a EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA 450/4-90-012
May 1990 "'
Air
TEST REPORT
Method Development and Evaluati
! s| !»"
of Draft Protocol for Measurement Of
Condensible Particulate Emissions
control
technology center
H -;<
-------
-------
EPA 450/4-90-012
TEST REPORT
METHOD DEVELOPMENT AND EVALUATION OF DRAFT PROTOCOL
FOR MEASUREMENT OF CONDENSIBLE PARTICIPATE EMISSIONS
CONTROL TECHNOLOGY CENTER
SPONSORED BY:
Emission Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Air and Energy Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Center for Environmental Research Information
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45266
MAY 1990
U.S. Environmental Protection Agency
Region 5, Library (5PL-16)
230 S. Dearborn Street, Room 1670
Chicago, IL 60604
-------
EPA 450/4-90-012
May 16, 1990
TEST REPORT
METHOD DEVELOPMENT AND EVALUATION OF DRAFT PROTOCOL
FOR MEASUREMENT OF CONDENSIBLE PARTICULATE EMISSIONS
by
William G. DeWees
Kathy C. Steinsberger
CEM/Engineering Division
Entropy Environmentalists, Inc.
P. 0. Box 12291
Research Triangle Park, North, Carolina 27709
EPA Contract No. 68D90055
Work Assignment Number 13
Project Officer
Candace Sorrel 1
Emissions Measurement Branch
Technical Services Division
Office of Air Quality Planning and Standards
U. S. Environmental Protection Agency
Prepared for:
Control Technology Center
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
ACKNOWLEDGEMENTS
The Information in this document has been funded in part by the U. S.
Environmental Protection Agency (EPA) under Contract Nos. 68-02-4336 and
68D90055 to Entropy Environmentalists, Inc. Technical assistance was provided
by Gary D. McAlister, Roger T. Shigehara, Candace B. Sorrel1 and Lori T. Lay
of EPA's Emission Measurement Branch, Technical Support Division, Office of
Air Quality Planning and Standards and Thomas E. Ward of EPA's Quality
Assurance Division, Atmospheric Research and Exposure Assessment Laboratory,
Office of Research and Development.
-------
PREFACE
The CTC was established by EPA's Office of Research and Development
(ORD) and Office of Air Quality Planning and Standards (OAQPS) to provide
technical assistance to State and Local air pollution control agencies. Three
levels of assistance can be accessed through the CTC. First, a CTC HOTLINE
has been established to provide telephone assistance on matters relating to
air pollution control technology. Second, more in-depth engineering
assistance can be provided when appropriate. Third, the CTC can provide
technical guidance through publication of technical guidance documents,
development of personal computer software, and presentation of workshops on
control technology matters.
The technical guidance projects, such as this one, focus on topics of
national or regional interest that are identified through contact with State
and local agencies. In this case, the CTC became interested in evaluating a
test method for measuring condensible particulate matter (CPU) in response to
a number of requests from State and local air pollution control agencies.
This document presents laboratory and field evaluations on the impinger catch
method. The impinger catch is the back half portion of the EPA Method 5
sampling train. The principal objective of this evaluation was to determine
the adequacy of the test method and produce documentation to support it. This
was accomplished through laboratory experiments and field tests.
-------
CONTENTS
Acknowl edgments i i
Preface i i i
Figures v
Tables vi
1. Introducti on 1
1.1 Background 1
1.2 Objecti ves 2
1.3 Basic Technical Approach 2
2. Conclusions and Recommendations 3
3. Laboratory Evaluation 4
3.1 Evaluation of Organic Matter Analysis 4
3.2 Evaluation of Inorganic CPM Sampling and Analysis 8
4. Field Evaluation Test 1 22
5. Field Evaluation Test 2 27
6. Field Evaluation Test 3 31
7. References 35
Appendix A. Draft of Method 202 - Determination of Condensible
Particulate Emissions from Stationary Sources 36
Appendi x B. Data Assessment 47
-------
Figures
Figure Page
Number Number
3.1 Scheme for evaluation of inorganic matter sampling and 10
analytical procedures
3.2 Laboratory setup for N2 purge and S02 sampling 11
4.1 Scheme for preparation and analysis of impinger samples 23
-------
Tables
Table Page
Number Number
3.1 Organic Results for Sample L from a Coal-Fired Boiler 6
3.2 Organic Results for Sample A from a Wood-Fired Boiler 9
3.3 Results from Evaluation of Inorganic Condensible Determination 13
3.4 Evaluation of Cl" and S04= Loss on Evaporation of Aqueous 16
Matrix
3.5 Evaluation of Cl" and S04= Loss on Evaporation of Aqueous 17
Matrix
3.6 Air Purge Experiments To Determine Removal of S03= 19
3.7 Summary of EMB Tests Conducted November 17-18, 1988 at 20
Coal-Fired Boiler
4.1 Determination of Ammonia Addition and Mass of Condensible 24
Matter from Coal Fired Boiler
4.2 Ion Chromatographic Results for the Sulfate, Chloride, and 25
Nitrate In Coal-Fired Boiler Samples
5.1 Determination of Ammonia Addition and Mass of Condensible 28
Matter from Wood-Fired Boiler
5.2 Wood-Fired Boiler Ion Chromatographic Results for Sulfate 29
and Chloride
6.1 Determination of Ammonia Addition and Mass of Condensible 32
Particulate Matter from Coal-Fired Boiler
6.2 Ion Chromatographic Results for Sulfate, Chloride, and 33
Nitrate for Flue Gas Samples from a Coal-Fired Boiler
VI
-------
1.0 INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is planning to propose in
Appendix M, 40 CFR Part 51, the impinger catch method for measuring
condensible particulate matter (CPM). To support this action, the Emission
Measurement Branch (EMB), Office of Air Quality Planning and Standards, EPA,
contracted with Entropy Environmentalists, Inc., to evaluate the impinger
catch approach for measuring CPM. This report details the laboratory and
field evaluations done under EPA Contract Nos. 68-02-4336 (Work Assignments 23
and 34) and 68-D-90055 (Work Assignment 13).
1.1 BACKGROUND
EPA considers all CPM to be PM,0, i.e., particulate matter (PM) of 10 pm
or less in aerodynamic diameter. Since the current PM10 test methods measure
only in-stack PM, EPA concluded that a CPM method was vital.
A candidate method was formulated based on a literature review1"9 and
consideration of the ease of execution, availability of equipment, analytical
detection limits, sensitivity and range, and sample preparation. EPA
concluded that the impinger catch approach would be most desirable because it
(1) allows the determination of both filterable PM and CPM simultaneously,
(2) uses existing methodology and equipment, and (3) was being used by several
State agencies.
The impinger catch is the back half portion of the EPA.Method 5 train.
The determination of CPM involves: (1) purging the impinger solution with an
inert gas after the test run; (2) extracting the impinger solution to separate
the organic from the inorganic CPM fraction; (3) drying the organic fraction
at room temperature and the inorganic fraction at 105*C; and (4) weighing the
residues.
Before this method could be proposed, EPA felt that several areas needed
to be investigated. They were as follows:
(1) Formation of "false" CPM. Since gases are bubbled through water,
noncondensible gases may react with other gases or condensibles to form CPM
that would not have otherwise formed. The most notable case of this is the
oxidation of SO^ to form S04=. The S02 dissolves in water to form H2S03, which
may oxidize to form H2S04. This S04= would then be counted as CPM. One
document1 showed that purging the impinger solution with air immediately after
sampling effectively removed S0?. However the air purge raised questions
concerning the possible conversion of S02 to sulfur trioxide (S03); therefore,
further evaluation of the effectiveness of an air or N2 purge to remove S02
was needed.
(2) Extraction solvents. Three solvents are commonly specified by State
agencies for extracting CPM from impinger solutions: methylene chloride
(MeCl2), chloroform-ether (C-E), and Freon. The effectiveness of these
solvents needed to be evaluated.
-------
(3) Weighing to a constant weight. The hygroscopic nature of H2S04
makes weighing of the residue to a constant weight difficult. In one State
procedure, NH4OH is added to the impinger solution to form the S04". The
accuracy of this approach needed to be documented.
1.2 OBJECTIVES
The principal objectives of the evaluation were to determine the adequacy
and produce documentation to support the test method, revise the candidate
method based upon the results of the laboratory experiments, and validate the
method in field te*sts and revise the method, if necessary.
1.3 BASIC TECHNICAL APPROACH
The basic approach consisted of two phases. The first was to evaluate
the candidate method in the laboratory as follows: (1) Determine the
effectiveness of the three different solvents being used by some State
agencies to extract the organic fraction at three different pH levels;
(2) determine the effectiveness of the N2 purge to remove dissolved S02 from
the impinger solution; and (3) revise the candidate method based on the
laboratory evaluation.
The second phase was to evaluate the revised test method in the field as
follows: (1) Determine the precision of the revised method at three different
field sites; (2) conduct any other experiments in the laboratory or in the
field to resolve the resulting questions; and (3) finalize the revised test
method into Federal Register format.
-------
2.0 CONCLUSIONS AND RECONNENDATIONS
As a result of the laboratory and field evaluations of the impinger catch
approach for determining CPM, the conclusions and recommendations are as
fol1ows:
1. When S02 is present at significant levels, the sample should be purged
with N2 at a rate of 20 liters/min for 60 minutes immediately after the
end of the test run. An air purge is not recommended, but could be
allowed.
2. If S02 is not present at significant levels, a purge is not necessary.
3. If the pH of the sample is less than 4.5, then NH4OH should be added to
the sample to stabilize H2S04 and provide accurate weighing.
4. When NH4OH is added, the amount retained by the sample should be
determined from the sulfate (S04=) analysis using ion chromatography
(1C). The mass of NH4OH retained may also be determined by titrating the
sample to a pH of 7.0 using 0.1 N NH4OH and a pH meter.
5. If NH4OH is to be used, the sample should be dried at 105'C and the
residue resuspended before adding the NH4OH. The sample may also be
analyzed for chlorides (CT) following the final weight determination and
the amount of ammonium chloride (NH4C1) may be subtracted from the final
weight.
6.
MeCl2 is recommended as the organic extraction solvent. C-E is
acceptable as a secondary choice. Freon is unacceptable.
7. When both organic and inorganic CPM is being determined, the pH of the
sample need not be adjusted before extraction. The extraction should be
repeated three times.
8. Four Method 5 type impingers should be used in the sampling train as
follows: Two should be tipped (Greenburg-Smith design) and the last two
should have the modified tip. In each of the first three impingers,
100 ml of deionized distilled water should be added, and 200 g of silica
gel should be placed in the fourth.
9. The precisions (standard deviations) for the draft method based on tests
at a wood waste boiler and two coal-fired boilers are 13.0 ±2.1 mg/m3,
3.5 ±'1.1 mg/m3, and 39.5 ± 9.0 mg/m3, respectively. Further studies are
recommended to determine whether precision of weighing low loadings (less
than 10 mg) can be improved.
10. The method may not be applicable at sources that contain high levels of
ammonia (e.g., when ammonia injection is used as a control technique).
For such sources, the amount of NH4C1 may be subtracted from the final
weight. Additional studies at sources with ammonia emissions are
recommended to determine if the N2 purge is still effective in removing
the S02 from the impinger solution under these conditions.
-------
3.0 LABORATORY EVALUATION
This section describes the experimental design of the laboratory
evaluation, data, and conclusions.
To provide typical sample matrix effects, samples of impinger contents
from previously conducted Method 5 test runs (mostly from studies conducted
for EPA) were used. These samples had not been subjected to the N2 purge.
The sources of the samples were a coal-fired boiler, an oil-fired boiler,
a mass burn municipal waste combustor, a refuse-derived fuel municipal waste
combustor, and wood-waste boilers.
3.1 EVALUATION OF ORGANIC MATTER ANALYSIS
In the laboratory, two parameters of the organic CPM analysis were
evaluated: (1) the effectiveness of organic extraction by MeCl,, C-E, and
Freon, and (2) the effect of pH on extraction. The effect of pH was evaluated
because of the different solubilities of organics. An alternative technique,
solid phase extraction (SPE), was also evaluated because of its potential to
(1) reduce the time of extraction step, (2) reduce the amount of solvent to be
evaporated 50-fold, and (3) increase the analytical precision.
To provide sufficient volume of impinger contents for comparisons of the
solvent extractions, the impinger contents from all sample runs within a test
series were combined into a single sample. The source categories used were a
wood-fired boiler (high in organic matter and low in inorganic), and a
coal-fired boiler (low in organic matter and high in inorganic).
The steps used to compare the different solvents at different pH levels
are given below.
For the coal-fired boiler, the samples were treated as follows:
1. The sample was filtered through a 0.2 urn Teflon filter and then the
pH was determined.
2. The sample was shaken vigorously and divided into 8 equal aliquots.
3. Aliquots 1, 2, and 3 were treated as follows:
a. Each aliquot was extracted using separatory funnels at the pH
of the sample (2.0) with MeCl2, C-E, or Freon. Each aliquot
was extracted four times.
b. A 5-ml portion of the aqueous fraction was analyzed for CV,
nitrates (N03~), and S04= by 1C.
c. NH4OH was added to the remaining aqueous fraction. This
fraction was dried, desiccated for 24 hours, weighed, and then
reweighed after two additional hours. To assure that the
-------
residue was completely dry, they were reheated, cooled,
desiccated overnight, and reweighed.
d. The residue from Step 3c was then redissolved in water and
analyzed for Cl", N03", and S04*.
4. Aliquots 4 and 5 were treated as follows:
a. Each aliquot was extracted as in Step 3a above with MeCl2 or
Freon, and analyzed for anions as in Step 3b.
b. The aqueous fraction was then adjusted to a pH of 12 with NH4OH
solution, reextracted as described in Step 3a above, and then
treated as in Step 3c without further addition of NH4OH.
5. Aliquot 6 was treated as follows:
a. One sample aliquot was passed through a bonded octadecyl si lie
(C-18) sorbent.
b. The organic matter on the C-18 sorbent was eluted sequentially
with four separate 2-ml portions of methanol. Each portion was
analyzed separately.
c. The aqueous fraction (part passing through the C-18 sorbent)
was analyzed as in Step 3b - 3d above.
6. Aliquots 7 and 8 were treated as follows:
a. The sample aliquots were adjusted to a pH of 7 with NaOH.
b. The samples were then treated as in Step 5b above.
The results from the coal-fired boiler test are presented in Table 3.1.
As expected, the fraction of organic CPM in the samples was low. Because of
the small amount of mass in each organic sample fraction, the relative
precision was poor. In the cases where 0.0 mg is shown for sample weight, the
actual weight may have been negative. However, the organic CPM weights of 1
to 2 mg constituted only about 1 percent of the inorganic CPM weights of 170
to 183 mg and thus will not affect the precision of the total CPM
determinations. The precision of the analysis for total CPM from the
coal-fired boiler was 178 ± 5 mg. The standard deviation was 2.8 percent.
The variable results for samples extracted at a pH of 2 could not be
traced to any particular solvent. When the sample was adjusted to a pH of 12,
some additional organics were extracted but the amount was not significant.
The most significant discovery was that the level of the Cl" (40 to
57 mg, which was about equal to the mass of S04=) in the aqueous fraction both
before and after it was taken to dryness was unexpectedly high. No N03" were
found in the samples.
-------
TABLE 3.1. ORGANIC RESULTS FOR SAMPLE L FROM A COAL-FIRED BOILER
Aliquot No.
Fraction
First extract
Second extract
Third extract
Fourth extract
Total
1 4
Trichloro-
trifluoro-
e thane
2 5
Methylene
Chloride
3
Chloroform/
Diethylether
6
7 8
C-18
]?H 2
PH 7
Extraction at sample pH (2.0)a
o.o 0.5
0.2 0.5
0.4 0,4
0.4 0.5
1.0 . 1.9
0.2 0.3
0.5 0.4
0.2 0.4
0,5 0.7
1.4 1.8
0.4
0.2
0.7
0.6
1.9
0.4
0.5
0.2
0.2
1.3
0.6 0.6
0.6 0.7
0.6 0.4
0.2 0.0
2.0 1.7
Extraction at pH of 12
First extract
Second extract
Third extract
Fourth extract
Total
Total Organic
0.0
0.0
0.1
0.0
0.1
1.0 2.0
0.6
0.0
0.0
0.0
0.6
1.4 2.4
1.9
1.3
2.0 1.7
Gravimetric Results for Aqueous Fraction
Aqueous Residue
(NHft)2S04 Corr.
Total
Condensibles
192 195
175 180
176 182
196 198
178 183
179 185
193
173
175
193
170
172
Aqueous Fraction Ion Chromatography Results
Chloride
Sulfate
52.5 40.7
50.3 43.0
54.0 40.2
51.7 41.9
54.0
58.4
57-0
64.6
Ion Chromatography Results of Resuspended Residue
Chloride
Sulfate
46.5
49.5
47-9
51.5
47.9
52.2
aExcept for C-18 extractions, which were extracted at the pH shown above.
Note: All results are in mg and represent the total for a 150 ml sample aliquot.
-------
The combined mass of the SO/ (42 to 65 mg) and CT (40 to 57 mg)
totalled only about one-half of the inorganic CPM (170 to 183 mg). SO," were
anticipated to constitute the majority of the inorganic CPM. Since this was
not the case, determining only the amount of S03 in the emissions to represent
CPN is not a viable approach.
For the wood-fired boiler, the samples were treated as follows:
1. The sample was filtered through a 0.2 jim Teflon filter and then the
pH was determined.
2. The sample was shaken vigorously and divided into 6 equal aliquots.
3. Aliquots 1, 2, and 3 were treated as follows:
a. Each aliquot was extracted using separatory funnels at the pH
of the sample (4.0) with MeCl2, C-E, or Freon. Each aliquot
was extracted four times.
b. NH.OH was added to the aqueous fraction. This fraction was
dried, desiccated for 24 hours, weighed, and then reweighed
after two additional hours. To assure that the residue was
completely dry, they were reheated, cooled, desiccated
overnight, and reweighed.
c. Before adding the NH,OH, a 5-ml portion of the aqueous fraction
of one of the aliquots was taken and analyzed for Cl", N03",
and S04= by 1C. After this aliquot was dried and weighed, the
residue was redissolved in water and analyzed for the same
anions.
4. Aliquots 4 and 5 were treated as follows:
a. Each aliquot was adjusted to a pH of 2 with concentrated nitric
acid and were extracted as above with Freon or C-E.
b. The aqueous fraction was then adjusted to a pH of 12 with NaOH.
One aliquot was reextracted as described above and the other
was extracted only once.
5. Aliquot 6 was treated as follows:
a. The sample aliquot was passed through a bonded octadecyl silic
(C-18) sorbent.
b. The organic matter on the C-18 sorbent was eluted sequentially
with four separate 2-ml portions of methanol. Each portion was
analyzed separately.
c. The aqueous fraction (part passing through the C-18 sorbent)
was analyzed as in Step 3b above. After this fraction was
-------
taken to dryness and weighed, the residue was redissolved In
water and analyzed for Cl", N03", and S04" by 1C.
Table 3.2 presents the results from the wood-fired boiler test. The mass
of organic CPM was high as had been anticipated. However, the mass of
Inorganic CPM in the aqueous fraction was significantly higher than had been
expected. The amount of S04" and Cl~ was low, and N03" were not found.
The pH levels of 2 and 4 did not affect the solvent extractions. MeCl2
yielded 16 mg (pH of 4), C-E yielded 15.5 (pH of 4) and 16.5 mg (pH of 2), and
C-18 yielded 15.6 mg (pH of 4). Freon gave 1.9 (pH of 4) and 2.6 mg (pH
of 2). Extractions at a pH of 12 gave an additional 25 percent CPM for all
solvents.
Although Freon extracted only about 2 of the 16 mg of organics, the
majority of the organics left in the aqueous fraction were analyzed as
inorganics. The total CPM for the samples extracted with Freon was the same
for those samples extracted with the other solvents. Therefore, Freon was
considered to be an unacceptable solvent if only organic CPM is desired but
could be acceptable if total CPM was being measured. However, since Freon is
a chloroflorohydrocarbon, its use is not recommended.
The precision of the total CPM analysis for wood-fired boiler was
46.6 ± 2.0 mg, which translates to a relative standard deviation of
4.3 percent.
Based on the results of both the coal-fired and wood-fired boilers, three
extractions are considered to be sufficient, the pH of the sample need not be
adjusted if total CPM (i.e., organic plus inorganic CPM) is being determined,
but if only the organic CPM is being determined, the samples should be
extracted at both pH's of 2 and 12.
3.2 EVALUATION OF INORGANIC CPM SAMPLING AND ANALYSIS
This part of the CPM method evaluation focused on the following: (1) Is
organic material removed during the purging process? (2) Can the S02
converted to sulfites in the impingers be removed by purging? (3) Is S02
converted to S04= during sampling? (4) What portion of the inorganic matter
is S04=? (5) Can inorganic CPM be gravimetrically determined accurately when
H2S04 is a significant portion of the aqueous fraction?
The overall scheme of the laboratory experiment is shown in Figure 3.1
and is described below. Samples from combustion sources with high levels of
S02 were used for this experiment.
1. The laboratory setup (see Figure 3.2) included an S02 gas cylinder
and a N2 gas cylinder, two impingers for the sample evaluation, one
impinger to remove S02, one impinger to remove moisture, an S02
analyzer, and a hydrocarbon analyzer. Excess S02 was vented
outside. The evaluation was conducted on nine samples and a water
blank (see Table 3.3).
8
-------
TABLE 3.2. ORGANIC RESULTS FOR SAMPLE A FROM A WOOD-FIRED BOILER
Aliquot No.
Fraction
1 4
Trichloro-
trifluoro-
e thane
pH 4 pH 2
2
Methylene
Chloride
pH 4
3 5
Chloroform/
Diethylether
pH 4 pH 2
6
C-18
pH 4
Extraction at Sample pH 4 and Adjusted pH 2
First extract
Second extract
Third extract
Fourth extract
Total
1.4
0.4
0.1
0.0
1.9
1.7
0.0
0.7
0.2
2.6
10.8
2-3
1.6
1-3
16.0
11.8
1.4
1.4
0.9
15-5
12.2
2.4
0.8
1.1
16.5
14.0
1.5
0.0
0.1
15.6
Extraction at pH of 12
First extract
Second extract
Third extract
Fourth extract
Total
Total Organic
0.8
0.8
1.9 3.4
16.0
1.4
1.4
0.0
1.2
4.0
15-5 20.5
15-6
Gravimetric Results for Aqueous Fraction
Aqueous Residue
(NH4)2S04 Corr.
46.4 '
43.4
33-9
30.9
32.3
29-3
36.6
33-6
Aqueous Fraction Ion Chromatography Results
Chloride
Sulfate
NAa
NA
NA
NA
0.9
0.6
NA
NA
Ion Chromatography Results of Resuspended Residue
Chloride
Sulfate
Total
Condensibles
NA
NA
45.3
NA
NA
46.9
1-3
5-7
44.8
1.1
5-5
49.2
Note: All results are in mg and represent the total for a 150 ml sample
aliquot.
'Not analyzed.
-------
Method 5 impinger solution
3 or more samples combined;
filter through 0.2 urn Teflon filter
20 IpmNg purge
for 20 minutes;
monitor SO 2 and THC
in purge continuously
Sample 250 ppm SO 2
for 2 hours; follow
with a 20 Ipm N 2 purge
until SO2 in purge is
negligible as measured by
SO2 continuous monitor
Mix well and
split sample
1. Dry to constant weight
2. Weigh residue
Remove 4 ml from
each impinger,
combine, and
determine 503
andSOd= by 1C
Remove 4 ml from
each impinger,
combine, and
determine 503
and SO4= by 1C
If SO 3 andSO4=
are significant,
consult Task Manager
Treat with ammonium hydroxide
1. Dry to constant weight
2. Weigh residue
3613 3/90
Figure 3.1. Scheme for evaluation of inorganic matter sampling and analytical procedures.
10
-------
O
9
CO
Q.
a
U)
C\J
8
CD
01
Q.
CM
Q.
CM
CO
2
5
CJ
CMQ
2 2
_i
O
a:
°
o
n
-------
2. The volume of each sample was determined, then the sample was
divided in two and placed into the first two impingers in an ice
bath. The impinger train was purged with N2 at a rate of
20 liters/min. During the purge, the S02 and organics exiting the
train were monitored continuously. The purge was conducted for
60 minutes or until S02 was no longer detected by the analyzer.
3. Two 4-ml aliquots, one from each of the first two impingers, were
taken, combined with 1 ml of formaldehyde, and brought to volume in
a 10-ml volumetric flask using deionized (01) water. The combined
impinger/formaldehyde samples were analyzed in duplicate for CT,
sulfites, N03', and S04S by 1C.
4. S02 at a concentration of approximately 250 ppm was bubbled through
the impingers for 2 hours at a rate of 20 liters/min to saturate the
water.
5. At the conclusion of the 2-hour purge using S02, the flow of S02 was
shut off and the N? gas cylinder was attached. The impingers were
left in the ice bath. A N2 purge was conducted at a rate of
20 liters/min until S0? was low (near zero) as detected by the
analyzer. The time it took to remove the sulfites from the impinger
contents was recorded.
6. Two.4-ml aliquots, one from each of the first two impingers, were
taken, combined and diluted as above, and analyzed'in duplicate for
S04=, Cl", N03~, and sulfites by 1C. If, after conducting the SO/
and sulfite analyses on two samples, the results had indicated that
(1) S02 was converted to S04= during the sampling phase, or (2) S02
was converted to sulfite and the sulfite could not be removed using
a N2 purge, the laboratory experiments would have been suspended and
the results reported to EPA.
7. The sample remaining in the first two impingers was combined,
vigorously shaken, and divided in half. NH4OH was added to one of
the sample halves. The samples were kept separate and were taken to
dryness in an oven at IIO'C. They were then desiccated for 24 hours
and the residues weighed. They were reweighed approximately 2 hours
later and both the before- and after-weights were recorded. The
sample residues were then reheated, cooled, desiccated overnight,
and reweighed to assure that the samples were dry. Using the amount
of S04= in each sample, the weight contributed by the NH4OH was
determined and subtracted from the weight of that sample residue.
The results of this part of the evaluation are presented in Table 3.3.
The water blank sample showed that S02 was not removed even after 2 hours of
purging with N2. It was later determined that the removal of S02 from the
impinger contents is directly related to the pH of the solution. The SO? is
more easily removed at a pH of 2 than at above 4. Sources that contain S02
would typically yield impinger solutions with a pH of about 2.
12
-------
TABLE 3.3. RESULTS FROM EVALUATION OF INORGANIC CONDENSIBLE DETERMINATION
Residue Weights (mg)
Un- NH4OH NH4OHb
treated Treated Corrected
1C Results (mg)a
as NH3
salt S04m
Impinger Concentrations
(mg/ml)
SO,,' S03= Cl N03'
Run A: Blank - Required 120-minute N2 purge to reach 0.24 ppm SO,
Before
After
5-8 6.4 5-7
3.0 2
Difference 2
.4
ND
0.0099
.4
ND
0.0059
ND
ND
ND
ND
Run B: Coal-fired boiler - Required 30-minute N_ purge to reach 0.08 ppm SO,
Before
After
42.4 52.8 46.6
Difference
15
24.6 18
2
.8
.2
.4
0
0
0
.0613
.0752
.0139
NDC
NDC
0
0
0
.162
173
.011
0
0
.0026
.0023
Run C: Coal-fired boiler - Required 40-minute N, purge to reach 0.08 ppm S02
Before
After
48.2 53.0 46.5
Difference
17.6
25.7 18.5
0.9
0.0615
0.0684
0.0079
NDC
NDC
0.1524 0.0019
0.1642 0.0026
0.0118
Run E: Coal/oil-fired boiler - Required 60-minute N_ purge to reach 0.1 ppm SO-
Before
After
20.6 37-6 30.9
Difference
20.
26.5 19-
-0.
1
4
7
0
0
0
.0754
.0774
.002
NDC
NDC
0.
0.
0402
0402
0.0
ND
ND
Run F: Coal/oil-fired boiler - Required 40-minute N2 purge to reach 0.08 ppm S02
Before
After
29.8 48.4 42.0
Difference
16
25-4 17
1
.8
9
.1
0
0
0
.0665
.0760
.002
NDC
NDC
0
0
0
.1237
.1287
.005
ND
ND
Run H: RDF incinerator - Required 40-minute N2 purge to reach 0.1 ppm SO
Before
After
88.9 101 75.8
Difference
66
98.9 71
4
.4
3
.9
0.237
0.270
0.033
ND
ND
0.134
0.137
0.003
ND
ND
(Continued)
13
-------
TABLE 3.3. (Continued)
Residue Weights (mg)
Un- NH4OH NH4OHb
treated Treated Corrected
1C Results (mg)a
as NH3
salt S043
Impinger Concentrations
(rag/ml)
S(\" S03S Cl N03'
Run I: Coal-fired boiler - Required 45-minute N, purge to reach 0.16 ppm SO,
Before
After
137 138 112
Difference
75
103 74
-1
.9
.0
.9
0
0
0
.306
.319
.013
ND
ND
0
0
0
.500
.^90
.010
0
0
.0050
.0050
Run J: Coal-fired boiler sample (pH 2) - No purge conducted
Before
After
not analyzed
Difference
93
131 9^
0
.9
.4
.5
0.
0.
0.
301
319
018
ND
CPe
0
0
0
.486
.577
.091
0
0
.0056
.0062
Run K: Wood-fired boiler sample (pH 4) - No purge conducted
Before
After
not analyzed
Difference
1
4-3 3
1
.4
.1
.7
0
0
0
.0065
.0151
.0086
ND
0.0264
IPf
IP
0
0
.0027
.0026
aS04s was determined from 1C results and S04= as the NH, salt was calculated from the
SO
1C value multiplied by the empirical correction factor, 1.375-
"The NH^OH corrected residue weight is the uncorrected weight minus the NH, retained
in the residue based on empirical calculations from the 1C S0h' results.
cThe 1C conditions for the analysis of S03=, as formaldehyde bisulfite, were not
capable of resolving the formaldehyde bisulfite from the large amount of Cl" present
in the samples.
dSamples preserved with propionaldehyde, with the propionaldedyde bisulfite being
resolved from the Cl" and NO," .
aCompound present, but large Cl" peak prevented quantification.,
fInterfering peak present. Suspected native formaldehyde bisulfite present.
14
-------
Sample runs B, C, E, and F showed that most of the SO, was removed from
the samples. Run H showed an increase of 4.9 mg of S02, which is about
6 percent of the sample weight. The RDF incinerator sample was the only
sample that had a significant increase and there is no explanation.
For all combustion source samples, the SO, in the impinger sample was
removed within 60 minutes by purging with N2 at a rate of 20 liters/min.
Based on this, Method 202 specifies a 60-minute sample purge time.
The residue weights represent both samples that were untreated and
treated with NH4OH. The weights for the untreated samples was expected to be
greater than the corrected values for the treated samples because of the
moisture collected by the H2S04. Later, it was discovered that a significant
amount of Cl" was being retained in the treated samples (by the addition of
NH4OH), whereas the HC1 evaporated from the untreated samples. Therefore, no
direct comparisons between the untreated and treated weights could be made.
The 1C analysis showed that the CT concentration was equal to or greater than
the $04* concentration in the majority of cases. The HC1 that reacted with
NH4OH remained in the sample residue after the sample was taken to dryness and
was counted as part of the condensible emissions.
Since HC1 emissions are not considered condensibles, an experiment was
designed to determine if the HC1 could be removed from the sample by initially
taking the aqueous inorganic fraction to dryness, resuspending this residue
with deionized (DI) water, and then adding the NH4OH to stabilize the S04=.
The results of this experiment are shown in Table 3.4. A water blank and the
aqueous inorganic fraction of five impinger samples were used. A 5-ml aliquot
was taken from each aqueous sample for 1C analysis of S04= and Cl". Each
sample was then taken to dryness and the residue resuspended in 100 ml of DI
water. At this point, another 5-ml aliquot was taken for 1C analysis. The
sample was again taken to dryness and resuspended in 95 ml of DI water.
For coal-fired boilers (Runs C, E, and F) and the wood waste boiler
(Run B), the initial heating and drying appeared to effectively remove the
Cl~. For the wood boiler, very little of the Cl" were removed. Ammonia was
suspected to be present in this sample.
The results in Table 3.4 also showed a decrease in S04=. This was
attributed to incomplete resuspension of the residue. When the samples were
taken to dryness, a black film formed on the bottom of the beaker. Much of
this material could not be resuspended in the DI water and, hence, the anion
analysis was considered to yield low results.
Because of the uncertainty introduced by not being able to redissolve the
dried residue, it was decided to repeat the experiment. For the second
experiment, aliquots from the air purge experiment were used; one fraction of
the aqueous sample was treated with NH4OH and the other fraction was not
treated. The samples were then taken to dryness, resuspended in water, and
analyzed for CT and S04=. The results for the four samples analyzed in this
experiment are presented in Table 3.5.
15
-------
TABLE 3.4. EVALUATION OF Cl~ AND S0ft' LOSS ON EVAPORATION OF AQUEOUS MATRIX
Run
cr
Catch
Sample Description (mg)
S04
Catch
(mg)
DI H20 Control Sample
A
Before drying*
After initial drying1*
After final drying0
0.0
0.0
0.0
0.0
0.0
0.1
Method 5 Impinger Sample from Wood Waste Boiler
B
Before drying
After initial drying4
After final drying*
0.9
0.1
0.1
3.0
0.8
0.9
Combined Method 5 Impinger Sample from Wood Jind Coal Boiler
C
Before initial drying
After initial drying4
After final drying4
11.5
0.0
0.0
32.9
26.9
26.4
Method 5 Impinger Sample from Wood Boiler
D
Before initial drying
After initial drying
After final drying
14.1
11.3
10.4
2.8
2.9
4.0
Method 5 Impinger Sample from Coal Boiler
E
Before initial drying
After initial drying
After final drying
84.3
0.1
O.l
82.2
57-3
56.1
Method 5 Impinger Sample from Coal Boi er
F
Before initial drying
After initial drying
After final drying
75-2
0.7
1.2
228.4
208.7
125.7
aAliquot of 200 ml sample taken before drying without NH4OH.
bSample redissolved and aliquot taken for analysis. Remaining solution
taken to dryness with NH^OH.
cNHi( OH-treated residue redissolved and aliquot taken for analysis.
dDid not completely redissolve.
16
-------
TABLE 3.5. EVALUATION OF Cl' AND S04" LOSS ON EVAPORATION OF AQUEOUS MATRIX
Run
A
B
C
D
Sample Description
Sample
After S02 &
Untreated
Treated
Sample
After S02 &
Untreated
Treated
Sample
After S02 &
Untreated
Treated
Sample
After S02 &
Untreated
Treated
"0" from Coal Boiler
Air Purging
"K" from Wood Boiler
Air Purging
"L" from Coal Boiler
Air Purging
"N" from Coal Boiler
Air Purging
ci-
Catch
(mg)
14.5
0.1
4.8
10.4
2.5
6.1
71.4
1.3
8.0
61.0
0.0
44.4
so,,*
Catch
(nig)
191.1
81.3
78.6
39-2
51.7
49.2
70.2
60.5
57-9
147.0
164.5
149-3
17
-------
The comparison of the treated and untreated samples shows that heating
and drying effectively removed the Cl" from the coal-fired boiler samples
(Runs A, C, and D). The treated samples contained more Cl~ in the resuspended
residue solution than in the untreated samples. The untreated wood-fired
boiler sample (Run B) showed a significant reduction in Cl" after drying, but
the procedure still did not totally eliminate the Cl", which indicated that
the Cl" were not HC1, but perhaps NH.C1. The results also indicated that some
of the NH4C1 are lost during the drying process. The SO/ results for the
treated coal-fired boiler samples were similar to those of the previous
experiment in that all of the SO/ could not be recovered. The wood-fired
boiler and untreated sample of Run D showed unexplainable increases in S04=.
It was concluded that the HC1 in the sample could be effectively removed
unless the emissions also contained a significant amount of ammonia which
could react with the sample during collection. The modified sample
preparation procedure which involves initially taking the sample to dryness,
resuspending the residue with DI water, and then adding the NH4OH, was added
to the candidate protocol.
Since a posttest purge with N2 lasting 60 minutes at a rate of
20 liters/min would require the tester to bring a full-sized N2 cylinder to
the field, an air purge was evaluated. As shown in Table 3.6, eight samples
and a DI water blank were prepared and analyzed in the same manner as
described previously for the N2 purge samples. After the samples were
saturated with S02, they were purged with air. The results were again very
much pH dependent. At pHs greater than approximately 4, S02 was not
effectively removed. At low pHs, the data showed considerable scatter, much
more than the N2 purge. Because of the scatter in the data and because air
purging could introduce the potential for oxidation of sample constituents,
the N2 purge was adopted as the procedure of choice. The air purge is offered
as an option in the candidate protocol, with a warning that it has been shown
to be less effective.
Just before the field evaluation of the candidate protocol, determining
the amount of NH4OH added to the sample by titration was tried. The data are
not presented, but indicated that the titration technique was possible.
Therefore, the technique was evaluated with field samples.
To determine the magnitude of the N2 purge under field conditions, staff
from EMB conducted paired-train testing at a coal-fired boiler burning coal
with a sulfur content of about 1.5 percent. At the conclusion of each test
run, one train was purged with N2 for 60 minutes at a rate of 20 liters/min
and the other train was not purged. Aliquots were taken for S04= analysis
before evaporation and drying. The results of the paired-train testing are
presented in Table 3.7.
The total CPM measured by the purged train averaged 4.7 mg, while that
measured by the unpurged trains averaged 51.4 mg. For this test series, the
S04= constituted approximately three-fourths of the CPM emissions. The
residue was resuspended in water and analyzed for S04= and Cl". This initial
field evaluation, conducted under the laboratory phase, clearly demonstrated
18
-------
TABLE 3.6. AIR PURGE EXPERIMENTS TO DETERMINE REMOVAL OF S03*
Sample
' ID
0
K
L
N
Q
Q
R
S
Process
Type
Coal boiler
Wood boiler
Coal boiler
Coal boiler
Oil boiler
Oil boiler
Oil boiler
Oil boiler
DI water
Initial
Sulfate
(mg)
202.9
1.9
61.2
176.2
8.1
8.2
144.4
51-2
0.0
Post Purge
Sulfate
(mg)
191.1
39.2
70.2
147.0
22.5
8.8
128.0
41.5
53.8
Difference
(nig)
-11.8*
37. 3b
8.0a
-29. 2a
14. 4a
0.6a
-16. 4a
- 9-7a
53- 8b
aSample had a low pH prior to the addition of S02.
bSample had a neutral pH prior to the addition of S02.
Conclusions: For 'all samples that had a low pH (with the exception of
the first run of sample Q), the S02 was removed within the measurement
error. For the two samples that had a neutral pH, the S02 was not
removed quickly at the start of the purge; this apparently resulted in
the S03= being converted to sulfate.
19
-------
H
§
CC
M
t
O
<
CO
T-H
I
cc
Cd
CQ
O
a
Cd
a
o
o
Cd
b.
O
CC
en
CQ
0>
.w
i-l O
J-> tJ
o u
S
0)
rH 4J
4J O
O ti
«EH ti
bo O
S O
03 r-l
3 O
H SC
0) 2
0)
CC
i
i 1
.3-
JT
O
CO
c
-I CB -
CO T3 faO
W C S
iSa^
Organics
(mg)
Inorganic
Corrected
(ing)
Inorganic
Uncorrected
(mg)
t3
Q)
faO
3
CU
0)
M
a a
S M
CO
CO
^HCO
V£>
^-O^
J3-
in <-4
0 O
0 O
roco
.3- in
CM in
o ^r
in
o o
c\j in
6-31
in
oo r
w
O 0)
2 ><
< CQ
^w
o in
iH ro
r t
0 0
in in
0 0
O CN
o in
**O CO
co ro
-3-
vo in
00 JT
0 0
^o in
CO -3-
O cy\
in in
vo
w
O QJ
2 >
< CQ
CM CM
O ON
CO t-
in r-
ON in
T-H CO
ro o
CO CM
CM O
-3- r~
CM m
t in
-3-
-3" O
in in
in
0 O
-3- O
in m
in
rH rH
ro r
CO
o a;
2 >
< CQ
CO CO
.3- t-
VD
^-H tO
^o in
j-
(M O
<
CO ro
CM CM
rH rH
bO hi)
20
-------
the effectiveness of the posttest N2 purge in removing S02 from the impinger
solution.
21
-------
4.0 FIELD EVALUATION TEST 1
The first field evaluation test program was conducted at a coal-fired
boiler burning 2 percent sulfur coal. Based on an EPA test report of a
similar facility, the emissions were expected to have sufficient quantities of
inorganic CPM.
The primary purpose of the first field test was to evaluate the revised
method at a source with high inorganic CPM. Specific objectives were to:
(1) determine the precision for the combined sampling and analysis procedures,
(2) evaluate the effect of the posttest N2 purge in the field, (3) evaluate
the effect of heating the aqueous (inorganic) fraction in removing CT, and
(4) compare the S04* content of the inorganic fraction as determined by 1C
versus NH4OH titration.
The sampling and analytical procedures used are detailed in the draft
method in Appendix A. Five quadruplicate train (quad-train) runs were
conducted. Sampling time was 2 hours, followed by a N2 purge at 20 liters/min
for 1 hour. In two of the runs, two of the trains (B and C) were purged to
measure the effectiveness of the purge method. In the other three runs, all
four trains (A, B, C, and D) were purged to assess the precision of the purge
method. The samples were prepared and analyzed as shown in Figure 4.1. The
results are presented in Tables 4.1 and 4.2.
Excluding results for the unpurged samples because the inorganic fraction
weights (SO/) were shown to be significantly different from those for the
purged samples, the standard deviation for the sampling and analytical
protocol averaged 1.1 mg/m3 for an average stack concentration of CPM of
3.5 mg/m3.
The S04= concentration (see Runs 5 and 6 in Table 4.2) in the purged
samples was compared to that in the corresponding unpurged quad-train samples
using Student's "t" test and the null hypothesis (see Appendix B). The
S04= concentrations in the unpurged samples were significantly higher than
that in the purged samples at the 95 percent confidence level. The S04=
concentration results (Table 4.2) were higher than theoretically possible when
compared to the inorganic CPM weight (Table 4.1); therefore, the t-test was
considered invalid.
The 1C analyses (Table 4.2) of sample aliquots taken initially, after the
initial drying (pretitration), and after titration using NH4OH (final drying)
show that the CT concentration was reduced from a level of approximately
100 mg/m3 to a level of about 0.02 mg/m3 by drying the sample at 105°C before
adding NH4OH, which supported the previous laboratory results (see Table 3.4).
The results of the titration techniques for ammonia determination did not
agree as well as they had in the laboratory evaluation. The average ammonia
values for the coal-fired boiler samples were 0.95 mg/m3 by 1C, 0.67 mg/m by
titration using an endpoint indicator, and 0.41 mg/m by titration using a pH
meter. The results of the 1C analyses for S04= were determined to be in error
and were not used to make the NH4" correction.
22
-------
MS Impinger Sample
Approximately 350 ml
Dilute to 500 ml*
w/DI H2O
500 ml Sample*
Remove
S-ml Aliquot for
1C Analysis of
495 ml Sample
Extract 3X w/ 75-ml each MeCI2 (use MeCI2 rinse for 1st extraction)
_L
Fraction
w/ Organics
_L
Aqueous Fraction
w/ Inorganics
Evaporate at Room Temp
& Atm P to Dryness
Desiccate
Weigh to
Constant Weight
Residue
(Weight Contribution
of Organics)
Evaporate on Hot Plate
Followed by 105°C Oven
Redissolve Residue in
100 ml of Dl
100ml
Aqueous Fraction
w/ Inorganics
5-ml Aliquot
Analyze by 1C
Remove
Remove
10-ml Aliquot
Titrate to pH 7
W/0.1N NH4OH
using a pH meter
Aqueous Fraction
w/ Inorganics
Add 5 Drops of
Bromothymol Blue
(pH 6.0-7.6)
Titrate w/
0.1N NH4OH
Combine
Evaporate
Desiccate
Weigh to
Constant Weight
Residue
(Weight Contribution
of Inorganics)
Redissolve in 95 ml 01 H 2 O
95 ml Sample
I
Remove
Archive
'Note: The efficiency sample diluted to 250 ml.
1
5-ml Aliquot for
1C Analysis of
SO4 = .C1-.NO3~
3613 3/90
Figure 4.1. Scheme for preparation and analysis of impinger samples.
23
-------
cc
Cd
M
O
CQ
O
Cd
CC
O
i
Cd
CQ
M
co
Q
O
O
O
C/5
a
<
o
M
-H
M
a
a
o
tx,
O
O
M
ss
cc
Cd
td
Cd
J
m
M
0
JJ
JJ
03
35
CD
i 1
jQ
H
CQ
CD
'O
c
o
O
CM
0
c
o
*^H
jj
03
C
H
a
5
CD
JJ
CD
«
O
'2
4_)
CD
a
H
CO
M
O
03
H
C
O
a
a
M
0
c
o
JJ
o
73
^£
fl
o o *jj.
+ s
^^
-""^
i m
CD JJ a
Ll -H --»
0, H hO
a
ed M .
ho CD hO
M -o a
0 C ^
o
o
CD
O iH
H J3 *-»
C -H fn
CO CO S
hO C -~.
Li 0) hO
o -o a
c c ^
M O
0
c
o
4_) r*-) £j
cd p"i ****1^.
Li 2 hO
JJ E
H
£_<
Ll
0) -^
JJ l~>
CD ma
S S3 --.
2 hO
x a
a
ZT
(*-i Q
C5 ZC ^*-
M 2 ho
a
0 CD
rH 8
asm
S o **-"*
en >
CD
hD
tj
^3
a,
0
a a
S HH
CO
C/3
CO
-a
H CD
Li -O
O -H
rH X
/* rt
£* O
U Li
T3
CD >,
C X
CD
Tf B
>> 3
£ -H
1 ^ d^
** c
CD O
a B
.c 3
S J3
2 jj
H
-a s
*^ JJ C
CO O
C -H
H JJ
a a)
JJ JJ
C -H
O JJ
CJ
o
co jj
rt)
i-H Ll
a o
S *H
CQ tj
cn a
C^- O VD r4
rf*i (*o co \o
CO JT O O
*-H
Lf*N f^~ ^«O ^5
^5 O^ ro ^r
m m O O
^4
co m r^ cn
oo CN C*- c**
*
o o o o
co I-H t~- m
co co co m
O CTN iH ^O
CO CM CO CM
CO CQ CQ CO
CD CD 0 CD
< CQ O Q
1 1 1 1
CO CO CO CO
CO
CM JT invjo
CM in o> r*-
i-l CM t CO
o o o o
m co o o\
e
CM CO-3- in
CO CO O t
{f\ c*^ vo r**^
o o o o
1-1 co moo
co m-a- \jo
o o o o
CM -3- CM JJ-
^^ ^^N ^y^ c^i
*
O rH O rH
O f- t-t iH
CM COCO PO
1-1 o ^ c~~
CO PO POCM
CQ CO CO CQ
0) 0 0 CD
< CQ O Q
1 1 1 1
-=r ^3- -3- -a-
mro CM CM-
CD co
t~-JJ- VJO CO
eo o mco
^^ c^ ^~o *r
e
iH O O O
in CTN CM vo
o in invo
a
->£> CO in CM
CM oo in <-i
i-H V£) VO CO
^H O O O
.3- O v£J CO
CO -3- CO IH
O O O O
CM ^o ON-a-
vj2 co m ^~
co o o CM
o in CM co
JT VO iH CO
rH o CM r~
COCO CO CM
co co
O 0 0 O
2 >- > 2
< CQ O Q
1 1 1 1
in in in in
O> CM O^ QN m
co PO CM in in
co o CM co 1-1
JT iH *-* r-l CM
JS-
in co CM CM
cy\.3- in r
CM IH co -a-
* *
r-t C\J O O
^ ^^ VO ON
CO r-t *sD C*1-
*
.^ ^_{ ^_l ^.^
C*^ LTN ^O U*N
^r CM CM I-H
o o o C)
CO CO i 1 C"
CM O i-H O
*
O O O CD
r*-vo in m
JT C--CO O
CM O O eg
1-1 \£> in u^
^r CM oo co
oo TH o m
COJT -a- PO
co co
O 0 0 O
2 > > 2
< CQ 0 C3
III
vjO vO VO ViD
\o 1-1 co m IH
JJ- CO CM CM CO
co co T-I moo
^H 0 i-H O 0
IH jj- in coco
CO T-I «H i-l iH
CD
9
(^
0
CM in in j- >
CO iH iH i-l
CO CO O i-H
JT CM VO CO
* *
i-i^OO
CM PO invo
** o o o
1-^ .^f ^o ^J*
*H O O O
*
0 O O O
t-~^r CTN r--
o o o o
CD CD CD CD
1-1 in CM t-~
oo r \o r~-
o o o o
.a- r~- TH co
\,Q VO VO ^^
^3" vO ^O ^O
.3- JT -3- -3-
CO CQ CO CQ
0000
< CQ O Q
1 1 1 1
C*1 C^" C^ C^*
o
T3
T3
CO
CO
H
O
a
£-1
O
H
Ll
a
CD
CO
o
£,
H
24
-------
§
CL,
I
ex
H
M
z
o
Cd
a
o
CO
ex
S
1
OS
o
rH
O
ex
HUi
o
o
en
1 Drying
solution
W W
S £
H "o
fe£
b
0 -O
CM 3
<
faC
-S c
£.2
0 4J
f-H rH
H tn
U CD
H -H
C T)
M
g a
t4 "»»
41 hO
55^
m
5-5.
rH M
3 S
CO '
0> ^
-w . >, >H >H
< 03 U Q
1111
on oo ro ro
(N rH O O
O O O O
O O O O
O O O O
o o o o
m rH r\j o
O \O .a- CM
rH rH rH CM
CM rH O O
o o o o
o o o o
o o o o
o o o o
-=r oo co -=r
O C ro CM
rH rH rH CM
CO CM oo ro
o\ CM co CM
0 rH O O
LT\OO COCO
rH O ^O
CT\ rH J3- C~~
rH LPl CM CM
en en w to
(1) CD « >1 >H >.
< 03 0 Q
1 1 1 1
-3- -3- -3- ^3-
(\J r-l O O
0000
o o o o
o o o o
o o o o
o C-CM in
rH rH O C^-
fO rH rH O
CM rH o o
o o o o
O 0 0 0
o o o o
o o o o
V£> CO CO CO
CM CM o c
fO rH rH O
^3 rH r~- O
-a- rH r~ o
rH rH O O
\£3 VD CO CT\
o> CTN co r~
.=r r co jo-
as CTvco r~-
V£) O VO rH
^£) CO IT\ CO
ON CM rH f-
cn en
o CD CD o
2 > >- 2
< m o Q
1 1 1 1
in in ir\ in
fM rH O O
O O O O
O 0 0 0
0000
0 O 0 0
TH CM r- m
O CO -3" CO
rH O O O
CM rH O O
O O O O
o o o o
o o o o
o o o o
rH rH ^£> LTN
O ro.3- ro
rH O O O
VD CM CTs CM
-3- CM VO -a-
CM rH O O
\D ^f ^O CM
r- r~--sr CM
CM CM CM CM
rH rH rH rH
CO CM VO VO
m o CM -a-
vo CM CM m
tn tn
O CD 0) O
z > >H 2:
< CQ O Q
I I I I
UD \£> ^D ^O
O O O O
0 0 0 O
o o o o
o o o o
O CO CM ON
CO rH CO CM
0 0 0 O
rH rH O O
0 0 0 O
O O O O
o o o o
o o o o
rH CO COVD
ro rH oo CM
O O O O
vo o- oo r
JT rH Vfi -3-
OvJ rH O O
O O ro ro
M3 t-~ C C~-
rH rH CTN rH
rH rH rH
r o LO-a-
rH O VO O
CM CM rH CM
tn en en en
0 CD CD CD
> >< > >
< CQ 0 Q
1 1 1 1
r- r r r~
25
-------
To determine If the N2 purge was removing any of the organic matter, the
organic condensible results were compared for the purged and unpurged sample
pairs from quad-train Runs 5 and 6. The results shown in Table 4.1 indicate
that the organic values for the purged and unpurged samples were equivalent at
the 95% confidence level. Therefore, the N2 purge did not affect the
concentrations of the organic condensibles measured.
CONCLUSIONS
The results of the first field test (see Tables 4.1 and 4.2) indicated
that a coal-fired boiler with significantly greater emissions was needed to
adequately determine the precision of the combined sampling and analysis
protocol. The posttest N2 purge and the pretitration drying of aqueous
samples proved to be effective. Without removal of the dissolved S02 by
purging and the HC1 by heating, the CPM measured for the average sample would
have been about 120 mg/m3 instead of the actual value of about 4 mg/m3. Since
the total CPM results were much lower than had originally been anticipated,
comparison of the 1C results with the gravimetric determinations was not as
close as had been expected. An additional evaluation at an emission source
with high S04° levels was recommended to provide a more accurate comparison of
the measured values.
26
-------
5.0 FIELD EVALUATION TEST 2
The second field evaluation was conducted at a wood-fired boiler. The
sampling location was after an electrified fabric filter. Based on previous
tests at a similar facility, the emissions were expected to have sufficient
quantities of organic CPM.
The primary purpose of this field test was to evaluate the revised method
at a site high in organic CPM emissions. Specific objectives were to:
(1) determine the precision, (2) evaluate the effectiveness of the posttest
N2 purge, and (3) evaluate the effectiveness of heating the aqueous
(inorganic) fraction in removing Cl".
The sampling and analytical procedures used are detailed in the draft
method reproduced in Appendix A. Five two-hour quad-train runs were
performed, followed by a one-hour N2 purge of the impinger trains at
20 liters/min. In two of the runs, two or three of the trains were purged.
In the other three runs, all four trains were purged. The samples were
prepared and analyzed in the same manner as the samples from the first field
evaluation (see Figure 4.1), except only an initial anion analysis was made.
The results of the test are shown in Table 5.1. As seen, the S04a
concentration was near the limit of detection (<0.07 mg/m3), and thus, the N2
purge would not have been necessary. The results showed that N, purging did
not adversely affect the organic CPM determinations; the mass of CPM was the
same for the purged and unpurged samples.
Since the organic CPM was shown to be the same for the purged and
unpurged samples, all samples were used to calculate the precision of the
method, which turned out to be 13.0 ± 2.1 mg/m3.
The results of the 1C and titration techniques for ammonia determination
did not agree as well as they did in the laboratory evaluation. The average
calculated ammonia values (see Table 5.2) were 0.03 mg/m3 by 1C and 0.32 mg/m3
by titration. However, this difference is not significant when compared to
typical particulate concentration levels of 30 to 50 mg/m3 for a compliance
determination.
CONCLUSIONS
The results of the second field evaluation at a wood-fired boiler led to
several conclusions and recommendations. A N? purge conducted at a rate "of
20 liters/min for 60 minutes at the end of a test run does not significantly
affect the determination of organic or inorganic CPM when SO, emission levels
are near detection limit or are very low. Therefore, it would be acceptable
to omit the purge at sources not expected to have significant concentrations
of S02 (an incorrect judgement may cause a positive bias). If the sample pH
is greater than 4.5, indicating low concentrations of H?S04 and S02 in the
sample, the ammonia addition is no longer required to stabilize the SO/.
27
-------
os
Ed
§
Q
Ed
OS
M
tt,
I
o
cc
Ex.
I
m
*i
cn
a
o
o
cn
cn
o
z
2
a
a
<
h-H
2
O
S
d
2
O
i<
2
14
cc
Li
Cd
T^
m
Cd
en
Ll
0)
4J
0)
0)
CO
c
CD
o
§
O
o
c
o
H
4J
CO
c
H
s
5
CD
0)
a
u
H
Li
J__^
CD
E
H
co
tj
C3
H
r* r
o
s f
a
H
M '
O
S p
« r
^C *
" *
?H fl
co LI a
JJ M ^^.
O O bO
E-I O B
.
Li -^
r-l Ll m
CO O E
0 C hO
E~< 3 S
CD
1 1
O JS -^
H -H ro
CUE
CO C "^
bO CU bQ
Li 73 S
O C *
O
O
CD
O i-H >
H .O. f>
C -H E
CO W --~
bO C hO
Li CD E
C C
M O
O
c ~
C O Li
H -rl CD - >
JJ «J (**
O CO CD S
CD Li S ~^>
C JJ M
0 :H = =
c 5
j " <>
c co o e
CD Li I 1 ^
a j-> - ho
a -H E
=C r"
CD CD
i 1 S -
D. ~ n
E r-t S
CO 0 -
cn >
hO
Li
3
CL,
CD
a Q
C HH
5
cn
ja- in^a- CM NO
in ON c^- (N co
a- in.3- CN NO
in ON t-- CN co
05
bo
co
Li
CD
^
<£
LTs *H ^ t*--
CO CO ^ O
0
f\j v^ jy <7*\
CNJ fNJ ^H LTs
ufS r*"" ^^ i-T^
Csj (\J rO fNJ
CNJ .s" co ro
C\J rH rH fNJ
o o o o
fNJ fNJ (M fNJ
O O O O
d d d d
«-i NO CO CO
NO CO rH r~-
NO in m ro
e . .
,H ^ ^ ~4
cn co
CD CD O O
>+ >* z z
< m o Q
i i t i
00 CO CO CO
in in o o in
mo t «-t co
in in o o in
in o r- ^ co
^ ^ ^ 7
hO
co
Li
CD
^
- O C^
» .
CO ro fNJ ro
*-H <-1 ^-H
fNJ r-~ fNJ .3-
^r CO ON -H
C^* ^^ *S" C*^*
^ i-^ ro .^r
^H m fNj r\i
. *
o o o o
ro fNJ fNj CM
O O O O
o o o o
^r in r-=o
1 (M t~- fNI
....
r_| rH < 1 rH
cn cn cn cn
CD CD CD CD
^H ^H ^« ^N
< CQ O Q
1 1 1 1
ON O"N CJN fJN
in ON o NO co
CO t~- ON r CO
in ON o vo ro
CO r ON r CO
0)
hO
co
Li
CD
^
^c
ro CN in o
t ON fNJ ro
in.3- NO .3-
CO CO t -31
V
CNJ fNJ fNJ PO
ro fNJ fNJ «H
i 1 (N 1 r-l
. 0
O O 0 O
-3- fNJ fNJ (M
O O O 0
d d d d
O f~ LfN fNI
NO C^ IX^ fNJ
\o vO in vo
....
TH rH rH ^H
cn
O CD O O
z. >* -Z. -z.
< m o a
i i t i
0000
, 1 r-l i 1 T 1
.3- in»H TH co
CNJ ^r CM in co
^r m «-i '-i ro
CM ^3- fNj in co
CD
hO
CO
Li
Q)
^
<
m O^^£> fNJ
C O fNI vo
vo tr\ vC) tH
^H
co fNJ f- ro
vo rTj* co >^r
Lf*\ O*N LfN CO
:*
t** -^ vi^ r*""
T-H ON u"N ^~t
. *
0 0 0 0
rO-=r oo ro
0 O O O
d d d d
co^r c~vj CJN
LTN fO LH
roco ro ro
....
t-H O rH r-l
cn cn 1/1 cn
CD CD jB CD
< m cj a
i i i i
^H vH ,-t ^H
O JS- CM fNJ t
O CO ON O\NO
O 3" fM fNJ t-~
O CO ON ONV£>
0)
hO
CO
L,
CD
^
<
in o o o
vo in in o
...
JT ON f\l ro
T-l 1
-3- NO CM fNI
roco c fNj
....
in co NO NO
NO ro CTvO
-=r r NO CM
....
O O O 0
un-=r ^- m
o o o o
....
o o o o
NC in c -r-
co ro LTNNO
NO CJNCO CO
....
O 0 O O
CO 'Jl CO CO
CD CD CD CD
£M ^JH ^ >H
< CQ U Q
1 1 1 1
CM (M CNJ fNJ
r 1 1 T 1 t 1
LI
o
C3
73
O
CD
cn
ro
o
H
JJ
O
CD
Li
Li
O
O
*
-------
TABLE 5.2. WOOD-FIRED BOILER ION CHROMATOGRAPHIC RESULTS
FOR SULFATE AND CHLORIDE
Sample
ID
8-A
8-B
8-C
8-D
9-A
9-B
9-C
9-D
10-A
10-B
10-C
10-D
11-A
11-B
11-C
11-D
12-A
12-B
12-C
12-D
Purged
Yes
Yes
No
No
Yes
Yes
Yes
Yes
No
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Initial
Sulfate
(mg/m3 )
0.06
0.06
0.06
0.06
0.09
0.06
0.06
0.06
0.10
0.06
0.06
0.06
0.0?
0.11
0.07
0.07
0.13
0.10
0.11
0.14
Chloride
(mg/m3 )
1.11
1.19
1.44
1.44
0.06
0.06
0.76
0.06
1.21
1.09
1.23
0.06
1.29
1.92
0.07
0.07
0.13
2.62
2.32
2.12
29
-------
In conclusion, the evaluation of the draft sampling and analytical
protocol at a wood-fired boiler demonstrated adequate precision. The
modifications for testing at sources with low SO, emissions involving
elimination of (1) the posttest purge and (2) addition of the NH4OH when the
sample pH is greater than 4.5 were added to the draft protocol.
30
-------
6.0 FIELD EVALUATION TEST 3
Because the first coal-fired boiler tested did not have sufficient SO^
emissions, the revised test method was evaluated at a second coal-fired boiler
with S02 levels in excess of 2000 ppm.
The objectives of the second coal-fired boiler field test were to
(1) determine the precision of the draft method, (2) evaluate the
effectiveness of the initial sample drying step to remove CT, (3) compare the
two procedures (1C analysis and titrimetric) of determining the amount of
ammonia added to the sample, (4) determine the effect of the posttest purge,
and (5) determine the collection efficiency of the impinger for CPM.
The sampling and analytical procedures followed those described in
Appendix A and depicted in Figure 4.1, with the following modifications:
o Initial dilution of the impinger sample plus water rinse was to
800 ml, due to the high moisture content of the sample gas.
o The first MeCl2 extraction was performed using the MeCl2 rinse
catch.
o The two remaining MeCl2 extractions were performed with 50 ml each
of fresh MeCl2.
o After redissolving the residue in 100 ml of DI H20, only the 5-ml
aliquot for 1C analysis was removed. The NH4OH/bromothymol blue .
titration was rejected in favor of a single NH4OH titration of the
remaining 95-ml sample to pH 7.0 using a pH meter.
o After evaporating and desiccating of the titrated inorganic samples,
final weights were recorded and the samples were archived for
further analysis.
The quad-train run scheme was almost identical to that for the two
previous field tests. Run times were two hours, followed by a 60-minute N2
purge of the impinger trains at 20 liters/min. Train A in each quad-train run
was fitted with an extra impinger before the silica gel impinger. The catches
from each of these impingers were kept separate for the impinger efficiency
evaluations. In two of the quad-train runs, only two of the trains were
purged with N2. For the remaining runs, all four trains were purged with N2.
Tables 6.1 and 6.2 present the results of the third field evaluation.
The average concentrations for the gravimetric analysis of the total CPM
ranged from 25 mg/m3 to 48 mg/m3 with standard deviations ranging from
±7.3 to ± 9.9 mg/m3. The effect of the N, purge on the S04~ concentrations was
evident. In the initial 1C aliquots, before heating, the purged train S04=
concentrations ranged from 1.7 mg/m3 to 8.10 mg/m3. The S04= concentrations
for trains not purged ranged from 16.3 mg/m3 to 25.6 mg/m3. Thus, the S03=
concentrations in the unpurged samples were 280% and 178% higher than the
average S04= levels in the purged samples. These results agree with those
from the laboratory evaluations and the two previous field evaluation tests
31
-------
TABLE 6.1. DETERMINATION OF AMMONIA ADDITION AND MASS OF CONDENSIBLE
PARTICULATE MATTER FROM COAL-FIRED BOILER
Sample Purged Sample
ID Volume
(m3)
la Yes
Extra impinger
Ib Yes 2.335
Ic No 2.299
Id No 2.307
Average
2a Yes 2.453
2b Yes 2 . 480
2c No 2 . 48?
2d No
Average
3a Yes 2.507
Extra impinger
3b Yes
3c Yes 2.538
3d Yes 2.542
Average
4a Yes 2.476
4b Yes 2.519
4c Yes 2.492
4d Yes 2.514
Average
5a Yes 2.491
Extra impinger
5b Yes 2.537
5c Yes 2.516
5d Yes
Average
Addition of Ammonia
1C ph Meter
NH3 NH3
(mg/m3 ) (mg/m3 )
(Sample blown out of
(0.37) (0.0)
3.04 4.20
7-31 4-37
9-60 9-37
6.65 6.0
1-73 2.73
1.84 3-60
6.11 4.12
(Sample taken to dry
3-2 3-5
2.18 4.54
(0.43) (0.0)
(Sand got in sample
1.50 2.97
0.64 1.94
1.44 3-2
0 . 64 1 . 89
0.79 2.12
1.20 2.30
1.99 2.93
1.2 2.3
2.06 8.35*
(1.19) (0.0)
2.48 2.45
0.83 1.79
(Large leak in
1.8 4.2(2.1)
Determination of Condensible Particulate
Inorganic Organic
Condensible Condensible
(mg/m3) (mg/m3)
Total Total
Uncorr. Corr.
(mg/m3) (mg/m3)
out of impinger during purge)
3-9 0.3 4.2 (4.2)
43.1 0.5 43-6 40.6
43.2 0.3 43-5 36.2
45-7 1-9 47-6 38.0
38.8 0.3
39-4 0.6
44.2 1.4
ness on hot plate during
39-1 37.4
40.0 38.2
45-6 39.5
dry down)
44.6 2.5
5-9 2.7
during dry down)
25-5 1-2
35-8 1.2
47.1 44.9
8.6 (8.6)
26.8 25.2
37-0 36.4
35-5 ± 9-9
30.1 0.3
46.2 0.4
49.0 5.0
47-5 0.5
30.4 29.8
46.6 45.8
54.0 52.8
48.0 46.0
43.6 i 9.8
49.4 0.4
3.5 o.o
39-5 0.3
33-3 1.1'
sample train)
49.8 47-7
3-5 (3-5)
39-8 37-3
34.4 33-6
39-5 ± 7-3
''Sample showed. chloride contamination (see Table 6.2).
32
-------
TABLE 6.2. ION CHROMATOGRAPHIC RESULTS FOR SULFATE, CHLORIDE, AND NITRATE
FOR FLUE GAS SAMPLES FROM A COAL-FIRED BOILER
Sample
ID
1-b
1-c
1-d
2-a
2-b
2-c
3-a
3-c
3-d
4-a
4-b
4-c
4-d
5-a
5-b
5-c
5-d
Train
Purged
Yes
No
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Initial
Sulfate
(mg/m3 )
8.10
19.50
25-60
4.60
4.90
16.30
5-80
4.00
1.70
1.70
2.10
3.20
5-30
5-50
6.60
2.20
2.80
Chloride
(mg/m3 )
13.00
11.40
7-90
8.50
8.70
8.60
8.70
8.70
8.90
10.40
8.50
6.40
6.40
6.40
7.40
7.00
5.10
Nitrate
(mg/m3 )
O.OO1
O.OO1
O.OO1
O.OO1
O.OO1
O.OO1
O.OO1
O.OO1
O.OO1
O.OO1
O.OO1
0.80
0.50
0.70
0.40
0.80
0.60
Pre-Titration
Sulfate
(mg/m3 )
11.88
13-77
19.53
5.18
6.41
9.41
13.03
9.19
5.47
4.05
3.60
4.27
5.83
4.94
5.13
2.87
4.75
Chloride
(mg/m3 )
0.00
0.20
0.30
0.00
0.20
0.00
0.00
0.30
0.03
0.30
0.30
0.30
0.00
7.102
0.01
0.00
0.10
Nitrate
(mg/m3 )
0.00
0.00
0.07
0.00
0.80
0.00
0.00
0.60
0.07
0.80
0.70
0.50
0.00
2.00
0.00
0.00
0.00
Collection Efficiency (4th) Impingers
1-a
5-a
Yes
Yes
6.00
1.60
0.20
2.10
0-30
0.30
0.99
1.34
0.10
0.80
0.00
0.30
1 Phosphate ion chromatographic interference.
2Sample contaminated.
33
-------
and demonstrate further that the N2 purge satisfactorily removes the dissolved
S02 collected in the impingers.
In comparing the 1C results for the initial Cl~ concentration to the
concentration after the sample was taken to dryness (pre-titration, see
Table 6.2), the effectiveness of initially taking the sample to dryness to
remove Cl" is clearly evident.
An extra impinger was placed before to the silica gel impinger in Train A
of each quad-train group to determine the collection efficiency. The extra
impinger collected about 10 percent of the total sample catch, which
demonstrated a need for an additional impinger efficiency.
The final comparison involved determining if the amount of ammonia added
to the sample could be determined by titration in lieu of 1C analysis (see
Appendix B for calculations). The average of the results for the two
procedures were within 1 mg/m3 for all but the last quad-train run where one
of the samples was contaminated with Cl" (see Table 6.2).
CONCLUSIONS
The procedure to remove the Cl" by taking the sample to dryness again
proved effective. The two techniques for determining the amount of ammonia
added to the sample were shown to be comparable. The EPA decided that the 1C
technique would be shown as the procedure of choice, with the tester permitted
to use the titration technique as an alternative procedure. The posttest
purge was again shown to be effective and is required by the draft method when
S02 is present in the stack emissions. The precision of the draft sampling
and analytical protocol was worse than had been anticipated. The impinger
train did not demonstrate sufficient collection efficiency; therefore, the
train configuration was modified to include two tipped impingers and the
sampling reagent (water) in the first three impingers.
34
-------
7.0 REFERENCES
1. Nothsteln, Greg. Masters Thesis. University of Washington, Department
of Environmental Health, Seattle, Washington.
2. Texas Air Control Board, Laboratory Division. "Determination of
Particulate in Stack Gases Containing Sulfuric Acid and/or Sulfur Dioxide."
From Laboratory Methods for Determination of Air Pollutants. Modified
December 3, 1976.
3. "Particulate Source Test Procedures Adopted by Puget Sound Air Pollution
Control Agency Board of Directors." Puget Sound Air Pollution Control
Agency, Engineering Division, Seattle, Washington, August 11, 1983.
4. Commonwealth of Pennsylvania, Department of Environmental Resources.
Chapter 139, Sampling and Testing (Title 25, Rules and Regulations, Part I,
Department of Environmental Resources, Subpart C, Protection of Natural
Resources, Article III, Air Resources). January 8, 1960.
5. Wisconsin Department of Natural Resources. Air Management Operations
Handbook. Revision 3. January 11, 1988.
6. "Determination of Particulate Emissions from Source Tests," DER Analytical
Procedures, Attachment No. 2.
7. "Determination of SOX as S02 from Source Tests," DER Analytical Procedures.
8. "Determination of S02 and S03 from Source Tests," DER Analytical
Procedures.
9. "Sampling Condensible Emissions from Stationary Sources," Department of
Environmental Quality, State of Oregon.
10. Federal Register. Vol. 53, No. 68, p. 11688, April 8, 1988.
11. "Estimation of the Importance of Condensed Particulate Matter to Ambient
Particulate Levels," report prepared for U. S. Environmental Protection
Agency, EPA Publication No. EPA 450/3-81-005a.
12. "Determination of Particulate After Removal of Sulfuric Acid and Waters
of Hydration," Entropy Environmentalists Inc., internal memo.
13. DeWees, W. G., "PM10 Emissions from Stationary Sources - How Do We Define,
Measure, Control, and Regulate Them?" Presented at the 78th Annual Meeting
of the Air Pollution Control Association, Detroit, Michigan, June 1985.
April 8, 1988.
35
-------
APPENDIX A
Method 202 - Determination of Condensible Particulate Emissions
from Stationary Sources
1. APPLICABILITY AND PRINCIPLE
1.1 Applicability. This method applies to the determination of
condensible participate matter (CPM) emissions from stationary sources. It is
intended to represent condensible matter as material that condenses after
passing through an in-stack filter (Note: The filter catch can be analyzed
according to Method 17 procedures). This method may be used in conjunction
with Method 201 or 201A if the probes are glass lined. This method may also
be modified to measure material that condenses at other temperatures by
specifying the filter temperature.
1.2 Principle. The CPM is collected in the impinger portion of a Method
17 (Appendix A, 40 CFR Part 60) type sampling train. The impinger contents
are immediately purged after the run with nitrogen (N2) to remove dissolved
sulfur dioxide (S02) gases from the impinger contents. The impinger solution
is then extracted with methylene chloride (MeCl?). The organic and aqueous
fractions are then taken to dryness and the residues weighed. The total of
both fractions represents the CPM.
2. PRECISION AND INTERFERENCE
2.1 Precision. The precisions based on method development tests at a
wood waste burner and two coal-fired boilers are 13.0 + 2.1 mg/m3,
3.5 + 1.1 mg/m3, and 39.5 + 9.0 mg/m3, respectively.
2.2 Interference. Ammonia (e.g., in sources that use ammonia injection
as a control technique) interferes by reacting with the hydrogen chloride
(HC1) in the gas stream to form ammonium chloride (NH4C1) which would be
measured as CPM. The sample may be analyzed for chloride and the equivalent
amount of NH4C1 can be subtracted from the CPM weight.
3. APPARATUS
3.1 Sampling Train. Same as in Method 17, Section 2.1, with the
following exceptions noted below (see Figure 202-1). Note: Mention of trade
names or specific products does not constitute endorsement by EPA.
3.1.1 The probe extension shall be glass-lined.
3.1.2 A Teflon filter support shall be used.
3.1.3 Both the first and second impingers shall be of the Greenburg-
Smith design with the standard tip.
36
-------
c
1-t
03
bO
a
en
en
a)
aJ
C3
«J
Q.
4J
iI
A
-r-l
01
-------
3.1.4 All sampling train glassware shall be cleaned prior to the test
with soap and tap water, water, and rinsed using tap water, water, acetone,
and finally, MeCl?. It is important to remove completely all si 11 cone grease
from areas that will be exposed to the MeCl2 during sample recovery.
3.2 Sample Recovery. Same as in Method 5, Section 2.2, with the
following additions:
3.2.1 N2 Purge Line. Inert tubing and fittings capable of delivering
0 to 28 liters/min of N, gas to the impinger train from a standard gas
cylinder (see Figure 202-2). Standard 0.95 cm (3/8-inch) plastic tubing and
compression fittings in conjunction with an adjustable pressure regulator and
needle valve may be used.
3.2.2 Rotameter. Capable of measuring gas flow at 20 liters/min.
3.3 Analysis. The following equipment is necessary in addition to that
listed in Method 5, Section 2.3:
3.3.1 Separatory Funnel. Glass, 1-liter.
3.3.2 Weighing Tins. 350-ml.
3.3.3 Drying Equipment. Hot plate and oven with temperature control.
3.3.4 Burette. 5-ml size with 0.01-ml graduations.
3.3.5 Pipets. 5-ml.
3.3.6 Ion Chromatograph. Same as in Method 5F, Section 2.1.6.
4. REAGENTS
Unless otherwise indicated, all reagents must conform to the
specifications established by the Committee on Analytical Reagents of the
American Chemical Society. Where such specifications are not available, use
the best available grade.
4.1 Sampling. Same as in Method 5, Section 3.1, with the addition of
deionized distilled water to conform to the American Society for Testing and
Materials Specification D 1193-74, Type II.
4.2 Sample Recovery. Same as in Method 5, Section 3.2, with the
following additions:
4.2.1 N2 Gas. No gas at delivery pressures high enough to provide a flow
of 20 liters/min for 1 hour through the sampling train.
4.2.2 Methylene Chloride.
4.2.3 Water. Same as in Section 4.1.
38
-------
e
,
03
ti
a
a>
bo
o
01
CD
03
O
a
o
CJ
u
03
I
CNJ
O
-------
4.3 Analysis. Same as in Method 5, Section 3.3, with the following
additions:
4.3.1 Methylene Chloride.
4.3.2 Ammonium Hydroxide. Concentrated (14.8 M) NH4OH.
4.3.3 Water. Same as in Section 4.1.
4.3.4 Phenolphthalein. The pH indicator solution, 1.0 percent in
60 percent alcohol.
5. PROCEDURE
5.1 Sampling. Same as in Method 5, Section 4.1, with the following
exceptions:
5.1.1 Place 100 ml of water in the first three impingers.
5.1.2 The use of silicone grease in train assembly is not recommended.
Teflon tape or similar means may be used to provide leak-free connections
between glassware.
5.2 Sample Recovery. Same as in Method 17, Section 4.2 with the addition
of a post-test N2 purge and specific changes in handling of -individual samples
as described below.
5.2.1 Post-test N? Purge for Sources Emitting SO,. (Note: This step is
recommended, but is optional. When no or little S02 is present in the gas
stream, i.e., the pH of the impinger solution is greater than 4.5, purging has
been found to be unnecessary.) As soon as possible after the post-test leak
check, detach the probe and filter from the impinger train. Leave the ice in
the impinger box to prevent removal of moisture during the purge. If
necessary, add more ice during the purge to maintain the gas temperature below
20*C. With no flow of gas through the clean purge line and fittings, attach
it to the input of the impinger train (see Figure 202-2). To avoid over- or
under-pressurizing the impinger array, s.lowly commence the N2 gas flow through
the line while simultaneously opening the meter box pump valve(s). Adjust the
pump bypass and N2 delivery rates to obtain the following conditions:
(1) 20 liters/min or AH@ and (2) an overflow rate through the rotameter of
less than 2 liters/min. Condition (2) guarantees that the N, delivery system
is operating at greater than ambient pressure and prevents that possibility of
passing ambient air (rather than N2) through the impingers. Continue the
purge under these conditions for 1 hour, checking the rotameter and AH
value(s) periodically. After 1 hour, simultaneously turn off the delivery and
pumping systems.
5.2.2 Sample Handling.
5.2.2.1 Container Nos. 1. 2. and 3. If filter catch is to be determined,
as detailed in Method 5, Section 4.2.
40
-------
5.2.2.2 Container No. 4 (Implnger Contents). Measure the liquid in the
first three impingers to within 1 ml using a clean graduated cylinder or by
weighing it to within 0.5 g using a balance. Record the volume or weight of
liquid present to be used to calculate the moisture content of the effluent
gas. Quantitatively transfer this liquid into a clean sample bottle (glass or
plastic); rinse each impinger and the connecting glassware, including probe
extension, twice with water, recover the rinse water and add it to the same
sample bottle. Mark the liquid level on the bottle.
5.2.2.3 Container No. 5 (MeCl2 Rinse). Follow the water rinses of each
impinger and the connecting glassware, including the probe extension with two
rinses of MeCl2; save the rinse products in a clean, glass sample jar. Mark
the liquid level on the jar.
5.2.2.4 Container No. 6 (Water Blank). Once during each field test,
place 500 ml of water in a separate sample container.
5.2.2.5 Container No. 7 (MeCl2 Blank). Once during each field test,
place in a separate glass sample jar a volume of MeCl2 approximately
equivalent to the volume used to conduct the MeCl2 rinse of the impingers.
5.2.2.6 Container No. 8 (Acetone Blank). As described in Method 5,
Section 4.2.
5.3 Analysis. Record the data required on a sheet such as the one shown
in Figure 202-3. Handle each sample container as follows:
5.3.1 Container Nos. 1. 2. and 3. If filter catch is analyzed, as
detailed in Method 5, Section 4.3.
5.3.2 Container Nos. 4 and 5. Note the level of liquid in the containers
and confirm on the analytical data sheet whether leakage occurred during
transport. If a noticeable amount of leakage has occurred, either void the
sample or use methods, subject to the approval of the Administrator, to
correct the final results. Measure the liquid in Container No. 4 either
volumetrically to ±1 ml or gravimetrically to ±0.5 g. Remove a 5-ml aliquot
and set aside for later ion chromatographic (1C) analysis of sulfates. (Note:
Do not use this aliquot to determine chlorides since the HC1 will be
evaporated during the first drying step; Section 8.2 details a procedure for
this analysis.)
5.3.2.1 Extraction. Separate the organic fraction of the sample by
adding the contents of Container No. 5 (MeCl2) to the contents of
Container No. 4 in a 1000-ml separatory funnel. After mixing, allow the
aqueous and organic phases to fully separate, and drain off most of the
organic/MeC!2 phase. Then add 75 ml of MeCl2 to the funnel, mix well, and
drain off the lower organic phase. Repeat with another 75 ml of MeCl?. This
extraction should yield about 250 ml of organic extract. Each time, leave a
small amount of the organic/MeC!2 phase in the separatory funnel ensuring that
no water is collected in the organic phase. Place the organic extract in a
tared 350-ml weighing tin.
41
-------
Moisture Determination
Volume or weight of liquid in impingers
Weight of moisture in silica gel ~
ml or g
9
Sample Preparation (Container No. 4)
Amount of liquid lost during transport
Final volume
pH of sample prior to analysis
Addition of NH4OH required?
Sample extracted 2X with 75 ml MeCl2?
For Titratlon of Sulfate
Normality of NH4OH
Volume of sample titrated
Volume of titrant
ml
ml
N
ml
ml
Sample Analysis
Container
number
Weight of Condensible Particulate, mg
Final Weight Tare Weight Weight Gain
4 (Inorganic)
4 & 5 (Organic)
Total
Less Blank
Weight of Condensible Particulate
Figure 202-3. Analytical data sheet.
42
-------
5.3.2.2 Organic Fraction Height Determination (Organic Phase from
Container Nos. 4 and 5). Evaporate the organic extract at room temperature
and pressure In a laboratory hood. Following evaporation, desiccate the
organic fraction for 24 hours in a desiccator containing anhydrous calcium
sulfate. Weigh to a constant weight and report the results to the nearest
0.1 mg.
5.3.2.3 Inorganic Fraction Weight Determination. Using a hot plate, or
equivalent, evaporate the aqueous phase to approximately 50 ml; then evaporate
to dryness in a 105'C oven. Redissolve the residue in 100 ml of water. Add
five drops of phenolphthalein to this solution, then add concentrated
(14.8 M) NH4OH until the sample turns pink. Any excess NH4OH will be
evaporated during the drying step. Evaporate the sample to dryness in a 105*C
oven, desiccate the sample for 24 hours, weigh to a constant weight, and
record the results to the nearest 0.1 mg. (Note: The addition of NH4OH is
recommended, but is optional when no or little S02 is present in the gas
stream, i.e., when the pH of the impinger solution is greater than 4.5, the
addition of NH4OH is not necessary.)
5.3.2.4 Analysis of Sulfate by 1C to Determine Ammonium Ion (NH4~)
Retained 1n the Sample. (Note: If NH4OH is not added, omit this step.)
Determine the amount of sulfate in the aliquot taken from Container No. 4
earlier as described in Method 5F (Appendix A, 40 CFR Part 60). Based on the
1C S04° analysis of the aliquot, calculate the correction factor to delete the
NH4" retained in the sample and to add the combined water removed by the acid-
base reaction (see Section 7.2).
5.3.3 Analysis of Water and MeClz Blanks (Container Nos. 6 and 7).
Analyze these sample blanks as described above in Sections 5.3.2.3 and
5.3.2.2, respectively. The sum of the values for the water blank and the
MeCl2 blank must be less than 2 mg or 5 percent of the mass of the CPM
(m0 + mr), whichever is greater. If the sum of the actual blank values is
greater, then subtract 2 mg or 5 percent of the mass of the CPM, whichever is
greater.
5.3.4 Analysis of Acetone Blank (Container No. 8).
Section 4.3.
Same as in Method 5,
6. CALIBRATION
Same as in Method 5, Section 5, except calibrate the 1C according to the
procedures in Method 5F, Section 5.
7. CALCULATIONS
Same as in Method 5, Section 6, with the following additions:
7.1 Nomenclature. Same as in Method 5, Section 6.1 with the
following additions.
ccpn) = Concentration of the CPM in the stack gas, dry basis, corrected to
standard conditions, g/dscm (g/dscf).
43
-------
CS04 = Concentration of S04" in the sample, mg/ml.
mb - Sum of the mass of the water and MeCl2 blanks, mg.
mc - Mass of the NH4" added to sample to form ammonium sulfate, mg.
mi - Mass of inorganic CPM matter, mg.
m0 = Mass of organic CPM, mg.
mr = Mass of dried sample from inorganic fraction, mg,.
m
re
ric
Mass of dried sample from inorganic fraction corrected for volume
of aliquot taken for 1C analysis, mg.
Volume of aliquot taken for 1C analysis, ml.
Volume of impinger contents sample, ml.
7.2 Correction for NH4" and H,0. Calculate the correction factor to
delete the NH4" retained in trie sample and to add the combined water removed
by the acid-base reaction based on the 1C S04=.
mc = K CS04
where:
K = 0.020502
7.3 Mass of Inorganic CPM.
m.
m.
1C
ic
- m
m
Eq. 202-1
Eq. 202-2
7.4 Concentration of CPM.
m
m
"cpm
std
Eq. 202-3
8. ALTERNATIVE PROCEDURES
8.1 Determination of NH4" Retained in Sample by Titration.
8.1.1 An alternative procedure to determine the amount of NH4" added to
the inorganic fraction by titration may be used. After dissolving the
inorganic residue in 100 ml of water, titrate the solution with 0.1 N NH4OH to
a pH of 7.0, as indicated by a pH meter. The 0.1 N NH4OH is made as follows:
44
-------
Add 7 ml of concentrate* (14.8 M) NH4OH to 1 liter of water. Standardize
against standardized 0.1 N H2S04 and calculate the exact normality using a
procedure parallel to that described in Section 5.5 of Method 6 (Appendix A,
40 CFR Part 60). Alternatively, purchase 0.1 N NH4OH that has been
standardized against a National Institute of Standards and Technology
reference material.
8.1.2 Calculate the concentration of S04" in the sample using the
following equation.
48.03 Vt N
Cso4 = Ec>- 202'4
100
where:
N - Normality of the NH4OH, mg/ml.
Vt - Volume of NH4OH titrant, ml.
48.03 - mg/meq.
100 = Volume of solution, ml.
8.1.3 Calculate the CPM as described in Section 7.
8.2 Analysis of Chlorides by 1C. At the conclusion of the final
weighing as described in Section 5.3.2.3, redissolve the inorganic fraction in
100 ml of water. Analyze an aliquot of the redissolved sample for chlorides
by 1C using techniques similar to those described in Method 5F for sulfates.
Previous drying of the sample should have removed all HC1. Therefore, the
remaining chlorides measured by 1C can be assumed to be NH4C1, and this weight
can be subtracted from the weight determined for CPM.
8.3 Air Purge to Remove S02 from Impinger Contents. As an alternative
to the post-test N2 purge described in Section 5.2.1, the tester may opt to
conduct the post-test purge with air at 20 liter/min. Note: The use of an
air purge is not as effective as a N2 purge.
9. BIBLIOGRAPHY
1. DeWees, W.D., S.C. Steinsberger, G.M. Plummer, L.T. Lay,
G.D. McAlister, and R.T. Shigehara. "Laboratory and Field Evaluation of the
EPA Method 5 Impinger Catch for Measuring Condensible Matter from Stationary
Sources." Paper presented at the 1989 EPA/AWMA International Symposium on
Measurement of Toxic and Related Air Pollutants. May 1-5, 1989. Raleigh,
North Carolina.
2. DeWees, W.D. and K.C. Steinsberger. "Method Development and
Evaluation of Draft Protocol for Measurement of Condensible Particulate
Emissions." Draft Report. November 17, 1989.
45
-------
3. Texas Air Control Board, Laboratory Division. "Determination of
Participate in Stack Gases Containing Sulfuric Acid and/or Sulfur Dioxide."
Laboratory Methods for Determination of Air Pollutants. Modified December 3,
1976.
4. Nothstein, Greg. Masters Thesis. University of Washington
Department of Environmental Health. Seattle, Washington.
5. "Particulate Source Test Procedures Adopted by Puget Sound Air
Pollution Control Agency Board of Directors." Puget Sound Air Pollution
Control Agency, Engineering Division. Seattle, Washington. August 11, 1983.
6. Commonwealth df Pennsylvania, Department of Environmental
Resources. Chapter 139, Sampling and Testing (Title 25, Rules and
Regulations, Part I, Department of Environmental Resources, Subpart C,
Protection of Natural Resources, Article III, Air Resources).
January 8, 1960.
7. Wisconsin Department of Natural Resources. Air Management
Operations Handbook. Revision 3. January 11, 1988.
46
-------
APPENDIX B
DATA ASSESSMENT
Statistical evaluations conducted using the results from the field
evaluations included: (1) determination of the precision for combined sampling
and analysis; (2) comparison of the sample results for the trains subjected to
a posttest purge to those having undergone no purge to evaluate the effect of
the purge on the reduction of sul fates, chlorides, and organics; (3)
comparison of the amount of sul fate in the final residue as determined by ion
chromatography (1C) to the stoichiometric amount of sul fate in the residue as
determined by ammonium hydroxide (NH4OH) titrations; (4) comparison 6f the
chloride content of each sample before and after it is initially taken to
dryness; and (5) calculation of the collection efficiency of the impingers for
condensible emissions. The methodology for performing these evaluations are
explained below. Equations are also provided for calculation of the sul fate
content by titration.
B.I PRECISION OF COMBINED SAMPLING AND ANALYTICAL PROCEDURES
The precision of the combined sampling and analytical procedures was
calculated from the standard deviation, and expressed as the relative standard
deviation (coefficient of variance), between the results for the four trains
in each quad-train run. The equations for the calculation of precision are
shown below.
(x, -
1/2
where
s = Standard deviation,
x1 = Result for ith sample,
x * Mean of sample results, and
n = Number of samples.
s
RSD(CV) =
x
where
RSD = Relative standard deviation or coefficient of variance.
47
-------
In order to obtain a more accurate estimate of precision, an F test was
run on the results from different quad-train runs to determine if the sample
variances were homogeneous. If the variances were homogeneous, the standard
deviations were pooled. Otherwise, a different precision was reported for
each run.
B.2 EFFECT OF POSTTEST PURGE
The effect of the posttest purge on the results for sulfates, chlorides,
and organics was determined using a matched-pair t-test. The null hypothesis,
H , was ud - 0 and Ha: ud = 0, where ud was the mean of the.differences between
tne results for the purged trains and the unpurged trains for each quad-train
run. The percent difference between the pairs of results was calculated.
B.3 COMPARISON OF SULFATES IN FINAL RESIDUE BY 1C VERSUS NH4OH TITRATION
The comparison of the results for sulfates in the final residue using 1C
versus NH4OH titration for measurement was also determined using a matched-
pair t-test. The null hypothesis, H , was ud = 0 and Ha: ud = 0, where ud was
the mean of the differences between the sulfates in the final residue as
measured by 1C and NH4OH titration. The difference between each pair of
results was also calculated as percent difference.
B.4 EFFECT OF DRYING ON CHLORIDES
The effect on the chloride content of initially taking the inorganic
fraction to dryness (using heat) was determined using a matched-pair t-test.
The null hypothesis, H0, was ud = 0 and Ha: ud = 0, where ud was the mean of
differences between the chlorides in each sample before and after heating.
The differences of the results was also calculated for each sample using the
before heating result as the basis:
% D = CT (after) - CT (before) / CT (before)
B.5 COLLECTION EFFICIENCY OF IMPINGERS FOR CONDENSIBLE EMISSIONS
The collection efficiency of the candidate train for collecting
condensible emissions was calculated from the results of the third field test
at a coal-fired boiler. The normal number of impingers (three) should collect
greater than 99% of the emissions. Thus, an extra (collection efficiency)
impinger placed after the third impinger should collect less than 1% of the
sample. The equation used for calculation of collection efficiency is shown
below:
Collection efficiency =
Meff
48
-------
where
M^ - Mass of sample caught in first three impingers, and
Meff - Mass of sample collected in extra impinger.
B.6 SULFATE CONTENT BY TITRATION WITH AMMONIUM CHLORIDE
In some cases, the amount of sulfate in the sample as HjSO^H.O was
determined by titrating with 0.1 N NH4OH to an endpoint pH 7.0. The titration
produces the reaction shown below:
H2S04 + 2 NH4OH ---> (NH4)2S04 + 2 H20
The amount of sulfate in the sample was calculated using the above
reaction and Technique 1 as follows:
Technique 1
H2S04 - 98 mg,
2 NH4OH = 70 mg,
(NH,),S04 = 132 mg, and
2 H20 = 36 mg.
normality NH4OH (meq) 35 mg NH4OH 17 mg NH3
mg NH, = ml NH4OH x : x x
ml NH/ meq NH4OH 35 mg NH4OH
The amount of sulfate was calculated using the above reaction and
Technique 2 as follows:
Technique 2
S04"2 = 96 mg, and
2NH/ = 36 mg.
normality NH4OH (meq) 35 mg NH4OH 18 mg NH/
mg NH/ = ml NH4OH x
ml NH/ meq NH4OH 35 mg NH4OH
96 mg SO/2
mg SO/2 = mg NH/ x
36 mg NH/
Both calculation techniques were found to be accurate for determination of
sulfate content in the sample. The weight of ammonia or ammonium added as
titrant was subtracted from the gravimetric determinations to give the
"corrected weights."
49
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA 450/4-90-012
4. TITLE AND SUBTITLE Method Development and Evaluation of
Draft Protocol for Measurement of Condensible Parti cul ate
Emissions
7. AUTHOR(S)
William G. DeWees
Kathy C. Steinsberger
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Entropy Environmentalists, Inc.
Post Office Box 12291
Research Triangle Park, NC 27709
-2. SPONSORING AGENCY NAME AND ADDRESS
Emission Measurement Branch and Control Technology Center
U.S. Environmental Protection Agency
Research Triangle Park, NC
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
May 16, 1990
6. PERFORMING ORGANIZATION CODE
8. PE-RFORMING ORGANIZATION REPORT MI
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVEREI
14. SPONSORING AGENCY CODE
EPA 200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The report describes the development and evaluation of a draft protocol for
measuring condensible particulate matter (CPM) emissions. An impinger catch approach
was evaluated in the laboratory and at wood-fired and coal-fired boilers. The report
summarizes the approach and major findings and recommendations which resulted from the
study. Test results and statistical analysis of the results are presented.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
18 DISTRI BUTION STATEMENT
Release unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
19 SECURITY CLASS (T/nr Repurt)
Unclassified
20 SECURITY CLASS 1 This page)
Unclassified
c. COSATI Field/Group
21 NO OF PAGES
55
22 PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |