 5 minutes 20 0.68 0.37 0.79 0.17 0.56 maximum concentration can be
1 hour 17 0.65 0.30 0.74 0.13 0.48- expressed as a function of averag-
8 hours 14 0.39 0.11 0.43 0.10 0.35-
-' 1 day 8 0.22 0.09 0.29 ing time to an exponent.
0.06 0.25
1 week 5 0.09 0.05 0.13 0.03 0.14 (2) For a given averaging time (one-
i 1 month 3 0.07 0.04 0.11 0.03 0.11 hour averages are shown) con-
\ 1 season 3 0.05 0.03 0.08 0.03 0.10 centration versus frequency of
1 year (mean) 3 0.03 0.03 0.06 0.01 0.06 occurrence plots as a straight line

566 Journal of the Air Pollution Control Association


on logarithmic by normal (log- AVERAGING TIME

normal) probability paper.6"8


i n i i i i n i i n i i n i i r m r
Some of the data fit these trends very
well, some only fairly. Some curve
slightly upward, some downward. On
the average, the data tend to fit these
trends, especially the most polluted half
of the samples. Some of the lower con-
centrations are so near thel owest detect-
able limit that they do not plot well.
Some of the maximum concentrations
for averaging times less than 30 minutes
fall below expected levels. This is
especially true of nitrogen oxides, since
the analyzer has a long averaging time.1 OOOI O.OI I 10. 100 1000 10,000 100,000

The time constant for nitrogen oxides AVERAGING TIME, JHOURS

samples is 10 minutes. Thus, if the Fig. 1. Frequency of various nitrogen oxides concentrations for various
sampled concentration were abruptly averaging times in Washington, D. C , March 1, 1 962 to March 1, 1 963.
changed from zero to one ppm, the
values would read only 0.63 ppm after
a 10-minute exposure. None of the
data have been corrected for slow re-
sponse.
REDUCED VARIATE.Z, NO. OF STANDARD DEVIATIONS
Pollutant Parameters
Frequency distributions that tend to'
be normal can be characterized by the
average and standard deviation. For a
lognormal distribution these values are
the geometric mean (approximately the
50 percentile) and the standard geo-
metric deviation (approximately the 16
percentile divided by the 50 percentile) .8
To demonstrate a technique, both pa- GEOMETRIC DEVIATION.

rameters will be calculated from a line


fitted to two points of the cumulative
frequency distribution of one-hour-
average concentration. The 30 and 0.1
percentiles will be used here. The 30-
percentile concentration is usually high
enough above the minimum detectable
value to be read with reasonable ac-
curacy. The 0.1 percentile is close 999 99.99
enough to the maximum to fit it well.
Thus a line fitted to these two points Fig. 2. Frequency of various one-hour averaging time nitrogen oxides
describes the upper half of the distribu- concentrations in Washington, D. C , March 1, 1962 to March 1, 1963.
tion quite well. For oxidant, only a
little more than 10 percent of the values
may be significantly above the minimum
detectable level. In this case, fairly
good fits have been obtained with the
10 percentile and 0.1 percentile.
The horizontal scale on probability
graph paper (Fig. 2) may be expressed
as percentage points of the cumulative
frequency distribution or as standard
deviation units (z values). The 30 per-
centile corresponds to z = 0.52, and the
0.1 percentile is at z = 3.09 (Table IV).
The concentration at any z value can
be calculated.7- 8
c = nigSg* (1)
where c is the concentration at z, mg is
the geometric mean, and sg is the stand-
ard geometric deviation. If two points
of the cumulative frequency distribu-
tion are known, sg may be calculated.

Z1
Fig. 3. Concentration vs. averaging time and frequency for nitrogen oxides
Sn = e
(2) in Washington, D. C , March 1, 1 962 to March 1, 1 963.

December 1965 / Volume 15, No. 12 567


NEW SAN WASHINGTON,
CHICAGO CINCINNATI LOS ANGELES ORLEANS PHILADELPHIA FRANCISCO D.C.
AVERAGING T I M E , hours
I 100 10,000 I 100 10,000 I 100 10,000 I 100 10,000 I 100 10,000 I 100 10,000 I 100 10.000
100 l l l I I I I I I I I I I I I I I I I I I I I I
- 30ppm
••••••••>.•'• •••••• ...•'
CALIFORNIA
STANDARD
CARBON MONOXIDE

"* ""*.*••• " • ••••".«••*


I -
HYDROCARBONS

100 -

NITRIC OXIDE
' • • • • « » *

0.01 -

NITROGEN <
DIOXIDE £ * "'•••"«**
••••%•*
5 o.oi -

NITROGEN
OXIDES - 0 . 1 ppm
0.01 - COLORADO
STANDARD

I -
OXIDANT
-0.15 ppm
EYE IRRITATION;
0.01 - THRESHOLD

SULFUR DIOXIDE • 0.3 ppm


TASTE
- THRESHOLD
1111111 rrtTT mini I till 11 iijiiii
SO ,50 50 50 50 50
FREQUENCY, % OF TIME

Fig. 4. Maximum concentration for various averaging times, and frequency of various pollutants in various cities.

where e is the base of natural logarithms, in = (0.047)(2.16)°- 51n2 - 16 (8)


2.718, In indicates natural logarithms,
CJJ is the concentration at the first z m = 0.063 ppm b =
. x2
value, cZ2 is concentration at the second In —
z value, and zx and z2 are the z values. The expected frequency of the extreme xl (12)
Substituting into Equation 2, using the concentrations (maximum and mini-
0.1 and 30 percentiles for one-hour mum) is approximately 9 In m
averaging time values of nitrogen oxides b =
s __ 60% 8760 hours
in Washington, D. C. (Table I), (9) ln
N 1 hour

sa = e
[ ,
ln
0T07
0.51 -i

3.09-0.52 J (3)
where / is percent frequency and N is
the number of samples. For the maxi-
Or, using Equations 1 and 7,
ln s z
mum one-hour concentration this fre- In " ~
s0 = 2.16 (13)
quency corresponds to z = 3.81 (Table ln 8760
IV). Substituting in Equation 1,
Equation 1 can be rearranged to solve b = -0.29
for the geometric mean. = (0.047) (2.16)3-81 (10) The coefficient of the equation is the
: = 0.9 ppm expected one-hour maximum (Equation
mg = (4) 10). Thus the equation is
V
c min = (0.047) (2.16) ~3-81 01) = o.9r 0 - 29 (14)
0.07
mg (5) cmin = 0.0025 ppm
~ 2.16 where t is averaging time in hours.
mg == 0.047 ppm Similarly, the minimum is
Concentrations for other percentiles can
be calculated and plotted (Table IV and c min = 0.0025i0-36 (15)
The arithmetic mean can now be cal- Fig. 1).
culated.7 The equation for the maximum con- The frequency distributions for other
In m = In mg + 0.5 In2 sg (6) centration line (Fig. 1) can be calculated averaging times can be plotted by draw-
from two points. The one-hour-average ing lines through the one-hour values
Or, taking antilogs,
maximum and the annual mean will be and the maximum concentration line
m = nigsg0-5 l n S9
(7) used here. The slope of the line is10 (at averaging times noted in Table IV).

568 Journal o f the Air Pollution Control Association


REDUCED VARIATE , Z, NO. OF STANDARD DEVIATIONS
Frequency distributions for other
averaging times can be calculated if two
points are available. One point is the
arithmetic mean, which is the same for
all averaging times. The other is the
maximum concentration. Equations 1
and 7 can be combined.
_ „ z — 0.5 In sg (16)
m
Taking logarithms, solving the resulting
quadratic equation, and taking antilogs,

V m )
s, = e (17)
Solving sg for 24-hour averaging time,
2.94-( 2 . 942 -2 1 n ^ ) 1 / 2 (18)
sg — e
10 16 20 30 40 50 60 70 80 99.9 99.99
sa = 1.94 FREQUENCY, % OF TIME

As mentioned earlier, the standard Fig. 5. Frequency of various 24-hour averaging time nitrogen oxides con-
centrations in Washington, D. C , March 1, 1962 to March 1, 1963.
geometric deviation describes concen-
tration spread. If the standard geo- AVERAGING TIME
HOURS DAYS MONTHS
metric deviation were one, Fig. 1 would
appear as one horizontal line. All sam-
ples for all averaging times would be
the same. The higher the deviation,
the greater the spread of the fan in
Fig. 1.
By the above calculations, all of the
lines on Fig. 1 have been constructed
from only two points, the 0.1 and 30
percentile concentrations of one-hour
averaging time. The lines could have
been calculated from any two points,
at the same or at different averaging-
times. Observed values (Table I) have
000 10,000 100,000
been plotted for comparison. Although AVERAGING T I M E . HOURS
most of the values fit very well, almost
all of the observed maxima are less than Fig. 6. Expected frequency of various nitrogen oxides concentrations for
various averaging times in Washington, D. C , March 1, 1962 to March 1,
expected. Just the reverse occurs for 1963.
sulfur dioxide in Chicago; most of the
observed maxima are slightly greater
than expected.
26 Samples per Year
Figure 1 is based on 70,000 five-minute For the 22 samples available the and r is the rank order (highest, second
values. The National Air Sampling geometric mean is 0.054 (Equation 19) highest, third highest, etc.). The con-
Network (NASN)5- » and National Gas as compared with 0.051 calculated from centration is calculated with Equation 1.
Sampling Network (NGSN)12 take 26 the one-hour data (Equation 4). The For Fig. 6, the one-hour maximum is
random 24-hour-duration samples per standard geometric deviation is 1.76 0.9 ppm. The fourth highest concen-
year at each station. How well would (Equation 20) as compared with 1.94 tration (/ = 0.041%, z = 3.34) is about
such a sampling program have approxi- calculated from the one-hour data two-thirds this value, 0.62 ppm. The
mated Fig. 1? A set of 24-hour-dura- (Equations 1, 14, 17, and 18). Thus a 12th highest (exceeded an average of
tion random values were selected from fairly good approximation to Fig. 1 once per month) is 0.53 ppm (z — 3).
the CAMP data by use of an NASN could have been obtained with these 22 All of the foregoing analyses have
sampling schedule (Fig. 5). The random samples. Future analyses are showed results for only one year of data.
logarithms of these concentrations were planned to determine how well various But air pollution is worse in some years
used to calculate the geometric mean sampling patterns would approximate than in others. In fact there may even
and standard geometric deviation. the continuous data. be a cyclic trend. The 1930 Meuse
Valley air pollution incident, the Decem-
High Concentrations ber 1952 London incident, the November
L AT J ,in, 1953 New York incident, and the De-
Concentrations can be ranked by fre-
mg = e (19) cember 1962 London incident all oc-
quency (Fig. 1) or order (Fig. 6).
where mg is the geometric mean, 2 In x curred during periods of low sun spot
Ordered values plot almost parallel to
is the summation of the logarithms of all activity (11-year cycle).10 Several years
the maximum concentration line. The
the observed concentrations, and N is
frequency is determined as9 of data (possibly one solar cycle, 11
the number of observations. For the years) would increase the reliability of
standard geometric deviation, predictions. The graph paper shown
Sin *x - N (In J»B) 2 "1V2 (Fig. 1) allows the plotting of data for
N - 1 (20) where N is the total number of samples averaging times from a second to a
sa = e

December 1965 / Volume 15, No. 12 569


century. If many years of data are Table IV—Plotting Position of Extreme Concentrations and Various Per-
used, the overall average effect of year- centiles for Various Averaging Times
by-year change in pollutant source
strength must be adjusted before com- ?
Plnttimr Pr
X 1\J V t l l l g , X UOA \JJL\Jli. •»

bining years. Reduced


If all sources are controlled by ap- Variate z
(No of std.
proximately equal percentages, then the Averaging time No. of samples Frequency, 60%/iV deviations
distribution after control will be shaped (hours) in Year (% of time) from mean)
the same as before control (Fig. 1), only 1 sec 0.000278 31,500,000 0.0000019 5.50
the concentration axis will be shifted.8

1—1
min 0.0166 525,000 0.0001142 4.73
For instance, if 90 percent reduction of 5 0.0833 105,000 0.000571 4.39
all nitrogen oxides sources in Washing- 8 .8 0.146 60,000 0.001 4.27
ton, D. C, occurred, one ppm on the axis 10 0.166 52,500 0.001142 4.24
15 0.25 35,000 0.001715 4.14
would become 0.1 ppm, 0.1 ppm would 30 0.5 17,500 0.00343 3.98
become 0.01 ppm, etc.

i—i
hour 1 8,760 0.00685 3.81
1 .46 1.46 6,000 0.01 3.72
Air Quality Standards 2 2 4,380 0.0137 3.63
4 4 2,190 0.0274 3.46
The ultimate purpose of air sampling 8 8 1,095 0.0548 3.26
is to determine if source reduction is 12 12 730 0.0822 3.14
needed, and if so, how much. The pur- 14 .6 14.6 600 0.1 3.09
1 day 24 365 0.1644 2.94
pose of this section is to demonstrate a 2 48 183
method for calculating the degree of air 0.328 2.72
4 96 91 0.657 2.48
quality control needed to meet any 5 .90 146 60 1 2.33
standard. It is not the purpose of this '7 168 52 1.153 2.27
14 346 26 2.31 1.99
paper to recommend any particular air
i—i

month 730 12 5 1.64


quality standard. 2 1460 6 10 1.28
The California standard for adverse 3 2190 4 15 1.04
air quality tends toward levels that are 6 4380 2 30 0.52
1 year 8760 1 50 0
not expected to significantly harm or
impair man or plant (Table V).8> 13
The levels allow some sensory irritation analyses could be tailored to the stand- would achieve the Colorado standard
(SO2 and oxidant). Two of the values ard instead. Also note that observed at the CAMP site in Washington, D. C.
are close to those for air in an industrial data are used here. Calculated data Mathematical meteorological modeling
environment as promulgated by the could have been used.8 may be used to determine how much air
ACGIH threshold limits (eight-hour The degree of overall average air quality would be improved by selectively
averages). The ACGIH recommends quality control required to meet a given controlling individual sources.18- 19
0.1 ppm for ozone; California recom- standard can be calculated. Negative numbers in Table VI indi-
mends 0.15 ppm (one-hour average) for cate the percent increase in source
oxidant. California recommends 30 (100%)(c-s) strength that would just exceed the
ppm (eight-hour average) for carbon R = (22)
standard. Note that a 20 percent in-
monoxide; the ACGIH tentative value crease in carbon monoxide emissions in
is 50 ppm. where R is the percent reduction that
will achieve the standard, c is the present Los Angeles could occur at the CAMP
The Colorado14 and West Germany station site (Tables II and VI) before
(VDI)15 standards tend toward values or future predicted concentration, and
s is the standard concentration.8 By use the California standard would be ex-
that would allow only infrequent sensory ceeded. California plans indicate that
irritation. The USSR values are de- of values for nitrogen oxides in Wash-
ington, D. C. and the Colorado stand- a 60 percent reduction in average carbon
signed to prevent the first human re- monoxide emissions will be needed by
sponse.15"17 Biologic instruments have ard,
1970.13'20 One reason for this difference
been used to determine decreased adap- (100%) (0.51 - 0.1) in calculated reduction is that 40 percent
tation to dark, shift in optical chronaxie, R = (23) more carbon monoxide is expected to be
or any other altered human response. 0.51
emitted in 1970 than in 1962.13 A
Air quality standards can be com- R = 80% second very important reason for this
pared with either tabular or graphical difference is that the CAMP site is not
air quality data. The Colorado stand- Thus an overall average reduction of 80 as polluted by carbon monoxide as are
ard for nitrogen oxides is 0.1 ppm for a percent in nitrogen oxides emissions many other locations in Los Angeles.
one-hour average for no more than one
percent of the hours in any three months
(Table V). Comparison of this standard Table V—Air Quality Standards for Various Pollutants
with Washington, D. C, data shows that n nTir , entratior1
20 percent of the one-hour average > VPP111/
concentrations exceed 0.1 ppm (Table Percent of Nitro-
Time Standard Carbon gen Nitro-
I, Figs. 1 to 4). The longest averaging State or Averaging Can Be Mon- Di- gen Sulfur
time for which 0.1 ppm was exceeded Country Time Exceeded oxide oxide Oxides Oxidant Dioxide
was one month (1000 hours). California14 8hr. 30 0.3
1 hr. 120 0.15 1
Calculating the Degree of Colorado13 24 hr. 1% of any 0.1
Air Quality Control 3 months
1 hr. 1% of any 0.1 0.1 0.5
Note that the air quality standard 3 months
parameters (Table V) have been altered W. Germany 30 min. 0.5 0.2
5
slightly (Table VI) to fit the available (VDI)i
16 17 short 1 0.3
data (e.g., Tables I and II). If just one USSR - 24 hr. 0.8 0.06 0.06
20 min. 4.8 0.18 0.18
standard were being used, the air quality

570 Journal of the Air Pollution Control Association


Table VI—Source Reduction That Would Achieve Various Air Quality Standards in Various Cities

Standard Compared with Source Reduction (%) That Would


Pollutant, Parameters, Measured Concentration (ppm) Achieve Various Air Quality Standards
and City Calif. Col. W. Germ. USSR Calif. Col. W. Germ. USSR
Carbon Monoxide
Parameters Used
Averaging time 8hr. 24 hr.
Percent of time exceeded 0.1% 1%
Standard Concentration 30 1
Measured Concentration
Chicago 22 15 -36 93
Los Angeles 25 19 -20 95
Philadelphia 35 23 14 96
Nitrogen Oxides
Parameters Used
Averaging time 1 hr. 1 hr. 24 hr.
Percent of time exceeded 0.1% 0.1% 1%
Standard Concentration 0.1 0.5 0.06
Measured Concentration
Chicago 0.63 0.63 0.31 84 21 81
Cincinnati 0.42 0.42 0.17 76 -19 65
Los Angeles 0.88 0.88 0.48 89 43 88
New Orleans 0.31 0.31 0.10 68 -61 40
Philadelphia 1.83 1.83 0.43 95 73 86
San Francisco 0.60 0.60 0.24 83 17 75
Washington, D. C. 0.51 0.51 0.22 80 2 73
Oxidant
Parameter Used
Averaging Time 1 hr. 1 hr.
Percent of time exceeded 0.1% o.i%
Standard Concentration 0.15 0.1
Measured Concentration
Chicago 0.09 0.09 -67 -11
Cincinnati 0.10 0.10 -50 0
Los Angeles 0.32 0.32 47 69
New Orleans 0.15 0.15 0 33
Philadelphia 0.09 0.09 -67 -11
San Francisco 0.17 0.17 12 31
Washington, D. C. 0.11 0.11 -36 9
Sulfur Dioxide
Parameters.Used .
Averaging time 8hr. 24 hr. 1 hr. 24 hr.
Percent of time exceeded 0.1% 1% 0.1% 1%
Standard Concentration 0.3 0.1 0.2 0.06
Measured Concentration
Chicago 0.73 0.49 0.93 0.49 59 80 78 88
Cincinnati 0.19 0.10 0.27 0.10 -58 0 26 40
Los Angeles 0.08 0.05 0.11 0.05 -275 -100 -82 -20
New Orleans 0.04 0.03 0.07 0.03 -650 -233 -186 -100
Philadelphia 0.47 0.27 0.64 0.27 36 63 69 78
San Francisco 0.07 0.04 0.09 0.04 -330 -150 -120 -50
Washington, D. C. 0.26 0.17 0.34 0.17 -15 41 31 65

Control must be based on the worst Summary Acknowledgments


situation that is to be rectified, whether Air quality is shown as a function of
Sincere thanks are expressed to Ed
by control of individual sources or by averaging times of five minutes to one
Diephaus, Supervisor, and Paul Britton
overall measures. Probably in no case year for carbon monoxide, hydrocar-
and Charles Hampel, Programmers, of
bons, nitric oxide, nitrogen dioxide,
the Sanitary Engineering Center Data
is the CAMP station measuring the
nitrogen oxides, oxidant, and sulfur
Processing Unit for many of the com-
worst air quality in a city. Other sites puter programs and all of the data
must be located, possibly with less dioxide in Chicago, Cincinnati, Los
Angeles, New Orleans, Philadelphia,
processing for this study.
frequent sampling, to find the worst air
quality situation that is to be controlled. San Francisco, and Washington, D. C. REFERENCES
Either this, or the worst situation can be Concentrations are approximately log- 1. R. I. Larsen, F. B. Benson, and G. A.
normally distributed for all pollutants in Jutze, "Improving the Dynamic Re-
estimated from the data from a single sponse of Continuous Air Pollutant
station.21 all cities for all averaging times. Maxi- Measurements with a Computer," J.
mum concentration is inversely propor- Air Poll. Control Assoc, 15: 19-22
The values in Table VI are based on tional to averaging time to an exponent. (January 1965).
1962 pollutant concentrations. In plans The exponent is a function of the stand- 2. R. C. Wanta and A. C. Stern,
for the future, growth must be con- "Classification of Air Pollution Ex-
ard geometric deviation. General air posures," Amer. Industr. Hyg. Assoc.
sidered.8 If present trends continue, quality and control parameters are Quarterly, 18: 156-160 (June 1957).
by the year 2000 sulfur oxides emissions derived and shown for one example, 3. R. I. Larsen, "Parameters of Aero-
will double, nitrogen oxides emissions nitrogen oxides in Washington, D. C. metric Measurements for Air Pollution
will triple, and carbon monoxide emis- Research," Amer. Industr. Hyg. Assoc.
These values are compared to one air J., 22: 97-101 (April 1961).
sions will quadruple.22 quality standard. 4. T. R. Oliver, "Mathematical Methods

December 1965 / Volume 15, No. 12 571


of Measuring Exposure to Air Pollut-
ants," presented in Houston at the
annual American Industrial Hygiene
NOMINATING COMMITTEE SEEKS
Conference (May 6, 1965).
5. Air Pollution Measurements of the
National Air Sampling Network, 1957-
MEMBERSHIP'S RECOMMENDATIONS
1961, Public Health Service Publica-
tion No. 978, Government Printing
Office, Washington 25, D. C. (1962).
6. C. E. Zimmer and G. A. Jutze, "An This year for the first time APCA's (2) Five Directors to serve accord-
Evaluation of Continuous Air Quality members will have an opportunity to ing to the following schedule: Two in-
Data," / . Air Poll. Control Assoc,
14: 262-266 (July 1964). nominate candidates for First Vice dustry representatives—to serve three
7. R. I. Larsen, "United States Air President instead of President. In years. One representative from the
Quality," Arch. Environ. Health, 8: accordance with the newly amended research, education, and consultant
325-333 (Feburary 1964). By-laws approved by the membership category—to serve three years. One
8. R. I. Larsen, "A Method for De-
termining Source Reduction Required this year, the Nominating Committee representative from the research, edu-
to Meet Air Quality Standards," J. may recommend multiple nominees for cation, and consultant category—to
Air Poll. Control Assoc, 11: 71-76 vacant offices. You, as an individual serve two years. One representative
(February 1961). member or company representative of from the research, education, and con-
9. E. S. Pearson and H. O. Hartley, APCA, are requested to submit to the
Biometrika Tables for Statisticians, 1: sultant category—to serve one year One
175, Cambridge University Press, chairman or any member of the Nom- control official—to serve two years.
London (1956). inating Committee, names of individuals Retiring Board Members are: Dr.
10. R. I. Larsen, "Determining Basic and the offices for which you feel they L. C. McCabe, Dr. E. R. Hendrickson,
Relationships Between Variables," are qualified. These recommendations
Symposium on Environmental Meas- Gordon W. Barr and James V. Fitz-
urements: 251-263, PHS Publication should be received before February 1, patrick.
999-AP-15, Sanitary Engineering Cen- 1966. The members of the Nominating Due to a change in the nature of the
ter, Cincinnati, Ohio (July 1964). Committee, all of whom have been activity of the U.S. Public Health Serv-
11. R. I. Larsen, "Choosing an Aero- members of the Association for a min-
metric Data System," J. Air Poll. ice since the passage of the 1963 Clean
Control Assoc, 12: 423-430 (Septem- imum of five years, are as follows: Air Act, the U.S. Public Health Service
ber 1962). is now deemed to be an agency with en-
12. E. C. Tabor and C. C. Golden, "Re- forcement powers. Accordingly, Mr.
sults of Five Years' Operation of the Dr. E. R. Hendrickson, Chairman
National Gas Sampling Network," Director of Research Arthur C. Stern is now considered to
J. Air Poll. Control Assoc, 15: 7-11 309 Tigert Hall have had his employment category
(January 1965). University of Florida changed from that of the Research Ed-
13. Technical Report of California Stand- ucation and Consultant category to that
ards for Ambient Air Quality and Gainesville, Florida 32603
Motor Vehicle Exhaust, California R. E. Hatchard, Director of a control official. In accordance with
Department of Public Health, Berk- Air Quality Control Division the by-laws, he may serve until the next
eley (1960). election in his present category and he
14. S. M. Rogers and S. Edelman, 1964 Portland Bureau of Health
Supplement to A Digest of State Air 1225 Southwest Third Avenue may be placed in nomination for a two-
Pollution Laws: 4: Public Health year term as a control official.
Portland, Oregon 97204
Service Publication No. 711, Wash- Mr. John S. Lagarias has changed his
ington, D. C. (1964). Arthur J. Robert employment category from industry to
15. A. C. Stern, "Summary of Existing Chief Investigator and consultant and in accordance with the
Air Pollution Standards," J. Air Director of Air Pollution
Poll. Control Assoc, 14: 5-15 (Jan- By-laws he may serve until the next
uary 1964). City of Saint Paul election in his category as consultant
16. A. C. Stern, Air Pollution, II, Aca- 100 East 10th Street and he may be placed in nomination
demic Press, New York (1962). St. Paul, Minnesota 55101 for a one year term in the consultant
17. V. A. Ryazanoy, "Sensory Physiology Elmer P. Wheeler
as Basis for Air Quality Standards," category.
Arch. Environ. Health, 5: 480-494 Manager, Environmental Health
(November 1962). Medical Department
18. F. Pooler, Jr., "A Prediction Model Monsanto Company
of Mean Urban Pollution for Use with 800 North Lindbergh Boulevard
Standard Wind Roses," J. Air and
Water Poll, 4: 199-211 (1960). St. Louis, Missouri 63166 DREXEL ANNOUNCES
19. D. B. Turner, "Relationships Between John H. Easthagen FACULTY EXPANSION
24-Hour Mean Air Quality Measure- Staff Engineer
ments and Meteorological Factors in Drexel Institute of Technology has
Nashville, Tennessee," / . Air Poll. Chevron Research Company
Control Assoc, 11: 483-489 (October 567 Standard Avenue expanded the faculty in the Environ-
1961). Richmond, California 94801 mental Engineering and Science Gradu-
20. J. A. Maga and G. C. Hass, "The ate Program.
Development of Motor Vehicle Ex- George T. Minasian Dr. H. C. Wohlers, former Director
haust Emission Standards in Cali- Consulting Engineer of Technical Services, Bay Area Air
fornia," J. Air Poll. Control Assoc, 104 Clark Street
10: 393-396 (October 1960). Pollution Control District, joins the
21. R. I. Larsen, W. W. Stalker, and C. R. Glen Ridge, New Jersey 07028 staff as Associate Professor of Environ-
Claydon, "The Radial Distribution Jean J. Schueneman mental Science; Dr. Norman Trieff,
of Sulfur Dioxide Source Strength and Chief, Technical Assistance Branch Assistant Professor of chemistry, was
Concentration in Nashville," «/. Air Division of Air Pollution
Poll. Control Assoc, 11: 529-534 previously connected with the Michigan
(November 1961). Taf t Sanitary Engineering Center Department of Health; Dr. Robert
22. V. G. MacKenzie, "The Role of the 4676 Columbia Parkway Stiefel will be concerned with the engi-
Scientist and the Citizen—a Case Cincinnati, Ohio 45226 neering aspects of water resources
Study, Air Pollution," presented at
the Scientists' Institute for Public development and water qualitative
Information Symposium of the Annual The offices for which this committee management; Dr. Alonzo Lawrence
Meeting of the American Association is to select nominees are: comes to Drexel as a specialist in bio-
for the Advancement of Science, logical treatment methods as applied to
Montreal, Canada (December 29, (1) First Vice President, who must
1964). be a control official—to serve one year. waste disposal systems.

572 Journal of the Air Pollution Control Association

You might also like