Collection Efficiency Evaluation Of Mercury-trapping Media For The Sass Train Impinger System


United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-84-089  Sept. 1984
Project Summary
Collection  Efficiency
Evaluation  of  Mercury-Trapping
Media  for  the SASS Train
Impinger System

A. D. Shendrikar, Ashok Damle, and W. F. Gutknecht
  This report reviews results of a project
that involved generation  of mercury
atmospheres of known and stable con-
centrations and evaluation of the ef-
ficiency of mercury collection using the
following trapping media:

 • 10 percent hydrogen peroxide.
 • 0.2M ammonium persulfate.
 • 0.2M ammonium persulfate+0.025M
   silver nitrate.
 • 1.5  percent potassium perman-
   ganate in 10 percent sulfuric acid.
  This Project Summary was developed
by EPA's Industrial Environmental Re-
search Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back}.

Introduction
  For purposes of monitoring and devel-
opment of control technology, it is im-
portant that accurate and precise methods
for sampling  and  analysis of source
emissions of mercury be available. The
primary focus of this project has been the
evaluation of such sampling methods.
  Commonly used sampling methods for
determination of mercury emissions from
stationary soruces include the use of the
Source Assessment Sampling System
(SASS), the EPA Method 5 train, or
similar sampling devices. In all  such
mercury collection devices, impingers
containing chemical solution effective in
collecting and retaining mercury and
other volatiles are used. Other mercury
collecting procedures utilize solid mater-
ials for collection. In general, collection
media considerations have been reagent
purity, possible interferences during
quantification, collection efficiency  at
high volume sampling, reagent stability,
cost, and compatibility with other imping-
ers in the sampling train.
  In spite of a number of  available
collection media consisting of metals,
sorbants, and absorbing solutions, poor
collection efficiences have been reported
during the stationary source  sampling.
The methodology for evaluating the collec-
tion efficiency of these various media has
been given less attention in the past—in
particular, analytical accuracy of standard
"sources" of mercury. Subsequently, this
work was undertaken (1) to develop an
accurate standard mercury source, and
(2) using this source, to test several of the
popularly used mercury collection media.

Experimental


Description of Mercury Test
Atmosphere Generation System
  Figure 1 shows the mercury generator
schematically. It is essentially based on a
saturation technique reported by E.  P.
Scheide et al. in 1979, with some modi-
fications. A stream of carrier gas (nitro-
gen) is passed through a dryer and a 60
jum filter to remove impurities. The nitro-

-------
gen stream then passes through a flow
meter and a regulating valve before it is
split into two streams. One stream goes
through a regulating valve and a high
flow rotameter and then to a dilution flask
(Dilution Stage I in Figure 1). The other
nitrogen stream flows through a regula-
ting valve, through a low flow rotameter
and into the generator which consists of:

• Mercury reservoir/vaporizer. A spe-
cially designed flask which contains a
pool of liquid mercury (30-50 g) that is
electrically heated and temperature
controlled.

• Condensers. Two Graham condensers
wrapped in insulating material and
connected in series, through which
cold water is continuously circulated.
The temperature of the condensers is
controlled to at least ± 0.1 °C by
connecting them to a water source
from a constant temperature bath
capable of maintaining temperature
down to -5°C. The temperature of the
exit gas containing mercury is moni-
tored with a thermometer at the top of
the second condenser.
• After leaving the condensers, the gas
stream containing mercury passes
into a 1000 ml flask where it is mixed
with additional nitrogen to produce the
desired levels of airborne mercury.
• The gas stream exiting from the dilu-
tion flask may be passed through a
quartz cell placed in the light path of an
atomic absorption spectrometer (AAS).
This arrangement allows a direct meas-
urement of generated mercury vapor
concentrations (following calibration
of the AAS system). The exhaust gases
from the quartz absorption cell are
passed into an absolute mercury trap
which contains a mixture of activated
charcoal and Hopcalite.

Generator Performance
Evaluation
Initial experimental work with the
generator included evaluation of its cap-
ability to produce stable atmospheres of
mercury. For this, the output from the
dilution flask was passed directly through
the quartz cell placed in the light path of
the atomic absorption spectrometer (AAS)
which provided readings in absorbance
units. These absorbance readings were
monitored as a measure of the stability of
the mercury atmosphere generation.
Through repeated attempts for steady
mercury atmosphere generation, the fol-
lowing observations were made:

• It is not necessary to heat the mercury
reservoir for generation of mercury
atmospheres in the range of 10 to 100
fjg/m3, which is the estimated range
of source emissions.
• A steady (± 5 percent) low flow of
nitrogen through the mercury reservoir
is required (range 0.25 to 1.85 L/min)
for the desired mercury concentration
ranges.
• Maintenance of a precise temperature
of 2°C ± 0.1 °C at the condensers is
found to be essential for stable mer-
cury atmosphere generation.

Using the above conditions, test mer-
cury atmospheres were generated and
monitored using the on-line AAS. Figure
2 shows the steady generation of mercury
for more than a 5-hour period, as mon-
itored using a strip chart recorder.
Exhaust
Mercury
Trap
Carrier
Nitrogen
•*• Dilution Nitrogen -*•
-* Carrier Nitrogen -«.
Mercury
Reservoir
and Heater
Exhaust
Pressure
Gauge
Mercury
Trap
Figure 1. Schematic of mercury generation system.

-------
Quantification of Mercury
Levels in the Test Atmospheres
To determine concentrations of mer-
cury in the generated test atmospheres,
the system connections and tubing after
the dilution flask were modified by in-
corporating two three-way valves, one
after the dilution flask and one before the
quartz absorption cell. With these chang-
es the test atmospheres containing mer-
cury could be passed through the quartz
cetl or, when required for concentration
determinations and/or collection effici-
ency evaluations, through a series of
impingers. The exit gas from the last
impinger could be directed to the quartz
cell via a moisture trap and a second
three-way valve.
The mercury vapor concentrations were
experimentally determined by collecting
vapor for a known period of time in two
impingers in series and then analyzing
the impinger solutions. The collection
media was 1.5 percent potassium per-
manganate in 10 percent (v/v) sulfuric
acid; 50 mL of this media was contained
in each 75-mL capacity impinger. The
sampling rate was 980 cc/min. The exit
gas from the second impinger was fed
into the quartz absorption cell during the
sampling process via a moisture trap
which contained magnesium perchlorate.
The mercury in the impinger solutions
was determined using the procedure in
EPA-600/7-78-201. This procedure in-
volves chemical conversion of all mercury
present in solution to the elemental form
and sweeping this Hg° into the quartz cell
of the spectrometer with a stream of
nitrogen. Standards were prepared by
spiking collection media with known
quantities of mercury ion. Quantification
was performed through comparison of
absorption peak areas obtained with the
impinger and standard solutions. All
samples from the second impingers were
found to contain mercury levels below the
detection limits of the method which was
0.05 jug of mercury.

Further experimental work included
generation of test atmospheres which
contained varying levels of mercury. This
was achieved by increasing and/or de-
creasing the dilution ratio. Samples were
collected and analyzed to establish a
relationship between experimentally de-
termined mercury concentrations in the
test atmospheres and absorbances of the
test atmospheres. Figure 3 shows an
approximately linear relationship be-
tween experimentally determined mer-
cury concentrations and AAS absorbance
readings for mercury concentration below
600 jug/m3.
Results and Discussion
With the successful generation of
known and stable mercury atmospheres,
evaluation of its collection efficiency at
high volume sampling rates became the
objective of the project work. For this
experimental work some changes were
made, including:

• Addition of a second 1000 mL dilution
flask, following the first dilution flask
(see Figure 1), through which filtered
compressed air was passed at a high
volume rate, 3-6 cfm (85-170 L/min).
• Use of three SASS train impingers in
series, replacing impingers in the
sampling system discussed earlier.
• Incorporation of a high volume cali-
brated sampling pump.
• Incorporation of one additional three-
way valve after the second dilution
flask so that test atmospheres with or
without mercury could be passed
through the impingers and then into
the quartz cell via a moisture trap.
With the changes made in the system
for sampling at SASS sample collection
rates, the following media were tested for
mercury collection efficiency:
oo
01
§
b

-------
0.300
0.200
I
I
0.100
(NOTE: Mercury concentration determined by collec-
tion into two impingers each containing 50 mL of
1.5% KMnOt in
0 100 200 300 400 500 600 700 800 900 1.000

Mercury Concentration, vg/m3

Figure 3. Absorbance vs. generated mercury concentration.
(1) 10 percent hydrogen peroxide (v/v).
(2) 10 percent hydrogen peroxide in 10
percent sulfuric acid.
(3) 0.2M ammonium persulfate (w/v).
(4) Freshly prepared solution of 0.2M
ammonium persulfate + 0.025M
silver nitrate (w/v).
(5) Aged (48 hours) solution of 0.2M
ammonium persulfate + 0.025M
silver nitrate.
(6) 1.5 percent potassium perman-
ganate in 10 percent sulfuric acid.
Figure 4 depicts typical sampling and
medium collection efficiency events as
recorded on the strip chart recorder when
using 1.5 percent KMnO4 in 10 percent
H2SO<. Sampling time was 1 hour, and
each impinger had 250 mL of trapping
medium.

Collection Efficiency
Measurements
Utilizing the mercury atmosphere gen-
eration system and the sampling ap-
proach described, the collection efficien-
cies of the above mentioned media were
investigated. Data obtained are summar-
ized in Table 1.

10 Percent Hydrogen Peroxide
Three SASS impingers in series, each
containing 250 mL of 10 percent hydro-
gen peroxide solution, were used for
collection efficiency evaluations. Three
sampling runs were made at test atmos-
phere mercury concentrations of 70
/jg/m3, and one run was made at a
mercury concentration of 10 //g/m3. The
collection efficiency of the medium was
poor; i.e., only about 20 percent. Further
experiments at low mercury levels with
acidification of the hydrogen peroxide
with 10 percent sulfuric acid did not
improve its collection efficiency, which
remained about 20 percent.
During this effort, the procedure given
in EPA-600/7-78-201 was followed for
analysis of the hydrogen peroxide med-
ium. This involved taking a 10 mL aliquot
of the hydrogen peroxide solution and
adding 5 percent acidified permanganate
solution to destroy excess peroxide. Ex-
perience indicates that a considerable
volume of permanganate solution needs
to be added to destroy the excess per-
oxide; this point should be emphasized in
EPA-600/7-78-201, the EPA Level 1
manual.
Although inefficient, the collection of
about 20 percent of the mercury vapor by
the hydrogen peroxide impingers must be
considered significant in terms of mercury
distribution in the SASS train impinger
system. Generally, hydrogen peroxide
impingers in the SASS train or EPA
Method 5 train precede mercury collec-
tion impingers. This project work indi-
cates that these hydrogen peroxide im-
pingers also need to be analyzed to get
true concentrations of mercury emitted
from a stationary source.

0.2M Ammonium Persulfate
The collection efficiency in this medium
at the high mercury concentration was
29* percent and at the low mercury
concentration was 25* percent.

0.2M Ammonium Persulfate with
0.025M Silver Nitrate
Duplicate sampling runs at both mer-
cury levels showed better than 99 percent
collection efficiency with the three-im-
pinger SASS sampling system. However,
one impinger showed only 74* percent
collection efficiency. This means that, for
'Estimated coefficient of variation is 10 percent.
-------
1
m
I
!
.1
Q
"o

!
o
I
9
O)
.c ^
5. c
S.S
I
I
I
I
|

fe
N
]'

|9
S-o
$ 6
1 1
w
£
I
v>
\
o
05
§
O
\
^
%
1

\
J
s
9
S?
8
o
\

8
9
8
b
\.

O
c
£

\
\~—

05
§
b
J
1 s
cS 9
I1 §
•| 8
•5
u
Time

Figure 4. Recording of absorbance vs. time at various points in the mercury vapor generation/sampling system.
better than 98 percent collection effi-
ciency, three impingers of this medium
are required.
Further collection efficiency evalution
work included aging of the0.2M ammon-
ium persulfate/0.025M silver nitrate
solution for 48 hours in a refrigerator and
then using it in the three impingers. The
collection efficiency of the medium at
high mercury concentration was found to
have gone down to 82* percent and for
low concentration was 79* percent.

1.5 Percent KMnO4 with
10%H2S04
Three SASS impingers, each contain-
ing 250 mL of this medium at high (70
fjg/m3) and low (10 fig/m3) mercury
levels in test atmospheres, were tested.
For both concentrations, the collection
efficiency was found to be better than 99
percent with the three-impinger sampling
system. One impinger alone, containing
250 mL of acidified potassium perman-
ganate, was found to collect 94* percent
of mercury from the test atmospheres
containing high levels of mercury; this
two-impinger system appears adequate
for optimum collection efficiency.
'Estimated coefficient of variation is 10 percent.
Table 1. Summary of Mercury Collection Efficiency of Various /Wed/a"
Average Average
Number of Sampling Rate Collection
Medium
10% hydrogen peroxide
Acidified 10% hydrogen peroxide
0.2M ammonium per su If ate
Freshly prepared 0.2M ammonium per su If ate
+ 0.025 M silver nitrate
Freshly prepared 0.2M ammonium persulfate
+ 0.025M silver nitrate
Aged 0.2M ammonium persulfate + 0.025 M
silver nitrate'
1.5% potassium permanganate in 10%
sulfuric acid '
1.5% potassium permanganate in- 10%
sulfuric acid
Runs
4"
1
2*

1"
cfm
4.25
3.93
3.75

4.00

4.33

3.75

4.67

5.24
Efficiency, %
20.9
18.2
26.8

>99.0

73.3

80.1

>99.0

94.4
^Three-impinger system was used unless otherwise specified. Sampling times in most cases were
60 minutes; exceptions are mentioned in the text. Room temperature was in the range of20-25°C,
and data on efficiency are based on relative absorbance readings at the inlet and outlet of the
impingers.
''Includes one run at low mercury levels.
"Includes two runs at low mercury levels.
''Only one impinger containing 250 mL of medium was used.
"Aged for 48 hours in the refrigerator.
'Sampling times less than 60 minutes. See details in text.
-------
The colleciton efficiency calculations
are based only on the relative absorbance
readings of the AAS taken at the inlet and
outlet of the impingers. This method of
calculating collection efficiency is con-
sidered valid in view of the fact that a
linear relationship between AAS absorb-
ance readings and mercury concentra-
tions in the test atmospheres was estab-
lished.

Conclusions and
Recommendations
In conclusion, a simple mercury gen-
erator, capable of producing known and
stable mercury-containing test atmos-
pheres, was constructed and operated
successfully. Incorporating an on-line
atomic absorption spectrometer with a
quartz absorption cell provided a method
for continuously monitoring the perform-
ance of the mercury vapor generator and
also a method for direct measurement of
collection efficiences. Of the mercury
collection media investigated, the 1.5
percent KMnO4 in 10 percent H2S04
appeared promising when tested under
laboratory conditions. The collection effi-
ciency of the 1.5 percent acidified KMnCU
solution needs to be verified through field
testing. Similarly, collection efficiency
needs to be verified by performing mate-
rial balance; i.e., by analyzing individual
impinger solutions. Although acidified
KMnCU appears to be an efficient medium
for mercury collection, it has stability and
storage problems; these need to be fully
investigated before the method can be
considered for field use.
Hydrogen peroxide was tested as a
collection medium following EPA Level 1
procedures. Although collection effici-
ency was low, two relevant observations
were made: (1) a considerable volume of
permanganate solution was necessary to
destroy the excess peroxide; and (2) since
hydrogen peroxide results in some mer-
cury collection, hydrogen peroxide im-
pingersfwhich normally precede mercury
collecting impingers in the SASS or EPA
Method 5 trains) should be analyzed for
mercury.
Finally, other mercury collection media
(e.g., iodized charcoal, Hopcalite, iodine
monochloride, and acidified potassium
dichromate) need to be studied.
A. D. Shendrikar, Ashok Damle. and W. F. Gutknecht are with Research Triangle
Institute. Research Triangle Park, NC 27709.
Frank E. Briden is the EPA Project Officer (see below).
The complete report, entitled "Collection Efficiency Evaluation of Mercury-
Trapping Media for the SASS Train Impinger System." (Order No. PB 84-243
112; Cost: $8.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
•&U. S. GOVERNMENT PRINTING OFFICE: 1984/759-102/10709
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
PS 0000329
U S ENVIR PROTECTION
HE 1,1 UN 5 LIBRARY
23$ S OEARBORN STREET
CHICAGO IL 60604
-------