DOC
EPA
United States
Department of
Commerce
National Oceanic and
Atmospheric Administration
Silver Spring MD 20910
United States
Environmental Protection
Agency
EPA-600/7-81-006
January 1981
Research and Development
Demonstration of a
Long Range Tracer
System Using
Perfluorocarbons
Interagency
Energy/Environment
R&D Program
Report
-------
DEMONSTRATION OF A LONG-RANGE ATMOSPHERIC
TRACER SYSTEM USING PERFLUOROCARBONS
FINAL REPORT
Gilbert J. Ferber
Kosta Telegadas
Jerome L. Heffter
National Oceanic and Atmospheric Administration
Air Resources Laboratories
Silver Spring, Maryland 20910
C. Ray Dickson
National Oceanic and Atmospheric Administration
Air Resources Laboratories Field Research Office
Idaho Falls, Idaho 83401
Russell N. Dietz
Environmental Chemistry Division
Brookhaven National Laboratory
Upton, New York 11973
Philip W. Krey
Environmental Measurements Laboratory
Department of Energy
New York, New York 10014
-------
NOTICE
Mention of a commercial company or product does not constitute
an endorsement by NOAA Environmental Research Laboratories or the
Environmental Protection Agency. Use for publicity or advertising
purposes of information from this publication concerning proprietary
products or the tests of such products is not authorized.
11
-------
Participants
H. Myers
Agronomy Research Station, Oklahoma State University
R. Dickson F. Mahoney
G. Ferber B. Olson
J. Heffter K. Telegadas
Air Resources Laboratories, NOAA
E. Cote R. Goodrich
R. Dietz
Brookhaven National Laboratory
I. Haskell R. Lagomarsino
P. Krey F. Wilson
Environmental Measurements Laboratory,, DOE
M. Alex M. Fowler
J. Banar J. Frank
S. Barr P. Guthals
J. Cappis R. Perrin
D. Curtis D. Rokop
W. Shields
Los Alamos Scientific Laboratory
C. Clark L. Showell
J. Lee G. Wardius
J. Weaver
National Severe Storms.Laboratory, NOAA
J. Loveless B. Spittler
T. Sinclair D. Whitman
National Weather Service Central Region, NOAA
E. Chapman T. Heimbigner
R. Hannigan R. Lee
Pacific Northwest Laboratory, Battelle
Col. Van Louven
Chief M/SGT. T. Greening, T/SGT. Milgrom
6th Air Weather Squadron, Tinker AFB, USAF
iii
-------
TABLE OF CONTENTS
Page
ABSTRACT 1
1. INTRODUCTION 1
2. PERFLUOROCARBON TRACER SYSTEM 2
3. 600-KM EXPERIMENT 6
4. 100-KM EXPERIMENT 29
5. EVALUATION OF PERFLUOROCARBON TRACER SYSTEM 39
6. SUMMARY 52
7. ACKNOWLEDGMENTS 53
8. REFERENCES 54
v
-------
LIST OF TABLES
No. Page
1. Comparative data on SF^ and perfluorocarbons 2
b
2. Tracer releases on July 8, 1980 7
3. Location of sampling sites at the 100 km arc 10
4. Sampling sites at the 600 km arc 13
5. Tinker AFB rawinsonde data for the July 8 experiment *-'
6. Aircraft wind observations at 1250 meters (MSL) along the
100 km arc 19
7. Aircraft wind observations at 1525 meters (MSL) along the
600 km arc i9
8. Tracer concentrations along the 100 km arc, July 8, 1980 24
9. Dual-Trap Sampler results at Site 20 (100 km arc), July 8, 1980... 26
10. Airborne whole-air sample concentrations ^'
11. Tracer concentrations along the 600 km arc 30
12. Tracer releases on July 11, 1980 37
13. Tinker AFB rawinsonde data for the July 11 experiment ^
14. Tracer concentrations along the 100 km arc, July 11, 1980 ^
15. Performance of BATS sampling-analysis system
16. Comparison of BATS sequential sampler with whole-air sampler
at the 100 km arc (July 8, 1980) 51
vi
-------
LIST OF FIGURES
No. Page
1. Automatic sequential sampler (BATS) ......... .... ............... 4
2. Location of the sequential air samplers (BATS) and aircraft
sampling path at 100 km from the tracer release site ........... 9
3. Location of sequential samplers (BATS), LASL samplers, and
aircraft sampling flight path at 600 km from the tracer re-
lease site. The locations of rawinsonde stations are also
shown ............................. . ..... . ........... ........... 12
4. Surface weather map for 1200 GMT, Tuesday, July 8, 1980 ........ 15
5. Surface weather map for 1200 GMT, Wednesday, July 9, 1980 ---- .. 16
6. Wind observations at 1250 meters (MSL) along the 100 km arc
aircraft sampling path ................. . . ...................... 1°
7. Wind observations at 1525 meters (MSL) along the 600 km air-
craft sampling path
8. Calculated transport layer trajectories to the 100 km arc for
the 3-hour tracer release on July 8 22
9. Comparison of the transport layer trajectory with the trajec-
tory in a layer 150 to 600 meters above terrain 23
10. Average 45-min. PMCH concentrations along the 100 km arc from
the July 8 experiment 25
11. Comparison of PMCH concentrations aloft with surface con-
centrations 28
12. Average 3-hour PMCH concentrations along the 600 km arc 3^
13. Average 3-hour PMCH concentrations along the 600 km arc for
the period July 9, 0800 GMT to July 11, 2000 GMT 36
14. Surface weather map for 1200 GMT,Friday, July 11, 1980 38
15. Calculated transport layer trajectories to the 100 km arc
for the 3-hour tracer release on July 11
41
16. Average 45-min PMCH concentrations along the 100 km arc from
the July 11 experiment
17. Comparison of PMCH and PDCH concentrations from the 100 km
BATS samples on July 8 ........................................ 46
vii
-------
LIST OF FIGURES (cont'd)
No. Page
18. Comparison of PMCH and PDCH concentrations from the 600 km
BATS samples 47
19. Comparison of PMCH and PDCH concentrations from the 100 km
BATS samples on July 11 48
20. Comparisons of tracer concentrations in whole-air samples
collected in the flight over the 100 km arc on July 8 50
viii
-------
DEMONSTRATION OF A LONG-RANGE ATMOSPHERIC TRACER
SYSTEM USING PERFLUOROCARBONS
FINAL REPORT
Abstract. Regional-scale tracer experiments are needed to validate
atmospheric dispersion aspects of air pollution models. The capa-
bility of a new system, using perfluorocarbon tracers (PFTs), for
long-range dispersion experiments at reasonable cost was demonstrated
in two experiments. Two PFTs (CyF;^ and CsFig) were released simul-
taneously with SFg and two heavy methanes.
The PFT system provides automatic sequential samplers and rapid,
inexpensive analyses down to 2 parts per lO^ of air. PFT concen-
trations were measured 600 km away, up to three days after release.
Performance of the PFT system was excellent and a consistent set of
tracer data was obtained.
1. INTRODUCTION
Atmospheric transport and dispersion models are being used extensively to
simulate the behavior of air pollutants and to estimate regional air concentra-
tions. Increased concern over regional and international aspects of air pollution
has created a need for reliable model calculations of concentrations as far as
1000 km from pollutant sources. Experimental verification of these calculations
is essential to establish the credibility of the models and 'environmental assess-
ments based on model simulations. ;
Attempts to verify model calculations with air quality data are complicated
by the presence of multiple sources and imprecise knowledge of emission amounts.
There is a need for nonreactive, nondepositing tracers that could be released at
precisely controlled rates and measured accurately at very low concentrations.
This would allow us to conduct tracer experiments which isolate atmospheric trans-
port and dispersion from other variables and provide data for verification of this
basic aspect of model calculations. Regional-scale experiments require tracers
that can be unambiguously identified and measured as far as 1000 km from the re-
lease point. Sulfur hexafluoride, SFg, has been used out to 100 km but its rela-
tively high and variable background concentration militates against its use to
much greater distances. Even at shorter distances, a tracer system is needed that
would provide automatic sequential sampling and rapid, inexpensive sample analysis.
A new atmospheric tracer system, using perfluorocarbons, has been developed to meet
this need.
The capabilities of the perfluorocarbon tracer (PFT) system were successfully
demonstrated in two long-range experiments described in this report. The experi-
ments were designed to provide a proof-test of the perfluorocarbon tracer release,
sampling, and analysis techniques and to demonstrate the feasibility of conducting
long-range atmospheric dispersion experiments at reasonable cost. Each experiment
involved simultaneous release of two PFT tracers along with SFg over a 3-hr period
with concentrations measured 100 km downwind. In the primary experiment, two
heavy methanes, new tracers being developed at the Los Alamos Scientific Labora-
tory (LASL) were also released and the perfluorocarbons and methanes were measured
at a distance of 600 km as well as 100 km. Intercomparison of the PFT, SFg, and
heavy methane results has established the validity of the new tracer systems.
-------
The perfluorocarbon tracer data on the 600 km sampling arc present an inter-
esting case of very fast transport by a night-time low-level jet and the reappear-
ance of tracer over the arc on the day following its first arrival. Tracer con-
centrations were still measurable three days after release. This experiment
provides a useful case study for verification of long-range transport and disper-
sion models.
2. PERFLUOROCARBON TRACER SYSTEM
Investigations by Lovelock (1974) indicated that a perfluorocarbon tracer
system could be developed that would be ideal for long-range dispersion studies.
The NOAA Air Resources Laboratories (ARL) contracted with Lovelock to develop
three different samplers as the first step in the development of the new tracer
system. Prototype instruments were delivered by Lovelock in 1976. Since then
ARL has been working closely with the Department of Energy's Environmental
Measurements Laboratory (EML) and Brookhaven National Laboratory (BNL) in a co-
operative effort to develop a practical perfluorocarbon system.
The perfluorocarbons are extremely stable non-toxic compounds, measurable at
very low concentrations by gas chromatography and electron-capture detection. At
present, we are working with two perfluorocarbons, perfluoromonomethylcyclohexane
(PMCH; CyF]^) and perfluorodime thyIcyclohexane (PDCH; C£F16). Comparative data
on SF5, PMCH and PDCH are shown in Table 1. The atmospheric background concen-
tration of PDCH is about 0.026 parts per trillion by volume (26x10-15)., about 1/25
of the SFg background. Background of PMCH is an order of magnitude lower than
PDCH. The amount of tracer released in any experiment must be sufficient to dis-
tinguish the plume from background at the maximum sampling distance. The required
release rate (by weight) for PDCH is about 10% that for SF6; for PMCH it is about
1% of the SFfc rate. Taking the higher price of the perfluorocarbons into account,
the PDCH required for an experiment would cost about 20% more than SF6; the cost
of PMCH would be about 10% of the SF6 cost.
Another factor in favor of the perfluorocarbons over SFg is their very uni-
form background concentration. SFg has a highly variable background because of
many local sources throughout the country and the world.
Table 1. Comparative Data on SF, and Perfluorocarbons.
Tracer
Formula
Mol. Wt.
Background (pptv)
Cost/kg
Relative Release
Rate (by wt.)
Relative Cost/
Release
Sulfur-
Hexa-
f luoride
SF6
146
0.6
$11
100
1.0
Perf luoro-
Dimethyl-
cyclohexane
(PDCH)
C8F16
400
0.026
$110.
12
1.2
Perf luoro-
Monomethyl-
cyclohexane
(PMCH)
C7F14
350
/ 0.0024
$110.
1.0
0.1
-------
2.1 Tracer Release Mechanisms
The two perfluorocarbon tracers, which are liquids at ordinary temperatures,
were released as aerosol sprays. Each tracer is held in a 210-liter tank on a
trailer. Compressed nitrogen provides pressure to force the liquid out of the
tracer tank.
The mechanics of the spray system are simple. The spray nozzle has two hoses,
one from the tracer tank, and the other from a construction-type air compressor
that delivers 100 psi at 100 cfm. The tracer is introduced into the fast-moving
air stream, atomized through a small orifice, and released into the atmosphere.
Tracer release rate is monitored with a calibrated rotometer.
A newly designed release system, which was not completed in time for these
experiments, has since been tested and performed well in the DOE Atmospheric
Studies in Complex Terrain (ASCOT) experiments in California in September 1980.
This system, also trailer-mounted and designed to be completely self-contained
(no air compressor required), vaporizes the tracer before release.
The tracer is mixed with a stream of N2 gas to evaporate it and to carry the
tracer through the system. This mixture of nitrogen and perfluorocarbon gas flows
through a tube furnace. Temperature of the tube furnace is kept above the boiling
point of PDCH, 105°C, to assure that the tracer is completely vaporized. From the
tube furnace the mixture of N£ and tracer gas passes through a mass flow meter
where the volume is accurately metered. From there the tracer is released to the
atmosphere.
The design of this system provides back-up measurements of the actual amount
of released tracer. The mass flowmeter provides both instantaneous and total
volumes, and also supplies a 0-5 volt dc output which is connected to a stripchart
recorder. The recorded release rate shows the constancy of tracer release and pro-
vides a measurement of total output over the time of release. The system also has
a large set of crane scales (0-450 kg) and a small balance (0-40 kg) to provide
accurate weighings of the tracer tanks before and after release.
Both release systems were designed and built by the NOAA Air Resources Labora-
tories Field Research Office in Idaho Falls, Idaho.
2.2 Automatic Sequential Sampler
Based on Lovelock's prototype, R. Dietz at BNLj developed an improved sequen-
tial sampler dubbed the Brookhaven Atmospheric Tracer Sampler (BATS). The sam-
pler consists of an Air Flow Module (lid) and a Power Control Module (base). The
entire unit, shown in Figure 1, measures 36x25x20 cm and weighs 7 kg. The lid
contains 23 sampling tubes filled with 150 mg of 20-50 mesh-type 347 Ambersorb*
which traps all the perfluorocarbons in the air flowing through the tube. The
base contains a constant volume pump which draws air through each sampling tube
in a sequence controlled by an internal digital clock. Flow rates, controlled by
critical orifices, are selectable from 2 to 50 cc/min. The base also contains a
digital printer that records the tube number, start time and number of pump
strokes (which can be converted to air volume) for each sample. Controls in the
base provide for automatic start at a preselected day and time for a preselected
*Trade name of Rohm and Hass Company.
-------
Figure 1. Automatic sequential sampler (BATS),
4
-------
number of samples and duration of sampling (1 min to 1 week per tube), as well as
for automatic analysis with a gas chromatograph. Internal rechargeable batteries
provide sufficient power for unattended operation for up to a month. After 23
samples have been collected, the lid unit can be removed for sample analysis (in
the laboratory) and a fresh lid attached in its place to continue the sampling
program.
The Air Resources Laboratories contracted with Gilian Instrument Corp. for
final design and production of 60 complete BATS samplers which were delivered in
May 1980 for use in the July experiments. An operations manual was also prepared
by Gilian (1980).
2.3 Sample Analysis System
The determination of perfluorocarbon tracer concentrations from the BATS sam-
ples is accomplished with an analysis apparatus designed, built and operated at
BNL. The tracer is recovered by thermal desorption from the BATS tubes with sub-
sequent gas chromatographic separation prior to electron capture detection. The
scheme also includes chemical processing of the recovered constituents in order
to destroy and remove interfering components, such as chlorofluorocarbons, which
are present in the air at concentrations order of magnitude higher than that of the
PFTs.
Before the sample is thermally desorbed, the BATS tube is purged with carrier
gas (5% H£ in N£) for a short period of time to remove any traces of oxygen which
otherwise would react with the PFTs during the 400°C desorption recovery. Desorp-
tion is accomplished by direct ohmic heating of the thin stainless steel wall of
the BATS tube. The sample is purged from the BATS tube through a Pd catalyst bed at
260°C and then through a 120 cm Porasil F pre-cut column. The 10-cm long catalyst
bed reduces any chlorofluorocarbon compounds, as well as any remaining oxygen, to
their hydrogenated form, thus rendering these interfering constituents non-
electron-capturing. After the surviving PFTs elute from the pre-cut column,
heavier molecular weight constituents,still within the column, are purged to the
atmosphere by reversing the direction of flow. Meanwhile, the eluted PFTs are re-
concentrated within a 10-cm long bed of Porapak QS adsorbent. The purpose of the
bed is two-fold. First, only the PFTs are retained in the Porapak QS; any lighter
constituents which might ultimately interfere are flushed away. Secondly, once
the Porapak QS-trapped PFTs are released into the main analytical column, the next
BATS tube recovery cycle can be initiated, thus halving the overall PFT recovery
and analysis time by overlapping the stages.
When the Porapak QS trap has been heated to 200°C, the PFTs are released into
a second catalyst bed (2.5 cm long) for a final clean-up and flushed through a
Nafion permeation dryer to remove traces of moisture before entering the main
column, 6 meters of Porasil F, which is at the same temperature as the pre-cut
column, 90°C. The 22 mL/min flow of carrier gas at .this column temperature pro-
vides good resolution of the two PFTs. Automation is accomplished by interfacing
the timing capability of the BATS with the INJECT command of a Varian CDS-111
integrator-controller, which provides the control capability for the involved
valving and heating sequences within a Varian 3700 series gas chromatograph.
Analyses of the 23 tubes on a BATS unit can be completed in just under 3 hours.
The present system incorporates a °-%i electron-capture detector which pro-
vides a measurement accuracy within ±10% at concentrations as low as 2 parts per
-------
10^ (approximate ambient concentration of PMCH) with a sampled volume of 8 liters
of air. This is the approximate volume collected in the 600 km arc samples (3-hr
duration). The uncertainty in measurements near the PMCH background level is some-
what greater (about ±25%) on the 100 km arc where the volume sampled was about 2
liters (45-minute duration).
2.4 Dual-Trap Sampler
Another prototype instrument, the Dual-Trap sampler, was designed by Lovelock
to combine the sampling and analysis functions into a single unit. The unit con-
tains two sampling tubes which are automatically cycled so that one tube samples
while the other is being analyzed. This instrument provided readout of PDCH tracer
concentrations (no PMCH) every five minutes at the sampling site.
The original prototype has been modified at EML and BNL to provide a more
rugged instrument for field use, to collect and measure PMCH and PDCH simultan-
eously, and to improve its detection limit by more than two orders of magnitude.
Ambient PDCH (about .03 ppt) can be measured with ±15% precision and PMCH can
be measured at concentrations slightly above its ambient level of about .003 ppt.
The attainment of this degree of sensitivity in a real-time field instrument is a
major advance which will add significantly to long-range tracer capability.
2.5 Continuous Tracer Monitor
The third prototype sampler developed by Lovelock is a real-time continuous
monitor intended primarily for use in aircraft sampling. Ambient air is drawn
through a catalytic reactor that reduces the 02 and other electron-absorbers,
leaving the perfluorocarbons and nitrogen. This is passed directly to an electron-
capture detector providing continuous concentration readout with only a 3-second
delay.
Many problems have been encountered in the operation of this instrument, but
the concept appears to be sound and work is continuing on the development of this
sampler. If successful, it should be able to provide a continuous in-flight re-
cord of tracer concentrations down to 0.1 ppt or better.
3. 600-KM TRACER EXPERIMENT
A long-range tracer experiment was conducted on July 8, 1980 with the simul-
taneous release of two perfluorocarbons, SF6, and to heavy methane tracers at the
NOAA National Severe Storms Laboratory (NSSL) at Norman, Oklahoma. Samplers were
deployed to measure tracer concentrations along arcs at 100 km and 600 km north of
the release point. The objectives of the experiment were:
1) to provide a proof-test of the perfluorocarbon release, sampling
and analysis techniques,
2) to test the concept of using the National Weather Service sub-
station network for cross-country sampling,
3) to compare measurements of five different tracers to establish
the validity of the new tracer techniques, and
-------
4) to demonstrate the capability to perform long-range atmospheric
transport and dispersion experiments, at reasonable cost, for
verification and improvement of air pollution models.
3.1 Tracer Release
The five tracers were released simultaneously over a 3-hr period from 1900 to
2200 GMT (1400-1700 CDT) from an open field at NSSL. Release nozzles were about a
meter above ground level. Flowrates were carefully monitored to assure a nearly
constant release rate for each tracer. Release amounts are shown in Table 2. The
amounts of perfluorocarbon and heavy methane released were calculated to produce
concentrations well above the detection limit at the 600 km sampling arc. The
amount of SFg released was sufficient to be detected at the 100 km arc for compari-
son with the new tracers.
Table 2. Tracer Releases on July 8, 1980
Tracer
PMCH
PDCH
SULFUR
HEXAFLUORIDE
METHANE-20
METHANE-21
Formula
C7F14
C8F16
SF,
12
CD,
13
CD,
Molecular
Wt.
350
400
146
20
21
Release Amount
(kg)
192
186
273
0.153
0.084
Tracer Release Ratios
(by Volume)
PMCH/PDCH
SF,/PMCH
D
SF./PDCH
o
PMCH/Me-20
PMCH/Me-21
SF6/Me-21
1.18
3.41
4.02
72
137
467
It should be noted that although very small amounts of heavy methanes are
required, they are relatively expensive to produce. When the costs of tracer
materials and sample analysis are taken into account, the cost per experiment is
comparable for perfluorocarbons and heavy methanes.
-------
The two perfluorocarbons (PMCH and PDCH) were released as aerosol sprays from
separate tanks mounted on trailers a few feet apart. The tanks were weighed im-
mediately before and after the experiment to determine the amount released from
each tank. Since the commercially available PDCH contains about 8% (by weight)
PMCH and the commercial PMCH has about 2% impurities, samples of the purchased
tracers were assayed at BNL prior to the experiment, and samples from the release
tanks were assayed after the second experiment. The release tank weighings and
the chemical assays were used to calculate the PMCH and PDCH release amounts shown
in Table 2. These values are accurate within ±4%.
SFg was released as a gas from pressurized cylinders positioned between the
perfluorocarbon trailers. The release amount was determined by weighing the
cylinders before and after release and is accurate within ±2%.
The two heavy methane tracers were released as a calibrated mixture of gases
from a single pressurized cylinder. The mixture was prepared at LASL and the
ratio of the two methanes was determined by mass spectrometry. The total amount
released was determined by weighings before and after release. Release amounts are
accurate within ±1%.
The lower part of Table 2 gives the tracer release ratios, by volume, as cal-
culated from the release amounts and molecular weights shown above. Ideally, if
the tracer systems worked perfectly, these same ratios should be found in all air
samples collected within the tracer plume (after ambient background concentrations
are removed).
3.2 Sampling Array
Sampling arcs were established at 100 km and 600 km from the release point.
Sites were selected in a sector to the north of the release site^ based on a 5-
year climatology of July trajectories.
3.2.1 100 km arc
Thirty sampling sites were selected at 4-5 km intervals along the roadway of
HWY 51 and HWY 33 as shown in Figure 2. The latitude-longitude azimuth, and dis-
tance from the release site of each sampling site are listed in Table 3. The
operational center for the 100 km arc was set up at the Agronomy Research Station,
Oklahoma State University at Stillwater, OK. National Weather Service instrument
shelters were set up at each location to house the BATS sequential sampler. Only
seventeen samplers were available, so the sites to be instrumented had to be se-
lected just prior to the start of the tracer release. Based on the latest trajec-
tory forecast, two EML sampling teams deployed the BATS samplers to Sites 12-28.
The tracer release began at 1900 GMT (1400 CDT) and the samplers were set to take
ten 45-minute samples starting at 2130 GMT, before the tracer was expected to
arrive.
A whole-air sampler (pump and plastic bag enclosed in a barrel) was co-
located with each BATS sampler to collect a single sample starting when the BATS
was placed at the site and ending when the 'BATS sampling was terminated. The
purpose of the whole air samplers was to provide intercomparisons among the five
tracers and aliquots were taken from each bag for perfluorocarbon, heavy methane,
and SFg analyses.
-------
36°
35.5°
35'
22 i 23 24 2526N ^7 28 29/30
N
•+• + + .Surface Sampling Sites
® ® Aircraft Flight Path
Norman
10 0
111111111
10
20
I
30 40
50
I
Kilometers
98°
97.5U
97°
96.5°
Figure 2. Location of the sequential air samplers (BATS) and aircraft
sampling path at 100 1
-------
Table 3. Location of sampling sites at the 100 Km arc,
Station
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Tinker AFB
Release Site
KTVY Tower
Latitude
°N
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.12
36.11
36.10
36.12
36.12
36.12
35.99
35.98
35.99
35.99
35.98
35.96
35.97
35.97
35.97
35.42
35.24
35.58
Longitude
°W
98.10
98.05
98.00
97.94
97.89
97.84
97.79
97.73
97.68
97.63
97.59
97.54
97.48
97.42
97.36
97.31
97.26
97.21
97.15
97.09
97.05
97.05
97.00
96.94
96.89
96.84
96.77
96.72
96.66
96.59
97.38
97.46
97.48
(a)
Distance
Km
115
113
111
108
106
105
103
102
100
99
98
98
97
97
98
98
98
98
101
103
104
91
93
95
97
100
104
108
110
114
Azimuth (a)
deg
328
330
333
335
338
340
342
345
347
350
352
355
358
001
003
006
008
Oil
014
017
019
022
025
028
030
033
036
038
040
043
10
-------
3.2.2 600 km arc
Sampling sites on the 600 km arc, in Nebraska and Missouri are shown in
Figure 3. Deployment and operation of samplers over the long distances on this
arc could have presented difficult logistic problems. Fortunately, we were able
to secure the cooperation of the NOAA National Weather Service (NWS) to allow us
to use their substation network as a fixed sampling array. This network is com-
prised of over 12,000 locations in the U.S. where cooperative observers, mostly
volunteers, gather weather data for the NWS.
The BATS samplers were delivered, in advance of the experiment, by NWS sub-
station specialists to the sites shown i,n Figure 3. At the time of delivery, the
samplers were set to take 22 three-hour samples. On the day of the experiment,
after the tracer release had begun, all observers were notified by telephone to
set the samplers to start automatically at 0800 GMT (0300 CDT) on July 9. The
station locations and cooperative observers are listed in Table 4.
The Los Alamos Scientific Laboratory had 6 cryogenic samplers available for
deployment on the 600 km arc for the collection of heavy methanes. On the evening
of July 8, based on the latest wind data and forecasts, they were advised by ARL
to deploy the samplers to the site indicated by double circles in Figure 3. Five
sequential samples were taken at these locations at 3-hour intervals beginning at
1100 GMT (0600 CDT) on July 9.
3.3 Airborne Sampling
The Battelle Pacific Northwest Laboratory provided a DC-3 aircraft and crew
for sampling missions over the 100 km and 600 km arcs. It was intended to obtain
plume profiles aloft with the Lovelock real-time continuous perfluorocarbon monitor
and a modified version of this instrument developed at BNL. However, neither in-
strument was operational on the day of the experiment. Whole-air samples were col-
lected in plastic bags, and analyzed for all five tracers. Frequent wind measure-
ments were also made aboard the aircraft during both sampling flights.
Three sampling passes were made at the 100 km arc along the flight path shown
in Figure 2 at an altitude of 900 meters above the ground (1250 m MSL) between
2300 GMT and 0000 GMT (6-7 PM). On each pass, a plastic bag was filled with out-
side air along each segment of the flight path.
The aircraft returned to Wiley Post Field in Oklahoma City, refueled, and then
took off for Kansas City in preparation for the 600 km sampling flight the next
morning. The plume had been forecast to arrive about 1300 GMT (8 AM) but the 0600
GMT wind data indicated faster plume travel and the aircraft was rescheduled for
take-off at 1230 GMT (7:30 AM) and a sampling flight path north of the 600 km arc,
shown in Figure 3 was chosen to compensate for the stronger winds. Bag samples
of about 12-minute duration were collected along each segment of the flight path
from about 1240 to 1630 GMT at an altitude of 1200 meters above the ground (1525
meters MSL). Aliquots were transferred from each bag for later analysis by BNL
and LASL.
3.4 Meteorology
On July 8-9 a broad area of high pressure dominated most of the U.S. A
west-to-east oriented stationary front just north of the 600 km sampling arc was
11
-------
100 W
43° N
40'
90°
Figure 3. Location of sequential samplers (BATS) 3 LASL samplers, and
aircraft sampling flight path at 600 km from the tracer release
site. The locations of rawinsonde stations are also shown.
12
-------
Table 4. Sampling sites at the 600 Km arc
Station
No.
NEBRASKA
A
1
2
3
4
5
6
7
9
10
11
12
13
MISSOURI
Location
Hastings
Clay Center 5W
Bradshaw
Fairmont
Friend
Western
Crete
Lincoln (WSO)
Firth
Sterling
Tecumseh
Table Rock 4N
Auburn 5NNE
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Fairfax
Skidmore
Maryville 2E
Conception
King City
Pattonsburg
Hamilton 2 W
Chillicothe
Coloma
Carrollton
Brunswick
Marshall
New Franklin
Boonville
Columbia (WSO)
Jefferson City
Freedom
Vienna
Vichy (FAA)
Rolla
Cook Station
Salem
Bunker
Ellington
Van Buren
Observer
Ralph A. Powell
Jim Chapman
Jack Pugh
Andrew Anderson
Jim Hannon
Kenneth Roesler
Dr. Delbert King
Orval Jurgena
Roland Beach
Raymond Zink
Arthur Lempke
Betty Vrtiska
Daryl Obermeyer
Dillard Price
Donald Brown
George Wolfe
Br. Datnian Larson
John Martin
Mrs. Kenneth Mason
William Kuhnert
Sam Bowling
Mrs. Freda Trussel
Harold Finley
John M. Smith
Steve Hilton
Mrs. Ronda Thiessen
Rolland Goode
Dave. Horner
Robert Block
Mrs. Velma Niewald
Henry Kaiser
Newton Lipplitt
Dr. Al Spreng
Mrs. Ozella Brand
Warren Sellers
Mrs. Grace Shaffer
Billy Swyres
Gerry Whittle
Latitude
°N
40.60
40.53
40.88
40.63
40.65
40.40
40.62
40.85
40.53
40.47
40.37
40.23
40.45
40.33
40.28
40.35
40.25
40.05
40.05
39.75
39.77
39.53
30.37
39.42
39.12
39.00
38.97
38.65
38.58
38.47
38.20
38.12
37.95
37.82
37.63
37.45
37.20
36.98
Longitude
°W
98.35
98.15
97.75
97,58
97.28
97.20
96.95
97.75
96.60
96.38
96.18
96.08
95.80
95.40
95.08
94.83
94.68
94.52
94.13
94.03
93.55
93.53
93.50
93.12
93.18
92.77
92.75
92.22
92.15
91.70
91.98
91.77
91.77
91.43
91.53
91.22
90.93
91.02
Azimuth
deg.
352
354
358
000
001
002
004
005
007
008
010
Oil
012
015
018
020
022
023
026
028
031
033
034
036
038
040
041
046
048
052
053
055
057
060
061
064
066
068
(a) Measured from Release Site.
13
-------
associated with a weak low pressure center moving slowly eastward (see Figures 4
and 5). The wind flow in the boundary layer (surface to about 2500 m) over the
central U.S. was predominantly from the south-southwest around a strongly persis-
tent high pressure system centered in the southeastern U.S. This weather pattern
was associated with the severe "heat wave" in the central U.S. during July 1980.
Afternoon surface temperatures in the experimental area generally rose above 38°C
(100°F) during the entire period of the experiment.
3.4.1 Forecast tracer trajectories
In order to alert the sites in advance to prepare for sampling, forecast
trajectories were prepared on a daily basis. Trajectories starting at 6-hour
intervals were determined from a computer program using the NOAA National Meteoro-
logical Center forecast gridded wind fields. The forecast obtained the morning
of July 8, based on 0000 GMT data, was for trajectories starting 18 to 24 hours
later (for a planned tracer release time of 1900 GMT). The plume centerline was
forecast to move to the northeast across the eastern part of the 100 km arc and
then continue northeast-to-north crossing near the center of the 600 km arc.
Based on the forecast, preparations continued for a 1900 GMT release. The last
forecast before release was obtained at noon (based on 1200 GMT data) for a tra-
jectory starting at 1800 GMT. The plume centerline was forecast to be in about
the same position as before with a slight northeast shift at the 600 km arc. The
tracer was released with the knowledge that backing (counter-clockwise shifting)
of the local winds was forecast during the afternoon turbulent mixing. This in-
sured that the plume would cross the 100 km arc shifting from east to west as time
progressed.
3.4.2 Special rax^insonde observations
Special rawinsonde observations up to 500 mb were taken at Tinker AFB, about
20 km north-northeast of the release site, starting the morning of July 8. Data
are given in Table 5 for observations from July 8, 1700 GMT to July 9, 0300 GMT.
The data include height, wind direction, wind speed, and temperature. In addition,
a transport layer height was computed from each temperature sounding (Heffter,
1980). The transport layer height, TLH, and average wind speed and direction in
the layer,are given at the bottom of the table for each sounding.
Special upper-air observations were also requested from several stations in
the regular National Weather Service rawinsonde network. Extra soundings were
taken at Omaha, NE; Topeka, KS; and Monett, MO, on July 8, 1800 GMT, July 9, 0600
GMT and 1800 GMT. All special rawinsonde data collected for this experiment, in-
cluding the Tinker AFB soundings, have been included in the NAMER-WINDTEMP data
tapes available at the National Climatic Center, Asheville, NC (see Appendix C,
Heffter, 1980).
3.4.3 Aircraft winds
The PNL sampling aircraft took wind observations at the 100 km arc along the
flight path shown in the lower part of Figure 6. The upper figure shows a plot
of the winds by longitude (along the flight path) versus time. To locate the
geographic position of any wind, read directly down (along a constant longitude)
from the plotted wind position of the upper figure to the intersection along the
flight path in the lower figure. The winds are tabulated in Table 6.
14
-------
'0861 f8
OOST
dviu
> rtu / /«o N »Sr Nv^___--"'
^lilt ' -^^ /o^£l*
* ;*^ -".^"v '«*
K/^&- .as/
^$t°--»
086t'« AIM '
-------
'&&:l.^r%'«^ i?H *
(•V', ^ ^i \'$»<% * tl" ^**} *4*5" !O»?' "**
- TO '*r
00
Oi
Si
I
CO
ca
CiJ
Ci
C5
00
^~l
SH
16
-------
LI
3 30
voro(MiN>cntM3X .pa»rocMv>o-j.p>->CDcnrovooNCM3x
or* oroMU»-«-gr~ ooootnomouiot-»ocno-jf c
rooo n
voenaj ocovooovo co omcnuicncncnai(jiooouiooc)xC o
o
xco
<"> x-o °n 2
3 O -«
*. a> »
\0-PIUtM009CnCM3X -JO»IN}CW\>-J.pl-»09Ull\>vOCINCM3X CD
Cf^ "Wl"* fO O 0(>CIIODCO*H vOOB"*JON|«J^-POIOICItlN5H*Hi O)CO*^ f f~
or~ OOM \o-poo-jf~ cnooo«-*cnocnooooio^jf~ c m
•o r
(-*ao uicr>-^^o>a»cpvO3?i iN>rv>rv»i-»»->roroiN>rorv>roi-»H»i->oo en
ooo^ot-'cnciNGom K*CMi-*vO|>ooi-'H*i->ooo>^->jmi-i CD •
3 O MMMM 3 t-» IH
xco wroi\>i\)i\>ro!v) CMI^H cfNcnociNciii^vocooNQN(jiONcnwx co vo ^
n ??TJ n -i x
O CD
o
MOO
a>o3J O^IOCMJIOON ex) cnoouiuicncncnocncnooooocn^} o o
CD CO I CMvOCDvOODlNJCDCO-l
or-
3 O t-»M H»H>t-» !-»»-« 3 IV) O
VOCOTJ -si->jcovo\ovoocnm •••••••••••••••• CO*D en
n XTJ o H >
3 3 3 O T|
ro- -4 cnwroroi-» •> ui-P-PCMCMCMiv>roiv>i->i->i->i-' •* vo o
-p\oo>cnc»i)->o>co-H voce-JONUivo-P-PtriCMCMroroi-'i-'coco-i c
T3 r- i
H»OO cn-J-»J-JCt>vOvO3;o fN)i\>rofoiN>roiN>H>i-»i-»t-»H»MMH'»-»oo m
o>m>-4 oo-PcocnoicT'CDn <3>0"»ocMi\>roi-*\oo30&-'4-j—j^j-jCDni—i CD
0033 oowMorooN co oocnoooacncnouicnoaicnocn^] o c_
c
3 O MI-»l-»MI-*l-»l-»t-» 3 O P"
OCO"D -^o>oo\ovooocnm ................ COTD en
o XTJ n -i
m
x
TJ
3 3 o m
cnwcMwroroM - cn-p-PO4O»(N3i\jfV)Mi-»H» -. \o 3J
-po>rooNM~j^M09cnrv>vO^M3X M
c_ 3
c m
r z
oiroi\>iv>rororoi\>roi->MH>Mi->oo H
oo»oON co cncncfioooocnooocnoocn>o o
Q f->|->|_>|_>|_lf^^ 3 O
rororofv>ropofv>CMwcwm-« (N>i-*pi\>-pvOoMi->rocMcn(Mv>xco CM
COTJ en
3
O H
-------
Latitude (°IM)
Time (GMT)
00
ft K
o
g
CO
S-i O
^. C^
S Co
«Q (S
'Xj ^
a a
o
s
Co
a
to
Ol
Ci
Co
ft
r-i
O
Ǥ
cf
00
DO
(D
10
in
o
^
10
Z+
a.
to
-------
Table 6. Aircraft wind observations at 1250 meters (MSL)
along the 100 km arc.
Time
(GMT)
2304
2315
2326
2332
2344
2347
2353
0000
0008
0009
Direction
(deg.)
186
205
182
181
178
176
194
205
188
198
Speed
(m/sec)
6.7
2.6
13.9
19.6
17.5
6.2
3.1
7.2
6.7
5.7
The sampling aircraft also took wind observations at the 600 km arc along the
flight path shown in Figure 7 (plotted similar to Figure 6). These winds are also
tabulated in Table 7.
Table 7. Aircraft wind observations at 1525 meters (MSL)
along the 600 km arc.
Time
(GMT)
1304
1313
1316
1327
1328
1340
1346
1352
1358
1404
1416
1427
1440
1449
1452
Direction
(deg.)
267
271
271
272
272
327
332
318
312
309
279
254
277
267
255
Speed
(m/sec)
20.1
19.0
19.0
19.0
19.0
10.3
13.4
12.9
12.4
15.4
11.8
18.0
12.4
13.4
17.0
Time
(GMT)
1504
1511
1516
1526
1528
1533
1540
1544
1550
1554
1556
1604
1607
1617
1622
Direction
(deg.)
265
263
262
265
269
265
259
261
256
262
263
255
258
260
244
Speed
(m/sec)
13.9
16.0
17.0
14.
13.
12.9
10.3
8.8
12.4
14.4
15.4
18.5
16.0
8.8
8.8
19
-------
Time (GMT)
N>
O
ft
CO
ft,
o
tr1
Co
^Q ft
t3 ^.
ft O
«- s
5r> Co
On
to
01
I
ft
^-~J
o
CO
Latitude (°N)
Ci
-------
3.4.3 Post-facto tracer trajectories
Tracer trajectories to the 100 km arc were hand-calculated using average winds
in the computed transport layer as determined from the Tinker AFB soundings. Tra-
jectories for the start and end of the tracer release are shown in Figure 8 with
times (GMT) indicated along each trajectory. Also shown is the expected plume
width. The calculated plume position and arrival time at the 100 km arc agreed
well with the tracer data although the actual plume extended further to the west
(see Section 3.5.1).
Tracer trajectories to the 600 km arc were computed using the ARL-ATAD model
(Heffter, 1980). Meteorological input data were obtained from the NAMER-WINDTEMP
data base. The computed trajectories are shown in Figure 9. The solid trajectory
is determined from winds averaged in a computed variable transport layer; the
dashed trajectory is from winds averaged in a constant layer 150 to 600 m above
terrain. The calculated plume centerline at the 600 km arc using the variable
transport layer was about 200 km east of the measured peak concentration; the cal-
culated position using the 150-600 m layer was in better agreement, about 100 km
east of the actual position.
3.5 Sampling Results
3.5.1 100 km sampling results
The BATS sequential samplers and whole air samplers were installed at Sites
12 through 28. The 45-minute sequential sample concentrations are given in Table
8. Due to analysis problems, no data are available for Site 17.
The PMCH sampling results on the 100 km arc are shown graphically in Figure
10. The sampling sites are plotted as a function of the azimuth from the release
site. The scale gives the distance in kilometers between sampling sites projected
onto the 100 km arc.
During the initial sampling period (2100-2145 GMT), the PMCH concentrations
at all sampling sites are at or slightly above the background concentration of
about 2.4 parts per 1015. During the second sampling period (2145 to 2230 GMT),
approximately 3 hours after the start of tracer release, concentrations had in-
creased by three orders of magnitude with the plume centered between Sites 12 and
16. The backing of the winds with time carried the tracer plume further west than
expected and the portion of the plume west of Site 12 was not sampled.
The next samples (2230 to 2315 GMT) show the peak PMCH concentrations. Later
samples show decreasing concentration with the plume centerline shifting toward
the west. As will be seen later, aircraft sampling data indicated that plume con-
centrations west of Site 12 probably decreased very rapidly.
During the sampling period 0130-0215 GMT (July 9), about 4 hours after the
end of the release, concentrations along the 100 km arc had returned to near-
background levels. Sites 23 through 28 had background concentrations during the
entire sampling period.
The Dual-Trap sampler, described in Section 2.4, was operated along the 100 km
arc but the only non-background data obtained was at Site 20 from 2227 to 2314 GMT
as shown in Table 9. The average PDCH concentration for this period was 228 parts
21
-------
zz
Latitude (°N)
in
T
w
C>
o
V)
a -1
£2 3 g
•*
(O
09
b
-------
42
41
40
39
0)
•o
38
37
36
35
1800
A
O
600 KM
Sampling
Sites
Layer 150 to 600 M
Above Terrain
Computed
Transport
Layer
1800
99
98
97
96 95
Longitude (°W)
94
30/
93
92
Figure 9. Comparison of the transport layer trajectory with the trajec-
tory in a layer 150 to 600 meters above terrain.
23
-------
TABLE 8
100 KM AKC
TRACER CONCENTRATIONS (PARTS PER 10 )
STATION
START
TIME(GMT)
JULY 08
2100
21*3
2230
2315
JULY 09
0000
0045
0130
0215
0300
0345
12
PHCH PDCH
650.
4000.
2170.
43?
4!&
4.1
3.6
3.1
25
580
2980
2160
1700
67
32
30
26
26
13
PMCH** PDCH**
4.
1300.
5900.
2700.
500.
4,
4.
4,
28
890
3900
2000
390
52
32
*1
31
26
14
PMCH PDCH
4,7 26
1010, 920
4670, 3500
1650. 1370
1.82.
4.5
5.0
4.4
4.6
4.4
28
28
27
27
28
15
PMCH PDCH
4.9
860.
2730.
1260.
ee.
4.0
4.8
5.*
5,*
10,*
25
760
2380
1110
96
25
23
2fi
26
25
STATION
16
PMCH PDCH
START
TIME (GMT)
JULY 08
2100 3.0
2145 1110.
2230 2810.
2315 1000.
JULY 09
0000 90.
0045 9.*
0130 3.*
0215 4.*
0300 f,,*
0745 10.*
26
980
2440
920
101
28
27
27
27
28
18
PMCH PDCH
3,0
290.
2100.
340.
3.*
3,*
3.*
4.*
3.*
31
2'o
1780
310
23
23
i1*
26
2«f
19
PMCH PDCH
3.0
130,
560,
50,
3.*
It*
3.*
3,*
'*
27
150
460
66
25
26
26
26
27
26
20
PMCH PDCH
3,7
16.
215,
3.8
4.5
4,1
4.2
4.2
4.1
26
40
238
41
30
29
31
31
31
31
STATION
START
TIME(SMT)
JULY 08
2205
2250
2335
JULY 09
0020
0105
0150
0235
0320
0405
04SO
21
PMCH PDCH
27.1
3.6
2.8
2.7
3.1
9.9
2.8
3.4
2.8
2.5
44
24
23
25
25
24
25
25
25
25
START
TAME(GMT)
JULY 06
2130
2?15
2300
2345
JULY 09
0030
0115
0200
0245
0330
0315
22
PMCH PDCH
5,7
4,1
3,4
3,8
3.7
5,5
3,5
4,2
4.3
3.9
37
24
25
27
26
27
25
29
26
27
23-28
PMCH PDCH
SEE
FOOT
NOTE
A
- NO DATA
* VALUE UNCERTAIN DUE PRIMARILY TO CONTAMINATION IN LAB ANALYZER.
** POOR DESORPTION POWER.CORRECTION ESTIMATED,
A SAMPLING SITES 23-28 HAD BACKGROUND PMCH AND PDCH
CONCENTRATIONS IN ALL SAMPLES,
24
-------
350°
000°
—r~
Azimuth From Release Site
010°
020°
5000
1000
500
in
^
O
£
o
I
o
o
I
o
100
50
10
T
T
100 KM Arc
July 8,1980
Sampling Period (GMT)
10
(Kilometers)
I I
J L
J L
_L
J L
2100-2145
2145-2230
2230-2315
2315-0000
0000-0045
0045-0130
0130-0215
12 13 14 15 16 17 18
Sampling Sites
19
20 21
Figure 10. Average 45-min PMCH concentrations along the 100 km arc from
the July 8 experiment.
25
-------
per 1015. The PDCH results from the BATS sequential sampler at Site 20 for the
2230 to 2315 GMT sampling period (see Table 8) show an average PDCH concentration
of 238 parts per 10^, in very good agreement with the Dual-Trap sampler.
Table 9. Dual-Trap Sampler Results at Site 20
(100 km Arc), July 8, 1980.
Sample
Mid-Time PDCH
(GMT) Parts per 1015
2227 60
2231
2236 180
2241 250
2246 660
2251 320
2255 400
2300 280
2304 160
2309 45
2314 35
A single whole air bag sample for the entire sampling period was collected at
each BATS sampling site. Laboratory analysis of aliquots from these samples, per-
formed at BNL, indicated nearly all were severely contaminated and could not be
used. It appears that the contamination (concentrations of SFg^ PMCH, and PDCH
all were too high) most likely occurred while the aliquots (in small plastic bags)
were stored at the BNL laboratory. Pin-hole leaks in the bags could have allowed
a slow penetration of laboratory air which often has very high concentrations of
all three tracers. Fortunately, aliquots from some of the same whole air samples,
taken by LASL for analysis, showed no evidence of contamination. Comparison of
their results with the BATS data at the same sites is shown in Section 5.5.
3.5.2 100 km aircraft samples
Tracer concentrations measured on three passes over the flight path shown in
Figure 2 are given in Table 10. Figure 11 shows the average PMCH concentrations
on each segment of the flight path (solid bars) along with the average PMCH
measurements obtained at the ground with the BATS samplers at about the same
time (2230-0000 GMT). Concentrations aloft are quite comparable to those at the
ground.
Since good PMCH data were not obtained at either end of the flight path, es-
timated PMCH concentrations (dashed bars) were calculated from the measured SF6.
First, the general SF£ background of 600xlO~15 was subtracted from the measured
SFg. Then the SFg concentration was divided by the SFg/PMCH release ratio of 3.41.
For the three passes along flight segment B to E, the PMCH estimated from the SFg
value is very close to the measured PMCH. The estimated PMCH along segments A-B
and E-F indicate a sharp drop in concentration. In fact, these concentrations were
probably very close to background, since other analyses done at LASL suggest that
26
-------
Table 10. Airborne Whole-Air Sample Concentrations
(parts per 1015),
Sampling Time (GMT)
(July 8, 1980)
Path A to B
PMCH
(1)
PDCH
(1)
(1) BNL analysis.
(2) LASL analysis.
? Bad data (contamination).
No analysis performed.
-91(2)
Me-21
2342-2346
2348-2353
2302-2305
2337-2342
2353-2357
2306-2311
2332-2337
2357-0002
2311-2316
2327-2332
0002-0007
2316-2321
2324-2327
0007-0012
7
7
990
930
880
3200
2800
5400
1300
7
7
7
7
7
7
Path B to C
835
985
810
Path C to D
3300
2400
4600
Path D to E
1400
7
7
Path E to F
7
1200
1300
3600
3100
3900
12700
8500
14500
4300
2100
1600
1200
1300
1300
—
8.87
6.21
6.87
36.6
21.8
29.6
9.92
—
27
-------
Azimuth From Release Site
340°
5,000
1,000
500
in
^
o
I
fi
£
o
8 100
o
O
I
i 10
350° 000° 010° 020°
"1 ' I ' 1 ' T
030°
(3)
— ' ' PMCH Calculated from
X
(#)
Average surface PMCH Cone.
(2230-0000 GMT)
Number of samples
Average Aircraft PMCH Cone.
100 KM ARC July 8,1980
6 8 10 12
14 16 18
Sampling Sites
20 22 24 26
Figiwe 11. Comparison of PMCH concentrations aloft with surface concentrations.
28
-------
SFg values on the order of 1100x10-15, obtained with their analyzer, are actually
background values. Thus, the aircraft samples suggest that only a small portion
of the plume extended west of Site 12, where no surface samplers were deployed.
The whole-air samples obtained in this flight also provided critical data for
intercomparison of PFT's, heavy methanes and SFg measurements (see Section 5.4).
3.5.3 600 km surface samples
Forecast trajectories based on 1200 GMT (July 8) wind data indicated that a
tracer release starting at 1900 GMT would arrive at the 600 km arc at about 1300
GMT (July 9) with the center line of the plume crossing over Site 20 (Hamilton,
MO). All sampling sites (note that Site 8 had been eliminated) along the 600 km
arc (Figure 3) were alerted to start sampling at 0800 GMT (July 9), five hours
before the expected arrival time. A low level wind jet developed during the night
transporting the tracer material faster than expected and further to the west.
The 3-hour sample concentrations along the 600 km arc are given in Table 11.
Sites that are not listed (14, 21, 22) failed to obtain samples.
The PMCH concentrations for the first 6 sampling periods are shown in Figure
12. The peak plume concentrations arrived at about the time sampling commenced.
The plume centered between Site 9 (Firth, NE) and Site 15 (Skidmore, MO). By
2300 GMT (sampling period 6) the PMCH at all sites was near the background level
of about 2 parts per lO-". Sampling sites east of Site 19 are not included in this
figure since all samples at those sites showed background levels during the first
6 sampling periods.
The entire record of PMCH concentrations at all sites between July 9 (0800
GMT) and July 11 (2000 GMT) along the 600 km arc is shown in Figure 13. The or-
dinate shows the sampling sites plotted as a function of azimuth from the release
site.
Solid dots indicate a measured PMCH concentration less than 3 parts per 10-L5
(for all practical purposes they can be assumed to be background concentrations)
while crosses indicate concentrations at or above 3 parts per 10^-5.
The initial plume probably arrived at the 600 km arc just before sampling
began at 0800 GMT on July 9 with a duration of about 15 hours (2300 GMT, July 9)
before background levels are seen at all locations. Background concentrations are
seen for the next 15 hours, whereupon the July 11, 1400 to 1700 GMT samples (about
40 hours after release) show a secondary plume arriving at the 600 km arc.
The maximum concentrations of this secondary plume are about two orders of
magnitude lower than the initial plume but they cover a much larger area. Although
PDCH concentrations are close to background (26xlO~-'-5) they confirm, the presence
of the secondary plume. The duration of this plume on the arc was about 30 hours.
At present we are not sure whether this is a return of the initial plume or, possi-
bly, tracer material that lagged behind the main plume. These data provide an
interesting meteorological case study and will be investigated further.
4. 100-KM EXPERIMENT
A second, more limited, tracer experiment was conducted on July 11, 1980 to
provide another test of the perfluorocarbon system.
29
-------
TABLE 11
600 KM ARC
15
TRACFR CONCENTRATIONS (PARTS PER 10 )
STATION
START
TIME(GMT)
JULY 09
0800
1100
1400
1700
2000
2300
JULY 10
0200
0500
0800
1100
1400
1700
2000
2300
JULY 11
0200
0500
0800
1100
1400
1700
PMCH POCH
?,5
29
27
27
26
23
01
PMCH* PDCH*
2.0
2.0
1.7
1,7
1.7
1.5
1.7
1.7
1.6
1,5
1.6
1.8
1.4
1.6
2.2
2.6
?.o
1.8
1.8
1,6
21
19
18
17
17
X5
16
16
17
16
m
15
15
16
16
13
17
15
17
16
02
PMCH POCH
3.6
3.5
3.1
2.8
2.4
2,4
2.3
2.3
2.3
2.6
2,4
2.1
2.5
2.3
5.0
5,4
3,2
2.6
2,2
2.2
28
29
27
22
21
21
22
24
20
24
23
21
23
19
23
22
25
24
20
21
03
PMCH POCH
3.7
2.3
2.3
2.2
3.1
2.8
26
25
?6
24
28
STATION
04
PMCH PDCH
START
TIME(GMT)
JULY 09
0800
1100
1400
1700
2000
2300
JULY 10
0200
0500
0800
non
1400
1700
2000
2300
JULY 11
0200
0500
0800
1100
1400
1700
2.6
2*7
t.6
? ,3
' , 5
9 5
2*5
2.4
?*4
' .4
3.4
8.6
3,0
3*2
9*5
2,4
2,5
i ij V. • .
30
30
30
28
27
28
28
29
28
28
28
27
29
33
31
30
29
30
30
05
PMCH PDCH
2,3
2.9
2.9
2.8
2.7
2.8
2.7
2.7
2.4
2,9
6.4
3,9
2,5
2.9
2,4
2,4
22
J*9
39
31
32
?3
Si
•SO
30
34
32
30
30
2°
30
06
PMCH PDCH
17.2
10.5
2.6
2,6
2.9
3,0
2,4
2.6
sli
2,6
3,5
3,8
6,0
9.0
6,6
4.0
3.2
2,7
2,5
37
70**
26
27
26
26
27
27
28
27
27
27
29
32
30
29
28
28
27
- NO DATA
* CONCENTRATIONS TOO LOW (DESORPTION PROBLEM!,
** CONCENTRATION TOO HIGH,
07
PMCH PDCH
3.2
2.8
2^8
2.4
3,2
2,3
2,4
2.3
2.5
3.3
8.2
10.1
9.8
4,0
5,6
4.8
2.7
2.4
2.3
30
32
27
28
27
27
2fl
28
28
29
31
34
34
29
30
30
28
2b
26
30
-------
TABLE 11 (CON'T)
600 KM ARC
15
TRACER CONCENTRATIONS (PARTS PER 10 »
STATION
START
TIME(GMT)
JULY 09
0800
1100
1400
1700
2000
2300
JULY 10
0200
0500
0600
1100
1400
1700
2000
2300
JULY 11
0200
0500
0600
1100
1400
1700
STATION
START
TIME(GMT)
JULY 09
0800
1100
1400
1700
2000
2300
JULY 10
0200
0500
0800
1100
1400
1700
2000
2300
JULY 11
0200
0500
0800
1100
1400
1700
627.
63;
^ 2
2.6
2.2
? . 3
2.4
O
4.1
6.2
" _
? « 3
2.2
2.1
13
PMCH
960,
500*.
350
66.
_
«
2.9
** .4
7.9
16*.
6,8
7,0
7,9
10*
7*.3
6.3
09
PMCH PDCH
598
82
26
24
22
25
25
25
25
25
25
26
*»
25
27
26
25
24
23
POCH
790
420
310
83
28
27
28
32
39
31
31
33
35
35
34
33
10
PMCH PDCH
11
PMCH POCH
12
PMCH
POCH
1280.
820.
26.
2.7
3.6
2,8
2.8
2,8
2.8
2.7
2.6
4.3
4.5
5.1
5.7
5.3
4.2
2.5
2.3
2,4
970
6"0
44
25
25
24
25
25
26
27
26
26
26
28
29
29
29
26
26
25
15
PMCH PDCH
16.
2.6
-------
TABLE 11 (CON'T)
600 KM ARC
15
TRACER CONCENTRATIONS (PARTS PER 10 )
STATION
START
TIME(GMT)
18
1100
1400
1700
2000
2300
JULY 10
0200
0500
0800
1100
1400
1700
2000
2300
JULY.11
1100
1400
1700
9.4
2*5
PDCH
27
2.6
27
19
PMCH PDCH
2.6
2.6
2.5
2.5
2,4
2.4
2.4
2.4
2.4
2.5
5.1
10.
6.7
7.4
8.7
1:1
20
PHCH* PDCH*
23
1.8
1.7
2.4
2.9
2,4
2.3
2.2
2.0
2.1
2.1
2.3
2.4
1.6
2.4
8.7
?•!
f . o
5,3
3.1
2.4
18
18
22
24
23
22
22
20
22
22
21
22
14
23
29
31
27
26
11
PHCH
2.5
PDCH
2,6
2,5
2,4
2,5
2.3
2.3
2,4
2,4
2.4
2.4
2.4
2.3
2.3
2.3
28
23
27
26
26
25
26
27
27
27
27
26
26
26
27
18
5>7
STATION
START
TIHE(GHT)
JULY 09
0800
1100
1400
1700
2000
2300
JULY 10
0200
0500
0800
1100
1400
1700
2000
2300
JULY 11
0200
0500
0800
1100
1400
1700
- NO DATA
24
PMCH PDCH
25
PMCH PDCH
2.0
2.0
?.0
2*1
2.1
2.0
9.1
9.1
?:.i
?'.!
1 .9
1.8
2ll
4.4
7.9
9.3
27
27
28
28
28
27
26
25
26
27
27
27
27
26
24
25
26
29
33
34
2.5
2.7
2.6
2.5
2.4
2.4
2.5
2,5
2.6
2.4
2.4
2,7
4.1
2,6
2.5
2,4
2.4
27
28
28
*5
25
27
28
28
*9
28
27
27
28
29
30
29
29
28
28
26
PHCH PDCH
2.2
2.2
2.2
1.9
2.1
2.1
2.2
2.0
1.9
1.9
1.9
3.0
5.7
2.4
2.2
27
PHCH POCH
2,6
2,6
2,4
2.4
2.5
2,5
2,5
26
27
26
27
26
27
27
27
* VALUES UNCERTAIN,SAMPLE VOLUMES HAD TO BE ESTIMATED.
32
-------
TABLE 11 (CON'T)
600 KM ARC.
15
TRACER CONCENTRATIONS (PARTS PER 10 )
STATION
START
TIME(GMT)
JULY 09
0800
1100
1400
1700
2000
2300
JULY 10
0200
0500
OflOO
1100
moo
1700
2000
2300
JULY 11
0200
0500
0800
1100
moo
1700
STATION
START
TIME(GMT)
JULY 09
0800
1100
1400
1700
2000
2300
JULY 10
0200
0500
oaoo
1100
1400
1700
2000
2300
JULY 11
0200
0500
0800
1100
1400
1700
2.3
2.7
2,0
2.2
2.1
2.1
2.3
2.3
2.3
?,2
?.2
'Jo
2,0
1^9
3,1
4.6
•*,3
1,9
1,7
1.7
32
PMCH
2,8
28
2.7
2.3
2.6
?.6
2.9
2.7
2.7
2.8
2,7
2.7
2.6
2.6
?,7
?.8
? 6
2,7
?,7
2.6
28
PMCH PDCH
27
29
25
25
25
25
28
28
28
26
26
25
23
22
22
23
21
21
PDCH
30
30
30
29
29
29
29
30
30
31
30
30
29
29
30
30
30
30
30
29
29
PMCH PUCH
30
PMCH PDCH
31
PMCH PDCH
2.5
2.5
2*4
2.3
2^5
2.5
2.4
2.5
2.5
2.6
2*4
2,6
2.7
3.0
2.8
2,5
2.7
33
PMCH
2.6
2.6
2.6
2.4
2.5
2.5
2.4
2.6
2.5
2.6
2.7
3.3
2.6
2.5
2.5
2.6
2.6
2.7
3,1
2.5
28
^9
29
28
29
29
29
•5ft
30
29
29
29
28
29
a°
50
29
30
PDCH
27
28
26
26
26
26
25
27
28
28
28
33
27
27
28
28
29
29
29
28
2.0
2.4
1.9
1.9
2.0
27
29
26
25
26
27
2.0
2.1
2.0
2^0
2,0
2.0
2.0
2.2
2.3
2.2
2.1
2.1
27
27
27
26
26
26
26
27
28
27
27
26
34
PMCH PDCH
3.4
2,9
2,6
2.6
2.5
2.6
2,7
2,7
2,9
2.8
2,7
2.6
2.3
2.6
2.6
2.7
2.6
277
2.5
27
27
26
25
25
25
27
26
26
27
26
26
23
25
25
26
24
26
24
2.9
3.3
2.8
2.9
3.1
2.8
2.5
2.7
2.6
3.3
2,6
2.6
2,6
2.6
2,7
2.8
2.8
2.8
2,7
2.7
25
26
26
22
25
27
28
28
27
27
27
27
27
27
26
28
29
?9
?7
27
35
PMCH POCH
2,3
2.4
2,5
273
2,3
2.4
2,4
2.3
2.5
2.5
30
30
29
29
3D
30
30
30
30
30
NO DATA
33
-------
TABLE 11 (CON'T)
KM ARC
15
TRACER CONCENTRATIONS (PARTS PER 10 )
STATION
36
PMCH PDCH
START
TIME(GMT)
JULY 09
0800
1100
moo
1700
2000
2300
JULY 10
0200
0500
0800
1100
moo
1700
2000
2300
JULY 11
0200
0500
0300
1100
moo
1700
?.l
?.o
9.0
9.0
9.0
?.o
2,2
2.1
?.l
9.2
9.2
9.0
1.9
1.8
?..!
?.0
1.9
?0
9.0
1.9
26
26
25
26
26
26
25
26
26
26
26
26
26
23
24
24
23
25
24
24
• NO OATA
37
PMCH PDCH
2.6
2.3
2.3
2.5
2.1
2.0
2.3
2.0
2.6
9.8
2.5
2.7
2.7
2.6
2.2
2.6
2,8
27
27
26
27
28
27
26
26
28
26
26
26
28
27
27
38
PMCH PDCH
2.1
2.0
1.8
1.7
1.8
1.9
2.0
2.4
2.0
2.0
1.9
1.8
1.9
2.0
2.0
2.2
2.1
2.0
1.9
27
26
26
24
24
25
25
31
25
24
24
25
25
26
26
27
26
25
25
34
-------
352°
356°
000°
Azimuth From Release Site
004° 008° 012° 016°
020°
024°
028°
I
I
I
I
1000
500
600 KM Arc July 9,1980
Sampling Period (GMT)
in
*b
a>
a.
£ 100
to
• 50
re
h.
+^
c
01
u
I
U
10
1
2
3
4
5
6
0800-1100
1100-1400
1400-1700
1700-2000
2000-2300
2300-0200
L_L
' '
L_L
_L
_L
45 679 1011 1213
Sampling Sites
14
1516 17 18 19
Figure 12. Average 3-hour PMCH concentrations along the 600 ten arc.
35
-------
OJ
V)
a.
E
(0
V)
38
36
34
32
30
28
26
24
22
•t 20
o
8
18
16
14
12
10
8
6
4
PMCH Concentration (Parts per 1015)
Denotes Cone. < 3
Denotes Cone. >3
-[ 1
~i r
&
i i l L
JL
5 GMT 8 11 14 17 20 23
U July 9 »
j l i L
8 11 14 17
July 10
20 23
J I 1 1 L
070°
060°
050°
040°
01
tr
030° 2
020°
010°
000°
350°
8 11 14 17 20 GMT
_ July 11
Figure 13. Average 3-hour PMCH concentrations along the 600 fan arc for the period July 9, 0800 GMT
to July 11, 2000 GMT.
-------
4.1 Tracer Release
The two perfluorocarbons and SFg were released over a 3-hour period (1900-
2200 GMT) using the same release systems at the same site as in the first experi-
ment. Release amounts, shown in Table 12, were calculated to produce concentra-
tions well above the detection limits at the 100 km arc. Also shown are the tracer
release ratios (by volume).
Table 12. Tracer Releases on July 11, 1980.
Tracer Release Amount (kg)
PMCH 21
PDCH 26
SULFUR 283
HEXAFLUORIDE
Tracer Release Ratios
(by Volume)
PMCH/PDCH: 0.91
SF6/PMCH: 33
SF6/PDCH: 30
4.2 Sampling Array
In this experiment, sampling was done only at 100 km downwind of the release
site, using the same array as in the first experiment. Based on the latest tra-
jectory forecast, three EML sampling teams deployed the BATS sequential samplers
to Sites 13-30. The tracer release began at 1900 GMT (2 PM) and the samplers were
set to start at 2200 GMT (5 PM) and take nine 45-minute samples. The same sam-
plers were used in both experiments with sampling tubes 1-10 being exposed in the
July 8 experiment and tubes 12-13 exposed on July 11. As in the first experiment,
a whole-air sampler was co-located with each sequential sampler to collect a single
sample over the entire period for comparison of SF^ and perfluorocarbon concentra-
tions.
4.3 Meteorology
As shown in Figure 14, on July 11 the broad area of high pressure continued
to dominate most of the U.S. The wind flow in the boundary layer over the 100 km
experimental area remained from the south-southwest.
37
-------
WtlDAY. JULY U, 1980. •
uo
00
•aj,.jt c -
Figure 24. Surface weather map for 1200 GMT, Friday., July 11, 1980.
-------
4.3.1 Special rawinsonde observations
Special rawinsonde observations up to 500 mb were again taken at Tinker AFB
starting on the morning of July 11. Data are given in Table 13 for observations
from July 11, 1800 GMT to July 12, 0000 GMT. The calculated transport layer height
(TLH) and average wind speed and direction in the layer are given for each sounding.
4.3.2 Post-facto tracer trajectories
Tracer trajectories shown in Figure 15 were calculated for the start and end
of the release period using the Tinker AFB soundings. Since the plume was still
passing over the 100 km arc at the time of the last Tinker sounding, an average
wind of 180 deg and 6 m/sec was estimated to complete the 2200 GMT trajectory. The
estimated plume width is shown in the figure. The calculated plume position and
arrival time at the 100 km arc agree well with the measured tracer data (see
Figure 16).
4.4 Sampling Results
The BATS sequential samplers were installed at Sites 13 through 30. The 45-
minute tracer concentrations are given in Table 14. Data for sites not listed were
lost due to sampler malfunction (Site 13) or analysis problems (Site 25).
The PMCH results are plotted in Figure 16. The initial sampling period (2200-
2245 GMT) showed concentrations near background at all sampling locations. The
next sampling period (2245-2330 GMT) shows concentrations at Sites 14 through 24
at about 50 times background levels. During the third sampling period (2330-0015
GMT) peak plume concentrations are reached at Sites 14 through 21 with decreasing
concentrations to the east.
During subsequent sampling periods, an orderly decrease in the PMCH concen-
tration occurs at all sampling sites and by the eighth sampling period (0315-0400
GMT) the concentration are approaching background levels again.
Since there was not sampling west of Site 14, the plume width could not be
determined but the trajectories (Figure 15) suggest that the plume did not extend
much beyond Site 14.
Whole-air samples again were unusable due to contamination which apparently
occurred in the BNL Laboratory.
5. EVALUATION OF PERFLUOROCARBON TRACER SYSTEM
These experiments were designed primarily as a proof-test of the perfluorocar-
bon tracer system. Our evaluation will focus on the performance of the release,
sampling, and analysis systems, and the reliability of the tracer concentration
measurements.
5.1 Tracer Release
The two perfluorocarbons were released via separate, but identical, aerosol
spray mechanisms. In both experiments, 3-hour releases were accomplished without
any problem and the actual release amounts were within 10% of the intended amounts.
39
-------
K O Q_*
£ OUIOvOvOvOvOsOsOCMvOvOvOT-iOvOO zc
CD CLCO ••••••••••••••• LJCOt-IO^cOCOcOcor*-
O CO XtOinvD vOvDNDsD^-^-NOin^K5fOFO t—LJIOCMCMCNiCMCMCM
o s: Q
O
QLJ
CLCOVO
COX
£
Ul
s:
»-»
at
LJ
a
x
LJ
o cccDoooooinooinmtnoooin
eo i-tLu<-icr>cocoo>o>o*-4CMK>;*-atr»3-w)
QOCMi-l«-liHiH.-4CMCMCMCMCVIC\KMCMCM
.jr-oinocooinomomoooin
co d-Oi-finoi-io
LJ
oca
£L
CM
»-COeOr-ieOoeOK)in
• t-»CMK)K)fOin
s:
XCOK)
_is:eo
LJ
X
QC
O
<
h-
O
o
CO
CO
u.
O
2. OLJOi-lCMr-iHr-lt-ICMvOvOsOtf'fHvOvO ZZ tHCOCO^r^-vOsDO OLJ
CD CLCO ••••••••••••••• LJCDiHOO^cocbcocO^ CLCOvO
CM co xrovoh-h-sOsOvor* \o \oin^-into in KLJIOIOCMCMCMCMCMCM cox
«M z: as:
o cccoininomoinoinincooinin co
CO i-lLJ:J-OOi-tr-ti-IOi-lT-li-ti-ICMC\jrOeO
OOCM CM CM CM CM CM CM CM CM CM CM CMCM CM CM
a.
DO CM
X f. IO vD (T> CM in «O iH d- r>~ «HIO \O-lr-ICMi-ICMi-ICMCMCMCMCMCM
-J
O _ar^omoi-ioinoinoioino
~3 1-COcOiHi-ICMtOIOIOiJ-d-inCMnvl)
xs: 10 \OONCM in eo 1-1 d-h-oi-iiovo
i-l • »-»»Hr-ICMCMCM»OtOlOM
xcom
_is;co
Q.X. (J
5: Oh-COcONflNOvOi-l QLJ
LuCDf-IO>COCCCOCOCOI^ CUCONi,
l-LJIOC\lCViC\iC\iCMCMCM COX
Q 2:
CO vDOvOvDOCTiOO OTOh-
LJcoNOinina-ocor-o i-tLjcr
Q.
_lf~iHc»-ecioioiOo _lo
XETfOininNOi-tiomo* _is:r^
• i-ICMCMK)K)»Oin I- »CM
Si £
cr
LJ
K O
z: OLjoi-tNor-NOt-ii-ivOiHi-iNCNOsc
eo CLCO •••••••••••••
omocMoinoinioinoo
"3 >-cocoi-ii-icM*»o»od-*oinNor>-
u
OLJ
a.cotn
I-LJIOCMCMCMCMCMCMCM COX
O £
CO
l-lLJiH
a:z:
-------
36.0
|
35.5
1
100 KM
Sampling
Sites
10
0000
Plume Width
2000
1900 O 2200 GMT (July 11)
Release
Site
30
35.0
98.0
97.5
97.0
Longitude (°W)
96.5
Figure 15. Calculated transport layer trajectories to the 100 km arc for the 3-hour tracer
release on July 11.
-------
TABLE IH
100 KM ARC
TRACER CONCENTRATIONS (PARTS PER 10
STATION
START
TIMEJGMT)
JULY 11
2200
2245
2330
JULY 12
0015
0100
0145
0230
0315
0400
STATION
START
TIME(GMT)
JULY 11
2200
2215
2330
0015
JULY 12
0100
0145
0230
0315
0100
STATION
START
TIME(GIT)
JULY 11
2200
2245
2330
JULY 12
0015
0100
0145
0230
071?
0400
It
PMCH
1.6
38.
270.
178.
65.
37.
1?.
•5.4
18
PMCH
4,5
71 1
259.
216.
99.
49.
24,
10!
7.1
22
PMCH*
in.
300.
200.
100.
40.
201
5.
?,
PDCH
27
61
337
221
97
70
44
34
w
PDCH
33
111
349
299
149
87
57
40
36
PDCH
34
400
240
84
65
39
30
30
28
15
PMCH
3.*
50,*
280.
190.
_
It*
2.*
2.*
3.*
19
PMCH
?.*
40.*
330.
-
_
_
_
-
23
PMCH
4,0
219.
145.
46.
20.
6.*
3.*
9.*
PDCH
26
«7
33?
235
70
38
26
*5
PDCH
36
'4
4"0
•
—
•
m
m
m
PDCH
28
2&3
175
67
40
25
^•2
24
16
PMCH
3,8
110.
241,
142,
68.
25.
9,5
5.3
4,7
20
PMCH
4,2
38.
302.
236,
66,
30.
5>3.
7,7
24
PMCH
a. 2
56.
76.
33.
25.
9,3
5.5
5.4
5.2
l-'DCH
26
157
323
207
117
54
3R
30
30
PDCH
27
60
357
288
95
55
47
30
w
PDCH
28
84
109
60
51
31
30
28
28
17
PMCH**
2,
12,
180,
90,
50,
28,
12.
•
21
PMCH
3.5
127.
360.
126.
42,
21.
11.*
1.*
11.*
26
PMCH
3, -5
3,9
14.
7,9
3.5
3,1
3.1
3.3
3,1
PDCH**
23
9
150
70
50
24
13
-
••
POCH
28
186
430
164
71
*5
33
27
25
PDCH
25
26
37
31
25
25
24
24
24
SEE FOOT NOTE A
• NO DATA
* VALUE UNCERTAIN DUE TO CONTAMINATION IN LAB ANALYZER,
** POOR DE«?ORPTION» CORRECTION ESTIMATED.
A SAMPLING SITES 27-30 HAD BACKGROUND PflCH AND PDCH
CONCENTRATIONS IN ALL SAMPLES,
42
-------
000°
010°
T~
Azimuth From Release Site
020°
1
T
030°
—T~
040°
~T~
T
500
100
r 50
S.
o
I
J
o
a.
10
Sampling Period (GMT) _
100 KM Arc
July 11,1980
1
2
3
4
5
6
7
8
2200-2245
2245-2330
2330-0015
0015-0100
0100-0145
0145-0230
0230-0315
0315-0400
0 10 20
(Kilometers)
I I
J I
I
I
J I
I
I
J I
I
l I I
14 15 16 17 18 19
20 21 22 23
Sampling Sites
24 25 26 27 28 29
30
Figure 16. Average 45-min PMCH oonoentrations along the 100 km arc from
the July 11 experiment.
43
-------
A newly designed release mechanism, in which the tracers are vaporized, was
not ready in time for these experiments but it was tested and used successfully in
September, 1980. The new system provides more precise control and continuous re-
cording of release rates.
5.2 BATS Sampling and Analysis System
The heart of the perfluorocarbon tracer system is the BATS automatic sequen-
tial sampler and the associated analysis apparatus.
The over-all performance of the BATS system was excellent in this first field
trial. As shown in Table 15, 72% of the 1121 scheduled samples provided good tracer
concentration data. Of the 28% lost, or poor data, 5% was due to human error (e.g.,
failure to turn on the sampler) and the remainder was about equally divided between
sampler malfunctions and sample analysis problems. Sampler malfunctions were due
most often to pump failures. Some units developed problems in the electronic con-
trol circuitry. Modifications to the design of the BATS sampler are under consid-
eration to alleviate these problems.
Table 15. Performance of BATS Sampling-Analysis System.
Scheduled Samples
Good Data
Sampling Failures
Analysis Failures
Human Error
Total Lost or Poor Data
Number
1121
810
134
123
54
311
Percent
100
72
12
11
5
28
Several problems in the analysis apparatus and procedures became evident
during analysis of the large number of samples. The most troublesome was the
presence of a contaminant that interferred with the PMCH chromatograph peak in
many analyses. Eventually, it was discovered that the contamination was coming
from a screen used in the Porapak QS trap in the analysis apparatus. The resolu-
tion of this, and other minor problems, should reduce the amount of data lost in
analysis to well below the 10% level experienced in this experiment.
Another problem that complicates the determination of tracer concentrations
is the non-linear response of the present electron-capture detector. Various
attempts to reduce this non-linearity, which shows a change in response factor
as much as 2-fold, depending on the size of the sample being analyzed, have not
succeeded. Evidence in the literature suggests that the strength of the 8 mCi
"•%! foil in the electron capture detector is about an order of magnitude too
intense for the strongly electronegative PFTs. Lower activity foils will be sub-
stituted in an attempt to correct this problem.
44
-------
5.3 Reliability of BATS Concentration Measurements
The reliability of tracer concentrations obtained with the BATS sequential
samplers can be checked by comparing the PMCH and PDCH measurements. If the tra-
cers behave identically in the atmosphere and the tracer release, sampling, and
analysis systems function perfectly, the measured PMCH/PDCH concentration ratio
in every sample (after backgrounds are subtracted) would be the same as the ratio
of the release amounts of the two tracers. Comparison of the measured tracer
ratios with the release ratio therefore provides a good test of the entire tracer
system.
Figure 17 shows a plot of PMCH versus PDCH concentrations measured with the
sequential samplers on the 100 km arc in the July 8 experiment. Background con-
centrations (2.4 parts per 10^ for PMCH and 26 parts per 10^^ for PDCH) have been
subtracted out. Only those 22 samples where both tracers had concentrations at
least twice background were used for this comparison. When the concentrations are
near background, uncertainties in the background value can have a large effect on
the tracer ratio. The straight line in Figure 17 represents the tracer release
ratio of 1.18. The measured mean ratio in the 22 samples is exactly what it should
be and there is remarkably little scatter about the true ratio.
Very similar results were obtained at the 600 km arc (Figure 18). There are
13 samples with concentrations at least twice background and the mean measured
PMCH/PDCH ratio is 1.19, again with very little deviation from the release ratio
of 1.18. The inset shows a plot of PMCH versus PDCH for the highest concentra-
tions observed during the second appearance of tracer on the 600 km arc on July 10.
The PDCH concentrations were less than twice background in all of these samples.
It is, therefore, not surprising that the data show more scatter. The mean ratio
of 1.09 is still quite close to the 1.18 release ratio.
Many samples obtained at the 600 km arc over the 3-day period showed back-
ground concentrations. The background to be subtracted from each concentration
was estimated separately from each sampler. Estimated background values varied
from 2.2 to 2.5xlO~15 for PMCH and from 24 to 28xlO~15 for PDCH. The consistency
of the background measurements attests to the good precision of tracer measurements
at these levels.
Figure 19 shows a plot of PMCH against PDCH concentrations on the 100 km arc
for the July 11 experiment. In this experiment there were 31 samples with concen-
trations at least twice background. The mean measured PMCH/PDCH ratio was 0.88,
with very little scatter about the release ratio of 0.91.
In the two experiments there were a total of 66 BATS sequential samples with
concentrations at least twice background. Most of the measured PMCH/PDCH ratios
are within ±5% of the release ratio, all are within ±20%, over a concentration
range from 20 to 5000 parts per 10 . These results are excellent but they do not
constitute a complete test of the BATS samplers. Any inaccuracies in sample vol-
ume or mechanical problems (e.g., timing errors) would have the same effect on
both tracers. Therefore, as a further test, duplicate BATS samplers were set up
at three of the sampling sites. As luck would have it, data from one sampler in
each pair, were lost due to failure of the sampler or analysis problems. Some
degree of independent verification of the BATS result was achieved by comparison
with whole-air samples (Section 5.5).
45
-------
PMCH Concentration (Parts per 1015)
g
o
o
' I I'"I
n—r
Ul
o
o
o
o
o
§
o
o
|g
O
O
3
O
IB
3
O
3
13 ui
1 §
o
o
o
s co
* 3
a> 3
3 "5.
30 g
a> r:
-» N)
'_» W
00
UI
S
o
o =. x
0 •< 1
^ oo 5
J L
I I I I
J I L
-------
1000
500
in
r-
a
o
'^
£
*-»
c
8
I
100
50
10
\ I
n i r
Experiment # 1 - July 8,1980
600 KM Arc
Samples: 13
Mean Ratio: 1.19
Samples: 10
Mean Ratio: 1.09
Release Ratio
PMCH/PDCH = 1.18
I I I I I I I
I
10 20
PDCH (Parts per 1015)
I I I I I I I I
10
50 100 500
PDCH Concentration (Parts per 1015)
1000
Figure 18. Comparison of PMCH and PDCH concentrations from the 600 km BATS
samples.
47
-------
500
in
"o
a>
a
100
o
c
a>
o
I
O
50
10
Samples: 31
Mean Ratio: 0.88
Release Ratio
PMCH/PDCH = 0.91
Experiment #2
July 11,1980
100 KM Arc
J L
1
J I L_L
_L
J L
10
50 100
PDCH Concentration (Parts per 1015)
500
1000
Figure 19. Comparison of PMCH and PDCH concentrations from the 100' hn BATS
samples on July 11.
48
-------
5.4 Comparison with Other Tracers
Whole-air samples were collected in plastic bags at sequential sampling sites
on the 100 km arc in the July 8 experiment in order to compare SF6, and heavy
methane tracer measurements. Unfortunately, most of the whole-air sample aliquots
sent to BNL for analysis became contaminated with PFTs and SF5 (apparently at BNL)
to an extent that made them useless for tracer intercomparisons. It should be
noted that the BATS sequential samplers were designed to avoid the contamination
problems that had been encountered previously in handling whole-air samples.
In spite of the contamination problem, some good data have been salvaged from
the whole-air samples. Aliquots were sent to LASL for analysis of heavy methanes
and SFg. Seven of the samples collected in the aircraft flight over the 100 km arc,
that were analyzed at LASL, appeared to be free of PFT contamination when analyzed
at BNL. After subtraction of appropriate background values: 2.4x10"-^^ for PMCH,
26xlO-15 for PDCH and 600xlO~15 for SFf, (Me-21 background is nil), the tracer con-
centrations were plotted in Figure 20. On the upper left, PMCH is plotted against
PDCH. The PMCH/PDCH ratios are quite good in these samples with a mean value of
1.08 and little scatter about the line representing the release ratio of 1.18.
On the upper right, PMCH concentrations determined at BNL are plotted against
SF6 concentrations determined at LASL. The mean SFg/PMCH ratio in these samples
is 3.08 compared to the releae ratio of 3.41 and the individual sample ratios show
only slightly more scatter than the PMCH/PDCH ratios.
The lower graph in Figure 20 shows PMCH concentrations plotted against
Methane-21. The mean PMCH/Me-21 ratio is 131, very close to the release ratio of
137 and again the scatter is small. Considering that the heavy methane and per-
fluorocarbon determinations are made by totally different analysis techniques (mass
spectrometry for the methanes, gas chromatography for the PFTs) these excellent
results inspire confidence in both tracer systems.
In all three comparisons shown in Figure 20, the mean of the measured tracer
ratios is within 10% of the release ratio and all individual sample ratios are well
within a factor of two of the release ratio. We can conclude from these data that
all tracers behaved the same in the atmosphere, faithfully following the air mo-
tions with no significant depletion mechanism out to 100 km from the source. We
would expect the same to hold true at 600 km and beyond. We hope to confirm this
with a total of ten whole-air samples, collected at two LASL cryogenic sites within
the plume at the 600 km arc, which have not yet been analyzed.
5.5 Comparison of BATS with Whole-Air Samples
It was intended that PFT concentrations from the BATS sequential samplers on
the 100 km arc would be compared with those from the co-located whole-air bag sam-
ples. Contamination of the aliquots sent to BNL from the bag samples rendered them
useless for intercomparison of perfluorocarbon measurements. However, the analyses
of SF6 and Me-21 concentrations, done at LASL, are available for five of the bag
samples, along with PFT analyses from the BATS sequential sampler at the same lo-
cations. The PMCH and PDCH concentrations from the 45-minute BATS sequential
samples were averaged over the time interval that the whole-air sample was col-
lected at each location. The concentration of each tracer is shown in the upper
part of Table 16. Background concentrations were subtracted from each value and
tracer ratios were determined as shown in the lower portion of the table. The mean
49
-------
5,000
in
^
o
r-
I
43
I
I
U
1,000
700
Samples: 7
Mean Ratio: 1.08
J I
I
700 1,000
. Release Ratio
PMCH/PDCH= 1.18
I I
5,000
PDCH (Parts per 1015)
5,000
in
t™
O
0!
a
o
a.
1,000
700
Samples: 7
Mean Ratio: 3.08
2,000
SF-/PMCH = 3.41
O
I I
5,000 10,000
SF- (Parts per 1015}
20,000
Experiment #1 July 8,1980
100 KM Aircraft Samples
5,000
in
^
O
CD
a
£
o
i 1,000
500
i I IT
Samples 7
Mean Ratio: 131
I I I
Release Ratio
PMCH/M-21 =137
I
J I
5 10
Methane-21 (Parts per 1015)
40
Figure 20. Comparisons of tracer concentrations in whole-air samples collected
in the flight over the 100 kn arc on July 8.
50
-------
Table 16. Comparison of BATS sequential samples with whole-air
samples at the 100 km arc (July 8, 1980).
Site
Whole Air Bag
Sampling Period
Tracer Concentrations (parts per 10 )
PMCH
(1)
PDCH
(1)
SF,
(2)
Me-21
(2)
12
13
14
16
17
(1)
(2)
(GMT)
2130-0435
2121-0427
2012-0418
2032-0401
2001-0357
1030
1090
700
450
700
u
810 3250 7.93
950 3530 8.10
570 4030 9.37
410 2060 3.92
600 2720 5.50
BATS (sequential air sampler) concentration averaged over the period of
the whole air bag sample (analysis by BNL) .
Whole air sample
(analysis by LASL) .
Tracer Ratios
Site
12
13
14
16
17
Mean
PMCH/PDCH
1.31
1.17
1.28
1.15
1.21
Ratio 1.22
Release
Ratio 1.18
SF, /Me-21
o
332
362
366
372
385
363
467
SF,/PMCH PMCH/Me-2L
D
2.58 129
2.69 134
4.93 74
3.29 113
3.04 127
3.31 115
3.41 137
51
-------
PMCH/PDCH ratio from the BATS samplers is very close to the release ratio and the
individual values are within ±10% of the mean. The SF6/Me-21 ratios from the
whole-air samples are consistent though they are about 20% lower than the release
ratio. The mean ratio between SFg from the whole-air samples and PMCH from the
BATS samplers is 3.31, very close to the release ratio of 3.41 although the scatter
of individual ratios is relatively large. The mean ratio between PMCH (BATS) and
Me-21 (whole-air) is 115, about 15% lower than the release ratio. One reason for
the discrepancy between the BATS results and the whole-air sampler results may be
the failure of the whole-air sampler to pump air at a constant rate. It was dis-
covered during the experiment that the bag sampler pumping rate was erratic, pro-
bably because it was not designed for the extreme heat encountered in this experi-
ment. In spite of this problem, the correspondence between the BATS and whole-air
sampler results is good; all measured tracer ratios are well within a factor of two
of the release ratios.
5.6 Performance of Real-Time Samplers
The real-time continuous PFT monitor, intended for use in the sampling air-
craft, was not available because of various malfunctions. Efforts to repair the
instrument in the field were unsuccessful. The difficulties appear to be correct-
able and efforts are continuing to develop this instrument into a reliable con-
tinuous airborne monitor.
The Dual-Trap sampler functioned well and was used to provide field analyses
of some of the whole-air bag samples as well as real-time tracer concentration
measurements on the 100 km arc. Some difficulty was experienced in positioning
the Dual-Trap sampler within the tracer plume because of shifting wind conditions
at the arc. However, an excellent set of 5-min plume concentrations was obtained
alongside of a BATS sampler (see Table 9). Concentrations obtained with these two
samplers show very good agreement. The Dual-Trap sampler was later used exten-
sively in the ASCOT drainage wind experiments and provided hundreds of 5-minute
samples with real-time readout of PMCH and PDCH concentrations.
6. SUMMARY
These experiments have successfully demonstrated the capabilities of the per-
fluorocarbon tracer system and the feasibility of carrying out atmospheric trans-
port and dispersion experiments over distances of 1000 km or more. A release of
65 kg/hr of PMCH produced concentrations at the 600 km arc almost three orders of
magnitude above the background value of 2.4 parts per 10-*-^. This suggests that
a PMCH release rate of about 10 kg/hr should be sufficient to provide plume measure-
ments out to 100 km from the release point.
Reliability of the BATS sequential samplers is judged to be very good for the
first trial of a completely new system. About 12% of the 1121 scheduled samples
were lost because of sampler malfunctions. Another 11% were lost in analysis.
Modification of the BATS pump and relatively minor changes in the analysis appara-
tus should significantly improve the reliability of the BATS system.
Simultaneous measurements of PMCH and PDCH concentrations with the BATS system
were remarkably consistent. The PMCH/PDCH ratios in all samples were very close
to the tracer release ratio; most measured ratios were within ±5% of the release
ratio.
52
-------
Most of the whole-air samples, intended for comparison of PFT measurements
with SFg and heavy methane tracers, were of no use because of contamination of the
bag samples. However, a small number of samples, that could be analyzed for all
five tracers, showed generally good agreement, sufficient to establish that all
tracers behaved the same in the atmosphere.
Deployment of many samplers over the large area involved in a long-range ex-
periment can be very costly and present difficult logistics problems. One of the
objectives of this experiment was to test the feasibility of using the National
Weather Service sub-station network of over 12,000 sites to deploy the BATS sequen-
tial samplers. Substation specialists delivered the samplers to 39 selected sites
on the 600 km arc where cooperative observers, who take routine temperature and
precipitation measurements for the NWS, operated the samplers. The 600 km sampling
program was very successful as the cooperative observers carried out their assigned
role with competence and enthusiasm. Future long-range tracer experiments should
take advantage of the capability inherent in the NWS sub-station network.
7. ACKNOWLEDGMENTS
This work was supported by the Office of Health and Environmental Research,
Department of Energy and the Environmental Protection Agency.
Development of the perfluorocarbon tracer system has been carried out by the
Air Resources Laboratories, NOAA, in collaboration with the Dept. of Energy's
Environmental Measurements Laboratory and Brookhaven National Laboratory.
We wish to acknowledge our debt to Dr. James E. Lovelock who first conceived
the perfluorocarbon tracer system and designed the prototype samplers and analyzer.
The success of this experiment would not have been possible without the co-
operation of the many individuals, from the different agencies and laboratories,
listed at the beginning of this report.
To Dr. Edwin Kessler, Director, National Severe Storms Laboratory, we owe a
debt of gratitude not only for the support he and his staff provided but also for
the hospitality extended to the experimenters stationed at Norman, OK and the spirit
of cooperation that prevailed.
To Dr. Harold Meyers, Superintendent of the Agronomy Research Station, Oklahoma
State University, we wish to express our appreciation for the assistance he and the
OSU students provided in shipping, storing, and operating sampling equipment along
the 100 km arc and also for providing work space at considerable personal incon-
venience.
Without the excellent cooperation and dedication of the National Weather
Service, and the cooperative observers listed in Table 4, this experiment would
not have been possible. We extend our thanks to Bernard Spittler, Chief Substation
Management Branch, NWS, and his associates who were instrumental in setting up the
600 km sampling program and instructing the cooperative observers in the operation
of the sequential samplers.
We wish to express our appreciation to Dr. Jeremy Hales, Battelle Pacific
Northwest Laboratories, for his cooperation in providing a DC-3 aircraft and
knowledgeable crew for airborne tracer sampling missions.
53
-------
We wish to thank Paul Guthals and his colleagues at the Los Alamos Scientific
Laboratory for participating in our 600 km experiment and providing heavy methane
and SF5 analyses for comparison with perfluorocarbon measurements.
Special thanks are extended to Col. Van Louven, Chief M/Sgt. Greening and
the members of the 6th Air Weather Squadron, USAF, for taking special rawinsondes
that provided data vital to the experiments.
8. REFERENCES
Gilian Instrument Corp. (1980): Brookhaven Atmospheric Tracer Sampler (BATS),
Operations/Parts Manual. Gilian Instrument Corp., 1275 Route 23, Wayne,
NJ 07470, 34 pp.
Heffter, J.L. (1980); Air Resources Laboratories Atmospheric Transport and
Dispersion Model (ARL-ATAD). NOAA Tech. Memo. ERL-ARL-80, 25 pp.
Lovelock, J.E. (1974): Improvements in the Experimental Methods for a Long-
Range Tracer Experiment. Unpublished report. Available at the Air
Resources Laboratories, NOAA, Silver Spring, MD 20910.
54
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
EPA-600/7-81-006
3. RECIPIENT'S ACCESSIOr+NO.
4. TITLE AND SUBTITLE
Demonstration of a Long-Range Tracer System Using
Perfluorocarbons - Final Report
5. REPORT DATE
January 1981
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Gilbert J. Ferber, Kosta Telegadas, Jerome L. Heffter,
C.Ray Dickson, Russell N. Dietz, Philip W. Krey
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
National Oceanic and Atmospheric Administration
Air Resources Laboratories
8060 13th Street
Silver Spring, MD 20910
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA-IAG-D5-0693
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Research & Development
Office of Energy, Minerals & Industry
Washinaton, D,C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
This project is part of the EPA-planned and coordinated Federal Interagency
Energy/Environment R&D Program.
16. ABSTRACT
Regional-scale tracer experiments are needed to validate atmospheric dispersion
aspects of air pollution models. The capability of a new system, using perfluoro-
carbon tracers (PFTs), for long-range dispersion experiments at reasonable cost,
was demonstrated in two experiments. Two PFTs were released simultaneously with
SFg and two heavy methanes.
The PFT system uses automatic sequential samplers and provides rapid, inexpensive
analyses down to .002 ppt. PFT concentrations were measured 600 km away, up to
three days after release. Performance of the PFT system was excellent and a very
consistent set of tracer data was obtained.
17.
(Circle One or More)
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Ecology
Environments
Earth Atmosphere ^
Environmental Engineering
Geography
Hydrology, Limnology
Biochemistry
Earth Hydrosphere
Combustion
Refining
Energy Conversion
Physical Chemistry
Materials Handling
Inorganic Chemistry
Organic Chemistry
Chemical Engineering
Atmospheric Tracers
Dispersion Experiments
6F 8A 8F
8H 10A 10B
7B 7C 13B
13. DISTRIBUTION STATEMENT
Release unlimited.
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
64
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
55
US GOVERNMENT PRINTING OFFICE'1981—777-002/1216 Region No 8
------- |