"trace determinations" 930 Kinnear Road, Columbus, Ohio Phone: 614-299-9229
Mailing Address: P.O. Box 5093, Columbus, Ohio 43212
THE DEVELOPMENT OF THE GOLD AMALGAMATION SAMPLING AND ANALYTICAL PROCEDURE FOR
INVESTIGATION OF MERCURY IN STACK GASES
by
G. William Kalb
and
Charles Baldeck
for
The ENVIRONMENTAL PROTECTION AGENCY
Contract Number 68-02-03U1
June 8, 1972
-------
TABLE OF CONTENTS
Page
INTRODUCTION
1
ANALYTICAL PARAMETERS
s
1- Design of the Amalgamator
2- Absorption Cell Design
3- Amount and Type of Gold
4- The Shape and Peak Height of the Absorption
Curve as a Function of the Firing Temperature
of the Amalgam .
5- Air Flow Rate
ANALYTICAL PROCEDURE
STACK GAS SAMPLING
27
33
1- Preparation of the Amalgamators
2- Operation of the Sampling Train
FIELD ANALYSES
STACK SAMPLING RESULTS
37
41
1- Peak Shapes
2- Interferences
3- Gold Deterioration
4- Absorption Cell
S- Stack Turbulence
6- Sampling Period
7- Flow Rates
8- Permanganate vs. Gold Trains
9- Efficiency
10- Particulate Analyses
CONCLUSIONS
75
77
FUTURE INVESTIGATIOnS
APPENDIX I
81
I .
Mercury-Gold Amalgam Shipping Characteristics
APPENDIX II
83
Conversion Graph for Flow Rates
-------
ACKNO\m:DGUENTS
The authors wish to acknowledge their appreciation to the
Columbus and Southern Ohio Electric Company and the Ohio Edison
Company for the use of their facilities which made this study
possible.
,
-------
LIST OF FIGURES
Fi~e
1. Schematic Diagram of the Analytical Apparatus
2. Illustration of the 15 mm AmalgaJllator
3. illustration of the 3 Absorption Cells
. 4. Peak Shape as a Function of Firing Temperature and Flow Rate
5. Peak Shapes Obtained with the ~ Inch Cell
6. Peak Height as a Function of Firing Temrrature and Air Flow
Rate (15 rom Amalgamator, th. Inch Cell
7. Peak Height as a Function of Firing Temperature and Air Flow
Rate (15 rom Amalgamator, 3 Inch Cell) . . .
8. Peak Height as a Function of Firing Temperature and Air Flmr
Rate (25 mm Amalgamator, 3 Inch Cell) .
9. Peak Height as a Function of Firing Temperature and Air Flow
Rate (15 rom Amalgamator, ~ Inch Cell)
10. Illustration of Procedure Used to Determine Aeration Time of
Standard Solutions
ll. Peak Height as a Function of Flow Rate and Sample Size
(15 mm Amalgamator, 6 Inch Cell)
12. Peak Height as a Function of Flow Rate and Sample Size
(15 ram Amalgamator, 3 Inch Cell)
13. Peak Height as a Function of Flow Rate and Sample Size
(15. JUJI1 Amalgamator, ~ Inch Cell)
14. Analytical Curve Obtained with ~ Inch Cell and 15 11111
Amalgamator
15. Analytical Curve Obtained with 3 Inch Cell and 15 11111
Amalg8Jll8tor
16~ Analytical Curve Obtained with 3 Inch Cell and 25 mm
Amalgamator . .
17. Physical Arrangement and Location of Ports at Site 1
~
6
9
II
14
16
17
18
19
.
20
22
24
2S
.
26
30
31
32 I
39
-------
Figure
18. Illustration of Differences Observed in Peak Shape due to
Ama.l.gams.ting the Standard Fr;om the Bottom up or the top down .
19. Illustr~~ion of ~ high Prepeak due to Water Absorption
Resul ting from the Chemical Breakdown of Sulfuric Acid
during Firing
20. The Observed Mercury Concentration as a Function of SampJ.ing
Time.
21. The Observed Mercury Concentration as a Function of Sampling
Rate
22. The Percent Recovery of Mercury on the First Amalgamator
Appendix II Air Fl.ow Curve: Perkin-Elmer Flowroeter Units
Page
S8
61
67
68
72
83
, .
-------
LIST OF TABLES
Table
Page
1. Reactions and Oxidation Potentials
2. Minutes of Aeration Required at Various Flow Rates
2
21
3. Daily Log of Runs:
Parameters
Mercury Concentration and Control
42-
4. Daily Log of Runs:
Control Parameters
so
s. Results Obtained from KMn04-Au and Au-KMn04 Sampling
Trains
70
6. Particulate Analyses by Induction Furn.ace-Permanganate
Digestion
7. Results of the Effect of Storing the Amalgam at 100°C for
16 Hours in the Amalgamator
74
82
, .
-------
INTRODUCTION
Limited investigations of mercUry concentrations in coal have shown a
national range of 0.05 to 0.50 ppm with some regionalized exceptions.l,2,),4,S
This mercury, naturally occurring in coals, is volatilized during combustion
of the coal in fossil-fuel fired steam generating plants.
The present.
consumption of coal could result in the release of a significant quantity
of mercury to the atmosphere.
The mercury volatilized during the combustion stage is not entirely lost
to the atmosphere.
During the gas cooling stages in generating plants,
involving the economizer, air preheater, cyclones, considerable ductwork,
--and. the stack, volatilized mercury appears to condense onto the flyash.
This conclusion, based on a few limited studies,6,7,8 shows that an
average of 10% of the original mercury is recovered in the fiyash m th
undetectable amounts present in the bottom ash.
Because mercury is not
observed in the bottom ash it is assumed that all the mercury is volatil-
ized and the flyash concentrations result from recondensation.
This
theoretical conclusion has shown that we cannot directly e~rapolate coal
concentrations to atmospheric releases. To .~derstand the behavior of
mercury in stack gases it is necessary to determine a mass balance, showing
the amount of mercury in the original coal and the resultant concentrations
in the bottom ash, fly ash and stack gases.
This requires an analytical
method by which we can reliably measure mercury in stack gas concentrations.
,.
Present methods for determining mercury concentrations in stack gases are
1
-------
2
wet scrubber sampling trains using acidic-oxidizing solutions to absorb
the volatile mercury from the gas stream.
The adaptability of solution
sampling trains to mercury stack gas sampling is dependent upon the ability
of the solution to oxidize elemental mercury to the +2 valence state thus
reraoving it from the gas stream as it bubbles through the solution.
Acidic
permanganate and/or iodine mono chloride solutions have been utilized by the
Environmental Protection Agency, TV A and TraDet.
The reactions and oxida-
tion potentials are shown in Table 1. As can be observed from the oxidation
potentials, 502 from the stack gases will reduce both the Mn0r; and IC1.
Because of this, excess permanganate and iodine mono chloride must be used
and a constantly decreasing concentration of oxidant vill be observed.
Table 1. Reactj,ons and oxidation potentials.
502 + 2H20 ~ 504 + ~+
2Hg ~ 2Hg++ + 4e-
C1- + ~2 ~ IC1aq.+ e-
Mn ++ + 2H20 ~ Mn02 + Wi+
++ 4H20 ~ Mn04 +
}in '+ + OR
+
213-
-0.17
-0.92
-1.19
+
2e-
-1.23
+
Sa-
-1.$.1.
Deterioration of the oxidant has been observed with both acid KMn04 and ICl
sampling trains during sampling.
In addition, acidic KMn04 undergoes
autodecomposition. After a representative sample is obtained in the
sampling train the mercury concentration is detemined by reducing the
excess oxidant and the mercury, and aerating the solution, carrying the
-------
J
mercury through an inclosed quartz tube in an atomic absorption spectra-
photometer.
Hydroxylamine hydrochl.oride and stannous chloride are used to
reduce the acidic K}!n04 samples and hydroxyl81Tline sulfate is used to
reduce the ICI samples.
Because of the potential condensation of mercury in the fly ash it is desirable
to simultaneously collect a particulate samp~e.
This requires the use of
an isokinetic sampling train for an extended sampling period.
Because of
this and the requirement that an extended sampling period is required for
a representative sample it is desirable to ha';e at least a ~ hour sampling
period.
Because of the S02 concentrations in the gas stream this is not
feasible with wet oxidizing solution trains.
It is the objective of this investigation to study other methods of sampling
for mercury in. gas streams containing high SOi concentrations.
The ultimate
achievement of this study is to develop a sampJ.ing procedure adaptable to
copper-zinc smelters as well as fossil-fuel fired plants.
Smelters roast
sulfide ores in a partially reducing environment releasing large quantities
. of S02 which mayor may not be recovered in an acid plant.
trations may exceed 6% in these plants.
The S02 concen-
Silver and gold amalgamation reactions with mercury chemically favor the
complete recovery of mercury in a gas streplIl. Silver reacts with S02
forming a sulfide surface layer (common tarnish) eliminating its application
to stack gases.
One of several possible reactions for this would be:
502 + 6Ag
) Ag2S
+ 2Ag20
In addition gold forms a strong amalgam with mercury resulting in the use
of it in this study to quantitatively adsorb mercury froM a gas stream.
-------
4
Mercury amalgamation accompanied by a flameless atomic absorption mercury
procedure is utilized in this investigation.
-------
ANALITrCAL PARAm:rERS
Go~d amalgamation has been used as a concentration and interference elim-
ination step in the flaroeless atomic absorption procedure for determining
mercury in water, fish, coal, etc.. The digested sample is reduced and the
elemental mercury is aerated onto the gold foil, where it forms a stable
amalgam. Since other compounds 1IJh1ch would absorb at the mercury resonance
wavelength of 254 m\.lo are not collected onto the gold, they are vented from
the system during the amalgmnation step. The amalgam is then fired in an
induction furnace, revolatilizing the mercury which is carried by an air
stream through a quartz cell in the optical path of an atomic absorption
spectrophotomete!". The instantaneous heating of the .gold in the induction
field produces a -very rapid release- of--mercurytrom-the gold and a sharp
peak is obtained on the recorder readout of the spectrophotometer with the
peak height proportional to the mercury concentration.
In the modification of this procedure for stack gas analyses, the amalgam
is obtained by passing the gas through a quantity of gold supported in a
tube (called an amaigamator) placed in a standard EPA samplin~ train.
When sufficient gas has been drawn through the amalgamator, it is p~aced in
the induction furnace and fired as described above. A schematic diagram
ot the anal.yt,ical. apparatus utilii-ed in this procedure is shown in Figure 1.
Immediately after' analyzing the amalgam obtained from the sampling train,
standards are run on the same amalgamator by placing aliquots of a standard
mercury solution in the aeration cell shown in the figure. A 100 or 200 ppb
5
-------
~
Figure 1.
6
'I
Schematic Diagram of the Analytical Apparatus
-------
7
mercury standard is prepared fresh daily from a 1 ppm stock solution.
The stock and workin~ standard solutions are fixed with 5% concentrated
nitric acid to retard deterioration. The 1 ppm stock solution is pre-
, .
pared weekly from a 1000 ppm stock mercury solution prepared from mercuric
chloride and containing 5% concentrated nitric acid.
After the standard
is placed in the aeration cell it is reduced by the addition of 1 ml. of
2~ SnCl2 in 50% HCl with a syringe inserted through the septum cover. The
SnCl2 reduces the oxidized mercury to elemental mercury which is readily
volatilized by an air stream that is bubbled through the solution. The
.' . .'.
volatilized mercury is carried by the air stream through the gold foil where
it is amalgamated. The amalgam prepared in this manner from standards is
then fired in the sameJl18JlIler as the sample collected in the gas stream.
Negligible blanks are obtained during the running of standards and since
there are no chemicals used in the sampling procedure the stack samples do
not have' a background'level requiring a blank determination. This is 'a
major advantage over the present wet chemical sampling trains.
There are several parameters that must be optimized in the analytical
procedure.
The parameters studied in this investigation include:
(l) design of the amalgamator,
(2) absorption cell design,
(3) amount and type of gold,
{4}
the shape and peak height of the absorption
curve as a function of the firing temperature
of the 8r.1algam,
(5) the air flow rate for aerating the standard
solutions and for transporting the volatilized
mercur,y from the amalgamator to the quartz cell.
The air flow rates are given in the units used on
the Perkin-Elmer atomic absorption spectrophoto~
meter flowmeter: for conversion to ml/min. refer
to Appendix II.
-------
8
l- Design of the Amalgamator
The optimized amalgamator designed to support from 5 to 20 grams of gold
is illustrated in Figure 2. I~ is adaptable to both the standard EPA iso-
kinetic sampling train and the induction furnace. This design permits
inserting the amalgamator into the sampling train in place of an impinger
during gas sampling and then inserting it into the induction furnace for
firing the amalgam.
This eliminates both (l) the necessity of having to
modify the existing standard sampling train and (2) repacking the gold
amalgam in a different holder for firing in the induction furnace. Two
amalgamators we~e investigatedJ (l) a l5 mm design, shown in the illustra-
tion, and (2) a 25 mm design containing a 25 mm bulge about ~~ high where
the gold is located, and a perforated glass plate to support the gold.
The large amalg.!l.mator effectively decreases the flow rate of the gas stream
through the supported gold. This decreased flow rat'e was found to be
unnecessary under the conditions investigated. The large amalgamator was
found to heat unevenly in the induction field producing irregularly shaped
peaks.
In addition, local overheating of the gold would result in fusing some'
of the gold chips together and the amalgamator would break from thermal
stress around the perforated plate. In order to obtain reasonably symmetrical
peaks using the large amalgamator a rather low flow rate (to minimize
turbulence of the air flow through the perforations in the plate) and a low
firing temperature (to avoid fusing the gold and breaking the glass) were
necessary. At low flow rates, slight random variations in the flow can
cause a comparatively large variation in sensitivity, especially with large
sampJ.es (see Figures 10, il, 12). This makes it undesirable to work at
.
. .
-------
9
o
,
Figure 2. Illustration of the IS 111m Amalgamator
-------
10
the low flow rates, and was one of the reasons why the large amalgamator
was rejected in favor of the straight type.
Both the illustrated three finger support for the gold and a perforated
plate support were investigated.
The perforated plate, when rigidly affixed
to the sides of the tube, tends to break with extensive sudden heating in
the induction field.
Addi tional problems were observed with the flow
patterns developed by the disk.
The three finger support did not experience
this extensive breakage. When the three finger support is used, a 2 gram
gold wire plug supports the gold foil chips.
The optimized amalgamator
was constructed of pyrex.
Quartz would be a better construction material
but the cost for four ground glass quartz' joints or one joint and a graded
seal becomes more expensive than the breakage.
The 15 rom three finger pyrex
amalgamators ha~-e been used extensively (several hundred firings) without
breakage but they do show glass strain when investigated optically.
Addi-
tional study may show that a sintered glass support would be acceptable.
This has not been investigated at this time.
2- Absorption cell design
Three absorption cells were designed as illustrated in Figure 3.
The
first two cells were B1' and )" long with quartz endp1ates.
The third
cell was ~t wide with two quartz windows.
The first two cells are high
volume peak height cells ie. more and more mercury enters the ceil during
firing, reaching a maximum concentration (peak height on recorder strip
chart) corresponding to measuring all the mercury at once.
The third ceil
approximates an integrating cell, theoretically measuring each increment
-------
q
d
3"
.
Figure 3.
.~
6 1/2/1
1/2/1
Illustration of the 3 Absorption Cells
11
-------
12
of mercury concentration as it passes through the cell.
An ideal peak
height cell theoretically measures all the mercury at one time, which would,
if a large enough cell were us~d, result in a plateau at the greatest ab-
sorption corresponding to the time involved before the cell starts eluting.
itself.
The small cell measures concentration as a function of time, which
is extremely dependent on all the above parameters including the exact
alignment of the amalgamator in the induction furnace.
It does not have a
sufficient volume to average out small non-uniformities in the release of
mercury, and the curves obtained vi th it show doublets, triplets and pre-
peaks.
Due to the impossibility of maintaining all of these parameters
constant, this cell is not suitable for p~ak height measurement.
Instead
the curve should be integrated which requires a recorder and a spectre-
._photometer1:1hich reads out in absorbance.
The integrating cell does decreE.se
sensitivity considerably, permitting determinations to be made at higher
mercury levels ie. longer stack sampling periods.
Since the available
instrumentation was not suitable for integration of the peaks obtained with
the ~ inch cell, the 3 inch peak height cell was utilized in most of the
studies.
This limited the amount of mercury that could be measured in the
linear section of the analytical curve (approximately 2.0 ~g.).
3- Amount and Type of Gold
Compressed gold wire and gold chips, both 0.007" thick were studied.
The
gold chips, between 1/6" and 1/16" square, packed more evenly in the
amalgamator resulting in a more uniform temperature in the induction field,
giving better reproducibility.
Ten grams of gold chips were found to be
-------
13
sufficient in all the sampling studies.
In tests with standards of up to
10 ~g. of mercury, 10 grams of gold quantitatively removed the mercury from
the air stream without leakage! which, at the observed stack concentrations,
is sufficient for one hour runs.
4- The shape and peak height of the absorption curve as a function of the
firing temperature of the amalgam .
Because the temperature achieved by the gold is induced, it is extremely
difficul t to measure.
Thermocouples, etc., are self inducing in the
induction field and optical pyrometers are only usable at the higher
temperatures.
Because of this the percent variac setting of the induction
furnace is recorded instead of the temperature.
The temperatures obtained
in the induction field are a function of the density of packing of the gold,
the amount of gold, the cross-sectional area of the gold in the amalgamator,
and the air flow rate through the amalgamator.
In order to eliminate any
variations in heating due to individual differences in the amalgamators, the
firing temperature studies for each absorption" cell length were performed
using one BlIlalg8lllator with the standard packing: . a 2.00 gram gold wire plug
supporting 10.00 grams of gold chips.
In order to examine the relationsldp
between air flow and firing temperature the study was performed at several air
flow rates.
This was done for each of the absorption cells investigated.
Figure 4 shows the typical effect on peak shape as the firing temperature is
varied at both the high and low flow settings, for the 3 inch cell.
Similar results are .observed with the ~inch cell. When the ~ inch cell is
used, the combination of the low sensitivity (large concentration of mercury
required to be on scale) and the short path length causes pre-peaks or
-------
Variac
F low - 5
Variac
Flow - 2
100%
100%
14
Va riac
70%
F I ow - 5
Variac
70%
Flow - 2
, .
Figure 4.
Peak Sha.pe as a Function of Firing Temperature and F lev Rate
-------
15
doublets to appear, l~ting the usefullness of this cell with the present
instrumentation.
At 10" firing temperatures, the pre-peak is higher than
the main peak; as the firing temperature is raised the main peak becomes
higher. Figure 5 shows typical peaks obtained with the ~ inch cell as the
firing temperature is varied.
In general, either a decrease in firing temperature or an increase in air
flow rate makes the peak lower and broader.
The combination of high air
flow rate and low firing temperature results in the peaks tending to trail.
The change in peak height observed as the firing temperature is varied at
several flow rs,tes is sUJ11JJ18.rized in Figures 6, 7, 8 and 9.
In general, the
optimum combinE-tion of firing temperature and air flow rate is one that
gives a symmetx'ioal Gaussian peak with no dOublets or pre-peaks, representing
a uniform relense of the mercury from the amalgam during firing.
This was
achieved over a fairly wide range when either the ~ inch or the 3 inch cell
was used.
The best setting for the firing temperature is one at which the
gold just starts to glow at the end of the firing period. This penni ts a
visual check of the degree of firing, but is sufficiently 10" to eliminate
fusing of the gold. Firing temperatl1res are not a critical parameter with
the longer cells as long as they are high enough to volatilize all the
mercury and avoid excessive trailing of the peaks.
A wide range of
temperatures were observed to release the mercury; the lower settings require
a longer period of firing.
Tl1e completeness of mercury volatilization is
observable on the strip chart recording. With the standard 15 mm amalgamator
containing lOgrams of gold chips and a 2 gram gold wire plug settings of
80 - 100% were utilized.
Firing tempera.ture can be used advantageously for
-------
Figure 5.
Variac 100%
Flow-4
Va riac 70%
Flow-4
Peak Shapes obtained with the ¥ CeJ.l
16
-------
Figure 6.
17
15 mm Amalgamator
6 1/2 II Cell
0.25 JJg Hg
Abs.
0.8
Flow
-01
~o-
/0 ---02
o ------- 0
~p 03
o~ ----------o~o 4
----0 ------::::0:::::------0 5
0----- ---- 0::::----- 0
0:::::::=---0
o
0.4
o
70
80
100 %
90
, .
. Variac
Peak Height as a Function of Firing Temperature and Air Flow Rate
(15 rom Amalgamator, ChI Cell) .
-------
Figure 7.
18
Small Amalgamator
3" Cell
Abs.
0.5 ~g Hg
0.8
Flow
0.4
/01
o
~2
/0----0 3
/ ~O 04
o O~ 0~5
O~O~O
0::::::=---0
o
o
o
70
80
Variac
100
90.
%
Peak Height as a Function of Firinp, Temperature and Air J<'low Rate
( 15 mm Amalgamator, 3" Cell)
-------
Figure 8.
0.8
0.4
o
19
Large Amalgamator
3 II Cell
0.5 JIg Hg
Abs.
Flow
/1
o
°
O/~;
0----- ----0;::::::::----
8-;:::::::::::--°
40
60
%
50
Variac
,
Peak Height as a Function of Firing Temperature and Air Flow Rate
. ( 25 mm. Amalgamator, )" Ceil)
) .I'~
-------
Abs.
0.2
0.1
20
Flow 4
/D Main Peak
~:/ Prepeak
o/8~ .
D -
--0 I I I I
70 80 90 100 %
Abs.
0.2
Flow 5
0.1
Figure 9.
, .
o
70
90
100 %
80
Variac
Peak Height as a Function of Firin~ Temperature and Air F10w Rate
( 15 rom Amalgamator, ~, Ce11)
-------
21
slight adjustment of sensitivity.
S- Air Flow Rate
For ease of operation and the manual reproducibility of settings it is
advantageous to utilize the same air flow rate for aerating the standard
solutions and carryin~ the revolatilized mercury from the amalgam through
the quartz absorption cell.
Slower flow rates just increase the aeration
time required to volatilize the mercury from the standard solution, permitting
us to optimize the air flow rate solely as a function of the sensitivity and
peak shape observed from the absorption of the mercury at the 254 ~ line in
the quartz cell. The tirn.e required for complE:te aeration of the standard
solution at a given air flow rate was deterrrrl.ned by volatilizing the mercury
from the standard solution directly through the quartz absorption cell in the
spectrophotometer (bypassing "the amalgamation step). Figure 10 shows typical
curves obtained and illustrates the dependence of complete volatilization
on aeration rate.
These curves were utilized to determine how long to aerate
the standards at the selected flow rate.
A summar.y of required aeration
times for various flow rates is given in Table 2 and shows that complete
aeration is a function of flow and time and nearly independellt of sample size.
Table 2. 1-1inutes of aeration required at various flow rates
1 2 3 4 5
{}.J' (O.~g). 12 5 3 2.5 2.5
,
3" (l.QJ.g) 12 5 3 3 2.5
~ (lO.QJ.g) 12 6 4 4 3
3" (l.~g) 12 2.5
(lg. amalgamator)
-------
Flow -1
o
3
4
5
1
2
Flow..3
o
2
3
4
1
6
7
8
9
10
Flow-S
o
2
1
Figure 10.
Illustration of Procedure used to determine Aeration Time of Standard Solutions
11
12
3
N
N
-------
23
The dependence of peak height, or absorbance, on air flow rate was studied
as a function of mercury concentration for each of the cells. The results
are shown in Figures 11, 12, and 13. The slower flow rates. greatly' increase
the sensitivity, and the effect is more pronounced with the shorter cells
and higher mercury concentrations.
(Same caution should be exercised when.
aerating the more concentrated standard solutions at the higher f~ow rates
to insure that there is no bypass during the first few seconds of aeration,
when the mercury concentration passing over the gold is very high.)
This
can be avoided by only partially opening the bypass valve on the aeration
cell for the first 30 seconds. In effect, this aerates the highest con-
centration of mercury at a slower rate. This will slightly alter the shapes
of the curves shown in Figure 10, increasing the total aeration time by
)0 seconds. Air flow utilized for the majority of the. standards and samples
analyzed in this study was a total of ~ minutes aeration time (30 seconds
slow, 3 minutes at the full flow) at an air flaw setting of ) (l.4 liters/min.)
with the white plastic ball on the Perkin-Elmer gas control regulator.
,
-------
Abs.
1.0
0.5
o
Figure li.
24
15 mm Amalgamator
6 1/2" Cell
0.5 ~g
o
Variac 100%
o~ .
o o~
~ O~o.
0'---0
0.05 ~g -------0
~o .
~o-
0.25 ~g
o
o
1
2.
F low Rate
3 .
4
5
, .
Peak Height as a Function of Flow Rate and Sample Size
( 15 rnm Amalgamator, t>Y' Celi)
-------
25
Abs.
1.5
15 mm Amalgamator
I ~g Hg
o .
3/1 Cell
Variac 100%
0.5 ..,g
o
1D
0.5
o~
" .~
~. 0
o~o .
0.1 ~g .-------
~ 0
~
-0-
0 0
o '
2 3 4 5
Flow Rate
Figure 12.
Peak Hei~ht as a Function of Flow Rate and Sample Size
( 15 JmIl Amalcaroator, 3" Ceoll)
-------
Abs.
1.0
0.5
o
Figure 13.
26
15 mm
Amalgamator
1/2" Cell
Variac
100%
5 JJg
o
10 JJg
o
Hg
o
o~ 0
. o~
o
~ . . ""'--0
. . 0------0-
1 JJg
o
o
1
2
3
4
5
, .
Air Flow Rate
Peak Hei~ht as a Function of F .low Rate and Sample Size
( .l5 JI\I11 Arrlalgronator, !~I Cell)
-------
ANALYTICAL PROCEDURE
The studies optimizing the analytical determination of mercury resulted in
the following settings and analytical procedure:
Atomic Absorption Spectrophotometer, Perkin Elmer Model 303
Noise suppression
254 mli
1
Mercury line
Scale expansion
1
Chart speed
(as prescribed)
60 mm.min. (during analyses)
Lamp setting
Induction Furnace, Leco ModeL 521
Variac setting
80 .. 100%
High
Grid
Amalgamator
15 mm dia. with three finger support
2 gram gold wire plug
10 grams of 0.007" thick gold chips
Air Flow
1.4 liters/minute
Aeration Time
~ min. @ approx. 0.5 liters/minute
3 min. @ 1.4 liters/minute
27
-------
28
The solutions and standards used in this study were prepared as follows;
Stock Hg solution.' 1000 ppm in 5% HNO)*
Prepared by weighing out 1.3535 g. of reagent grade HgCl2'
addingi t to a .1000 mr volumetric flask containing 50 ml
of cone. HN03 and diluting with distilled water.
Stock Hg sOlution, 1 pp1'1 (prepared weekly)
Prepared by taking 1.00 m1 of the 1000 ppm stock Hg solution,
adding it to a 1000 ml volumetric flask containing 50 ml of
cone. 1iN03 and diluting with distilled water.
Working Hg solution, 100 ppb (prepared daily)
Prepared by taking 50 m1 of the 1 ppm stock Hg solution,
adding it to a 500 m1 volumetric flask containing 25 ml
cone. liNO) and diluting with distilled water.
SnC12 solution, 20% in 50% HC1
Dissolve 100 grams of SnC12.2H20 in 250 ml of cone. HCl (12M).
Dilute to 500 m1 with distJ.lled wate::-. Add a few pieces of
meta1J.ic tin to the solution. This :301ution may have a cloudy
appearance when first prepared; this does not affect its use.
In" the optimized procedure, ~ the straight .mnalgmnator containing the stack
sample is placed in the induction furnace; the top and bottom ball joints
are connected so that the system is sealed with a constant air flow rate
through the mnalgamator into the quartz absorption cell. The gold is
aligned in the furnace so that the tube i8 concentric with the induction coil
and then fired at about 90%. The resulting peak appears on the recorder
within a few seconds. A standard, approximating the observed peak, is
*ASTM, Mercury in Coal Methodolo~ Group has tentatively proposed. that all
mercury standards be made up to contain 5% lIN03 as a preservative. The.
1000 PIJ!1 stock solution prepared in this way is stable for at least a y.ear.
-------
29
placed in the aerator and reduced with 1 ml of 20% SnCl2 in 5~ HCl,
introduced by means of a syringe through the septum cover. A sufficient
time, at least l~ to 2 minutes, is allowed to assure that the gold has
cooled to approximately room temperature since the last firing. The
standard is then aerated for three minutes, (at an airflow of 3)
amalgamating the volatilized mercury onto the gold foil.
The amalgam is
then fired at the same variac and air flow settings as the sample, producing
a standard peak.
Experience has shown that the standard more exactly
reproduces the peak shape of the stack sample if the standard is aerated
onto the. top of the gold. This duplicates the direction of gas flow through
the amalgamator during the sampling operation and pl.aces the bulk of the
mercury onto the top layer of gold, promoting uniform release of mercury
from both standard and sample.
Figures 14 and 15 show typical working curves obtained using. the straight
amalgamator and the ~ and 3 inch cells respectively. The ~ inch cell was
useful up to about 2.0 ~g of mercury. Figure 16 shO\fS a typical working
curve obtained using the large (25mm) amalgamator in conjunction with the
3 inch cell.
,
-------
0.7
. ',5
"_3
0.1
.a.bs.
o
/
9
30
Ana Iytical
Curve
15 mm
Amalgamator
6
1/2 II Cell
Flow Rate 3
Variac
100%
o
0.5
1.0
, .
JJg Hg
Figure 14. Analytical Curve obtained with 6\:}' Cell and 15mm Amalgamator
-------
31
Analytical Curve
15 mm Amalgamator
3 II Cell
Flow Rate 3
Variac 100%
Abs.
0.8
~o
o
0.4
.rl
o
o
2.0
1.0
..,g Hg
, .
Figure 15. Analytical Curve obtained with 3" Cell and lSmm Amal~arnator
-------
32
Analytical Curve
25 mm Amalgamator
3" Cell
Flow Rate 1.5
Variac
45%
Abs.
0.7
0.1
o
/0
/
0.5
0.3
o
1.0
2.0
~g Hg
Figure 16. Analytical Curve obtained with 3" Cell and 25mm Amalgamator
-------
STACK GAS SM{PLING
As discussed in the previous section, the amalgamators were designed to
be inserted into a standard EPA isokinetic sampling train in place of the
impingers.
The interchangeable impingers and amalgamators permit using
any combination of amalgamation and wet absorption techniques to verify
collection efficiency of the various components.
The isokinetic train
is utilized to obtain a particulate sample simultaneously with the gas
sample.
Detectable levels in the particulate sample yield information on
the condensation of the volatile mercury.
1- Preparation of the Amalgamators
When handling the gold wire and chips used i:1 the amalgamators, care
should be taken to keep the hands clean and free of oil or grease (such
as stopcock grease picked up from handling the impingers).
If grease is
accidentally transferred to the gold it will be slowly released as the
gold is fired, giving a false reading and a very irregular peak.
The gold
wire plug and chips to be used in the amalgamators should be cleaned before
use.
The best way of cleaning is to place the gold into small crucibles and
fire it at about 7500C for three plus hours (or overnight).
The cleaned
gold may be stored for an indefinite period of time before use in small
sealed (mercury free) vials.
To . prepare the gold wire plug, 2 grams of gold wire are worked into a bail
with the fingers. . The wire is then rolled until it is slightly larger than
the inside diameter of the amalgamation tube.
The plug is then flattened into
JJ
-------
34
a wafer and forced into the tube against the supporting fingers. A length
of~' dowel rod and a, stiff (about #10) wire are handy to help wedge the wad
in place and flatten out any B~ray ends of wire.
The gold plug should fit
tightly enough in the tube to eliminate the bypass of any gold chips.
After a wad has been used a few times, it may become loose in the tube.
This may be corrected b.1 removing it from the tube and spreading it slightly
with the fingers and then repacking it with the dowel. After this has been
done a few times the wad will become quite dense and tend to cI"UJ$le. When
this happens the wad may be pulled apart until it is separated into its
original strands of wire.
Enough new \-lire is added to make up 2 grams, and.
the plug is re-formed.
In this way the plugs can be used almost indefinetE~ly
with only occasional addition of gold wire.
The gold chips may be prepared by cutting up a sheet of the metal, 0.007"
thick, with a scissors or cutting board. The chips~of gold should be about
1/8 to 1/16 inch square. ' Ten grams of gold chips are weighed and
poured into the amalgamator on top of the plug, using a plastic funnel.
The amalgamator is held vertically and gently tapped with the dowel to
slightly compress the chips. \oJhen the amalgamator is packed, it is placed
in the induction furnace and f,ired in the air stream to verify that the
gold is "clean" (no absorption peak is registered on the recorder).
After
tiring, the amalgamator is sealed by placing rubber stoppers in the ends of
the tube. The plugs are removed just before the amalgamator is placed in
the sampling train.
,
2- Operation of the Sampling Train
The sampling train predominantly utilized in this study consisted of a
-------
35
5 foot glass probe heated to l$OoF. with a 0.2$on diameter tip.
Samples
vere not collected isokinetically in the study permitting the probe tip to
be pointed downstream (particulates were not routinely analyzed).
The
cyclone and filter were maintained at 200~.
Five impinger bases were used
in an ice bath.
The first impinger was a Greenburg-Smi th type containing
300 ml. of an approximately 1~ SnC12 in 4% HCl solution.
This solution was
prepared by adding 25 rul. of the stock SnCl2 solution (used for reducing
the standard solutions) and then diluting with distilled \-later.
This
reducing solution decreases ~ potential mercury absorption.
The
Greenburg-Smith impinger is used to decrease bubble' size thus increasing
gas contact with the solution.
The solution is used to' reduce the gas
temperature below the dew point of sulfuric acid mist, removing SO 3' which
is an apparent interference in the analytical procedure.
The use of a
SnC12 solutior.. in the first impinger improved the original analyses, but the
concentration and amount of the solution bave not-.been optimized.
The
second impinger was empty and helped to remove some of the moisture from
the gas stream.
The third and fourth impingers each contained' a gold
amalgamator, one serving as a backup. .. The fifth impinger contained silica
gel to remove the remaining vater vapor before it entered the vacuum pump
of the console unit.
After the sample is collected, the probe should be disconnected from the
cyclone, the meter reading noted, and approximately one cubic foot of
ambient air pulled through the unit (through the SnC12 solution) to assure
complete volatilization of merCl~ from the solution.
An aliquot of the
SnC12 scrubbing solution should be preserved for analysis.
-------
36
Although the particulate samples were not routinely analyzed in this study,
'. .
the filter was replaced after each run and the probe rinsed periodically,
depending on the particulate concentration of the stack. When samples are
collected isokinetically the probe and cyclone should be rinsed with a
1:3 nitric acid-water solution, and the rinsings saved for analysis.
.'
..
-------
FIELD ANALYSES
After the analytical procedure'was optimized, the following five areas
were investigated with the stack gas samples:
(1) The efficiency of the gold procedure compared to an acidic
permanganate solution absorption train.
This was achieved by
running Au-Au, Au-KMn04' and KMn04-Au trains. The pennanganate
solutions were prepared with nitric acid.
ICl was not used because
mercury- volatilized from ICl solutions will not amalgamate onto
gold requiring the elimination of the gold amalgamation step.
(2)' A check for interferences, either other volatile compounds that
also absorb 254 ~ radiation or compounds present in the system
that abE:orb or adsorb mercury. . The reaction of the gold with the
gas stream must be considered to determine .if it chemically absorbs
any compounds from the gas stream or if volatile compounds condense
on the gold.
(3) Analyses were performed at 5, 10, 15 and 30 minute sampling periods.
(4) Samples were collected at flows of 0.2, 0.3, 0.4, 0.5, 0.6, 0.7
and 0.8 CFM.
Samples were not collected isokinetically due to the
location of the samp~ing ports and plant conditions at the first
plant.
(5) A limited number of particulate samples were analyzed by (1) perchloric
nitric digestion and flameless atomic absorption and (2) volatiliza-
tion in the induction furnace acconpanied by flameless atomic
absorption.
37
-------
38
Sampling sites at two different power plants were utilized for obtaining
stack samples to investigate the above parameters under field conditions.
A mobile laboratory permitting on site and instantaneous analyses of the
samples was located at the first site for six weeks a.11d at the' second site
for three dayso
The objective of the second site was to verify the results
of the first site under different field conditions.
The first power plant utilized a Mixed strip mine coal. source predominant~
from southeastern Ohio.
Information obtained from the purchasing department
showed the coal sources as: Ohio #5 seam 10%, Ohio #6 seam 40%, and Ohio
#7 seam 40%.
Samples were obtained in a 90,000 Kw single-train generator
boiler.
The unit is used as a peaking unit with a highly variable ~oadiDg.
A 50% change in loading was achievable within a few minutes.
Maximum air
utilization WEIS 900,000 ~bs./hr.
Particulates in the stack gas were cont.rolled
b,y. cyclone separators.
Upon leaving the cyclones the gas passed through the
ID fan, and then up the stack.
The sampling site, illustrated in Figure 17,
vas located in a horizontal duct between. the ID fan and the stack about 30
feet above the ground. A damper, used to control the draft, was located six
feet upstream ,from the sampling ports w1d did result in a high degree of
turbulence.
The duct had a negative pressure ranging from'l to 2 inches of
vater.
At maximum plant operating conditions the gas temperature was .:
approximately 32SoF. at the sampling 5i tee
The mobile laboratory was located
,beneath the sampling site penni tting the control conso.l.e of the sampling
train to be operated from within the laboratory.
, ,
The second site also utilized a mixed coal source but the mines were located
-------
Stack
Stack
Figure 17.
Ports
Side View
I D Fan
Ports
Top View
o
o
....
......
--
....... Damper
.'
Cylcones
Physical Arrangement and Location of Port5 at Site 1
....
--
--
39
, I D
Fan
Preheaters
-------
40
in eastern Ohio.
During the past few months the composition of this coal
has averaged: Ohio #5 seam 10%, Ohio #6 seam 33%, Ohio #1 seam 3%, and
Ohio-West Virginia #8 seam 5L%.
The unit that was sampled contained two
boilers supplying one 86,000 Kw generator.
The unit is normally operated
at a constant loading resulting in a more advantageous sampling situation.
Air usage uas maintained between 300,000 and 350,000 lbs./hr.
Particulates
in the stack gas were controlled by a Western Precipitation electrostatic
precipi tat,or.
Two banks of 4" sampling ports had been previously installed
in rectangular ducts immediately before and after the electrostatic pre-
cipitator.
The outlet ports were .located between the preCipitator and the
ID fan.
The perpendicu.Lar cross-section of the inlet duct is 5'0" x 22'8"
with a velocity range of .l2.2 to 5.l.6ft./sec., averaging 29.3 rt./sec.
The outlet duct was 4'1 3/4" x 13'0" with a velocity range of 31.3 to 62.2
rt./sec., averaging 51.8 rt./sec.
The static stack pressure was negative
between 4 and 5 inches water with a slightly higher negative pressure on
the outlet side.
Under normal operating conditions the gas teMperature at
the ports was approximately 330or~
Samples were obtained at both the inlet and
outlet ports.
-------
STACK SAMPLING RESULTS
The observed mercury concentrations and the control parameters are presented
in Tables 3 and 4 as a daily log of the runs. The results will be inter-
preted in relation to peak shapes, gold deterioration, interferences,
absorption cell size, stack turbulence, length of runs, flow rate, observed
concentrations and reproducibility.
1- Peak Shapes
As mentioned previously, a uniform release 0.£ mercury from the amalgam
w1.ll result in a Gaussian or 'bell' shaped p~ak. For accurate detenninations
of the mercury concentration, the peaks resulting from standards should
match the peaks resulting from the stack samples.
In practice the standards
took longer to be released from the amalgam and resulted in a lower, broader
peak.
This was a result of the stack gas samples being amalgamated from the
top down l stack gas flow in the impingers of the train is down) and the
standards being amalgamated up from the bottom. This situation is partially
corrected by the additional step of reversing the direction of the air
flow in the amalgamator while aerating the standard solutions. This results
in amalgamating the volatile mercury by flow down through the amalgamator
while both obtaining tne sample and amalgamating the standard. Figure 16
illustrates an identical standard amalgamated from the top and from the
bottom of the amalgamator and a typical stack gas sample. The potential
error in the calculation of the stack sample is obvious.
The firing of
the induction furnace produces and induction field which inter~6ts with
41
-------
Table 3. Daily log 01' runs&
mercury concentration and control parameters
SITE 1
DATE TIME SAMPLING SAMPLING VOLUHE IMPINGER }lliRCURY PRE PEAKS % RECOVERY
TIME RATE ORDER !MAL. #1 AMAL.#2 TOTAL CONCENTRATION 1 2 !MAL. #1
minutes CFN liters iJ.g iJ.g iJ.g iJ.g/1. % abs.
3/9/12 2:30 5 .762 107.8
5 ~734 105.0
5 .66h 94.0
5 .638 90.4 >
5 .767 108.5
5 .509 72.1
5 .579 82.0
5 .586 83.2
3/10/72 2:30 5 .931 131.7
2.5 .349 24.7
2.5 .269 19.0
3:45 2.5 .297 21.0
4:05 2.5 .324 22.9
2.5 .307 21.7
3/13/72 11:00 4.00 .372 41.7 A 1.3 1.3 .031 ?
11:31 4.00 .391 44.5 A 0.5 0.5 .011 -
. 11: 56 . 4.00 .344 36.4 A 0.74 0.74 .019 x
12113 8.00 ~484 108 A 0.21 0.21 .002 x
2:20 4.00 .452 A
4:15 5.00 .527 A
3/J.W72 11: 34 5.00 .580 81.0 A 0.31 0.31 .0038 -
1:45 30.00 .565 475 A 0.3 0.3 .0006 -
2:41 30.00 .540 454 A 0.8 0.8 .0018 ?
::-
I\)
-------
SITE 1 (con~.)
DATE TIl-tE SAMPLING SAMPLING VOLUME IMPING ER MERCURY PREPEAKS % RECOVERY
TIME RATE ORDER ANAL. #1 AHAL. #2 TOTAL CONCENTRATION 1 2 AMAL. #1
minutes CFl-1 liters ~g ~g ~g . ~/g1 % abs.
3/15/72 10:44 5~1 .067 9.6 AA
11:57 5.0 .065 9.2 AA
12:58 5.0 .064 9.1 AA
3:50 5.0 .073 10.3 EEA
4:33 5.0 .075 10.6 EEA
4:55 5.0 .076 10.7 EEA
5:15 5.0 0073 10.3 SEA
5:35 5.1 .011 10.2 SEA
5:51 5.5 .011 11.0 SEA
3/16/72 4:15 5.0 .164 23.0 SEAA 0.50 0.10 0.60 .025 - x 83
5:36 5.0 .156 21.9 SEliA 0.21 0.16 0.43 .019 - x 63
6:33 5.0 .067 9.h SEAA 0.19 0.36 0.55 .058 - - 35
3/17/72 1:20 5.0 .018 10.9 SEAA 0.35 0.17 0.52 .048 x x 67
2:17 5.0 .081 11.3 SEAA 0.29 0.20 0.49 .044 5 - 59
3:55 5.0 .080 11.2 SEAA 0.21 0.19 0.40 .036 - - 52
4:43 5.2 .080 11.6 SEAA 0.29 0.36 0.65 .056 - - 45
3/20/72 11:08 5.0 .029 4.1 SEAA 59 83
11:46 5.0 .046 6.5 SEAA 0.146 0.059 0.205 .031 6 x 73
12:18 5.0 .191 27.0 SEAA 0.48 0.006 0.49 .018 9 x 98
12:58 5.0 .198 28.0 SEAA 0.29 0.054 0.34 .012 - - 85
1:35 5.0 .326 46.1 SEAA 1.1 0.074 1.17 .025 - x 94
2:26 5.0 .338 47.7 SEAA 0.94 0.093 1.03 .022 - - 91
3:03 5.0 .487 68.9 SEAA 1.35 0.06 1.41 .020 - - 96
3:45 5.0 .t,87 68.8 SEAA 1.57 0.096 1.67 .024 - - 94
4:18 5.0 .619 81.6 SEliA 1.52 0.25 1.71 .020 - - 86
4:52 5.0 .611 86.4 SEAA 1.49 0.29 1.18 v020 - - 84
5:32 5.0 .720 102.0 SEAA 1.58 0.54 2.12 .021 - - 75
6:03 5.0 .709 100.4 SEAA 0.67 0.10 1.37 .014 - - 49
6:35 5.0 .216 30.6 SEliA 0.18 0.12 0.30 .010 - x 60
~
\.A)
-------
SITE :. (con;.:
DATE TIME SAMPLHTG SAMPLING VOLUME IMPINGER MERCURY PREPEAKS % RECOVE4tY
TIME RATE ORDER AMAL. #1. AMAL.#2 TOTAL CONCElITRATION 1 2 AYtAL. #1
minutes CFM liters ~g ~g \J.g ~g/l % abs.
3/21/72 11: 00 5.0 .185 26.1 SEAA 0.44 0.017 0.46 .024 - - . 96
11:44 5.0 .179 25.2 SEAA 0.14 0.025 0.17 .007 - - 82
12:27 5.0 .414 58.6 SEAA 0.55 0.038 0.59 .010 - - 93
1:08 5.0 .411 58.1 SEM 0.61 0.086 0.70 .012 - - 87
1:43 5.0 .554 78.4 SEAA 1.13 0.12 1.25 .016 - - 99
2:30 5.0 .551 78.0 SEAA 1.26 O.lh 1.40 .018 - - 90
3:11 5.0 .631 89.2 SEAA 1. 3.~ 0.12 1.47 .017 - - 92
3:50 5.0 .631 8902 SEAA 1.28 0.13 1.41 .016 - - 91
4:48 5.0 .787 111.2 SEM 0.22 0.22 0.44 .004 - - 50
5:29 5.0 .788 1ll.3 SEM 0.27 79 x -
. . .
3/22/72 10:21 5.0 .182 25.8 SEM 0.46 0.22 0.68 .026 - x 68
11:16 5.0 .170 24.0 SEAA 0.24 0.70 0.94 .039 - - 26
12:02 5.0 .418 59.2 . SEAA 0.81 0.10 0.91 .015 - x 89
2:44 5.0 .379 53.6 SEM 1.1.0 0.10 1.50 .028 - x 93
3:22 5.0 .384 54.6 SEM 1.04 0.16 1.22 ;.:022 - - 65
3:49 5.0 .549 77.6 SEAA 1.23 0.35 1.58 .020 - - 78
3/23/72 9:50 5.0 .213 . 30.2 SEAA 0.54 0.048 1.02 0.020 10 27 53
10:18 5.0 .219 31.1 SEAA 0.30 - 34
10:42 5.0 .3112 48.5 SEM 1.20 0.04 1.24 0.026 - 6 97
11.:.17 5.0 .343 46.6 SEAA 1.20 0.055 1.26 0.026 - 10.4 95
1l.:57 5.0 .440 62.3 SEAA 1.29 0.09 1.36 0.022 - 1 94
12:30 5.0 .h42 62.5 SEAA 1.21 0.17 1.38 0.022 - x 86
1:00 5.0 .436 62.0 SEM 1.3 0.20 1.50 0.024 - 5 67
, 1:31 5.0 .506 71..8 SEM 1.14 0.25 1.39 0.01.9 - - 62
2:00 5.0 .494 69.9 SEAA 1.1 0.25 1.35 0.019 ? - 82
2:32 5.0 .608 66.0 SEAA -
3:05 5.0 .587 63.9 SEAA - gold deteriorating
.4:00 10.0 .395 1ll.7 SEll 2.6 I I 0.023 70
5:12 10.0 .384 106.6 SEM - gold reteriOrjting
g:
-------
~ITE J. (cont.)
DATE TIHE SIuV.PLIHG SAJ!iPLING VOLUME IMPINGER MERCURY PRE PEAKS % RECOVERY
TD1E RATE ORDI<1\ AHAL. #1 A.,.J.fAL. (12 TOTAL CONCENTRATION 1 2 AHAL. #1..
J1U.nlltes CI<'U liters l-Lg ~g l-Lg ~g/1 % abs.
3/2'4/72 10:31 5.0 .488 69.1 SEAA 1.24 0.047 1.29 0.019 - - 96
11:14 5.1 .350 49.6 SEAA 1.04 0.025 1.06 0.021 - ~ 98
11:46 5.0 .583 82.6 SEAA 1..73 O.03~ 1.80 0.021 - - 96
12:32 5.0 .204 28.9 SEAA 0.30 0.088 0.39 0.013 - - 11
1:09 5.0 .103 99.6 SEM 1.3 0.035 1.34 0.013 - - 97
1:'48 5.0 .184 110.9 SEAA 2.2 0.054 2.25 0.020 - -. 98
2:41 10.0 .202 51.2 SEA-I\. 0.55 25 50 -
3:22 10.0 .213 60.3 SEAA 0.56 0.13 0.69 0.011 - - 81
4:23 10.0 .350 99.1 SEM 2.5 0.035 2.54 0.026 - - 98
5:00 10.0 .632 118.5 SEAA )5.0 0.26 - - -
5:)'4 10.0 .110 218.0 SEM >5.0 0.15 x x -
6:40 10.0 .h95 139.8 SEAA 0.17 - -
3/25/72 10: 30 10.0 .209 59.1 KEAA x x -
11:13 10.0 .335 94..7 KEAE x -
12:10 10.0 .331 95.h KEAE x -
1:40 10.0 .338 94.5 KEAE 2.3 2.3 0.024 -
3:20 10.0 .330 93.8 KEAE 1.96 1.96 0.021.. 80 -
h:03 10.0 .318 90.0 KSEA 1.89 0.078 1.97 0.022 - 96
5:10 1.0 .378 74.9 SEAK 1.9 0.22 2.12 0.028 16 90
5:30 10.0 .243 68.6 KSEA 1.51 ? 1.51 0.022 51 -
6:20 1.0 .311 74.6 St:AK 1.1 0.21 1.3 0.018 x 85
3/27/72 12:15 15.1 .304 130.0 KSEA 2.59 0.69 3.28 0.025 - 19
1:18 15.0 .2?1 123.4 SEAK 2.3 '0.5 '2.8 (0.023 - - -
2:11 15.0 .h43 188.0 SEAK 4.7 0.4 5.1 0.027 - 92
3:05 15.0 .438 186.0 KSEA 3.93 0.59 4.52 0.024 4 87
4:04' 15.0 .575 244.0 KSEA 5.96 0.08 6.35 0.025 - 94
5:34 15.0 .519 244.8 SEAK 4.68 0.65 5.33 0.022 - 88
::-
V\
-------
SITE 1 (cont.)
DATE . TIME SkHPLnlG SAMPLING VOLUME IMPINGER MERCURY PREPEAKS % RECOVERY
TIME RATE ORDER ANAL. #1 AMAL.#2 TOTAL CONCENTRATION 1 2 AMAL. #1
minutes CFH liters \Jog \Jog \Jog \Jog/1. % abs.
3/28/72 9:53 15.0 .681 289.5 SEAK 1.69 0.39 2008 0.006 - 81
10:49 15.0 0622 264.0 KSEA 4.43 0016 4.59 0.025 - 97
1:17 1.5.0 .478 203.0 KSEA 2.50 0.29 2.79 0.014 x 90
2:00 15.0 .494 210.0 SEAK 1.85 0.9 2.95 0.013 - 68
2:43 15.0 .320 136.0 SEAK 1.93 <0.3 1.93 0.0l4 x "'"'. 90+
:;. .-
3/29/72 8:20 1.7.0 .395 189.9 SEAK 3.94 0.60 4.54 0.024. - - 87
9:02 17.0 .393 189.0 SEAK - -
9:55 10.1 .395 113.0 SEAA x. x -
11: 00 10.0 .397 112.2 SEAA x x -
11:55 30.0 .386 328.0 SEAA -
2:16 5.0 .479 61.8 SEAA 1.19 0.161 1.36 0.020 10 x 88
3:07 5.0 .469 66.4 SF..AA 1.25 0.189 1.44 0.022 x ? 87
3:52 5.0 .312 44.8 SEAA 0.58 0.04 0.62 0.014 5 14 93
3/30/72 il: 10 5.0 .304 43.0 EEAA 0.06.3 0.046 0.109 0.00] 15 17 58
1:09 5.0 .297 42.0 HEAA 0.24 hidden 0.000 60 70 -
2:3h 5.0 .300 42.5 SEAA 0.46 hidden O.Oll 20 62 -
3:33 5.0 .297 42.0 AgEAA 0.12 0.21 0.33 0.008 64 x 36
4:43 5.0 .510 72.2 SEAA 1.0 0.14 1.14 0.016 - 40 88
3/31/72 11:58 5.0 .340 48.1 HEAA 0.28 - 30.5 -
12:42 5.0 .342 48.3 HEAA 0.29 - 16.5 -
1:23 5.0 .340 48.1 HEAA 8 19.8 -
2:15 5.0 .340 48.1 SEAA 0.79 0.073 0.863 0.018 - 15.2 92
3:15 5.0 .328 46.5 SEAA 0.30 0.075 0.375 0.0081 - 5.6 81
4:26 5.0 .329 46.5 SEAA 0.355 0.057 0.412 0.0088 - 4.5 86
5:36 5.0 .328 46.5 HEAA 0.50 0.046 0.55 0.012 8.e 5.1 91
4/6/72 12:25 5.0 .473 40.8 SEAA 1.5. 0.04 1.54 0.038 - - 97
~
~
-------
SITE 1 (cont.)
DATE TIME SAMPLING SAMPLING VOLUME IMPINGER MmCTffiY PRE PEAKS % RECOVERY
TIME RATE ORDER AHAL. #1 AHAL.#2 TOTAL COUCr~NTRA TION 1 2 AMAL. #1
minutes CFM liters ~g ~g ~g ~g71 %abs.
4/1/12 10:50 5.0 .531 15.8 SEll 0.60 0.29 0.89 0.012 - x 68
11:32 5.0 .568 80.4 SEAA 0.50 0.063 0.56 0.001 - 3 89
12: 03 5.0 .564 19.8 SEAA 0.44 0.62. 1.06 0.013 x x 41
12:39 5.0 .559 19.1 SEll 1.6 0.15 1.75" 0.022 . - . - 92
1:10 5.0 .557 78.8 SEAA 1.44 0.32 1.76 0.022 - 15 82
1:51 5.0 .546 1103 SEAA 1.46 0.14 1.60 0,,021 ... 12 91
2:28 5.0 .548 17.5 SEAA 0.55 0.23 0.18 0.010 x 11
3:05 5.0 .538 16.1 SEAA 1.41 0.19 1.66 0.022 89
3:40 5.0 .542 16.1 SEAA 1.57 0.43 2.00 0.026 ~ - 19
4:40 5.0 .409 58.0 SEAA 101 0.41 1.57 0.027 ? ? 70
5:10 5.0 .390 55.3 SEAA 0.95 0.61 1.62 0.021 5 ? 59
5:45 5.0 .361 52.2 SEAA 0.94 0.46 1.40 0.021 15 ? 61
6:30 5.0 .493 10.0 SEAA 1.13 0.19 1.32 0.019 - 86
1:05 5.0 .488 69.0 SEAA 1.06 0.23 1.29 0.019 83
4/10/12 10:51 5.0 .382 5h..1 SEAA 0.99 0.12 1.11 0.021 - x 89
11:39 5.0 .195 21.6 SEAA 0.55 x - -
12:11 5.0 .531 15.1 SEll 1.5 0.11 1.61 0.022 x 11 90
12:48 5.0 .619 81.6 SEAA 1.43 0.41 1.85 0.021 - x 18
2:08 5.0 .161 108.3 SEAA 1.59 0.51 2.10 0.019 - - 16
2:42 5.0 .304 43.0 SEAA 0.48 0.32 0.80 0.019 - 25 60
3:31 5.0 .304 43.0 SEAA 0.19 0.23 1.02 0.024 - 13 11
4:00 5.0 .621 88.1 SEAA 0.89 0.62 1.51 0.011 - - 59
4:32 5.0 .110 . 109.0 SEAA 0.96 0.10 1.66 0.015 - - 58
5:10 5.0 .392 55.5 SEAA 0.29 0.66 0.95 0.016 - x 31
5:51 5.0 .399 56.5 SEAA 0.40 0.48 0.88 0.016 - 17 46
6:21 5.0 .396 56.1 SEAA 0.45 0.58 1.03 0.018 - 17 44
~
~
-------
SITE 1 (cont:~)
DATE TIME SAMPLING SAMPLING VOLUHE IMPINGER MERCURY , PREPEAKS % RECOVERY
TIME RATE ORDER Mill. 111 A1-'JAL.#2 TOTAL CONCENTRATION 1 2 AMAL. #1
minutes CFM liters i-Lg i-Lg i-Lg ~g/l %abs.
4/11/72 11:04 5.0 .402 56.9 SEAA 0.935 0.073 1.008 0.018 - x 86
11:45 5.0 .737 105.0 SEAA 1.20 0.046 1.25 0.012 - - 96
12:21 5.0 .734 104.0 SEAA - - -
1.:40 5.0 .714 1.01.0 SBAA 0.52 0.16 0.68 0.0066 - - 77
3:26 5.0 .799 113.1. SEAA 0.86 0.11 0.97 0.0086 - :t 89
4:21 5.0 .508 71.9 SEAA 1.1 0.09 1..19 0.017 - 6 93
5:31 5.0 .259 36.6 SEAA 0.37 0.15 0.52 0.014 x x 71
4/12/72 10:20 5.0 0.625 88.3 SEAA 1.84 0.18 2.02 0.023 .. x 91
11:25 5.0 0.379 53.5 SEAA 0.78 0.02 0.80 0.015 x x 97
1:00 5.0 0.841 118.9 . SEAA 1.56 0.27 1.83 0.015 - - 85
1:52 5.0 0.307 43.5 SEAA 0.27 0.h2 0.69 0.016 - x 39
2:37 5.0 0.415 58.7 SEAA 0.85 0.0485 0.898 0.015 X - 94
3:21 10.0 0.423 119.7 SEAA 1.32 0.41 1.73 0.014 - x 77
4:10 15.0 0.383 162.h SEAA 2.21 0.33 2.54 0.016 15 8 83
5:10 5.0 0.427 60.3 SEAA 0.20 0.44 0.64 0.011 - 6 31. .
4/13/72 10:55 5.0 0.489. 69.3 SEAA 1.33 0.33 1.66 0.024 - - 80
11:30 5.0 0.484 68.5 SEAA 0.42 0.35 0.77: 0.011 - - 55
12:33 5.0 0.488 . 69.0 SEAA 0.84 0.08 0.92.. 0.013 '8!' - 91
3:25 5.0 0.488 69.0 SEAA 0.46 0.18 0.64 0.0093 x - 72
4:17 5.0 0.500 70.6 SEAA 0.78 0.052 0.83 0.012 - - ' .. 94
5:05 15.0 0.470 199.0 SEAA 3.4 0.55 3.95 0.020 - - 86
:::-
(X)
-------
SITE 2
DATE TIME SAMPLING SAMPLING VOLUME IMPINGER MERCURY PREPEAKS %RECOVERY
TIME RATE ORDER AMAL. #1 AMAL.#2 TOTAL CONCENTRATION 1 2 AMAL. #1
minutes CFM liters l-Lg l-Lg l-Lg l-Lg/l %abs.
4/19/72 10: 35 5.0 0.500 77.6 SEAA 0.30 0.01 0.30 0.0039 - - 97+
11:19 5.0 0.494 70.6 SEAA 0.21 0.01 0.21 0.0030 - - 95+
11:55 10.0 0.511 144.0 SEAA 0.46 0.021 0.48 0.0033 -' - 96
12:15 15.0 0.498 211.0 SEAA 2.06 0.031 2.09 0.0099 - x 99
2:11 15.0 0.510 216.0 S!::AA 2.10 0.45 2.55 0.0120 x - 83
3:23 10.0 0.511 144.7 SEAA 0.65 0.50 1.15 0.0079 - I - 57
8:30 5.0 0.509 72.1 SEAA 0.20 1'0.01 0.20 0.0028 - - 95+
9:00 5.0 0.506 71.6 SEAA 0.45 0.0063 - x -
9:33 5.0 0.519 73.4 SEAA 0.37 0.016 0.39 0.0052 x 10 95
4/20/72 9:47 10.0 0.1192 139.1 SEAA 0.90 0.054 0.95 0.0068 - 20 95
10:33 10.0 0.506 143.2 SEAA 0.89 0.05 0.94 0.0066 - 47 95
11:17 10.0 0.322 85.0 SEAA 0.46 --- 0.46 0.0054 10 40 -
12:04 10.0 0.328 92.7 SEAA 0.29 --- 0.29 0.0031 6 40 -
12:51 10.0 0.723 20h.5 SEAA 0.78 0.47 1.25 0.0061 - - 62
2:16 10.0 0.804 228.0 SEAA 1.06 0.05 1.11 0.0049 - x 96
2:59 5.0 0.734 104.'0 s EA.A 0.46 0.033 0.49 0.0047 - 4 94
3J39 5.0 0.1.~96 70.2 SEAA 0.345 0.012 0.357 0.0050 - - 97
4:33 5.0 0.394 55.9 SEAA 0.274 0.056 0.33 0.0059 - 3 82
5:13 5.0 0.721 102.0 SEAA 0.43 0.02 0.45 0.0044 - - 95
Imp1nger OrderJ
A - gold.
Ag- AgN03 scrubber
E - empty
H - water scrubber
K - K}mD4 scrubber
S - SnC12 scrubber
Prepeaks I
- - absent
x - present, but not measurable
l:""
'0
-------
I ~a):.e 4. )aily. .og 0, ~ runs: control parameters
DATE TIME DRY GAS AHALGA11ATOR GOLD STACK COlfDITIONS II{STR{W~NT PARAHETERS AMBIENT
TEMP. CONDITION AHOUNT STATIC TEHP. AIR FLOri AIR FIDI CELL FIRING TEMP.
PRF..5SURE P~rkin-E1mer SIZE TEMP.
of rom grams inches H?O of 1bs./hr. units inches Of
3/9/72 2:30 63 2.3 330
69 2.2 330
77 1.9 330
83 1.6 330
89 1.2 330
92 0.h6 325
94 0.38 320
95 0.11 325
3/10/72 2:30 66 1.7 335 3 . 75%
63 1.6 330
64 1.5 325
3:45 66 1.6 325
4:05 67 1.6 320
68 1.6 325
3/13/72 11:00 70 25 10 1.5 320 920,000 4 3" 75% 71
11:31 75 15 1~S 310 760,000 85%
11:56 77 25 1.7 305 700,000 65%
12:13 83 15 .1..6 305 820,000 85%
2:20 79 15 1.7 310 620,000 80%
4:15 14 1.5 305 770,000
3/14/72 l.l: 34 66 15 1.5 340 . 900,000 1.5 3" 82% 35
1:h5 92 15 1.6 340 700,000
2:41 97 15 1.4 340 760,000
3/J.5/72 .LO: 4h 63 15 F 15, 20 2.1 330 840,000 1.5 3" 82% 35
11:57 64 - 15, 20 2.2 320 800,000
. 12: 58 67 - 15, 20 2.0 320 810,000
3:50 69 F 15 1.9 320 740,000
4:33 70 - 15 1.4 320 610,000
'cS-
-------
DATE TIME DRY GAS AMAI.G A11ATOR GOLD STACK CONDITIONS INSTRUMENT PARAMETERS A1-1BIENT
T El'{p . CONDITION AMOUNT STATIC TEt-1P. AIR . FIDI AIR FLOW CELL FIRING TEMP.
PRESSURE Perkin-Elmer SIZE TEHP
Of mm grams inches H20 Of 1bs./hr. units inches ~
3/15/72 4:55 71 15 - 15 1.5 320 555,000 1.5 3" 82% 35
5:15 71 - 15 1.5 320 555,000
5:35 71 - 15 104 320 555,000
5:57 71 - 15 109 320 680,000
3/16/72 11:15 75 15 F 1.5 315 760,000 1.5 3" . 82% SO
5:36 71 1.3 320 540,000
6:33 70 1.3 315 540,000
3/17/72 1:20 - 15 F 2.0 325 740,000 1.5 3" 82% 40
2:11 78 1.6 335 540,000 .
3:55 77 1.8 330 650,000
4:43 78 1.6 330 550,000
3/20/72 11: 08 65 15 F 10 1.5 325 6.l0, 000 1.5 3" 82% -
11:46 66 - 1.9 325 670,000
12:18 68 - 1.9 325 610,000
12:58 70 - 1.5 310 610,000
1:35 74 - 1.5 310 620,000
2:26 75 - 1.2 280 1140,000
3:03 79 - 1.1 275 2110,000
3:45 81 - 1.1 270 170,000
4:18 84 - 1.0 260 170,000
4:52 85 - 1.0 260 300,000
5:32 85 - 1.0 265 300,000
6:03 85 - 1.0 270 280,000
6:35 19 - 1.2 280 450,000
3/21/72 11:00 68 25 F 12 1.2 270 420,000 1.5 3" 45:£ -
11: 44 73 0.9 250 260,000
12:27 77 0.9 250 280,000
1:08 79 1.2 275 6110,000
V\
~
-------
DATE TIME DRYGAS AMALGAMATOR GOLD STACK CONDITIONS INSTRUlID"T PARAMETERS AMBIENT
TEl1P. COlIDITION AMOUNT STATIC TEMP. AIR FLOVl AIR FLOW CELL FIRING TEMP.
PRESSURE Perkin-Elmer SIZE TEHP.
~ rnm grams inches H20 ~ 1bs./hr. units inches Of
3/21/72 1:43 83 25 F 12 1.3 285 630,000 1.5 3" 45%
2:30 84 1.2 280 420,000
3:11 86 1.0 210 390,000
3:50 85 1.1 210 380,000
4:48 84 1.8 305 700,000
5:29 81 1.4 305 580,000
3/22/72 10:21 &J 25 F 24, 12 1.5 315 670,000 1.5 3" 45% -
11:16 63 F 12, 24 :J..4 325 580,000 45%
12:02 65 lIN03 12 1.3 320 550,000 45%
.2:44 65 IIN03 1.4 325 590,000 . 50%
3:22 65 lINo) 1.3 325 560,000 50%
3:49 67 IINO. 1.3 325 560,000 50%
3
3/23/72 9:50 56 15 F 10 1.8 340 680,000 3 3" 80% 30
10:18 58 F 2.0 350 830,000
10:42 60 F 2.0 350 900,000
11:17 62 lIN03 2.0 355 860,000
11:51 62 2.0 355 760,000
12:30 63 1.9 355 840,000
1:00 64 1.9 350 800,000
1:31 64 2.0 340 900,000
2:00 65 2.0 355 790,000
2: 32 65 1.9 .350 B40,000
3:05 66 2.1 350 800,000
4:00 66 2.0 365 920,000
5:12 64 1.8 360 670,000
3/24/72 11:31 56 15 F 10 2.1 355 760,000 3 3" 85% -
11:14 58 F 2.1 345 810,000
11:46 63 F )45 810,000
V\
N
-------
DATE TIME DRY GAS AMAWAMATOR GOLD STACK COrIDITIONS INSTRUMENT PARAI1ETERS AMBIENT
TEMP . CONDITION AHOUNT STATIC TEf-1P. AIR FLOW AIR FLOW CELL FIRING TEMP.
PRF..5SURE Perkin-Elmer RT7.F. TF.MP -
~ MI'I1 grams inches H20 ~ lbs./hr. units inches <7
,
3/2'4/72 12:32 66 15 F 10 2.1 3ho 870,000 3 3" 85%
1:09 75 mm3 1.9 3h5 820,000
1:48 76 2.0 340 840,000
2:41 76 2.1 3US 820,000
3:22 - 1.4 330 530,000
4:23 71 2.0 34:0 910,000
5:00 80 1.5 330 520,000
5:54 83 1.5 330 530,000
6:40 76 I 1.7 330 720,000
.
3/25/72 10:30 56 15 F 10 2.1 360 980,000 3 3"' 85 33
11:13 67 F 2.1 360 980,000
12:10 71 mm3 2.0 355 980,000
1:40 75 2.1 360 820,000
3:02 74 2.1 345 730,000
4:03 75 1.8 330 700,000
5:10 75 2.1 .340 900,000
5:30 76 2.1 345 820,000
6:20 77 . - 355 960,000
3/27/72 12:15 64 15 F 10 1.9 350 940,000 3 ¥' 9r:J/. 47
1:18 71 F 1.9 345 970,000
2:17 76 F 1.9 340 970,000
3:05 76 F 1.8 350 970,000
4":04 80 - HN03 1.8 350 970,000
5:34 79 lIN03 1.8 355 970,000
3/28/72 9:53 70 15 F 10 2.1 360 980,000 3 ¥" 9CJ1, 42
10:49 74 F" 2.1 360 980,000
1:17 82 F 1.9 340 900,000
2:00 81 lIN03 2.0 345 980,000
2:"43 80 t 2.0 350 980,000
V\
W
-------
I
DATE TIME DRY GAS AMAIDAHATOR GOLD STACK CONDITIONS INSTRUHENT PARANETERS AMBIENT
TEMP . CONDITION AJ.iOUUT STATIC TEHP. AIR FIDI AIR FLOW CELL FIRING TEMP.
PRESSURE Perkin-Elmer 5T7.F. 'T'-:<'MP.
vy mm grams inches H20 Of Ibs./hr. units inches vF
3/29/72 8:20 73 15 F 10 109 340 870,000 2. ~ 43
9:02 02 F. 1.9 3hO 940,000 ~,
9:55 79 F 2.0 _.J~ ~,
J~U 920,000
11: 00 75 liNO 1.8 340 850,000 ¥
11: 55 92 HN03 1.9 340 820,000 ¥
2:16 82 mm3 1.8 330 780,000 3"
):07. 77 HH03 1.9 320 180,000 3"
3:52 75 m~03 1..8 335 3"
3 550,000
3/30/72 11: 10 59 15 F 10 1.5 320 600,000 2 >3" 100% 47
1:09 66 F 1.4 315 620,000
2:34 72 HNO 102 310 540,000
3:33 74 rom3 1.3 310 560,000
4:43 76 HN03 1.3 300 440,000
3
3/31/72 11:58 65 15 F 10 1.2 270 240,000 3" 52
12:42 64 F 1.2 265 240,000
1:23 82 F 1.1 270 240,000
2:15 86 HNO) 1.1 270 240,000
):15 87 HNO) 1.2 270 240,000
4:26 68 HNO) 1.1 265 240,000
5:36 66 HNO) 1.2 270 240,000
4/6/72 12:25 60 15 F 10 1.1 280 490,000 double sma gsmation
-
4/7/72 10:50 53 15 F 10 - )20 600,000 ) )" ~ 26
11:32 56 F 1.9 320 600,000
12:03 - F 1.5 330 600,000
12:39 64 HNO) - )40 760,000
1:10 63 1 1.6 )40 700,000
1:51 6) 1.5 )20 680,000
, ..
V\
~
-------
DATE TIME DRY GAS AMALGAMATOR GOLD STACK COtIDITIONS INSTRUMENT PARAMETl!.'RS AMBIENT
TEHP. CONDITION AMOUNT STATIC TEl1P . AIR FLOW AIR FLO\i . CELL FIRING TEMP.
PRESSURE Perkin-1!.:lmer SIZE . TEl-II'.
O}i' rom grams inches H20 Of 1bs./hr. units inches v.F
4/1/12 2:28 62 15 HN03 10 1.5 325 680,000 3 3" 9CJ1, 26
3:05 63 1.8 340 770,000
3:40 64 1.7 340 770,000
4:40 61 1.5 335 650,000
5:10 62 1.4 325 640,000
5:45 62 1.6 335 650,000
6:30 62 .L08 340 710,000
7:05 62 ,I, - 360 980,000
4/10/72 10:57 64 15, 25 F 10 1.7 310 705,000 3 ~tf 100%- 58
11: 39 66 F 1.8 310 705,000 55%
12:11 13 F 1.8 )10 705,000
.12:48 71 HNO) 1.8 310 710,000
2:08 81 .1.7 315 700,000
2:42 79 1.8 320 850,000
3: 31. 75 .1.9 320 850,000
4:00 78 .L.8 325 850,000
4:32 80 .L.tS 330 850,000
5:10 75 1.9 330 780,000
5:57 72 1.9 335 780,000
6:27 72 " .l.9 330 740,000
4/ti/72 .ll:04 72 15 F 10 1.9 320 720,000 3 3" 100% 65
11:45 71 F ' 1.7 315 700,000
12:21 79 F 1.8 305 650,000
1:40 78 TJ .l.9 305 680,000
3:26 77 .l.6 305 520,000
4:21 76 .1..9 320 710,000
5:13 75 .L.8 320 700,000
VI.
VI.
-------
DATE TilIE DRY GAS AMAlGAHATOR GOLD STACK COtIDITIONS INSTRill1ENT PARPJ1ETERS AMBIENT
TlliP. CO?IDITIOU ANOUNT STATIC TEMP. AIR FLOU AIR FLOW CELL FIRING TEMP.
PRES0URE Perkin-Elmer SIZE TF.MP.
Of mm grams inches H20 ~ 1bs./hr. units inches ~
u/12/72 10: 20 64 15 F 10 2.0 325 850,000 3 3" 10CY;t 64-73
11:25 72 F 1.8 325 880,000
.1:00 80 F 1.9 315 830,000
.1:52 79 TJ 1.8 320 680,000.
2:37 79 1.6 305 690,000
3:21 83 1.9 310 69U,000
4:10 90 .1.9 310 700,000
5:10 84 1.3 300 520,000
h!l.3/72 10:55 74 15 F 10 1.7 330 960,000 3 3" 100% 72
11:30 79 F 10 325 960,000 .
-
12133 81 F 10 1.9 320 820,000
3:25 84 F ', 10 1.9 305 780,000
u:17 86 liNO'j 17.5, lC 1.9 310 800,000
5:05 9Z HN03 17.5, 10 1.9 320 800,000
h!19/72 10:35 83 15 F 15 4.9 320 340,000 3 3" 80%
11:19 85 F 4.8 320 350,000
11: 55 88 F 4.9 320
12:51 94 mr03 5.0 320
2:11 94 HN03 4.9 320
3:23 91 HN03 h.8 320
8:30 86 F 4.7 310
9:00 88 F 4.9 310
9:33 89 F 4.9 315 \V
h!20/72 9:u7 75 15 F 15 4.2 340 300,000 3 3" 8D.'
10:33 84 F 4.2 330 I
11:17 86 F 4.3 330
12:04 84 F 4.2 330
112: 51 88 F 4.1 330
2:16 86 F" 4.3 330
\1\
~
-------
DATE TIME DRY GAS A1.1ALGAMATOR GOLD STACK COtIDITIONS INSTRUMENT PARAMETEHS AMBIENT
TEHP. CONDITION AMOUNT STATIC TEMP. AIR FLOW AIR FLOW CELL FIRING TEMP.
PRESSURE Perkin-Elmer SIZE TEMP.
~ mm grams inches H20 VF lbs./hr. units inches Of
4/20/72 2:59 85 15 F 15 4.1 330 300,000 3 3" 80%
3:39 83 F 4.3 330 1
4:33 81 F 3.9 340
5:13 83 F 3.8 330
Gold Condition:
- - not cleaned
F - fired at 7S0oC
HNO) - rinsed in boiling nitric acid for 5 minutes
V\
-4
-------
0.5 pg .Hg
Standard amalgamated
from bottom up
Typic;al stack
gas sample
Figure 18.
58
0.5 pg
Hg
Standard amalgamated
from top down
Illustration of Differences observed in Peak Shape due to
amalgamatinG the Standard from the bottom up or the top down
-------
59
the electronics of the recorder resulting in the initial jump of 1 to 3%
of the recorder paper observed in the figure.
This permits an accurate
way to measure the time involved from irP.tial firing until the curve peaks.
The figure illustrates the increased time required for the bottom amalgamated
standard to peak.
The difference in the direction of amalgamation does
verify that we are not saturating the gold with mercury.
These initial results have shown that the manner in wbich the mercury is
amalgamated is critical.
Considering that the sample is amalgamated at a
high velocity and a relatively low concentration and the standard is
amalgamated at a low velocity and a high concentration it is obvious that
the standards and samples cannot be perfectly' matched. '~'l'he st~dards and
samples can be matched exactly if a double arr,algamation procedure is adoptf;d.
Periodically doublets and prepeaks occur.
It, is presently the opinion of the
investigators that a sharp prepeak on either the. first or second amalgamator
1s the result of water. Water is known to absorb 254 mjJ. radiation.
The
water is present from two sources; (1) condensation, and (2) chemical break-
down during firing in the induction furnace. Water was observed as a severe
interference during cold weather.
During these climatic conditions water
would condense and run down the sides of the impinger-amalgamator during
sampling and during the initial removal of the amalgamator from the train.
This water was trapped by the gold foil and wire.
This condition would also
be a factor of the percent moisture in the stack.
When the amalgamator was
fired the water volatilized before the mercury and formed the prepeak.
This
,
1s eliminated by drying the amalgamator before analyses.
The second source
-------
60
appears to be from the firing of H~04 condensed on the gold. At times these
prepeaks occur repeatedly, but geners.lly they a!>pe8r to occur randomly,
usua1~ from only one of the two amalgamators.
,
If we are near the dew point
of sulfuric acid mist this would account for this phenomenon occurring on
only one of the amalgamators.
The prepeaks are distingushed from the true
mercury peaks by their sharpness and quickness in being volatilized during
firing. Figure 19 contains a sample peak illustrating the prepeak, and a
peak obtained from a standard solution.
The theoretical water peak may
be distinguished from ,the mercury by measuring the time evolved from the
start of firing.
This is illustrated in the figure permitting easier inter-
pretation of the results.
A slightly lower firing temperature tends to
, further separate the two peaks.
A magnesium perchlorate drying tube was
inserted between the amalgamator and the absorption cell which did eliminate
the prepeak while in use.
After several good runs it also absorbed the
mercury.
Periodically doublets will occur near the top of the mercury peak.
Both
peak tops are similarly shaped and do correspond to the standard in both
shape and time evolved after firing.
The doublets are simi1ar to those
observed with standards in the ~, cell.
These are a function of the
amalgamation reaction and the accompanying release during firing of the
amalgam.
The doublet would be eliminated by either using a larger volume
cell or by double amalgamations.
2- Interferences
Interferences may be classed as (1) also absorbing 254 m ~ radiation, or
-------
Figure 19.
H20
61
Hg
Illustration of a hiCh Prepeak due to Water Absorption resulting
from the chemical breakdown of Sulfuric Acid dllrin~ firing.
-------
62
(2) prohibiting the mercury from reaching the absorption cell b.1 either a
chemical reaction or. physical adsorption. Water, resulting from condensation
on and. above the gold in the ~gamator, does produce a prepeak which
may be distinguished from mercury or eliminated by drying. Condensed
sulfuric acid mist evolves water vapor during firing and also appears
to absorb a portion of the volatile mercury. When a sample containing
this interference is fired the prepeak is observed and a white precipitate
forms on the amalgamator wall above the induction coil. The white precipitate
is water soluble and the wash solution is very acidic and contains mercury
when analyzed by the same procedure as the standard. It is our theory
that the white material is recondensed sulfuric acid mist containing some
mercuric sulfate. Sulfuric acid has been adued dropwise to the gold and
then fired producing identical results. These washings from the amalgamator
precipitated with barium produced a sul1'ate x-ray pattern. Verification
of this interference has been difficult due to its irrational periods of
occurrence. The SnC12 scrubber solution in the first impinger was added to
the sampling train to eliminate this interference. It greatly decreased
the observable amounts of the white precipitate but did not completely
eliminate it or the prepeak.
Insufficient time .was available to optimize
the scrubber solution or to try several other procedures including rinsing
the gold with water and redrying it between sampling and firing in the
induction furnace. The United States Geological Survey developed a method
for directly amalgamating mercury out of water solutions strongly suggesting
that the amalgam is stable in water.10
,
-------
63
3- Gold Deterioration
Gold deterioration can result from (1) chemical reactions with constituents
of the stack gas stream and (2) pl~sical adsorption or condensation of
volatiles in the gas stream which are either chemically broken down or not
completely revolatilized during firing.
Gold is a very stable element and
forms very few compounds. It does fonn the chloride.
AuCl decomposes to
AuCl) at l700C.
AUC1) melts at 254°C and sublimes at 26$oC.
To verify the
existence of chlorine and/or chloride in the stack gases the first impinger
was filled with a silver nitrate solution. The solution turned white
immediately upon the start of sampling. Joensuu9 has mentioned chloride
as an interfere~ce in the double gold procedure for mercury in coal, and
he eliminates it by collection on heated silver foil as the chloride.
(The silver chl.oride is stable to much higher temperatures than the gold
chloride.)
The results obtained in this study have not shown any reaction
specifically designating chloride as adversely affecting the gold.
The
SnCl2 solution in the first impinger would theoretically reduce any chlorine
to the water soluble chloride. A large amount of chlorine could deplete
the reducing capability of the stannous chloride accounting for the small
mercury residue observed in the scrubber solution after sampling.
The gold does deteriorate with increased usage in the stack.
The peaks of
standards run on gold used extensively in the stack show a lowering and
broadening of the peak.
Additional firing of the gold after the mercury peak
has returned to base level with old gold results in a second low broad peak.
Initially the gold was boiled in concentrated nitric acid after each usage
-------
64
in the sampling train.
This did not rejuvenate the gold. Firing in an oven
at 7500c for three plus hours did rejuvenate the gold. The reaction
accounting for this is unknown.
Condensation of volatile organics, etc.,
on the gold is a possibility. Gold deterioration does not increase at a
constant rate but appears to be a function of the coal used and the firing
efficiency of the boiler.
Some days the gold will retain its sensitivity for
the complete day, other days it will be good for only two to three runs.
From the unpredictable results that were observed it is suggested that the
gold should be fired in an oven after each usage in the stack. Based on
the studies at these two power plants this precaution will eliminate any
problems from gold deterioration..
4- Absorption Coll
The maximum concentration obtainable in the linear portion of the analytical
curve with the 3" and Ch!' absorption cell is approximately 1 ~g per sample.
This corresponds to a 5 to 10 minute sampling period at isokinetic rates.
The ~, cell permits analyzing up to 7 - 10 ~g, which would be acceptable
for extended isokinetic sampling. This ceil, which approximates measuring
concentrations instead of total mercury, accentuates any dou~lets or other
idiosyncrasies in the peak shape.
For this cell to be used as a peak
height cell the mercury must be distributed identically and evenly with
each sample and standard.
This cannot be achieved with the present procedure.
Modifications permitting this desensitized cell to be used would include
(1) double amalgamation and (2) a mixing chamber between the induction
,
furnace and the absorption cell.
Alternate methods, decreasing the
-------
6$
sensitivity, include a high volume peak height cell vith a short optical
path and a cycl~ng system.
Complete 1JI1 Ti nc is cri ticaJ. in the high
volume ceil, and some fOnl! of .bafflesto assure mixing would probably be
required.
As has been shown, the standard method of desensitizing an anaJytical optical
system by shortening the path length is not directly applicable in these
circumstances.
Based on these results the Laboratory Data Control mercury
unit would require some extensive modifications and studies before it could
be used. This problem and some of the other problems encountered may be
solved by using a cycling system where the mnasured mercury concentration
would be a function of the available space ill the cycling system.
In
practice, the sample would be fired identic~LlY to the present procedure.
The mercury would then by cycled through the cell with a small pump until
equilibrium was achieved.
The sensitivity of the system would be directly
related to the volume that the sample is cycled through. This would
permit using any of the described cells and also the Laboratol"1 Data
Control unit.
This would eljminate the need for double amalgamation and
would greatly desensitize tne system pennitting extended sampling periods.
5- Stack Turbulence
Stack turbulence, very prevalent at the first plant, may affect the
effioiency of removal of mercury on the. first amalgamator.
Increased tur-
bulence ma.r cause a proportional increase in bypass of the first amalgamator,
,
resulting in a greater mercury recovery on the second or backup amalgamator.
No attempt was made to obtain a quantitative measure of the turbulence.
-------
66
The turbulence was qualitatively interpreted fram fluctuations observed in
the static stack pressure" during sampling and from the air flow recordings
obtained from the plant. Increases or decreases in loading occurred
frequently at the first plant. and resulted in changes in the speed of the
ID fan and the damper setting. A:ny problems resulting from turbulence
may be corrected by increasing the amount of gold.
6- Sampling Period"
Samples were obtained and analyzed at 5, 10, 15, and 30 minute sampling
periods. The analytical interpretation of the data became progressively
more difficult the longer the saropling period due to (1) working in the
nonlinear section of the analytical. curve, a:ad (2) the increased likelihood
of turbulence and interferences. Gold deterioration was not observed ~th
the longer runs, but this ~ have been masked by other problems. Figure 20
illustrates the effect of increased sampling periods. Under optimum conditions
.
the values are very consistent. The results must be compared on a daily
basis due to bin loading occurring every one or two days resulting in a
different grade of coal at both of the plants investigated. Analyses of
coal samples from these seams obtained through the United States Bureau
of Mines 'round robin' have shown a range of 0.1 to 0.44 ppm mercury, a
four fold difference.
7- Flow Rates
Samples were collected non-isokinetically at 0.2, 0.3, 0.4, o.S, 0.6, 0.7,
,
and 0.6 Cn!. Figure 21 illustrates the effect of flow rate as a function
of the observed concentration (concentrations are not corrected for
-------
67
IJgjli.tel' Hg
0.03
0.0
Site 1
o
0------:/0
0.0
0- - - - C
Site 2
__0
c--
o
o
5
10
15
30
, .
Sampling Time -Minutes
.
Figure 20. The Observed Hercury Concentration as a Function or Sampling Time
-------
0.2
March 23
C
J,lg/Iiter Hg
0.1
o
o
0.2
0.1
o
C 0
o [C
.
April 12
.
.
.
G
o
cC
Site 1
~A
A
Site 2
April 20
0.5
Figure 21. The Observed Mercury Concentration as a Function of Sampling Rate.
~
A
~
0.3
0.4
Sampling Rare - C;:;y',
ex>
0.6
March 20
o
.0
L:iL:i
0.7
.
.0
0.8
0',
co
-------
69
standard atmospheric conditions).
Although there is a general fluctuation
of the concentrations, there appears to be no significant difference in
the recovery ratios at the various sampling rates. This should be reverified
at a plant operating at a constant loading and with a consistent grade of
coal before final acceptance.
Under optimum conditions there will be no
problems with sampling at isokinetic rates.
8- Permanganate vs. Gold Trains
Samples were collected using trains contidning gold backed by permanganate,
and permanganate backed by gold.
The results of these studies are extracted
from Table 3 and are presented in Table 5.
During short sampling periods
there were no significant differences observed ,iith either the permanganate
or the gold.
TiJere are some problems in comparing, the data from the runs
containing a mixed train since only an aliquot of the permanganate solution
is analyzed and all of the gold is analyzed.
The sampling period and flow
rate cannot be optimized for using both recovery systems simultaneously.
Sensitivity problems With the gold we:reobserved during the high volume
.. . .
(290 liters) extended sampling period of the first run on Mar~ 28.
This
was expected with the present analytical procedure.
The results of the mixed trains and the permanganate and gold trains showed
that the gold recovery is as efficient as the permanganate.
Based on the
assumption that the permanganate does recover essentially all the mercury
when
reduced by 502' it is concluded that the gold also recovers
essentially all the mercury when a backup amalgamator is used.
The gold
-------
70
is not affected by high 502 concentrations and has been satisfactorily used
to collect mercury in'6~ 502.
Table 5. Results obtained from KMn04-Au and Au-KMn04 sampling trains
Date Time Position 1 Position 2 Total Concentration
~g ~g ~g . ~g7l..
3/25/72 1:40 K 2.3 Au - 2.3 0.024
3:20 K 1.96 Au - 1.96 0.021
4:03 K 1.89 Au 0.078 1.97 0.022
5:10 Au 1.9 K 0.22 2.12 0.028
5:30 K 1.51 Au - 1.51 0.022
6:20 Au 1.1 K 0.21 1.3 0.018
3/27/72 12:15 K 2.5Y Au 0.69 3.28 0.025
1:18 Au 2.3 K ~0.5 <.2.8 <0.023
2:17 Au 4.7 K 0.4 5.1 0.027
3:05 K 3.93 Au 0.59 4.52 0.024
4:04 K 5.96 Au 0.08 6.35 0.025
5:34 Au 4.68 K 0.65 5.33 0.022
3/28/72 9:53 Au 1.69 K 0.39 2.08 0.006
10:49 K 4.43 Au 0.16 4.59 0.025
1:17 K 2.50 Au 0.29 2.79 0.040
2:00 Au 1.85 K 0.90 2.95 0.013
2:43 Au 1.93 K "0.3 1.93 0.. 0.J.4
9- Efficiency
The efficiency of recovery of volatile mercury by gold was investigated by
interpreting the percent recovery of mercury on the first amal~amator where
two amalgamators were used in series. The data is presented in Table 1.
The gold was fired each night resulting in rejuvenated gold at the start of
each day. Four amal.gamatorswere prepared each day and were cycled through
the runs using two during each sampling. Under optimum conditions, 90-98%
recovery was obtained on the first amalgamator. The parmueter. most seriously
-------
71
affecting the percent recovery was gold deterioration. Figure 22 contains
the percent recoveries from two days runs as a function of the usage of the
gold. The recovery of the eff~ciencies on April 19 is a result of
rejuvenating the gold by firing at 7500C for ~ hours between the runs.
The relationships are not immediately obvious in each days run due to the
effect of other parameters on the results and different procedures utilized
in the treatment of the gold during various stages of the study.
10- Particulate Analyses
Particulate samples collected on April 13, at Site 1 and on April 19 and 20
at Site 2 were analyzed for mercury by the gold amalgamation flameless
atomic absorption procedure identical to that used for the standarde. The
fiber glass filter sRnlples collected in the standard EPA sampling train
were digested by (1) perchloric acid-nitric acid digestion in a Bethge
apparatus, and (2) volatilization of the mercury by combustion in an
induction furnace accompanied by absorption of the volatilized mercury in
a nitric-perroanganate solution.
The perchloric acid digestion procedure consisted of adding 5 ml. of
concentrated HNO) Rnd 10 rol. of perchloric acid to the filter paper in a
Bethge apparatus.
Digestion was carried out by collecting the condensed
volatiles until the solution reached 198°C and then refluxing for ten
minutes.
These samples were reduced with 5 ml. of the standard stannous
chloride solution and analyzed by the same procedures as the standards.
The induction furnace procedure consisted of folding filter paper and
-------
% Recovery
72
100
90
80
0
70
60
50
40
30
20
March 20-0
10 April 19 -0
0
1 2 3 4 5 6 7 8 9 10 11 12
Run Number
,
Figure 22. The Percent Recovery of Mercury on the Firat /unalgamator
-------
73
placing it in the quartz-graphite crucible whieh was fired in the induction
field at an 80% variac setting.
An air strearo carried the volatilized
mercury through the bubbler i~ustrated in Figure 1 where the mercury was
oxidized by an acidic permanganate solution t20 ml. saturated KMn04' 5 ml.
cone. HN03' and 25 mi. distilled water) which removed the merc\~ from the
air stream.
The excess permanganate was reduced by adding 5 rol. of a
10% (W/V) hydro~lamine hydrochloride solution.
The mercury was then
reduced with 1 JUl. of the standard stannous chlorid~ solution and analyzed
by the same procedure used for the standards and the perchloric acid
digestions.
Because a whole. filter paper was used for each analysis the reproducibility
of the results could not be checked. Although similar results were obtained
with the perchJ.oric acid and induction furnac:e procedures, the induction
furnace method was .favored by the analysts. This preference was based on
the inability to adequately control the reduction reaction.
The results
from the induction furnace procedure are in Table 6.
The mercury concen-
trations are stated as a function of the gas volume because the weight of
particulates collected during a 5 minute sampling period was too small to be
measured accurately.
The erratic results on April 13 are due to a fire in the
coal bunker at the power plant res\~ting in very high partic\uate concentrations
when the water soaked coal was burned. The results frOM April 19 and 20
illustrate the differences before and after an electrostatic precipitator.
-------
74
Table 6. Particulate analyses by induction furnace-permanganate digestion
Date Time Mercm-y Location
Concentration
~g/filter ~g/M3
paper
April 13, 1972 10:55 A.M. .07 .98 Cyclone Outlet
11: 30 A.M. .03 .40
12:33 P.M. .40 5.7
:.4:1.7 P.M. .12 1.7 , ,
April. 19, 1972 9:30 P.M. .075 .95 Precipitator
Outlet
April 20, 1972 9:47A.M. .05 .35
10: 33 A.M. .075 .51
11:17 A.M. .06 .68
12:04 P.l-i. .06 .62
12:51 P.M. .02,. .60
2 :16 P .Jot. .09 .40
2:59 P.M. .09 .86
3:39 P.M. .03 .43 'v
4:33 P.M. .185 3.3 Precipitator
Inlet
5:13 P.M. .119 1.4 ~
,
-------
CONCLUSIONS
A method for collecting an~ analyzing mercury in power plant stack gases
has been developed.
The collection procedure consists of modifying a
-standard EPA isokinetic sampling train, utilizing two modified impingers
containing 10 grams of gold chips to collect the mercury vapor as an amalgam.
Particulates containing mercury are collected on a fiber glass filter
upstream of the gold amalgam.
The analytical procedure consists of
transferring the modified impinger containing the mercury amalgam to an
induction furnace.
The induction field instantaneously volatilizes the
mercury and an air stream carries it through a quartz optical cell
positioned in the 254 mlJ. mercury line of an atomic absorption spectro-
photometer.
The mercury concentration is calculated from a peak obtaine,d
on the strip chart recording of the absorption pattern.
The sampling train utilizes two amalgamators in series as a method of
verifying the efficiency of recoverY of the first amalgamator.
Under
optimum conditions the percent recovery on the first amalgamator i5 about
95~. Gold deterioration and, pos~ibly, stack turbulence will lower this
recovery ratio. A very recent study by the U. S. Bureau 01" Mines 7 has also
suggested that when the gold is warmed by the stack gas, small amounts of
mercury are revolatilized by a flowing air stream and are carried to the
second ama1ga.'1lato~. This would be in agreement with the observed results
suggesting a higher recovery when the system is cooler.
, .
7$
-------
. 16
The analytical procedure was found to be too sensitive resulting in a
maximum sampling period of 5 to 10 minutes. Several methods of
desensitizing the procedure have been suggested.
,
Although the procedure does require further refinement, accurate results
are obtainable from stack gases at this time. Presently the authors are
not able to differentiate between the range of the analytical values
observed and the range of the results due to fluctuations in the coal
consumption and the coal composition.
The sampling and analytical
procedure has been developed to the position where an accurate mercury
concentration can be obtained within two days of on site sampling. The:
first day is spent in finding the range rir the mercury concentration and
the second dai1 in obtaining the usable re:;ults. The procedure has; not
been sufficiently developed to permit shipping the saroples back to the
stationar,r laboratory for analyses after the sampling is completed.
,
-------
77
FUTURE INVESTIGATIONS
utilizing the procedure recommended in this study, gold amalgamation has
been shown to be an efficie,nt method of collecting several micrograms
of volatile mercury under the S02 conditions observed in power plants.
This investigation has shown the analytical procedure to be the
limiting factor: the peak height celis are too sensitive and the
integrating or f101l through cell is too dependent on the exact amal-
gamation reaction prohibiting matching standards and samples. Immediate
future work should be oriented toward either improving the present
system or redesigning the system to permit a lower sensitivit.Y.
Present System
(1) The present. system may be improved by utilizmg
a large volume-short :9ath length peak height cell.
This cell will desensitize the system, but baffles
~y be required to approximate a homogeneous gas
mixture.
( 2) A gas mixing chamber between the amalgamator and
the optical cell mBiY provide a homogeneous enough
mixture to permit using the present integrating
cell or the large volume cell described in item 1
without baffles.
(3) Double amalgamation would permit amalgamating the
standards and sarop1es identically, theoretically
permitting the use of any of the systems investi-
,
gated. The procedure would consist of transferring
-------
78
the gold from the ir.1pinger into another amalgamator
tube, firing the amalgam, recollecting the mercury
on another 8.1119lgamator above it, and then firing the
second amalgamator carrying the volatilized mercury
into the optical cell. The standard would then be
aerated onto the first amalgamator, fired onto the
second amalgamator and then fired into the optical
cell. Both the standards and samples would be
amalgamated onto the upper gold at the same fl~
rate and relative concentrations resulting in a
similar release when the second amalgam is fired.
:t-todified System
(4) A cycling system would permit desensitizing the
system by increasing the volUITle that the volatilized.
mercury is cycling through. The procedure would
consist of volatilizing the mercur.y by firing the
. .
amalgamator and then, by the use of a one way valve
and a pump, cycling the mercury through a closed
loop containing the optical cell.
Instead of a peak,
a line would result which, after the gas had become
homogeneous, would correspond to the mercury concentra-
tiono A method of desensitizing the system during
cYcling could be done by increasing the volume of the
system with a large modified syringe. This unit wou.ld
permit greater flexibility with an individual sample
since one sample can be diluted during analyses if it
has a high concentrationo
-------
79
Additional study of the following areas would result in a more efficient
sampling and analytical procedure.
Analytical Method
1- The behaVior of sulfuric acid mist, including the
chemical breakdown of sulfuric acid during firing,
should be more complete~ investigated.
2- Additional study of the behavior of the gold during-
amalgamation coul.d result in a smaller quantity of
gold being required and, possibly, a higher recover.y
efficiency on the first amalgamator.
3- The induction furnace combustion procedure for
measuring mercury in the particulates appeared to
work very well, but additional work should be perforn-.ed
at higher particulate concentrations (longer sampling
period or higher particulate loading in the gas
stream). This would permit obtaining splits of the
filter paper for duplicate analyses.
4- The parameters investigated for sample storage and
shipment should be expanded and more thoroug~ studied.
$- The feasibility of incorporating the Laboratory Data
Control mercury monitor into the analytical procedure
should be investigated.
Sampling Train
1- A better support system for the gold in the amalgar.1ator
,
-------
80
should be developed.
2- The effect of the probe and the filter box ternperature3
should be investigated under controlled conditions
in the laboratory.
3- Additional work is required to verify the need for and
then optimize the scrubber solution in the sampling
train.
In conjunction with the modification of the analytical procedure a better
source facility containing a constant loading and a consistent fuel supply
will be advantageous for additional studies.
-------
81
APPENDIX I
MERCuRY-GOLD .A1-1ALGAM SHIPPING CHARACTERISTICS
The adaptation of the procedure investigated in this stu~ for routine
monitoring of mercury in stack gases is dependent upon the capability of
shipping the amalgam from the sampling point to the analytical facilities.
To test the feasibility of shipping the amalgam under adverse conditions
gold foil was amalgamated with known quantities of mercury and stored in an
oven at lOOoC for 16 hour periods.
These amalgSl'ls were stored in either
the original amalgamators or in porcelain crucibles.
After the 16 hour
period, the amalgamators were run directly by flarneless atomic absorption
followed by identical standards run on the same amalgamator. The smnples
stored in the c:rucibles were poured back into an amalgamator and analyzed
followed by a standard.
Table 7 contains the peak he~ghts of the samples and standards stored in the
amalgamators. '\-lith the exception of two of the runs the results are with-
in the present experimental error.
The amalgams that were transferred into
the crucibles showed considerable variation reflecting the redistribution
of the amalgamated gold foil in the non-amalgamated foil.
The peak heights
were in between. the same standards amalgamated from the top down and from
the bottom up.
These results verified the initial conclusion that the stand-
ards must be amalgamated onto the gold by a procedure closely approximating
the stack sampling procedure.
, .
It is concluded from this limited stu~ that under the conditions investip,ated
-------
82
the amalgam is stable for extended periods of time. If the amalgam is to be
transferred into another container for shipping a double amalgamation or
cycling procedure will have to, be adopted to account for the redistribution
of the amalgam.
Table 7. Results of the effect of storing the amalgam at 100°C for 16 hours
in the amalgamator.
Percent Absorption
Sample Standard
O.:hJ.g 34.7 19
33.5 26 .
o.~g 39.8 38.9
4408 \ 35
O.5~ 73.2 70.8
54.4 60.2
1.0~ 85.5 86
2.0~ 89.5 82.5
,
-------
/
/0
o
/
/0
/0
/0
/0
A ir Flow
liters/min.
4.0
3.2
2~
1.6
0.8
83
APPENDIX. II
o
1
2
7
3
4
5
6
Air Flow
Perkin Elmer Flowmeter
-------
84
REFERENCES
1.
Pollock, Gene.
Kennecott Copper Company - personal cOmmunication.
2.
Knapp, Norman.
Ohio Geological Survey - personal communication.
3. Rueh, R. R., Gluskoter, H. J., and Kennedy, E. J. Mercury Content of
Illinois Coals, Illinois Geological Survey, Environmental Geology Notes,
Number 43 (1911).
4.
Kalb, G. vi. The Determination of Hercury in Coal by Flameless Atomic
Absorption, presented 163rd. National Meeting American Chemical Society,
Division of Fuel Chemistry 16(3), 1972.
S. unpublished studies by the authors
6. Booth, M. R. Mercury Emissions from Generating Stations, W. P. Dobson
Research Laboratories, Ontario Hydro, Toronto, Canada (1971).
7.
Diehl, R. C.~ Hattrnan, E. A., Schultz, H., and Haren, R. J. Fate of
'I'race Mercury in the Combustion of Coal, Bureau of Mines Technical
Progress Report TPR 54, 1972.
8.
unpublished studies by the authors
9.
Joensuu, o. J.
Mercury Vapor Detector, Applied Spectroscopy l2(5), 1971.
10. Hinkle, M. Eo, and Learned,. Ro E. Determ1Dation of Mercury in Natural
Waters by Collection of Silver Screens, U. S. Geological Survey Prof.
Paper 6S0-D, 196y.
------- |