Information is our business.
FIELD VALIDATION OF THE DNPH METHOD FOR
ALDEHYDES AND KETONES
RADIAN CORP., RESEARCH TRIANGLE PARK, NC
APR 96
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
EPA/600/R-96/050
April 1996
DCNNo.: 95-650-049-12-04
EPA No.: 68-D4-0022
Field Validation of the DNPH Method for Aldehydes and Ketones
Final Report
Work Assignment 12
Prepared for;
Joseph E. Knoll
National Exposure Research Laboratory
Air Measurements Research Division
Methods Branch
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Prepared by:
Gerald S. Workman, Jr.
Joette L. Steger
Radian Corporation
P. O. Box 13000
Research Triangle Park, North Carolina 27709
January 30,1996
-------
-
TSCHHICAL REPORT DATA
1. REPORT NO.
EPA/600/R-96/05Q
2.
t. im£ AND SUBTITLE
Field Validation of the DHFH Method for Aldehydea
and Ketones.
5.REP0W DATE
April 1996
6.PERFORMING ORGANIZATION CODE
1. AUTHORIS)
Cerald 6. Workman, Jr. and Joette L. Stager
B.PERFORMING ORGANIZATION REPORT HO.
9. PERFORMING ORGANIZATION NAME AMD ADDRESS
Radian Corporation
P.O. Box 13000
Research Triangle Park, North Carolina 27709
10.PROGRAM ELEHEHT NO.
11. CONTRACT/GRANT NO.
68-04-0022
12. SPONSORING AGENCY NAKE AND ADDRESS
national Expoaure Reaearch Laboratory
Office of Reaearch and Development
U.S. Environmental Protection Agency
Reaearch Triangle Park, HC 27711
13.TYPE OF REPORT AND PERIOD CCVERED
Project Report
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The
- A stationary source emission teat method for selected aldehydes and ketones has been validated
siethod employs a aanpling train vhith impingera containing 2,«-dinitrophenylhydrarine (DHPHI to
derlvatlze the analytea. The resulting hydrazones are recovered and analyzed by high performance liquid
chromatography. Nine analytea vera studied: formaldehyde, acetaldehyda, propionaldehyde, methyl ethyl
ketone, methyl laobutyl ketone, acetophenone, iaophorone, and quinone.' The study employed the
validation techniques deacribed in EPA method 301, which uaea train spiking to determine biaa, and
collocated aampling traina to determine precialon. The atudiea vera carried out at a plywood veneer
dryer and a polyester manufacturing plant.
On the basis of the Method 301 criteria for preciaion f£50l relative standard deviation) and blaa
(correction factor of 1.00 ~0.30jlthe nethod waa validated"for formaldehyde, aeetaldehyde,
propionaldehyde, acetophenone and lsophorone. Acrolein, ssethyl ethyl ketone, aethyl isebutyl ketone,
and quinone did not Mat the validation criteria.
17.
KIT WOKS AND KXU4DIT AMAL1S1S
uininoM
b.iDtmxritM/ omm kkdcs nw
C.C0MTI
la. nmieurioM tnmunT
RELEASE TO f-'BtlC
i». security clmi trot* upon)
UNCLASSIFIED
31 .MO. or IASU
BA.
28. ataimrr cmi mt* »
-------
Acknowledgements
Under EPA Contract No. 68-D4-0022, Radian Corporation prepared this report with
the supervision and guidance of Merrill D. Jackson, EPA Project Officer and Joseph E.
Knoll, EPA Work Assignment Manager, in the National Exposure Research
Laboratory, Air Measurements Research Division, Research Triangle Park, North
Carolina. The Radian Corporation Project Manager was Joan Bursey; Gerald
Workman was the Principal Investigator; Joette Steger and James Howes provided the
technical review; and Robert Petty and Joe Fanjoy were the editors.
iii
-------
TABLE OF CONTENTS
Plage
1.0 INTRODUCTION . 1
2.0 CONCLUSIONS AND RECOMMENDATIONS 6
3.0 FIELD TEST I 8
FIELD SAMPLING 8
ANALYSIS 14
STATISTICAL ANALYSIS 22
4.0 FIELD TEST II 27
FIELD SAMPLING 27
ANALYSIS 33
STATISTICAL ANALYSIS 37
5.0 FIELD TEST PROCEDURES 43
EQUIPMENT 44
Probe 44
Sampling Trains 45
Dynamic Spiking Apparatus 45
PREPARATION 49
Glassware Preparation 49
DNPH Preparation 49
Method 001 r Equipment Preparation 49
SAMPLING OPERATIONS 50
Preparation of Sampling Trains 50
Sample Recovery 51
Container 1 - Probe Rinse, First and Second Impinger Contents- .. 53
Container 2 - Third and Fourth Impinger Contents- 53
Field Train Blank(s) 53
Field Reagent Blank(s) 53
Sample Storage and Shipping 54
v
-------
TABLE OF CONTENTS - Continued
Itage
ANALYTICAL PROCEDURES 54
Sample Preparation .54
Extraction 56
Solvent Exchange 56
Chromatographic Analyses 58
Standard Preparation- 58
Instrument Calibration- 59
Sample Analysis- 60
Laboratory Method Blanks 60
QUANTITATION 60
SPIKING 61
PRECISION AND ACCURACY ASSESSMENT 63
Assessment of Precision According to Method 301 64
Assessment of Bias According to Method 3011 66
6.0 QUALITY ASSURANCE/QUALITY CONTROL 69
QUALITY CONTROL 69
Sampling QA/QC Procedures 69
Data Quality Objectives- 69
Manual Method Performance Criteria- 70
Field Equipment Calibrations- 71
Sampling Operation/Recovery Procedures— 72
Representative Sampling- 72
Documentation— 77
Sample Custody- . 77
Laboratory QA/QC Procedures 81
Sample Custody/Tracking- 82
Calibration Curve- 82
Daily QC Checks- 82
System Blanks- 82
LaboratoryJS^ethodBlanks-* 88
Laboratory Method Spikes and Method Spike Duplicates- 88
Field Train and Field Reagent Blanks- 94
7.0 REFERENCES 99
vi
-------
TABLE OF CONTENTS - Continued
APPENDICES
A Results From Preliminary Laboratory Study
B Sampling and Analytical Methods
B. 1 Aldehyde and Ketone Sampling Checklist
B.2 Aldehyde and Ketone Sampling Method
C Site Survey Analysis Results
vii
-------
LIST OF FIGURES
Page
1 Sampling Train for Aldehydes and Ketones 9
2 Sampling Location for Aldehyde and Ketone Method Development
Testing Program 11
3 Diagram of Sampling Location for the Second Carbonyl Field Test 28
4 Quad-Train Probe and Pitot Arrangement 46
5 Upper and Lower Sampling Probes (Side View) 47
6 Dynamic Spiking Apparatus 48
7 Sample Identification Code . 78
8 Example of Sample Label and Integrity Seal 79
9 Chain-of-Custody Record 80
viii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
. 3
. 4
12
15
16
18
20
23
26
30
32
34
36
38
40
42
52
55
57
59
LIST OF TABLES
Results of the EPA Method 301 Statistical Evaluation
Aldehydes and Ketones Included on the Clean Air Act Title m List
Sampling Parameters, Field Test I (August 1994)
Spike Quantities for First Field Test (August 1994)
Analytical Results for First Two Impingers for Field Test I (August 1994)
Analytical Results for All Fractions (Field Test I, August 1994)
Spike Recovery for Field Test I (August 1994)
Breakthrough Analysis
Summary of Method 301 Statistical Analysis (Field Test I, August 1994) ,
Sampling Parameters, Field Test n (April 1995)
Spike Quantities
Analytical Results, Impingers 1 and 2
Spike Recovery
Analytical Results, All Fractions
Breakthrough Analysis
Statistical Analysis Using First Two Impingers
Sample Recovery Scheme
Hold Time Between Sample Collection and Sample Extraction ........
Solvent Exchange and Dilution Procedures
HPLC Gradient for Analysis of DNPH-Derivatized Aldehydes
ix
-------
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
59
62
70
70
73
75
81
83
84
86
89
90
91
92
95
96
97
98
LIST OF TABLES - Continued
Retention Times of Aldehyde Derivatives
Compounds Spiked and Nominal Spike Concentrations
Field Sampling Quality Control Objectives
Summary of Acceptance Criteria, Control Limits, and Corrective Action . .
Leak Rates, Field Test I
Leak Rates, Field Test H
Laboratory Quality Control Procedures
Calibration Data
Calibration Check Standard Recoveries for Field Test I
Calibration Check Standard Recoveries for Field Test n
Laboratory Method Blank Results for Field Test I
Laboratory Method Blank Results for Field Test U
Percent Recovery for Method Spike Samples for Field Test I
Percent Recovery for Method Spike Samples for Field Test II
Field Train Blank Results in Total Micrograms of Carbonyl for Field Test I
Field Train Blank Results in Total Micrograms of Carbonyl for Field Test II
Field Reagent Blank Results for Methylene Chloride
Blank (Field Test I, August 1994)
Field Reagent Blank Results in Total Micrograms of Carbonyl for
Field Test n
x
-------
SECTION 1.0
INTRODUCTION
Radian Corporation, while assisting the Method Branch of the National Exposure
Research Laboratory (NERL), has evaluated and validated a multiple pollutant sampling and
analytical method for aldehydes and ketones in emissions from stationary sources. This study
is part of an EPA program to develop stationary source emission test methods for the 189
hazardous air pollutants listed in the Clean Air Act Amendments of 1990, and which are
needed to determine risk to the public and to support the regulatory process.
The method in the present study employs an impinger train containing acidified
2,4-dinitrophenylhydrazine (DNPH) to capture and derivatize aldehyde and ketone
compounds. Validation of the test method was needed to demonstrate applicability to different
source types. Test sites known to emit relatively low concentrations of both acetaldehyde and
formaldehyde were selected. Under Work Assignment 67 of EPA Contract 68-D1-0010, the
method was evaluated at a plywood veneer dryer vent at a pressboard manufacturing plant;
under Work Assignment 12 of EPA Contract 68-D4-0022, method evaluation was conducted at
a spinning machine exhaust vent at a polyester fiber manufacturing plant. Site parameters and
aldehyde concentrations were confirmed with information gathered during pretest site surveys.
The present report covers both of these field validation studies.
The method was evaluated using procedures described in EPA Method 301,1 Protocol
for the Field Validation of Emission Concentrations from Stationary Sources, in which bias is
determined by spiking sample trains and precision is determined by collocating sampling
trains. In the present study, spiking was carried out by a dynamic method in which measured
quantities of analyte were introduced into the flue gas being sampled.
1
-------
Precision and bias of the test method for each compound tested are summarized in
Table 1. For Field Test I data is shown for both two and four impingers. Precision and bias
calculations were completed using all four impingers for Field Test I because of the high
breakthrough values that occurred during Runs 3, 4, 6, and 7. Two-impinger data also was
reported for Field Test I to demonstrate that formaldehyde and acetophenone passed with only
two impingers. For Field Test II, data is shown for two impingers only because breakthrough
levels for all of the trains were low and there was little difference in total amounts recovered
between the two- and four-impinger data sets.
For Field Test I and Field Test II, four sampling trains were operated simultaneously
(quadruplicate sampling train) to collect flue gas samples. The configuration of each sampling
train was the same as that described in SW-846 Method 00112 for formaldehyde, except that
the first impinger contained 200 mL of reagent to increase sample capacity, and an additional
impinger containing DNPH was added to check for breakthrough. The actual method
evaluated is included in Appendix B. In this sampling method, gaseous and particulate
pollutants are collected from an emission source in aqueous, acidic DNPH. Aldehydes and
ketones present in the stack gas stream react with the DNPH to form dinitrophenylhydrazones.
Samples are then extracted with organic solvent. The resulting organic extract is concentrated
as necessary and exchanged into an appropriate solvent for analysis by high performance liquid
chromatography (HPLC).
Ten aldehydes and ketones listed in Title III of the Clean Air Act were studied as part
of this project. These compounds are listed in Table 2. Nine of the ten compounds listed in
Table 2—formaldehyde, acetaldehyde, quinone, acrolein, propionaldehyde, methyl ethyl
ketone, acetophenone, methyl isobutyl ketone, and isophorone—were spiked into the sampling
trains during sample collection as part of the method evaluation procedure at the first field test
site. The compound 2-chloroacetophenone was excluded from the list of compounds
quantifiable by this method because a purified DNPH derivative of this compound could not be
successfully made during the initial laboratory studies. Furthermore, because
2
-------
Table 1. Results of the EPA Method 301 Statistical Evaluation
Parameter
Form-
aldehyde
Acet-
aldehyde
Propion-
aldehyde
Aceto-
phenone
Methyl
Ethyl
Ketone
Methyl
Isobutyl
Ketone
Isopboroiie
Quinone
Acrolein
Field Test 1*
RSD Spiked (56)
7.36
7.18
7.20
7.94
26.1
17.2
7.94
40.0
12.1
RSD Unspiked (%)
10.2
10.6
21.0
42.5
74.3
32.2
211
39.7
17.3
Bias CF
1.11
1.26
1.25
1.08
2.55
2.22
1.08
1.84
2.00
Disposition
Pass
Pass
Pass
Pass
Fail
Fail
Fail
Fail
Fail
Field Test lfc
-
RSD Spiked <%)
7.32
8.15
NR
7.79
NR
NR
NR
NR
NR
RSD Unspiked (%)
9.95
10.3
NR
43.5
NR
NR
NR
NR
NR
Bias CF
1.10
1.34
NR
1.11
NR
NR
NR
NR
NR
Disposition
Pass
Fails
NR
Pass
NR
NR
NR
NR
NR
Field Test V
RSD Spiked (%)
8.8
16.7
12.9
10.4
18.8
21.2
9.0
NT
NT
RSD Unspiked (%}
20.7 '
12.4
48.5
-
_
-
-
NT
NT
Bias CF
1.10
1.24
1.29
1.09
2.45
4.33
0.93
NT
NT
Disoosition
Pas?
Pass
Pass
Pass
Fail
Fail
Pass
NT
NT
NR = Not Repotted
NT® Not Tested
RSD = Relative Standard Deviation
CF = Conection Factor
"Statistics calculated from 4-impinger results in Field Test 1.
^Statistics calculated from 2-impinger results in Field Test I.
'Statistics calculated from 2-impinger results in Field Test 2,
-------
2-chloroacetophenone can be determined by Method 0010,3 there was no need to include it in
the Method 00112 validation study.
Table 2. Aldehydes and Ketones Included on the Clean Air Act Title III List
»SBSg= 1 i ri" i"ll I ii' 'in - ¦
Formaldehyde
Acetaldehyde
Quinone
Acrolein
Propionaldehyde
Methyl Ethyl Ketone
Acetophenone
Methyl Isobutyl Ketone
2-Chloroacetophenone
Isophorone
For Field Test II, acrolein and quinone *.vere not included in the spiking solution.
Acrolein is chemically unstable under the acidic reaction conditions because of its double
bond. Acrolein is a highly reactive substance and is known to dimerize by the Diels-Alder
reaction. Acrolein may also react with other aldehydes, causing their recoveries to be low.
Therefore, acrolein was considered inappropriate to study as part of a multiple pollutant
aldehyde and ketone method test. A pollutant-specific method may be required to measure
acrolein emissions. Quinone appears to be collected in the impingers but does not react well
with the DNPH under the conditions specified in the method. Quinone is also a strong
oxidizing agent having the potential to oxidize formaldehyde, and its addition to the spiking
solutions may have caused low recoveries of some aldehydes during the first field test. For
these reasons, quinone was also excluded from the second field study. Of the compounds that
were spiked, the laboratory studies indicated the method would perform satisfactorily for five:
formaldehyde, acetaldehyde, propionaldehyde, acetophenone, and isophorone. Methyl ethyl
ketone and methyl isobutyl ketone in the impingers and do not react rapidly enough with the
4
-------
DNPH to be quantitatively collected. The two compounds are volatile and are swept through
the sampling train before they have time to react.
This test report is divided into seven sections. Section 2 is a summary of the validation
test results including the conclusions and recommendations based on the results of the field
validation tests and laboratory studies. Sections 3 and 4 present the results of Field Test I and
Field Test n, respectively. Sampling and analytical procedures are detailed in Section 5.
Quality assurance/quality control (QA/QC) data are described in Section 6 and references are
provided in Section 7.
5
-------
SECTION 2.0
CONCLUSIONS AND RECOMMENDATIONS
Based on the work performed in the laboratory studies and the field evaluation of the
aldehydes and ketones, and using Method 3011 criteria as revised in December, 1994, the
following conclusions may be drawn regarding the proposed sampling method.
• Acetophenone, Formaldehyde, Isophorone, Acetaldehyde, and
Propionaldehyde Using the criterion of 70-130% recovery for the dynamically
spiked compounds, acetophenone, formaldehyde, isophorone, acetaldehyde, and
propionaldehyde meet the minimum recovery criterion.
• Quinone, Acrolein, Methyl ethyl ketone, and Methyl isobutyl ketone The
test method is not appropriate for the measurement of quinone, acrolein, methyl
ethyl ketone, and methyl isobutyl ketone, due either to poor collection
efficiency or analytical problems.
• Formaldehyde, Acetaldehyde, Propionaldehyde, Methyl Ethyl Ketone,
Acetophenone, and Methyl Isobutyl ketone are all stable in the aqueous
spiking solution for up to 62 days.
• All Compounds Except Formaldehyde Dynamic spiking allowed the
collection efficiency of the train to be more adequately evaluated than static
spiking and is the preferred spiking technique especially when very volatile,
water-purgeable compounds are being tested,
• AU Compounds Keeping the first two impingers in an ice bath results in higher
compound recoveries with less breakthrough into the second impinger and less
tautomer formation than when the first two impingers are kept warm.
Based on work performed in the laboratory and in the field evaluation, the following
recommendations are made:
* Subject to the number of impingers used for various compounds (as stated
below), the sampling and analytical method tested is recommended for adoption
as a standard EPA method for the determination of formaldehyde,
6
-------
acetophenone, isophorone, acetaldehyde, and propionaldehyde emissions from
stationary sources.
To obtain quantitative recoveries of formaldehyde, acetophenone, and
isophorone, use 200 mL of DNPK reagent in the first impinger followed by one
impinger containing 100 mL and keep the impingers iced. To obtain
quantitative recoveries of acetaldehyde and propionaldehyde, use 200 mL of
DNPH reagent in the first impinger followed by two impingers containing
100 mL and keep the impingers iced.
Recoveries for acrolein in the laboratory studies were low, probably due to the
reactive nature of the double bond. Alternative sampling and analytical methods
should be pursued for acrolein or modifications should be made to Method
00112 to stabilize acrolein. Potential modifications to Method 00112 include
using hexane to recover the sample trains instead of methylene chloride.
Method 00112 yields inconsistent results when used to determine quinone.
Alternative sampling and analytical methods should be investigated for quinone.
Methyl isobutyl ketone and methyl ethyl ketone are not efficiently collected by
the aqueous DNPH reagent. Alternative sampling and analytical methods,
possibly using sorbents, should be investigated for these compounds.
Alternatively, modifications to Method 001 lz such as using five or more reagent
impingers, sampling at lower flow rates, using a lower pH reagent (>2N HC1),
may improve the performance of Method 00 ll2 for these compounds.
7
-------
SECTION 3.0
FIELD TEST I
The first Method 00112 field evaluation study was conducted at a plywood veneer
manufacturing plant during the weeks of July 26 and August 1, 1994. Ten runs were
performed using quadruplicate aldehyde and ketone sampling trains. The sampling train that
was evaluated is shown in Figure 1. Dynamic analyte spiking was used for method evaluation.
The dynamic spiking apparatus and procedure are described in detail in Section 5.
Samples were analyzed and the analysis results were used to determine the method
precision and bias for each of the spiked compounds by EPA Method 301.1 Two fractions
from each individual sampling train were recovered and used to detect and quantify the amount
of breakthrough of the nine test compounds through the DNPH solution in the first two
impingers. Laboratory results in total micrograms of each compound were summed for the
probe rinse, first impinger contents, and second impinger contents (Fraction 1) and for the
third impinger and knockout rinse (Fraction 2). Breakthrough was calculated as the percentage
of the total that was found in Fraction 2. Recovery efficiency of the sampling and analytical
method for the aldehyde and ketone compounds was determined using the data from the 20
dynamically spiked trains.
Details of the sampling runs and results of the laboratory and statistical analyses are
presented in the following subsections.
FIELD SAMPLING
Flue gas samples for caibonyl analysis were collected at a plywood veneer
manufacturing plant from a dryer used to dry the plywood veneer before shipping. Samples
were collected from the first dryer stack. The sampling ports were 6-inch (152 mm) diameter
8
-------
vo
Temperature
Sensor
V
S-Type
Pilot tube
Orifice
Manometer
Temperature
Sensor
Figure 1. Sampling Train for Aldehydes and Ketones
-------
pipe nipples located approximately 2.6 meters above the sampling platform. The sampling
platform was approximately 12 meters above ground level. Figure 2 is a diagram of the
sampling location. The ports were located at least 4 stack diameters downstream and 1 stack
diameter upstream of the nearest flow disturbances. Preliminary samples were collected from
the dryer stack during a pretest site survey. Formaldehyde, acetaldehyde, propionaldehyde,
and acrolein were all detected in the dryer stack gas at levels over 10 times the method
detection limit. Other aldehydes and ketones, including methyl ethyl ketone and methyl
isobutyl ketone, were also identified and determined to be present at low concentrations in the
samples.
Ten quad train runs were completed at the test site. The quad-train probe is described
in detail in Section 5. Trains A and D were spiked and Trains B and C were unspiked.
Table 3 summarizes the sampling parameters recorded for each run. The diameter of all the
sampling nozzles was 5.72 mm. The static pressure in the stack was positive, and remained
constant at approximately 15 mm of water during all test runs. The target sample volume for
each run was 0.85 cubic meters. The sampling time was 75 minutes.
Because of the additional liquid spiked into Trains A and D, only Trains B and C were
used to calculate the percentage of moisture in the stack gas. Moisture values were in the
range of 19 to 28% by volume because of the high level of moisture expelled from the
product.
Hie stack temperature and velocity for each run were measured using a single
thermocouple and S-Type pitot tube on the sampling probe assembly. Individual stack gas
temperature and pitot tube differential pressure measurements were taken for each of the four
trains at the time the other sampling data were recorded. This measurement scheme resulted in
some slightly different temperature and velocity data associated with individual trains for the
same run, even though measurements were made with a common probe. These temperature
and differential pressure measurement differences did not affect the test data because the
10
-------
4 - 12 Pitch Roof
Figure 2. Sampling Location for the First Aldehyde and Ketone Method Development Field Test
-------
Table 3. Sampling Parameters, Field Tert I (August 1994)
Sampling Standard Stack Stack Gas
Duration Moisture Meter Volume Temperature Velocity Percent
Run
(miB
(dscm)
r a
tow)
Isokinetic
1A
75
0.871
160
881.0
98.97
IB
75
27,4
0.820
160
887.4
97.09
1C
75
25.8
0.924
160
884.6
107,4
ID
75
-
0.864
160
881.5
98.70
2A
IS
-
0.944
194
1014
101.9
2B
15
26.6
0.904
194
1018
99.60
2C
15
26,2
0.906
194
1017
99.27
2D
15
-
0.932
194
1015
101.1
3A
75
1.00
197
1042
107.0
3B
75
27,7
0.908
197
1046
99.19
3C
75
27.2
0.926
197
1045
100.5
3D
75
_
0.914
197
1043
98.40
4A
75
-
0.849
189
980.4
97.52
4B
75
28.4
0.830
189
982.4
96,47
4C
75
27.7
0.849
189
980.9
97.78
4D
75
-
0.837
189
978.9
95.30
5A
IS
-
0.889
189
986.0
97.53
SB
15
26.0
0.852
189
989.2
95.21
5C
15
25.2
0.852
189
987.4
94.19
5D
IS
_
0.872
189
985.3
95.34
6A
IS
~
0.919
191
988.5
104.3
6B
75
29.4
0.860
191
993.7
100.5
6C
75
29.0
0.858
191
992.7
99.70
6D
15
-
0.862
191
990.6
99.01
7A
IS
-
0.912
203
1022
95.21
7B
75
22.7
0.885
203
1026
93.89
7C
75
21.6
0.892
203
1023
93.61
7D
IS
0.873
203
1020
90.15
12
-------
Table 3. (Continued)
Run
Sampling
Duration
(mill
Moisture
(%>
Standard
Meter Volume
fdscm)
Stack
Temperature
CC)
Slack Gas
V>lnrifv
V wlvwl%^
(mDm)
Percent
Isokinetic
8A
75
-
0.978
204
1049
98.09
8B
75
23.6
0.924
204
1057
96.47
8C
75
20.9
0.943
204
1051
95.62
8D
75
-
0.954
204
1050
96.09
9A
75
-
0.931
204
988.8
96.98
9B
75
20.3
0.891
204
993.3
95.02
9C
75
19.7
0.868
204
992.1
91.91
9D
75
-
0.882
204
989.2
92.11
10A
75
-
0.859
203
980.8
91.11
10B
75
19.7
0.852
203
981.8
90.81
IOC
75
19.5
0.863
203
981.4
91.83
10D
75
-
0.856 '
203
982.1
91.42
13
-------
sample for all four trains was collected from the same point, the volumes collected were
recorded, and the data was corrected for the slight differences in sample volume.
The spiking system was operated to inject approximately equal quantities of spiking
solution into Trains A and D during each sampling run. The actual amounts spiked varied
from train to train because the syringe pumps used did not always deliver exactly the same
amount of spiking solution. The results of the laboratory study indicated that dynamic spiking
was preferable to static spiking even though it resulted in variable spike amounts. Table 4
shows the quantity of each compound spiked into Trains A and D during each run. Spiked
quantities were determined by weighing the spiking syringes before and after each test run.
Spike weights were recorded in a field notebook.
ANALYSIS
The samples from each train were collected and analyzed in two fractions. The first
fraction contained the probe rinse and contents of the first two impingers. The second fraction
contained the contents of the third and fourth impingers. Table 5 shows the results of the
analysis of the first fraction from each run.
Table 6 shows cumulative analytical results for both fractions combined (all impingers)
of each sampling train. Table 7 shows the percentage of each spiked compound recovered in
all four impingers. The recovery is calculated as follows:
R = 100% x S " M
CS
where:
R = percent recovery,
S = measured quantity in the spiked sample,
M = mean value of the unspiked samples in the run, and
CS = calculated spike quantity.
14
-------
Table 4. Spike Quantities for First Field Test (August 1994)
Quantity Spiked (tig)
Run
Formaldehyde
Acetnldehvde
Acrolein
ProDinnnldchvde
Acctoohenone
MEK
MIBK
Isoohorone
1A
20,700
12,500
4,220
4,420
8,900
5,310
7,100
9,650
ID
18,700
11,300
3,820
4,000
8,050
4,810
6,430
8,730
2A
20,400
12,400
4,180
4,370
8,810
5,260
7,030
9,550
2D
22,400
13,600
4,580
4,800
9,660
5,770
7,710
10,500
3A
22,900
13,900
4,690
4,910
9,880
5,900
7,890
10,700
3D
23,800
14,400
4,850
5,080
10,200
6,110
8,180
11,100
4A
22,400
13,600
4,580
4,800
9,660
5,770
7,710
10,500
4D
21,900
13,300
4,470
4,680
9,430
5,630
7,530
10,200
5A
20,300
12,300
4,160
4,350
8,760
5,230
7,000
9,510
5D
21,400
13,000
4,370
4,570
9,210
5,500
7,350
9,990
6A
23,500
14,200
4,790
5,020
10,100
6,030
8,070
11,000
6D
26,000
15,700
5,300
5,550
11,200
6,680
8,920
12,100
7A
23,500
- 14,200
4,790
5,020
10,100
6,030
8,070
11,000
7D
22,000
13,300
4,490
4,710
9,480
5,660
7,570
10,300
8A
21,700
13,100
4,430
4,640
9,340
5,580
7,460
10,100
8D
22,100
13,400
4,520
4,730
9,520
5,690
7,600
10,300
9A
21,600
13,100
4,410
4,620
9,300
5,550
7,430
10,100
9D
22,400
13,600
4,580
4,800
9,660
5,770
7,710
10,500
10A
21,300
13,000
4,350
4,550
9,160
5,470
7,320
9,940
10D
21.200
12,800
4,320
4,530
9,120
5,450
7,280
9,890
-------
Table 5. Analytical Results for First Two Impingers for Field Test I (August 1994)
Quantity Measured (ug)
Run
Formaldehyde
Acetaldehvde
Ouinnne
Acrolein
Proplon-
nldehvde
Acetoohenone
MEK
MIBK
Isoohorone
1A
24,400
14,400
6,850
5,250
4,020
7,740
2,110
4,400
8,250
IB
11,900*
6,340
2,320
1,900
614
150
183
307
42.3
1C
11,800
6,670
2,390
1,860
503
180*
228*
282"
ND
ID
27,100
14,850
3,640
5,100
4,400
8,530
1,500
2,960
8,550
2A
24,500
16,000
7,370
2,980
3,900
8,730
1,330
2,690
9,190
2B
6,290
6,740
2,360
457
231
20.2'1*
46.4*
61.8
11.4*
2C
5,480
5,500
427b
396
217**
<0.84
260*
45.5"
ND
2D
27,100
17,400
6,480
3,030
4,140
9,280
2,030
2,990
7,770
3A
24,500
13,200
6,690
1,680
2,800
8,200
518
1,110
9,220
3B
4,950
5,060
2,300
230
106'
39.5
13.3*'*"
88.8
8.46*
3C
4,020
4,030
1,220
195
til*
20.5'*
21.6*
<0.23
ND
3D
29,800
17,000
7,920
2,440
3,830
10,600
702
1,900*
10,800
4A
27,000
16,200
5,240
2,660
4,060
9,010
850*
1,940
9,250
4B
5,780*
4,670
1,630*
490
143
12.3'*
34.6»
<0.12
8.16*
4C
5,420*
4,260
3,010
452
182
51.6
27.2*
<0.12
8.14*
4D
29,200
16,100
7,650
2,050
3,450
9,580
696
1,720*
9,460
5A
25,100
15,800
6,290
2,220
3,850
8,310
1,400
3,020
8,300
SB
5,350
4,280
1,240
289
182
<0.84
25.9*
<0.23
ND
5C
5,750
4,530
993
318
219
46.2
28.9*
<0.23
9.46*
5D
26,200
15,700
10,800
2,200
3,810
8,900
763
2,360
9,300
6A
18,400
10,700
8,110
1,270
2,700
7,190
832
1,580
7,650
6B
4,420*
3,000
1,500
98.6
77.5'
43.8
15.6*A
285
8.14
6C
4,530
2,860
1,690
98.3
65.5*
45.6
13.8'-*
274
11.7
6D
24,200*
9,800
5,890
2,920
2,410
8,070
419
1,050
9,280
7A
18,500
9,780
3,690
1,310
2,530
6,630
837
1,350
7,430
-------
TableS. (Continued)
Quantity Measured fug)
Run
Formaldehyde
Acetaldebvde
Ouinone
Acrolein
Propion-
aldehvde
Acetoobenone
MEK
MIBK
IsonhoroM
7B
3,330
3,150
1,180
128
132
7.60*-"
20.4*
38.4
13.4"
7C
3,900
3,590
1,290
116
117
13.4**
18.8*"
31.2
15.6"
7D
18,200
10,100
7,580
1,360
2,640
7,080
834
1,280
7,820
8A
25,000
14,600
6,440
2,070
3,510
8,380
1,060
2,510
9,240
8B
3,230
4,290
1,380
183
174
ll.fi**
30.4*
34.7
8.12b
8C
3,340
3,990
785
158
132
tl.2'*"
18.6**
34.1
9.27"
8D
26,700
15,700
6,500
1,870
3,450
8,920
771
2,140
9,810
9A
26,800
13,800
4,830
1,990
3,730
8,830
1,510
2,890
9,560
9B
3,010
2,940
1,410
136
136
9.21,b
21.7*
26.6
9.24"
9C
3,140
3,230
1,550
141
160
12. r*
20.7*
33.9
8.91"
9D
27,700
14,400
2,160
2,470
3,720
9,300
1,450
2,530
10,700
10A
24,200
12,600
869s
1,650
3,370
8,200
1,370
2,050
9,490
10B
3,230
3,630
893
158
212
20. rb
51.2*
38,1
166
10C
2,850
3,180
725
154
189
IS.O*1"
53.3*
36.4
8.87"
10D
24.400
14.400
7.220
1.760
3.480
8.860
1.130
2.650
10.100
NOTE; Final values are not corrected for the field train blank.
ND = Not Detected.
* Less than 10 times the field train blank.
"Below calibration curve, quantified by extrapolation.
'Above calibration curve, quantified by extrapolation.
-------
Table 6. Analytical Results for All Fractions (Field Test I, August 1994)
Quantity Measured (eg)
Run
Formaldehyde
Aeetaldehvde
Ouinone
Acrolein
Propion-
aldehvde
MEK
Aceto-
Dheiwoe
MIBK
IsoDinrotx
1A
24,500
15,000
6,870
5,410
4,250
2,640
7,860
5,090
8,520
IB
12,100
6,730
2,370
1,940
638
202
155
314
42.3
1C
11,800
7,(»0
2,390
1,860
503
228
180
282
ND
ID
27,100
15,500
3,640
5,250
4,620
2,310
8,680
3,910
8,800
2A
24,500
16,900
7,410
3,090
4,250
2,400
8,890
3,880
9,610
2B
6,300
7,640
2,370
469
279
60.9
20.2
61.8
11.4
2C
5,510
6,200
427
396
273
304
ND
45.5
ND
2D
27,100
18,200
6,550
3,130
4,450
2,970
9,430
3,850
8,270
3A
24,500
16,900
6,730
1,820
3,720
2,180
8,560
2,710
9,740
3B
4,990
6,460
2,330
242
158
28.3
39.5
105
8.46
3C
4,040
5,420
1,220
207
166
30.8
24.9
ND
ND
3D
29,900
19,500
7,940
2,600
4,480
1,880
11,000
3,150
11,500
4A
27,100
18,000
5,250
2,780
4,520
1,960
9,220
3,180
9,750
4B
5,800
5,790
1,630
504
189
47.6
12.3
ND
8.16
4C
5,450
5,660
3,020
488
245
45.8
53.7
ND
8.14
4D
29,300
19,900
7,680
2,240
4,370
2,200
9,990
3,490
10,100
5A
25,200
16,300
6,310
2,310
4,040
2,200
8,430
3,170
8,610
SB
5,380
4,770
1,240
310
210
39.4
ND
ND
0.32
sc
5,760
4,960
998
334
243
40.6
46.2
8.41
9.46
5D
26,200
16,800
10,800
2,290
4,120
15,900
9,060
3,230
9,720
6A
18,400
13,200
8,130
1,380
3,330
2,080
7,500
2,910
8,240
6B
4,440
4,510
1,500
108
131
30.0
43.8
299
8.14
-------
Table 6. (Continued)
Quantity Measured (tig)
Run
Formaldehyde
Acetaldehvde
Ouinone
Acrolein
Propion-
aldebvde
MEK
Aceto-
DbeooiK
M1BK
Isoohorone
6C
4,560
5,690
1,690
110
128
27.1
45.6
293
11.7
6D
24,300
14,900
5,940
3,090
3,610
1,240
8,660
2,240
10,200
7A
18,500
10,900
3,710
1,390
2,860
1,620
6,820
2,190
7,860
7B
3,340
3,750
1,180
137
176
35.6
7.60
53.9
13.4
7C
3,910
4,470
1,290
126
170
29.6
13.4
46.9
15.6
7D
18,200
12,100
7,610
1,460
3,300
2,280
7,340
2,560
8,260
8A
25,000
15,300
6,470
2,150
3,760
1,900
8,530
3,260
9,670
8B
3,240
4,810
1,390
190
199
46.0
11.6
34.7
8.12
8C
3,350
4,570
785
167
159
30.6
11.2
34.1
9.27
8D
26,700
17,200
6,520
1,980
3,910
1,920
9,140
3,340
10,400
9A
26,800
14,200
4,860
2,060
3,890
2,360
8,950
3,580
9,880
9B
3,(SO
3,370
1,410
144
163
33.0
9.21
26.6
9.24
9C
3,140
3,590
1,550
148
183
31.*
15.8
33.9
8.91
9D
27,800
15,000
2,190
2,550
3,940
2,270
9,450
3,300
11,100
10A
24,200
13,200
900
1,710
3,590
2,090
8,330
2,760
9,860
JOB
3,240
4,090
893
166
249
65.2
21.6
38.1
166
IOC
2,860
3,600
725
162
228
70.8
15.0
36.4
8.87
10D
24.500
14.900
7-230
1,?QQ
3.680
1.930
8.960
3.240
10.500
NOTE: Final values are not corrected for the field train blank.
-------
Table 7. Spike Recovery for Field Test I (August 1994)"
Percent Recovered
Form- Proplon-
Run
aldehyde
Acetaldehvde
Ouinone
Acrolein
aldehvde
Acefoohenone
MEK
MIBK
Isonborone
1A
60.0
66.4
54.4
82.3
81.8
86.6
45.9
67.3
87.8
ID
82.1
74.7
16.8
88.6
103
106
43.2
56.4
101
2A
89.2
75.0
61.4
62.7
90.7
101
44.4
54.3
100
2D
96.3
88.4
68.1
59.7
87.0
97.7
46.2
49.3
79.0
3A
85.1
75.2
47.9
33.7
72.6
86.3
36.4
33.0
90.8
3D
109
97.7
70.5
49.4
84,9
107
30.3
38.6
104
4A
94.8
89.6
40.3
49.6
90.2
95.3
33.2
41.2
93.0
4D
109
107
53.1
39.1
88.0
105
38.3
46.3
98.8
SA
97.2
93.7
62.1
48.1
88.1
96.2
41.2
45.3
90.6
5D
95.6
91.7
114
44.9
84.7
97.9
28.2
43.8
97.2
6A
59.5
61.5
70.5
26.5
63.8
73.8
33.9
32.4
75.1
6D
76.1
58.7
40.8
56.3
62.8
77.1
18.2
21.9
83.8
7A
64.8
50.3
26.9
26.1
53.6
67.4
26.2
26.4
71.6
7D
65.0
57.5
71.6
29.6
66.4
77.3
39.8
33.2
80.2
8A
101
79.6
58.4
44.1
76.8
91.2
33.2
43.2
95.4
8D
106
94.3
64.8
40.2
79.3
95.9
33.2
43.5
100
9A
110
82.7
39.8
43.5
80.6
96.1
41.9
47.8
97.8
9D
110
84.0
7.09
52.5
78.3
97.7
38.8
42.4
106
10A
98.5
70.3
0.09
35.6
73.4
90.7
37.0
37.1
97.5
10D
102
88.4
76.7
37.9
76.2
98.1
34.2
44.1
106
-------
Table 7. (Continued)
Percent Recovered
Form- Propion-
Rnn aldehyde Acttaldehvde Ouinone Acrolein aldehyde Aeetonhtnone MEK MTBK Isoohoronc
Minimum 59.5 50.3 0.09 26.1 53.6 67.4 18.2 21.9 71.6
Maximum 110 107 114 88.6 103 106 46.2 67.3 106
Average 90S 793 52J. £7j5 79J 92.2 36.2 42.4 92.8
'Based on the analysis of aU impingere.
-------
Recovery of quinone, acrolein, MEK, and MIBK was poor, as expected. The average
recovery of the other five compounds was acceptable.
Analysis of the second fractions enabled examination of breakthrough of individual
compounds into third and fourth impingers. Any amount of compound detected in the second
fraction was classified as having broken through the first two impingers. Breakthrough for
each compound is shown in Table 8. Average breakthrough of the spiked MEK and MIBK
was over 30 percent. Average breakthrough of the acetaldehyde and propionaldehyde was
greater than 10 percent. Average measured breakthrough of the other five spiked compounds
was less than 10 percent. Except for formaldehyde and quinone, the compounds follow a
consistent trend with high breakthroughs for Runs 3, 4, 6 and 7. The high breakthroughs do
not appear to correlate to moisture levels in the source, source temperature, or sampling rate.
These results indicate that some of the compounds, especially MEK and MIBK, may be carried
beyond the fourth impinger, especially at high flow rates. Measured breakthrough in the
unspiked samples is also shown in Table 8, but many of the values have a wide margin of
error because the concentration of these compounds was close to the detection limit.
STATISTICAL ANALYSIS
Data using all impingers from all ten runs were used to generate the method validation
statistics. Two-impinger data for formaldehyde, acetaldehyde, and acetophone were also
evaluated. Before statistical analysis, all compound quantities from the analytical reports were
normalized using the gas volume sampled by each train, using the equation below:
/ V
m - m * —
V
m
where:
m' = normalized quantity;
m = measured quantity;
22
-------
Table 8. Breakthrough Analysis
Percent Breakthrough
Run
Form-
aldehyde
Acet-
aldehyde
Quinone
Acrolein
Proplon-
aldehyde
MEK
Actlo-
phenone
MIBK
Isopborone
1A
0.13
4.29
0.29
3.10
5.46
20.2
1.56
13.7
3.10
IB
1.18
5.78
2.13
2.12
3.75
9.06
3.30
2.19
0.00
1C
0.33
5.91
0.00
0.00
0.00
0.00*
0.00*
0.00*
ND
ID
0.21
4.49
0.00
2.78
4.74
34.9
1.72
24.2
2.85
2A
0.17
5.57
0.44
3.32
8.01
44.7
1.79
30.6
4.33
2B
0.24
11.8
0.22
2.54
17.4
23.8
0.00*
0.00
0.00*
2C
0.55
11.2
0.00*
0.00
20.7*
14.4'
ND
0.00*
ND
2D
0.14
4.53
1.00
3.14
6.79
31.7
1.64
22.4
6.07
3A
0.1S
22.0
0.48
7.60
24.8
76.2
4.20
59.2
5.35
3B
0.68
21.7
0.98
4.86
32.9
53.2*
0.00
15.3
0.001
3C
0.50
25.7
0.40
5.79
33.3
29.7
17.8*
ND
ND
3D
0.18
12.9
0.20
6.15
14.5
62.8
3.64
39.6*
6.47
4A
0.1S
10.0
0.26
4.36
10.1
56.6
2.27
39.1
5.18
4B
0.26
19.3
0.15
2.79
24.2
27.3
0.00*
ND
0.00*
4C ¦
0.50
24.7
0.44
7.36
25.7
40.6
3.92
ND
0.00*
4D
0.22
* 18.9
0.41
8.43
20.9
68.4
4.10
50.6*
6.60
5A
0.14
3.27
0.28
3.81
4.86
36.2
1.47
4.61
3.63
SB
0.47
10.3
0.38
6.65
13.2
34.3
ND
ND
100
5C
0.21
8.65
0.52
4.77
9.96
28.9
0.00
100,
0.00*
5D
0.13
6.55
0.06
3.88
7.43
52.0
1.83
26.9
4.33
6A
0.23
19.4
0.26
7.71
19.1
60.0
4.10
45.7
7.21
6B
0.46
33.4
0.12
8.83
41.0
48.1'
0.00
4.62
0.00
6C
0.61
49.7
0.37
10.8
48.9
48.9
0.00
6.31
0.00
6D
0.29
34.3
0.76
5.47
33.4
66.4
6.79
53.3
8.72
7A
0.16
10.3
0.60
5.80
11.7
48.2
2.77
38.0
5.45
7B
0.44
16.0
0.13
6.57
24.8
42.7
0.00*
28.7
0.00*
-------
Table 8. (Continued)
Percent Breakthrough
Form-
Acet-
Propion-
Aceto-
Run
aldehyde
aldebyde
Quinone
Acrolein
aldehyde
MEK
phenone
MIBK
Isophorone
7C
0.23
19.7
0.00
7.48
31.0
36.6
0.00"
33.5
0.00*
7D
0.19
16.8
0.40
6.93
20.0
63.5
3.47
49.8
5.31
8A
0.14
4.52
0.43
3.61
6.55
43.8
1.75
22.9
4.50
8B
0.43
10.8
0.15
4.14
12.7
33.9
0.00*
0.00
0.00*
8C
0.29
12.6
0.00
5.37
17.1
39.2
0.00*
0.00
0.00*
8D
0.20
8.89
0.34
5.54
11.7
59.8
2.45
35.8
5.46
9A
0.11
3.03
0.66
3.44
4.09
36.0
1.29
19.2
3.22
9B
0.32
12.8
0.00
5.64
16.8
34.1
0.00*
0.00
0.00*
9C
0.21
10.2
0.00
5,07
12.9
34.9
23.6*
0.00
0.00*
9D
0.12
3.88
1.52
3.10
5.60
36.1
1.58
23.3
3.65
10A
0.15
4.32
3.47
3.87
6.13
34.3
1.56
25.5
3.83
10B
0.25
11.1
0.00
5.12
14.7
21.6
6.87*
0.00
0.00
IOC
0.28
11.7
0.00"
4.63
16.9
24.7
0.00*
0.00
0.00"
I0D
0.12
3.76
0.19
2.34
5.39
41.4
1.04
18.2
3.08
Spike Average
0.17
10.1
0.60
4.72
11.6
48.7
2.55
324
4.92
Maximum
0.29
-34.3
3.47
8.43
33.4
76.2
6.79
59.2
8.72
Minimum
0.11
3.03
0.00
2.34
4.09
20.2
1.04
4.61
2.85
Unspiked Average
0.42
16.6
0.30
5.02
20.9
31.3
2.77
9.53
5.00
Maximum
1.18
49.7
2.13
10.8
48.9
53.2
23.6
100
100
Minimum
0.21
oo
0.00
0.00
0.00
0.00
0.00
0.00
0.00
ND = Component not delected in either fraction.
'Levels measured were below the calibration curve.
-------
V = sample volume; and
Vn = mean sample volume (all runs).
Normalization of the data was required because each train collected slightly different sample
volumes.
Results for the statistical analysis for each compound are shown in Table 9, Hie RSD
and bias correction factor were calculated using the EPA Method 3011 with the typographical
errors corrected as posted on the EPA bulletin board. Using the criteria of 50% maximum for
the RSD and 1.00 ± 0.30 for the bias correction factor, the method validation test was
successful for formaldehyde, acetaldehyde, propionaldehyde, and acetophenone. Quinone,
acrolein, MEK, and MIBK did not meet the bias criterion, so the method was shown to be
invalid for these four compounds. MEK and isophorone did not meet the relative standard
deviation criterion for the unspiked samples. Low levels of MEK are challenging to identify
and quantitate because low levels of other four-carbon carbonyl compounds can interfere with
the identification and quantification of MEK by HPLC. For isophorone, one of the unspiked
samples contained approximately 200 /xg of isophorone while all the other unspiked samples
contained 40 ng or less. However, when analyte concentrations in the stack effluent are very
low, the relative standard deviation criterion is unrealistic. Because the native isophorone
concentration was very low, isophorone is judged to have performed acceptably using this
method.
25
-------
Table 9. Summary of Method 301 Statistical Analysis (Field Test I, August 1994)
Parameter
Form-
aldehyde
Acet-
aldehyde
Quinone
Acrolein
Propion-
aldehyde
MEK
Accto-
phenone
MIBK
Isophorone
Statistic; Calculated from ComDPiinds Collected in fmnincers 1 thrmich 4
RSD Spiked (%)
7.36
7.18
40.0
12.1
7.20
26.1
7.94
17.2
7.94
RSD Unspiked (55)
10,2
10.6
39.7
17.3
21.0
74.3
42.5
32.2
211
Bias CF
1.11
1.26
1.84
2.00
1.25
2.55
1.08
2.22
1.08
Disposition
Passes
Passes
Fails
Fails
Passes
Fails
Passes
Fails
Fails
Statistics Calculated from Coraoounds Collected in First Two IniDincers
RSD Spiked (%)
7.32
8.15
NR
NR
NR
NR
7.79
NR
NR
RSD Unspiked (%)
9.95
10.3
NR
NR
NR
NR
43.5
NR
NR
Bias CF
1.10
1.34
NR
NR
NR
NR
l.U
NR
NR
Disposition
Passes
Fails
NR
NR
NR
NR
Passes
NR
NR
RSD = Relative Standard Deviation
CF = Correction Factor
NR = Not Reported
-------
SECTION 4.0
FIELD TESTE
Ten test runs were completed during testing at a polyester fiber manufacturing plant
during the week of April 24 through April 28, 1995. The sampling trains were each recovered
into two sample fractions.
Samples were analyzed for seven target compounds. Results were reported for the two
sample fractions from each test run; the first two impingers and all four impingers. Results
were normalized by the sample gas volumes before statistical analysis, in order to remove
variability attributable to the small differences in the volume of gas extracted from the stack
through each train. Statistical analysis was performed according to the latest revisions to EPA
Method 301.'
Details of the sampling runs and results of the laboratory and statistical analyses are
presented in the following subsections.
FIELD SAMPLING
Flue gas samples were collected from a spinning machine exhaust stack at a polyester
fiber manufacturing plant. Sampling was performed from a concrete slab roof surface,
approximately 22 meters above ground level. The sampling port was a 4-inch (102 mm)
diameter pipe nipple, 1.4 meters above the sampling platform. Figure 3 is a diagram of the
sampling location. Preliminary samples were collected from the spinning machine exhaust
duct in a pre-test site survey. Formaldehyde, acetaldehyde, and propionaldehyde were all
detected in the preliminary and validation test samples.
27
-------
Flow
60"
Hanging Pipe
(Nearest obstruction)
14' Minimum clearance to edge of slab
4" Sampling Port
81"
5614"
IliliWi
Figure 3. Diagram of Sampling Location for Second Carbonyl Field Test
-------
Ten quad train runs were completed at the test site. The quad-train probe is described
in detail in Section 5. Trains A and D were spiked and Trains B and C were unspiked.
Table 10 summarizes the sampling parameters recorded for each run. The diameter of all the
sampling nozzles was 6.30 mm. The static pressure in the stack was negative, and remained
constant at approximately -130 mm of water during all test runs.
The target sample volume for each run was 0.85 cubic meters. The sampling time was
normally 100 minutes. However, some runs were extended to allow collection of the full
0.85 cubic meters when the stack gas velocity dropped slightly.
Because of the additional liquid spiked into trains A and D, only trains B and C were
used to calculate the percentage of moisture in the stack gas. Moisture values were generally
in the range of 4-5 percent by volume. The average moisture content indicated by trains B and
C was used in subsequent calculations for trains A and D.
The stack temperature and velocity for each run were measured using a single
thermocouple and S-Type pitot tube on the sampling probe assembly. Individual stack gas
temperature and pitot tube differential pressure measurements were taken for each of the four
trains at the time the other sampling train data were recorded. This measurement scheme
resulted in some slightly different temperature and velocity data associated with individual
trains for the same run, even though measurements were made with a common probe. These
temperature and differential pressure measurement differences did not affect the test data
because the sample for all four trains was collected from the same point, the volumes collected
were recorded, and the data was corrected for the slight differences in sample volume.
The spiking system was operated to inject approximately equal quantities of spiking
solution into trains A and D during each sampling run. The dynamic spiking apparatus and
procedure are described in detail in Section 5. Table 11 shows the quantity of each compound
spiked into Trains A and D during each run. Spiked quantities were determined by weighing
the spiking syringes before and after each test run. Spike weights were recorded in a
29
-------
Table 10. Sampling Parameters, Field Test II (April 1995)
Run
Sampling
Duration
(min)
Moisture
m
Standard Metered
Volume
(dscm)
Stack
Temperature
(deg. C)
Stack Gas
Velocity
(mpm)
Percent
Isokinetic
1A
100
-
0.891
34.4
295
108.6
IB
100
5.21
0.877
34.4
293
107.8
1C
100
4.82
0.861
34.4
294
105.2
ID
100
-
0.861
34.4
294
105.3
2A
100
-
0.858
36.1
293
109.0
2B
100
5.11
0.846
36.7
293
104.6
2C
100
5.02
0.824
36.1
293
101.7
2D
100
-
0.821
36.7
293
101.4
3A
100
-
0.876
36.7
304
102.5
3B
100
4.01
0.872
37.2
305
102.0
3C
100
4.20
0.873
36.1
304
102.0
3D
100
-
0.853
37.2
305
99.8
4A
100
-
0.879
37.2
302
103.9
4B
100
4.32
0.860
37.2
302
101.6
4C
100
4.26
0.867
36.7
302
102.4
4D
100
-
0.845
37.2
302
39.8
5A
100
-
0.823
36.7
259
103.1
SB
100
4.85
0.831
36.7
259
104.5
5C
100
3.83
0.842
36.7
261
104.3
5D
100
-
0.807
36.7
261
100.3
6A
100
-
0.816
37.2
289
101.4
6B
100
4.79
0.851
16.7
289
105.8
6C
100
4.82
0.847
37.2
289
105.2
6D
100
-
0.844
37.2
289
105.0
7A
100
-
0.861
37.8
302
101.6
7B
100
4.00
0.868
37.2
302
102.4
7C
100
4.20
0.859
37.2
302
101.5
7D
100
-
0.853
37.2
302
100.6
8A
110
-
0.892
37.8
283
102.4
8B
110
4.19
0.906
37.8
283
104.1
8C
110
4.24
0.880
37.2
283
101.0
8D
110
-
0.874
37.2
283
100.4
30
-------
Table 10. (Continued)
Run
Sampliag
Duration
(min)
Moisture
(%)
Standard Metered
Volume
(dscm)
Stack
Temperature
(deg. C)
Stack Gas
Velocity
(mpm)
Percent
Isokinetic
9A
106
-
0.855
37.2
286
102.1
9B
106
5.18
0.858
37.2
285
102.S
9C
106
5.05
0.866
36.7
285
103.3
9D
106
-
0.849
37.2
286
101.4
10A
100
-
0.799
37.8
284
101.6
10B
101
4.97
0.832
37.8
284
104.9
IOC
102
4.60
0.840
37.8
283
104.5
10D
103
-
0.824
37.2
283
101.7
31
-------
Table 11. Spike Quantities
Run
Form-
aldehyde
Acet-
aldebyde
Propion-
aldehyde
Acetophenooe
(us)
Methyl
Ethyl
Ketone
(jit)
Methyl
Isobutyl
Ketone
Isophorone
(na)
1A
1621.2
6006.9
2945
6483.4
3773.4
5267.1
7375.4
ID
1707.7
6327.3
3102
6829.2
3974.6
5548
7768.7
2A
1372.6
5085.8
2493.4
5489.3
3194.8
4459.5
6244.5
2D
1426.7
5286.1
2591.6
5705.4
3320.6
463S
6490.3
3A
1750.9
6487.5
3180.5
7002.1
4075.3
5688.5
7965.4
3D
843
3123.6
1531.4
3371.4
1962.2
2738.9
3835.2
4A
1329.4
4925.7
2414.9
5316.4
3094.2
4319
6047.8
4D
1242.9
4605.3
2257.8
4970.6
2892.9
4038.1
5654.4
5A
1405
5206
2252.3
5619
3270.3
4564.8
6392
5D
1513.1
5606.4
2748.6
6051.2
3521.8
4916
6883.7
6A
1437.5
5326.1
2611.2
5748.7
3345.7
4670.2
6539.5
6D
1351
5005.8
2454.1
5402.9
3144.5
4389.2
6146.1
7A
1437.5
5326.1
2611.2
5748.7
3345.7
4670.2
6539.5
7D
1405
5206
2552.3
5619
3270.3
4564.8
6392
8A
1523.9
5646.5
2738.3
6094.4
3547
4951.1
6932.8
8D
1599.6
5926.8
2905.7
6397
3723.1
5196.9
7277
9A
1448.3
5366.2
2630.8
5791.9
3370.9
4705.3
6588.6
9D
1491.5
5526,3
2709.4
5964.8
3471.5
4845.7
6785.3
10A
1351
5005.8
2454.1
54Q2J
3144.5
4389.2
6146.1
10D
702.5
2603
1276.1
2809.5
1635.1
2282.4
3196
32
-------
field notebook. Review of the spiking data indicated that there may have been a spiking enor
associated with runs 3 and 10. Recorded weights from both of these runs show a discrepancy
between the amounts spiked in the A and D trains.
ANALYSIS
The samples from each train were collected and analyzed in two fractions. The first
fraction contained the probe rinse and contents of the first two impingers. The second fraction
contained the contents of the third and fourth impingers. Table 12 shows the results of the
analysis of the first fractions from each run. This sample is the fraction intended for analysis
using Method 0011.2 Acetaldehyde and formaldehyde were present in the unspiked samples,
along with trace amounts of propionaldehyde.
Table 13 shows the percentage of each spiked compound recovered in the first two
impingers. The recovery is calculated as follows:
R = 100% x
cs
where:
R = percent recovery;
S = measured quantity in the spiked sample;
M = mean value of the unspiked samples in the run; and
CS = calculated spike quantity.
Recovery of methyl ethyl ketone and methyl isobutyl ketone was poor, as expected. The
average recovery levels of the other five compounds were acceptable. The recovery level of
all compounds calculated for runs 3 and 10 are inconsistent with the values calculated for the
other runs. These are the same two runs for which a spiking error is suspected. These two
runs, therefore, were eliminated from subsequent statistical analysis for method validation.
33
-------
Table 12. Analytical Results, Impingers 1 and 2
Run
Form-
aldehyde
Acet-
aldehyde
Propion-
oldehyde
tog)
Aceto-
phenooe
Methyl
Ethyl
Ketone
(xr)
Methyl
Isobutyl
Ketone
Odd
Isophorone
0»8>
1A
1656.9
6343.4
2730.8
6791.2
1478.4
1068*
9128.6
IB
lS.Slb
463.2
SSTf4
<4.40
<1.27
<1.39
<3.01
1C
12.02*
433
SAT*
<4.40
<1.27
<1.39
<3.01
ID
1592.2
5412.2
2222
5690.1
1284.7
1055*
7680.2
2A
1149.6
5077.3
1886.8
4653.6
1570.4
1136*
6552.8
2B
19.8b
836.8
9.84^
<8.80
<2.55
<2.79
<6.02
2C
16.43b
834.3
5.21w
<4.40
<1.27
<1.39
<3.01
2D
1168.6
4971.1
1854.2
4872.1
1089.9
'837*
6385.5
3A
970.7
4361
1662.3
4291.2
1120.8
336*
5486.2
3B
19,63b
849.7
4.52""
<8,80
<2.55
<2.79
<6.02
3C
13,14b
739.7
2.96"4
<4.40
<1.27
<1.39
<3.01
3D
1398.5
5439.1
2051.3
5635.3
1093
623'
7341
4A
1532.3
6453.5
1736.1
5839.6
1423.9
964*
7693.5
4B
10.5b
449.4
<1.12
<4.40
<1.27
<1.39
<3.01
4C
22.7b
839.7
3.12M
<4.40
<1.27
<1.39
<3.01
4D
1280.7
5025.1
1916.3
4854.9
1341.5
956*
6305.6
SA
1401.2
5670.9
2133.9
5502.5
1474
1170*
7382.6
5B
20.32"
908.9
<1.12
<4.40
<1.27
<1.39
<3.01
5C
27.4k
898.9
<2.24
<8.80
<2.55
<2.79
<6.02
5D
1218
4583.6
2031.6
5263.4
1440.8
1691*
7040.6
6A
1232.8
4801.9
2009.5
5009.5
1293.8
812*
6751.4
6B
18.55"
1013.4
<1.12
<4.40
<1.27
<1.39
<3.01
6C
I6,29b
860.7
3.01w
<4.40
<1.27
<1.39
<3.01
6D
1379.3
5176.8
2122
5552.7
1136.3
764*
7329.7
7A
1335.3
5061.7
2016.1
5310.9
1267.4
977*
6868.5
7B
19.9b
1003.3
3,77w
<4.40
<1.27
<1.39
<3.01
7C
17.99"
982
3.33w
<4.40
<1.27
<1.39
<3.01
7D
1249.8
2544.1
2019.9
5022.3
1327.6
1041*
6665.1
8A
1205.9
5361.1
1967.8
5246
1478.1
1181*
7171.3
SB
16.52"
1060.8
3.87m
<4.40
<1.27
<1.39
<3.01
8C
17.97b
973.4
3.44w
<4.40
<1.27
<1.39
<3.01
8D
1296.2
S690.1
1990.1
5143.4
1483.2
1339*
7081.9
34
-------
Table 12. (Continued)
Run
Form-
aldehyde
G»g)
Acet-
aldehyde
0«R)
Propion-
aldehyde
0»t0
Aceto-
phenone
Gig)
Methyl
Ethyl
Ketone
0.R)
Methyl
Isobutyl
Ketone
0»R)
Isophorone
G»r)
9A
1286.4
4937.1
1859.7
4860.9
1646.2
1191*
6438.3
9B
17.48b
987.4
2.5W
<4.40
<1.27
<1.39
<3.01
9C
18.04b
967.8
2.79b'4
<4.40
<1.27
<1.39
<3.01
9D
1459.7
S830.1
2141.9
6048.5
1344.4
1170*
7355.9
10A
1312.1
1520
1766.27
4990
1821.6
1448'
6919.5'
10B
17.22b
840.7
2.21"
<4.40
<1.27
<1.39
• <3.01
IOC
16.24b
825.3
2.47m
<4.40
<1.27
<1.39
<3.01
10D
112 969
4423.37
1510.8
4058.4
1503
1106'
Lh
00
o
00
NOTE: Final values are not corrected for the field train blank.
'Method spike recoveries outside acceptable range.
bLess than 10 times field train blank.
'Calibration check standard outside acceptable range.
dBelow calibration curve.
'Above calibration curve.
35
-------
Table 13. Spike Recovery
Methyl Methyl
Form- Propion- Etfiyl Isobutyl
aldehyde Acetaldebyde aldehyde Acetopheoone Ketone Ketone Isopbcrose
Run
(%)
m
(%>
(%)
m
1A
101
98
93
105
39
20*
124
ID
92
78
71
83
32
19*
99
2A
82
83
75
85
49
25*
105
2D
81
78
71
85
33
18*
98
3A
55
55
52
61
28
6*
69
3D
164
149
134
167
56
23*
191
4A
114
118
72
no
46
22'
127
4D
102
95
85
98
46
24*
112
5A
98
92
95
98
45
26*
115
5D
79
66
74
87
41
34*
102
6A
85
73
77
87
39
17*
103
6D
101
85
86
103
36
17*
119
7A
92
76
77
92
38
21*
105
7D
88
30
79
89
41
23*
104
8A
78
77
72
86
42
24*
103
8D
80
79
68
80
40
26'
97
9A
88
74
71
84
49
25*
98
9D
97
88
79
101
39
24*
108
10A
96
98
72
92
58
33*
113"
10D
158
138
118
144
92
48*
182"
Maximum*
114
118
95
110
49
34
127
Minimum*
78
30
68
80
32
17
97
Average*
91
81
78
92
41
23
108
~Does not Include Runs 3 and 10 (see text).
•Method spike recoveries outside acceptable range.
"Calibration check standard outside acceptable range.
36
-------
Table 14 shows analytical results for both fractions (all impingers) of each sampling
train combined. Analysis of the second fractions enabled examination of breakthrough of
individual compounds into third and fourth impingers. Breakthrough for each compound is
shown in Table 15. Breakthrough of the spiked MEK and MIBK was over 20 percent.
Breakthrough of all other spiked compounds in the spiked samples was less than 10 percent
Measured breakthrough in the unspiked samples is also shown in Table 15, but values for
formaldehyde and especially propionaldehyde have a wide margin of error since the
concentration of these compounds was close to the detection limit.
STATISTICAL ANALYSIS
Data from eight of the ten runs were used to generate the method validation statistics.
Runs 3 and 10 were eliminated from the data set because of suspected spiking errors. Before
statistical analysis, all compound quantities from the analytical reports were normalized using
the gas volume sampled by each train. This was done using the equation
m' = m * —
vro
where:
m' = normalized quantity;
m = measured quantity;
V = sample volume; and
Vm = mean sample volume (all runs).
37
-------
Table 14. Analytical Results, All Fractions
Run
Form-
aldehyde
ton)
Acet-aldebyde
Proploii-
aldehyde
(HR>
Aceio-
phenone
0«8>
Methyl
Ethyl
Ketone
test)
Methyl
Isobutyl
Ketone
(fiti
Isophorone
(ffO
1A
1656.9
6639.2
2866.1
6969.2
2245.5
1875.2"
9445.2
IB
17.56*
506.8
10.25*
ND
ND
ND
ND
1C
14.48*
471.6
10.53*
ND
ND
ND
ND
ID
1630.9
5673.8
2359.25
5878.5
1980.1
1699.2*
7927.9
2A
1183.7
5215.5
1960.2
4782.3
1931.7
1378.5"
6718.5
2B
22.36*
886.7
14.37*
ND
ND
ND
ND
2C
17.62*
903.6
10.16*
ND
ND
ND
ND
2D
1168.6
5119
1913.3
4985.5
1386.6
1047.1"
6630.9
3A
970.7
4620.1
1791.2
4423.4
1583.2
723.6"
5679.1
3B
22.14*
906.9
4.52*
ND
ND
ND
ND
3C
17.71*
805
4.9*
ND
ND
ND
ND
3D
1398.5
5690.8
2176.4
5814.6
1610.9
1164.2"
7820.6
4A
1557.8
6669.7
1818.3
5983.9
1807.1
1246,5"
7986.1
4B
12.93*
503.6
ND
ND
ND
ND
ND
4C
25.03*
910.2
3.12*
ND
ND
ND
ND
4D
1280.7
5163.3
2020.1
4854.9
1692
1198.7"
6471.2
5A
1423.7
5845.8
2205.9
5625.9
1946.3
1576.8"
7625
5B
23.01*
967.7
ND
ND
ND
ND
ND
5C
31.52*
952.9
ND
ND
ND
ND
ND
5D
1218
4673.3
2141.9
5263.4
1794.4
1981.8"
7040.6
6A
1232.8
4956.7
2110.3
5137.8
1675.2
1080.1"
6751.4
6B
21.62*
1077.6
ND
ND
ND
ND
ND
6C
19.39*
930.3
3.01*
ND
ND
ND
ND
6D
1379.3
5361.4
2210.1
5662
1448.9
968.2"
7488.9
7A
1356.5
5243.6
2134.1
5452.4
1692.3
1336.4"
7103.1
7B
22.5*
1060.5
3.77*
ND
'ND
ND
ND
7C
20.24*
1049.7
5.27*
ND
ND
ND
ND
7D
1249.8
2691.9
2118.7
5147.3
1674.4
1313.4"
6812.4
8A
1227.2
5505.7
2063.2
5363
1794.3
1444.5"
7415.8
SB
18.79*
1116.3
5.86*
ND
ND
ND
ND
8C
20.29*
1039.1
5.42*
ND
ND
ND
ND
8D
1321.6
5841.4
5052.7
5280.4
1874.9
1692.6"
7242
38
-------
Table 14. (Continued)
Run
Form-
aldehyde
(mr)
Acet-aldehyde
0itO
Propion-
aldebyde
(mr)
Aceto-
phenone
(mr)
Methyl
Ethyl
Ketone
(MR)
Methyl
Isobutyl
Ketone
(MR)
¦ Isopborone
(MR)
9A
1309.1
5120.6
1970
5028.1
2040
1519.2*
671X6
9B
19.76*
1036.1
. 4.13*
ND
ND
ND
ND
9C
20.2*
1025.3
5.97*
ND
ND
ND
ND
m
1486.3
5935.6
2259.6
6190.6
1572.7
1394.3'
7594*
10A
1312.1
5844.27
1570
5102.4
2163
1731.5*
7089.7*
10B
19.04'
882.9
4.27*
ND
ND
ND
ND
10C
19.26'
865.8
4.04'
ND
ND
ND
ND
10D
115?.?
4560.8
1582.3
1860.7
1371,t
NOTE: Final values are not corrected for the field train blank.
ND «= Not Detected
'Less than 10 tunes field train blank.
'Method spike recoveries outside acceptable range.
'Calibration check standard outside acceptable range.
39
-------
Table 15. Breakthrough Analysis
Run
Form-
aldehyde
(%)
Acet-
aldebyde
1%)
Propion-
aldehyde
(%)
Aceto-
phenone
(%)
Methyl
Elliyl
Ketone
m
Methyl
Isobutyl
Ketone
<%)
Isopbor
m
2A
0.0
4.5
4.7
2.6
34.2
43.1
3.4
IB
11.7
8.6
41.8
ND
ND
ND
ND
1C
17.0
8.2
48.1
ND
ND
ND
ND
ID
2.4
4.6
5.8
3.2
35.1
37.9
3.1
2A
2.9
2.7
3.7
2.7
18.7
17.6
2.5
2B
11.4
5.6
31.5
ND
ND
ND
ND
2C
6.8
7.7
48.7
ND
ND
ND
ND
2D
0.0
2.9
3.1
2.3
21.4
20.1
3.7
3A
0.0
5.6
7.2
3.0
29.2
53.5
3.4
3B
11.3
6.3
0.0
ND
ND
ND
ND
3C
25.8
8.2
39.6
ND
ND
ND
ND
3D
0.0
4.4
5.7
3.1
32.2
46.5
6.1
4A
1.6
3.2
4.5
2.4
21.2
22.7
3.7
4B
18.8
10.8
ND
ND
ND
ND
ND
4C
9.3
7.7
0.0
ND
ND
ND
ND
4D
0.0
2.7
5.1
0.0
20.7
20.3
2.6
5A
1.6
3.0
3.3
2.2
24.3
25.8
3.2
SB
11.7
6.1
ND
ND
ND
ND
ND
5C
13.1
5.7
ND
ND
ND
ND
ND
5D
0.0
1.9
5.2
0.0
19.7
14.7
0.0
6A
0.0
3.1
4.8
2.5
22.8
24.8
0.0
6B
14.2
6.0
ND
ND
ND
ND
ND
6C
16.0
7.5
0.0
ND
ND
ND
ND
6D
0.0
3.4
4.0
1.9
21.6
21.1
2.1
7A
1.6
3.5
5.5
2.6
25.1
26.9
3.3
7B
11.6
5.4
0.0
ND
ND
ND
ND
7C
11.1
6.4
36.8
ND
ND
ND
ND
7D
0.0
5.5
4.7
2.4
20.7
20.8
2.2
8A
1.7
2.6
4.6
2.2
17.6
18.3
3.3
8B
12.1
5.0
34.0
ND
ND
ND
ND
40
-------
Table 15. (Continued)
Run
Form-
aldehyde
(%)
Acet-
aldehyde
(%)
Propion-
aldehyde
Aceto
phenone
(%)
Methyl
Ethyl
Ketone
(%)
Methyl
Isobutyl
Ketone
(%)
Isophor
m
8C
11.4
6.3
36.5
ND
ND
ND
ND
8D
1.9
2.6
60.6
2.6
20.9
20."
2.2
9A
1.7
3.6
5.6
3.3
19.3
21.6
4.1
9B
11.5
4.7
39.5
ND
ND
ND
ND
9C
10.7
5.6
53.3
ND
ND
ND
ND
9D
1.8
1.8
5.2
2.3
14.5
16.1
3.1
10A
0.0
2.0
3.1
2.2
15.8
16.4
2.4
10B
9.6
4.8
48.2
ND
ND
ND
ND
10C
15.7
4.7
38.9
ND
ND
ND
ND
10D
2.0
3.0
4.5
2.5
19.2
19.3
2.3
Average Spiked
1.1
3.2
8.2
2.2
22.4
23.3
2.6
Maximum
2.9
5.6
60.6
3.3
35.1
53.5
6.1
Minimum
0.0
1.8
3.1
0.0
14.5
14.7
0.0
Average Unspiked
12.4
6.7
23.1
ND
ND
ND
ND
Maximum
25.8
10.8
53.5
ND
ND
ND
ND
Minimum
6.8
4.7
0.0
ND
ND
ND
ND
Averages, maximums, and minimum*; do not include Runs 3 and 10 (see text).
ND = Not Delected
41
-------
Results of the statistical analysis for each compound collected in the first two impingers are
shown in Table 16. Statistical analysis results for Impingers 1 through 4 are not reported
because they did not significantly differ from the results with two impingers. The RSD and
bias correction factor were calculated using the EPA Method 3011 with the typographical
errors corrected as posted on the EPA bulletin board. Using the criteria of 50% maximum for
the RSD and 1.00 ± 0.30 for the bias correction factor, the method validation test was
successful for formaldehyde, acetaldehyde, propionaldehyde, acetophenone, and isophorone.
Collection of MEK and MIBK did not meet the bias criterion, and therefore the method was
not shown to be valid for these two compounds.
Table 16, Statistical Analysis Using First Two Impingers
Parameter
Form-
aldehyde
Acet-
aldehyde
Propion-
aldehyde
Aceto-
phenone
Methyl
Ethyl
Ketone
Methyl
Isobutyl
Ketone*
Isophorone
RSD Spiked
(%)
8.8
16.7
12.94 •
10.43
18.75
21.17
8.99
RSD
Unspiked
(%)
20.71*
12.35
48.54*
Bias CF
1.1
1.24
1.29
1.09
2.45
4.33
0.93
Disposition
Pass
Pass
Pass
Pass
Fail
Fail
Pass
'Method spike recoveries were outside acceptable range.
''Measured amounts were less than 10 times the field train blank.
42
-------
SECTION 5.0
FIELD TEST PROCEDURES
The purpose of the sampling programs was to evaluate the proposed aldehyde and
ketone sampling and analytical methods and to determine the performance (precision and
accuracy) of the methods. Replicate, independent flue gas samples were collected
simultaneously from an aldehyde/ketone emission source to determine precision. For bias
determination, known concentrations of aldehydes and ketones were dynamically spiked only
into Trains A and D. Various blank samples were collected and analyzed to identify sources
of contamination in the method.
Both field tests consisted of 10 quadruplicate sampling runs. Each test run used four
independent sampling trains to collect four samples from essentially the same location during
each test run.
The nozzle and probe rinse and the contents of the first two impingers comprised the
first of two samples collected from each sampling train. The contents of the third and fourth
impingers made up the second sample collected from each train. Samples were processed and
analyzed at Radian's PPK laboratory following procedures detailed in this section of this
document. Both samples collected from each of the four trains were analyzed to determine
carryover into the third impinger.
43
-------
Sample collection during both field validation field tests was performed using
procedures similar to those detailed in SW-846 Method 0011,2 "Sampling for Formaldehyde
Emissions from Stationary Sources." The method that was evaluated was modified based on
the results of the laboratory studies (reported in Appendix A) and to enable information on
compound breakthrough to be collected. This sampling method is a modification of the EPA
stationary source test Method 5. Gas was extracted isoMneticaUy from the source through a
heated glass nozzle and probe system as shown earlier in Figure 1. The gas was passed
through a five-bottle impinger train, a sample pump, a dry gas meter and an orifice differential
pressure meter. The following modifications were made to the SW-846 Method 00112 for the
aldehyde and ketone sampling method validation tests:
• Four co-located sampling trains were used per Method 3011 to allow
determination of precision and bias of the proposed sampling and analytical
method.
• The trains were dynamically spiked with a solution of aldehydes and ketones.
• The first impinger contained 200 mL of DNPH reagent to increase the sample
capacity.
• The second impinger contained 100 mL of DNPH.
• A third reagent impinger containing 100 mL of DNPH was added to the train
between the second reagent impinger and the empty impinger to enable
compound breakthrough to be determined.
EQUIPMENT
Ecafce
A special probe assembly was required to allow simultaneous sampling at essentially
the same point with four independent sampling trains. Proposed Method 3011 describes field
evaluation procedures and details the criteria for the quadruple sampling probe tip
arrangement. The quad-probe arrangement is designed to minimize velocity variations at the
nozzles of the four sampling probes.4
44
-------
Figures 4 and 5 illustrate the configuration of the sampling probe used during the
aldehyde and ketone test program. Note that the probe inlets are in the same plane
perpendicular to the gas stream, allowing the probe tip openings to be exposed to the same gas
conditions.
EPA Method 301 specifies that the inside edge of sampling probe tips shall be situated
in a 6.0 cm x 6.0 cm square area, and that the area encompassed by the probe tip arrangement
should occupy less than 5% of the stack cross-sectioned area. If this criterion is met, then the
flow at each of the four probe tips can be considered similar. Radian used a probe tip
assembly with a cross-sectional area of 19 square centimeters as measured from the
probe/nozzle centerlines. The criterion that the probe tip area not exceed 5% of the stack area
was met at both field test sites.
Sampling Trains
Four independent impinger trains comprised the quad-train assembly. Each train used
five glass impingers. Each train had its own meter box and pump. The trains were designated
"A," "B," -C," and "D." Spiking compounds were dynamically added to trains A and D in
the field for bias determination.
Dynamic Spiking Apparatus
Spiked compounds were introduced to the sampling system in gaseous form using
liquid syringe injection through a heated glass elbow mounted at the outlet of the probe as
shown in Figure 6. The Teflon® line from the syringe pump was connected to a piece of
glass-lined stainless steel tubing with a beveled tip. The liquid spike was maintained as a
droplet at the tip of the glass-lined stainless steel tubing, from which point the spike volatilized
and became a gaseous spike as it entered the heated gas stream. The spiking liquid was not
allowed to drip into the sampling line. Liquid feed rates of the spiking solution were metered
45
-------
Upstream View (bottom)
Upstream View (front) •ni*g*d for daun
Figure 4. Quad-Train Probe and Pitot Arrangement
46
S
n
3
5
-------
Drawing not to accte. For doty, prob«dataJlia«xagg*rttt0d.
Figure 5. Upper and Lower Sampling Probes (Side View)
5
-------
From
Probe
1/16* Glass Lined
Stainless Steel
Needle
Motor Driven Sydnga Pump
Figure 6. Dynamic Spiking Apparatus
48
-------
by means of motor driven syringe pumps. The quantity of liquid spiked was measured
gravimetrically by recording the syringe weights before and after each test run.
PREPARATION
Glassware Preparation
All glassware used for sampling, including the probe, impingers, all sample bottles,
and all utensils used during sample recovery, was thoroughly cleaned prior to use. All
glassware was washed with hot soapy water, rinsed with hot tap water, rinsed with distilled
water, and dried. The glassware was triple rinsed with methanol followed by triple rinsing
with methylene chloride (MeClj). No acetone was used in glassware preparation.
Reagent bottles used for the storage of DNPH derivatizing solution were rinsed with
acetonitrile and dried before use.
DNPH Preparation
The DNPH reagent was prepared and purified within five days of sampling. The
reagent was prepared at Radian's Perimeter Park (PPK) laboratory in North Carolina using the
procedure described in the test plan. Each reagent container was properly labeled, tightly
capped, and sealed with Teflon® tape. The reagent was delivered directly to the test locations
via Radian vehicle. Once a container of prepared DNPH was opened in the field, the contents
were used within 48 hours to minimize the possibility of the reagent becoming contaminated
from the ambient air.
Method 00112 Equipment Preparation
Reference calibration procedures were followed when available for all the train
equipment, including meterboxes, nozzles, pitot tubes, and thermocouples. The results were
49
-------
properly documented and retained. A discussion of the techniques used to calibrate this
equipment is presented in Section 7 of this document.
SAMPLING OPERATIONS
Flue gas samples were collected isokinetically from a single sampling point identified
from a preliminary velocity traverse. Preliminary information obtained during the pre-site
survey was used for selecting the proper nozzle size. Prior to testing, a leak check of pitot
lines was performed according to EPA Method 2.5
Preparation of Sampling Trains
Impingers for the four sampling trains were filled and assembled in the recovery trailer.
The impinger buckets were clearly marked as Train A, B, C, or D. All impingers used were
tared to obtain the initial weight. Approximately 200 mL of purified DNPH reagent were
transferred into the first impinger of each train, and 100 mL of reagent were added to the
second and third impingers. The fourth impingers remained empty, and 200 to 300 grams (g)
of silica gel were placed in the fifth impingers. Openings were covered with Teflon® film or
aluminum foil.
Final assembly of the sampling trains took place at the sampling location, as shown in
Figure 1. Thermocouples were attached to measure the stack, probe outlet, and impinger
outlet temperatures. Crushed ice was added to each impinger bucket, and the probe heaters
were turned on and allowed to stabilize at 120 ±14°C (248°F: ±25°F).
The sampling trains were leak checked before and after sampling. To leak check the
assembled train, the nozzle end was capped off and a vacuum was pulled in the system. With
the system evacuated, the volume of gas flowing through the system was timed for 60 seconds.
50
-------
The leak rate is required to be less than 0.5,66 L/min (0.02 aefm), or 4% of the average
sampling rate, whichever is less. After the leak rate was determined, the cap was slowly
removed from the nozzle end until the pressure equalized, and then the pump was turned off.
The leak rates and sampling start and stop times were recorded on the sampling task
log. Also, any other events that occurred during sampling were recorded on the task log (such
as pitot cleaning, thermocouple malfunctions, heater malfunctions, and any other unusual
occurrences).
A checklist for aldehyde/ketone sampling is included in Appendix B-l. Sampling train
data were recorded every five minutes on standard data forms. Actual data forms are provided
in Appendix B-2. With the single-pitot arrangement used in the quad-test, the pitot tube was
connected to only one of the four DGM boxes (Box A).
Simple Recovery
Recovery of the sampling trains is summarized in Table 17. The sample bottles
containing the probe and nozzle washings and each of the sampling trains were moved to the
recovery trailer. Each impinger was carefully removed from the impinger bucket, the outside
was wiped dry, and the final impinger weight was measured and recorded. The
aldehyde/ketone sample was collected in the following fractions:
• First and second impinger contents, water and MeCl2 rinses from the
nozzle/probe liner and the first and second impingers; and
* Contents and MeCI2/water rinses from the third and fourth impingers.
No methanol or acetone was used in the field.
51
-------
Table 17. Sample Recovery Scheme
Probe
Nozzle Extension
Rinse with D! H20 and
Brush
Rinse with MeCl2 and
Brush
Collect Contents into
Sample Container
First and Second
DNPH Impingers
Weigh for Moisture
Gain
Empty Contents into
Sample Container
Rinse with DI H20
Rinse with MeCl2
Combine Contents
with Probe Rinse
Third and Fourth
Impingers
Weigh for Moisture
Gain
Empty Contents into
Sample Container
Rinse with Di H30
Rinse with MeCl2
Collect Contents
into Sample
Container
Silica Gel Impingers
Weight for Moisture
Gain
Inspect and Discard
if Spent
Fraction 1 Fraction 2
52
-------
Container 1 - Probe Rinse, First and Second Impinger Contents-
The contents of each of the first two impingers and first two impinger connectors were
included with the probe/nozzle rinse solution. A small portion of MeCl2 was used to rinse the
impingers and connectors three times. Exposed glassware surfaces were brushed to ensure
recovery of fine particulate matter. A final rinse of the impinger and Teflon® brush with
MeClj was also necessary as the two-phase DNPH/MeCl2 mixture does not pour well, and a
significant amount of impinger catch was left on the impinger wall.
Container 2 - Third and Fourth Impinger Contents-
The contents of the third and fourth impingers of each train were recovered in the same
manner as described in Table 17. The contents of these impingers were analyzed separately
from the contents collected in the first and second impingers to check for breakthrough. Care
was taken to avoid physical carryover from the first and second impingers to the third and
fourth.
Field Train Blankfs)
Two sets of field train blanks were prepared. A sampling train was assembled in the
staging area, taken to the sampling location, and leak-checked before and after the test period.
No gaseous sample passed through the sampling train. The blank sampling trains were
recovered into two containers in the same manner as the other trains. These samples were
returned to the laboratory, processed, and analyzed with the flue gas samples collected.
Field Reagent Blanks
Aliquots of each lot of DNPH, MeCl2, and deionized water were collected for analysis
as field reagent blanks. These samples were returned to the laboratory, processed, and
analyzed with the flue gas samples collected.
53
-------
Sample Storage and Shinning
Sample containers were checked to ensure that complete labels were affixed. The
labels identified Trains A, B, C, or D as appropriate. Teflon®-lined lids were tightened and
secured with Teflon® tape. The sample bottles were stored in a cooler on ice, and returned to
the Radian PPK laboratory.
ANALYTICAL PROCEDURES
All analyses were performed at Radian's PPK laboratory. This section describes the
procedures that were used.
Sample Preparation
The samples were received in the laboratory in screw-capped glass bottles with
Teflon®-lined caps, and stored in coolers on ice. Samples were extract within 12 days of
collection and analyzed within 30 days of extraction. Actual times between sample collection
and extraction are provided in Table IE.
All lab ware was washed with detergent and tap water and rinsed with organic-free
water, followed by a methanol and methylene chloride solvent rinse prior to use. Because
acetone is an analytical interferant, glassware was not rinsed with acetone, and care was taken
to minimize acetone contamination. Methanol and methylene chloride used were HPLC grade
or equivalent.
54
-------
Table 18. Hold Time Between Sample Collection and Sample Extraction
Field Test
Samples
Hold Time
(Days)
1
MS 3,4, 8 and 9; MB 3, 4, 8 and 9; Run 4; Run 5
Train A
1
1
Runs 1, 6 and 7; Run 5 Trains B, C, and D; MS 5-7;
MB 5-7; Run 8 Trains A and B, FTB A
2
1
MS 1 and 10; MB 1 and 10; Run 8 Trains C and D;
Run 9
3
2
MB 7 and 8; MS/MSD 7 and 8
1
Runs 2 and 10; MS 2; MB 2; Run 3 Trains A and B;
FTB B ; FRB 2
4
2
MB land 2; MS/MSD land 2
1
Run 3 Trains C and D; FRB 1
5
2
MB 3 and 4; MS/MSD 3 and 4
2
Runs 1-3; Run 4 Trains A and B; Run 4 Train C P/I;
FTB B ; MB 5, 6 and 9; MS/MSD 5, 6 and 9
6
2
Run 4 Train CI/K; Run 4 Train D; Runs 5-7, and 9;
MB 10; MS/MSD 10
7
2
Runs 8 and 10; FTB A
8
2
MeC!2 Bl; DNPH B12; H20 B12
11
2
DNPH Bl 1: H,0 Bi 1
12
MS/MSD = Method Spike/Method Spike Duplicate
MB = Method Blank
FTB = Field Train Blank
FRB = Held Reagent Blank
55
-------
Extraction
The samples were extracted into methylene chloride using separatory funnels. The
separately funnel was shaken for at least three minutes. Three separatory funnel extractions
were performed. The methylene chloride extracts were added to a volumetric flask (100,250,
or 500 mL), which was then filled to the line with methylene chloride. The organic extract
was then transferred to a bottle for storage at 4°C.
Solvent Exchange
The samples were solvent exchanged into acetonitrile before HPLC analysis. Table 19
summarizes the solvent exchange ratios used for the samples. To solvent exchange the
samples, an aliquot of the methylene chloride extract was evaporated to near dryness at room
temperature under a stream of pure nitrogen. Eight milliliters of acetonitrile was added when
the sample just reached dryness. For some of the train samples, a 1:5 solvent exchange was
performed by transferring a 1 mL aliquot of the methylene chloride extract into a graduated
test tube, evaporating the solvent until only 0.5 mL remained, and bringing the solvent volume
back up to 8 mL with acetonitrile. This step was repeated a second time, followed by a third
evaporation step. The solvent volume was brought up to a final volume of 5 mL. For most of
the samples and all of the blanks, a 15:4 solvent exchange was performed by transferring a
15 mL aliquot of the methylene chloride extract to a graduated test tube and following the
sample procedures as for the spiked samples, except that the final solvent volume was brought
up to 4 mL. The exchanged samples were transferred to vials with Teflon®-lined screw caps
and stored at 4°C until analysis.
56
-------
Table 19. Solvent Exchange and Dilution Procedures
Samples
Solvent Exchange Ratio
Field Test 1
Run 1 Trains A and B; Run 2 Trains A, B, and D; Runs 3-
10; MB 1-10; FTB A and B; FRBs
Field Test 2
Runs 1-10 Trains B and C; FTB A and B; FRBs; MB 1-10
15 mL MeCl2:4 mL ACN
Field Test 1
Run 1 Trains C and D; Run 2 Train C
Field Test 2
Run 1, 4 and 10 Trains A and D; Runs 2, 5, 7 and 8 Trains
A and D I/K; Runs 3, 6 and 9 Train A I/K; Runs 3, 6 and 9
Train D;
1 mL MeCl2:5 mL ACN
Field Test 1
MS 1-10
Field Test 2
MS/MSD 1-7
1 mL MeCJ2:1 mL ACN
Field Test 2
Runs 2, 5, 7 and 8 Trains A and D P/I; Runs 3, 6 and 9
Train A P/I;
2 mL MeCk5 mL ACN
Field Test 2
MS/MSD 8-10
1 mL MeCI2:2 mL ACN
MS/MSD = Method Spike/Method Spike Duplicate
MB = Method Blank
FTB = Field Train Blank
FRBs = Field Reagent Blanks
P/I = Fraction 1 (Probe Rinse and First Two Impinger Contents)
I/K = Fraction 2 (Third and Fourth Impinger Contents)
MeCl2 = Methylene Chloride
ACN = Acetonitrile
57
-------
Qirnmatograpbtc Analyses
Standard Preparation—
A multicomponent stock aldehyde derivative standard was prepared at a concentration
of 200 ngtfih by weighing 40 mg (± 0.01 mg) of purified derivatized aldehyde crystals into
small vials, dissolving the crystals in acetonitrile, quantitatively transferring the solutions to a
200-mL volumetric flask and diluting to the line with acetonitrile. This stock solution was
aliquoted into 1-mL glass ampules, sealed and stored at 0°C.
Calibration standards were prepared by diluting 12.5, 25, 150, 300, and 500 pL of the
multicomponent stock solution to 5 mL with acetonitrile to provide a standard curve with
calibration points at 0.5, 1.0, 6, 12, and 20 ng/^L of derivative.
A check standard was prepared at 5 ng/juL of derivative by taking 125 /iL of the
200 ngffiL multicomponent stock standard and diluting to 5 mL with acetonitrile. The check
standard was used to check the instrument response and the calibration curve.
The HPLC system operating parameters for analysis of standards and samples were as
follows:
Instrument: Varian 5000 LC with autosampler
Data System: Nelson 2600 or Turbochrome
Column: Zorbax ODS (4.6 mm ID x 25 cm), or equivalent with pellicular
Flow Rate:
Detector:
Mobile Phase:
Gradient:
ODS (2 mm ID x 2 cm) guard column, or equivalent
Acetonitrile/Water/Methanol
Table 20
Perldn Elmer LC 95, ultraviolet at 360 nm
0.9 mL/min
Injection Volume: 25 t*L
Retention Time: See Table 21
58
-------
Table 20. HPLC Gradient for Analysis of DNPH-Derivatized Aldehydes
Time
Acetonltrile
Water
Methanol
(min)
(%)
(%)
(%)
0
20
40
40
12
5
25
70
18
5
23
72
28
10
15
75
35
10
15
75
37
20
40
40
47
20
40
40
Table 21. Retention Times of Aldehyde Derivatives
Component
Retention Time
(min)
Formaldehyde
8.38
Acetaldehyde
11.48
Quinone
13.86
Acrolein
15.08
Propionaldehyde
16.41
Methyl ethyl ketone
21.40
Acetophenone
28.99
Methyl isobutyl ketone
30.51
IsoDhorone
38.22
Instrument Calibration-
Calibration standards were prepared at five levels as described earlier. Each calibration
standard was injected in duplicate. Linear regression analysis of peak area response vs.
concentrations of derivatized aldehyde or ketone was used to prepare a calibration curve, and
the linearity was confirmed by visual inspection and a "orrelation coefficient to be at least
0.995. After an initial calibration curve was obtained, the calibration check standard described
59
-------
earlier was analyzed. The standard was injected periodically throughout the analysis of
samples (i.e., after every six to eight samples), and was used for daily calibration.
Sample Analysis-
Samples were analyzed by HPLC. An acetonitrile blank was analyzed at least once per
day to ensure that the system was not contaminated. A check standard was analyzed prior to
sample analysis, after 6-8 samples, and at the end of the sample analysis. Samples were
diluted as necessary to keep concentrations within the calibration range.
Analytes were identified by retention time. The width of the retention time window
used for identification was based on the standard deviation in retention time for multiple
injections of a standard.
Laboratory Method Blanks
After DNPH preparation was completed, an aliquot of the solution was retained at the
laboratory and analyzed with the samples, as an indicator of any aldehyde/ketone contributions
attributable to laboratory procedures.
QUANTITATION
A least squares linear regression analysis of the calibration standards data was used to
calculate a correlation coefficient, slope, and intercept. Concentrations were used as the
X-variable, and response was used as the Y-variable.
The concentration of aldehyde in the samples was calculated as follows:
Concentration Aldehyde in Sample = SamPle ' lnl£rceP' x ^ ,MeM'
Slope MW derivative
60
-------
where;
MW aldehyde = the molecular weight of the aldehyde or ketone; and
MW derivative = the molecular weight of the derivative.
The total weight of aldehyde in the sample was calculated from the concentration, the volume
of methylene chloride into which the derivative was originally extracted, the volume of
methylene chloride that was used for the solvent exchange, and the final volume of acetonitrile
into which the sample was solvent exchanged.
Total Concentration Total Volume mL of ACN
ACN in _ ACN in of MeCl2 (after solvent exchange)
Sample Sample Extract mL of MEC12
(tig) (pg/mL) (mL) (before solvent exchange)
SPIKING
Two of the four trains making up the quad assembly were dynamically spiked during
each test run. Ten complete runs resulted in a total of 20 spiked and 20 unspiked trains. For
the first field test, nine different spiking compounds were used: formaldehyde, acetaldehyde,
quinone, acrolein, propionaldehyde, methyl ethyl ketone, acetophenone, methyl isobutyl
ketone, and isophorone. For the second field test, quinone and acrolein were excluded.
Spiking compounds were added at a level indicated in Table 22. Spiking compounds were
added at a level approximately five times that determine in the site survey samples
(Appendix C) of the flue gas stream or at 2 ppmv for compounds that were present at
0.4 ppmv or less.
61
-------
Table 22. Compounds Spiked and Nominal Spike Concentrations
Compound Nominal Concentration Spiked
Field Test I
Field Test U
ppmv
total mg
ppmv
total mg
Formaldehyde
20
21
2.0
2.1
Acetaldehyde
8.6
13
4.4
6.4
Quinone
2.1
8.1
NT
NT
Acrolein
2.1
4.2
NT
NT
Propionaldehyde
2.1
4.4
2.1
4.4
Methyl ethyl ketone
2.1
5.3
2.1
5.3
Acetophenone
2.1
8.9
2.1
8.9
Methyl isobutyl ketone
2.0
7.1
2.0
7.1
IsoDhorone
2.1
10
2.1
10
NT = Not Tested
The compounds dynamically spiked into the designated trains were prepared from neat
materials in water at a nominal concentration of 0.2 to 1 mg/mL. The concentrations were
verified in the laboratory and an aliquot removed and stored in the laboratory at 4°C. During
each run, the spiking solution was introduced to two of the four Method 00112 trains through
glass-lined stainless steel tubing via motor-driven syringe pumps. The flow rate of the liquid
spike into each train was set to 0.25 mL/min to allow the collection of a nominal 2 to 20 mg
of each compound in each Method 00112 train over a 1-hour sampling period. The spike was
introduced to each train at a point immediately after the probe and before the first impinger.
The probe and glass tubing leading to each train was maintained at a temperature of 130°C
(266°F).
62
-------
PRECISION AND ACCURACY ASSESSMENT
Precision is defined as the estimate of variability in the data obtained from the entire
system (i.e., sampling and analysis). At least two paired samples are needed to establish
precision.
Accuracy (bias) is defined as any systematic positive or negative difference between the
measured value and the true value. Percent recovery is defined as any gain or loss of a given
compound compared to a known spiked value.
Ten quadruplicate sampling runs (i.e., 40 sampling trains) were conducted during each
testing program. Acceptability criteria for the runs are detailed in Section 6 of the test plan.
Completion of at least six quad runs (24 independent trains) is required for statistical analysis
by Method 301.1 For the second field test, two runs were eliminated from the data set because
of suspected spiking errors. The following data treatment approach is written based on the
completion of all 10 quad runs. Adjustments to calculations were made based on the number
of runs actually performed and accepted.
The latest version of the Method 3011 describes the data analysis method necessary to
evaluate both the bias and the precision of emission concentration data from stationary sources.
Method 3011 was used for the statistical evaluation of the test data for this work assignment.
Method 301' assumes that the spike amounts for each train are equal. A problem
encountered in this study was that the calculated value of the spiked level was not constant for
every train. In order to complete the Method 3011 statistical analysis, the variability of the
spiked data was calculated from Equation 1:
di = YiA " YiB ~ SiB -
63
-------
where:
dj = the difference for Run i;
= the measured concentration of spiked sample A for Run i;
Ym = the measured concentration of spiked sample B for Run i;
Sm = the amount spiked into sample B for Run i; and
Sja = the amount spiked into Sample A for Run i.
Assessment or Precision According to Method 301
Precision of the spiked compounds was calculated using the difference between the
measured concentration, dj, of each spiked compound for each spiked train as calculated in
Equation 1. Precision is reported as the standard deviation between the paired measurements
of spiked compounds, SD„ given by the following equation:
sd^n
2n
where:
SD, = the standard deviation between the paired measurements of each spiked
compound;
n = the number of paired samples used in the calculation (n = 8 or 10); and
dj = the difference of paired sampling train measurements as calculated in
Equation 1.
The percent relative standard deviation (%RSD) of the proposed spiked method was calculated
as:
SD
%RSD = —- * 100
S„
64
-------
where:
S„
= measured mean of the spiked samples.
The proposed method is acceptable if the %RSD is not greater than 50 percent.
Precision of the unspiked compounds was calculated using the difference between the
measured concentration, dit of each spiked compound for each unspiked train. Precision (SDJ
is reported as the standard deviation of the differences between the paired measurements of
unspiked compounds, given by the following equation:
SDu = N
JX
2n
where:
SDU = the standard deviation between the paired measurements of unspiked
compounds;
n = the number of paired samples used in the calculation (n = 10); and
d; = the difference of paired unspiked sampling train measurements.
The %RSD of the unspiked trans was calculated as:
SD
o/0rsd = —a- * 100
M
m
where:
M,,, = measured mean of the unspiked samples.
The proposed method is acceptable if the 9SRSD is not greater than 50 percent.
65
-------
Assessment of Bias According to Method 3011
The experimental design allows for the determination of bias for each spike compound.
Bias for each spike compound was calculated using 16 or 20 spiked field samples, 16 or
20 unspiked field samples, and the calculated value of each spike. Because of differing spiked
amounts, the equation as given in Method 301' was modified to calculate bias for each spiked
train. Bias, b, of the method for each spiked compound for each spiked train of each run is
defined as:
bii = s« "
' Mj, + MaN
-CSfl
where:
i = run number (i.e., 1, 2, 3,
j = 1 or 2 (to indicate the first sample or the second sample);
by = bias for the j"1 spiked sample of the i"1 run;
Sjj = reported amount of the compound in thej"1 spiked sample of the i* run; and
M„ = reported amount of the compound in the first unspiked sample for the i* run;
M2 = reported amount of the compound in the second unspiked sample for the i*
run; and
CSjj = calculated (or theoretical) value of the spiked compound in the j* spiked
sample of the i® run.
The overall bias was then defined as:
B =
66
-------
where;
ba = bias for the j* spiked sample of the i* run; and
n = the number of samples used in the calculation (i.e., 2*the number of runs).
The standard deviation of the bias was then calculated as follows:
sn =
W Jmw
\
E £"' -(E E bg>'
n-1
The bias, B, calculated above was tested to determine if it was statistically different
from 0.0. A /-test was used to make this determination. He /-test compared the calculated
/-statistic of the test data with the critical / value for the degrees of freedom in the test data and
the desired level of significance. For the test matrices in this plan, there were 8 or 10 data
points, which were tested using a two-tailed /-distribution at the 95% confidence level. The
/-statistic was calculated as shown below:
t - _§!_
SD
fn
This /-test evaluates the hypothesis that the bias is not equal to zero. If the calculated absolute
value of the /-statistic is greater than the two-tailed critical value for the specified degrees of
freedom and level of significance, then there is significant bias. If the calculated absolute
value of the /-statistic is less than the critical value for the specified degrees of freedom and
level of significance, then the average difference of the concentration between paired sampling
trains is assumed to be zero and the measured concentration can be pooled for statistical tests.
The critical value of the /-statistic for the two-tailed /-distribution at a 0.05 level of
significance (95% confidence level) with 18 degrees of freedom is 2.101.
67
-------
When the /-test showed that the bias was statistically significant, the correction factor
(CF) was calculated as follows:
1
CF
i-S.
cs
where:
CF = the correction factor;
B = the bias; and
CS = the average calculated (or theoretical) spiked amount.
When the CF was within the range of 0.70 to 1.30, the data and method were considered
acceptable.
68
-------
SECTION 6.0
QUALITY ASSURANCE/QUALITY CONTROL
The quality assurance/quality control (QA/QC) activities for the sampling and analytical
procedures are presented in this section,
QUALITY CONTROL
The quality control procedures for field and laboratory activities are described in this
section. In addition to sampling and analytical QA/QC procedures, the project staff was
organized to allow review of project activities and provide QC coordination throughout the
term of the evaluation program.
Sampling QA/QC Procedures
The sampling QA/QC program for this project included data quality objectives, manual
method sampling performance criteria, field equipment calibrations, field spiking consistency,
sampling and recovery procedures, representative sampling, complete documentation of field
data and abnormalities, and adequate field sample custody procedures.
Data Quality Objectives-
Precision, bias, and completeness objectives were determined for manual sampling
operations and are listed in Table 23. The completeness objective was met with sampling runs
completed in the field. The precision and bias objectives were met for five of the seven
compounds tested. As expected from previous testing, MEK and MIBK did not pass the
method bias tests.
69
-------
Table 23. Field Sampling Quality Control Objectfves
Precision
Accuracy
Completeness
(%RSD)*
(%)
(%)
Aldehyde/Ketone Concentration
50"
70-130%e
100
Flue Gas Temperature
+ 1°C
+3°C
100
•Relative standard deviation.
bMethod 301,1 Section 1.2.2, precision objectives for method validation.
Method 3011 bias objectives for method validation.
Manual Method Performance Criteria-
Acceptance criteria, control limits and corrective actions for sample collection using the
Method 00112 sampling train are provided in Table 24.
Table 24. Summary of Acceptance Criteria, Control Limits,
and Corrective Action
Criteria
Control Limits*
Corrective Action
Final Leak Rate
s0.00057 acmm or 4% of
sampling rate, whichever is
less
None: Results are
questionable and should be
reviewed and compared with
other (3) train results
Dry Gas Meter Calibration
Post average factor (A.) agree
±5% of pre-factor
Adjust sample volumes using
the factor that gives smallest
volume
Individual Correction Factor (a)
Agree within 2% of average
factor
Redo correction factor
Average Correction Factor
1.00 ±1%
Adjust the dry gas meter and
recalibrate
Intermediate Dry Gas Meter
Calibrated every six months
against EPA standard
—
Analytical Balance (top loader)
0.1 g of NBS Class Weights
Repair balance and recalibrate
Barometric Pressure
Within 2.5 mm Hg of
mercurv-in-elass barometer
Recalibrate
'Control limits are established based on previous test programs conducted by the EPA.
70
-------
Field Equipment Calibrations--
S-Type Pitot Tube Cnlihratlon-The EPA has specified guidelines concerning the
construction and geometry of an acceptable S-Type pitot tube. Information pertaining to the
design and construction of the Type-S pitot tube is presented in detail in Section 3.1.1 of the
Quality Assurance Handbook.6 Pitot tubes were inspected and documented as meeting EPA
specifications prior to field sampling. A pitot tube coefficient of 0.84 was used for velocity
calculations.
Sampling Nozzle Calibration-Glass nozzles were used for isokinetic sampling. All
•lozzles were thoroughly cleaned, visually inspected, and calibrated according to the procedure
outlined in Section 3.4.2 of EPA's Quality Assurance Handbook.6
Drv Gas Meter Calihration-Dry gas meters (DGMs) were used in the
aldehyde/ketone sample trains to measure the sample volume. All DGMs were calibrated to
document the volume correction factor prior to the departure of the equipment to the field.
Post-test calibration checks were performed after the equipment was returned to Radian's PPK
laboratory. All dry gas meters met the acceptance criteria listed in Table 24.
Dry gas meter calibrations were performed at Radian's PPK laboratory using an
American® wet test meter as an intermediate standard. The intermediate standard is calibrated
every six months against the EPA spirometer at EPA's Emission Measurement Laboratory in
Research Triangle Park (RTP), North Carolina.
Prior to calibration a positive pressure leak check of the system was performed using
the procedure outlined in Section 3.3.2 of EPA's Quality Assurance Handbook.6 The system
was placed under approximately 250 mm of water pressure and a gauge oil manometer
demonstrated that no decrease in pressure occurred over a one-minute period.
After the sampling console was assembled and leak checked, the pump was allowed tu
run for 15 minutes to allow the pump and DGM to warm up. The valve was then adjusted to
71
-------
obtain the desired flow rate. For the pre-test calibrations, data were collected at the orifice
manometer settings (AH) of 13, 25, 38, 51, 76, and 102 mm H20. Gas volumes of 0.14 mJ
were used for the two lower orifice settings, and volumes of 0.28 m3 were used for the higher
settings. The individual gas meter correction factors (Yj) were calculated for each orifice
setting and averaged. The method requires that each of the individual correction factors fall
within ±2% of the average correction factor or the meter will be cleaned, adjusted, and
recalibrated. In addition, Radian requires that the average correction factor be within 1.00
±1 percent. For the post-test calibration, the meter wzz calibrated three times at the average
orifice setting and vacuum which were used during the actual test.
Sampling Operation/Recovery Procedures—
To ensure consistency between trains/runs, two individuals conducted the manual
sampling, and one person was assigned to clean up, recover, and reassemble the glassware.
This protocol serves to eliminate propagation of multiple operator variance. All team
members were familiar with the procedures detailed in the test plan. Sampling trains were
leak checked before and after each run. The leak rate of each train was within the specified
limits. The recorded leak rates for each train are presented in Tables 25 and 26. All samples
were withdrawn at a rate within 10 percent of isokinetic with the stack gas velocity. Isokinetic
rate data axe presented with the sampling parameters in Tables 3 and 10.
Representative Sampling—
The uniformity of sampling between trains was verified by comparing gas volumes and
moisture content values. Velocity head and flue gas temperature were compared between runs
to assess the variability in stack gas conditions.
-------
Table 25. Leak Hates, Field Test I
Pre-Test Post-Test
Run
Leak Rate
(irf/mJirt
Vacuum
(mm He)
Leak Rate
fmVmirrt
Vacuum
(mm He)
1A
0.00017
203
0.00017
127
IB
0.00023
203
0.00028
102
1C
0.00034
203
0.00023
178
ID
0.00011
203
0.00023
152
2A
0,00017
229
0.00017
127
2B
0.00040
203
0.00025
127
2C
0.00014
178
0.00014
152
2D
0.00011
203
0.00037
127
3A
0.00017
178
0.00031
152
3B
0.00028
203
0.00037
203
3C
0.00025
178
0.00037
152
3D
0.00017
203
0.00011
279
4A
0.00011
229
0.00008
127
4B
0.00023
203
0.00028
178
4C
0.00008
178
0.00011
152
4D
0.00020
305
0.00011
305
5A
0.00011
152
0.00011
127
5B
0.00042
178
0.00034
203
5C
0.00017
178
0.00023
203
5D
0.00017
305
0.00025
279
6A
0.00040
152
0.00031
127
6B
0.00040
178
0.00028
178
6C
0.00011
203
0.00008
127
6D
0.00023
203
0.00025
152
7A
0.00023
152
0.00011
152
7B
0.00023
178
0.00042
203
7C
0.00017
203
0.00011
127
7D
0.00008
178
0.00014
178
8A
0.00006
152
0.00017
203
8B
0.00025
178
0.00037
152
73
-------
Table 25. (Continued)
Pre-Test
Post-Test
Run
Leak Rate
CmVmln)
Vacuum
(nun He)
Leak Rate
(mVmin)
Vacuum
(mm He)
8C
0.00008
178
0.00011
178
8D
0.00020
203
0.00011
178
9A
0.00011
203
0.00006
152
9B
0.00040
254
0.00045
254
9C
0.00023
178
0.00008
127
9D
0.00025
178
0.00011
229
10A
0.00008
152
0.00014
203
10B
0.00034
254
0.00028
305
IOC
0.00028
178
0.00113
152
10D
0.00014
254
0.00011
127
74
-------
Table 26. Leak Rates, Field Test II
Pre-Test Post-Test
Run
Leak Rate
(m3/min)
Vacuum
(nun Hg)
Leak Rate
(mVmin)
Vacuum
(mm Hg)
1A
0.00028
254
0.00017
203
IB
0.00017
254
0.00014
203
1C
0.00028
254
0.00025
229
ID
0.00040
254
0.00031
178
2A
0.00023
203
NR
NR
2B
NR
NR
0.00014
178
2C
0.00020
254
0.00025
178
2D
0.00017
254
0.00011
127
3A
0.00031
178
0.00020
203
3B
0.00011
178
0.00008
127
3C
0.00025
203
0.00017
178
3D
0.00034
203
0.00025
203
4A
0.00014
127
0.00006
127
4B
0.00023
178
0.00020
178
AC
0.00020
178
0.00020
203
4D
0.00031
178
0.00017
254
5A
0.00020
203
0.00031
203
SB
0.00020
203
0.00031
203
5C
0.00017
152
0.00000'
203
5D
0.00011
203
0.00025
178
6A
0.00014
152
0.00025
178
6B
0.00028
178
0.00011
152
6C
0.00023
178
0.00017
152
6D
0.00020
127
0.00020
178
7A
0.00031
229
0.00025
203
IB
0.00017
203
. 0.00011
127
7C
0.0014
127
0.00017
178
ID
0.00020
152
0.00011
152
8A
0.00014
178
0.00011
127
8B
0.00008
178
0.00017
178
8C
0.00017
178
0.00000*
178
8D
0.00031
254
0.00025
178
75
-------
Table 26. (Continued)
Pre-Test Post-Test
Leak Rate
Vacuum
Leak Rate
Vacuum
Run
(m'/min)
(mmHg)
(m'/niin)
(mmHg)
9A
0.00025
203
0.00017
178
9B
0.00020
178
0.00011
178
9C
0.00023
152
0.00008
127
9D
0.00020
203
0.00008
178
10A
0.00017
178
0.00011
178
10B
0.00017
178
0.00023
229
IOC
0.00011
152
0.00017
229
10D
0.00017
254
0.00020
203
NR = Not recorded.
*Leak check performed after tightening impinger clamp which was knocked loose during removal of probe
assembly from stack.
76
-------
Documentation-
Field data sheets were completed and checked alter each test run. Test progress and
any notable events affecting the sampling or process were recorded in the field log notebook.
Documentation of pre- and post-test calibrations and inspections was maintained.
Sample Custody-
Sample custody procedures for this program are based on EPA-recommended
procedures. The custody procedures emphasize careful documentation of sample collection
and field analytical data and the use of chain-of-custody records for samples being transferred.
These procedures are discussed below.
The sample recovery task leader was responsible for ensuring that all samples taken
were accounted for and that proper custody and documentation procedures were followed for
the field sampling efforts. A master sample logbook was maintained by the recovery task
leader to provide a hard copy of all sample collection activities. Manual flue gas sampling
data were also maintained by the recovery task leader.
Following sample collection, all samples were given a unique alphanumeric sample
identification code as shown in Figure 7. Sample labels and integrity seals, similar to those
shown in Figure 8, were completed and affixed to the sample containers. The sample volumes
were determined and recorded and the liquid levels were marked on each bottle. The sample
identification code was recorded on the sample label and in the sample logbook. The samples
were stored in a secure area until they were packed.
As the samples were packed for travel, chain-of-custody forms (Figure 9) were
completed for each shipment container. The chain-of-custody forms and written instructions
specifying the treatment of each sample were enclosed in the sample shipment container.
Shipping containers were labeled with "up arrows" to clearly indicate the upright position of
sample bottles.
77
-------
EPA
1118
t t
Cliant Sampling
Duignsiion OaJa
MdKot
~
— PR/lat
*
— 1118
~
I
Sampla
Typa
I
• Train
Component
I
Run
Nufflbar
{1-12 cr blank)
t
Train kSartifceilon
or
Quad Auambly
PR/1* combinad prob« rtnta
and first impingar contants
2nd/3rd comblnad contents of
s«cond and third tmp4r>3«rs
Figure 7. Sample Identification Code
78
-------
soRPoemTBoea PRELIM. NO:
900 P«rlm»t«r Park
Morrl»vlll», NC 27580
(919)*i1-0212
SAMPLE TYPE
LOCATION:
DATE: _
REMARK:
CONTRACT:
FINAL WT:
TARE:
SAMPLE WT:
ATTENTION: ATTENTION:
BEFORE OPENING ¦».¦!«««« BEFORE OPENING
NOTE IF BOTTLE WAS SAMPLE CODE N0TE ,p BomE WAS
TAMPERED WITH. FIELD NO. CONTAINER NO. TAMPERED W»TH.
Figure 8. Example of Sample Label and Integrity Seal
79
-------
Chain of Custody Racord
F1EUJ SAMPLE LO.
RECEIVED BY:
BEUNOUtSHEDBY;
RELINQUISHED BY:
Figure 9. Chain-of-Custody Record
80
-------
Laboratory OA/OC Procedures
The laboratory QA program for this project included proper handling, logging and
tracking of incoming samples, procedure validations including calibration curves, daily QC
checks, and collection and/or analysis of field train and field reagent blanks, and method
spikes as well as laboratory spikes. A summary of Radian's laboratory QC procedures Is
provided in Table 27.
Table 27. Laboratory Quality Control Procedures
Parameter
Analytical
Method
Quality
Control
Check
Frequency
Acceptance
Criteria
Corrective
Action
Linearity
Check
HPLC
Run 5-point
curve
At setup or
when check
std, is out-of-
range
Correl. coeff.
a0.995
Check integ.,
reinteg. If
necessary,
recalibrate
Retention
Time
HPLC
Analyze
check
standard
1/6-8
injections
±15% day-
to-day; ±5%
within one
day
Check instr. fund,
for plug, etc. Heat
column; Adjust
gradient
Calibration
Check
HPLC
Analyze
check
standard
1/6-8
injections
min. 2/set
±15% of
calibration
curve
Check integ.,
remake std. or
recalib.
System
Blank
HPLC
Analyze
acetonitrile
1/day
sO. 1 level of
expected
analyte
Locate source of
contam.; reanalyze
Method
Spikes
HPLC
Analyze
spiked DNPH
1/10 samples
or 1/set
±20% of
spiked
amount
Check integ.,
check instr.
function,
reanalyze,
reprepare if
possible
81
-------
Sample Custody/Tracking—
Upon receipt of samples at Radian's PPK laboratory, the chain-of-custody forms and
sample bottle labels were compared to verify receipt of samples. A copy of the sample log
notebook was provided to the laboratory representative. After logging the samples into the
Radian tracking system, they were stored at 4°C to prevent decomposition of derivatives.
Calibration Curve--
A five-point calibration curve as described in Section 5 was prepared and analyzed after
initially setting up the instrument. The calibration data are presented in Table 28. All of the
calibration curves used for both field tests had correlation coefficients greater than 0.998.
Daily QC Checks—
A check standard as described in Section 5 was prepared and used to check instrument
response and the calibration curve. Hie check standard was analyzed before and after all
sample analyses and after each sixth to eighth sample. The check standard recoveries are
presented in Tables 29 and 30 for Field Tests I and II, respectively. All of the check standard
responses fell within the 85 to 115% of known value criterion for Field Test I. Two MEK and
most of the isophorone check standard responses fell outside the 85 to 115% criterion for Field
Test II because of a calculation error during sample analysis. In some cases the data were not
affected by the high check standard responses because only diluted samples were being
analyzed for acetaldehyde.
System Blanks-
Neat acetonitrile (system blank) was analyzed at least once per day to ensure that the
analytical instrument was not contaminated. None of the analytes were detected in the system
blanks.
82
-------
Table 28. Calibration Data'
Meets
Correlation Acceptance
Compound
Date
Slope
Intercept
Coefficient
Criteria
Formaldehyde
4/94
8.14 x 10*
0.0422
0.9999
Yes
5/95
8.44 x 10*
0.0916
0.9989
Yes
Acetaldehyde
4/94
7.98 x lO*
-0.00133
0.9999
Yes
5/95
8.18 x 10*
0.0529
0.9989
Yes
Quinone
4/94
1.30 x 10 s
-0.191
0.9997
Yes
Acrolein
4/94
7.08 x 10*
-0.00088
0.9998
Yes
Propionaldehyde
4/94
8.37 x 10*
0.00642
0.9999
Yes
5/95
8.46 x 10*
0.0690
0.9987
Yes
Methyl Ethyl Ketone
4/94
8.91 x 10*
0.0809
0.9996
Yes
5/95
9.66 x 10*
0.0669
0.9989
Yes
Acetophenone
4/94
9.96 x lO*
0.0158
0.9999
Yes
5/95
1.12 x 10 s
0.165
0.9987
Yes
Methyl Isobutyl Ketone
4/94
9.75 x 10*
0.00486
0.9999
Yes
5/95
9.79 x 10*
0.0585
0.9989
Yes
Isopherone
4/94
1.13 x 10"5
-0.0108
0.9999
Yes
5/95
1.23 x 10 s
0.104
0.9988
Yes
* Concentration of Derivative Qig/mL) = Area x Slope + Intercept
83
-------
Table 29. Calibration Check Standard Recoveries for Field Test I
Percent of Target'
Sample
ID
File#
Date
Time
Form-
aldehyde
Aeet-
aldehyde
Quinoae
Acrolein
Propion-
aldebyde
MEK
Aeeto-
phcnone
MIBK
Isophor-
one
QC 1
T4HA004
080194
16:53
90.3
91.4
97.3
99.8
96.9
90,0
92.9
92.2
93.4
QC 2
T4HA014
080294
00:53
96.5
98.7
97.7
102.
102.
97.9
99.5
98.1
101.
QC3
T4HA025
080294
09:40
96.2
98.6
97.8
101.
94.7
97.7
101.
100.
104.
QC 1
T4HB001
080294
14:52
98.4
101.
104.
104.
103.
98.7
100.
101.
102.
QC 2
T4HB012
080294
23:40
97.8
97.8
97.8
101.
102.
95.6
100.
97.9
102.
QC3
T4HB020
080394
06:03
88.8
95.6
99.5
102.
102.
95.8
96.6
97.3
100.
QC 1
T4HC001
080394
16:30
95.9
97.4
101.
100.
98.3
97.0
101.
99.1
102.
QC2
T4HC013
080494
02:06
97.3
96.2
94.1
98.7
95.1
94.9
95.8
96.2
101.
QC3
T4HC019
080494
06:53
96.6
99.5
100.
102.
101.
98.2
101.
98.5
101.
QC 1
T4HD010
080494
17:12
97.3
97.9
97.8
100.
99.1
98.6
97.3
98.3
101.
QC2
T4HD022
080594
02:47
98.2
97.7
96.6
101.
99.5
98.3
99.3
98.2
100.
QC3
T4HD027
080594
06:47
93.6
95.7
95.3
101.
99.2
93.9
95.2
96.0
99.9
QC 1
T4HE001
080594
17:24
93.6
95.3
96.1
99.0
99.7
92.6
94.6
94.1
101.
QC2
T4HE013
080694
02:59
94.9
96.7
95.9
99.2
98.1
96.4
98.6
97.2
98.1
QC3
T4HE020
080694
08:35
96.4
99.6
98.6
104.
102.
98.4
102.
99.8
101.
QC 1
T4HH001
080794
16:04
98.5
98.1
96.5
101.
101.
98.4
103.
100.
102.
QC2
T4HH013
080894
01:40
99.4
100.
98.0
103.
102.
97.9
98.4
98.8
103.
QC 3
T4HH022
080894
08:51
94.6
98.8
95.5
99.6
99.0
96.4
98.6
99.1
100.
QC 4
T4HH022H
080894
15:21
96.3
97.1
101.
103.
106.
94.3
97.4
98.5
100.
QC 1
T4HI007
080894
20:44
94.4
93.6
93.9
98.0
98.6
94.0
93.2
94.1
96.6
QC2
T4HI019
080994
06:19
99.1
98.5
97.1
99.9
98.5
99.2
96.6
99.6
100.
QC 3
T4HI031
080994
15:55
95.3
97.3
93.5
99.1
99.5
98.4
96.1
93.9
101.
QC4
T4HI041
080994
23:58
98.4
97.0
97.8
104.
104.
95.6
100.
100.
ioi.
QC5
T4HI046
081094
03:58
97.7
100.
93.4
100.
99.0
97.0
100.
96.6
101.
QC 6
T4HI046K
081094
12:46
95.1
96.0
94.6
100.
102.
96.3
95.5
96.3
100.
-------
Table 29. (Continued)
Percent of Target'
Sample Form- Acet- Propion- Aceto- Isophor-
ID
File#
Date
Time
aldehyde
aldehyde
Quinone
Acrolein
aldehyde
MEK
pfaenone
MBS
one
QC1
T4HK001
081094
14:00
95.8
97.2
96.6
103.
103.
97.1
94.9
98.0
99.2
QC2
T4HK013
081094
23:35
97.7
97.9
95.8
100.
99.2
97.0
96.7
100.
100.
QC3
T4HK022
081194
06:47
96.5
98.3
95.0
99.2
98.6
99.5
97.6
100.
102.
QC 1
T4H0002
081394
10:35
94.5
99.4
97.6
100.
99.6
97.0
95.2
97.3
101.
QC2
T4HO014
081394
20:11
95.3
98.3
95.9
99.4
100.
97.0
94.0
99,8
101.
QC 3
T4HO026
081494
05:46
99.0
98.8
96.3
101.
101.
97.1
95.0
96.7
102.
QC4
T4HO036
081494
13:46
95.7
98.3
98.6
102.
106.
98.1
93.7
97.9
99.4
QC l
T4HY001
082594
10:55
99.5
99.1
101.
103.
101.
99.7
95.8
98.8
101.
OC2
T4HY009
082594
17:18
97.6
97.2
97.0
101.
98,2
96.6
93.3
95.0
99.3
MEK = Methyl ethyl ketone
MIBK = Methyl isobutyl ketone
* Acceptable range is 85 to 115 percent.
-------
Table 30. Calibration Check Standard Recoveries for Field Test II
Percent of Target'
Sample
ID
File#
Dale
Time
Form-
aldehyde
Acet-
aldehvde
Propion-
aldehvde
MEK
Aceto-
Dhenone
MIBK
Iso-
ohorone
QC 1
J5EC001
04-May-95
05:59 am
102.
112.
108.
109.
109.
104.
116.b
QC2
J5EC015
04-May-95
05:11 pm
93.5
98.0
99.3
101.
102.
96.0
110.
QC 1
T5ED002
04-May-95
07:12 pm
104.
106.
109.
110.
105.
102.
114.
QC 2
T5ED014
OS-May-95
04:48 am
92.3
101.
101.
106.
102.
97.9
109.
QC 3
T5ED024
0S-May-9S
12:47 pm
96.2
101.
102.
108.
101.
96.5
109.
QC 1
J5EH001
06-May-95
01:04 pm
98.4
108.
107.
110.
107.
104.
116.b
QC 2
J5EH016
07-May-95
02:49 am
98.1
103.
102.
110.
100.
99.4
118.b
QC 3
J5EH028
07-May-9S
12:25 pm
96.8
105.
104.
108.
102.
100.
116.b
QC 4
J5EH040
07-May-9S
10:00 pm
103.
107.
109.
112.
104.
103.
116.b
QC 5
J5EH051
08-May-95
06:48 am
90.8
98.3
95.6
102.
94.0
91.0
104.
QC 1
T5EJ001
08-May-95
01:02 pm
101.
108.
107.
111.
104.
103.
116.b
QC 2
T5EJ013
09-May-95
12:17 am
97.4
104.
106.
109.
101.
100.
117.h
QC 3
T5EI025
09-May-9S
09:53 am
• 88.1
103.
98.8
107.
97.7
98.0
116."
QC 1
T5EO001
15-May-95
11:09 am
100.
108.
111.
116/
106.
106.
119.'
QC 2
T5EO010
15-May-95
07:08 pm
97.3
101.
101.
105.
97.9
96.8
114.
QC 3
T5EO019
16-May-95
02:20 am
87.6
95.8
96.9
97.5
97.7
89.9
99.8
QC 4
T5EO031
16-May-95
11:56 am
94.7
103.
105.
104.
97.0
98.3
117.c
QC 5
T5E0043
17-May-9S
12:36 am
96.1
107.
109.
113.
102.
103.
116.b
QC 6
T5EO055A
17-May-95
01:08 pm
97.2
111.
111.
112.
102.
105.
116."
QC 7
T5E0068
17-May-95
11:31 pm
100.
111.
108.
111.
99.3
101.
115.
-------
Table 30. (Continued)
Percent of Target'
Sample
ID
File#
Date
Time
Form-
aldehyde
Acel-
aldehvde
Propion-
aldehvde
MEK
Aceto-
ohenone
MIBK
Iso-
nhorone
QC 8
T5EO082
18-May-95
10:43 am
97.7
105.
106.
113.
98.4
98.7
117."
QC9
T5EO084C
18-May-95
02:42 pni
96.4
112.
109.
117.«
97.5
104.
U8.c
QC 1
T5ED002A
22-May-95
11:33 am
94.6
103.
105.
108.
97.1
99.5
119."
QC 2
T5EV006
22-May-95
03:37 pra
94.2
103.
104.
109.
96.6
101.
115.
MEK = Methyl ethyl ketone
MIBK = Methyl isobutyl ketone
* Acceptable range is 85 to 1 IS percent.
b Outside range, data flagged.
* Outside range, data not affected.
-------
Laboratory Method Blanks-
One method blank (MB) was prepared for every quad run for both field tests. The
MBs indicated contamination that occurred in the laboratory during the sample preparation
process. The MB data is presented in Tables 31 and 32 for Field Tests I and II, respectively.
Laboratory Method Spikes and Method Spike Duplicates-
For the first field test, one method spike (MS) for every quad train was prepared. For
the second field test, one MS and method spike duplicate (MSD) for every quad train were
prepared. Thus, for both field tests a total of 30 MS samples were prepared and analyzed.
The recovery criterion for MS and MSDs was 100±20 percent. The MS recovery data are
presented in Tables 33 and 34 for Field Tests I and II, respectively.
Formaldehyde MS/MSD recoveries were within the acceptable range in every case.
One isophorone MSD recovery during the second field test was just barely outside the upper
limit (121 versus 120). Two acetophenone MSD recoveries during the second field test were
also outside the upper limit. For these three compounds the MS/MSD recovery criteria were
achievable greater than 90% of the time.
During the first field test, acetaldehyde and propionaldehyde MS recoveries were
within the acceptable range for 9 of the 10 samples. MEK and MIBK MS recoveries were
within the acceptable range for 8 of the 10 samples and 7 of the 10 samples, respectively.
However, during the second field test, acetaldehyde MS/MSD recoveries were within range
for only 14 out of the 20 samples; propionaldehyde was within the acceptable range for 3 of
the 20 samples; MEK was within the acceptable range for 6 of the 20 samples; and MIBK was
out of the acceptable range for all 20 samples. The poorer performance of these compounds
during the second field test may have resulted from the longer time that the samples were
stored between being spiked and extracted.
88
-------
Table 31. Laboratory Method Blank Results for Field Test I
Total micrograms
Form- Acet- Prop»D- Aceto- bo-
Sample aldehyde aldehyde Quinone Acrolein aldehyde MEK phenone MIBK phorone
MB 1
1.03
ND
ND
ND
ND
1.90
ND
0.86
ND
MB 2
1.05
0.19
ND
ND
ND
ND
ND
ND
ND
MB 3
1.21
ND
ND
ND
0.76
1.33
ND
ND
ND
MB 4
0.70
0.28
ND
0.22
0.47
1.11
ND
ND
ND
MBS
1.01
0.26
ND
ND
ND
ND
ND
ND
ND
MB 6
1.13
ND
ND
ND
0.46
1.30
ND
ND
ND
MB 7
0.70
0.31
ND
0.70
0.24
1.28
ND
ND
ND
MB 8
10.6
0.28
4.26
0.51
0.51
2.83
ND
ND
ND
MB9
1.95
ND
ND
ND
ND
ND
ND
ND
ND
MB 10
2.96
ND
0.39
0.59
2.29
1.31
ND
ND
ND
Avenge
2.23
0.13
0.47
0.20
0.47
1.106
0.00
0.09
0.00
Standard
Deviation
3.03
0.14
1.34
0.29
0.69
. 0.91
NA
0.27
NA
Relative
Standard
Deviation
135%
108%
285%
145%
147%
82%
NA
300%
NA
-------
Table 32. Laboratory Method Blank Results for Field Test II
Total micrograms
Samole
Formaldehyde
Acetaldehvde
Prouionaldehvde
Methyl Ethyl
Ketone
Acetoohenone
Methyl
Isobutyl
Ketone
Isoohorone
MB 1 RERUN
0.82
0.89
2.07
ND
ND
ND
ND
MB 2
0.92
0.64
1.27
ND
ND
ND
ND
MB 3
0.99
0.50
ND
ND
ND
ND
ND
MB 4
1.09
0.28
ND
ND
ND
ND
ND
MBS
1.95
0.61
ND
ND
ND
2.12
ND
MB 6
1.34
0.58
ND
ND
ND
1.43
ND
MB 7
0.81
0.28
ND
ND
ND
ND
ND
MBS
1.18
0.59
ND
ND
ND
ND
ND
MB 9
1.09
0.73
ND
ND
ND
ND
ND
MB 10
1.12
0.58
ND
ND
ND
ND
ND
Average
1.13
0.57
1.67
ND
ND
1.78
ND
Standard Deviation
0.33
0.19
0.57
NA
NA
0.49
NA
Relative Standard
Deviation
29.31%
32.91 %
34.03%
NA
NA
27.40%
NA
Note: Final values are not Method Blank corrected.
ND = Not Detected
-------
Table 33. Percent Recovery' for Method Spike Samples for Field Test I
Form- Acel- Propion-
Sample
aldehyde
aldehyde
Quinone
Acrolein
aldehyde
MEK
Acetophenone
MIBK
Isophorone
MS 1
97.3
91.3
93.1
81.0
97.3
92.6
102
81.4
107
MS 2
95.2
89.4
85.5
80.0
101
74.6"
93.8
73.8"
99.2
MS 3
101
92.4
68.6"
71.4b
93.6
97.5
102
90.5
105
MS 4
98.1
90.9
57.4b
64.5b
90.1
80.9
98.1
80.0
105
MS 5
110
100
68.3b
7l.8b
107
74.6"
110
78.9"
115
MS 6
113
99.3
39,9b
83.2
102
94.7
104
92.2
112
MS 7
108
97.4
56.6"
85.4
108
1&2.8
107
97.5
115
MS 8
99.0
92.9
81.6
65.2b
94.5
95.5
104
94.6
111
MS 9
83.4
76.8''
46.6*
48.3b
73.4b
81.4
82.3
69.9b
86
MS 10
101
89.0
36.0"
59.9b
90.1
95.5
98.8
82.4
104
Maximum
113
100
93.1
85.4
108
102.8
107
97.5
115
Minimum
83.4
76.8
36.0
48.3
73.4
74.6
82.3
69.9
86
Average
101
92.0
63.4
71.1
95.8
89.0
100
84.1
106
Standard Deviation
8.45
6.69
19.5
11.8
10.2
10.2
7.79
9.18
8.53
Relative Standard
8.40%
7.27%
30.8%
16.6%
10.7%
11.4%
7.78%
10.9%
8.06%
Deviation
'
MEK - Methyl ethyl ketone
MIBK = Methyl isobutyl ketone
_ . n Total its Recovered „ tnn
Percent Rccovciy = p - * 100
Total |ig Spiked
* Acceptable range is 80 to 120 percent.
k Outside range, data flagged.
-------
Table 34. Percent Recovery8 for Method Spike Samples for Field Test II
Sample
Formaldehyde
Acelnldehvde
ProDionaldehvde
MEK
Acctaohenone
MIBK
Isoohorone
MS 1
96.6
91.8
79.5b
86.0
118
54.2b
112'
MSD 1
87.2
89.0
77.4b
72.4b
114
36.5b
112*
MS 2
92.4
92.8
84.5
93.1
119
47.3b
117*
MSD 2
89.7
88.0
77.9b
86.5
114
39.8b
106'
MS 3
91.6
84.3
77.9b
78.4b
119
43.2b
110'
MSD 3
93.8
82.6
77. lb
78.9b
121b
40. lb
109"
MS 4
94.2
81.2
82.1
85.5
119
37.6b
107*
MSD 4
91.2
93.9
76.9b
80.0
117
35.1b
121b,i
MS 5
87.2
91.0
73.4b
65. lb
113
38.2b
114'
MSD 5
95.4
90.3
84.2
69.4b
124b
41.9b
115'
MS 6
88.0
82.3
74.6b
79.7b
111
30.2b
107*
MSD 6
82.5
82.4
69.9"
77.8"
104
30.5b
104'
MS 7
82.8
79.2"
67.6b
70.5"
108
30.0b
106'
MSD 7
91.7
81.5
73.6b
72. lb
114
28.6b
108'
MS 8
84.1
76.8b
68.8V
80.6
98.0
43.3b
103'
MSD 8
87.5
78.6b
71.7s
77.9b
99.5
45.5"
104'
MS 9
90.8
81.6
73.7*
72.4b
104
30.0b
109*
MSD 9
82.3
71.8b
58.7*
67.8h
89.7
35.6b
98.4'
MS 10
88.2
78.4b
75.6b
99.0
30.7b
105*
MSD 10
86.2
76.2b
67.3b
70.5b
97.3
30.0b
104*
-------
Table 34. (Continued)
Samole
Formaldehyde
Acetaidehvde
Fronionaldehvde
MEK
Acetoohenone
MIBK
Isoohorone
Maximum
96.6
93.9
84.5
93.1
124
54.2
121
Minimum
82.3
71.8
58.7
65.1
89.7
28.6
98.4
Average
89.2
83.7
74.3
77.0
110
37.4
109
Standard Deviation
4.29
6.23
6.30
7.16
9.65
7.05
5.55
%RSD
4.81%
7.44%
8.48%
9.30%
8.77%
18.8%
5.11%
MEK = Methyl ethyl ketone
MIBK = Methyl isobutyl ketone
Percent Recovciy = Total pg Recoyenxl „ 1Q0
Total jig Spiked
* Acceptable range is 80 to 120 percent,
' Outside range, data flagged,
* Calibration check standard outside range (116%, 117%),
-------
Quinone and acrolein were only Included during the first field test. The MS recoveries
were usually outside of the acceptable limits. Quinone seems to react with the DNPH reagent
at a slower rate than the other carbonyl compounds and acrolein, because of its reactive double
bond, tends to tautomerize.
Field Train and Field Reagent Blanks-
Two field train blanks were collected as described in Section 5. These field train
blanks were collected on the first and fourth day of sampling and were processed in the same
manner as collected samples. One field train blank was collected using a spiked train
(Train A) and the other field train blank was collected using an unspiked train (Train B). The
field train blank results are reported in Tables 35 and 36 for Field Tests I and II, respectively.
Formaldehyde and acetaldehyde were detected in all four field train blanks. MIBK and
isophorone were not detected in any of the field train blanks.
Field reagent blanks of recovery solvents and unused DNPH reagent were collected in
the field and shipped to Radian's PPK laboratory. The field reagent blank results are reported
in Tables 37 and 38 for Field Tests I and II, respectively. Field train and field reagent blank
analytical results serve as indicators of contamination that may have occurred during sampling
and recovery operations.
94
-------
Table 35. Field Train Blank Results in Total Micrograms of Carbonyl for Field Test I
Field Train Blank A Field Train Blank B
Coraoound
Probe,
Iraptngers 1
and 2
Impinger 3
and 4
Total
Probe,
Impingers 1
and 2
Impinger 3
and 4
Total
Averase
Formaldehyde
6.90*
2.08'b
8.98
5.61*b
2.14,,b
7.75
8.36
Acelaldehydo
2.69th
1.82,b
4.51
3.55,b
1.40**
4.95
4.73
Quinone
2.88b
ND
2.88
ND
ND
ND
1.44
Acrolein
0.97,b
1.68,b
2.65
ND
ND
ND
1.32
Propionaldehyde
1.07*b
2.19",b
3.26
7.36,b
4.11'*
11.5
7.36 •
Methyl ethyl ketone
2.77b
2.88b
5.65
15.5b
<1.54
15.5
10.6
Acetophenono
2.26b
<0.42
2.26
<0.84
<0.42
<0.84
1.13
Methyl isobutyl ketone
<0.58
<0.58
<0.58
<1.16
<0.58
<1.16
<0.87
Isophorone
ND
ND
ND
ND
ND
ND
ND
Note: Final values are not corrected for (be Field Reagent Blank.
ND = Not Detected
'Less than 10 tiroes the level measured in the Field Reagent Blank.
''Below calibration curve.
-------
Table 36. field Train Blank Results in Total Micrograms of Carbonyl for field Test II
Field Train Blank A Field Train Biank B
Comoound
Probe,
Impinge rs 1
and 2
Impinger 3
and 4
Total
Probe,
Impingers 1
and 2
Impinger 3
and 4
Total
Avenue
Formaldehyde"
3.56'"
3.36"1
6.92
4.23'4
3-95"1
8.18
5.82
Aeetaldehyde1"
3.3 T4
1.60s4
4.98
2.96'4
2.6V4
5.58
4.03
Propionaldehyde*
<1.12
<1.12
<1.12
1.41'4
i.68'4
3.09
1.54
Methyl ethyl ketone
<1.27
<1.27
<1.27
<1.27
<1.27
<1.27
<1.27
Acctophenone
<4.40
<4.40
<4.40
<4.40
<4.40
<4.40
<4.40
Methyl isobutyi ketone
<1.39
<1.39
<1.39
<1.39
<1.39
<1.39
<1.39
Isophorone
<3.01
<3.01
<3.01
<3.01
<3.01
<3.01
<3.01
Note: Final values arts not corrected for the Field Reagent Blank.
ND « Not Detected
* More than 10% of the lowest sample value, data flagged,
h Less than 10% of the lowest sample value, meets criterion.
'Less than 10 times the level measured in the field reagent blank,
d Below calibration curve.
-------
Table 37. Field Reagent Blank Results for Methylene Chloride
Blank (Field Test I, August 1994)
Total micrograms
Compound
WIM5
WIL-86
Average
Formaldehyde
1.94
1.95
1.95
Acetaldehyde
1.67
0.57
1.12
Quinone
ND
ND
ND
Acrolein
0.78
ND
0.39
Propionaldehyde
1.18
1.96
1.57
Methyl ethyl ketone
ND
ND
ND
Acetophenone
ND
ND
ND
Methyl isobutyl ketone
ND
ND
ND
Isophorone
ND
ND
ND
97
-------
Table 38. Field Reagent Blank Results in Total Micrograms of Carbonyl for field Test II
Methylene
Chloride
Compound
Blank
DNPH Blank
Water Blank
4/28/95
4/27/95
4f2ms
Averaec
4/27/95
4/2S/95
Ave rase
Formaldehyde
1.34"-b
3.25'
1.58'*
2.41
0.%'"
1.21**
1.08
Acetaldehyde
<0.69
1.05*b
1.25*b
1.15
0.77*'b
0.50**
0.63
Propionaldehyde
1.54b
<0.45
<0.45
<0.45
<0.45
<0.45
<0.45
Methyl ethyl ketone
<1.27
<0.51
<0.51
<0.51
<0.51
<0.51
<0.51
Acetopbenooe
<4.40
<1.76
<1.76
<1.76
<1.76
<1.76
<1.76
Methyl isobutyl ketone
3.65b
1.11*
1.28b
1.20
1.00"
1.21"
1.10
Isopherone
<3.01
<1.20
<1.20
<1.20
<1.20
<1.20
<1.20
Note: Final values are not Laboratory Method Blank corrected.
'Less than 10 times the level measured in the method blank.
''Below calibration curve.
-------
SECTION 7.0
REFERENCES
1. U. S. Environmental Protection Agency. Method 301, in Code of Federal Regulations.
Title 40, Part 63. Washington, D.C. Office of the Federal Register, July 1, 1987.
2. U.S. Environmental Protection Agency, Method 0011, in "Test Methods for Evaluating
Solid Waste, Physical/Chemical Methods, SW-846 Manual, 3rd ed." Document
No. 955-001-0000001. Washington, D.C. November 1986.
3. U.S. Environmental Protection Agency, Method 0010, in "Test Methods for Evaluating
Solid Waste, Physical/Chemical Methods, SW-846 Manual, 3rd ed." Document
No. 955-001-0000001. Washington, D.C. November 1986.
4. Mitchell, William J., Midgett, M. Rodney. "Means to Evaluate Performance
Stationary Source Test Methods." Environmental Science and Technology. 10:85,
January 1976.
5. U.S. Environmental Protection Agency. Method 2, in Code of Federal Regulations.
Title 40, Part 60, Appendix A.
6. U.S. Environmental Protection Agency. Quality Assurance Handbook for Air
Pollution Measurement Systems. Volume III, Staionary Source Specific Methods
(EPA 600/4-77-027b).
99
-------
Appendix A
Results from Preliminary Laboratory Study
-------
APPENDIX A
This appendix provides a description of the technical activities and results obtained for
the laboratory studies conducted on Work Assignment No. 67 on Contract No. 68-D1-0010,
entitled "Improvement and Testing of the DNPH Method for Aldehydes & Ketones," for the
period of performance between August 1993 and September 1994.
CONCLUSIONS AND RECOMMENDATIONS
Based on the work performed in the laboratory, the following conclusions may be
drawn from the results:
• Formaldehyde, acetaldehyde, propionaldehyde, methyl ethyl ketone,
acetophenone, and methyl isobutyl ketone are all stable in the aqueous spiking
solution for up to 62 days.
• Because 5 % or less of the recovered formaldehyde was found in the second
impinger regardless of whether the trains were dynamically or statically spiked,
the spiking procedure used does not significantly affect the results obtained for
formaldehyde.
• For all of the compounds studied other than formaldehyde, dynamic spiking
allowed the collection efficiency of the train to be more adequately evaluated
than static spiking and is the preferred spiking technique especially when very
volatile, water-purgeable compounds are being tested.
• Keeping the first two impingers in an ice bath generally resulted in higher
compound recoveries with less breakthrough into the second impinger and less
tautomer formation than when the first two impingers were kept warm.
Based on work performed in the laboratory, the following recommendations are made:
• Recoveries for acrolein were low probably due to the reactive nature of the
double bond. Alternate sampling and analytical methods should be pursued for
acrolein or modifications should be made to Method 00111 to stabilize acrolein.
Potential modifications to Method 00111 include using hexane to recover the
sample trains instead of methylene chloride.
A-l
-------
• Quinone performs inconsistently by Method 0011'. Alternate sampling and
analytical methods should be investigated for quinone,
• Methyl isobutyl ketone and methyl ethyl ketone are not efficiently collected by
the aqueous reagent. Alternate sampling and analytical methods, possibly using
sorbents should be investigated for these compounds. Alternatively,
modifications to Method 00111 such as using five or more reagent impingers,
sampling at lov/er flow rates, using a lower pH reagent (>2N HC1), etc., may
improve the performance of Method 00111 for these compounds.
• To obtain quantitative recoveries use 200 mL of reagent in the first impinger
followed by two impingers containing 100 mL when sampling high levels
(above 10 ppmv) of aldehydes and ketones and keep the impingers iced.
INTRODUCTION
Title I of the Clean Air Act (CAA) identifies 189 substances as toxic air pollutants
which must be monitored under several provisions of the CAA Amendments. Title I identifies
several members of the class of organic compounds consisting of aldehydes and ketones as
toxic compounds emitted from stationary sources. No test method for aldehydes and ketones
is currently validated to perform the required stationary source monitoring under 40 CFR
Part 60.
Radian Corporation is assisting the Methods Branch of the National Exposure Research
Laboratory (NERL) in evaluating sampling and analytical methods for measuring aldehyde and
ketone emissions from stationary sources. All aldehydes and ketones listed in Title I of the
CAA have been studied as part of this project.
Sampling and analytical methods that were evaluated in the laboratory were based on
the SW-846 Method 0011 for formaldehyde. SW-846 Method 0011 uses the EPA Method 52
sampling trains modified to collect gaseous and particulate pollutants from an emission source
in aqueous acidic 2,4-dinitrophenylhydrazine (DNPH). Aldehydes and ketones present in the
stack gas stream react with DNPH to form the dinitrophenylhydrazone derivative. Samples
A-2
-------
are then extracted with organic solvent, dried, concentrated, and exchanged into an appropriate
solvent for analysis by high performance liquid chromatography (HPLC).
Background
Prior activities on the aldehyde/ketone sampling and analysis program include the
following efforts:
• Synthesis of all of the hydrazone derivatives for the aldehydes and ketones listed
in the CAA, as well as the analytes listed in SW-846 Method 0011;1
• Study of the effect of pH on hydrazone derivative formation efficiency in
DNPH solution, at a pH of 0, 0.5, 1-0, and 2.0;
• Optimization of the HPLC analytical method to effectively separate the
hydrazones from one another for accurate quantification and to select an internal
standard for the analysis; and
• Confirmation of the chemical composition and purity of the hydrazone
derivatives which had been synthesized.
The following conclusions could be drawn from the previous studies:
• . The 2-chloroacetophenone hydrazone was not readily purified following the
standard derealization and recrystallization procedures. However,
2-chloroacetophenone has shown acceptable performance in laboratory
validation studies using the SemiVOST method,3 and in one field validation
study using the semiVOST method where 2-chloroacetophenone was
dynamically spiked in the field.4
• The acrolein hydrazone derivative converted to another form (referred to as "x-
acrolein", possibly a tautomer) during recrystallization using ethanol and in
contact with 02 in the air. This conversion also occurred during pH dependent
reaction rate studies.
• Three pairs of carbonyl compounds coeluted under the analytical conditions
chosen for the HPLC analysis: butyraldehyde and isobutyraldehyde,
acetophenone and c-tolualdehyde, and methyl isobutyl ketone (MIBK) and
jttolualdehyde.
A-3
-------
The solubility of the DNPH reagent in hydrochloric acid solution decreases
rapidly as the pH is increased.
At pH 0 (2N HC1), formaldehyde, acetaldehyde, acetone, propionaldehyde,
methyl ethyl ketone, valeraldehyde, m-tolualdehyde, jhtolualdehyde, MEBK,
hexaldehyde, and 2,5-dimethylbenzaldehyde average recoveries were betwecr.
80 and 120 percent.
Average recoveries for formaldehyde, acetaldehyde, propionaldehyde,
valeraldehyde, m-tolualdehyde, jBolualdehyde, MlBK, hexaldehyde, and
2,5-dimethylbenzaldehyde were not changed when the pH was increased to 0.5
(0.3N HC1).
The average recoveries for quinone and acrolein increased when the pH was
increased from 0 to 0.5.
The average recoveries for acetone and methyl ethyl ketone decreased when the
pH was increased from 0 to 0.5.
At pH 1 and 2 where the DNPH reagent was exhausted as indicated by the lack
of a DNPH peak in the HPLC chromatogram, the recoveries of the aromatic
aldehydes-benzaldehyde, m-tolualdehyde, and 2,5-dimethylbenzaldehyde-were
greater than 80%, indicating that the aromatic aldehydes effectively competed
with the more reactive aldehydes (formaldehjde and acetaldehyde) for DNPH
reagent, probably because the aromatic aldehydes are more stable in solution
than the other compounds studied.
Information on the reaction of aldehydes and ketones to form hydrazones under
different pH conditions is available, and information on the ability of the various aldehydes
and ketones listed in Title I of the CAA to form hydrazone derivatives is also available.
Under Work Assignment 13 (Contract No. 68-D1-0010), a successful field study was
completed at a fiberglass coating plant. However, during the laboratory and field studies,
several problems were observed:
• Ketones are not collected as efficiently as aldehydes. Also, ketones are more
likely to tautomerize than aldehydes.
• Certain polymeric substances containing formaldehyde are reported to
decompose in the absorbing solution and react with the DNPH.
A-4
-------
These questions were addressed in controlled laboratory studies and another field test was
conducted to provide a validated stationary source test method. Other laboratories have
encountered difficulties in the application of SW-846 Method 00111 to extensive lists of
analytes.
Objectives
The EPA Methods Branch is developing a test method for aldehydes and ketones in
emissions from stationary sources for use by the Office of Air Quality Planning and Standards
(OAQPS) in the regulatory process. The object of Work Assignment 67 was to provide a fully
validated source test method.
To achieve this goal, Radian carried out the following tasks:
• Determined the collection efficiency of the SW-846 Method 00111 sampling
train for the aldehydes and ketones listed in Title I of the CAA and studied the
effect of changing sampling conditions, including pH of the DNPH solution and
volume and temperature of the DNPH solution.
• Studied the stability of the DNPH solution and the derivatives in the DNPH
solution and in the methylene chloride extract.
• Studied the potential for interference from formaldehyde-containing polymeric
substances, including hexamethylenetetramine, paraformaldehyde, and trioxane.
Project Description
Studies have been performed to establish the purity of the hydrazone derivatives that
have been synthesized. The purity information is summarized in Table A-l. Tne purity of the
hydrazone derivatives was confirmed by melting point, HPLC analysis, GC analysis, and
GC/MS analysis. Melting points were determined for all the hydrazone derivatives. Most of
the hydrazones melted within one to four degrees of the values reported in the literature.
Hydrazones of 21 aldehydes and ketones were analyzed by HPLC to check purity. Seventeen
A-5
-------
Table A-l
AJdehyde/Ketone Hydrazone Derivative Purity Data
Analyzed ¦
Purity
(%>
Melting Point (°C)
HPLC
Retention
Time
(min)
Hydrazone
Carbonyl Compound
Formula
Measured
Literature
Aeetaldebyde
CHjCHO
>99.5
150
147
12.6
Acetone
CH,COCH,
>99.5
121
126
17.5
Acetopheoone
C,HjCOCH,
>99.5
243
NA
25.5
Acrolein
CHj=CHCHO
>99.5
162
165
14.8
Benzaldehyde
C,HjCHO
>99.5
235
237
23.0
ButyraJdehyde
CHjCHjCHjCHO
>99.5
117
122
21.6
Crotonaldehyde
CHjCH=CHCHO
>99.5
183
190
19.5
2,5-Dimethylbeazaldehyde
C.H^CHjJjCHO
>99.5
233
NA
28.1
Formaldehyde
HCHO
162
166
Heptaldehyde
C,H„CHO
>99.5
99
108
28.4
Hexaldehyde
CH,(CHj)4CHO
>99.5
100
104
26.5
Isobutyraldehyde
(CHj)jCHCHO
>99.5
171
182
20.6
Isophorone
C,HltO
>99.5
140
NA
29.9
Methyl ethyl ketone
CHjCOCHjCH,
97.5
110
117
21.4
22.0
4-Methyl-2-pentanone
(methyl isobutyl ketone,
MIBK)
CH,COCHjCH(CHj)j
>99.5
81
95
26.4
Propionaldehyde
CHjCHjCHO
>99.5
149
154
23.5
Quinone
c.HA
92.9
16.2
m-Tolualdehyde
C.H.O
>99.5
212
211
26.0
o-Tolualdebyde
C,HtO
>99.5
189
195
16.2
p-ToIualdehyde
C,H,0
>99.5
241
239
26.0
Valenddehvde
CH,(CH,),CHO
>99.5
104
106
24.4
Note: Data from Shriner, R.L., Fuson, R.C., Curtin, D.Y., Morrill, T.C. The Systematic Identification of Organic
Compounds. Sixth Edition. John Wiley & Sons, Inc., New York, New York. 1980.
A-6
-------
of the derivatives are 99% pure based on HPLC analysis at 360 nm. Because the hydrazone of
2-chloroacetophenone could not be purified to a level of more than 66% and because
2-chloroacetophenone shows acceptable performance in the semiVOST method,1,2 we
recommend that this compound be omitted from further study by SW-846 Method 00111
sampling methods.
A further check of the purity of the hydrazones was performed by gas chromatography
with flame ionization detection. Ten of the hydrazones (formaldehyde, butyraldehyde,
benzaldehyde, valeraldehyde, acetaldehyde, hexaldehyde, acetone, methyl ethyl ketone, and
propionaldehyde) were greater than 86% pure. The tolualdehydes, 2,5-dimethylbenz-
aldehyde, and acetophenone did not elute from the chromatographic column. The rest of the
aldehydes and ketones were less than 76% pure.
Several aldehyde/ketone hydrazones were synthesized in Radian's Specialty Chemicals
Group in Austin. The compounds shown in Table A-2 are currently available. In the
Specialty Chemicals Group, all hydrazones derivatives are purified by multiple
recrystallization and analyzed by HPLC, GC, GC/MS, NMR, IR, and melting point; all
standards are > 99 % pure.
Studies have also been performed to establish the optimum pH for reaction of
aldehydes/ketones to produce the hydrazone derivatives. From the pH studies, pH 0.5
appeared to be the best for most of the compounds studied. Raising the pH from 0 to 0.5 did
not appear to significantly affect the recoveries for formaldehyde, acetaldehyde,
propionaldehyde, valeraldehyde, m-tolualdehyde, p-tolualdehyde, MIBK, hexaldehyde, and
2,5-dimethylbenzaldehyde. Raising the pH from 0 to 0.5 appeared to increase the recovery of
butyraldehyde, acetophenone, o-tolualdehyde, benzaldehyde, quinone, and acrolein. Only the
recoveries of acetone and MEK decreased when the pH was raised to 0.5. In the laboratory
experiments which were performed, pH was 0.5 based on previous studies.
A-7
-------
Table A-2
Crystalline AJdehyde/Ketone-DNPH Derivatives
Purity
(%)
Acetaldehyde-DNPH
CAS No. 1019-57-4
CtH,N404
M.W. 224.18
99
Acetone-DNPH
CAS No. 1567-89-1
QH.oNA
M.W. 238.20
99
Acrolein-DNPH
CAS No. 888-54-0
QHtN404
M.W. 236.19
99
Benzaldehyde-DNPH
CAS No. 1157-84-2
c„h10n4o4
M.W. 286.25
99
2-Butanone (MEKJ-DNPH
CAS No. 958-60-1
C,oH,2N404
M.W. 252.23
99
n-Butyraldehyde-DNPH
CAS No. 1527-98-6
C10Hj2N4O4
M.W.252.23
99
Crotonaldehyde-DNPH
CAS No. 1527-96-4
C.oH.^NA
MW 250.21
99
Formaldehyds-DNPH
CAS No. 1081-15-8
MW 210.15
99
Hexanal-DNPH
CAS No. 1527-97-5
c„h14n404
M.W. 280.28
99
Methaerolein-DNPH
CAS No. 5077-73-6
C10Hl(N4O4
M.W. 250.21
99
Propionaldehyde-DNPH
CAS No. 725-00-8
c»hi0n4o4
M.W. 238.20
99
m-ToIualdehyde-DNPH
CAS No. 2880-05-9
c14h12n4o4
M.W. 300.27
99
o-Tolualdehyde-DNPH
CAS No. 1773-44-0
c14hi2n4o4
M.W. 300.27
99
p-Tolualdchyde-DNPH
CAS No. 2571-00-8
C„HiaN404
M.W. 300.27
99
Valeraldehyde-DNPH
CAS No. 2057-84-3
c„hI4n404
M.W. 266.26
99
A-8
-------
The following activities were performed for Work Assignment 67:
• DNPH stability and derivative stability tests;
• Interference study; and
• Method 0011 train studies.
The following sections summarize the experimental results. The experimental
procedures are described in at the end of this appendix.
PRELIMINARY STUDIES
The preliminary studies included a reverse stability study and an interference study.
The stability study will be discussed first.
Stflhilitv Study
A reverse time study was conducted to evaluate the stability of pH 0.5 DNPH over
time. A test solution consisting of nine of the CAA aldehydes and ketones was used to test
reactivity: 2-chloroacetophenone was omitted from the list of carbonyl compounds in the
CAA. The reaction of 2-chloroacetophenone with DNPH appears to yield multiple products
and a pure derivative could not be obtained in derivatization studies. In addition, the
compound has shown acceptable results in laboratory and field studies using Method 0010.5
Table A-3 shows the experimental design of the stability study. In the reverse time
study, DNPH reagent was prepared. On Day 30, 8 aliquots of the DNPH solution were
selected. Four aliquots were designated as blanks; two were refrigerated and two were held at
room temperature. Four aliquots were spiked with the test solution and refrigerated. The
spiking procedure was repeated at Day 15, Day 7, Day 4, and Day 0. All samples were then
extracted, solvent-exchanged, and analyzed by HPLC to determine the effect of time upon the
reactivity of DNPH.
A-9
-------
Table A-3
Experimental Design for Studying the Stability of DNPH Solution
and Derivatives in the DNPH Solution
Number of Aliquots
Day
Spiked'
Unspikedb
4°C
Ambient 4°C
30
4
2 2
15
4
2 2
7
4
2 2
4
4
2 2
0
4
2 2
'All spiked samples will be stored in 500 mL wide-mouth amber bottles with
Teflon®-lined caps and sealed with Teflon® tape.
bAll unspiked aliquots will be stored in 250 mL narrow-neck amber bottles with
Teflon®-lined caps and sealed with Teflon® tape. (Reagent is generally stored in 1L
bottles with minimal headspace.)
The results of the DNPH stability test allowed the evaluation of the amount of time that
DNPH solution which has been prepared can be held until used, as well as the amount of time
that the collected samples can be held before extraction.
The spiked samples were solvent exchanged using the 15:4 method. The unspiked
samples were solvent exchange using the 1:1 method. Half of the spiked samples and half of
the unspiked samples were analyzed 3 times to allow for a statistical evaluation of the data.
Only half of the samples were analyzed in triplicate to save time and money. The results for
spiked sample results are presented as percent recovered in Table A-4. The results for the
unspiked samples are presented in total ng in Table A-5.
A-10
-------
Table A-4
Results of the Stability Study of the Derivatives in pH 0.5 DNPH
Average Recovery (%)
Compound Day 0 Day 4 Day 7 Day 15 Day 30
Formaldehyde
86
78
72
80
70
Acetaldehyde
94
89
88
86
68
Quinone
<1
2
3
13
51
Acrolein
31
26
29
28
45
Propionaldehyde
73
69
75
71
66
MEK
11
5
4
6
4
Acetophenone
38
102
106
101
99
MIBK
16
15
10
16
11
Isophorone
3
22
26
44
47
-------
Table A-5
Results of the Stability Study of the DNPH Reagent at pH 0.5
Total Micrograms
Stored at Ambient Temperature Stored Refrigerated at 4°C
Compound Day 0 Day 4 Pay 7 Pay If Day 30 Day 0 Day 4 Day 7 Pay 15 Pay 30
Formaldehyde
16
7
9
46
19
4
28
1
6
30
Acetaldehyde
<1
2
2
14
25
ND
<1
<1
2
7
Quinone
18
ND
<1
ND
<1
17
ND
ND
ND
<1
Acrolein
ND
ND
1
<1
ND
<1
ND
ND
ND
ND
Propionaldehyde
ND
ND
ND
ND
ND
I
ND
ND
ND
ND
MEK
<1
ND
19
6
ND
ND
ND
ND
ND
ND
Acetophenone
ND
ND
ND
ND
1
ND
ND
ND
ND
ND
MIBK
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Isophorone
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND = Not Detected
-------
Interference Study
Duplicate aliquots of DNPH at pH 4 were challenged with potential interferences such
as hexamethylenetetramine, trioxane, and paraformaldehyde. The DNPH aliquots were then
extracted, solvent-exchanged, and analyzed by HPLC. Blank DNPH was used as a control for
laboratory interferences. The results are reported in Table A-6. Saligenin and s-trioxane did
not interfere under the conditions tested. Dimethylolurea created a slight interference and
hexamethylenetetramine and paraformaldehyde significantly interfere with the determination of
formaldehyde. No other potential interferences were studied.
Table A-6
Results of Interference Study at pH 4.0
Formaldehyde Measured
Sample 1 Sample 2
Interferant
Area
Bias frig)
Area
Bias (me)
Dimethylolurea
88277
+6.4
82328
+5.6
Hexamethylenetetramine
331391
+36
382432
+42
Paraformaldehyde
315908
+34
534753
+61
Saligenin
ND
0
ND
0
s-Trioxane
ND
0
ND
0
ND = Not Detected
Spiking Solution Stability Studies
Recoveries from the sample trains using pH 4 reagent were consistently low. Several
explanations were possible: the spiking solution could be deteriorating, the dynamic spiking
apparatus could be failing to properly deliver the spiked aldehydes and ketones to the
impingers, or the reagent could be ineffective at efficiently converting the aldehydes and
A-13
-------
ketones to the hydrazines. To determine the cause for the low recoveries, an investigation
into the stability of the spiking solution was initiated and additional train experiments were
conducted.
Stability of the spiking solution was evaluated in two ways. First, a freshly prepared
and a two-month-old spiking solution were analyzed by GC/FID and the relative peak areas for
each component were compared. The results are shown in Table A-7. The percent bias
ranged from -9 for MIBK to +12% for acetaldehyde.
Second, the recoveries of reference spike samples using the old spiking solution at 41,
55, and 62 days were compared with reference spike sample recoveries of the new spiking
solution prepared at Day 0. These results are shown in Table A-8. Except for quinone and
acrolein, the recoveries on day 62 were equal or larger than the recoveries on Day 0. Quinone
was only detected on Day 0 and acrolein recoveries decreased by 40% after 62 days. Thus,
formaldehyde, acetaldehyde, propionaldehyde, methyl ethyl ketone, acetophenone, and methyl
isobutyl ketone derivatives were all stable in the aqueous spiking solution for up to 62 days.
Comparison of Dynamic and Static Spiking
To perform train studies for SW-846 Method 0011,1 a dynamic spiking system for
aldehydes/ketones was developed, constructed, and evaluated. Two approaches were
considered for spiking of an aqueous solution of the nine compounds: static spiking of an
aqueous solution, and dynamic spiking of an aqueous solution using a syringe pump. Dynamic
spiking was performed immediately after the probe.
After the dynamic spiking apparatus was constructed and installed in the SW-846
Method 00111 train, dynamic and static spiking procedures were compared using the
experimental design shown in Table A-9. Two trains were spiked statically by directly adding
tiie spiking solution to the first impinger. Another two trains were spiked dynamically using a
A-14
-------
Table A-7
Spiking Solution Stability Based on GC/FID Analysis*
Compound Peak Area*
Old Spiking Solution
(WA67-CDK-113093)
New Spiking Solution
(WA67-DST-013194)
Bias1
%Biasd
Acetaldehyde
1794557.3 + 400405.8
1608230.7 + 207361.6
+ 186,326.
6
11.59
Propionaldehyde
1653741.0 ± 272985.3
1512187.7 ± 317168.8
+ 141,553.
3
9.36
Acrolein
1855555.0 ± 380143.2
1833751.7 ± 188208.0
+21,803.3
1.19
Methyl Ethyl Ketone
Formaldehyde
3246638.3 ± 691584.8
2985691.7 ± 334098.6
+260,946.
6
8.74
Methyl Isobutyl Ketone
2672018.3 ± 481260.3
2929719.3 ± 229216.3
-257,701.0
-8.80
Acetophenone
Isophorone
3610432.7 ± 525417.9
3900593.0 ± 246395.9
-290,160.3
-7.44
* Quinone did not chromatograph under the conditions used.
b Average of triplicate analyses.
e Bias = Old Peak Area - New Peak Area
4 %Bias = Bias/New Peak Area x 100
-------
Table A-8
Comparison of Spiking Solution Recoveries with Time
Compound Old Spiking Solution (WA67-CDK-113093) New Spiking
Solution
(WA67-DST--
' 013194)
41 Days 55 Pays . 62 Days
Recovery
(%)
Bias*
toBias"
Recovery
<%>
Bias*
%Biask
Recovery
<%)
Bias*
%Biasb
DayO
Formaldehyde
86-106
-12 to +8
-12 to +8
102
+4
+4
101
+3
+3
98
Acetaldehyde
89-112
-2 to +21
-2 to +23
104
+ 13
+ 14
103
+ 12
+ 13
91
Quinone
ND
-35
-100
ND
-35
-100
ND
-35
-100
35
Acrolein
44-54
-33 to -23
-43 to -30
49
-28
-36
46
-31
-40
77
Propionaldehyde
81-99
+10 to +28
+ 14 to +39
84
+ 13
+18
91
+20
+28
71
Methyl Ethyl Ketone
10-20
-1 to +9
-9 to +82
6
-5
-45
11
0
0
11
Acetopbenone
16-36
-21 to -1
-57 to-3
.25
-12
-32
43
+6
+ 16
37
Methyl Isofautyl
Ketone
7-14
-4 to +3
-36 to +27
7
-4
-36
14
+3
+27
11
Isophorone
ND
0
NA
ND
0
NA
ND
0
NA
ND
* Bias = Day X - Day 0
%Bias = Bias/Day 0 x 100
NA = Not Applicable
-------
Table A-9
Experimental Design for the Comparison of Dynamic and Static Spiking
Procedures Using pH 0 Reagent Prepared with HC1
Sample Name
Temperature C°C)
Spike Amount
(mg)
Spiking Procedure
Train 1
0
1.5
Static
Train 2
0
1.5
Static
Train 3
0
1.5
Dynamic
Train 4
0
1.5 .
Dynamic
Reference Spike
RT
1.5
Static
Blank
RT
0.0
NA
RT = Room Temperature (approximately 20°C)
NA = Not Applicable
syringe pump. For quality control purposes, a reference spike and method blank sample were
also analyzed.
Results for static spiking are presented in Table A-10. Recoveries based on the
concentration of the spiking solution and volume of solution spiked were above 50% for
•formaldehyde, acetaldehyde, propionaldehyde, acetophenone, MIBK, and isophorone.
Quinone was either not detected or detected at levels that were too low to quantitate. Only
30% of the MEK was recovered and just slightly less than 50% of the acrolein. Over 94% of
the compounds recovered were recovered in the first impinger. When percent recoveries are
calculated versus the reference spike, recoveries range from 70 to 120% for formaldehyde,
acetaldehyde, acrolein, propionaldehyde, acetophenone, MEK and isophorone.
Results for dynamic spiking are presented in Table A-l 1. Recoveries based on the
concentration of the spiking solution and volume of solution spiked were above 50% and less
A-17
-------
Table A-10
Static Spike Train Recoveries Using pH 0 Reagent and Spiking at a Nominal 1.4 mg for
Each Compound
Percent of Spike Recovered (based on spiking solution concentration)
Reference -
Spike
Train 1
Train 2
Compound
Impinger 1
Impinger 2
Total
Impinger X
Impinger 2
Total
Formaldehyde
74
82
<1
82
86
<1
86
Acetaldehyde
82
73
3
76
79
4
83
Quinone
25
BQL
ND
BQL
ND
BQL
BQL
Acrolein
41
46
ND
46
49
ND
49
Propionaldehyde
70
66
<1
66
68
1
69
Methyl Ethyl
Ketone
91
30
1
31
31
2
33
Acetophenone
171
137
BQL
137
135
BQL
135
Methyl Isobutyl
Ketone
67
55
BQL
55
56
BQL
56
Isophorone
86
72
4
76
78
5
83
BQL = Below the quantitation Limit
ND = Not Detected
-------
Table A-ll
Dynamic Spike Train Recoveries Using pH 0 Reagent and Spiking at a Nominal 1.4 mg
for Each Compound
Percent Recovered (based on spiking solution concentration)
Reference •
Spike
Train 1
Train 2
Compound
Impinger 1
Impinger 2
Total
Impinger 1
Impinger 2
Total
Formaldehyde
74
257
13
270
118
6
124
Acetaldehyde
82
57
24
81
48
22
70
Quinone
25
BQL
BQL
BQL
BQL
BQL
BQL
Acrolein
41
30
7 "
38
23
7
30
Propionaldehyde
70
40
18
58
38
17
55
Methyl Ethyl
Ketone
91
14
20
34
8
11
19
Acetophenone
171
114
27
141
170
17
187
Methyl Isobutyl
Ketone
67
6
14
20
7
9
16
Isophorone
86
57
11
68
57
10
67
BQL = Below the quantitation Limit
-------
than 150% for acetaldehyde, propionaldehyde, and isophorone. Formaldehyde and
acetophenone had recoveries greater than 150% for one train out of the pair. Quinone was
detected at levels too low to be quantitated. Less than 40% of the acrolein, MEK, and MIBK
was recovered. Significant quantities of all of the compounds except for formaldehyde were
detected in the second impinger. For MEK and MIBK over 50% of the compound recovered
was recovered in the second impinger. When recoveries were calculated compared to the
reference spike, 73 to 99% of the acetaldehyde, acrolein, propionaldehyde, acetophenone, and
isophorone were recovered. Formaldehyde recoveries were greater than 150% and quinone,
MEK, and MIBK recoveries were less than 40 percent.
Table A-12 compares the average results for static and dynamic spiking. When
dynamically spiking the trains, a large positive bias in formaldehyde was observed. There are
at least two possible sources for this high bias-contamination of the sample during spiking,
sampling, recovery, preparation, or analysis and decomposition of one or more of the other
compounds into formaldehyde. If decomposition of one or more of the other compounds into
formaldehyde was occurring, a high positive bias would also be expected to be present for the
static trains. Because the static trains did not exhibit a high positive bias for formaldehyde, the
high positive bias for the dynamic trains was contributed to contamination. For the remaining
dynamic spiking trials, the glassware and spiking apparatus was cleaned thoroughly with
methylene chloride to eliminate any possible traces of methanol which can be contaminated
with formaldehyde. Regardless of the spiking procedure used, 5% or less of the recovered
formaldehyde was found in the second impinger, indicating that the spiking procedure does not
significantly affect the results obtained for formaldehyde.
Total recoveries for acetaldehyde were equivalent by the two spiking methods.
Interestingly, the distribution of the acetaldehyde in the train shifted. When dynamic spiking
was used, 30% of the recovered acetaldehyde was present in the second impinger versus only
4% when static spiking was used. Thus, although the spiking procedure does not affect the
overall performance of the train in recovering acetaldehyde, it does affect any conclusions
regarding breakthrough of acetaldehyde.
A-20
-------
Table A-12
Comparison of Dynamic and Static Spike Train Recoveries Using pH 0 Reagent
and Spiking at a Nominal 1.4 mg for Each Compound
Percent of Total Percent Recovered
Percent Recovered (based on Recovered in Second (based on reference
spike amount) Impinger spike)
Compound
Reference
Spike
Static
Trains"
Dynamic
Trains"
Static
Trains"
Dynamic
Trains"
Static
Trains"
Dynamic
Trains"
Formaldehyde
74
84
197
<1
5
114
266
Acetaldehyde
82
80
76
4
30
97
92
Quinone
25
BQL
BQL
NA
NA
BQL
BQL
Acrolein
41
48
34
0
20
116
83
Propionaldehyde
70
68
56
<2
31
96
81
Methyl Ethyl Ketone
91
32
26
4
58
35
29
Acelophenone
171
136
164
0
14
80
96
Methyl Isobutyl
Ketone
67
56
18
0
63
83
27
Isophorone
86
80
68
6
16
92
78
• Average of two trials
BQL - Below the quantitation Limit
NA = Not Applicable
-------
Quinone was not detected at this spike level by either spiking procedure although it was
detected in the reference spike. Additional tests were done at higher spike levels to determine
whether there was a threshold level at which quinone would react.
For acrolein, propionaldehyde, MIBK, and isophorone, the total recoveries were less
with dynamic spiking than with static spiking and significant quantities of the recovered
compounds were found in the second impinger. For these compounds, static spiking would
overestimate the performance of the train and could lead to false conclusions that the sampling
procedure is adequate for these compounds when in reality significant quantities of the
compound would not be recovered. Thus, for acrolein, propionaldehyde, MIBK, and
isophorone, dynamic spiking should be used for any evaluation and validation activities.
For MEK the overall recoveries for the dynamically spiked train varied from 19% to
34% so the recoveries were equivalent to or less than the recoveries for the statically spiked
trains and much more variable. Most of the recovered MEK (58%) was found in the second
impinger for the dynamically spiked trains, indicating that the impingers and DNPH reagent
do not collect MEK efficiently. The variability in the overall recoveries for the dynamically
spiked trains also indicate a lack of precision of this sampling method for MEK.
The acetophenone results were biased high in all of the samples. Interestingly, higher
recoveries were obtained for acetophenone when using dynamic spiking rather than static
spiking. However, with dynamic spiking 14% of the recovered acetophenone was found in the
second impinger, indicating that breakthrough occurs.
Because different results were obtained with some of the compounds when dynamic
spiking was used and dynamic spiking is more representative of what occurs in an actual
sampling situation, dynamic spiking was used for the remaining studies.
A-22
-------
Evaluation of Reagent Amount in First Impinger and Impinger Temperature on
Carbonyl Recoveries
The effect of the amount of reagent in the first impinger and impinger temperature
were evaluated using the experimental design shown in Table A-13, Four trains were
dynamically spiked with 15 mg of each carbonyl compound. For two of the trains, the first
impinger contained 100 mL of reagent and was kept in an ice bath during the entire sampling
period. For the other two trains, the first impinger contained 200 mL of reagent. For one of
these trains the first impinger was kept at room temperature during sampling and for the other
train the impinger was maintained in an ice bath. For quality control purposes, a reference
spike and method blank sample were also analyzed.
Table A-13
Experimental Design for the Evaluation of the Amount of Reagent
in the First Impinger and the Impinger Temperature Using pH 0 Reagent
Prepared with HC1 and Spiking 15 mg of Each Carbonyl
Sample Name
Temperature of First
Impinger <°€)
Reagent Amount in the
First Impinger (mL)
Train 5
0
100
Train 6
0
100
Train 11
RT
200
Train 12
0
200
Reference Spike
RT
100
Blank
RT
100
RT - Room Temperature (approximately 20 °C)
Results for comparison of the amount of reagent in the first impinger are reported in
Table A-14. Recoveries based on the concentration of the spiking solution and volume of
solution spiked improved for all of the compounds except quinone when the volume of reagent
A-23
-------
Table A-14
Spike Train Recoveries Using pH 0 Reagent and Spiking at a
Nominal 14 mg for Each Compound
Perecent of Spike Recovered
(based on spiking solution concentration)
100 mL in First Impinger 200 mL in First Impinger
Compound
Train 5
Train 6
Mean
Train 12
Bias
%Bias
Formaldehyde
45.5
53.8
49.6
106
+56.4
114
Acetaldehyde
27.0
37.9
32.4
61.8
+29.4
90.7
Quinone
50.5
57.9
54.2
54.5
+0.3
0.6
Acrolein
30.1
39.9
35.0
49.9
+ 14.9
42.6
Propionaldehyde
24.3
33.7
29.0
59.9
+30.9
107
Methyl Ethyl
Ketone
4.57
6.88
5.72
13.0
+7.28
127
Acetophenone
34.4
49.4
41.9
54.7
+ 12.8
30.5
Methyl Isobutyl
Ketone
5.26
8.88
7.07
14.6
+7.53
107
Isophorone
15.4
14.0
14.7
79.9
+65.2
444
BQL = Below the quantitation Limit
ND = Not Detected
in the first impinger was increased from 100 to 200 mL. The recovery for isophorone
quadrupled. Recoveries for formaldehyde, acetaldehyde, propionaldehyde, MEK, and MIBK
doubled. The recoveries for acrolein and acetophenone increased by 40 and 30%,
respectively. Thus, for sampling high levels (above 10 ppmv) of aldehydes and ketones, using
200 mL of reagent in the first impinger is recommended.
Results for comparison of the temperature of the first impinger reagent solution are
presented in Table A-15. Recoveries based on the concentration of the spiking solution and
volume of solution spiked were above 70% in the first impinger for formaldehyde and
A-24
-------
acetophenone regardless of whether the impinger was kept warm or cold. Recoveries and
breakthrough into the second impinger were unaffected by impinger temperature for
acetophenone, formaldehyde, and quinone. For isophorone the recoveries were unaffected by
impinger temperature but the breakthrough into the second impinger was lower when the
impingers were kept cold. For acetaldehyde and propionaldehyde the recoveries were higher
and the breakthrough was less when the impingers were kept cold. For acrolein, cold
impingers resulted in slightly better recoveries. In addition, less tautomer formed in the cold
impingers (5.78% versus 15.8% in the warm impingers). For MEK and MIBK the two cold
impingers recovered more compound. Interestingly, the breakthrough into the second
impinger was also higher when the impingers were cold. In general, for all of the compounds,
the train performs better (higher recoveries, less breakthrough, higher compound stability)
with the impingers cold.
EXPERIMENTAL PROCEDURES
This section focuses on the preparation, sampling and analysis procedures used during
the laboratory studies. The procedures will be discussed in an order consistent with the order
they would be performed in an actual situation: reagent preparation, sampling, sample
preparation, and finally analysis.
Preparation of 0.5 pH Reagent
To prepare the DNPH reagent used for the pH 0.5 laboratory studies, a 4 liter
container is placed under a fume hood on a magnetic stirrer. A large stir bar is added and the
container is filled half full with organic-free reagent water. A pipet is used to measure 6.5 mL
of concentrated sulfuric acid, which should be added to the stirring water slowly. Fumes may
be generated and the water may become warm. Approximately 15 to 20 g of DNPH crystals
are weighed on a one-place balance and added to the stirring acid solution. The 4 liter
container is filled with organic-free reagent water and allowed to stir overnight. If all the
A-25
-------
Table A-15
Spike Train Recoveries Using pH 0 Reagent and Spiking at a Nominal 14 mg for
Each Compound
Percent of Spike Recovered (hased on spiking solution concentration)
Impingers at Room Temperature Impingers in Ice Bath
Compound
Impinger 1
Impinger 2
Total
Breakthrough
(%)
Impinger 1
Impinger 2
Total
Breakthrough
(%)
Formaldehyde
95.9
2.9
98.80
2.94
106
2.5
108.
5
2.3
Aeetaldehyde
33.2
14.1
47.30
29.81
61.8
14.2
76.0
18.7
Quinone
55.3
2.2
57.50
3.83
54.5
1.7
56.2
3.0
Acrolein
40.3
0.2
40.50
0.49
49.9
0.6
50.5
1.2
Propionaldehyde
42.5
13.4
55.90
23.97
59.9
14.7
74.6
19.7
Methyl Ethyl Ketone
4.4
3.7
8.10
45.68
13.0
16.9
29.9
56.5
Acetophenone
52.7
13.1
65.80
19.91
54.7
11.6
66.3
17.5
Methyl Isobutyl
Ketone
6.2
6.1
12.30
49.59
14.6
19.2
33.8
56.8
Isophorone
74.5
15.7
90.20
17.41
79.9
10.2
90.1
11.3
-------
DNPH crystals have dissolved overnight, additional DNPH is added and the solution is stirred
for two more hours. The process of adding DNPH is continued with additional stirring until a
saturated solution is formed. The DNPH solution is filtered using gravity filtration and set
aside for the next step.
Approximately 1.6 liters of the DNPH reagent is placed in a 2 liter separatory funnel.
Approximately 200 mL of cyclohexane was added to the funnel. The stopper is then placed in
the funnel. The stopper is wrapper with paper towels to absorb any leakage. The funnel is
inverted and vented. The funnel is shaken vigorously for three minutes, venting initially every
10-15 seconds and then irregularly. After the layers have separated, the upper (organic) layer
is discarded.
The DNPH is extracted a total of three times. The clean DNPH solution is stored in
amber bottles that have been rinsed with acetonitrile and allowed to dry. The top of the amber
bottle has been capped with a teflon lined top and then sealed around the edges with teflon
tape.
Sample Preparation
The samples were prepared using the same method as the reagent preparation with a
few modifications. The sample was placed into an appropriate size separatory funnel (a
250-mL sample would be placed into a 500-mL separatory funnel). A small amount of
methylene chloride was added to the funnel. The funnel was stoppered. Paper towels were
wrapped around the stopper to absorb leakage. The funnel was inverted and venuxl. The
funnel was shaken vigorously for three minutes, venting initially every 10-15 seconds and then
irregularly. After the layers separated, the lower (organic) layer was placed into a volumetric
flask. The sample was extracted a total of three times. The extract solution was brought to
volume with methylene chloride and stored in an amber bottle rinsed with methylene chloride
and allowed to dry.
A-27
-------
The samples were then solvent exchanged. Fifteen milliliters of sample were placed
into a graduated centrifuge tube. The tube was placed on an N-evap and the solvent was
evaporated under nitrogen at room temperature to a volume of 0.5 mL. Volume was adjusted
with acetonitrile to a preordained volume. The solvent was again evaporated under nitrogen at
room temperature to a volume of 0.5 mL. Volume was readjusted with acetonitrile to a
preordained volume. This volume depended on the type of solvent exchange being performed.
The usual solvent exchange technique was abbreviated as 15:4. One starts with 15.0 mL
sample evaporates to 0.5 mL, adjusts volume to 8.0 mL, evaporates to 0.5 mL, and adjust
volume to 4.0 mL. Another technique is abbreviated as 1:1. One starts with 15.0 mL sample,
evaporates to 0.5 mL, adjusts volume to 15.0 mL, evaporates to 0.5 mL, and adjusts to
15.0 mL.
Sample Analysis
The samples were analyzed by HPLC components consisting of a Rainin HPLX solvent
delivery system, a Waters autosampler, and a Rainin Dynair.ax absorbance detector. The
mobile phase gradient is shown in Table A-16. The HPLC operating parameters are shown in
Table A-17. The analytes were located using retention times found in Table A-18.
REFERENCES
1. Test Methods for Evaluating Solid Waste-Phvsical/Chemical Methods. EPA SW-846,
3rd Edition, Method 0011, U.S. Environmental Protection Agency, Washington, DC.
2. Method 5 —Determination of Particulate Emissions from Stationary Sources, Federal
Register, Part 60, Appendix A, 742-766, July 1, 1991.
3. Laboratory Validation of VQST and Semi-VQST for Halogenated Hydrocarbons from
the Clean Air Act Amendment?? List, Volumes 1 and 2, U.S. Environmental Protection
Agency, 600/4-93/123a and b. NTIS PB93-227163 and PB93-227171, July 1993.
4. McGaughey, J.F., J.T. Bursey, and R.G. Merrill. "Field Test of Generic Method for
Halogenated Hydrocarbons." Prepared for the Atmospheric Research and Exposure
Assessment Laboratory, U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711, EPA-600/SR-93/101, September 1993, NTIS PB93-212181.
A-28
-------
5. Test Methods for Evaluating Solid Waste-Phvsical/Chemical Methods. EPA SW-846,
3rd Edition, Method CX)I0, U.S. Environmental Protection Agency, Washington, DC.
A-29
-------
Table A-l 6
HPLC gradient for analysis of DNPH-derivatized aldehydes
Time (min)
Acetonltrile (%)
Water (%)
MetbanoI(%)
0
20
40
40
25
5
25
70
40
5
15
80
62
5
15
80
64
20
40
40
74
20
40
40
Table A-17
HPLC operating parameters
Instrument
Rainin HPLX solvent delivery system
Data System
Nelson 2600
Column
Zorbax ODS (4.6 mm ID x 25 em), or equivalent with
pellicular ODS (2 mm ID x 2 cm) guard column, or equivalent
Mobile Phase
Acetonitrile/Water/Methanol
Gradient
See Table A-l6
Detector
Rainin Dynamax Absorbance Detector UV at 360 nm
Flow Rate
0.8 mL/min
Retention Time
See Table A-l8
A-30
-------
Table A-18
Retention Times for the Analytes
Analyte Retention Time Cmio)
Formaldehyde 11.3
Acetaldehyde 15.9
Quinone 19.5
Acrolein 21.6
Propionaldehyde 23.5
Methyl Ethyl Ketone 31.8
Acetophenone 41.7
Methyl Isobutyl Ketone 43.0
Isophorone 52.7
A-31
-------
Appendix B - Sampling and Analytical Methods
B.l Aldehyde and Ketone Sampling Checklist
-------
Table B.l-1. Aldehyde and Ketone Sampling Checklist
Before test starts:
1. Check impinger sets to verify the correct order, contents, orientation, and number of
impingers.
2. Check that the correct pieces of glassware are available and in good condition. Have at
least one spare probe liner, probe sheath, and meterbox ready to go at location.
3. Verify that a sufficient number of appropriate data sheets are available." Complete
required preliminary information including ambient temperature, barometric pressure,
and static pressure.
4. Examine meter boxes - level as necessary, zero the manometers and confirm that pumps
are operational.
5. Clean the stack access port to eliminate the chance of sampling deposited material.
6. Add probes to quad-train. Verify that the pitot tube and probes are properly positioned.
7. Check thermocouples - make sure they are reading correctly.
8. Perform initial leak checks; record leak rate and vacuum on sampling log.
9. Turn on variacs/heaters and check to see that the heat is increasing.
10. Add ice to impinger buckets.
11. Record the initial dry gas meter reading.
During test:
1. Notify crew chief of any sampling problems ASAP. Train operator should fill in
sampling log and document any abnormalities.
-------
Table B.l-1. Aldehyde and Ketone Sampling Checklist
6. Probe recovery (use 500-mL amber flint glass bottles)
a) Move the probes to a clean area, protected from wind to reduce chances of
contamination or losing sample. Recover sample probe using care to segregate
the four probes and trains.
b) Wipe the exterior of the probe to remove any loose material that could
contaminate the sample.
c) Carefully remove the nozzle/probe liner and cap it off with aluminum foil or
Teflon® tape.
d) Recover samples from each train as follows:
• Rinse the inside surface of the probe/nozzle assembly with deionized
water (DI H20). Brush with a Teflon® bristle brush until rinse shows no
visible particles or yellow color. Make a final rinse of the inside
surface.
• Recover DI H20 into a pre-weighed, pre-labelled sample container.
• With recovery bottle positioned at end of probe, wet all sides of probe
interior with DI H20.
• While holding the probe in an inclined position, put pre-cleaned Teflon®
brush down into probe and brush it in and out.
• Rinse the brush, while in the probe, with DI H20.
• Rinse at least 3 times until all the particulate has been recovered.
• Rinse the brush with DI H20 and collect these washings in the sample
bottle.
• After brushing, make a final rinse of the probe with DI H20.
-------
Table B.l-1. Aldehyde and Ketone Sampling Checklist
e) Rinse the nozzle/probe liner thoroughly with methylene chloride (MeClJ.
• With recovery bottle positioned at end of probe, wet all sides of probe
interior with MeCl2.
• Rinse the brush with MeCl2 and collect these washings in the sample
bottle.
7. Cap both ends of nozzle/probe liner for the next test, and store in a dry safe place.
8. Make sure data sheets are completely filled out legibly and give them to the Crew
Chief.
-------
Appendix B - Sampling and Analytical Methods
B.2 Aldehyde and Ketone Sampling Method
-------
METHOD XXXX
METHOD XXXX - SAMPLING AND ANALYSIS FOR
ALDEHYDE AND KETONE EMISSIONS FROM STATIONARY SOURCES
1.0 SCOPE AND APPLICATION.
1.1 Method XXXX is applicable to the collection and analysis of the aldehydes and
ketones listed in Table XXXX-1. This method has been validated for these pollutant
compounds at wood pressboard and polyester fiber manufacturing processes and is believed to
be applicable to other processes where these aldehydes and ketones may be emitted.
TABLE XXXX-1. LIST OF ANALYTES, CAS NUMBERS RETENTION TIMES,
AND DETECTION LIMITS
Method Detection
Compound Name
CAS No.'
Retention Time
(minutes)b
Limits (MDL)
(ppbv)c
Acetaldehyde
75-07-0
11.48
40
Acetophenone
98-86-2
28.99
10
Formaldehyde
50-00-0
8.38
90
Isophorone
78-59-1
38.22
10
Propionaldehyde
123-38-6
16.41
60
* Chemical Abstract Services Registry Number
b HPLC conditions: Reverse phase CI8 column, 4.6 x 250 mm; gradient elution using
acetonitrile/methanol/water (20:40:40, v/v/initial); flow rate 0.9 mL/min.; UV detector
360 nm, injector volume 25 jtL.
5 For an 849 Liter (30 cubic foot) sample, based on 10 times the levels detected in field
train blanks, or 10 times the instrument detection limit.
1.2 When this method is used to analyze unfamiliar sample matrices, compound
identification should be supported by at least one additional qualitative technique. A gas
XXXX-1
Revision 0
September 1995
-------
METHOD XXXX
The conditions permit the separation and measurement of aldehydes and ketones in the extract
by absorbance detection at 360 nanometers (nm).
3.0 DEFINITIONS. Reserved
4.0 INTERFERENCES.
4.1 A decomposition product of DNPH, 2,4-dinitroaniline, can be an analytical
interferant if the concentration is high. 2,4-Dinitroaniline can coelute with the
2,4-dinitrophenylhydrazone of formaldehyde under the HPLC conditions used for the analysis.
High concentrations of highly-oxygenated compounds, especially acetone, that have the same
retention time or nearly the same retention time as the dinitrophenylhydrazone of
formaldehyde, and that also absorb at 360 nm, will interfere with the analysis. Formaldehyde,
acetone, and 2,4-dinitroaniline contamination of the aqueous acidic 2,4-dinitrophenylhydrazine
(DNPH) reagent is frequently encountered. The reagent must be prepared within five days of
use in the field and must be stored in an uncontaminated environment both before and after
sampling in order to minimize blank problems. Some acetone contamination is unavoidable,
because background levels of acetone are widespread in laboratory and field operations. In
spite of these background levels, the acetone contamination must be minimized.
4.2 Dimethylolurea creates a slight interference. Hexamethylenetetramine and
paraformaldehyde significantly interfere with the determination of formaldehyde. O-
Tolualdehyde interferes with the determination of acetophenone because their hydrazones
coelute under the analytical conditions used. Acetone can interfere with the determination of
propionaldehyde if the hydrazones of the two compounds are not well resolved. High levels of
nitrogen dioxide can interefere by consuming all of the reagent.
XXXX-3
Revision 0
September 1995
-------
METHOD XXXX
4.3 Method interferences may be caused by contaminants in solvents, reagents,
glassware, and other sample processing hardware that lead to discrete artifacts and/or elevated
baselines in the chromatograms. All of these materials must be routinely demonstrated to be
ftee from interferences under the conditions of the analysis by analyzing laboratory reagent
blanks.
4.3.1 Glassware must be scrupulously cleaned. Clean all glassware as soon as
possible after use by rinsing with the last solvent used. This rinse should be followed
by washing with hot water and detergent, and rinsing with tap water and distilled
water. Glassware should then be drained and heated in a laboratory oven at 130°C
(266°F) for several hours before use. Solvent rinses using acetonitrile may be
substituted for the oven heating. After drying and cooling, glassware should be stored
in a clean environment to prevent any accumulation of dust or other contaminants.
4.3.2 The use of high purity reagents and solvents helps to minimize
interference problems. Purification of solvents by distillation in all-glass systems may
be required.
4.4 Formaldehyde analysis is expeeially complicated because, like acetone,
background levels are constantly encountered in laboratory and field operations.
4.5 Matrix interferences may be caused by contaminants that are coextracted from the
sample. The extent of matrix interferences will vary considerably from source to source,
depending upon the nature and diversity of the matrix being sampled. If interferences occur in
subsequent samples, some additional cleanup may be necessary.
4.6 The extent of interferences that may be encountered using liquid chromatographic
techniques has not been fully assessed. Although the HPLC conditions described allow for a
XXXX- 4
Revision 0
September 1995
-------
METHOD XXXX
resolution of the specific compounds covered by this method, other matrix components may
interfere.
5.0 SAFETY.
5.1 The toxicity or carcinogenicity of each reagent used in this method has not been
precisely defined; however, each chemical compound should be treated as a potential health
hazard. From this viewpoint, exposure to these chemicals must be reduced to the lowest
possible level by whatever means are available. The laboratory is responsible for maintaining
a current awareness file of Occupational Safety & Health Administration (OSHA) regulations
regarding the safe handling of the chemicals specified in this method. A reference file of
material safety data sheets (MSDSs) should also be made available to all personnel involved in
the chemical analysis. Additional references to laboratory safety are available.
5.2 Formaldehyde has been tentatively classified as a known or suspected human or
mammalian carcinogen.
6.0 EQUIPMENT AND SUPPLIES.
6.1 A schematic diagram of the sampling train is shown in Figure XXXX-1. This
sampling train configuration is adapted from EPA Method 4 procedures. The sampling train
consists of the following components: Probe Nozzle, Pitot Tube, Differential Pressure Gauge,
Metering System, Temperature Sensor, Barometer, and Gas Density Determination
Equipment.
6.1.1 Probe Nozzle. Quartz or glass with sharp leading edge at a tapered
30° angle. The taper shall be on the outside to preserve a constant inner diameter.
XXXX-5
Revision 0
September 1995
-------
METHOD XXXX
Figure XXXX-1. Sampling Train for Aldehydes and Ketones
XXXX-6
Revision 0
September 1995
-------
METHOD XXXX
The nozzle shall be of a buttonhook or elbow design. A range of nozzle sizes suitable
for isokinetic sampling should be available in increments of 0.16 cm (1/16 in), e.g.,
0.32 to 1.27 cm (1/8 to 1/2 in), or larger if higher volume sampling tains are used.
Each nozzle shall be calibrated according to the procedures outlined in Section 10.1.
6.1.2 Probe Liner. Borosilicate or quartz glass shall be used for the probe
liner. The tester should maintain the temperature in the probe at 120 ± 14°C
(248 ± 25°F).
6.1.3 Pitot Tube. Type S, as described in Section 2.1 of Promulgated EPA
Method 2 (Section 6.1 of Reformatted EPA Method 2), or other device approved by
the Administrator. The pitot tube shall be attached to the probe to allow constant
monitoring of the stack gas velocity. The impact (high pressure) opening plane of the
pitot tube shall be even with or above the nozzle entry plane (see EPA Method 2,
Figure 2-6b) during sampling. The Type S pitot tube assembly shall have a known
coefficient, determined as outlined in Section 4 of Promulgated EPA Method 2 (Section
10.0 of Reformatted EPA Method 2).
6.1.4 Differential Pressure Gauge. Two inclined manometers or equivalent
devices as described in Section 2.2 of Promulgated EPA Method 2 (Section 6.2 of
Reformatted EPA Method 2). One manometer shall be used for velocity-head readings
and the other for orifice differential pressure readings.
6.1.5 Temperature Sensor. A temperature sensor capable of measuring
temperature to within ± 3°C (± 5.4°F) shall be installed so that the temperature at the
impinger outlet can be regulated and monitored during sampling.
XXXX-7
Revision 0
September 1995
-------
METHOD XXXX
6.1.6 Impinger Train. The sampling train requires a minimum of five
impingers, connected as shown in Figure XXXX-1, with ground glass (or equivalent)
vacuum-tight fittings. For the first, third, fourth, and fifth impingers, use the
Greenburg-Smith design, modified by replacing the tip with a 1.27 cm (1/2 in.) inside
diameter glass tube extending to 1.27 cm (1/2 in.) from the bottom of the flask. For
the second impinger, use a Greenburg-Smith impinger with the standard tip.
6.1.7 Metering System. The necessary components are a vacuum gauge, leak-
free pump, temperature sensors capable of measuring temperature within 3°C (5.4°F),
dry gas meter (DGM) capable of measuring volume to within 1 %, and related
equipment as shown in Figure XXXX-1. At a minimum, the pump should be capable
of 4 cubic feet per minute (cfm) free flow, and the DGM should have a recording
capacity of 0-999.9 cubic feet with a resolution of 0.005 cubic feet. Other metering
systems may be used which are capable of maintaining sample rates within 10 percent
of isokinetic and of determining sample volumes to within 2%, subject to the approval
of the Administrator. The metering system may be used in conjunction with a pitot
tube to enable checks of isokinetic sampling rates.
6.1.8 Barometer. Mercury, aneroid, or other barometer capable of measuring
atmospheric pressure to within 2.5 mm Hg (0.1 in. Hg).
NOTE: The barometric pressure reading may be obtained from a nearby National
Weather Service Station. In this case, the station value (which is the absolute
barometric pressure) shall be requested and an adjustment for elevation
differences between the weather station and sampling point be made at a rate of
minus 2.5 mm (0.1 in.) Hg per 30 meters (100 ft.) elevation increase or plus
2.5 mm (0.1 in.) Hg per 30 meters (100 ft.) elevation decrease.
yyyy q
AAAA*O
Revision 0
September 1995
-------
METHOD XXXX
6.1.9 Gas Density Determination Equipment. Temperature sensor and pressure
gauge (as described in Sections 2.3 and 2.4 of Promulgated EPA Method 2 as well as
Sections 6.3 and 6.4 of Reformatted Method 2), and gas analyzer, if necessary, as
described in EPA Method 3. The temperature sensor shall, preferably, be permanently
attached to the pitot tube or sampling probe in a fixed configuration so that the tip of
the sensor extends beyond the leading edge of the probe sheath and does not touch any
metal. Alternatively, the sensor may be attached just prior to use in the field. Note,
however, that if the temperature sensor is attached in the field, the sensor must be
placed in an interference-free arrangement with respect to the Type S pitot openings (as
illustrated in Promulgated EPA Method 2, Figure 2-7, as well as Reformatted
Method 2, Figure 2-4). As a second alternative, if a difference of no more than 1 % in
the average velocity measurement is to be introduced, the temperature sensor need not
be attached to the probe or pitot tube (This alternative is subject to the approval of the
Administrator).
6.1.10 Viton A O-ring.
6.1.11 Heat Resistant Tape.
6.1.12 Teflon tape.
6.2 Sample Recovery. The following items are required for sample recovery.
6.2.1 Probe Liner and Probe Nozzle Brushes. Teflon bristle brushes with
stainless steel wire handles are required. The probe brush must have extensions of
stainless steel, Teflon, or inert material at least as long as the probe. The brushes must
be properly sized and shaped to brush out the probe liner, the probe nozzle, and the
impingers.
XXXX- 9
Revision 0
September 1995*
-------
METHOD XXXX
6.2.2 Wash Bottles. Three wash bottles are required. Teflon or glass wash
bottles are recommended; polyethylene wash bottles should not be used because organic
contaminants may be extracted by exposure to organic solvents used for sample
recoveiy.
6.2.3 Graduated Cylinder and/or Balance. These will be used to measure
condensed water to the nearest 1 mL or 0.5 g. Graduated cylinders must have divisions
not >2 mL. Laboratory balances capable of weighing to ±0.5 g are required.
6.2.4 Amber Flint Glass Storage Containers. One-liter wide-mouth amber flint
glass bottles voth Teflon-lined caps are required to store impinger water samples. The
bottles must be sealed with Teflon tape.
6.2.5 Rubber Policeman and Funnel. To aid in the transfer of material into
and out of containers in the field.
6.2.6 Cooler. To store and ship sample containers.
6.3 Reagent Preparation.
6.3.1 Bottles/Caps. Amber 1- or 4-L bottles with Teflon-lined caps are
required for storing cleaned DNPH solution. Additional 4-L bottles are required to
collect waste organic solvents.
6.3.2 Large Glass Container. At least one large glass container (8 to 16 L) is
required for mixing the aqueous acidic DNPH solution.
XXXX -10
Revision 0
September 1995
-------
METHOD XXXX
6.3.3 Stir Plate/Large Stir Bars/Stir Bar Retriever. A magnetic stir plate and
large stir bar are required to mix the aqueous acidic DNPH solution. A stir bar
retriever is needed for removing the stir bar from the large container holding the
DNPH solution.
6.3.4 BQchner Filter/Filter Flask/Filter Paper. A large filter flask (2-4 L) with
a buchner filter, appropriate rubber stopper, filter paper, and connecting tubing are
required for filtering the aqueous acidic DNPH solution prior to cleaning.
6.3.5 Separatory Funnel. At least one large separatory funnel (2 L) is required
for cleaning the DNPH prior to use.
6.3.6 Beakers. Beakers (150 mL, 250 mL, and 400 mL) are useful for
holding/measuring organic liquids when cleaning the aqueous acidic DNPH solution
and for weighing DNPH crystals.
6.3.7 Funnels. At least one large funnel is needed for pouring the aqueous
acidic DNPH into the separatory funnel.
6.3.8 Graduated Cylinders. At least one large graduated cylinder (1 to 2 L) is
required for measuring organic-free reagent water and acid when preparing the DNPH
solution.
6.3.9 Top-Loading Balance. A top loading balance readable to the nearest
0.1 g is needed for weighing out the DNPH crystals used to prepare the aqueous acidic
DNPH solution.
XXXX - 11
Revision 0
September 1995
-------
METHOD XXXX
6.3.10 Spatulas. Spatulas are needed for handling DNPH crystals when
preparing the aqueous DNPH solution.
6.4 Crushed Ice. Quantities ranging from 10-15 lb may be necessary during a
sampling run, depending upon ambient temperature. Samples must be stored and shipped
cold; sufficient ice for this purpose must be allowed.
6.5 Analysis.
6.5.1 Separatory Funnel. 250 mL, with Teflon stopcock.
6.5.2 Concentrator Tube. 15 mL graduated or equivalent. A ground glass
stopper may be used to prevent evaporation of extracts.
6.5.3 Vials. 10, 25 mL, glass with Teflon lined screw caps or crimp tops.
6.5.4 Analytical Balance. Capable of accurately weighing to the nearest
0.1 mg.
6.5.5 Glass Ampules. 1 mL in size. Used for storing stock aldehyde
derivative standard.
6.5.6 High Performance Liquid Chromatograph (modular).
6.5.6.1 Pumping system. Gradient with constant flow control capable
of 0.9 mL/min.
6.5.6.2 High Pressure Injection Valve with 25 ph loop.
XXXX -12
Revision 0
September 1995
-------
METHOD XXXX
6.5.6.3 Column. 250 mm x 4.6 mm ID, 5 pm particle size, €18 (or
equivalent).
6.5.6.4 Ultra-Violet (UV) Absorb an ce detector. 360 nm.
6.5.6.5 Strip Chart Recorder Compatible With Detector. Use of a data
system for measuring peak areas and retention times is recommended.
6.5.7 Volumetric Flasks. 250 or 500 mL.
6.5.8 Nitrogen blow down apparatus.
7.0 REAGENTS AND STANDARDS.
7.1 Reagent grade chemicals shall be used in all tests. Unless otherwise indicated, all
reagents shall conform to the specifications of the Committee on Analytical Reagents of the
American Chemical Society, where such specifications are available. Other grades may be
used, provided that the reagent is of sufficiently high purity to use without jeopardizing
accuracy.
7.2 Organic-free reagent water. All references to water in this method refer to
organic-free reagent water, as defined in Chapter One of SW-846 (see Reference 2 in
Section 16.0).
7.3 Silica Gel. Indicating type, 6-16 mesh. If the silica gel has been used previously,
dry at 180°C (350°F) for 2 hours before using. New silica gel may be used as received.
Alternatively, other types of desiccants (equivalent or better) may be used.
XXXX - 13
Revision 0
September 1995
-------
METHOD XXXX
7.4 2,4-Dinitrophenylhydrazine (DNPH), [2,4]-(02N)2C6H3]NHNH2 - The moisture
content may vary from 10 to 30%.
7.4.1 The DNPH reagent must be prepared in the laboratory within five days
of sampling use in the field. DNPH can also be prepared in the field, with
consideration of appropriate procedures required for safe handling of solvent in the
field. When a container of prepared DNPH reagent is opened in the field, the contents
of the opened container should be used within 48 hours. All Moratory glassware must
be washed with detergent and water and rinsed with water, methanol, and methylene
chloride prior to use.
NOTE: DNPH crystals or DNPH solution should be handled with plastic gloves at all
times with prompt and extensive use of running water in case of skin exposure.
7.4.2 Preparation of Aqueous Acidic DNPH Derivatizing Reagent: Each batch
of DNPH reagent should be prepared and purified within five days of sampling,
according to the procedure described below.
NOTE: Reagent bottles for storage of cleaned DNPH derivatizing solution must be
rinsed with acetonitrile and dried before use. Baked glassware is not essential to
prepare DNPH reagent. The glassware must not be rinsed with acetone or
methanol or an unacceptable concentration of acetone or formaldehyde
contamination will be introduced. If DNPH is prepared in the field, exercise
caution to avoid acetone contamination.
7.4.2.1 Place an 8 L container under a fume hood on a magnetic stirrer.
Add a large stir bar and fill the container half full of organic-free reagent water.
Save the empty bottle from the organic-free reagent water. Start the stirring bar
XXXX - 14
Revision 0
September 1995
-------
METHOD XXXX
and adjust it to stir as-fast as possible. Using a graduated cylinder, measure
1.4 L of 12N hydrochloric acid. Slowly pour the acid into the stirring water.
Fumes may be generated and the water may become warm. Weigh the DNPH
crystals on a one-place balance (see Table XXXX-2 for approximate amounts)
and add to the stirring acid solution. Fill the 8-L container to the 8-L mark
with organic-free reagent water and stir overnight. If all of the DNPH crystals
have dissolved overnight, add additional DNPH and stir for two more hours.
Continue the process of adding DNPH with additional stirring until a saturated
solution has been formed. Filter the DNPH solution using vacuum filtration.
Gravity filtration may be used, but a longer time is required to filter the DNPH
solution. Store the filtered solution in an amber bottle at room temperature.
TABLE XXXX-2. APPROXIMATE AMOUNT OF CRYSTALLINE DNPH USED
TO PREPARE A SATURATED SOLUTION
Amount of Moisture
Weight Required per
in DNPH
8 L of Solution
10 weight percent
36 g
15 weight percent
38 g
30 weight percent
46 g
7.4.2.2 Within five days of proposed use, place about 1.6 L of the
DNPH reagent in a 2-L separatory funnel. Add approximately 200 mL of
methylene chloride and stopper the funnel. Wrap the stopper of the funnel with
paper towels to absorb any leakage. Invert and vent the funnel. Then shake
vigorously for 3 minutes. Initially, the funnel should be vented frequently
(every 10-15 sec). After the layers have separated, discard the lower (organic)
layer.
XXXX -15
Revision 0
September 1995-
-------
METHOD XXXX
7.4.2.3 Extract the DNPH a second time with methylene chloride and
finally with cyclohexane. When the cyclohexane layer has separated from the
DNPH reagent, the cyclohexane layer will be the top layer in the separatory
funnel. Drain the lower layer (the cleaned extract DNPH reagent solution) into
an amber bottle that has been rinsed with acetonitrile and allowed to dry.
7.4.3 Shipment to the Field; Tightly cap the bottle containing extracted DNPH
reagent using a Teflon-lined cap. Seal the bottle with Teflon tape. After the bottle is
labeled, the bottle may be placed in a friction-top can (paint can or equivalent)
containing a 1-2 inch layer of granulated charcoal and stored at 4°C until use.
7.4.3.1 If the DNPH reagent has passed the Quality Control criteria in
Section 9,2.4, the reagent may be packaged to meet necessary shipping
requirements and sent to the sampling area. If the Quality Control criteria are
not met the reagent solution may be re-extracted; or, the solution may be re-
prepared and the extraction sequence repeated.
7.4.3.2 If the DNPH reagent is not used in the field within five days of
extraction, an aliquot may be taken and analyzed as described in Section 11.3.
If the reagent meets the Quality Control requirements in Section 9.2.4, the
reagent may be used. If the reagent does not meet the Quality Control
requirements in Section 9.2.4, the reagent must be discarded and new reagent
must be prepared and tested.
7.5 Field Spike Standard Preparation. To prepare a formaldehyde field spiking
standard at 4.01 mg/mL, use a 500 ptL syringe to transfer 0.5 mL of 37% by weight of
formaldehyde (401 mg/mL) to a 50 mL volumetric flask containing approximately 50 mL of
water. Dilute to 50 mL with water.
XXXX - 16
Revision 0
September 1995
-------
METHOD XXXX
7.6 Hydrochloric Acid, HC1. Reagent grade hydrochloric acid (approximately 12N) is
required for acidifying the aqueous DNPH solution.
7.7 Methylene Chloride, CH2C12. Methylene chloride (suitable for residue and
pesticide analysis, GC/MS, HPLC, GC Spectrophotometry or equivalent) is required for
cleaning the aqueous acidic DNPH solution, rinsing glassware, recovery of sample trains, and
extracting samples.
7.8 Cyclohexane, C6H12. Cyclohexane (HPLC grade) is required for cleaning the
aqueous acidic DNPH solution.
NOTE: Do not use spectroanalyzed grades of cyclohexane if this sampling methodology
is extended to aldehydes and ketones with four or more carbon atoms.
7.9 Methanol, CH3OH. Methanol (HPLC grade or equivalent) is required for the
HPLC analysis.
7.10 Acetonitrile, CH3CN. Acetonitrile (HPLC grade or equivalent) is required for
rinsing glassware, solvent exchanging of the samples, and the HPLC analysis.
7.11 Purified derivatized aldehyde crystals are required for preparation of standards.
7.12 Ethanol (absolute), CH3CH2OH. HPLC grade or equivalent.
7.13 2,4-Dinitrophenylhydrazine (DNPH) (70% (W/W)), [2,4-(02N)2C6H3]NHNHJ,
in organic-free reagent water.
7.14 Formalin [37.6 percent (w/w)], formaldehyde in organic free reagent water.
XXXX -17
Revision 0
September 1995
-------
METHOD XXXX
7.15 Stock standard solutions.
7.15.1 Preparation of Calibration Standards for Chromatographic Analyses.
7.15.1.1 Stock Aldehyde Derivative Standard. Prepare a multi-
component stock aldehyde derivative standard at a concentration of 200 ng/^L
by weighing 40 mg (± 0.01 mg) of purified derivatized aldehyde crystals into
small vials, dissolving the crystals in acetonitrile, quantitatively transferring the
solution to a 200 mL volumetric flask and diluting to the line with acetonitrile.
From this stock solution, prepare 1-mL aliquots using 1-mL glass ampules.
Seal and store the aliquots at 0°C (32°F).
7.15.1.2 Calibration Standards. Prepare calibration standards by
diluting 12.5, 25, 150, 300, and 500 fiL of the multi-component stock solution
to 5 mL with acetonitrile to provide a standard curve with calibration points at
0.5, 1.0, 6, 12, and 20 ng//iL of derivative.
7.15.1.3 Check Standard. Prepare a check standard of 5 ng/^L of
derivative by taking 125 ftL of the 200 ng/jtL multi-component stock standard
and diluting to 5 mL with acetonitrile.
7.15.2 Standard solutions must be replaced after six months, or sooner, if
comparison with check standards indicates a problem.
8.0 SAMPLE COLLECTION, PRESERVATION, STORAGE AND TRANSPORT.
8.1 Because of the complexity of this method, field personnel should be trained in and
experienced with the test procedures in order to obtain reliable results.
XXXX - 18
Revision 0
September 1995
-------
METHOD XXXX
8.2 Laboratory Preparation.
8.2.1 All the components must be maintained and calibrated according to the
procedure described in APTD-0576 (Reference 3 in Section 16.0), unless otherwise
specified.
8.2.2 Weigh several 200 to 300 g portions of silica gel in airtight containers to
the nearest 0.5 g. Record on each container the total weight of the silica gel plus
containers. As an alternative to preweighing the silica gel, it may be weighed directly
in the impinger or sampling holder just prior to train assembly.
8.3 Preliminary Field Determinations.
8.3.1 Select the sampling site and the minimum number of sampling points
according to EPA Method 1 or other relevant criteria. Determine the stack pressure,
temperature, and range of velocity heads using EPA Method 2. Check the pitot lines
for leaks according to Promulgated EPA Method 2, Section 3.1 (Reformatted EPA
Method 2, Section 8.1). Determine the stack gas moisture content using EPA
Approximation Method 4 or its alternatives to establish estimates of isokinetic
sampling-rate settings. Determine the stack gas dry molecular weight, as described in
Promulgated EPA Method 2, Section 3.6 (Reformatted EPA Method 2, Section 8.6).
If integrated EPA Method 3 sampling is used for molecular weight determination, the
integrated bag sample shall be taken simultaneously with, and for the same total length
of time as, the sample run.
8.3.2 Based on the range of velocity heads, select a nozzle size that will
maintain isokinetic sampling rates below 28 L/min (1.0 cfm). Do not change the
nozzle during the run. Ensure that the proper differential pressure gauge is chosen for
XXXX -19
Revision 0
September 1995
-------
METHOD XXXX
the range of velocity heads encountered (as described in Section 2.2 of Promulgated ,
EPA Method 2, as well as Section 8.2 of Reformatted EPA Method 2).
8.3.3 Select a suitable probe liner and probe length so that all traverse points
can be sampled. Consider sampling from opposite sides of large stacks so a shorter
probe can be used.
8.3.4 Select a total sampling time greater than or equal to the minimum total
sampling time specified in the test procedures for the specific industry. A total
sampling time must be selected so that (1) the sampling time per point is not less than 2
minutes (or some greater time Interval as specified by the Administrator), and (2) the
sample volume taken (corrected to standard conditions) will exceed the required
minimum total gas sample volume. The latter is based on an approximate average
sampling rate.
8.3.5 The sampling time at each point shall be the same. It is recommended
that the number of minutes sampled at each point be an integer or an integer plus one-
half minute, in order to avoid timekeeping errors.
8.3.6 In some circumstances (e.g., batch cycles) it may be necessary to sample
for shorter times at the traverse points and to obtain smaller gas-volume samples. In
these cases, careful documentation must be maintained in order to allow accurate
concentration calculation.
XXXX-20
Revision 0
September 1995
-------
METHOD XXXX
8.4 Preparation of Collection Train.
8.4.1 During preparation and assembly of the sampling train, keep all openings
where contamination can occur covered with Teflon film or aluminum foil until just
prior to assembly or until sampling is about to begin.
8.4.2 Place 200 mL of purified DNPH reagent in the first impinger and
100 mL of reagent in the second and third impinger?. Leave the fourth impinger
empty. Transfer approximately 200 to 300 g of pre-weighed silica gel from its
container to the fifth impinger. Be careful to ensure that the silica gel is not entrained
and carried out from the impinger during sampling. Place the silica gel container in a
clean place for later use in the sample recovery. Alternatively, the weight of the silica
gel plus impinger may be determined to the nearest 0.5 g and recorded. For moisture
determination, weigh all of the impingers after filing them with reagent.
8.4.3 With a glass or quartz probe liner, install the selected nozzle using a
Viton A O-ring when stack temperatures are <260°C (500°F) and a woven glass-fiber
gasket when temperatures are higher. See Reference 3 in Section 16.0 for details.
Other connection systems utilizing either 316 stainless steel or Teflon ferrules may be
used. Mark the probe with heat-resistant tape or by some other method to denote the
proper distance into the stack or duct for each sampling point.
8.4.4 Assemble the train as shown in Figure XXXX-1. During assembly, do
not use any silicone grease on ground-glass joints upstream of the impingers. Use
Teflon tape, if required. A very light coating of silicone grease may be used on
ground-glass joints downstream of the impingers, but the silicone grease should be
limited to the outer portion [see APTD-0576 (Reference 3 in Section 16.0)] of the
ground-glass joints to minimize silicone grease contamination. If necessary, Teflon
XXXX-21
Revision 0
September 1995-
-------
METHOD XXXX
(1 in. Hg) vacuum. Alternatively, leak-cheek the probe with the rest of the
sampling train in one step at 381 mm Hg (IS in. Hg) vacuum. Leakage rates in
excess of (a) 4% of the average sampling rate or (b) >0.00057 m3/min (0.020
cfm), are unacceptable.
8.5.1.3 The following leak check instructions for the sampling train
described in APTD-0576 and APTD-0581 (References 3 and 4 of Section 16.0,
respectively) may be helpful. Start the pump with the fine-adjust valve fully
open and coarse-adjust valve completely closed. Partially open the coarse-
adjust valve and slowly close the fine-adjust valve until the desired vacuum is
reached. Do nol reverse direction of the fine-adjust valve, as liquid will back
up into the train. If the desired vacuum is exceeded, either perform the leak
check at this higher vacuum or end the leak check, as shown below, and start
over.
8.5.1.4 When the leak check is completed, first slowly remove the plug
from the inlet to the probe. When the vacuum drops to 127 mm (5 in. Hg) or
less, immediately close the coarse-adjust valve. Switch off the pumping system
and reopen the fine-adjust valve. Do not reopen the fine-adjust valve until the
coarse-adjust valve has been closed to prevent the liquid in the impingers from
being forced backward in the sampling line and silica gel from being entrained
backward into the fourth impinger.
8.5.2 Leak Checks During Sampling Run.
8.5.2.1 If, during the sampling run, a component change (i.e.,
impinger) becomes necessary, a leak check shall be conducted immediately after
the interruption of sampling and before the change is made. The leak check
XXXX - 23
Revision 0
September 1995
-------
METHOD XXXX
tape may be used to seal leaks. Connect all temperature sensors to an appropriate
potentiometer/display unit. Check all temperature sensors at ambient temperatures.
8.4.5 Place crushed ice around the impingers.
8.4.6 Turn on and set the probe heating system at the desired operating
temperature. Allow time for the temperature to stabilize.
8.5 Leak-Check Procedures.
8.5.1 Pre-test Leak Check.
8.5.1.1 After the sampling train has been assembled, turn on and set the
probe heating system to the desired operating temperature. Allow time for the
temperature to stabilize. If a Viton A O-ring or other leak-free connection is
used in assembling the probe nozzle to the probe liner, leak-check the train at
the sampling site by plugging the nozzle and pulling a 381 mm Hg (15 in. Hg)
vacuum.
NOTE: A lower vacuum pressure may be used, provided that the lower vacuum
pressure is not exceeded during the test.
8.5.1.2 If a heat resistant string is used, do not connect the probe to the
train during the leak check. Instead, leak-check the train by first attaching a
carbon-filled leak check impinger to the inlet and then plugging the inlet and
pulling a 381 mm Hg (15 in. Hg) vacuum. (A lower vacuum pressure may be
used if this lower vacuum pressure is not exceeded during the test.) Next
connect the probe to the train and leak-check at approximately 25 mm Hg
XXXX-22
Revision 0
September 1995
-------
<
METHOD XXXX
shall be performed according to the procedure described in Section 8.5.1, (
except that it shall be performed at a vacuum greater than or equal to the
maximum value recorded up to that point in the test. If the leakage rate is
found to be no greater than 0.00057 mVmin (0.020 cfm) or 4% of the average
sampling rate (whichever is less), the results are acceptable. If a higher leakage
rate is obtained, the tester must void the sampling run.
NOTE: Any correction of the sample volume by calculation reduces the integrity of the
pollutant concentration data generated and must be avoided.
8.5.2.2 Immediately after a component change and before sampling is
reinitiated, a leak check similar to a pre-test leak check must also be conducted.
8.5.3 Post-test Leak Check.
8.5.3.1 A leak check of the sampling train is mandatory at the
conclusion of each sampling run. The leak check shall be performed in
accordance with the same procedures as the pre-test leak check, except that the
post-test leak check shall be conducted at a vacuum greater than or equal to the
maximum value reached during the sampling run. If the leakage rate is found to
be no greater than 0.00057 m3/min (0.020 cfm) or 4% of the average sampling
rate (whichever is less), the results are acceptable. If, however, a higher
leakage rate is obtained, the tester shall record the leakage rate and void the
sampling run.
XXXX-24
Revision 0
September 1995
-------
METHOD XXXX
8.6 Sampling Train Operation.
8.6.1 During the sampling run, maintain an isokinetic sampling rate to within
10% of true isokinetic, below 28 L/min (1.0 cfm). Maintain a temperature around the
probe of 120° ± 14°C (248° ± 25°F).
8.6.2 For each run, record the data on a data sheet such as the one shown in
Figure XXXX-2. Be sure to record the initial DGM reading. Record the DGM
readings at the beginning and end of each sampling time increment, when changes in
flow rates are made, before and after each leak check, and when sampling is halted.
Take other readings indicated by Figure XXXX-2 at least once at each sample point
during each time increment and additional readings when significant adjustments (20%
variation in velocity head readings) necessitate additional adjustments in flow rate.
Level and zero the manometer. Because the manometer level and zero may drift due to
vibrations and temperature changes, make periodic checks during the traverse.
8.6.3 Clean the stack access portholes prior to the test run to eliminate the
chance of collecting deposited material. To begin sampling, remove the nozzle cap,
verify that the probe heating systems are at the specified temperature, and verify that
the pitot tube and probe are properly positioned. Position the nozzle at the first
traverse point with the tip pointing directly into the gas stream. Immediately start the
pump and adjust the flow to isokinetic conditions. Nomographs, which aid in the rapid
adjustment of the isokinetic sampling rate without excessive computations, are
available. These nomographs are designed for use when the Type S pitot tube
coefficient is 0.84 ± 0.02 and the stack gas equivalent density (dry molecular weight)
is equal to 29 ± 4. APTD-0576 (Reference 3 in Section 16.0) details the procedure
for using the nomographs. If the stack gas molecular weight and the pitot tube
XXXX-25
Revision 0
September 1995
-------
METHOD XXXX
11 o i
SIM! 1*1
Z 3 5 & * * 3 a O f
* *.
}ff E
I1!
.5 f
ill t
*" I o
C ~
lfc
If
si
if
IS
•
3
i £
i
r £
i
*
1 9
if E
: e
o
t
.Si _
1 "i °-
fii *
le| £
r
I l
I i
-I *
• e l,
*1 t*
j
i *
5 £
!* t
m
&
e
1
s
|!
» c
I
1
Figure XXXX-2. Field Data Sheet
Revision 0
XXXX - 26 September 1995
-------
METHOD XXXX
coefficient are outside the above ranges, do not use the nomographs unless appropriate
steps are taken tc compensate for the deviations.
8.6.4 When the stack is under significant negative pressure (equivalent to the
height of the impinger stem), take care to close the coarse-adjust valve before inserting
the probe into the stack in order to prevent liquid from backing up through the train. If
necessary, the pump may be turned on with the coarse adjust valve closed.
8.6.5 When the probe is in position, block off the openings around the probe
and stack access porthole to prevent unrepresentative dilu'ion of the gas stream.
8.6.6 Traverse the stack cross-section, as required by EPA Method 1. To
minimize the chance of extracting deposited material be careful not to bump the probe
nozzle into the stack walls when sampling near the walls when removing or inserting
the probe through the access porthole.
8.6.7 During the test run, make periodic adjustments to keep the temperature
around the probe at the proper levels. Add more ice and, if necessary, salt, to maintain
a temperature of <20°C (68°F) at the silica gel outlet. Also, periodically check the
level and zero of the manometer.
8.6.8 A single train shall be used for the entire sampling run, except in cases
where simultaneous sampling is required in two or more separate ducts; at two or more
different locations within the same duct; or, in cases where equipment failure
necessitates a change of trains. Additional train(s) may also be used for sampling when
the capacity of a single train is exceeded.
XXXX-27
Revision 0
September 1995'
-------
METHOD XXXX
8.6.9 When two or more trains are used, components from each train shall be
analyzed. If multiple trains have been used because the capacity of a single train would
be exceeded, first impingers from each train may be combined, and second impingers
from each train may be combined.
8.6.10 At the end of the sampling run, turn off the coarse adjust valve, remove
the probe and nozzle from the stack, turn off the pump, record the final dry gas meter
reading, and conduct a post-test leak check as outlined in Section 8.5.3. Also, leak
check the pitot lines as described in EPA Method 2 (Section 8,1 of Reformatted
Method 2). The lines must pass this leak check in order to validate the velocity-head
data.
8.6.11 Calculate percent isokineticity (as described in Section 6.11 of
Method 5, as well as see Section 12.11 of Reformatted Method 5) to determine whether
the run was valid or another test should be performed.
8.7 Sample Recovery.
8.7.1 Preparation.
8.7.1.1 Proper cleanup procedure begins as soon as the probe is
removed from the stack at the end of the sampling period. Allow the probe to
cool. When the probe can be handled safely, wipe off all external particulate
matter near the tip of the probe nozzle and place a cap over the tip to prevent
losing or gaining particulate matter. Do not cap the probe tip tightly while the
sampling train is cooling because a vacuum will be created drawing liquid from
the impingers back through the sampling train.
XXXX-28
Revision 0
September 1995
-------
METHOD XXXX
8.7.1.2 Before moving the sampling train to the cleanup site, remove
the probe from the sampling train and cap the open outlet, being careful not to
lose any condensate that might be present. Remove the umbilical cord from the
last inipinger and cap the impinger. If a flexible line is used, let any condensed
water or liquid drain into the impingers. Cap off any open impinger inlets and
outlets. Ground glass stoppers, Teflon caps, or caps of other inert materials
may be used to seal all openings.
8.7.1.3 Transfer the probe and impinger assembly to an area that is
cleaned and protected from wind so that the chances of contaminating or losing
the sample are minimized.
8.7.1.4 Inspect the train before and during disassembly, and note any
abnormal conditions.
8.7.1.5 Save a portion of all washing solutions (methylene chloride,
water) used for cleanup as a blank. Transfer 200 mL of each solution directly
from the wash bottle and place each in a separate prelabeled sample "blank"
container (see Section 8.7.2.2).
8.7.2 Sample Containers.
8.7.2.1 Container 1: Probe and Impinger Catches. Using a graduated
cylinder, measure to the nearest mL, and record the volume of the solution in
the first four impingers. Alternatively, the solution may be weighed to the
nearest 0.5 g. Include any condensate in the probe in this determination.
Transfer the impinger solution from the graduated cylinder into the amber flint
glass bottle. Taking care that dust on the outside of the probe or other exterior
XXXX-29
Revision 0
September 1995
-------
METHOD XXXX
surfaces does not get into the sample, clean all surfaces to which the sample is
exposed (including the probe nozzle, probe fitting, probe liner, all impingers,
and impinger connectors) with methylene chloride. Use less than 500 mL for
the entire wash. Add the washing to the sample container.
8.7.2.1.1 Carefully remove the probe nozzle and rinse the inside
surfats with methylene chloride from a wash bottle. Brush with a
Teflon bristle brush, and rinse until the rinse shows no visible particles
or yellow color. Make a final rinse of the inside surface. Brush and
rinse the inside parts of the Swagelok fitting with methylene chloride the
same way.
8.7.2.1.2 Rinse the probe liner with methylene chloride. While
squirting the methylene chloride into the upper end of the probe, tilt and
rotate the probe so that all inside surfaces will be wetted wiJi methylene
chloride. Let the methylene chloride drain from the lower end into the
sample container. The tester may use a funnel (glass) to aid in
transferring the liquid washes to the container. Follow the rinse with a
Teflon brush. Hold the probe in an inclined position, and squirt
methylene chloride into the upper end as the probe brush is being pushed
with a twisting action through the probe. Hold the sample container
underneath the lower end of the probe, and catch any methylene
chloride, water, and particulate matter that is brushed from the probe.
Run the brush through the probe three times or more. Rinse the brush
with methylene chloride or water, and quantitatively collect these
washings in the sample container. After the brushing, make a final rinse
of the probe as described above.
XXXX - 30
Revision 0
September 1995
-------
METHOD XXXX
: Two people should clean the probe in order to minimize sample losses.
Between sampling runs, brushes must be kept clean and free from
contamination.
8.7.2.1.3 Rinse the inside surface of each of the first three
impingers (and connecting tubing) three separate times. Use a small
portion of methylene chloride for each rinse, and brush each surface to
which the sample is exposed with a Teflon bristle brush to ensure
recovery of fine particulate matter. Water will be required for the
recovery of the impingers in addition to the specified quantity of
methylene chloride. There will be at least two phases in the impingers.
This two-phase mixture does not pour well and a significant amount of
the impinger catch will be left on the walls. The use of water as a rinse
makes the recovery quantitative. Make a final rinse of each surface and
of the brush, using both methylene chloride and water.
8.7.2.1.4 After all methylene chloride and water washings and
particulate matter have been collected in the sample container, tighten
the lid so the solvent, water, and DNPH reagent will not leak out when
the container is shipped to the laboratory. Mark the height of the fluid
level to determine whether leakage occurs during transport. Seal the
container with Teflon tape. Label the container clearly to identify its
contents.
8.7.2.2 Container 2: Sample Blank. Prepare a blank by using an
amber flint glass container and adding a volume of DNPH reagent and
methylene chloride equal to the total volume in Container 1. Process the blank
in the same manner as Container 1.
XXXX-31
Revision 0
September 1995
-------
METHOD XXXX
8.7.2.3 Container 3: Silica Gel. Note the color of the indicating silica
gel to determine whether it has been completely spent, and make a notation of
its condition. The impinger containing the silica gel may be used as a sample
transport container with both ends sealed with tightly fitting caps or plugs.
Ground-glass stoppers or Teflon caps may be used. The silica gel impinger
should then be labeled, covered with aluminum foil, and packaged on ice for
transport to the laboratory. If the silica gel is removed from the impinger, the
tester may use a funnel to pour the silica gel and a rubber policeman to remove
the silica gel from the impinger. It is not necessary to remove the small amount
of dust particles that may adhere to the impinger wall that are difficult to
remove. Since the gain in weight is to be used for moisture calculations, do not
use water or other liquids to transfer the silica gel. If a balance is available in
the field, the spent silica gel (or silica gel plus impinger) may be weighed to the
nearest 0.5 g.
8.7.2.4 Sample containers should be placed in a cooler, cooled by
(although not in contact with) ice at a temperature not to exceed 4°C. Sample
containers must be placed vertically and, because they are glass, protected from
breakage during shipment. Samples should be cooled during shipment so they
will be received at the laboratory at 4°C. It is recommended that samples be
extracted within 30 days of collection and that extracts be analyzed within 30
days of extraction.
8.8 Alternative Procedure.
8.8.1 Addition of a Filter to the Sampling Train. As a check on the survival of
particulate material through the impinger system, a filter can be added to the impinger
XXXX - 32
Revision 0
September 1995
-------
METHOD XXXX
train either after the second impinger or after the third impinger. Since the impingers
are in an ice bath there is no reason to heat the filter at this point.
NOTE: Any suitable medium (e.g., paper, organic membrane) may be used for the
filter if the material conforms to the following specifications.
1) The filter has at least 95% collection efficiency (<5% penetration) for 3 fim
dioctyl phthalate smoke particles. The filter efficiency test shall be conducted in
accordance with ASTM standard method D2986-71. Test data from the
supplier's quality control program are sufficient for this purpose.
2) The filter has a low aldehyde blank value (<0.015 mg formaldehyde/cm2 of
filter area). Before the test series, determine the average formaldehyde blank
value of at least three filters (from the lot to be used for sampling) using the
applicable analytical procedures.
8.8.2 Recover the exposed filter into a separate clean container and return the
container over ice to the laboratory for analysis. If the filter is being analyzed for
formaldehyde, the filter may be recovered into a container or DNPH reagent for
shipment back to the laboratory. If the filter is being examined for the presence of
particulate material, the filter may be recovered into a clean dry container and returned
to the laboratory.
9.0 QUALITY CONTROL.
9,1 Sampling. Sampling quality control procedures are listed in Table XXXX-3. See
Reference 5 in Section 16.0 for additional Method 5 quality control.
XXXX - 33
Revision 0
September 1995 •
-------
METHOD XXXX
TABLE XXXX-3. SAMPLING QUALITY CONTROL PROCEDURES
Criteria
Control Limits*
Corrective Action
Final Leak Rate
Dry Gas Meter Calibration
Individual Correction
Factor (y)
sO.00057 acmm or 4% of
sampling rate, whichever is
less.
Post average factor y agree
±5% of pre-factor.
Agree within 2% of average
factor.
Average Correction Factor 1.00 ± 1 %.
Intermediate Dry Gas Meter
Analytical Balance (top
loader)
Barometer
Calibrated every six months
against EPA standard.
0.1 g of NBS Class Weights.
Within 2.55 mm Hg of
mercury-in-glass barometer.
None: Results are
questionable and should be
compared with other (3)
train results.
Adjust sample volumes using
the factor that gives the
smallest volume.
Redo correction factor.
Adjust the dry gas meter and
recalibrate.
Repair balance and
recalibrate.
Recalibrate.
'Control limits are established based on previous test programs conducted by the EPA.
9.2 Analysis. The quality assurance program required for this method includes the
analysis of the field and method blanks, procedure validations, and analysis of field spikes.
The assessment of combustion data and positive identification and quantitation of formaldehyde
are dependent on the integrity of the samples received and the precision and accuracy of the
analytical methodology. Quality assurance procedures for this method are designed to monitor
the performance of the analytical methodology and to provide the required information to take
corrective action if problems are observed in laboratory operations or in field sampling
activities. Table XXXX-4 lists laboratory quality control procedures.
XXXX - 34
Revision 0
September 1995
-------
METHOD XXXX
TABLE XXXX-4. LABORATORY QUALITY CONTROL PROCEDURES
Parameter
Analytical
Method
Quality
Control
Check
Frequency
Acceptance
Criteria
Corrective
Action
Linearity
Check
HPLC
Run 5-
point
curve.
At setup or
when check
standard is
out-of-
range.
Correlation
coefficient
*0.995.
Check integ.,
reinteg. If
necessary
recalibrate.
Retention
Time
HPLC
Analyze
check
standard.
1/10
injections.
Within three
standard
deviations of
average
calibration
relative retention
time.
Check instr.
funct. for plug,
etc. Heat
column: Adjust
gradient.
Calibration
Check
HPLC
Analyze
check
standard.
1/10
injections
min. 2/set.
±15% of
calibration curve.
Check integ.,
remake std. or
recalib.
System
Blank
HPLC
Analyze
acetonitrile
1/day.
sO.l level of
expected analyte.
Locate source of
contam.;
reanalyze.
Method
Spike/
Method
Spike
Duplicate
HPLC
Analyze
spiked
DNPH.
1/set or
1/20 samples
±20% of spiked
amount.
Check integ.,
check instrument
function,
reanalyze,
reprepare if
possible.
Replicate
Analyses
HPLC
Re-inject
sample.
1/10 samples
or 1/set
±15% of first
injection
Check integ.,
check instrument
function,
reanalyze.
Method
Blank
HPLC
Analyze
DNPH
1/set or 1/20
samples
sO. 1 level of
expected analyte
Locate source of
contamination,
reanalyze,
reprepare if
possible.
XXXX-35
Revision 0
September 1995
-------
METHOD XXXX
9.2.1 Field Train Blanks. Field blanks must be submitted with the samples
collected at each sampling site. The field blanks include the sample bottles containing
aliquots of sample recovery solvents, methylene chloride and water, and unused DNPH
reagent. At a minimum, one complete sampling train will be assembled in the field
staging area, taken to the sampling area, and leak-checked at the beginning and end of
the testing (or for the same total number of times as the actual sampling train). The
probe of the blank train must be heated during the sample test. The train will be
recovered as if it were an actual test sample. No gaseous sample will be passed
through the blank sampling train.
9.2.2 Laboratory Method Blanks. A method blank must be prepared for each
set of analytical operations, to evaluate contamination and artifacts that can be derived
from glassware, reagents, and sample handling in the laboratory.
9.2.3 Field Spike. A field spike is performed by introduction of 200 fiL of the
Field Spike Standard into an impinger containing 200 mL of DNPH solution. Standard
impinger recovery procedures are followed and the spike is used as a check on field
handling and recovery procedures. An aliquot of the field spike standard is retained in
the laboratory for derivatization and comparative analysis.
9.2.4 Preparation of DNPH Reagent. Take two aliquots of the extracted
DNPH reagent. The size of the aliquots depends on the exact sampling procedure
used, but 100 mL is reasonably representative. To ensure that the background in the
reagent is acceptable for field use, analyze one aliquot of the reagent according to the
procedure in Section 11. Save the other aliquot of aqueous acidic DNPH for use as a
laboratory method blank when the analysis is performed.
XXXX - 36
Revision 0
September 1995
-------
METHOD XXXX
10.0 CALIBRATION AND STANDARDIZATION.
NOTE: Maintain a laboratory log of all calibrations.
10.1 Probe Nozzle. Probe nozzles must be calibrated before their initial use in the
field. Using a micrometer, measure the inside diameter of the nozzle to the nearest 0.025 mm
(0.001 in.). Make measurements at three separate places across the diameter and obtain the
average of the measurements. The difference between the high and low numbers shall not
exceed 0.1 mm (0.004 in.). When the nozzles become nicked, dented, or corroded, they must
be replaced. Each nozzle must be permanently and uniquely identified.
10.2 Pitot Tube Assembly. The Type S pilot tube assembly must be calibrated
according to the procedure outlined in Section 4 of Promulgated EPA Method 2 (Section 10.1
of Reformatted Method 2), or assigned a nominal coefficient of 0.84 if it is not visibly nicked
or corroded, and, if it meets design and intercomponent spacing specifications.
10.3 Metering System.
10.3.1 Calibration Prior to Use. Before its initial use in the field, the metering
system shall be calibrated according to the procedure outlined in APTD-0576 (see
Reference 3 of Section 16.0). Instead of physically adjusting the DGM dial readings to
correspond to the wet-test meter readings, calibration factors may be used to correct the
gas meter dial readings mathematically to the proper values. Before calibrating the
metering system, a leak check procedure may not detect leaks with the pump. For
these cases, the following leak check procedure will apply. Make a ten-minute
calibration run at 0.00057 m3/min (0.020 cfm). At the end of the run, take the
difference of the measured wet-test and dry-gas meter volumes and divide the
XXXX-37
Revision 0
September 1995
-------
METHOD XXXX
difference by 10 to get the leak rate. The leak rate should not exceed 0.00057 mVmin
(0.020 efm).
10.3.2 Calibration After Use. After each field use, check the calibration of the
metering system by performing three calibration runs at a single intermediate orifice
setting (based on the previous field test). Set the vacuum at the maximum value
reached during the test series. To adjust the vacuum, insert a valve between the wet-
test meter and the inlet of the metering system. Calculate the average value of the
calibration factor. If the value has changed by more the 5%, recalibrate the meter over
the full range of orifice settings, as outlined in APTD-0576 (Reference 3 of
Section 16.0).
10.3.3 Leak check of metering system. The portion of the sampling train from
the pump to the orifice meter (see Figure XXXX-1) should be leak checked prior to
initial use and after each shipment. Leakage after the pump will result in less volume
being recorded than is actually sampled. Use the following procedure. Close the main
valve on the meter box. Insert a one-hole rubber stopper with rubber tubing attached
into the orifice exhaust pipe. Disconnect and vent the low side of the orifice
manometer. Close off the low side orifice tap. Pressurize the system to 13 - 18 cm
(5-7 in.) water column by blowing into the rubber tubing. Pinch off the tubing and
observe the manometer for 1 minute. A loss of pressure on the manometer indicates a
leak in the meter box. Leaks must be corrected.
NOTE: If the DGM coefficient values obtained before and after a test series differ by
>5%, either the test series must be voided or the test series must be calculated
using whichever meter coefficient value (i.e., before or after) gives the lower
value of total sample volume.
XXXX-38
Revision 0
September 1995
-------
METHOD XXXX
10.4 Probe Heater. The probe heating system must be calibrated before its Initial use
in the field according to the procedure outlined in APTD-0576 (Reference 3 of Section 16.0).
Probes constructed according to APTD-0581 (Reference 4 of Section 16.0) need not be
calibrated if the calibration curves in APTD-0576 (Reference 3 of Section 16.0) are used.
10.5 Temperature Sensors. Each temperature sensor must be permanently and
uniquely marked on the casting. All mercury-in-glass reference thermometers must conform
to ASTM E-l 63C or 63F specifications. Temperature sensors should be calibrated in the
laboratory with and without the use of extension leads. If extension leads are used in the field,
the temperature sensor readings at the ambient air temperatures, with and without the extension
lead, must be noted and recorded. Correction is necessary if using an extension lead produces
a change >1.5%.
10.5.1 Impinger and DGM Temperature Sensors. For the temperature sensors
used to measure the temperature of the gas leaving the impinger train, a three-point
calibration at ice water, room air, and boiling water temperatures is necessary. Accept
the temperature sensors only if the readings at all three temperatures agree to ± 2°C
(+ 3.6°F) with those of the absolute value of the reference thermometer.
10.5.2 Probe and Stack Temperature Sensor. For the temperature sensors used
to indicate the probe and stack temperatures, a three-point calibration at ice water,
boiling water, and hot oil bath temperatures must be performed. Use of a point at
room air temperature is recommended. The thermometer and thermocouple must agree
to within 1.5% at each of the calibration points. A calibration curve (equation) may be
constructed (calculated) and the data extrapolated to cover the entire temperature range
suggested by the manufacturer.
XXXX-39
Revision 0
September 1995'
-------
METHOD XXXX
10.6 Barometer. Adjust the barometer initially and before each test series to agree to
within ±2.5 mm Hg (0.1 in. Hg) of the mercury barometer or the correct barometric pressure
value reported by a nearby National Weather Service Station (same altitude above sea level).
10.7 Triple-Beam Balance. Calibrate the triple-beam balance before each test series,
using Class S standard weights. The weights must be within ±0.5% of the standards, or the
balance must be adjusted to meet these limits.
10.8 Analytical Calibration.
10.8.1 Establish liquid chromatographic operating parameters to produce a
retention time equivalent to that indicated in Table XXXX-1. Suggested
chromatographic conditions are provided in Section 11.2. Prepare derivatized
calibration standards according to the procedure in Section 7.15.1. Calibrate the
chromatographic system using thj external standard technique (Section 10.8.2).
10.8.2 External Standard Calibration Procedure.
10.8.2.1 Analyze each derivatized calibration standard using the
chromatographic conditions listed in Section 11.2, and tabulate peak area
against concentration injected. The results may be used to prepare calibration
curves for each analyte listed in Table XXXX-1.
10.8.2.2 The working calibration curve must be verified on each
working day by the measurement of one or more calibration standards. If the
response for any analyte varies from the previously established responses by
more than 15% (see Section 12.8), the test must be repeated using a fresh
calibration standard, but only after it is verified that the analytical system is in
XXXX-40
Revision 0
September 1995
-------
METHOD XXXX
control. Alternatively, a new calibration curve may be prepared for that
compound. If an autosampler is available, it is convenient to prepare a
calibration curve daily by analyzing standards along with test samples.
10.8.2.3 Periodically use the check standard prepared in Section
7.15.1.3 to check the instrument response and calibration curve.
11.0 PROCEDURES.
11.1 Extraction of Stack Gas Samples.
11.1.1 Pour the sample into a separatory funnel, rinse the bottle three times
with methylene chloride, adding the rinses to the separatory funnel, and drain the
methylene chloride into a volumetric flask.
11.1.2 Extract the aqueous solution with two or three aliquots of methylene
chloride depending l the initial volume of methylene chloride present. If more than
100 mL of methylene chloride is present in the sample, use two aliquots, otherwise use
three. Add the methylene chloride extracts to the volumetric flask.
11.1.3 Fill the volumetric flask to the line with methylene chloride. Mix well
and remove an aliquot.
11.1.4 If high levels of formaldehyde (>2000 /ig/mL, derivatized) are present,
the extract can be diluted with mobile phase, otherwise the extract must be solvent
exchanged as described in Section 11.1.5. If low levels of formaldehyde are present
(<0.5 /xg/mL, derivatized), the sample should be concentrated during the solvent
exchange procedure.
XXXX-41
Revision 0
September 1995
-------
METHOD XXXX
11.1.5 Solvent exchange the methylene chloride to acetonitrile for analysis.
11.1.5.1 Evaporate an aliquot of the methylene chloride extract to near
dryness (s0.5 mL) at room temperature under a stream of pure nitrogen.
11.1.5.2 Add acetonitrile when the sample just reaches dryness. Add
3 mL more than the final sample volume.
11.1.5.3 Evaporate the sample to near dryness again.
11.1.5.4 Repeat Steps 11.1.5.2 and 11.1.5.3. After the third
evaporation step, bring the volume up to the final volume with
acetonitrile.
11.1.6 Transfer the organic extract to a bottle and store at 4°C (39°F).
11.2 Chromatographic Conditions.
Column:
Mobile Phase:
Gradient:
Flow Rate:
UV Detector:
Injector Volume:
C18, 250 mm x 4.6 mm ID, 5 /*m particle size
Acetonitrile/methanol/water
See Table XXXX-5
0.9 mL/min.
360 nm
25 /xL
XXXX-42
Revision 0
September 1995
-------
METHOD XXXX
TABLE XXXX-5. HPLC GRADIENT FOR ANALYSIS OF
DNPH-DERIVATIZED ALDEHYDES
Time
Acetonitrile
Water
Methanol
(min)
(%)
(%)
(%)
0
20
40
40
12
5
25
70
18
5
23
72
28
10
15
75
35
10
15
75
37
20
40
40
47
20
40
40
11.3 Analysis.
11.3.1 Analyze samples by HPLC, using conditions established in
Section 11.2. Table XXXX-1 lists the retention times and MDLs that were obtained
under these conditions. Other HPLC columns, chromatographic conditions, or
detectors may be used if the requirements for Section 9.2. are met, or if the data are
within the limits described in Table XXXX-1.
11.3.2 The width of the retention time window used to make identifications
should be based upon measurements of actual retention time variations of standards
over the course of a day. Three times the standard deviation of a retention time for a
compound can be used to calculate a suggested window size; however, the experience
of the analyst should weigh heavily in the interpretation of the chromatograms.
11.3.3 If the peak area exceeds the linear range of the calibration curve, a
smaller sample volume should be used. Alternatively, the final solution may be diluted
with acetontrile and reanalyzed.
Revision 0
XXXX - 43 September 1995
-------
METHOD XXXX
11.3.4 If the peak area measurement is prevented by the presence of observed
interferences, further cleanup is required. However, no method has been evaluated for
this procedure.
12.0 DATA ANALYSIS AND CALCULATIONS.
Carry out calculations, retaining at least one extra decimal figure beyond that of the
acquired data. Sound off figures after final calculations.
12.1 Nomenclature:
ACN = Volume of acetonitrile after solvent exchange (mL)
AIC = Acceptable Impurity Concentration Q*g/mL),
ALDC = Concentration of aldehyde in sample Qxg/mL)
ALDX = Total aldehyde in sample (jig)
Cf = Concentration of aldehydes in stack gas (mg/dscm)
EAC = Expected Analyte Concentration (ppbv)
FW = Formula weight of analyte (g/mole)
MeCl2 = Volume of methylene chloride before solvent
exchange (mL)
MVOL = Total volume of MeCl2 extract (mL)
RVOL = Volume of DNPH reagent that will be used in the
impingers (mL)
SVOL = Volume of air sampled at standard conditions (L)
V = Organic extract volume (mL)
VB(Itd) = volume of gas sample a measured by dry gas
meter, corrected to standard conditions, dscm
(dscf)
XXXX - 44
Revision 0
September 1995
-------
METHOD XXXX
12.2 Concentration of Aldehyde in Sample. A least squares linear regression analysis
of the calibration standards shall be used to calculate a correlation coefficient, slope, and
intercept. Concentrations are the X-variable, and response is the Y-variable.
12.3 Calculation of Total Weight of Aldehydes in the Sample. To determine the total
aldehyde in ng, use the following equation:
ALDt = ALDC x MVOL x iggjl Eq. XXXX-1
12.4 Aldehyde concentration in stack gas. Determine the aldehyde concentration in
the stack gas using the following equation:
„ K (total formaldehyde, mg)
cr = 7T Efl- XXXX-2
where:
K = 35.31 ftVm3 if Vm(rtd) is expressed in English units
= 1.00 m3/m3 if is expressed in metric units
12.5 Average Dry Gas Meter Temperature and Average Orifice Pressure Drop are
obtained from the data sheet.
12.6 Dry Gas Volume: Calculate Vn(tld) and adjust for leakage, if necessary, using the
equation in Section 6.3 of EPA Method 5.
12.7 Volume of Water Vapor and Moisture Content: Calculate the volume of water
vapor and moisture content from equations 5-2 and 5-3 of EPA Method 5.
XXXX - 45
Revision 0
September 1995'
-------
METHOD XXXX
12.8 Calculate the Acceptable Concentrations of Impurities in DNPH Reagent as
follows.
EAC x SVOL x — x (FW + 180)
IC = 0.1 x — x (RVOLx 1,000 ^ XXXX-3
FW
where:
0.1 is the acceptable contaminant concentration,
22.4 is a factor relating ppbv to g/L,
ISO is a factor relating underivatized to derivatized analyte,
1,000 is a unit conversion factor.
13.0 METHOD PERFORMANCE.
13.1 Method performance evaluation: The expected method performance parameters
for precision, accuracy, and detection limits are provided in Table XXXX-6,
13.2 The MDL concentrations listed in Table XXXX-1 were obtained using field train
blank sample results (formaldehyde, acetaldehyde, propionaldehyde) or instrument detection
limits (acetophenone and isophorone).
14.0 POLLUTION PREVENTION. Reserved
XXXX-46
Revision 0
September 1995
-------
METHOD XXXX
TABLE XXXX-6. EXPECTED METHOD PERFORMANCE BASED ON EPA
METHOD 301 VALIDATION TESTS
Compound
Precision
(% RSD)*
Bias
(Correction
Factor)*
Detection
Limit
(ppbv)c
Concentration
Level
(ppmv)
Test
Matrix
Formaldehyde
±8
1.11
90
20
Plywood
Dryer Vent
±9
1.10
70
2
Polyester
Spinner
Vent
Acetaldehyde
±9
1.26
40
9
Plywood
Dryer Vent
±17
1.24
40
4
Polyester
Spinner
Vent
Propionaldehyde
±8
1.25
60
2
Plywood
Dryer Vent
±13
1.29
20
2
Polyester
Spinner
Vent
Acetophenone
±8
1.11
10
2
Plywood
Dryer Vent
±11
1.09
10
2
Polyester
Spinner
Vent
Isophorone
±8
1.08
10
2
Plywood
Dryer Vent
±9
0.93
10
2
Polyester
Spinner
Vent
* Relative Standard Deviation (%) for dual spiked trains as calculated by EPA Method 301,
k Bias Correction Factor for dual spiked trains as calculated by EPA Method 301.
e Based on ten times the levels measured in the field train blank samples for a 849 L (30 cubic
foot) sample.
XXXX-47
Revision 0
September 1995
-------
METHOD XXXX
15.0 WASTE MANAGEMENT.
15.1 Disposal of Excess DNPH Reagent. Excess DNPH reagent may be returned to
the laboratory and recycled or treated as aqueous waste for disposal purposes.
2,4-Dinitrophenylhydrazine is a flammable solid when dry, so water should not be evaporated
from the solution of the reagent.
16.0 REFERENCES.
1. Federal Register, 1986, 51, 40643-40652; November 7.
2. EPA Methods 6010, 7000, 7041, 7060, 7131, 7421, 7470, 7740, and 7841,
Test Methods for Evaluating Solid Waste: Physical/Chemical Methods.
SW-846, Third Edition. September 1988, Office of Solid Waste and
Emergency Response, U.S. Environmental Protection Agency, Washington,
D.C. 20460.
3. Rom, Jerome J. Maintenance, Calibration, and Operation of Isokinetic Source
Sampling Equipment. Environmental Protection Agency. Research Triangle
Park, NC., 27711. APTD-0576. March 1972.
4. Martin, Robert M. Construction Details of Isokinetic Source-Sampling
Equipment. Environmental Protection Agency. Research Triangle Park, NC.,
27711. APTD-0581. April 1971.
5. Quality Assurance Handbook for Air Pollution Measurement Systems. Volume
III: Stationary Sources of Specific Methods (Interim Edition). U.S.
XXXX-48
Revision 0
September 1995
-------
METHOD XXXX
Environmental Protection Agency. Office of Research & Development,
Washington D.C., 20460. EPA/600/R-94-038c. April 1994.
6. U. S. Environmental Protection Agency. Method 301-Protocol for the Field
Validation of Emission Concentrations from Stationary Sources. Code of
Federal Regulations, Title 40, Part 63. Washington, D.D. Office of the
Federal Register, July 1, 1987.
17.0 TABLES, DIAGRAMS, FLOWCHARTS, AND VALIDATION DATA.
XXXX-49
Revision 0
September 1995
-------
METHOD XXXX
Figure XXXX-3. Aldehydes and Ketones by High Performance Liquid Chromatography
(HPLQ
XXXX-50
Revision 0
September 1995
-------
METHOD XXXX
Figure XXXX-3. Aldehydes and Ketones by High Performance Liquid Chromatography
(HPLC) (continued).
Revision 0
September 1995-
XXXX-51
-------
METHOD XXXX
Figure XXXX-3. Aldehydes and Ketones by High Performance Liquid Chromatography
(HPLC) (continued)
Revision 0
XXXX - 52 September 1995
-------
Appendix C
Site Survey Analysis Results
-------
APPENDIX C
SITE SURVEY ANALYSIS RESULTS
This appendix provides the analysis results of the site survey samples collected on
Work Assignment No. 67 on Contract No. 68-D1-0010 and on Work Assignment No. 12 on
Contract No. 68-D4-0022.
FIELD TEST SITE 1
Flue gas samples for aldehyde/ketone analysis were collected at a plywood veneer
manufacturing plant. The unit tested at this facility is a plywood veneer dryer used to dry the
product veneer before shipping. Preliminary sampling was performed during the pre-test site
survey. Formaldehyde, acetaldehyde, propionaldehyde, and acrolein were all detected in the
dryer stack gas at levels over ten times the method detection limit. Low concentrations of
other aldehydes and ketones, including methyl ethyl ketone and methyl isobutyl ketone were
also identified. Average concentrations of these compounds in the pre-test samples are shown
in Table C-l. Method detection limits and reagent blank analysis results are also shown, for
comparison.
Table C-l. Average Aldehyde and Ketone Concentrations in
Pretest Samples for Site 1
Run 1
Run 2
Reagent Blank
Method
Concentration
Concentration
Concentration
Detection
Compound
(ppbv)*
(ppbv)'
(ppbv)*
Limit (ppbv)*
Acetaldehyde
1400
1700
0.5
2.1
Acrolein
120
120
ND
2.0
Formaldehyde
2800
3500
0.5
2.20
Methyl Ethyl Ketone
13
14
ND
1.9
Methyl Isobutyl Ketone
8.6
4.7
ND
1.7
Propionaldehyde
62
71
ND
2.0
Ouinone
100
130
ND
1.6
"Concentrations shown are for a 30 ft3 gas sample.
ND = Not Detected
C-l
-------
FIELD TEST SITE 2
Flue gas samples for aldehyde/ketone analysis were collected at a polyester fiber
manufacturing plant. The emission source tested is a duct which carries air exhausted from
two fiber spinning machines. Preliminary samples were collected from the spinning machine
exhaust duct in a pre-test site survey. Formaldehyde, acetaldehyde, and propionaldehyde were
all detected in the samples. Average concentrations of these compounds in the pre-test samples
are shown in Table C-2.
Table C-2. Average Aldehyde and Ketone Concentrations in
Pretest Samples for Site 2
Run 1
Run 2
Reagent Blank
Method
Concentration
Concentration
Concentration
Detection
Compound
(ppbv)"
(ppbv)*
(ppbv)'
Limit (ppbv)*
Acetaldehyde
120
100
ND
2.1
Formaldehyde
14
13
2
2.2
Propionaldehyde
8
7
2
2.0
'Concentrations shown are for a 30 ft3 gas sample.
ND = Not Detected
C-2
------- |