WATER POLLUTION CONTROL RESEARCH SERIES
                                  16020 GLY 05/71
     MONITORING  MERCURY VAPOR
        NEAR  POLLUTION SITES
U.S. ENVIRONMENTAL PROTECTION AGENCY

-------
WATER POLLUTION CONTROL RESEkRCH SERI
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of
pollution in our Nation’s waters. They provide a central
source of information on the research, development, and
demonstration activities in the ‘water research program of
the Environmental Protection Agency, through inhouse
research and. grants and contracts ‘with Federal, State, and
local agencies, research institutipns, and industrial
organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should. be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, D.C. 2O 6O.

-------
MONITORING MERCURY VAPOR NEAR  POLLUTION  SITES
                        by
       Environmental Measurements, Inc.
             215 Leidesdorff Street
       San Francisco,  California  94111
                     for the


          Office of Research and Monitoring
        ENVIRONMENTAL  PROTECTION AGENCY
               Grant No.  16020  GLY
                    May 1971
     For sale by the Superintendent of Documents, U.S. Government Printing Office
              Washington, D.C. 20402 - Price 70 cents

-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval, does not signify that the contents necessarily
reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recornrnenda—
tion for use.
ii

-------
ABSTRACT
Field and laboratory measurements were made to demonstrate
that mercury vapor in the aix near mercury-polluted water
or sediment can be detected using an extremely sensitive de-
tector, the Barringer Airborne Mercury Spectrometer.
Areas were visited where the presence of mercury was known
from fish, water, or sediment analyses; anomalous mercury
levels ranging from 50 to more than 20,000 nanograms per cubic
meter (ng/M 3 ) were detected. Ambient air contained from 0 to
50 ng/M 3 .
Anomalous concentrations of atomic mercury vapor in air may
be classified as natural or man-made. The largest anomaly
detected was natural, emanating from a steam vent at The
Geysers, California. The second largest was man-made, and
was measured downwind from a large chemical plant.
Laboratory studies demonstrated that the mercury spectrometer
is sensitive only to atomic mercury. By means of pyrolysis
or combustion, organic compounds could be converted to metal-
lic form and detected. To detect mercury pollution in water,
pyrolysis appears necessary to convert combined mercury to
the atomic state for measurement by rapid spectrophotometric
techniques.
This report was submitted in fulfillment of Project Number
16020 GLY under the partial sponsorship of the Water Quality
Office, Environmental Protection Agency.
iii

-------
CONTENTS
Section Page
I Conclusions 1
II Recommendations 3
III Introduction S
IV Ohjectives 7
V Mercury In Air 9
VI The Barringer Airborne Mercury Spectrometer 13
Theory of Design 13
Calibration 16
Interferences 18
VII Laboratory Experiments 19
Decomposition of Combined Mercury 19
Solar Radiation 19
Pyrolysis 20
Mercury Under Water 22
VIII Field Program 25
Measurements in Motion 25
Stationary Measurements 29
Data Presentation 30
Areas of Study 30
IX Measurement Results 33
Natural Anomalies, Ground-Based Surveys 34
Man-Made (Cultural) Anomalies, Ground-
Based Surveys 41
Helicopter Surveys 50
Boat Surveys 53
X Discussion 57
New Almaden 57
Clear Lake 58
San Francisco Bay 58
Natural Mercury Deposits 59
Cultural Sources of Water Contamination 61
Dispersion of Mercury Vapor 61
V

-------
Section Page
XI Acknowledgements 63
XII References 65
vi

-------
FIGURES
No. Page
1 Temperature Dependence of Saturated Mercury 9
2 Barringer Airborne Mercury Spectrometer (BAMS) 13
3 Simplified BANS System Block Diagram 14
4 Example of Mercury Record at Industrial Site 15
5 Typical Calibration Responses 17
6 Initial Pyrolyser Apparatus 20
7 Experimental Moveable Laboratory Pyrolyser 21
8 Mercury Vapor Dissemination Through Water 23
9 Vehicle Installation of BAMS 25
10 BANS on Boat 27
11 Near Surface Sampling Over Water 27
12 Helicopter Cockpit Installation of BANS Readout 28
13 Spectrometer, Bivalve Assembly and Power Con-
verter Stowed in Helicopter Luggage Compartment 28
14 Sample Air Intake With Baffle 29
15 Location Map of California Investigation Sites 33
16 New Almaden, California 35
17 Clear Lake, California; Buckingham Point 37
18 Clear Lake, California; Sulfur Bank Mine 38
19 Abbott Mine, California 39
20 The Geysers, California 40
21 Pittsburg, California 42
22 Richmond, California 44
23 Berkeley, California 46
vii

-------
No. Page
24 Oakiand/Eiueryville, California 48
24 Location During Stack Sampling 51
26 Helicopter Data, Chicago, Illinois 52
27 Boat Survey Routes 54
28 Richmond, California, Sewage Outfall 55
29 Peak Mercury Vapor Levels Near Natural Sources 60
30 Peak Mercury Vapor Levels Near Cultural Sources 62
viii

-------
TABLES
No. Page
1 Summary of Atmospheric Mercury Measurements 34
2 New Almaden, California 36
Measurei ent Locations and Data
3 Clear Lake and Abbott Mine, California 38
Measurement Locations and Data
4 The Geysers, California 41
Measurement Locations and Data
5 Pittsburg, California 43
Measurement Locations and Data
6 Richmond, California 45
Measurement Locations and Data
7 Berkeley, California 47
Measurement Locations and Data
8 Oakland-Emeryville, California 49
Measurement Locations and Data
ix

-------
SECTION I
CONCLUSIONS
1. The presence of mercury vapor in the atmosphere was meas-
ured with the Barringer Airborne Mercury Spectrometer (BANS).
Repeatable threshold sensitivity of two to five nanograms per
cubic meter (ng/M 3 ) was attained.
2. Mercury vapor was conveniently and rapidly measured in
the atmosphere from a variety of mobile platforms. Installa-
tions of the BANS were made in a compact automobile, a small
van, a helicopter, and on a cabin cruiser.
3. Surveys were conducted in the vicinity of natural sources
of mercury (mines, reported pollution sites, and geothermal
vents) and over cultural sources (outfalls of sewage treat-
ment plants, industrial sites, and waste disposal areas).
Urban ambient monitoring was undertaken and demonstrated di-
urnal variations of mercury in the atmosphere.
4. Ambient air levels ranged from the zero level to peaks ex-
ceeding 80 ng/M 3 in urban areas (downtown San Francisco).
The presence of anomalous peaks of mercury were narrowly con-
fined near the sources of the vapor. A level exceeding 28,000
ng/M 3 was located at a geothermal steam vent; levels near the
effluent of incineration smokestacks exceeded 50,000 ng/M 3 .
These high levels were uncommon; measured peaks were more fre-
quently a few thousand ng/M 3 .
5. The intent of the demonstration grant was to seek mer-
cury vapor’s presence to be used as a rapid means of surveying
for mercury water pollution sites. The results in this re-
gard were inconclusive for three reasons:
Measurements were made near reported sites of mercury
pollution; and no samples of water or bottom sediment
were gathered simultaneously. Such corroborative evi-
dence was not gathered because the reported sites were
not located with sufficient accuracy to allow efficient
sampling.
Originally, the program was designed to monitor during
hot and humid months in the vicinity of well-documented
mercury pollution sites in the Great Lakes. Delayed
funding precluded visitation of these sites (due to
Winter) and most measurements were undertaken in Califor-
nia.
The intensity of anomalies measured, which did not re-
late directly to water sites, was so much greater that
1

-------
these were investigated, as a tracing method to de-
tect sources, rather than to continue the search for
minute atomic mercury levels directly over water.
6. Mercury vapor was shown to pass easily through a water
overburden once equilibrium had been established.
7. Experiments were successful in showing that organic mer-
cury compounds could be decomposed to release free elemental
mercury. This was then directly measured with the spectrom-
eter. Because mercury pollution has been reported to be
present principally in organic form, and the vapor pressure
of this organic form is reported some 40,000 times greater
than the vapor pressure of elemental mercury, the ability
to precondition the air sample so that “total” mercury could
be measured was very encouraging.
In conclusion, the demonstration grant showed an ability to
measure mercury in the air and demonstrated a technique for
preconditioning this sample so that total mercury might be
measured in the same fashion.
2

-------
SECTION II
RECOMMENDAT IONS
This program was limited to the demonstration of an ability
to survey for atmospheric mercury pollution near water. Dur-
ing the course of the program, an evaluation of pyrolysis
techniques were undertaken to determine if total mercury
could be measured using the same rapid-response, high-sensi-
tivity instrumentation. However, it was not within the scope
of this program to do more than field demonstrate the tech-
niques developed.
The success of surveying for atmospheric mercury and the
ability to decompose its organic compounds leads to a recom-
mendation that areas that have been tested in California are
revisited using the pyrolyser to determine total mercury in
the air.
All of the measurements made during the course of this demon-
stration grant were accomplished in winter months. The de-
pendence of mercury vapor pressure on temperature suggests
that much greater volume of material would be available in
the summer months. It is recommended that measurements be
made under hot conditions and that mercury samples to taken
from the water and bottom sediments simultaneously while the
presence of mercury vapor in the air is monitored. Visitation
to sites in the Great Lakes’ area are still recommended, as
is a survey trip to investigate recent water sample measure-
ments made by the Geological Survey from the Merced River,
California.
Mercury plumes were measured emanating from industrial sites
and natural sources. It is recommended that a detailed sur-
vey of these plumes be undertaken to determine their dispo-
sition and fallout character in rural and urban areas. An-
cillary to this suggestion, a detailed survey in an area of
high relief would delineate the effect of topography on the
apparent channeling of mercury vapor to low-elevation reser-
voir collection areas.
The simple ability to measure mercury, much less total mercury,
provides a fast and efficient method of survey. It is recom-
mended that the technique be incorporated as a method to
catalog mercury’s storage or presence, at both natural and
cultural sites. A building library of data would help assess
the relative contribution of various sources of mercury pol-
lution.
3

-------
SECTION III
INTRODUCTION
Since the discovery of mercury in fish caught in Lake St. Clair,
Ontario In 1970 (1), a great deal of interest in the United
States has been centered on the presence of mercury in water
and the relationship of this mercury to industrial and natural
sources. A major problem in the study of mercury distribution
in the environment has been the difficulty of location and
chemical analysis of water and sediment samples.
Extremely sensitive instrumentation has been developed to de-
tect mercury associated with ore deposits (2), and this equip-
ment has been used in prospecting for ore deposits. Environ-
mental Measurements, Inc. of San Francisco made application
to the Federal Water Quality Administration for a grant to
demonstrate the application of this prospecting instrument,
the Barringer Airborne Mercury Spectrometer (BAMS), to locate
places of water pollution. It was proposed to use this in-
strument to detect mercury vapor in air over sites of contam-
ination.
Environmental Measurements, Inc. took delivery of the
Barringer Airborne Mercury Spectrometer on December 8, 1970.
A program of familiarization was first carried out. Subse-
quently, areas of known mercury pollution were visited; meas-
urements were made of the mercury vapor in the air in the
immediate vicinity of these sites. Natural and industrial
sources were selected and visited, and the measurements were
cataloged and displayed. In addition, a reconnaissance pro-
gram involved traversing near industrial areas where mercury
pollution had not previously been reported. This program
revealed at least two new areas of mercury contamination.
The following sections describe the details of the surveying
and environmental analysis carried out in this program.
5

-------
SECTION IV
OBJECTIVES
The objective of this program was to demonstrate the appli-
cability of a newly designed proprietary mercury monitor,
the Barringer Airborne Mercury Spectrometer, to the rapid
and efficient survey of mercury pollution in water. The pri-
mary purpose of this was to be able to locate polluted waters
by detecting the vapors given off from such waters.
It became clear that the technique could also be used to
detect the principal sources of the mercury which contri-
buted to the pollution in the water. The secondary ob-
jective of the program, therefore, became the definition of
specific sources of mercury pollution which could contribute
mercury to the water or sediment.
7

-------
SECTION V
C,,
(I )
L J
0
> .
• 306400.
• 336300
• 305600
• 005200
.034300
.304400
.304300
• 003600
• 303203
• 002300
.332430
.001600
• 301203
333803
• 030403
MERCURY IN AIR
Mercury is the only metal to exist in a liquid state at stan-
dard conditions of temperature and pressure. Because of its
toxicity and high vapor pressure, it presents special problems
in storage, usage, and disposal.
The vapor pressure of mercury is markedly temperature depen-
dent. Figure 1 illustrates the variation in vapor pressure
Figure 1
Temperature Dependence of Saturated Mercury
with temperature over the diurnal range commonly experienced
in North America (3). The graph illustrates that over a cold
night/warm day interval, the vapor pressure of mercury can
vary ten-fold. This property is extremely important in the
understanding of ambient levels of mercury detected in the
air anywhere.
TEJIPERATURE (°C)
4 3 12 16 20 24 28 32 36 40 44
9

-------
The vapor pressure of mercury is also reported to be depen-
dent on barometric pressure. McCarthy, Meuschke, Ficklin
and Learned (4) report work which indicates that as the bar-
ometric pressure drops, mercury vapor is released from the
ground. This property would influence background levels in
areas where mercury is widespread.
The customary unit of measure for concentration of mercury
vapor in the air is the nanogram per cubic meter (ng/M 3 ).
This is equivalent to parts per trillion by determining the
weight of a cubic meter of air at normal conditions of temper-
ature and air pressure. Since one cubic meter of air weighs
1205 grams, one ng/M 3 becomes 1.0 x 10-9/1.205 x or
approximately 1.0 x i012, a part per trillion.
Certain industries producing and using mercury are hazards
because of this high vapor pressure. The Socrates Mine in
Northern California is so rich in mercury that the mercury
vapors are above danger level, and the mine cannot be worked.
The felt-hat industry is notorious for its cases of mercury
poisoning.
In some chemical laboratories it is customary to cover mer-
cury with a thick layer of water in the same container to
seal off the poisonous mercury vapor. This technique has
become a standard rule for safety (5). In Section VII the
results of an experiment demonstrate that this rule is not
totally effective.
The high vapor pressure of mercury has led to its use in pros-
pecting not only for mercury but also for other metals with
which it naturally associates. Geologists suggested that if
an instrument could be developed to detect this mercury in
air, it could be used as an airborne prospecting tool. The
development of the BAMS was in response to this need; criti-
cal specifications were high sensitivity and rapid response.
Mercury vapor has been detected in ambient air near naturally
occurring mercury deposits, in rural areas of agricultural
activity, in urban areas and near certain specific industrial
developments (4) (6) (7).
The Oak Ridge paper (7) reviews the standards for permissible
mercury vapor in ambient air and under industrial conditions.
The variation is extreme, ranging from 50,000 ng/M 3 in the
United States to 1,000 ng/M 3 in Germany. This variation in-
dicates the general uncertainty which exists concerning the
importance of low levels of mercury vapor in the air.
The residence time of mercury vapor in air is not clearly
known. The studies described in this report may contribute
10

-------
to that knowledge. The high density of mercury (13.5955
g/ml at 0°C) suggests that it cannot be very long.
11

-------
SECTION VI
THE BARRINGER AIRBORNE MERCURY SPECTROMETER (BAMS)
Theory of Design
The mercury measurements on this project were all made using
the Barringer Airborne Mercury Spectrometer (Figure 2).
F
Figure 2
Barringer Airborne Mercury Spectrometer (BAMS)
This instrument is an atomic absorption spectrophotometer,
specifically designed and built to isolate the 2536.5-
angstrom emission and absorption spectra characteristic of
atomic mercury vapor. The equipment was originally developed
for exploration purposes and for laboratory soil sampling (2);
subsequent design improvements (8) led to the rapid second-
range response time and high nanogram-per-cubic-meter reso-
lution needed for airborne use.
A system block diagram is shown in Figure 3. The air to be
sampled is drawn through one side of a bivalve assembly into
the sampling chamber. By operator choice, the air may pass
freely or be moved through a palladium-chloride-saturated
filter to absorb its mercury content. Repeated calibrations
have demonstrated the efficiency of this filter’s absorption
to be in excess of 96 percent.
‘1
13

-------
BIVALVE ASSEIIBLY
AIR
[ XHAJST
Figure 3
Simplified BAMS System Block Diagram
Under ambient pressure, the sample passes through the op-
tical path of the spectrometer at about one-third cubic meter
per minute. The entrapped sample volume is 0.016 cubic meter
along a one-meter path. Mirrors allow light to pass along
this chamber several times between energy source and photo-
detector. An effective six-meter pathlength produces the
needed concentration-pathlength for high resolution dependent
upon the Beer-Lambert Law of Absorption:
I = I ea
0
WHERE = incident light intensity
a = absorption coefficient (as a function
of wave length)
c = concentration
1 = path length
A commercial mercury-neon lamp is used as an energy source (9);
it is operated at an elevated temperature to broaden the
emission spectrum, and visible light is filtered out. Se-
quentially, a saturated cell of mercury (Case A) or a narrow
OPTICAL ASS&II3LY
14

-------
band interference filter, allowing the passage of the
2 536.5-angstrom line (Case B), is placed in the optical path.
In Case A, the mercury cell absorbs all of the transmitted
light except the broadened line’s edges; any sample in the
chamber has no additional effect. In Case B, the chamber
sample absorbs the emission line. The difference in the en-
ergy transmitted in each case, and monitored by a photo-
detector, is in direct proportion to the concentration of
mercury present in the sampling chamber. This value is am-
plified and presented in the chart recorder (Figure 4).
_ tL - —
4* LL 1dL±1
ii -
- ___
-
I - -f H- -r- —-
______ tttc
44 L(
____ ___
- - - 4 r — - - I -
_____ - - -
i1t ElI
4 t f1
- i : ±tH I . f ’iI L-
-i - -t - L
- - - - —- • -+ - E:4LL.
÷ - : [ j -
_______ i -i -
4 4 4 [ 1i
t -t4
—
Figure 4
Example of Mercury Record at Industrial Site
15
I P -fIC C’.! N’TP IS
flItPF I 0 NFVti YflPK
= - 4_

-------
The lamp and mercury cell must be carefully held in thermal
control. This is accomplished by heaters controlling the
environment of the electro-optical end of the spectrometer
unit. Their control and all signal processing takes place
at an electronic console, separated by cable from the spec-
trometer. Only system test and monitor features exist; all
readings and scale changes are obtained on the separate chart
recorder (or on a separate digital voltmeter). In the heli-
copter installation, the chart display was used for data re-
duction and analysis, and a DVM was used by the pilot to
guide his positioning in the absence of visible plumes.
Calibration
Calibration is accomplished by injecting a known quantity of
mercury into the air intake. The amount is obtained from a
rubber-capped bottle, placed in a dewer for thermal stability,
which contains liquid mercury at a known temperature. A
cubic centimeter is drawn through the septum and injected by
syringe into the intake hoses. As this mercury cloud passes
through the chamber, the chart deflection is noted for cal-
ibration (Figure 5). (To give the feeling for the range pro-
vided by the vapor pressure of mercury, 1 cc is equivalent
to 1380 ng/M 3 at 78° F, but only 180 ng/M 3 at 38° F).
The approximate value for calibrations during these tests
was 10 nanograms per cubic meter for each millivolt. The
sensitivity of the BAMS varied, over the course of the pro-
gram, from 6 to 15 ng/M 3 per millivolt. This range, largely
thermally dependent, shifted with aging characteristics of
the lamp and with temperature of the air being drawn through.
Calibrations were carried out at the beginning, during each
hour or two, and at the end of each field day or measurement
trip. Calibration shifts were never abrupt, and the most
adjacent data were used for data reduction. The average
value of calibration adjacent to each data set was used for
analysis.
An alternate means of calibration was used in the laboratory.
Because it agreed with the injection “shot” method described
above, the faster procedure was used most frequently. This
slower alternative technique also used a syringe of mercury
vapor drawn through the septum. The vapor was then slowly
“leaked” into the intake hose at a rate definitely slower
than the system flow rate. An offset on the record, repre-
senting the calibration, would last during the input period.
One-half cubic centimeter of vapor was injected in 20 seconds
using a motorized syringe (to produce a constant stream of
vapor); the system flow rate was measured at one cubic meter
per 200 seconds. Therefore, the offset represents one-tenth
of the quantity of mercury present in the syringe (which is
16

-------
determined by the temperature of the mercury source vessel).
The palladium-chloride filter was used to determine a zero
presence of mercury. By switching the air through this cham-
ber, then back through the non-absorbing chamber, any minor
mercury present in the ambient air is measured.
f i EF
•tj- iJ H+- - Hi- 1 ± H - I
l-L -
ie \1 t }4t
I t 1 j
I i - -1 - - J -4--(Th -
L !1Z
- j i
]i ftL±Jt L I
- i [ f ±tT *T 4 ±L F T T TTT
t Ij -
± j t
I ih 111t I LI
1 1 _ FL Ii i L J_ -
- T_C 1 i:’ 4 :r
I L -. L i!ii H
4 f4 L 4 _
- ii - &
- —-I — -4_
1! L
- -l iT
__ __ _______
i:: { i - ----

•ON
Figure 5
Typical Calibration Responses
(8.4 ng/M 3 = 1 millivolt)
17

-------
System noise is reflected in the chart record as the ampli-
tude of the trace under essentially clean air conditions.
Under these conditions no deflection is noted in the record
when the air is passed through the zero filter compared to
when it goes directly into the spectrometer. Resolutions on
the order of 2 to 5 ng/M 3 were possible under a wide range
of environmental conditions.
Interferences
Some substances with absorption spectra in the ultraviolet
region increase system noise due to the inefficiency of op-
tical filters and the presence of additional emission lines
in the lamp discharge. A judicious selection of thermal
range allows the operator to adjust the system to minimize
potential interferences. This is possible because the char-
acter and amplitude of the lamp’s emission spectra is tem-
perature dependent (9). For example, at optimum setting,
sulfur dioxide absorption interference was measured to be
less than 0.02 millivolt for each part per million of S02.
This is the equivalent of about 0.2 ng/M 3 of mercury; it
could still be distinguished from mercury signals by use of
the zero filter (which absorbed the mercury).
During the course of the measurement program, sulfur dioxide
was the principle source of concern for interference. In
the vicinity of some sewage treatment plants, however, other
gases also were detected as interference, as were the direct
measurement of gasoline fumes or auto exhaust. Careful selec-
tion of appropriate thermal balance allowed a different
setting of the oven temperature (from S02) to reduce this
hydrocarbon interference to minor significance. The adjust-
ment is somewhat time consuming; therefore, on some occasions
the data were not used because of interferences. This hap-
pened, for instance, when the interference adjustment was
optimized for auto fumes and a cloud of sulfur dioxide was
encountered.
18

-------
SECTION VII
LABORATORY EXPERIMENTS
Decomposition of Combined Mercury
Mercury is present in other than free atomic state, and iner-
cury pollution is particularly present in organic combin-
ations, such as methyl mercury (CH3Hg ) and dimethyl mercury
[ (CH 3 ) Hg]. These compounds do not report molecular absorp-
tion in the 2537-angstrom region. Tests to determine the
practicality of decomposing these compounds by heat (pyrolysis)
or light radiation were in progress at the Barringer Research
laboratory in Toronto, Ontario. EMI participated in these
experiments to assess the feasibility of preconditioning the
atmospheric sample to allow measurement of total mercury.
The BAI”IS would, thus, measure any atomic mercury present in-
itially plus any amounts converted.
Organic mercury compounds reportedly have very much higher
vapor pressures (10). This characteristic, if detection was
convenient, could provide a large quantity of traceable mater-
ial directly related to the most toxic forms of mercury poi-
lution.
Laboratory tests were successful, but no field measurements
were made within the limited scope of this grant.
Solar Radiation
A sample of dimethyl mercury was placed in a quartz cylinder
(about 2 inches by 4 inches long) and exposed to direct
sunlight for periods up to two hours. No detectable quantity
of mercury was measured in portions drawn off this enclosure.
(Daylight was a winter sun around noon in Toronto).
A 500-watt ultraviolet lamp was directed on a similar sample
of dimethyl mercury. Mercury was detected within one minute.
(The amounts were not calibrated with care because the mere
existence of the atomic form was sought). In two minutes the
quantity had tripled. In 2.5 minutes, twenty-five percent
more atomic mercury was present, showing a linear response
with time. The strong irradiation did decompose the compound,
but the time required appeared to prohibit this approach for
rapid mobile surveys.
Another approach was then considered, the use of direct
heat.
19

-------
PyrolysIs
Experiments to decompose organic mercury compounds by heat
were in progress at the Barringer laboratory. The initial
apparatus (Figure 6) was not acceptable because of its re-
quirement for large power consumption. EMI cooperated in
the building of a small apparatus, a prototype mobile pyro-
lyser, used successfully to convert several combined mer-
cury compounds.
The apparatus (Figure 7) consists of a vitreous cylinder,
about four feet long, wrapped in insulation. Nichrome wire
was coiled and wound within, and ceramic beads filled the
void. A thermalcouple placed at one end monitored the in-
ternal temperature. A variac controlled the current through
Figure 6
Initial Pyrolyser Apparatus
20

-------
the wire. This assembledge was used to draw vapor through
and directly into the spectrometer. A series of empirical
tests were conducted using this array.
Figure 7
Experimental Moveable Laboratory Pyrolyser
The intent was to assess if decomposition occurred, not to
set forth the boundary coiTdlitions of the phenomena. It was
felt that much more care would be needed to do the latter.
Temperatures ranging to 1000°C, various flow rates, and se-
lected mercuric compounds were used.
Syringe quantities of 1 cc were injected into the pyrolyser
at increasing temperature plateaus. Below a few hundred
degrees, only modest amounts of elemental mercury were ob-
served. At a point, the breakdown became obvious and notice-
ably greater amounts of mercury were observed. Increases in
the quantity of mercury observed were on the order of 10 to
1 for methyl mercury and 15 to 1 for mercurous cloride.
Dimethyl mercury produced dramatic results. One twentieth
of the sample (0.05 cc) of the other vapors produced off-scale
results. Large amounts of atomic mercury were obviously present.
21
)
I ____ —

-------
There is no question that these evaluation experiments de-
serve more thorough analysis. They appear to demonstrate a
convincing means of monitoring “total” mercury.
Mercury Under Water
The question asked was if mercuryvapor could pass through
water. A simple experiment was conducted to demonstrate this
and to determine the extent to which metallic mercury vapor
emanating from 1 cc of liquid mercury can penetrate a column
of water at 62° F.
One cubic centimeter of metallic mercury (liquid) was placed
in a one liter graduated cylinder. Using the BAMS, the probe
was located directly over the mouth of the cylinder to meas-
ure the amount of mercury vapor present with 1) no water in
the cylinder, 2) immediately after 500 ml of water (at 62° F)
was added to the cylinder, and 3) with a total of one liter
of water (at 62° F) in the cylinder. Using one liter of
water, the water was stirred vigorously to note any changes
in the amount of vapor being measured.
In the second phase of the experiment, the cylinder containing
the mercury was stored and covered with one liter of water
to allow the system to reach equilibrium. It was then nieas-
ured as before (first without stirring and then with vigorous
stirring).
A tabular presentation of the experimental results showing
the condition and amount of mercury detected at the cylinder
mouth follows (see also Figure 8):
No Water in Cylinder
7,780 ng/M 3
6,500
500 ml Water over Mercury
1,430 ng/M 3
2,275
3,250
2,275
1,755
1,983
One liter Water over Mercury
65 ng/M 3
65
65
22

-------
One liter Water over Mercury; Fully Equilibrated
759 ng/M 3
635
8,832 (shake cylinder)
8200
7600
7000
6600
6000
5Lffl0.
4800
4200
3600
= 3000
2400
1800
1200
600
0 500 1000
ML H 2 0 IN 1 LITER CYLINDER
Figure 8
Mercury Vapor Dissemination Through Water
Metallic mercury vapor emanating from one cubic centimeter
mercury passes through 500 ml of water rather readily at
room temperature. Stirring increases the amount of vapor
emanating from the water significantly.
Covered by one liter of water, a longer period of time is re-
quired before significant levels of mercury are given off.
However, once equilibrium is reached, fairly high levels of
mercury are released.
Meteorological conditions would greatly effect this phenomena
in the nature state. Wind would blow the vapor, surely, but
the churning water of a stream or wave action would also
carry the vapor down stream or to shore.
23

-------
SECTION VIII
THE FIELD PROGRAM
Specific measurement techniques were established to achieve
the objectives of this program. The principal method of
measurement was with the BAMS installed in a mobile labora-
tory, a Volkswagon microbus. Data were gathered both in
“traversing” mode and in “stationary” mode. To track down
suspected mercury polluted waters, the instrument was trans-
ferred to a boat. To study the mercury emanations from
specific industrial sources, the instrument was installed in
a helicopter. Each of these modes required special operating
power and different methods of air in-take and measurement.
Measurements In Motion
For ground-level measurements, the instrument was installed
either in a small car (Figure 9) or in the mobile laboratory.
Power was supplied by a 60-cycle, 110-volt gasoline generator.
The air intake was extended through a window and forward of
the vehicle. The problem of a Venturi effect at the sampling
Figure 9
Vehicle Installation of BAMS
25

-------
intake. This was effective at speeds below 50 miles per
hour. Hence, while traversing, a speed of 30 miles per hour
or less was generally used to provide a margin of reliability.
Because of the sensitivity of the instrumentation to temper-
ature variations, the oven and the electronic console were
both packed in insulating material. This effectively controlled
sudden changes in temperature due to direct sunlight or erratic
breezes blowing in through the windows. Both oven and filter
were mounted on independent spring suspensions. This protected
them from vibrations of the vehicle and reduced the associated
noise in the data.
In reconnaissance, mercury was measured while the instrument
was driven along selected streets and highways. Significant
variations in mercury level were immediately apparent if they
were encountered. By marking locations at frequent intervals
on the chart records, it was possible to make maps showing
the location of mercury vapor anomalies as they occurred.
Once a mercury anomaly was detected, the direction of wind
was noted. The mercury was traced upwind to its source
whenever possible. This was carried out by traversing at
right angles to the wind across the mercury anomaly very
slowly to define the lateral boundaries of the plume. As each
profile was completed, the traverse line was offset by a
block or two in the upwind direction, and another traverse
profile was produced. Eventually, a traverse would detect no
mercury; it could be concluded that the source of the mercury
vapor was downwind of the last profile.
The waterborne surveys were carried out in cooperation with
the California Department of Water Resources, using their 30’
motor cruiser, the Blue Angel . The equipment was installed
in the forward cabin (Figure 10). Since no appropriate power
was available on the vessel, the 60-cycle, 110-volt gasoline
generator was installed at the stern. The instrument intake
tube was extended through the forward hatch. In the sampling
of specific effluent out-falls a boat hook was used to extend
the intake hose to a few inches above the water (Figure 11).
To measure mercury vapor from the air, the instrument was in-
stalled in the luggage compartment of a Jet Ranger Helicopter
(Figures 12 and 13). Power was obtained from the 24-volt sys-
tem in the aircraft. The air intake hose and baffle were ex-
tended outside through a window (Figure 14). The signal was
recorded on the stripchart recorder and displayed on a digital
voltmeter mounted adjacent to the pilot’s instruments. This
digital display permitted the helicopter pilot to monitor the
mercury level in the air and, hence, to monitor the passage of
the aircraft through an otherwise invisible plume. Traverses
were run back and forth through the plume to detect and map
the downwind anomalous mercury vapor.
26

-------
- .-
__ - . - —.--- --
- - __ —.
________________ - - . —— —-
-
- —
— — — w -- — --
— - - -

— -. - . -. -
__ - -=-
- -
rØ ” - . - - --
Fig. 11. Near Surface Sampling
4
I
I
-In ,. f I T
N
Fig. 10. BAMS-on Boat
- -
r
Over Water
27

-------
•: *
V -
r p
V
Fiç . 12. He1ico ter CockDit Installation of BAMS Readout
Fig. 13. Spectrometer, Bivalve Assembly and Power Converter
Stowed in Helicopter Luggage Compartment
•; : 4L
28

-------
- ———-- - -—

Figure 14
Sample Air Intake with Baffle
Stationary Measurements
To detect levels of mercury vapor between zero and SU to iuu
ng/M 3 , the vehicle was stopped to measure for a period of a
few minutes. This procedure was adopted for both highway and
boat traverses, whenever such small levels were being sought.
Hovering with the helicopter was used to monitor plumes eman-
ating directly from stacks.
29
F -.
rJ-
L
-I
p.
f —
LL.y •‘ I
.. A
-__
• _I iI

-------
To obtain these small measurements, the air was sequentially
passed through the palladium-chloride filter and filter by-
pass. Thus, a series of offsets was detected which indicated
the amount of mercury which either went around the filter and
through the instrument or was absorbed as the air passed
through the filter.
The data obtained from this mode of operation consisted of
a series of measurements made at selected grid points. It
is necessary to interpolate between these points, rather
than read directly from the profile, as is the case in the moving
traverses.
Data Presentation
The data obtained in these measurement programs are presented
in tabular form listing the maximum signal detected at each
site. Bar graphs are also used for purposes of comparison.
Where levels of mercury were detected at selected locations,
the values are plotted on a map to show their relative position.
Areas of Study
The demonstration grant requested that measurements be made
during the summer months of 1970, when the ambient atmosphere
was warm and humid, in The Great Lakes Region. The grant was
received during the winter months; therefore, the site of
measurement was moved to California to get the warmest weather
possible. Nevertheless, the highest ambient temperature
during any day of measurement was in the 70’s (°F); this is
still cool.
Areas which were known or suspected to be sites of water or
air contamination by mercury were visited (4) (7) (11) (12).
However, in the course of the reconnaissance program, pre-
viously unreported anomalies were detected.
Before the program was initiated, for purposes of a separate
contract, measurements were made by helicopter in the Midwest,
around industrial smokestacks and power plants (13). Because
of the extreme mobility of the installation it was possible
to visit a large number of sites in a short time. The meas-
urements were made directly in the plumes and near stack
exits. (The readings are probably higher than would have
been detected on the ground due to the high concentration of
the effluent close to the stack mouth.)
Areas of natural mercury occurrence in California were visited.
At all of these sites traverses were made by driving around
on the county roads and continuously recording the mercury
levels and location.
30

-------
In the San Francisco Bay Area, the industries which were
known to be manufacturers or users of mercury were visited
first and traverses were conducted on roads near the sites.
These industries were located in Pittsburg, Millbrae, and
San Francisco. Several community sewage treatment plants in
the area were also visited. In the course of the reconnais-
sance program other anomalies were detected in Richmond,
Berkeley, and Emeryville. These new areas were not thoroughly
studied because this was not the intent of the demonstration.
Waterborne surveys were made during the course of regular
investigations carried out by the California Department of
Water Resources. In addition, industrial sites in Pittsburg,
Selby, Martinez, and Richmond, and sewage outfalls from
treatment plants and industries in the San Francisco Bay Area
were visited.
31

-------
SECTION IX
MEASUREMENT RESULTS
During the Course of the program, specific sites at which
anomalous mercury vapor could be detected in the air were
isolated (Figure 15). Many of these sites were visited more
c—
Abbot Mine
Clear Lake
The GQvsers
Pitt s burg
Ri c h man d
Berkeley
Oakland
Emeryville
San Fx ancisco
New :\lflLidefl
Figure 15
Location Map of California Investigation Sites
33

-------
than once to confirm the measurements. The anomalies are
classified as to origin: natural or man-made . Table 1
lists each site and presents the maximum measurement obtained.
TABLE 1
SUMMARY OF ATMOSPHERIC MERCURY MEASUREMENTS
llercurv Mercury
Background Level Peak Value
Site* D e* Time Wind ( ng/M 3 ) ( ng/M 3 ) Comments
Natural Sources
Abbott Mine 12 February PM
Cultural Sources
E
0
470
Low Population Density
E
N
0
0
150
200
Resort
Resort
Area,
Area,
Low
Low
Pop.
I’op.
censitv
Uensitv
N
200-800
28100
Rural
Resort
Area
Neglig.
0
1500
Rural
NW
S-iS
449
Rural
Berkeley 22 March
24 March
29 March
30 March
Oakland!
Emeryville 30 March
5 April
6 April
Pittsburg 11 February
12 February
25 February
22 March
1 April
- - 0
- - 0
-- 0
-- 0
Natural Anomalies, Ground-Based
800 Light Industrial-Residential
449 Light Industrial-Residential
154 Light Industrial-Residential
1050 Light Industrial-Residential
Light Industrial
Light Industrial
Light Industrial
Industrial Area
Industrial Area
On Boat
Residential
Industrial and Residential
278 Commercial Area
152 Commercial Area
100 Ambient Measurements,
Financial District
35 Ambient Measurements,
Financial District
1400 Residential, Near Primary School
2000 Residential, Near Primary School
Surveys
21 Light Industrial
Measurements were made near the New Almaden mining district,
Santa Clara County; around Clear Lake, Lake County; at The
Geysers, Sonoma County; and near various active and inactive
mercury mines while traversing to and from these points.
In two separate trips made to the New Almaden area, the peak
anomalous value of mercury was 1500 ng/M 3 , measured along the
roadways adjacent to the operations of the New Almaden Mine.
Similar anomalies were also detected at the bridge on County
Highway G8 across Alamitos Creek (Figure 16; Table 2); this
stream drains the Almaden Reservoir (now closed to fishing
because of mercury contamination) (14) and passes directly
through the town of New Almaden. A third peak was detected
northeast of the mine area away from all known or suspected
Clear Lake 12 February PM
2 April PM
The Geysers 15 February P11
New Almaden 4 January PM
24 March PM
AM
N
10
All
N
- -
PM
N
10
PM
N
0
PM
NW
0
196
PM
N
0
688
PM
N
0
110
PM
1
5 1)
770
AM
NW
0
1000
AM
N
0
0
PM
WNW
5
10
All-PM
N
0
4141
18 March PM
19 March AM
19 January AM-PM
- 26 Jan.
7 9 8 April AM-PM
San Francisco
(Quicksilver
Products)
San Francisco
Richmond
San Carlos
(Boat Survey)
*111 locations
29 March AM NW 0
PM N S
OApril AM -- 0
in California in 1971
34

-------
deposits. On March 24, a traverse was run downwind of the
Calero Reservoir (also now closed to fishing) to determine if
the mercury from the upwind mine area remained airborne at
this distance. The data illustrate the general trend of the
ambient level of mercury between the reservoir and the Santa
Clara Valley. It was not possible to detect either a high or
low downwind of the reservoir.
Clear Lake, north of San Francisco and the site of mercury
contamination reported by the California Department of Water
Resources (15), was visited twice. Under the different wind
conditions quite different patterns of airborne mercury dis-
tribution were detected. On February 12 the wind was from
35
Figure 16
New Almaden, California

-------
TABLE 2
NEW ALMADEN, CALIFORNIA
MEASUREMENT LOCATIONS E DATA
Location Date
4 January 1971 24 March 1971
A 828 83
765 145
132
132
125
B 1500 165
C 257
383
317
449
139
139
D 8
9
8
11
8
12
11
9
15
15
20
16
12
20
E 7
13
16
16
15
13
F 13
13
15
11
12
36

-------
the northeast and an anomaly was detected at Buckingham Point
(Figure 17; Table 3). The maxI um signal at BuckIngham Point
was about 150 ng/M 3 . On April 2, the Sulfur Bank and Bucking-
ham Point sites (Figures 17, 18; Table 3) were specifically
visited. Near the Sulfur Bank area there was some interfer-
ence from the S02 emanations. (By readjusting the instrument
this could be reduced; however, time did not permit). At
the Buckingham Point site no mercury was detected on the sec-
ond day. The wind was blowing from the west; this may indi-
cate that the first mercury detected at Buckingham Point was
in fact blowing from the Sulfur Bank area which lies to the
northwest.
is’and ru t sIand_ ‘ suiptiu” Bank point -.
— - ‘ -

L
- ‘ ; •: — 2
HOneII ’ ’°°”
Cove • ... /
-
Figure 17
Clear Lake, California; Buckingham Point
shag Rock
Weekend
37

-------
TABLE 3
CLEAR LAKE ABBOTT MINE, CALIFORNIA
MEASUREMENT LOCATIONS DATA
Location
A [ Figure
B (Figure
C (FIgure
17)
18)
18)
Date
1971 2 April
63
162
171
140
99
153
171
221
126
Clear Lake,
\ ir” ” - - I -I
- - -
? N’ ’ b F ‘
)2 — -’ L - :
Surank 2’/1
—----— 6 - - - -
1 Lake - - -
-
- _ _ - _\_____ (
‘— ‘]D c —
SuIpht
I , - -
Figure 18
California; Sulfur Bank Mine
12 February
150
1971
D (Figure 18)
E (Figure 18)
F (Figure 18)
G (Figure 19) 470
S __j
\\ - \__ _; -. :‘ -
- - - - - -
I
L
StubbS .’

RaU - - D _
% gnd ‘ 1/ (T -- T
-
.lphur Baf k Pok t
AXE
c4w
38

-------
In the course of traversing to Clear Lake, the Abbott Mine
was passed on State Highway 20 (Figure 19; Table 3). An
anomaly of approximately 470 ng/M 3 was detected even though
the mine was across a valley from the road.
c ’--. . 5
• ‘

-. .

• r: - \
4.. : :
Ma w y
S. ’ SP S 5 S_t _J I . • FISt .
. “ - : -• ‘j’f j t. - -
-°c - - - . . - .
--• - - A--- --’ -- -
I - -
..- S S :: .- . $ -
I -S
.S5S —4__ 3 2s) •
- . (S 5. -4 ‘ . I
-
- 21 ( , , .
LJ” -
— - —
,._ —
— • • • S_ 5\ (S •S -
- . ‘ G . - . ’: - ’ ‘- . ‘
¼ ,- C — — ‘ -
- j r -
- ‘- . S... - -
— -- - I l r 4: ‘ ‘ -‘ T —
Figure 19
Abbott Mine, California
The Geysers area, a geothermal region north of San Francisco
used for power generation, was visited on February 15 (Fig-
ure 20; Table 4). The plumes from the Steam Wells were specif-
ically observed, as were ambient levels near the health resort
and local mercury mines. The ambient levels ranged from 5 to
10 ng/M 3 west of the area, to around 400 ng/M 3 in the valley
near the resort, to approximately 2,000 ng/M 3 near the geo-
thermal steam power plant and downwind of specific vents. The
maximum signal was detected from a single vent on the north
side of the valley; it was greater than 28,000 ng/M 3 .
Mercury was measured for several days in downtown San Francisco
in January and February 1971. Daytime levels on the order
of 35 ng/M 3 were common; peaks reaching 80 ng/M 3 were frequent.
Nighttime levels ranged from zero to 10 to 15 ng/M 3 . These
39

-------
C S
- . . ‘ ‘ 9 . ;
\L. - -
o•y.J . ,,,
I -‘ - .
-. . A . ‘: - . •.. .- .
‘is, U
I ) - J
- . . .. - -, ‘ T ’ ,‘ - ‘ • \ 5 ‘ - . ‘ .“ . .
B ;tT ’T :t T ” ‘
J ’
- - - “k
e ‘ .•, ‘.. .• , ,-. s -
I — . ,: — 1
C ‘s.” ‘ “ , . z - - s ,• ’
\ For • -: • ‘ i’ ,.’.
—S.- Ad4 , C’ ‘ I ’
- ‘ , V’ Pd . ” S ) ‘ C ,’’ ‘ - -\
, , ,( ‘, “ ( - t
5. ):. .S_ . ,, .--.. -. -
• -‘-2’. Se.; .. . & % , * jW’
‘ - ‘ “ -: ‘• -
- I I ( - - Isr - —
r G4yw Pe, Ic ‘ s , . - - • • . . S.
• •. • , / - - ,• . . . . . , , , .3 -‘--.--
i- S . i
(1
5 5 - ‘ ‘: ‘ \j . : ._ :k: 5 T :.
j t 7 ) / - ?
S ?WQJ :
Figure 20
The Geysers, California
data exceed similar measurements, made by Williston (6) in
the suburbs of San Francisco, by a factor of about three.
The site of measurement was in the traffic congested finan-
cial district one block from the well known intersection of
California and Montgomery Streets. (Certainly this mercury
is the result of cultural development; it is mentioned here
for comparison).
40

-------
TABLE 4
THE GEYSERS, CALIFORNIA
MEASUREMENT LOCATIONS DATA
Location
15 February 1971
A 51
B 108
322
C 888
874
438
430
508
572
620
918
3510
3600
6200
11080
16800
28100
2350
1106
While traversing over rural highways to or from target areas,
mercury anomalies were detected. As an example, a single
anomaly of 50 ng/M 3 was detected on State Highway 24 near
Lafayette, Contra Costa County. A later traverse was run on
the north side of the highway around the Briones and San
Pablo Reservoirs (where the USGS reported mercury in the
water) (11) but no anomalies were detected. Thus, the
Highway 24 result was not explained. Other peaks detected
enroute similarly went unresolved.
Man—Made (Cultural) Anomalies, Ground-Based Surveys
Three separate trips were made to Pittsburg, Contra Costa
County (Figure 21; Table 5). A noticeable mercury signal
was detected downwind of a large chemical plant. The signal
varied in amplitude and in direction, dependent upon the wind,
and a maximum value of 4,100 ng/M 3 was measured on April 1.
The plume appeared to be directly downwind of the settling
basin which this company uses for its effluent. A local rep-
resentative of the company indicated that a chlor-alkali
plant was in operation directly upwind of the site of meas-
urement until April 1970. Unfortunately, due to the company’s
reluctance to permit a traverse northward to the river, it
was not possible to delimit the boundaries of the plume.
41

-------
- -
- 4..

- 4
8 R- 0 W N S
- - , ‘ - I -
LJNr L _r
S
-


Figure 21
Pittsburg, California
A boat traverse was made to the river side of this plant
along New York Slough on February 25. On this day the wind
was from the northeast. No mercury anomalies were detected.
42

-------
TABLE 5
PITTSBURG, CALIFORNIA
MEASUREMENT LOCATIONS DATA
Location Date
11 February 1971 12 February 1971 1 April 1971
A 361
410
754
1968
631
B 320 274 238
308 223 287
460 426
492
836
853
1517
C 340 850 2706
606 394 1066
390 637 2419
418 109 1927
830 1968
464 1681
1000 295
1460
1804
2911
4141
2378
A reconnaissance through the City of Richmond, Contra Costa
County, revealed a large plume emanating from the North
Richmond area (Figure 22; Table 6). The maximum signal in
this plume was detected in the playground of the Peres Public
School, just downwind of a pesticide and d’efoliant plant. It
was possible to trace this plume for a distance of about 1.5
miles across the central Richmond area. A change in the wind
affected this distribution pattern so that by late afternoon
it was blowing directly over the residential are of North
Richmond.
Another noticeable mercury anomaly was detected downwind of
a large bay-fill site near the Berkeley Yacht Harbor, Contra
43

-------
I : .

I . . . c •. We de
— . ,
n c ! f -n ..
W TI : R-ICHMOND
HI ;I
- .— ... l.
7t:
.. ,c j
L M.ki N

T j j \..
I ± .±JT E:JT=: :
- . / - S _Jp1
__LLE .
-
/ I __
Figure 2
Richmond, California
44
I D E.F,
I

-------
TABLE 6
RICHMOND, CALIFORNIA
MEASUREMENT LOCATIONS DATA
Location Date
29 March 1971
A 1913
1313
B 495
C 413
360
D 1500
B 810
F 128
150
630
1050
1050
775
G 48
33
H 45
30
I 90
66
J 128
K 63
L 38
M 60
N 35
Costa County (Figure 23, Table 7). The maximum signal de-
tected downwind, just west of Interstate 80, was 1,000 ng/M 3 .
The plume was also detected on 1-80, although the high speed
of traverse required for safety made it difficult to assess
45

-------
Ti.
YACHT HAlaQR f
Figure 23
Berkeley, California
k•::T
MUD H r - % T •
-i — - / t- - - - ( - -- - -
-‘ \ • . -! j ) - - -‘.
% I -
‘ - - ‘ ‘ ‘. \
;:
1• ‘ ‘ -‘ N- - A
Fleming I - V ,_ ‘ ‘ I - I I :
Point = ‘ ‘ ‘ V , I V
- - :-
- — —-‘ ‘— t_\ . -r•
‘ : kt \ (r .f -1
_____- I -t’ ‘ - L.-; v i J -i t
\::9 t c :\
\QA C
D - - ; t L. •‘- A
E 1
liG -
t ¶1.
- J -Hii k
• 4 - ;

‘ 4 \. ) c: ‘: ti
\
Vt’,-. ri -
• • -r- --
,i % . _. 1 ._j •-,‘ 4
- \tT
- .\1t • ‘., -v %)
, -:. t. - ‘1 i66
I
• . - - S A
‘ ‘ ‘ \
\ \ ( - •• 0 ••? -: I
- v: :: \ t-
46

-------
TABLE 7
BERKELEY, CALIFORNIA
MEASUREMENT LOCATIONS DATA
Location Date
22 March 1971 24 March 1971 26 March 1971 30 March 1971
A 800 449 154 917
492 1050
945
875
1050
875
B 210
196
175
175
203
C 140
D 35
28
56
98
91
E 28
28
the width of the plume. Immediately downwind of the freeway
to the east, levels dropped off sharply. No more than 50 to
100 nanograms per cubic meter were detected on the local
streets of the residential area.
A separate area of variable mercury concentration was detected
in Emeryville, Alameda County (Figure 24; Table 8). Visits
on three occasions over one week revealed anomalies. The max-
imum value was over 600 nanograms per cubic meter. Local in-
dustries include a paint plant, a chemical plant, and a scrap
iron recovery facility. The specific source was not identi-
fied, however.
Visits were made to many of the local sewage treatment plants
around the San Francisco Bay Area. Downwind of the East Bay
Municipal District Sewage Treatment Plant flare (see Figure 24,
sewage disposal) , values as high as 110 ng/M 3 were detected;
immediately upwind of the flare the ambient level was 0. No
47

-------
I
- - ( [ _ .
3
‘
. %‘ ,•t
U - C
-
• °R o To o - 2
- -- J1 / &
! •‘ : T - / p - - -. b— --
I
N:
Figure 24
Oakland/Emeryville, California
48

-------
TABLE 8
OAKLAND-EMERYVILLE, CALIFORNIA
MEASUREMENT LOCATIONS DATA
Location Date
30 March 1971 5 April 1971 6 April 1971
A 35
42
24
24
30
26
B 63 688 45
119 23
112 70
175 56
38 18
31
14
C 196 12
147
D 28
E 14
28
35
42
39
25
F 12
15
9
110
G 21
11
14
12
H 10
49

-------
mercury vapor was detected in the air immediately over the
effluent water leaving the filtration and treatment plant,
even though measurements were made within two inches of the
surface of the water. The plume from the plant incinerator
did contain minor amounts of mercury.
No metallic vapor was detected downwind of other sewage
treatment plants in the area. However, the outfall from a
plant in San Carlos, San Mateo County, contained some mer-
cury as reported under Boat Surveys, below.
The area immediately around the Quicksilver Products Company
(7) of San Francisco was surveyed on a number of occasions
because it was the only reported source of mercury in the
City. The results were variable. Maximum anomalies on the
order of 250 ng/M 3 were detected within one to two hundred
feet of the site of the plant itself. Elsewhere, low ambient
levels were measured. A single visit was made to the
Garrett-Callahan Company (7) in Millbrae, San Mateo County;
no metallic mercury was detected in the area.
Helicopter Surveys
As originally conceived, the demonstration grant was to in-
corporate aircraft usage if anomalies of sufficient size
were observed. Fortuitously, a separate project (16) incor-
porated the use of a Bell Jet Ranger Helicopter to conduct
measurements from a variety of combustion plumes. This ex-
perience was observed and is reviewed in lieu of carrying
out separate flights to measure the ground-level findings
reported herein. Operation characteristics were learned,
and it was judged unnecessary to repeat flights to demon-
strate the procedures.
Some results of the helicopter measurements have been reported
(13). Data were gathered from plumes emanating from municipal
incinerators, petroleum refineries, chemical production plants
and power generation facilities. Figure 25 shows a view from
the aircraft during plume measurements.
Peak concentrations of mercury from smokestacks in Missouri
and Illinois (13) generally were less than 100 ng/M 3 from
power plants, ranged to several hundred ng/M 3 near industrial
sites, and reached several thousand ng/M 3 from incinerators.
One peak value from an incinerator drove the instrument off-
scale to exceed 50,000 ng/M 3 .
An example of the data obtained by helicopter from an incin-
erator is provided to illustrate the typical operation (Fig-
ure 26). The aircraft flew into the visible plume and hovered.
If a signal was measured, it was recorded for a minute or so.
50

-------
Location During Stack Sampling
(The measurement time was limited by the noxious character of
the gas cloud and the ability of the crew to endure this).
The helicopter then left the plume, giving the operators fresh
Figure 25
51

-------
air to breathe, and returned in a moment with the mercury
absorbing filter switched in. If no (or a very much reduced)
signal was observed, the prior record was considered a valid
measurement. Occasionally interference signals were observed,
and the data were labelled unsuitable or questionable.
—
,c .
Figure 26
Helicopter Data, Chicago, Illinois
Sensitivity limits were investigated. Generally speaking,
installation in the aircraft did not degrade the quality of
the equipment operation. Certainly higher flight speeds,
above 30 to 40 mph, prevented data collection principally
because of the Venturi effect limiting the air intake. Most
data were gathered while hovering or moving at low forward
speeds.
The helicopter rotor does disturb the air in the vicinity;
this effect is not significant at the door where the intake
was fitted. For plume work, the air to be measured was drawn
I
C
52

-------
past the craft so this “fan” effect was not a limitation.
For near-ground, high resolution measurements the helicopter
would have to be fitted with an intake which extended forward
of the rotor. During a survey of this type a suitable forward
motion would be required to avoid distrubing the air to be
measured.
Our experience with the helicopter would indicate that an
external appendage would be feasible, and survey would be
possible near sites suitable for low level flight operations.
Obviously buildings, strung lines and cables, stacks, trees,
and high density population areas are all limits to this
method of rapid measurement.
The use of fixed-wing aircraft seems limited for the meas-
urement of atmospheric mercury associated with pollution,
principally because the clouds of vapor detected to date are
so localized. The density of the vapor also suggests that
the plumes generally remain low to the ground. While the re-
sponse of the BAMS is controlled principally by the air flow,
the few second delay becomes critical during fixed-wing air-
craft flights where five seconds at near-stall speeds of 80
miles per hour represent a ground-covered distance of 587
feet (179 meters). Such a delay in reporting the presence of
the invisible vapor makes it very difficult to survey and
locate at these speeds. The helicopter, on the other hand,
enables slow traverses and rapid changes of direction.
Boat Surveys
In cooperation with the California Department of Water Re-
sources, four days of surveying were carried out with the
equipment installed on board the power boat, Blue Angel . Meas-
urements were made over the water in the immediate vicinity
of the outfalls from sewage treatment plants around the Bay,
the outfall from specific industries in the North Bay and
Delta, and for the ambient mercury levels over the whole area
(Figure 27)
Two days of traversing were devoted to the North Bay and
Delta, centering on the Richmond and San Pablo sewer outfalls
(Figure 28), the coastline near the recently closed sulfide
ore smelter in Selby and, finally, the area near the pre-
viously surveyed chemical plant in Pittsburg. The wind was
blowing gently from the north on February 24; the day’s tem-
perature was around 60°F. On February 25 the wind was strong
from the north, and it was cold (about 40-50°F). No mercury
vapor was detected anywhere in the North Bay area.
Two days of surveying in April concentrated on Southern San
Francisco Bay. April 7 was devoted to the southern most
53

-------
sloughs. Coyote Creek, Santa Clara County (San Jose), was
penetrated to a railroad bridge with no significant mercury
anomalies detected. The results were the same elsewhere --
negative.
KEY
February 24, 1971
February 25, 1971
April 7, 1971
April 8, 1971
Figure 27
Boat Survey Routes
54

-------
Richmond, California, Sewage Outfall
On April 8 the ambient levels on the morning were very low,
o to S ng/M 3 ; they increased later in the day to about 10 to
15 ng/M 3 . Superimposed on this background was a single anomal)
detected directly over the oil smear of the San Carlos Sewage
Treatment Plant outfall. The maximum signal was approximately
50 ng/M 3 .
On this day, a traverse was also made in the Islais Channel,
San Francisco, in hopes of detecting mercury vapor near the
site of a reported mercury anomaly in water (11). No clear
anomaly above the 10 to 20 ng/M 3 ambient level was detected.
(On April 9, measurements were made from adjacent roadways;
no significant mercury was detected above the ambient level.)
Figure 28
55

-------
SECTION X
DISCUSSION
The nature of the mercury vapor presence detected during the
course of this program is quite varied. Near mercury-polluted
waters, the vapor levels were low; they seemed to depend on
the general background level as much as on emanations from
the water. Sewage and waste treatment plants may emit vapor
both from flares and incinerators. Near naturally-occurring
mercury deposits, mercury vapor measurements can be distinctly
high; the highest levels detected during the ground measure-
ment program were from natural sources. In the vicinity of
industrial users or producers of mercury, the vapor level can
also be high.
New Almaden
The reservoirs near the New Almaden, California, mercury mining
area have reported fish with mercury concentration greater
than FDA limit of 0.5 ppm. On May 7, 1971, signs warning
against consumption of fish taken from these reservoirs were
posted (14). Because of the documented presence of mercury
in these reservoirs, reported by the California Department
of Water Resources, this area became a prime target for mer-
cury vapor measurements.
The data that have been released indicate there is little
mercury in the water itself (11) (15) (17). Further, sedi-
ment samples are reported to yield inconclusive results. The
question of just how the mercury entered the reservoirs, and
then the fish, is still being considered.
The anomalous peaks in mercury vapor detected at the bridge
over Los Alamitos Creek in New Almaden suggest that vapor
was being carried along by an evening air drainage pattern.
The movement of air downhill along the gullies and valleys
off New Almaden mining area (Figure 16) would carry mercury
vapor emitted by the deposit over the Almaden reservoir
through the town of New Almaden, and even across the Calero
Reservoir.
The peak observed in the Alamitos Creek bed was not explained,
or duplicated in measurement made on a later day. However,
its presence was certain on the evening of January 4. Dif-
ferences in the wind account for the change. Whether this
peak represented mercury vapor from mercury already there was
not discernable.
The similar peak near the Guadalupe Mine on Guadalupe Road and
the small peak at the south end of Guadalupe Road are also
57

-------
apparently at the base of air drainage channels. Their
origin and extent were not apparent from these preliminary
measurements.
The regional measurements made downwind of New Almaden on
March. 4 did not indicate a clear anomaly downwind or immed-
iately upwind of the Calero Reservoir. The general broad
ambient mercury level of 5 to 15 ng/M 3 appears to represent
a pattern of dispersal froni the ore deposit area. However,
it might represent water or sediment laden with mercury which
were emitting the vapor.
Since the mercury values detected in the water and sediment
are reported to be mostly in dissolved form, it seems likely
that greater levels of combined mercury vapor would be ex-
pected in the air. Pyrolysis would aid in detecting the
presence of “total tt mercury anomalies in water and sediment
which this survey program did not detect.
Clear Lake
Fish taken from Clear Lake have also been found to be high
in mercury (15). So far, however, no recommendations against
consumption of Clear Lake fish have been issued.
The mercury vapor detected near Clear Lake varied on the two
days of measurement because of the differing wind patterns.
On February 12, the wind was from the east, and Buckingham
Point experienced mercury vapor levels of 150 ng/M 3 . On
April 2, the wind was from the west and no mercury was de-
tected on Buckingham Point. At Sulfur Bank, the vagueness
of the data may have been due to the wind which was very
strong, and may have produced a narrow, gusting, hard-to-
locate plume.
The origin of the mercury vapor detected at Buckingham Point
is in question. It may have been blown across the Lake from
Sulfur Bank, or it might reflect a mercury concentration in
the water near the Point or in the adjacent Little Borax Lake.
The great distance from Sulfur Bank would suggest that any
mercury vapor emitted then would fall out and be deposited
in the water before it reached Buckingham. The lack of vapor
when the wind was from the west suggests that the water just
east of the Point could be the source.
San Francisco Bay
The surveys run on San Francisco Bay were aimed at detecting
mercury vapor near the sites of water sampling programs which
had previously yielded mercury values. Specific targets were
Coyote Creek (Santa Clara County) and Islais Channel (San
58

-------
Francisco County). In addition, the outfalls from various
San Mateo County sewage plants were tested since industries
were reported to use these sewer systems for their effluent.
The lack of significant anomalies in elemental mercury vapor
suggest that any mercury contamination is in combined form.
However, the anomaly detected directly over the San Carlos
sewer outfall was a positive indicator of metallic mercury.
The measurements downwind of the East Bay Municipal Util-
ities District flare, where waste gases are burned, showed
mercury vapor with peaks of 100 ng/M 3 . The flare might have
reduced combined mercury to metallic form, which the BAMS
could then detect. This contrasts with levels measured at
incinerator stack exits with the helicopter. At the Stickney
Dryer in Chicago several hundred ng/M 3 were measured (13)
in garbage waste burners, much higher levels were found.
In the case of the flare and the incinerator, the plumes
mostly blew inland, and any mercury they carried was de-
posited on an adjacent freeway and nearby industrial/residen-
tial areas. When the wind is from the east, however, these
plumes extend over the Bay; the mercury would be deposited
in the water.
Natural Mercury Deposits
The highest peak value of mercury vapor monitored on the
ground during this program was identified at The Geysers,
where greater than 25,000 ng/M 3 were measured emanating from
a steam well (Figure 29). The steam from all the vents made
a clear picture of dispersal downwind to the east on Febru-
ary 15.
Since the rate of fallout of mercury from a plume is not
known,. the possibility that mercury from The Geysers steam
vent is blowing 20 miles north to Clear Lake exists. Sim-
ilarly the extent to which the now-closed Sulfur Bank Mine
has contributed to the mercury contamination of Clear Lake
can only be speculated. Mining operations 50 years ago
could have contributed to the mercury contamination which has
been discovered just recently.
The natural mercury mineralization at New ‘Almaden is appar-
ently responsible for most of the mercury contamination in
Santa Clara County. The vapor above natural concentrations
of mercury has been known for a number of years (4). The
mode of contamination of the water is not known, however.
Certainly the air movement is a carrier of the gaseous mer-
cury. The extent to which ground or surface water also is a
carrier has not been determined.
59

-------
1 CLEAR LAKE
28,030- 2 ABBOTT tuNES
I 3 STEAN VENT , THE GEYSERS
4 CORAL MINE
5 flEW ALIIADEN
3000.
2000-
iooo
123 5
SOURCE
Figure 29
Peak Mercury Vapor Levels Near Natural Sources
60

-------
Cultural Sources of Water Contamination
The effluent from known or suspected users of mercury has
been monitored with some vigor; however, no concern has been
felt for the vaporous emanations from these same industrial
users. Due to the high vapor pressure of mercury, it is
reasonable to expect some air contamination around these sites.
The magnitude of air contamination around these sites varies
significantly (Figure 30) and depends on the topography and
local meteorology. Large industrial installations using
chior-alkali processes would be expected to yield higher
levels than the small manufacturer of electrical components.
The largest mercury vapor level near a cultural source was
measured in Pittsburg near a large chemical plant which
closed down its chior-alki operation April 1970. This signal
was detected downwind of an area which contains their settling
pond and the approximate location of the now-closed cell. A
company representative indicated that mercury had been de-
tected in the ground near the cell site, and that this was
probably due to spillage at the time of operation. It was
not possible to discriminate the source area. However, if
the settling basin now contains mercury deposits at its base,
the water cover above it is not effectively sealing the vapor.
The second largest level was measured downwind of a large
pesticide complex in Richmond. It was possible to map the
distribution pattern of this plume for over a mile. During
the days of measurement, the wind was from the north and
northwest, and the plume was dispersed over the city. Of
particular note, the plume passed directly through school
facilities. On days when the wind was from the south and
southeast, the plume would be spread over San Pablo Bay. The
extent to which mercury fallout on land contributes to the
mercury level in the water by subsequent drainage is not known.
Other industrial areas also showed mercury vapor. The iden-
tification of each source, and the extent to which this form
of emission contributes to the mercury contamination of water
requires more measurement and analysis.
Dispersion of Mercury Vapor
The parameters which control the dispersion of a mercury
vapor plume still require definition. No plume from an in-
dustrial source was tracked for more than 1.5 miles. On the
other hand, a major well-travelled highway seemed to be in-
strumental in dispersing the major plume in Berkeley.
Natural plumes may be of greater extent, perhaps because of
less air turbulence in a rural environment coupled with their
greater volume. More work is required in this area.
61

-------
—
U-
w
I-
w
L)
—
L)
>-
U i
U-
I
3000
2000
1000
5 6
SOURCE
Figure 30
Peak Mercury Vapor Levels Near Cultural Sources
4000
1 SAN FRANCISCO
2 OAKLAND
3 EMERYVILLE
4 BERKELEY
5 PITTSB&JRG
6 RICI*IOHD
1 2 3 14
62

-------
SECTION XI
ACKNOWLEDGEMENTS
The support of the project by the Office of Research and
Development of the Environmental Protection Agency’s Water
Quality Office and the help provided by Dr. A. F. Forziati
and Dr. Louis Swabe is acknowledged with sincere thanks.
The long hours and encouragement provided by Dr. A. R.
Barringer, Dr. II. R. Clews, and George Motycka of Barringer
Research, who assisted in adapting their geochemical instru-
mentation to pollution measurements, is greatfully acknowl-
edged.
The cooperation of the Committee for Environmental Informa-
tion project members, Dr. William Vaughan and Mr. Steven
Fuller, and the skill and cheerful cooperation of our heli-
copter pilot, John R. Murphey of St. Louis Helicopter Air-
ways, is much appreciated.
EMI staff are appreciative of Drs. Ephram Kahn and Fred
Ottoboni of the California Department of Public Health and
the adhoc Statewide Coinmittee on Mercury Pollution.
Dr. Teng-Chung Wu of the California Department of Water Re-
sources was instrumental and very helpful in arranging our
boat surveys. The high competence and skill of William R.
Macke was appreciated particularly in his ability to locate
our measurements with precision in the water.
Originally considered to be a two-month project, the enthus-
iasm and encouragement of all cooperating parties suggested
that we carry out the work as described herein; no additional
funds were requested.
63

-------
SECTION XII
REFERENCES
1. Jervis, R.E., et al, Summary of Progress, Canadian
National Health Grant No. 605-7-510 (Trace Mercury
in Environmental Materials) for the period Sept.
1969-May 1970. Partial summarized in Chemical Engin-
eering News , Oct 5 (1970).
2. Barringer A.R. , “Interference-free Spectrometer for
High-sensitivity Mercury Analyses of Soils, Rocks and
Air,” Applied Earth Science , Institute of Mining and
Metallurgy, Volume 75, B120-Bl24 (1966).
3. Weast, Robert C. (Editor), Handbook of Chemistry and
Physics , 51, p. D-l45 (1970).
4. United States Department of the Interior, Geological
Survey, “Mercury in the Environment,” Geological Sur-
vey Professional Paper 713, (1970).
5. Anonymous, “Mercury and Mud,” Scientific American , 223,
No. 3, pp 82 and 86 (1970).
6. Williston, Samuel H., “Mercury in the Atmosphere,”
Journal of Geophysical Research , 73, No. 22, pp 7051-
7055 (1968).
7. Wallace, Robin A., et al, “Mercury in the Environment:
The Human Element,” Oak Ridge National Laboratory,
ORNL NSF-EP-1 (January 1971).
8. Barringer, A. R., “Method and Apparatus for Detecting
Traces of Substances,” Canadian Patent No. 789, 106
(July 2, 1968).
9. Kaye, Wilbur, “Emission Spectra of Argon, Helium,
Krypton, Neon and Xenon Pen-ray Lamps,” Pen-ray Rare
Gas Lamp Spectra , Ultra-violet Products, Inc. (1964).
10. Thompson, H.W. and J.W. Linnett, “The Vapor Pressure
and Association of Some Metallic and Non-metallic
Alkyls,” Transactions of the Faraday Society, 32,
p. 681, (1936).
65

-------
11. Durum, W. H., et al, “Reconnaissance of Selected Minor
Elements in Surface Waters of the United States,
October 1970,” United States Department of the Interior,
Geological Survey Circula 643 (1971).
12. Egenhoff, Elisabeth L., “De Argento Vivo,” California
Journal of Mines and Geology (Supplement) (1953).
13. Staff, “Mercury in the Air,” Environment , 13 No. 4
(1971).
14. Anonymous, “Santa Clara Warns of Fish Mercury Hazards,”
San Jose Mercury , May 1, 1970.
15. Kahn, Ephram, Personal Communication, 1971.
16. Helicopter-borne mercury measurements were carried
out in cooperation with the Committee for Environ-
mental Information, St. Louis, Missouri.
17. Averitt, Robert, Personal Communication, 1971.
U. S. GOV}R’.\IFST RI T1SU OFF CL 1972—4 4—486/2 .5
66

-------
Accession Number
W
Subject Field & Group
05A 05B SELECTED WATER RESOURCES ABSTRACTS
‘ INPUT TRANSACTION FORM
Organization
Environmental Measurements, Inc.
San Francisco, California
Title
MONITORING MERCURY VAPOR NEAR POLLUTION SITES
Author(s)
Jepsen, Anders F.
Langan, Lee
Project Designation
EPA, WQO Contract No. 16020 GLY
J
Citation
Descriptors (Starred First)
*Mercury, *Pollutant identification, *Analytical techniques,
*Surveys, Spectrophotometry, Water pollution sources, Prototype
tests, Air pollution effects, Distribution patterns, Measurement
25 Identifiers (Starred First)
*California, Barringer Research
_ j Abstract
Field and laboratory measurements were made to demonstrate that mer-
cury vapor in the air near mercury-polluted water or sediment can be de-
tected using an extremely sensitive detector, the Barringer Airborne
Mercury Spectrometer.
Areas were visited where the presence of mercury was known from fish,
water, or sediment analyses; anomalous mercury levels ranging from 50 to
more than 20,000 nanograms per cubic meter (ng/M 3 ) were detected. Ambient
air contained from 0 to 50 ng/M 3 .
Anomalous concentrations of atomic mercury vapor in air may be classi-
fied as natural or man-made. The largest anomaly detected was natural,
emanating from a steam vent at The Geysers, California. The second largest
was man-made, and was measured downwind from a large chemical plant.
Laboratory studies demonstrated that the mercury spectrometer is sen-
sitive only to atomic mercury. By means of pyrolysis or combustion, or-
ganic compounds could be converted to metallic form and detected. To de-
tect mercury pollution in water, pyrolysis appears necessary to convert
combined mercury to the atomic state for measurement by rapid spectro-
photometric techniques. (Langan-Environfliental Measurements)
WR tO2 (REV. JULY 969)
WR5I C
SEND. WITH COPY OF DOCUMENT. TO: WATEN RESOdRCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. D. C. 20240
* GPO: 1970—38S930

-------