United States
Environmental Protection
Agency
Environmental Monitoring and Support
Laboratory
Research Triangle Park NC 27711
EPA 600 4 80-018
March 1980
Research and Development
v>EPA
An Evaluation
Study of EPA
Method 8
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2 Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
AN EVALUATION STUDY OF EPA METHOD
by
Joseph E. Knoll and M. Rodney Midgett
Quality Assurance Division
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection Agency, and approved for publication,
Mention of trade names or commercial products does not constitute endorsement
or recommendation for use.
11
-------
FOREWORD
Measurement and monitoring research efforts are designed to anticipate
potential environmental problems, to support regulatory actions by developing
an in-depth understanding of the nature and processes that impact health and
the ecology, to provide innovative means of monitoring compliance with regula-
tions and to evaluate the effectiveness of health and environmental protec-
tion efforts through the monitoring of long-term trends. The Environmental
Monitoring Systems Laboratory, Research Triangle Park, North Carolina has
responsibility for: assessment of environmental monitoring technology and
systems; implementation of agency-wide quality assurance programs for air
pollution measurement systems; and supplying technical support to other groups
in the Agency including the Office of Air, Noise and Radiation, the Office of
Toxic Substances and the Office of Enforcement.
This investigation was conducted at the request of the Office of Air
Quality Planning and Standards. A test method for the measurement of acid
mist and sulfur dioxide emissions from sulfuric acid plants was evaluated.
The work was carried out to answer certain questions that arose in earlier
studies in which high acid mist values occurred together with low sulfur
dioxide measurements. Those studies suggested that some conditions may have
existed that caused the oxidation of sulfur dioxide to sulfate in the sample
collection apparatus. The present investigation included a study of operating
characteristics and the efficiency of components in the sampling system as well
as of some of the analytical processes. Both laboratory and field investiga-
tions were conducted.
• C73
s K. Hai
Thomas R, Mauser, Ph.D.
Director
Environmental Monitoring Systems Laboratory
Research Triangle Park, North Carolina
-------
ABSTRACT
Techniques used In EPA Method 8, the source test method for acid mist
and sulfur dioxide emissions from sulfuric acid plants, have been evaluated.
Evidence was found that the presence of trace amounts of peroxides in isopropyl
alcohol results in the conversion of sulfur dioxide to sulfate and causes a
significant positive error in acid mist values. Such peroxide contamination
was found to be common. Methods are described for measuring peroxide spec-
trophotometrically and for removing them over a column of activated alumina.
No conversion of sulfur dioxide to sulfate was observed in peroxide-free
isopropyl alcohol, in sintered glass filter supports or on glass fiber filters.
Efficient collection of sulfur dioxide took place in the hydrogen peroxide-
containing impingers specified in Method 8. Sulfuric acid mist and sulfur
trioxide were efficiently collected by a single impinger containing isopropyl
alcohol in combination with a filter. Two alternate indicators, sulfanazo III
and xylene cyanol FF, were studied and found to be inferior to thorin. Solid
ammonium sulfate was shown to be useful for the preparation of audit samples.
In field testing, paired-probe techniques showed that, when sulfur
trioxide is absent, acid mist is efficiently collected by a single filter even
when the isopropyl alcohol-containing impinger is eliminated. Both ammonia
and diemthylanaline, which are employed as gas scrubbers, cause sulfur dioxde
to be retained in isopropyl alcohol and result in large positive interferences
in acid mist values. Ferric oxide, present in the effluents of steel pickling
operations, causes a large negative interference in acid mist values.
IV
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables vi
Acknowledgments vii
1. Introduction 1
2. Conclusions 3
3. Experimental Procedures 5
Laboratory procedures 5
Field testing procedures 8
4. Results and Discussions 9
Acid mist and SCL collection efficiency 9
Efficiency of S02 collection 13
Interaction of acid mist and SCL 17
Discussion of methodology 20
Field testing 21
Evaluation of solid ammonium sulfate
as an audit material 26
References 28
Appendices
A. Method 8 determination of sulfuric acid mist and sulfur
dioxide emissions from stationary sources 30
B. Acid mist performance standard relationships 33
-------
FIGURES
Number Page
1 Acid mist and sulfur trioxide effluent simulator 6
2 Sulfur dioxide test sampling port arrangement 6
3 Train used in acid mist collection efficiency study 10
TABLES
Number Page
1 S03 Sampling Data and Collection Efficiency 11
2 Acid Mist Sampling Data and Collection Efficiency 12
3 S02 Collection in Method 8 Train 15
4 Relative FLO^ Levels in I PA from Various Sources 16
5 Test for S02 Oxidation on Glass Fiber Filters
Containing Added H2S04 18
6 Effects of S02 on Acid Mist Collection 19
7 HpSCL Interaction With Glass Fiber Filters 20
8 Method 8 Train Compared with Other Configurations
Using Paired Probe Sampling 23
vi
-------
ACKNOWLEDGMENTS
The authors wish to acknowledge the assistance of Mr. William H. Maxwell
of the Midwest Research Institute, Kansas City, Missouri, in conducting the
field study. They also wish to thank Dr. William Mitchell, Environmental
Monitoring Systems Laboratory, and Mr. Roger Shigehara, Emissions Standards
and Engineering Division, for their helpful discussions.
-------
SECTION 1
INTRODUCTION
In its efforts to limit the discharge of harmful sulfur compounds into the
environment, the U.S. Environmental Protection Agency has promulgated standards
of performance for new stationary sources which include sulfuric acid plants
and which limit the discharge of sulfuric acid mist and sulfur dioxide into the
atmosphere (1). An appendix to that regulation (see Appendix A) specifies a
test procedure to determine compliance with the standard. This test, Reference
Method 8, is the subject of the present study. It was conducted in the Quality
Assurance Division of the Environmental Monitoring Systems Laboratory, which is
responsible for evaluating and standardizing EPA test methods. Since promulga-
tion of Reference Method 8, the need for a number of changes has become apparent.
As a result, several changes to further clarify the method and improve its
reliability and accuracy have been incorporated (2).
The test procedure under consideration is designed to separate and measure
sulfuric acid mist (including sulfur trioxide) (S03) and sulfur dioxide (S02).
A measured volume of gas sample is isokinetically extracted from the stack and
passed through an impinger-filter train, where sulfuric acid (FLSO.) mist is
collected in isopropanol (IPA) solution and onto a glass fiber filter, while
sulfur dioxide is oxidized to sulfate in hydrogen peroxide (HpOp) solution.
Both fractions are then analyzed by the barium-thorin method (3). This method
3
was reported to have a detection limit of 0.05 mg/m for sulfur trioxide and
3
1.2 mg/m for sulfur dioxide (4). However, the present study is primarily con-
cerned with the adequacy of the method in the concentration ranges determined
by the performance standard. The latter (1) is stated in terms of allowable
quantities of acid mist and S02 that may be emitted per ton of acid produced
(2 Kg S02/metric ton and 0.075 Kg of acid mist/metric ton). When these stand-
ards are related to the proportion of dilution air commonly employed in the
-------
manufacturing process (5,6), they correspond to gaseous concentrations of acid
3 3
mist and SC^ in the 15 to 30 mg/m and 400 to 800 mg/m ranges, respectively.
(See Appendix B for details.)
The present investigation was carried out to answer certain questions that
arose during a recent collaborative test (4). In that study, there were a
number of observations where high acid mist values occurred together with low
SOp measurements, at a frequency greater than due to chance alone. It was
felt that some conditions might have existed to cause the simultaneous occur-
rence o these values; for example, during S02 oxation to sulfate in the sample
collection apparatus. Some information about the operating characteristics
and collection efficiencies of various train components under simulated field
conditions was also expected to result from this investigation.
-------
SECTION 2
CONCLUSIONS
Evidence was found revealing that the presence of trace amounts of
peroxides in isopropyl alcohol (IPA) results in the conversion of sulfur
dioxide to sulfate. The observed conversion, although too small to effect
sulfur dioxide measurement, produced sufficient H^SO. to cause a significant
positive error in the acid mist values. Such contamination was common in
reagent grade IPA. The presence of peroxides was determined by shaking with
10 percent potassium iodide (KI) solution and reading the absorbance at 352 nm.
Peroxide contamination was removed from IPA by passing it through a column of
activated alumina.
No conversion of S02 to sulfate was observed in peroxide-free IPA, in the
sintered glass filter support disc, or on the glass fiber filter. Investiga-
tion of the latter component included the use of heated gas streams and the
selection of filter paper from several different suppliers. Furthermore, the
presence of acid on the filter paper did not catalyze the conversion of S0? to
sulfate.
Efficient collection of SO- took place in the HgO^-containing impingers
specified in Method 8.
Sulfuric acid mist and SO., were efficiently collected by a single IPA
impinger in combination with a filter. No variation was observed over a range
of flow rates, temperatures, and concentrations or with the various filter
papers employed. However, for efficient S0~ collection, the IPA level must be
maintained above the impinger impaction plate, since SO- must be hydrated to
acid mist to allow its collection on the glass fiber filter.
-------
Two alternate indicators were studied and found unsuitable. These were
sulfanazo III and mixtures of thorin and xylene cyanol FF. Both were strongly
absorbed on the barium sulfate (BaSO.) precipitate, producing color changes
that obscured the end point.
In field testing, paired probe techniques were employed to compare a stand-
ard Method 8 train with several train variations. These studies showed that,
when S03 is absent from the gas stream, acid mist is efficiently collected by
a single filter even when the IPA-containing impinger is eliminated. Ammonia
(NH.j) and dimethyl analine (DMA) vapors collected in the IPA-containing
impinger cause SCL retention and result in large positive interferences in
acid mist values. Both NHL and DMA are used in sulfuric acid plant tail gas
scrubbers. Ferric oxide (Fe^O-), which is present in the effluents of steel
pickling operations, causes a large negative interference in acid mist values.
Solid ammonium sulfate ([NH.LSCL) has proven to be a useful material in
the preparation of audit samples for use in Method 8 analysis.
-------
SECTION 3
EXPERIMENTAL PROCEDURES
LABORATORY PROCEDURES
Gaseous mixtures of S02 in dry air or nitrogen (Np) were prepared using
a model SP-492 Matheson Dyna-blender. The SO,, employed was Linde, anhydrous
grade. Compressed air was purified by passage through a glass wool filter,
a silica gel impinger and a CRS Drierite/Molecular Sieve filter. In some
instances when the generated atmosphere was required to have a moisture con-
tent, the silica gel impinger and CRS filter were replaced with an impinger
containing distilled water. Liquid air nitrogen was used without additional
treatment. The Dyna-blender was capable of preparing three component mixtures
in the 20 to 50,000 ppm range at flow rates of 200 1/min. Its calibration was
checked over a 4 to 200 1/min range using a model P-1900 chain compensated
Collins Spirometer. Deviations of less than 1 percent were encountered.
Atmospheres containing S03 were generated using the schematically shown system
in Figure 1. Dry nitrogen was passed through a Hastings model ALF2KX Mass
Flowmeter and into an impinger containing 30 percent oleum. The resulting SO.,-
containing gas was mixed with the output of the Dyna-blender and then passed
through a series of mixing chambers. A glass fiber filter removed any acid
mist from the gas stream. Acid mist was generated using the same system,
except that moisture was added to hydrate the S03 and the mist-knockout filter
was removed. One series of experiments examined the collection of acid mist
both before and after the addition of SOp. This study utilized the manifold
shown in Figure 2 from which parallel samples were extracted, one upstream
and one downstream, from the point of S02 addition. Appropriate correction
factors were applied to account for the dilution caused by the S02-
-------
EXHAUST
SAMPLING P>—I I—
-------
The analytical and sampling procedures specified in Method 8 (described
in Appendix A) were followed throughout this study. Only those modifications
noted below were employed. The essential features of Method 8 rely on the
following sequence.
• A measured volume of gas is passed through a Greenburg-Smith
impinger containing 80 percent isopropanol-water solution and
then through a glass fiber filter.
• Sulfuric acid mist is collected in these components; any SO.,
present is hydrated to H2SO. and also collected.
• The combined filter extract and impinger solution are analyzed
for sulfate by the barium-thorin method.
• The gas then passes through two additional Greenburg-Smith
impingers containing 3 percent H202 solution. This solution
absorbs S02 gas and oxidizes it to l^SO,.
t These solutions are combined and are also subjected to
barium-thorin analysis.
One variation practiced throughout this study consisted in the separate deter-
mination of the sulfate content of the individual train components. In some
instances, the sulfate collecting impinger-filter combination was supported
by an additional combination in order to determine if SO, had escaped
collection.
J. T. Baker analyzed grade thorin and desicchlora (anhydrous Ba[C104]2)
were employed in the analysis. The titrant was standardized using Fisher
Certified standard HpSO, solution. The isopropyl alcohol employed came from
several sources.
The pH of glass fiber filter paper was determined by placing a 9 sq in.
sample of the paper in a 125 ml glass stoppered Erlenmeyer flask together with
-------
a 1 in by 3/8 in teflon sheathed stirrer bar and 15 ml of 0.05 M potassium
chloride (KC1) solution made from carbon dioxide (CO,,) free, deionized water.
The mixture was held at 22 +_ 1°C and stirred for 10 min at 60 rpm, after
which the pH was measured using a Beckman Expandomatic TM pH meter using a
model 41263 glass electrode.
FIELD TESTING PROCEDURES
Field testing was conducted at a contact sulfuric acid plant. The
facility contained a series of three-stage sulfur burning contact plants of
2000 ton/day capacity. The stacks were 8.5 ft ID by 200 ft high and made of
steel. The sampling site was 160 ft above ground level and 80 ft above the
inlet breeching. Effluent control was achieved by using double pass absorp-
tion at each stage and a Brinks mist eliminator at the outlet.
Samples were isokinetically collected by a team from Midwest Research
Institute using the paired-probe method (7). This consisted of running two
sampling trains simultaneously with their probes in close proximity (30 mm)
so that they essentially sampled the same environment. The following train
variations were studied: inclusion of a second filter between the probe and
the IPA impinger; inclusion of a second IPA impinger-filter combination after
the first one; introduction of 1 ml of concentrated NH, solution into the IPA
impinger before sampling; introduction of 0.2 g of DMA into the IPA impinger;
and introduction of 50 mg of Fe^O, into the IPA impinger. Sulfate analyses,
in this series, were performed at Midwest Research Institute using the methyl
thymol blue procedure (8).
-------
SECTION 4
RESULTS AND DISCUSSION
ACID MIST AND S03 COLLECTION EFFICIENCY
A study of the Method 8 train efficiency for acid mist and S03 collection
was conducted in a system consisting of two impingerfilter combinations con-
nected in series, followed by a third impinger (Figure 3). This train sampled
from a stream containing acid mist or SO., and the sulfate collected in each of
these five components was measured. A collection efficiency was calculated
representing the percentage of the total sulfate found in the first impinger
and filter, since these are the components of the standard Method 8 train
designed to collect the acid mist and SO.,. Several of the parameters varied
were: the sulfate concentration and temperature of the gas streams under
analysis, the type of glass fiber filter used, and the flow rate through the
sampling train. These results are tabulated in Tables 1 and 2. The greater
part of this study was done using SO., rather than acid mist. It was felt that
SO., presented a greater challenge to the method, since any traces remaining in
the gas phase or forming very fine particulate would pass through the filter.
(Significant concentrations of SO., may be present in HUSO, plant stack gases
as a result of poor absorber operation [9].)
The data in Tables 1 and 2 shows that the first impinger-filter combina-
tion was highly efficient in collecting both acid mist and SO^. Furthermore,
the bulk of the sulfate was collected on the filter; less than 25 percent was
usually detected in the first impinger. Table 1 lists the type of glass fiber
filter paper employed in each run. The pH values were measured; MSA 1106BH
and Gelman AE had values in the 6.2 to 6.7 range, while Reeve-Angel 934AH had
a pH of 9.1. Except for one incidence yielding a low collection efficiency
value when a Reeve-Angel filter was used (run 9) effects resulting from the
-------
SAMPLING
PROBE
, TO DRY
—-T\ GAS METER
o
•100ml, 80 ptrwnt
IPA SOLUTION
Figure 3. Train used in acid mist collection efficiency study.
10
-------
TABLE 1. S03 SAMPLING DATA AND COLLECTION EFFICIENCY
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
Sampling
rate
1/min
21
21
16
31
21
21
21
30
35
16
20
20
20
Gas
temp.
20
20
20
20
75
75
75
75
75
75
75
95
77
Filter
paper
MSAa
MSA
MSA
MSA
MSA
Gelb
Gel
Gel
R-AC
R-A
R-A
MSA
MSA
1st IPA
impinger
39.8
36.6
3.3
0.7
1.2
46.8
3.9
38.7
0.6
30.5
1.3
11.0
10.2d
so3
Collected, H2S04 mg/m
1st 2nd IPA
filter impinger
143.6
59.4
19.1
11.1
3.6
213.2
14.5
248.9
2.5
385.9
17.2
21.8
16.7
0.7
0.1
0.1
0.5
0.1
0.7
0.4
0.2
0.2
0.1
0.1
0.4
3.1
2nd
filter
0.3
0.4
0.2
0.2
0.1
0.2
0.0
0.2
0.7
0.2
0.1
0.3
10.9
3rd IPA
impinger
0.0
0.0
0.0
0.0
0.0
9.9
0.0
0.0
0.1
0.0
0.0
0.1
0.0
Total % Recovery in 1st
collected impinger & filter
184.
96.
22.
12.
4.
270.
18.
288.
4.
416.
18.
33.
40.
/i.
5
/
5
9
8
8
0
0
7
7
6
9
99
100
99
95
98
96
98
100
76
100
99
98
66
aMine Safety Appliance 1106BH
Type Gelman AE
cReeve Angel 934 AH
Impinger filled with IPA solution to a level below the impaction plate
-------
TABLE 2. ACID MIST SAMPLING DATA AND COLLECTION EFFICIENCY
Acid mist collected, H2SO
Run3
1
2
3
4
5
Gas .
temp.
20
20
70
91
70
1st IPA
impinger
12.5
6.2
15.1
11.8
13. 8C
1st
filter
3.9
115.9
114.9
53.3
43.6
2nd IPA
impinger
1.0
0.2
0.5
0.2
0.0
2nd
filter
0.3
0.1
0.0
0.3
0.1
4 mg/m
3rd IPA
impinger
0.2
0.0
O.U
0.0
0.0
Total
collected
17.8
122.4
130.5
65.6
57.5
% Recovery
impinger &
92
100
100
99
100
in 1st
filter
3MSA 1106 BH filters and sampling rates of 21 1pm were used throughout this series.
clmpinger filled with IPA solution to a level below the impaction plate
-------
variation of filter type were undetected. The effect of flow rate through the
sampling train was also considered. Flow rate was varied from 16 to 34 1/min,
the range over which the Greenburg-Smith impinger has been shown to be effi-
cient (10). Low flow rates (runs 3 and 10) did not produce poor collection.
In one instance (run 9), poor collection resulted at a higher rate, but this
was unconfirmed; thus, the collection efficiency remained relatively constant
over the range of sampling flow rates studied. Sulfate concentrations
3
generated varied from 4 to 400 HpSO. mg/m without any apparent effect on the
collection efficiency.
When sampling was carried out from heated gas streams, it was noted that
the gas rapidly cooled upon entering the Greenburg-Smith impinger resulting in
the evaporation of the IPA solution. After two hours of sampling, the alcohol
was often depleted to a level below that of the impinger impaction plate.
Nonetheless, excellent collection of both acid mist and SO., resulted. During
a further evaluation, sampling was carried out with the IPA level below the
impaction plate throughout the entire run. Acid mist was still collected
efficiently (run 5 Table 2) but very poor collection of S03 resulted (run 13,
Table 1). The quantity of moisture in the first impinger was apparently in-
sufficient to cause the complete hydration of S03, since acid mist formation
was observed in the second impinger. The results indicated the necessity of
maintaining the IPA solution above the level of the impinger impaction plate
for efficient S03 collection.
EFFICIENCY OF S02 COLLECTION
The first part of the Method 8 collecting train, consisting of a Greenburg-
Smith impinger containing isopropanol-water solution plus a glass fiber filter,
is designed to collect acid mist and to separate it from SO^. For the most
part, S02 remains in the gas phase and is subsequently collected in 3 percent
H202 solution in a latter portion of the train. A small amount of S02 dis-
solves in the IPA solution, but is removed by purging. However, since the
quantity of S02 under analysis is large when compared with acid mist, measure-
ments were conducted to determine if S02 was retained in the train components
used for acid mist collection. If such were the case, spurious acid mist
13
-------
values would result. These studies were done using gas streams containing S02
and where acid mist was absent.
The results of these measurements are summarized in Table 3, which yielded
the following conclusions. Collection of S02 in the second and third impingers
was carried out with recoveries that varied between 96 and 102 percent. Glass
fiber filter extracts yielded measurable quantities of sulfate, but these
values were unrelated to the concentration of SO,, in the gas stream and were
negligible when compared with typical acid mist source values. Similar results
were obtained when the filter support disc was also extracted and measured for
sulfate (runs 27 and 28). However, sulfate was found in the IPA solution of
the first impinger in substantial amounts when compared with the quantity of
acid mist encountered in a typical compliance analysis. Acid mist compliance
values between 14 and 32 mg/m3 (see Appendix B) and values of approximately 4
mg/m3 were found in runs 26 through 32. (Roughly the same volume of gas, 0.57
m3 was collected in each of these runs.) Thus, the passage of S02 through the
sampling train resulted in spurious acid mist values ranging from 12 to 28 per-
cent of the compliance value.
The amounts of H2S04 collected remained relatively constant even though
the concentration of S02 in the gas stream under analysis was varied by a
factor of four. Increasing the flush time from 15 min (runs 26 through 33) to
30 min (runs 34 through 37) did not decrease the quantity of H2S04 collected.
Furthermore, no change was observed when additional H2S04 was added to the IPA
before sampling (runs 29 and 33) or when acetone was added (runs 36 A and B).
However, when a second IPA impinger was placed in series with the first (run
35B), the same amount of H2S04 was collected as in the first impinger. Also,
an impinger containing twice as much IPA solution (run 37B) collected twice as
much H2S04. These results indicated that the collection of S02 and its conver-
sion to H2S04 resulted from the oxidation of the S02 by a substance contained
in the IPA solution. This was confirmed when IPA which had been redistilled
was employed over SnCl2 (run 37A) and the amount of H2S04 collected was greatly
reduced.
14
-------
TABLE 3. S02 COLLECTION IN METHOD 8 TRAIN
S02 Concentration, mg/m
Run
26
27
28
29
30
31
32A
32B
33
34
35A
35B
36A
36B
37A
37B
Prepared
531
531
527
1030
983
1047
1084
0
813
2013 .
1030
1030
1030
1030
1030
1030
Measured3
512
522
526
1022
997
1028
1038
0
826
2057
-
-
-
-
-
-
H^SO. recovered, mg
I PA Solution
2.54
3.11
2.25
2.ZO
2.61
2.19
2.20
0.00
2.16
2.18
8.93d
2.37
2.20
2.38
0.04e
4.30
Filter
0.30
0.02b
0.16C
0.18
0.01
0.06
0.02
0.00
0.02
0.04
-
-
O.M2
0.02
0.04
0.04
aQuantity collected in two impingers containing 3 percent
0.15 mg also detected on filter disc.
C0.04 mg also detected on filter disc.
IPA solution not flushed.
eIPA solution distilled over SnCl,,.
15
-------
The occurrence of oxidants in the IPA observed here should not be regarded
as an isolated example. Diisopropyl ether, a substance which readily forms
peroxides (11) is employed in the manufacture of anhydrous IPA. When samples
of IPA from a number of commercial sources were tested for peroxides, a serious
contamination problem was often encountered. The test employed was the KI -
peroxide test (see Table 4).
The IPA solution producing the spurious acid mist values listed in
Table 4 gave an absorbance of 0.660 when tested. However, Table 4 also con-
tains another entry at least three times that value. It is therefore highly
probable that seriously contaminated IPA has been employed in the past in
Method 8 source testing and has resulted in spurious acid mist values. Thus,
is recommended that all samples of IPA be tested for peroxides prior to being
employed in Method 8 measurements, and that the method be modified to include
the KI test, herein described. Peroxides can be removed from IPA by distilla-
tion over SnC^ or by passage through an activated alumina column. The latter
procedure is more convenient and was employed for all subsequent measurements
made in this study.
TABLE 4. RELATIVE H LEVELS IN IPA FROM VARIOUS SOURCES
Commercial source
Fisher Spectranalyzed
Fisher Spectranalyzed
Arthur H. Thomas
Baker Analyzed
Mallinckrodt AR
Occidental Chemical Co.
.
Lot
755378
753162
?
607031
CSV
?
Absorbance3
0.660
O.D80
0.050
0.517
1.978
0.025
aAbsorbance resulting at end of 1.0 min in a 1 cm path cell containing a 1:1
mixture of IPA and 10 percent KI.
16
-------
INTERACTION OF ACID MIST AND S02
To further investigate the mechanisms that might produce spurious acid
mist production, the possibility was investigated that S02 was catalytically
oxidized by the H2$0, collected on the glass fiber filter. This study employed
the train shown in Figure 3; the sole modificatio involved the last impinger
which contained 3 percent hLOp solution to collect the S02> The gas stream
analyzed consisted of 5.5 percent 02 in N2, and had S02 concentrations as
listed in Table 5 and a temperature of 75°C. Solution containing H2$04 was
measured onto glass fiber filters and air dried. The quantity added corre-
sponded to the amount of H2S04 that would be collected from the gas stream at
the end of 2 hrs, if the acid mist concentration were in the 11 to 24 mg/m
range. The molar quantity of S02 that passed through the sampling train was
from 100 to 250 times greater than that of the H2SO. placed on the filter
paper.
The following results are evident from Table 5. The single H202 con-
taining Greenburg-Smith impinger collected S02 in good yield. Only traces
of H2SO, were found in the IPA containing impingers. Sulfuric acid recovered
from the glass fiber filters equalled the amounts added within experimental
error in three out of four instances. On the first filter of run 2, an excess
amount was detected corresponing d to an acid mist value of 0.4 mg/m . Thus,
if oxidation of S02 took place, it did so at a level at or below the detection
limits of these measurements. Furthermore, the S02 concentrations employed
were approximately four times greater than those encountered at a typical
source. Therefore, we must conclude that H2SO, does not promote the oxidation
of S02 to the extent of causing a significant acid mist interference.
A further attempt to find evidence for the interference of SO/, in acid
mist measurements was carried out as follows.
Suspended acid mist particulate was produced in heated gas streams
consisting of 5.5 percent 0? in N?. A gas manifold was utilized allowing S02
the addition of and enabling samples to be withdrawn both before and after the
point of S02 introduction. The gas manifold was described earlier. Parallel
17
-------
TABLE 5.
TEST FOR S02 OXIDATION ON GLASS FIBER FILTERS CONTAINING ADDED
Run
1
2
1st IPA
impinger
0.008
0.001
1st Fi-
Added
0.300
0.248
H2S04 mmoles
Iter
Found
0.297
0.261
2nd IPA
impinger
0.
0.
010
001
2nd Fi
Added
0.600
0.248
Iter
Found
O.b93
0.244
Vol ume
m
2.456
2.366
S02 mg/m3
Added Found
1801
1800
1811
1749
%
Recovery
100.6
97.2
00
-------
samples were withdrawn by two Method 8 trains operating simultaneously. Acid
mist concentrations were measured and the results are presented in Table 6.
TABLE 6. EFFECTS OF S02 ON ACID MIST COLLECTION
Run
1
2
3
4
5
6
7
8
9
10
H2S04
S02 Absent
4.12
10.95
30.8
72.1
98.9
39.6
18.0
33.1
18.8
17.8
mg/m
S02 Present
4.43
13.15
38.2
49.6
120.1
43.0
18.3
32.9
18.6
19.9
%
Deviation
8
20
24
-31
21
9
2
-1
-1
12
S02 Added
mg/m
1030
840
1230
1190
0
0
0
0
3200
1400
It seems impossible to draw definite conclusions from this data. Acid mist
values measured before and after the S02 addition neither agreed nor showed a
consistent trend. The observed deviations, ranging from -31 to +24 percent,
took place when S02 was absent and did not correlate with S02 concentrations
when the latter substance was added. Clearly, these differences represent
fluctuations introduced by experimental factors, rather than chemical inter-
actions between acid mist and S02- Thus, the data only allows us to place an
upper limit on such interactions. Consequently, when it was noted that these
measurements employed S02 concentrations as much as eight times greater than
typical source values, this series of experiments added weight to the other
results reported in this study. Based on this data and on our failure to find
evidence for the oxidation of S02 to H2$04 in IPA impingers or on glass fiber
filters, it must be concluded that S02 does not interfere in acid mist
measurements.
19
-------
DISCUSSION OF METHODOLOGY
During the course of the preceding study, it became necessary to prepare
filters containing a known amount of HpSO.. The initial filters made were
dried in an oven. Discoloration resulted and the added H,,S04 could only be
partially recovered. The variety of filter paper employed (MSA 1106BH) con-
tains approximately 0.5 percent organic binder. It became evident that oxi-
dation of the filter paper binder had taken place with a resultant loss of
H2S04. When H2S04 was added to filters that were subsequently dried at ambient
temperatures, no discoloration resulted and good recoveries of H2S04 were made.
These measurements were repeated several times and the results are listed in
Table 7. The filter employed in EPA Method 8 remains at or below ambient tem-
peratures where loss of H2S04 is absent. However, we have received suggestions
for the design of other techniques which the heating of filters for use with
sources possibly containing H,,S04. The results in Table 7 indicate that such
alternative methods would have to employ glass fiber filters practically free
of organic binder in order to avoid collected acid mist losses or a reduction
in the filter "tare."
TABLE 7. H2S04 INTERACTION WITH GLASS FIBER FILTERS
Sample
Pipette calibration
Pipette calibration
Oven dried, temp. & time
not recorded. Charring.
Pipette calibration
Pipette calibration
Dried at ambient temperature
no darkening of filter
Filter dried at 160°C, 2 h
Charring of paper observed
H2S04,
Added
0.300
0.300
0.300
0.300
0.250
0.250
0.250
0.250
0.250
0.250
0.250
0.250
mmoles
Found
0.301
0.299
0.189
0.208
0.248
0.251
0.245
0.248
0.240
0.136
0.133
0.154
%
Recovery
100.3
99.6
63.1
69.4
99.2
100.2
98.1
99.1
96.2
54.5
53.1
61.7
20
-------
The barium-thorin analytical method is based on the insolubility of barium
sulfate (BaSO^) and the formation of a colored complex between barium ions and
thorin indicator. During the course of this and related studies, both barium
chloride (BaCl2) and barium perchlorate (Ba[C104]2) have been employed as
titrants. Experience has shown that the use of BaCl2 results in end points
that are not as sharply lineated. This may be caused by the formation of a
complex between barium and chloride ions which competes with the barium-thorin
complex. The use of a Ba(C104)2 as the titrant is therefore preferable.
Difficulty in detecting the yellow-pink color change which occurs at the
barium-thorin end point prompted the investigation of other indicators.
Sulfanazo III and thorin-xylene cyanol FF mixtures were studied. Both of
these indicators give sharp, easily observable end points, when low sulfate
concentrations are measured. However, sulfanazo III and xylene cyanol FF dyes
are strongly absorbed on BaSCL precipitate, so that continuous color changes
take place during titration. This effect is especially pronounced when high
sulfate concentrations are being determined. Since high sulfate concentrations
often occur during source sampling, both of these indicators are unsuitable foi
such measurements.
FIELD TESTING
Field testing was carried out to study several variations of the Method 8
sampling procedure under the rigorous conditions of actual use. Each variation
was tested simultaneously with a standard Method 8 train, using the paired-
probe method. Sulfate analysis was carried out by the methyl-thymol blue
method previously mentioned (8). This procedure was employed because only the
sample collection part of Method 8 was under consideration in the field tests.
In addition, NH~, DMA, and Fe^O.,, which were employed in this study, do not
interfere in the methyl thymol blue procedure (12,13), but do interfere in
barium-thorin analyses under certain conditions (3). The results are listed
in Table 8; values obtained using the Method 8 train are designated as A and
train variations are designated as B.
21
-------
Effect of Purge Time Variation
In the first test, Train B was purged for only 4 min at the end of the
sampling run, while Train A was purged for the full 15 min specified in the
Method 8 procedure (sample 1 and 2). There is no evidence of S02 retention
in the IPA impingers flushed for only 4 min, nor do the S0? values from
Train A appear to be increased as a result of ambient SCL collection.
Acid Mist Collection Efficiency
To test the efficiency of acid mist collection, a second IPA impinger and
a filter were placed immediately after the first combination. Moderately high
vacuums were encountered on Train B due to the extra filter. Results of the
sulfate analysis of the probe, the first IPA impinger, and the first filter,
are listed under the acid mist column of Table 8 (3B and 4B). Results of the
analysis of the second IPA impinger and the second filter are designated 3B'
and 4B1. There is some indication that sulfate was detected in the second IPA
impinger-filter combination in sampling run no. 3. However, very low acid mist
values were encountered in this study. The analytical blank corresponded to
3.6 mg of h^SO^/m so that the value listed as 3B1 represented barely more than
a trace amount. The quantity of sulfate detected in Sample 4B' was too small
to measure. Thus, samples 3 and 4 do not provide conclusive evidence; however,
they do indicate that at least the bulk of the acid mist is collected in the
first IPA impinger-filter combination.
The configuration adopted in sampling runs no. 5 and 6 consisted of a
filter, an IPA impinger, and another filter, followed by the peroxide impingers.
The first filter and the probe were analyzed separately for sulfate and results
are reported in the acid mist column of Table 8 for samples no. 5B and 6B. The
IPA impinger and the second filter were analyzed together and the results are
reported as samples 5B' and 6B1. This data shows that all of the acid mist was
collected in the probe and on the first filter. Sample no. 6 is particularly
significant because sulfate values were encountered which were substantially
above trace levels.
22
-------
TABLE 8. METHOD 8 TRAIN COMPARED WITH OTHER CONFIGURATIONS
USING PAIRED PROBE SAMPLING
Sample
No.
l/£
IB
2A
2B
3A
3B
3B1
4A
4B
4B1
5A
5B
5B1
6A
6B
6B1
7A
7B
8A
8B
9A
9B
Special conditions
employed in Train B
4 min purge
4 min purge
Second IPA impinger
and filter
Second IPA impinger
and filter
Second filter
before IPA impinger
Second filter
before IPA impinger
NH., in IPA impinger
DMA in IPA impinger
Fe203 in IPA impinger
Gas vol . sampled,
standard m
1.087
1.073
1.156
1.172
1.130
1.134
-
1.163
1.146
-
1.122
1.110
1.255
1.251
1.276
1.235
1.323
1.304
1.275
1.266
Acid mist.,
H2S04 mg/mj
6.0
5.0
trace
trace
2-lh
!.?£
0.8-
0.7,
1.1\
trace-
1-"L
n od
1 . o—
trace^-
6.5,
4.4_
trace^
5.1
57.6
5.0
98.0
6.1
0.5
S02'3
mg/m
381
372
372
380
742
685
-
356
335
-
355
359
440
433
-
393
273
386
200
373
359
—Sample designation A denotes use of standard Method 8 train.
^Value obtained from first IPA impinger-filter combination.
%alue obtained from second IPA impinger-filter combination.
^Value obtained from first filter.
—Value obtained from second filter and IPA impinger.
23
-------
A method that has been employed for sulfuric acid plant analysis - the
Monsanto Method (12) - determines sulfate from the extract of the sampling
probe and a glass fiber filter. It is similar to EPA Method 8, except that the
latter method also employs an IPA impinger before the filter. From these re-
sults, combined with the laboratory evaluations we can conclude that a filter,
not preceded by an IPA impinger, will collect sulfuric acid mist efficiently;
however, accurate results can only be expected when SOg is absent from the gas
stream. For the latter compound to be collected on a glass fiber filter, a
hydration mechanism is required, such as the one provided in the IPA solution
specified in Method 8.
Efficiency of SOp Collection
The results in Table 8 show that the paired-probe trains agreed within +_ 1
percent in samples 1, 2, 5, and 6. Agreement in samples 3 and 4 was +_ 4 and 3
percent, respectively. These sampling runs employed a second impinger-filter
combination and a higher train vacuum. Consequently, this may have caused a
slight, intermittant leakage diluting the sample and resulting in lower values.
On the other hand, +4 percent is not considered large in source sampling; thus,
it is a fair judgement that good agreement was obtained between paired-probe
train S02 values in the entire series of samples 1 through 6. This is evidence
that the experimental train variations employed in this series did not disrupt
train functioning.
Effect of Interferents
Ammonia (NHj) has been employed in S02 scrubbers and their tail gas
streams may contain as much as 0.03 percent NH~ vapor (15). At that concen-
tration, gas collected at a 28 1/min sampling rate for 2 h would contain a
total of 40 mmoles of NHg. To approximate these conditions, 15 mmoles of NH~
were added to the IPA solution in the first impinger, prior to sampling. When
sampling was conducted the formation of a white solid was observed [NH.HSO^]
which eventually plugged the filter, requiring an interruption of the run plus
a filter change. At the conclusion of sampling, the odor of NH3 was no longer
24
-------
detected in the IPA impinger. Methylthymol blue analysis of the combined
extract of the IPA impinger and the filter from the train with added NH
O
yielded a sulfate value more than 10 times that of the control train (see
Table 8). The methylthymol blue method determines sulfite as well as
sulfate (8). The measurement value obtained from the NH- containing train
corresponded to 1.6 milliequivalents of S02- Since 15 milliequivalents of NH3
was added, this would indicate that much of the NH3 had been purged out of the
IPA impinger. However, the degree of S02 retention was sufficiently large
that serious interferences may be expected if Method 8 is used to measure
sulfate emissions in effluent streams containing NhL vapor.
Dimethlanaline is another material used in scrubbers to control SCL
emissions. In order to simulate conditions (16) that may occur after sampling
effluents from DMA scrubbers, 3.3 milliequivalents of DMA were added to the
IPA impinger and the results were compared with a standard Method 8 train (see
Table 8). Analysis indicated that 3.8 milliequivalents of excess S02 were
detected in the IPA impinger extract and the filter of the train with added
DMA. However, the total amount of S02 measured in the latter train was con-
siderably less than that measured in the standard Method 8 train; therefore,
the presence of DMA may have caused some analytical difficulty, in addition to
producing a large positive interference in the sulfate value. Sulfuric dioxide
values obtained when NH^ and DMA were present showed poor agreement with the
corresponding paired-probe train values. This remained true in samples 7 and
8 even when the "acid mist" values of Train B were added to the S02 values and
compared with the quantity of SCL determined in Train A. It can only be con-
cluded that the presence of NH3 and DMA caused a general disruption of both
the sampling and the analytical procedures and produced meaningless results.
The use of Method 8 at steel pickling plants prompted us to study the
effects of ferric oxide (Fe203) on Method 8 train operations. Approximately
1.2 milliequivalents of iron were added to the IPA impinger; Table 8 shows that
a negative interference in the sulfate measurement resulted. Agreement between
the S02 values of Trains A and B was fair, so it appears that the presence of
Fe0 affected only the acid mist determination.
25
-------
EVALUATION OF SOLID AMMONIUM SULFATE AS AN AUDIT MATERIAL
It would prove invaluaMe to have a solid substance to aid in the prepara-
tion of audit samples for barium-thorin analysis. Audit samples prepared from
a solid sulfate would be independent of the H2S04 solution used for calibration
and would be useful, not only for evaluating the operator, but also for
detecting calibration errors. For these reasons, we carried out the following
comparison study and selected solid ammonium sulfate (NH4)2S04 as the candidate
material.
Barium perchlorate (Ba[C104]2) solution was prepared as described in
Method 8 and standardized against standard acid. The results are listed in
the first column of Table 9. A stock solution of (NH4)2S04 was prepared from
reagent-grade salt that had been dried for 2 h at 110°C. The (NH4)2$04 solu-
tion had the same normality as the standard acid, 0.01013N. The results of its
barium-thorin analysis appear in the second column of Table 9. An additional
test was conducted employing (NH4)2S04 solutions produced by adding measured
amounts of NH3 to aliquots of standard H2S04 solution. These were analyzed by
the barium-thorin method (see Table 9).
TABLE 9. COMPARISON OF BARIUM THORIN ANALYSIS OF H2S04 AND (NH4)2S04 SOLUTIONS
Ba(C104)2 Solution, ml
Standard
H2S04
26.35
26.79
26.82
26.42
MEAN 26.613
(NH4)2S04a
Solution
25.60
25.66
25.62
-
25.627
NH3 Standard
H2S04
26.41
27.43
26.90
26.82
26.890
Prepared from solid (NH4)2S04-
26
-------
The following conclusions may be drawn from the data in Table 9. Measure-
ment of the solution prepared from solid (NHJpbO* yielded values that were
slightly low - 96.3 percent of the gravimetric value. The discrepancy is real,
because a statistical treatment of the pooled data in the first and second
columns shows the difference between the two means to be significant at the
95 percent confidence level. A similar treatment of the data in the first and
third columns shows that those differences are insignificant.
Thus, barium-thorin analysis of a solution prepared from solid
yielded results approximately (96.3%) equal to the prepared value. The discrep-
ancy from 100 percent was apparently real and may have been caused by impurities
in the salt or by a calibration bias. The presence of a stochiometric amount of
NH3 did not produce an interference. Therefore, solid (NH4)2S04 shows promise
for future use as an audit material in barium-thorin analysis, although
additional work will be necessary before this material can be established as a
calibration standard.
27
-------
REFERENCES
1. U.S. Environmental Protection Agency. Standards of Performance of New
Stationary Sources. Federal Register 36(247):24876-24893, December 23,
1971.
2. U.S. Environmental Protection Agency. Standards of Performance of New
Stationary Sources. Federal Register 41(111):23060-23090, June 8, 1976.
3. Fritz, J. S., and S. S. Yamamura. Rapid Microtitration of Sulfate.
Analytical Chemistry, 27:1461-1469, 1955.
4. Hamil, H. F. and D. E. Camann. Collaborative Study of Method for the
Determination of Sulfuric Acid Mist and Sulfur Dioxide Emissions from
Stationary Sources. EPA-650/4-75-003, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1975.
5. U.S. Public Health Service. Atmospheric Emissions from Sulfuric Acid Manu-
facturing Processes. Public Health Service Publication No. 999-AP-13,
Cincinnati, Ohio, 1965.
6. Office of Air Programs. Background Information for Proposed New Source
Performance Standards. Report No. APTD-0711, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1971.
7. Mitchell, W.J., and M. R. Midgett. Means to Evaluate Performance of
Stationary Source Test Methods. Environmental Science and Technology,
10:85-88, 1976.
8. Bergman, F. 0., and M. C. Sharp, Measurement of Atmospheric Sulfates:
Evaluation of the Methylthymol Blue Method. Appendix. Environmental
Monitoring Series, EPA-600/4-76-004, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1976.
9. Sittig, M. Sulfuric Acid Manufacture and Effluent Control. Chemical
Process Review No. 55. pp. 239-245. Noyes Data Corp., Park Ridge,
New Jersey, 1971.
10. Cooper, H. B. H., and A. T. Rossano. Source Testing for Air Pollution
Control, pp. 141-142. Environmental Research and Applications, Inc.,
Wilton, Connecticut, 1971.
11. Reddick, J. A., and W. B. Bunger. Organic Solvents, p. 651. Wiley-
Interscience, New York, 1970.
28
-------
12. Meites, L. Handbook of Analytical Chemistry. McGraw-Hill, New York,
New York, 1963. pp. 10-149
13. Appel, B. R., E. L. Kothny, E. M. Hoffer, and J. J. Wesolowski.
Comparison of Wet Chemical and Instrumental Methods for Measuring Air-
borne Sulfate. Environmental Monitoring Series. EPA-600/7/77-128.
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina 27711, 1977. p. 76
14. Patton, W. F., and J. A. Brink. New Equipment and Techniques for
Sampling Chemical Process Gases. Journal of the Air Pollution Control
Association. 13:162-167, 1963.
15. Kohl, A. L., and F. C. Riesenfield. Gas Purification. McGraw-Hill,
New York, New York, 1970. p. 255
16. Ibid. p. 207.
29
-------
APPENDIX A
MITKOD I— DlTIKMWAtlOM Of IClrUUC AOD HIM
AMD gvmt DlOIIOl IMUNOM VtOU ITiTlOMjktY
I. Prtnttpti t*t AppUetMlUt
l.l Principle A |U strnplt Is attracted laoklnttloally
from the ii*ci. The lulhirlc icld mill (Including niUur
trloilde) and ih* sulfur dloilde in separated, tnd both
tracilont tn measured separately by ib» be/lum-thorln
ttlntloa method.
1.1 ApplKiblllty. TU* method U applicable tor th«
extermination ot sulrurio «cld mlit (including wlfUr
Monde, tnd In (hi tbirnce of olbv paniculate matter)
tad tuUur dloildt tmlHloiu (rare stationary MurMi.
Collebofatl?s'iesu have ihovn that the minimum
skueltble llmlu of the method are O.OT mllUmmi/eublt
•Better (0.03 > 10-* noundi/cubic fool) for sulfur Irtotld*
tnd 1.1 mi/mi (0.74 10-' Ib/ll') for sulfur dioilda. No
•pper limits hiT* boon t«aUl»hcd. Utasd on thoontletl
Mkutelloni lor 200 rulUUUers of I percent hydrogen
ptvoilde tolullon, tbi upper concentration limit tor
Mlrur dloilde In i 1.0 m' (M.S ft>) iu iunple U kboat
UJOO mi/mi (7.7X10-< Ib/lt'). Tbe upper limit ran bo
esttndtd by Inertulnf the qiunllty of pcroilde tolulioo
I* the Impinf en.
f omlblt Inttrferirif vents of tbu method ere fluoridee,
trae tmmonle, knd dimethyl enlllne. If tny ol these
toUrterini iienli ire prtsrnt (ihli cen be determined by
knowledfe ol Ih* proeeie), »HemillTe method*. wb)eel
U th* Mmottl ol the AdmliilMimtor.U.S EPAire
Filterable partlcultte matter may be de-
termined along with SO. and SO, (*ub)ect to
the approval of the Admlnlatrator) by In-
•ertlnt a heated glaw fiber filter between
the probe and Uoproptnol Imptnger (see
Section 2.1 of Method 6), If thli option la
chosen, paniculate analysis is gravimetric
only: H.SO, acid mill la not determined aep-
•rately.t7
S. Apptrttut
2.1 *«mpUnf. A echemillo ot the luapUni train
uard la Ihli method U ihown In Flpire t-1; ft U Mmllir
to the Method 5 train circpt thai the fllter position la
dUfwtnt tnd th< filler holder don not hive to w hated.
Commercial oiodrli of thli train art available. For thoee
who desire to build their own, however, complete oon-
alrucllon detalll ar< dctcrlbrd Iu Al'TD-fx>l Chanfee
from the At'TU-U'41 doi'unieiit and »Uow»blo modi-
fication* to Figure t-1 are dlxuurd In the following
aubtcclloni.
Tho opmilnf and maintenance procedures for the
lampllng train are dotcilUedtn AfTu-OiTe. Since correct
uu|e Is lm|Kirtant In obiulium valid resulti, all uteri
ctiuuld rc»d Iho Al'TH-OJ?!! rtofurcciit and adopt the
operating and malntriuncc prixiJuri* outlined In it.
unleu otherwise spn'incd hcrrln. Further JttallJ and
guidelines on oitcruilon and mulnientnce arc pftn la
Method 5 and should \» read and tallowed whenever
they are applicable.
1.1.1 ProlHi Noiile. Same u Method i, Section 2.11.
J.1.2 Prolxi Uner. Uoroolllcate or i|Uaru (last, with a
heailrm system to prevent visible condriiMilou during
aunpUni Do not use metal proU linen.
1.1.1 fltol Tube, game at Method 4, Section 1.1.1.
11.4 Differential Pntrar* Otnje. Bam, H Method i,
gecilon 114.
IU Filter Bolder. BoroalUeaU ilaes, with a glua
frit niter support tnd t alllcone rubber gasket. Other
•tket mturltli, e.g., Teflon or VIton, may be wed sub-
Ret to the approval of the Administrator. The holder
MtD ibtll provide a poalUve teal agtlnit leakage from
tbe onuldt or around tbe fllUr. The Blur bolder iball
be placed between the flrtt and Meond Implngen. Not*:
Do not heat the flltn holder.
11.6 Imptngen-Tour. at abown In Tlgwe a-1. Tht
•nt tnd third shall be of tbe Orttnbun-amlth talin
with standard Upe. The second tnd fourth shall be o«
tbe Oraenburf-tmUh design, modified by replacing the
Insert with tn approilmately 11 millimeter (0.6 In.) ID
flas* tube, bavins an uneonilrtettd tip located 11 mm
(0.1 tn.) from the bottom of the flask, glmllar collection
trtlama, which have bean approved by tbe Adminis-
trator, may be used.
11.7 MeUrlng gyitem. lame at Method (, gectloo
ll'j Barometer. Bame at Method (.Section ll.t.
II .• Oas Density Determination Equipment, lame
at Method 5, Section l.UO. ¥ ^^
11.10 Temperature Oauie. Thermometer, or equlv*.
BevmpU Recovery-
TEMPERATURE SENSOR
PROIE
THERMOMETER
mOTTUIE
TEMrERATURESENSOR
FILTER HOLDER
.CHECK
/VALVE
REVERSE TYPE
UTOT TUBE
VACUUM
LINE
VACUUM
GAUGE
MAIN VALVE
DRY TEST MITER
•
Pifurt 1-1. Sulfurlc Kid mlit umpllng train.
30
-------
Wwb Bottle*, PeiyotbytaB* or ilaai, 100 mt a.1.4 In
SJLI ^Graduated .Cylinder*. MO ml.
UUr. **
Ibe tollowinf Ult for d« tec tint peroxldee la **cb lot •!
kntwvpaool: Sb*k* 10 ml of the teopropanol with 10 ml
fftMDjypn
U.4 OndMtod Cylinder. 100 mL
Mi Trip BitMos. 100 | ctpadty, to BMMur* to
IJ.I* Drapptaf BotU*. T» add indtatar MbjUta.
Bf-mlMa.
OnlM a«e«ye«Ud tofceta-aeaTil,
r prepared 10 percent poUnlum Iodide wlatloo.
• blank by aimllarly treat I n« 10 ml of dlMOIod
Aftor I minute, read the tbaorbtoe* on a naetre-
pbotometer *t 1&2 naaomeUn. II the abm baaea ueeede
1.1. the leopropanol ahaU not bo tued. Peroxide* may bo
renoted from leopropanol by radl*UIUn(. or by peeaage
Ibtoufh a column of actlreled alumina. However, ra-
•CanHtrade laopropanol with aultobly tow peroxide level*
k rawlly aTallable from oommuxlal eoorcee; tberelora,
Mlectlon of oaotomuatod lot* mar b* more amnam<
tbaa tollowlni the peroxide removal prooedurr.
(.14 Hydrogen Peroxide. I Parent. Dilute 100 ml
•I30 percent hydroten paroxldo to 1 Utor wttb ililnnlnl.
•bUDixS wfttcr Frepctfv fl*Mb OAlly.
•.I.* Crushed Ice.
U •ample Becorery.
•M.I Water. BUM u 3.1.3.
*JJ I*oprop»nol,30Percent»»mei«3.1.4.
U AnalyiU.
U.I Water. Same a« 3.14.
M.2 toi«op*iiol,100Peiwnt
Mldteabl* «r(kBtc
1.14 Bwtnia Poreblorttt (0.0100 Nonctl
I.Mtofbor1uBp*rehlor»U lrihjrdr»u (B»(CIO,)riniO)
ta TO ml delonlud diftllM »»ur. uid dllui* U) 1 Uui
with Inprapwtol. 1.22 | of bo/lnm chloride dlhydrMi
(BtClhlHiO) miy b* ooed IntttKl of th« bknum p*r-
ohl«rat« BuniUrdlM wHh •olfurlc ocld u In SwUon 8 >
TtLl« aohiUoa matt b* protaotod ifklim mponUon M
1.1.6 Sulftirte Acid SundVd (0.0100 N) Puitho* 01
(ludordlu to ±00001 N icklrut 00100 N NtOH tlu:
prerlouily boui ituidtrditod n*Jn«t prlmon
poualum icid phth«Ui«.
4.1 8unpUi«.
4.1.1 Pretest Praptntkm. Follow ih« procedur* out-
Unod In Method 5. gallon 411; mttn should b* In-
(Bfctfd, bat need not b» dMleeoiMl w*l|hMt, or ld*ntl-
nra. If the effluent |U can b* cotuidrrad dry , I.*. , rnoU-
tun fm. tb* tillcm f«l nr*d not b* weijhixl
4.1.2 PKlimint/y Dflrrmliuiioru Follow UM pro-
eodura outlined In Method}, Section 4.1.2.
4.1.3 Pnptntion of ColUction Trmln. Follow th* pro-
ndnn outlined In Method 5. Section 413 (eicept tor
lh« tecond pvwraph and other obvlotuly InappUctbW
pvti) tod me Flfure §-1 Initetd of Fifur* VI. Repltco
tb* Mcond piracritpb with: Pl*c« 100 ml of 80 pwnnt
hopropanol in th* flnt Implnfw. 100 ml of S percent
hydrocen peroild* In both ih* second and third 1m-
pinc*n; rouln » portloa of **eh miini tor no* M •
N«nk wlutloo. Pbo* •boot 100 1 of itUct, |tl In tb* tain*
Iraplnmr.
man.
ItCATIOR-
•KRATOM.
•ATI
•0.
SAMPLE MX «0,
WTEMANf
C FACTO*
HTOT TUU COEFFICIEHT. C» .
HATtC PMEttVM. M N ttk N)
AMIIENT TEMKMATUBf
IAROMETHIC PMESSUNI
ASSUMED MOISTURC. %
PROK LENGTH, • (ft)
SCHEMATIC OF STACK CROSS SECTION
NOZZLE IDENTIFICATION NO
AVERAGE CALIBRATED NOZZLE DIAMETER,
PROiE NEATER SETTING
LEAK RATE, »J/mi«,|e»i»)
PROiE LINER MATERIAL
FILTER HO.
TBAVERM POINT
NUMIEK
TOTAL
SAMFLM*
TME
U)).*to.
AVERAGE
VACUUM
MHt
te-ttoj
STACK
TEMFERATURI
.
-------
(plui abaorblngiolutlon) t
these weights. The weight
plus container) muit also I
NOTI.— It molitun content U to b« determined by
Implnger analysis, weigh each ol the flnt three implnfen
(plui abaorbtng_solutlon) to tbe nearest 0.5 I end record
i weight of the silica gel (or silica (el
lit also be determined to tbe nearest
0.5 1 and recorded.
4.1.4 Pretest Leak-Check Procedure. Follow the
basic procedure outlined In Method 5, Section 4.1.4.1,
noting that the probe better sbtll be adjiuted to tbe
minimum tempenture required to prevent condensa-
tion, and also that verbage such as, • • plugging the
Inlet to tbe (liter bolder • • • ," shell be replaced by,
..... plugging the Inlet to the flat Implnier • • *.''
Tha pretest leak -check U optional. 8/
4.14 Train Operation. Follow the bade procedures
outlined In Method 5, Section 4.1.5, In conjunction with
tbe following special Instructions. Data shall be recorded
on a sheet similar to the one In Figure 8-2. The sampling
rat* shall not esceed 0.030 m'/mfn (1.0 cfm) during tbe
ran. Periodically during the test, observe the connecting
Una between the probe and first Implnger for signs of
condensation. If It does occur, adjust the probe beater
sitting upward to the minimum temperature required
to prevent condensation. If component chanies become
necessary during a run, a leak-check shall be done Im-
mediately before each change, according to the procedure
outlined In Section 4.1.4.2 of Method 5 (with appropriate
modifications, as mentioned In Section 4.1.4 of this
method); record all leak rate*. If tbe leakage rated)
etoaed the specified rate, the tester shall either Told tbe
run or shall plan to correct the sample volume as out-
lined In Section 64 of Method ft. Immediately after com-
ponent changes, leak-checks are optional. If these
leak-checki an done, the procedure outlined In Section
4.1.4.1 of Method 5 (with appropriate modifications)
shall be used.
After turning off tbe pump and recording the final
readings at the conclusion of each run, remove the probe
from the stack. Conduct a post-test (mandatory) leak-
ebeck as In Section 4.1.4.1 of Method 5 (with appropriate
modification) and record the leak rate. If the post-test
leakage rate eiceeds tbe specified acceptable rate, the
tester shall either correct the sample volume, as outlined
la Section t.3 of Method 5. or ihall void the run.
Drain the Ice bath and, with the probe disconnected,
purge the remaining part of the train, by drawing clean
ambient air through tbe system for 15 minutes at the
avenge flow rate used for sampling.
NOTE.— Clean ambient air can be provided by passing
air through a charcoal filter. At tbe option of the tester,
ambient air (without cleaning) may be used.
4.1.8 Calculation of Percent Isoklnetlc. Follow tbe
procedure outlined in Method 5, Section 4.1.t.
4J Sample Recovery.
4J.1 Container No. 1. If a moisture content analysis
If to be done, weigh tbe flnt Implnger plus contents to
tbe nearest 0.5 g and record this weight.
Transfer the contents of tbe first Implnger to a ISO-mi
graduated cylinder. Rinse the probe, Ant Implnger. all
connecting glassware before the filter, and the front half
of the filter bolder with 80 percent taopropenol. Add the
rinse solution to tbe cylinder. Dilute to 250 ml with W
percent Isopropanol. Add tbe filter to the solution, mix,
and transfer to the storage container. Protect the solution
against evaporation. Mark tbe level of llojiid on the
aontalnerandldentlfytheaamplecontalner. V
4.U Container No. I. U a moisture content analysis
Is to be done, weigh the second and third Implngen
(plus contents) to the nearest O.S g and record thesa
weights. Also, weigh tbe spent silica gel (or alllca gel
pluslmplnger) to tbe nearest 0.5 g.
Transfer the solutions from the second and third
Impingen to a 1000-ml graduated cylinder. Rinse all
connecting glassware (Including back half of filter holder)
between the filter and silica geTlmplnger with delonlied,
distilled water, and add this rinse water to tot cylinder.
Dilute to a volume of 1000 ml with delonlsed, distilled
water. Transfer the solution to a storage container. Mark
the level of liquid on tbe container. Seal and Identify the
sample container.
«/ Analysis.
Note the level of liquid In contslnen t and 3, and con-
firm whether or not any sample was lost during ship-
ment; note thli on the analytical date sheet. If a notice-
able amount of leakage has occurred, either void the
sample or use methods, subject to the approval of the
Administrator, to correct the final results. •
44.1 Container No. 1. Shake the container holding
the Isopropanol solution and tbe Alter. If tbe filter
breaks up, allow the fragments to settle tor a few minutes
baton removing a sample. Pipette a 100-ml aliquot of
tbte solution Into a 150-ml Erlanmeyer flask, add 2 to 4
drops of thorin Indicator, and titrate to a pink end point
using 0.0100 N barium parcblorate. Repeat the tltratlon
with a second aliquot of sample and average the Utratton
, JUpUoste utratlons malt agree within 1 psntttt
r dl ml, whichever Is greater.
4JJ Container No. t Thoroughly mix tbe aslotton
n the oontainar holding the contents of the second and
Utd Implngers, Pipette a 10-ml allo.nct of sample Into a
Bft-ml Krlenmeyer flask. Add 40 ml of Isopropanol, I to
I drops of thorin Indicator, and titrate to a pick endpolnt
using 0.0100 N barium percblorate. Repeat the tltratlon
witb a second aliquot of aample and average tbe tltratlon
values. Replicate titntlons mustagree within 1 percent
or 0.9 ml, whichever 1s greater. 87
444 Blanks. Prepare blanks by adding 2 to 4 drops
of thorin indicator to 100 ml of 80 percent Isopropanol.
Titrate the blanks In the same manner as tbe samples.
(. CWfrarfen
(.1 Calibrate equipment using tbe procedures speci-
fied In tbe following sections of Method 6: Section 54
(metering system); Section 6.5 (temperature gauges);
Section 5.7 (barometer). Note that the recommended
leak -check of the metering system, described In Section
{.« of Method 5, also applies to this method.
(.2 Standard! w tbe barlnm perchlonte solution witb
U ml of standard suUurtc acid, to which 100 ml of 100
percent Isopropanol has been added.
eolate the moisture content of tbe stark gas, using Eo.ua-
tton {-« of Method 6. Tbe "Note" In Section 8.irfMetKod
( also applies to this method. Note that If the affluent su
stream can be considered dry, tbe volume of water vanor
and moisture content need not be calculated
6.6 Sulfuric acid mist (including BOO concentration
Maid)
Equation 8-2
JTi-0.04t04 g/mlUleqnlvalcnt for metric unite
-1.081X10-1 Ib/meq for English units.
8.8 Sulfur dioxide oonoentration.
Note.— Carry out calculations retaining at least one
extra decimal figure beyond that of the acquired data.
Round off figures after final calculation.
•.1 Nomenclature.
X.- Cross-sectional area of nostle, m'(ft')-
B«-WaUr vapor in the gas stream, proportion
by volume.
C,,r»^ -Sulluric acid (Including SOi) concentration,
g/dscm (Ib/dscf). 87^
C*J -Sulfur dioxide concentration, g/dscm (lb/
/•Percent of Isoklnetlc sampling.
^-Normality of barium parcblorate tltrant. g
0 equivalents/liter.
"o.r- Barometric pressure., At tbe sampling site,
mm Eg (In. Hg). «r
P.-AbsoluU stack gas pressure, mm Bg On.
"-(•id)
Equation 8-3
I for metric unite.
>/meq tor English 1
•.7.1 Calculation from raw data.
. 100 T.t
Imm Ii
-Standard absolute
n. Hg).B/
WtV.P.A,
.„ preesun, 710 mm Bg
(29.92 in. Hg). 87
T«—Average absolute dry gas meter temperature
(seeFlgure8-2).>KC R).
r.-Average absolute stack gas temperature (see
,_ Figure 8-2),* KCR).
1 «id -Standard absolute tempenture, M* X
(S2S» H).B7 "^
V.-Volume of sample aliquot titrated, 100 ml
for H»8Oi and 10 ml for 80i.
Vi,-Total volume of liquid collected In Implngers
and silica gel, ml.
V.-Volume of gas sample as measured by dry
., gas meter, dcm (del).
v mIII-Total volume of solution In which tbe
auUurlr acid or sulfur dioxide sample b
contained, ZW ml or 1,000 ml, nspectlvely.87
Vi-Volume of barium perchlonte Utrant used
for the sample, ml.
V»-Volume of barium perchlonte tltrant used
for tbe blank, ml.
V-Dry gas meter calibration factor.
Atf-Avtrue pressure drop across orlfies meter,
mm (in.) HiO.
• •Total sampling tuns, mln.
l».8-8peclflc gravity of mercury.
80-SK/mln.
100- Conversion to percent.
8.2 A verve dry gas meter temperature and average
trifle* pressure drop. See date sheet (Figure 8-2).
84 Dry Oat Volume. Correct the sample volume
measured by the dry gas meter to standard oondltloni
f»> C and 780 mm Hg or 88* T and 28.92 In. Bg) by using
Equation 8-1.
Equation 8-t
when:
JKi'0.003484 mm Hg-m'/ml-'K for metric unite
-0.002678 in. HgVmV'RtorEnlush mStT'
7 J CakulatUm from intermediate values?
87
817 J
-g.
Equation 8-5
JTi-4430 (or metric unite.
.
-0.08440 for English unite.
84 Acceptable Results. If (0 percent 04858 *xymm Hg for metric nnlts.
•17.84 *R/m. Bg for English unite.
NOT*.—If the leak nte observed during any manda-
tory leak-checks exceeds tbe specified acceptable rate,
the tester shall either correct the value of V. In Equation
1-1 (as described in Section 84 of Method e), or shall
Invalidate the test run.
M Volnma of Wate* Vapor and Moisture Content.
Calculate the volume of water vapor mini Equation
1-2 of Method «j the weight of water eoUerted intbe
miullltert (the specific gravity of water Is 1 g/ml). Cal-
t. Corbett. P. F. The Determination of SO
to Flue Oases. Jonmalolthelnstituteof Fuel I
1WI>
«. Martin, Robert If. Construction Details of laoklnetlc
•ourot Sampling Equipment. Environmental ProtecUon
Agency. ;ResearchJfrlaiigle Park, N^cTAlr PouSS"
Confrpl Office Publication No. APTD-O81. April 1J71
4. Patton. W. F. and J. A. Brink, JrjfSw Eqnlpm«it
tod Techniques for Sampling Chemical Prooass Daua
*o«rnaJofAfrPoUnUondon^trolA7soclaUon^82 SS
•.Rom. 1.1. Maintenance. Calibration, and OpanUon
«/ IsoUneUc Source-Sampling Bqulpment. Office of
Air Prognms, Environmental Protection Agency.
EesearohTriangle Park, N.C. APTD-0&78. March UA
8. Hamll. H7P. and D. ircuunn CoUaboraUve
Mudy of sietbod tor Determination of BulfuF Dtoilde
Emissions from Stationary Sources (Fossil Fuel-Flred
•team Oenenton). Environmental ProtecUon Agency.
bar
T. Annual Book of ABTIf Standards. Part«; Water,
32
-------
APPENDIX B
ACID MIST PERFORMANCE STANDARD RELATIONSHIPS
Federal regulations (1) state the acid mist performance standard in terms
of a maximum quantity which may be discharged into the atmosphere per ton of
acid produced; 0.075 Kg of mist/metric ton of H2S04. Method 8, however,
determines acid mist as a weight of H2S04 per unit volume of gas sampled and
in order to evaluate results obtained in the present study, it is necessary
to relate these quantities. The volume of gas exiting a contact acid plant
per ton of HgSO^ produced varies with the plant and the operating conditions
employed. However, in a survey of 40 plants (6), the minimum volume to weight
ratio occurred in a plant that produced 150 short tons/day and had an exit gas
discharge of 7,940 SCFM. This corresponds to:
7,940 ft3/min (1440 min/day)(0.02832 m3/ft3)(l day/150 short tons)
(1 short ton/0.908 m ton) = 2370 m3 of exit gas/metric ton of HgSO^
produced (Equation A-l)
Relating this to the standard:
2
75,000 mg of mist/metric ton of acid (1 metric ton/2370 m )
= 32 mg of acid mist/m3 (Fquatlnn A-2)
In the plant under consideration, no dilution air was in use. However, at
other plants employing dilution air, this figure can be reduced to as low as
14 mg/m .
33
-------
TECHNICAL REPORT DATA
EPA 600/4-80-018
ACC;5SiO>NO.
AN EVALUATION STUDY OF EPA METHOD 8
5. REPORT DATE
March 1980
6. PERFORMING ORGANIZATION CODE
•O '
3. PERFORMING ORGANIZATION REPORT NO.
Joseph E. Knoll and M. Rodney Midgett
IC^ NAME AND AOOR6SS
10. PROGRAM ELEMENT NO.
Quality Assurance Division
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Research Triangle Park. NC 27711 _
11. CON TRACT/GRANT NO.
12. SPONSORING ACENCV NAME AND ADDRESS
Environmental Monitoring Systems Laboratory
3ffice of Research and Development
J.S. Environmental Protection Agency
Research Triangle Park. NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA 600/08
, •-U*''"_t .lc'. r AnY NOTES
To be published as an Environmental Monitoring Series report.
16 «
Techniques used in EPA Method 8, the source test method for acid mist and sulfur
dioxide emissions from sulfuric acid plants, have been evaluated. Evidence is shown
that trace amounts of peroxides in isopropyl alcohol result in the conversion of
sulfur dioxide to sulfate and cause positive errors in acid mist values. Methods for
measuring and purifying IPA are described. No conversion of sulfur dioxide to sulfate
on filters or filter supports were observed. Collection efficiencies of train com-
jonents are described and two alternate indicators are evaluated. Solid ammonium
sulfates's use as audit samples is discussed.
Field testing is also describedJn which paired-probe techniques were employed.
They showed that, when sulfur trioxide is absent from the effluent streams,.acid mist
is efficiently collected by a single filter, even when the isopropyl alochol-containing
impinger is eliminated. Both ammonia and dimethyl analine, which are employed as gas
scrubbers, cause sulfur dioxide to be retained in the isopropyl alcohol and result in
large positive interferences in acid mist values. Ferric oxide, present in the
effluents of steel pickling operations, causes a large negative Interference in acid
mist values.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS
COSATI I icKI/GtOUp
air pollution
gas sampling
acid mist
sulfur dioxide
43F
68A
I'J it CuHl r Y CLAjS / I n,l lit purl,
21
------- |