EPA/600/R-92/003b
March 1992
EMISSIONS OF METALS, CHROMIUM AND NICKEL SPECIES, AND ORGANICS
FROM MUNICIPAL WASTEWATER SLUDGE INCINERATORS
VOLUME II: SITE 5 EMISSION TEST REPORT
HEXAVALENT CHROMIUM METHOD EVALUATION
Prepared by:
Robin R. Segall
Entropy Environmentalists, Inc.
Research Triangle Park, North Carolina 27709
William G. DeWees
DEECO, Inc.
Cary, North Carolina 27519
EPA Contract No. 68-CO-0027
Work Assignment No. 0-5
Technical Managers:
Harry E. Bostian, Ph.D.
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Eugene P. Crumpler
Office of Water
U.S. Environmental Protection Agency
Washington, D.C. 20460
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268

-------
DISCLAIMER
This material has been funded wholly or in part by the United States
Environmental Protection Agency's Risk Reduction Engineering Laboratory and Office
of Water under Contract No. 68-02-4442, Work Assignment No. 81; Contract No. 68-02-
4462, Work Assignment No. 90-108; and Contract No. 68-C0-0O27, Work Assignment No.
0-5. It has been subject to the Agency's review and it has been approved for publication
as an EPA document. Mention of trade names or commercial products does not
constitute an endorsement or recommendation for use.
ii

-------
FOREWORD
Today's rapidly developing and changing technologies and industrial products and
practices frequently carry with them the increased generation of materials that, if
improperly dealt with, can threaten both public health and the environment. The U.S.
Environmental Protection Agency is charged by Congress with protecting the Nation's
land, air, and water resources. Under a mandate of national environmental laws, the
agency strives to formulate and implement actions leading to a compatible balance
between human activities and the ability of natural systems to support and nurture life.
These laws direct the EPA to perform research to define our environmental problems,
measure the impacts, and search for solutions.
The Risk Reduction Engineering Laboratory is responsible for planning,
implementing, and managing research, development, and demonstration programs to
provide an authoritative, defensible engineering basis in support of the policies,
programs, and regulations of the EPA with respect to drinking water, wastewater,
pesticides, toxic substances, solid and hazardous wastes, and Superfund-related activities.
This publication is one of the products of that research and provides a vital
communication link between the research and the user community.
The problem of disposing of primary and secondary sludge generated at municipal
wastewater treatment facilities is one of growing concern. Sludge of this type may
contain toxics such as heavy metals and various organic species. Viable sludge disposal
options include methods of land disposal or incineration. In determining the
environmental hazards associated with incineration, the Risk Reduction Engineering
Laboratory and the Office of Water sponsored a program to monitor the emissions of
metals and organics from a series of municipal wastewater sludge incinerators. The
following document presents the final results from the Site 5 emissions test program.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
iii

-------
ABSTRACT
The U.S. Environmental Protection Agency (EPA) Office of Water (OW) has
drafted risk-based sewage sludge regulations under Section 405d of the Clean Water Act
and EPA's Risk Reduction Engineering Laboratory (RREL) has been assisting OW in
the collection of supporting data for the proposed regulations. Because of the associated
cancer risk, there is particular concern regarding chromium and nickel species in the
emissions from sludge incineration.
An RREL/OW research program was implemented to determine the ratios of
hexavalent to total chromium and nickel subsulfide to total nickel in sewage sludge
incinerator emissions under varied incinerator operating conditions. This report presents
the test results from the first of a series of five incinerator test sites. Four incinerators
tested under a previous project conducted by Radian Corporation are included in the
Site numbering convention used. Thus, the first site in the series tested under the
present project, covered by this report, is referred to as Site 5.
TTiree candidate sampling methods and two candidate analytical methods for
hexavalent chromium were assessed at this site. The conversion of hexavalent chromium
(Cr+S) to other valence states of chromium during sampling and sample storage was of
primary concern. All Method 5-type train samples and impinger train samples were
collected by PEI Associates, Inc. (PEI). Dilution train samples were collected by
Southern Research Institute.
Method 5-type train samples were analyzed by Technology Applications, Inc.'s
(TAI) staff under contract to EPA's Environmental Monitoring Systems Laboratory
(EMSL) in Cincinnati, Ohio. Dilution train samples, Method 5-type samples, and
impinger train samples were analyzed by Entropy Environmentalists, Inc. under contract
to RREL. TAI used an ion chromatograph with post column reaction (IC/PCR) and
inductively coupled argon plasmography/mass spectrometry (ICP/MS) to analyze the
Method 5-type samples. A stable chromium isotope (nCr+s) spiked onto the Method 5
filter prior to sample collection was used to assess conversion of Cr+S conversion during
sampling/sample recovery.
The samples analyzed by Entropy were collected using a dilution train on an 8.5
in X 11 in glass fiber filter, a Method 5-type sampling train on an 82 mm quartz fiber
filter, and an impinger sampling train with an alkaline impinger reagent. Since PEI did
not have a recirculating impinger train, the system that was evaluated in this test was an
impinger train without the recirculating system. The glass fiber filters and impinger
solutions were spiked with native hexavalent chromium (nCr+s) and a radioactively-
labeled chromium isotope (5ICr+s). The samples were analyzed for hexavalent chromium
by IC/PCR and for the radioactive isotopes by scintillation (gamma) counting.
iv

-------
Site 5 was a typical multiple hearth incinerator controlled by a
venturi/impingement tray scrubber system. Process samples were not collected for
analysis since the purpose of the test program was the evaluation of conversion of
hexavalent chromium during flue gas sampling.
Hexavalent chromium test data for the Method 5-type train samples analyzed by
TAI have not been released by EMSL, and are therefore not presented or discussed.
The preliminary method evaluation testing demonstrated that all sampling
methods had problems with conversion of hexavalent chromium during sample and
storage prior to analysis. EPA decided that the Method 5-type train and the
recirculating reagent impinger train would be further evaluated during testing at Site 6.
This report was submitted in fulfillment of Contract Nos. 68-02-4442, 68-02-4462,
and 68-C0-0027 with the Risk Reduction Engineering Laboratory under the sponsorship
of the U.S. Environmental Protection Agency.
v

-------
TABLE OF CONTENTS
Section	Page
Disclaimer		 ii
Foreword		 iii
Abstract	• • • 		 iv
List of Figures	 viii
List of Tables	 ix
Acknowledgements	 x
1.0 Introduction		1-1
2.0 Site 5 Test Summary		2-1
2.1	Testing program design		2-1
2.2	Test program results 		2-1
2.2.1	Dilution Train Sampling Approach	 2-3
2.2.2	Method 5-Type Sampling Approach	 2-3
2.2.3	Impinger Train Approach 	 2-7
2.2.4	Conclusion 	 2-10
3.0 Process Description and Operation	 3-1
4.0 Test Results	 4-1
4.1	Testing Program Design 			 4-1
4.2	Test Program Results 	 4-2
4.2.1	Flue Gas Conditions 	 4-2
4.2.2	Dilution Train Sampling Approach 	 4-3
4.2.3	Method 5-Type Sampling Approach	 4-3
4.2.4	Impinger Train Approach 	 4-8
4.3	Conclusions	 4-14
5.0 Sampling Procedures	 5-1
5.1	Dilution Train	 5-1
5.2	Method 5-Type Train 	 5-3
5.3	Impinger Train 		 5-7
5.4	EPA Methods 1,2,3, and 4	 5-11
vi

-------
TABLE OF CONTENTS (Continued)
Section	Page
6.0 Analytical Procedures 				6-1
6.1	First Test Day - June 9, 1989 		6-1
6.1.1	Dilution Train		6-1
6.1.2	Method 5-Type Train 		6-2
6.1.3	Impinger Train 		6-3
6.2	Second Test Day - August 3, 1989 		6-3
6.2.1	Dilution Train		6-4
6.2.2	Method 5-Types Train		6-4
6.2.3	Impinger Train 		6-4
7.0 Quality Assurance and Quality Control 		7-1
7.1	QA/QC Program Objectives 		7-1
7.2	Flue Gas Sampling and Analysis QC Results 	 7-1
7.2.1	Dilution Train Sampling 		7-2
7.2.2	Method 5-Type Sampling		7-2
7.2.3	Impinger Train Sampling		7-4
7.2.4	Native Hexavalent Chromium and slCr+6 Analysis 		7-5
References	 8-1
Appendix A Analytical Data	A-l
Appendix B Method for Determination of Hexavalent Chromium using
Recirculating Reagent Impinger Train	 B-l
vii

-------
LIST OF FIGURES
Number	Page
4-1	Radiochromatograms for the pretest and posttest spiked
Method 5-type samples 	 4-9
4-2	Radiochromatograms for the impinger train samples 	4-13
5-1	Dilution train sampling system	 5-2
5-2 Method 5-type sampling train	 5-5
5-3 Impinger sampling train	 5-9
5-4 Recirculating reagent sampling train 	5-10
viii

-------
LIST OF TABLES
Number	Page
2-1 Test program sampling matrix 			2-2
2-2 Hexavalent chromium recovery using dilution train from June 9, 1989 		2-4
2-3 Hexavalent chromium recovery using method 5-type train
from June 9 1989 		2-5
2-4 Hexavalent chromium recovery using method 5-type train
from August 3, 1989 		2-6
2-5 Hexavalent chromium recovery using impinger train from
the first development test		2-8
2-6	Hexavalent chromium recovery using impinger train from
second development test 		2-9
3-1	Site 5 incinerator operating conditions		3-1
4-1	Summary of flue gas conditions for site 5		4-2
4-2 Hexavalent chromium recovery using dilution train from June 9, 1989 	 4-4
4-3 Hexavalent chromium recovery using method 5-type train
from June 6, 1989 	 4-5
4-4 Recovery of hexavalent chromium for method 5-type train
from August 3, 1989 	 4-7
4-5 Hexavalent chromium recovery using impinger train from
the first development test		4-10
4-6	Recovery of hexavalent chromium from impinger train
for the second development test		4-12
5-1	Dilution train sampling conditions 	 5-4
5-2 Cr/Cr +6 glassware cleaning procedures	 5-6
5-3 Method 5-type train sample conditions	 5-7
5-4 Cr + 6/Cr teflon/glass components cleaning procedures	 5-9
5-5 Impinger train conditions		5-12
ix

-------
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the following invaluable contributions to the
efforts described in this report: Dr. Joseph E. Knoll of the Quality Assurance Division of
EPA for advice and assistance throughout the project and Dr. Scott C. Steinsberger
formerly of Entropy Environmentalists, Inc. for his tireless effort and ingenuity in
developing new methodologies.
x

-------
1.0 INTRODUCTION
The U.S. Environmental Protection Agency (EPA) Office of Water (OW) has
been developing new regulations for sewage sludge incinerators and EPA's Risk
Reduction Engineering Laboratory (RREL) has been assisting OW in the collection of
supporting data. There is particular concern regarding chromium and nickel species in
the emissions from incineration of municipal wastewater sludge because of the associated
cancer risk. OW has drafted risk-based sludge regulations under Section 405d of the
Clean Water Act which have been published for comment in the Federal Register.
Volume 54, No. 23, February 6, 1989. Final regulations are scheduled for promulgation
in January 1992.
The draft regulations are based on the risk incurred by the "most exposed
individual" (MEI). The MEI approach involves calculating the risk associated with an
individual residing for seventy years at the point of maximum ground level concentration
of the emissions just outside the incinerator facility property line. EPA's proposal for
regulating sewage sludge incinerators is based on ensuring that the increased ambient air
concentrations of metal pollutants emitted from sludge incinerators are below the
ambient air human health criteria. The increase in ambient air concentrations for four
carcinogenic metals, arsenic, chromium, cadmium, and nickel, are expressed as annual
averages. The concentrations are identified in the proposed regulations as Risk Specific
Concentrations (RSC). Both nickel and chromium emissions from sludge incinerators
presented a specific problem in establishing RSCs, because unknown portions of the
emissions of these metals are in forms which are harmful to human health. In
performing the risk calculations, EPA assumed that 1% of the emissions of chromium
from the sludge incinerators is in the most toxic form, hexavalent chromium. For nickel,
1-1

-------
EPA assumed that 100% of the nickel emissions are in the most toxic form, nickel
subsulfide.
Chromium is likely to be emitted in either the highly carcinogenic hexavalent state
(Cr+6) or in the noncarcinogenic trivalent state (Cr+3). Trivalent chromium has not been
shown to be carcinogenic and is toxic only at levels higher than those normally found in
sewage sludge incinerator emissions. Although hexavalent chromium (as the most
oxidized form) could be reasonably expected to result from combustion processes,
investigators speculate that most of the chromium is likely to be emitted in the trivalent
state.1 This is because hexavalent chromium is highly reactive, and thus likely to react
with reducing agents to form trivalent chromium.
Studies have been conducted to determine the potential for chromium in sewage
sludge to be converted to the hexavalent form. Analysis of laboratory combusted sludges
dosed with various levels of lime and ferric chloride revealed that the hexavalent to total
chromium ratio increased with lime dosage.1 One-hundred percent conversion of
trivalent chromium to hexavalent chromium was observed in several of the tests.1 These
tests indicate that when lime and ferric chloride are used as sludge conditioners, high
ratios of hexavalent to total chromium may be formed under certain incinerator
operating conditions.
EPA has previously sponsored emission testing studies for measurement of
hexavalent chromium at two sludge incinerators.2*3 For one site, the hexavalent
chromium concentrations were below the analytical detection limit; for the other site, a
hexavalent-to-total chromium ratio of 13% was calculated. The 1% value chosen for the
draft regulations may seem low. This is the result, however, of weighting various values
to give the most credible ones more influence. With this approach, lower values were
assigned a stronger contribution. The lack of a substantial data base on hexavalent
chromium emissions prompted the following statement in the EPA's Technical Support
Document for the Incineration of Sewage Sludge: "EPA plans to perform additional tests
of sewage sludge incinerator emissions for hexavalent chromium before this proposed
rule is finalized. The additional data should allow the Agency to better understand how
hexavalent chromium is generated in sewage sludge incinerators." There were no
1-2

-------
published EPA emission measurement test methods for the sampling and analysis of
hexavalent chromium. In addition, very little data exist on the conditions that may cause
its formation. The primary objectives of the RREL/OW research described in this
report were to conduct preliminary evaluations of available hexavalent chromium
sampling and analytical methodologies. The conversion of hexavalent chromium to other
valence states during sampling and sample storage was the primary concern. Three
sampling techniques were evaluated: 1) a Method 5-type sampling system that collects
the chromium species in a heated sampling probe and on a heated glass fiber filter, 2) an
impinger sampling train that collects the chromium species in an alkaline impinger
reagent, and 3) a dilution train that dilutes the sample gas 15-fold with ambient air and
collects the chromium species on an unheated glass fiber filter. The Method 5-type
samples and the impinger train samples were collected by PEI Associates, Inc. (PEI).
The dilution train samples were collected by Southern Research Institute (SRI). Half of
the Method 5-type train samples were analyzed by Technology Applications, Inc.'s (TAI)
staff under an EPA Contract with the Environmental Monitoring Systems Laboratory
(EMSL) in Cincinnati, Ohio. The remainder of the Method 5-type samples, and all of
the dilution train and the impinger train samples were analyzed by Entropy
Environmentalists, Inc. under contract to RREL.
TAI used an ion chromatograph with post column reaction (IC/PCR) and
inductively coupled argon plasmography/ mass spectrometry (ICP/MS) for analysis. A
stable chromium isotope (nCr+6) spike was used to assess hexavalent chromium
conversion. However, the analytical data were never released by EMSL, and are
therefore, not discussed in this report.
The samples analyzed by Entropy were collected using the dilution train, the
Method 5-type train, and the impinger sampling train. Since PEI did not have a
recirculating impinger train, the system evaluated in this test was an impinger train
without the recirculating system. A radioactively-labeled chromium spike (51Cr+6) was
used to assess hexavalent chromium conversion. The samples were analyzed by IC/PCR
for the hexavalent chromium and scintillation (gamma) counting for the radioactively-
1-3

-------
labeled spike. No process or control equipment operating data were collected during
this test program.
This report presents the method evaluation results for the hexavalent chromium
samples analyzed by Entropy. This test program was the first in a series of five test sites
(Sites 5, 6, 7, 8, and 9). Four incinerators tested under a previous project conducted by
Radian Corporation are included in the Site numbering convention used. Thus, the first
site in the series tested under the present project, covered by this report, is referred to as
Site 5.
The following sections present detailed descriptions of the testing and results from
the Site 5 program. Section 2.0 present a summary of the test results. Section 3.0
presents a short process description and process operating conditions. Section 4.0
provides a more detailed discussion of the sampling and analytical results. Section 5.0
describes the sampling procedures, Section 6.0 describes the analytical procedures, and
Section 7.0 describes the quality assurance/quality control (QA/QC) program.
1-4

-------
2.0 SITE 5 TEST SUMMARY
2.1	TESTING PROGRAM DESIGN
The primary objectives of the RREL/OW research described in this report were
to conduct preliminary evaluations of available hexavalent chromium sampling and
analytical methodologies. The conversion of hexavalent chromium to other valence
states during sampling and sample storage was of primary concern. Three sampling
techniques were evaluated: 1) a Method 5-type sampling system that collects the sample
in a heated sampling probe and on a heated filter, 2) an impinger sampling train that
collects the sample in an alkaline impinger reagent, and 3) a dilution train that dilutes
the sample gas 15-fold with ambient air and collects the sample on an unheated glass
fiber filter.
The emission testing at Site 5 was conducted on June 9 (first method evaluation
test day) and August 3, 1989 (second method evaluation test day). The test program
sampling matrix is shown in Table 2-1. Sampling was conducted at the outlet of the
venturi/impingement tray scrubber used to control the multiple hearth incinerator
emissions. The sampling and analytical methods used are described in detail in Sections
5.0 and 6.0, respectively.
2.2	TEST PROGRAM RESULTS
The emission results are summarized in this section; the run-by-run data are
presented in Section 4.0.
2-1

-------
TABLE 2-1. TEST PROGRAM SAMPLING MATRIX
Testing
Date
Sampling
Technique
Analytical
Techniques
No. of Samples Length of Run
Collected (min.)
Cr+< Spike
Timing
First Test Day (June 9, 1989)
6/9/89
Dilution
Train
IC/PCR
Scintillation
1
15
Before

n
n
1
30
Before

n
H
1
60
Before

n
n
1
120
Before

n
n
i
60
Before
6/9/89
Method 5-
type Train
IC/PCR
Scintillation
4
60
Before
6/9/89
Impinger
Train
IC/PCR
Scintillation
8
60
Before
Second Test Day (August 3, 1989)
8/3/89
Method 5-
type Train
IC/PCR
Scintillation
4
120
2-Unspiked
2-After

11
n
4
120
Before
8/3/89
Impinger
Train
IC/PCR
Scintillation
8
120
Before
2-2

-------
2.2.1 Dilution Train Sampling Approach
Five dilution train sample runs were conducted on the first test day. As shown in
Table 2-2, the recoveries seen for both the native and radioactively-labeled hexavalent
chromium spikes were similar, with averages of 76.6% and 77.4%, respectively. Thus,
approximately 25% of the native and labeled hexavalent chromium were converted to
trivalent chromium. The recoveries were fairly consistent from run-to-run and did not
appear to be related to the length of the sample run.
2.2.2 Method 5-Tvpe Sampling Approach
Method 5-type sampling runs were conducted on both test days. Some of the
Method 5-type samples collected were analyzed by TAI under contract to EMSL. The
data from these trains have not been released by EMSL and will not be presented in this
report.
Four samples from a Method 5-type quadruplicate "quad" train collected on June
9 were analyzed by Entropy. The filters were spiked with native and labeled hexavalent
chromium prior to sampling. As shown in Table 2-3, the average recoveries for these
chromium species were 29.5% and 54.8%, respectively, indicating that approximately
70% of the native hexavalent chromium and 45% of the labeled hexavalent chromium
were convened to trivalent chromium during sampling and/or sample recovery.
Because there were problems in recovering the native and labeled chromium from
the control samples (see Section 4) for the first Method 5-type sample run, the second
second set of Method 5-type test runs (August 3, 1989) included posttest spiking as well
as pretest spiking of the filters. Two quad-train runs were conducted. The analytical
procedures for measuring the labeled hexavalent chromium were also improved under a
different contract with EPA's Quality Assurance Division in the two months between the
first and second test dates. The results are presented in Table 2-4. For the flue-gas
exposed, spiked filters, the recoveries were 66.1% and 67.3% for the native Cr+<; and
91% of the recovered 51Cr was Cr+<. The recoveries for three exposed filter samples,

-------
TABLE 2-2. HEXAVALENT CHROMIUM RECOVERY USING DILUTION TRAIN
FROM JUNE 9,1989
Sample
Sample Time
% native Cr+< Recovered
ED
(min)
from Filter
Native Hexavalent Chromium Analytical Data
F-l
15
73.7
F-2
30
67.0
F-3
60 (interrupted)
64.5
F-4
120
99.2
F-5
60
76.8
Average

76.6

Sample
Sample Time
% 51Cr""6 Recovered
ED
(min)
from Filter
Radioactively-Labeled Chromium Analytical Data
F-l
15
67.1
F-2
30
76.5
F-3
60 (interrupted)
75.3
F-4
120
87.3
F-5
60
80.8
Average

77.4
2-4

-------
TABLE 2-3. HEXAVALENT CHROMIUM RECOVERY USING
METHOD 5-TYPE TRAIN FROM JUNE 9, 1989
Sample
Sample Time
% native Cr+6 Recovered
ID
(min)
from Filter
Native Hexavalent Chromium Analytical Data
F-A
60
19.1
F-B
60
34.3
F-C
60
38.5
F-D
60
25.9
Average

29.5

Sample
Sample Time
% 51Cr+6 Recovered
ID
(min)
from Filter
Radioactively-Labeled Chromium Analytical Data
F-A
60
43.0
F-B
60
57.8
F-C
60
63.2
F-D
60
55.0
Average

54.8
2-5

-------
TABLE 2-4. HEXAVALENT CHROMIUM RECOVERY USING
METHOD 5-TYPE TRAIN FROM AUGUST 3, 1989
Native Hexavalent Chromium
Sample Identity
Expected Found % of
(Hq) (M<3) Expected
Posttest Spiked Method 5-Type Filters (2 hr run)
Spiked Control
Spiked Control
Posttest Spike (B-2)
Posttest Spike (B-4)
Exposed Filter (B-l)
Exposed Filter (B-3)
10.8 9.4 87.0
10.8 8.9 82.4
13.7 9.0 66.1
13.7 9.2 67.3
0 2.9 NA
0 2.8 NA
Pretest Spiked Method 5-Type Filters (2 hr run)
Spiked Filter (1-1)
Spiked Filter (1-2)
Spiked Filter (1-4)
Spiked Filter FB
13.7 6.4 47.0
13.7 6.4 47.0
13.7	6.6 48.3
10.8	0.6 5.3
Radioactively-Labeled Hexavalent Chromium

Percent of Total
Sample Identity
51Cr+3 51cr+6
Posttest Spiked Method 5-Type Filters
Spiked Control
Spiked Control
Posttest Spike (B-2)
Posttest Spike (B-4)
Exposed Filter (B-l)
Exposed Filter (B-3)
2.3 97.7
1.5 98.5
11.5 88.5
9.0 91.0
NA NA
NA NA
Pretest Spiked Method 5-Type Filters
Spiked Filter (1-1)
Spiked Filter (1-2)
Spiked Filter (1-4)
Spiked Filter FB
38.5	61.5
37.4 62.6
26.7 73.3
68.6	31.4
2-6

-------
relative to the expected value of 13.7 ng, averaged 47.4%. Relative to the spiked control
filter samples, the recoveries averaged 70.9%. Of the recovered slCr, about 65% was
apparently in the hexavalent state. The results of the tests conducted on August 3 using
the improved technique indicated that, for the pretest and posttest spike respectively,
87% and 92% of the soluble radioactively-labeled chromium extracted from the filters
were in the hexavalent state.
The experimental results for the Method 5-type trains indicate that chromium
conversion does occur during sampling and/or sample recovery by extraction, and can be
measured, semi-quantitatively, using nCr+6 and/or native Cr+6 spikes.
2.2.3 Impinger Train Approach
Since PEI Associates, Inc. did not have a recirculating reagent impinger train, the
evaluation testing was conducted using only the impinger portion of the recirculating
reagent sampling train. On June 9, 1989, two quad-train runs were conducted. The
impinger reagent used was 80% isopropyl alcohol and 20% 2 N NaOH (IPA/NaOH).
As shown in Table 2-5, the average recoveries of the native and labeled hexavalent
chromium were 65.4% and 99.7%, respectively.
The results for the spiked impinger train sampling conducted on August 3 are
summarized in Table 2-6. The improved analytical technique, described in Section 4.0,
was used for the analysis of the labeled hexavalent chromium. Recoveries of spiked
native chromium (10.8 ng) from two IPA/NaOH samples with a 2-hr sampling period
averaged 78.3%. These recoveries constituted an average of 94.8% of a spiked
IPA/NaOH control sample where 8.9 ng of the native chromium was recovered. The
average recovery of spiked s,Cr+6 in the two samples at 78.4% was in agreement with the
average native Cr recovery; recovery in the control sample was higher at 89.8%.
A 0.5 M phosphate buffer used in the impinger train yield very low chromium
recoveries, about 25% for the native Cr+< and about 15% for s,Cr+< (see Table 2-6).
Both control and field blank samples yielded good recoveries of both native Cr+6 and
nCr+6. While the phosphate buffer did not prevent conversion during Sampling, the

-------
TABLE 2-5. HEXAVALENT CHROMIUM RECOVERY USING IMPINGER TRAIN
FROM THE FIRST DEVELOPMENT TEST
Sample
Sample Time
% native Cr+S Recovered
ID
(min)
from Solution
Native Hexavalent Chromium Analytical Data
1-1 (A-l)
60
87.0
1-2 (A-2)
60
59.0
1-3 (B-l)
60
24.5
1-4 (B-2)
60
73.7
1-5 (C-l)
60
69.2
1-6 (C-2)
60
74.4
1-7 (D-l)
60
74.0
1-8 (D-2)*
60
(21.1)
Average

65.4

Sample
Sample Time
% 51Cr+s Recovered
ID
(min)
from Solution
Radioactively Labeled Chromium Analytical Data
1-1 (A-l)
60
99.8
1-2 (A-2)
60
99.8
1-3 (B-l)
60
99.9
1-4 (B-2)
60
99.8
1-5 (C-l)
60
99.6
1-6 (C-2)
60
99.6
1-7 (D-l)
60
99.9
1-8 (D-2)'
60
(13.7)
Average

99.7
* The sample from run 1-8 had silica gel in the impinger solution; results not
included in average.
2-8

-------
TABLE 2-6. HEXAVALENT CHROMIUM RECOVERY USING IMPINGER TRAIN
FROM AUGUST 3, 1989
Native Hexavalent Chromium
Sample Identity
Expected Found % of % of
(Mg) (M9) Expected Control
Spiked 80% IPA/20% 1.0 N NaOH Impinger Reagent (2 hr run)
IPA/NaOH Control
IPA/NaOH Sample (1-2)
IPA/NaOH Sample (1-3)
IPA/NaOH Field Blank
10.8 8.9 82.5 NA
10.8 8.8 81.7 99.0
10.8 8.1 74.8 90.6
10.8 4.9 45.4 55.0
Spiked 0.5 M Phosphate Buffer Impinger Reagent (2 hr run)
0.5 M P04 Control
P04 Sample (P-l)
P04 Sample (P-2)
P04 Field Blank
10.8 11.2 103.5 NA
10.8 2.8 26.2 25.3
10.8 3.1 28.5 27.5
10.8 9.2 84.7 81.8
Radioactively-Labeled Chromium
Sample Identity
Percent of Total
5lCr+3 51cr+6
Spiked 80% IPA/20% 2 N NaOH Reagent (2 hr
IPA/NaOH Control
IPA/NaOH Sample (1-2)
IPA/NaOH Sample (1-3)
IPA/NaOH Field Blank
10.2	89.8
24.3	75.7
18.9 81.1
20.9 79.1
Spiked 0.5 M Phosphate Reagent (2 hr run)
0.5 M P04 Control
P04 Sample (P-l)
P04 Sample (P-2)
P04 Field Blank
6.8 93.2
87.4 12.6
85.0 15.0
9.7 90.3
2-9

-------
presence of both 51Cr+s and 51Cr+3 in the filtrate demonstrated the ability of the
improved analytical technique (see Section 4.0, Figure 4-2)) to separate soluble 51Cr
species.
In summary, the IPA/NaOH proved to be the best collection media, with the
native and radioactively-labeled chromium results agreeing quite well. The precision of
the impinger train measurements was also good. However, the IPA did cause long-term
problems with the analytical column on the IC/PCR.
2.2.4 Conclusions
The goal of this preliminary testing was to develop a sampling and analytical
method for hexavalent chromium that could attain approximately 100% recovery for both
native and labeled chromium. The test results demonstrated that all the candidate
methods showed some conversion of the hexavalent chromium during sampling and
sample storage prior to analysis. Consequently, it was decided that the Method 5-type
sampling train and the "recirculating reagent" impinger train would undergo further
evaluation testing at Site 6. Although the recirculating reagent impinger train was not
used at Site 5 because PEI did not have the equipment, studies previously conducted by
Entropy indicated that the conversion of hexavalent chromium during sampling was
significantly reduced by continuously recirculating the impinger reagent to the inlet of
the sampling probe. The dilution train was eliminated from further evaluations because
of the cost, operating difficulties, potential for filter contamination, and the conversion of
25% of the native and labeled hexavalent chromium during sampling and sample storage.
The data collected in this field evaluation was not emissions data and was not
intended to support the OW regulations. The study was conducted to evaluate the
conversion of internal spiked standards of hexavalent chromium. Therefore, none of the
data should be used for any standard setting purposes.
The cause of the conversion of the hexavalent chromium during sample collection
and/or recovery could not be determined from the method evaluation test. Additional
2-10

-------
work was conducted on the hexavalent chromium method by Entropy under a contract to
EPA's Quality Assurance Division (QAD) in the Research Triangle Park, North
Carolina. QAD plans to publish a report on the hexavalent chromium method
development at the conclusion of their work on the method. Therefore, a description of
the additional method development work is not discussed in this report.
2-11

-------
3.0 PROCESS DESCRIPTION AND OPERATION
Site 5 was a typical small multiple hearth incinerator with emissions controlled by
a venturi scrubber/impinger tray scrubber. The site was selected based on its close
proximity to PEI Associates, Inc. (one of the testing contractors). Sampling was
conducted in the discharge stack of the venturi scrubber/impingement tray scrubber.
Since the primary purpose of the testing was to assess the recovery of native and labeled
hexavalent chromium from the sampling trains, no process samples were taken for
analyses.
The first series of test runs was conducted on June 9, 1989, and the second series
of test runs was conducted on August 3, 1989. The average incinerator operating
conditions for these two days are shown in Table 3-1. These values are typical of
operating conditions for a small multiple hearth incinerator.
TABLE 3-1. SITE 5 INCINERATOR OPERATING CONDITIONS
Date
(1989)
Sludge Feed Rate-
(tons/hr, wet basis)

Average Hearth Temperature (°F)

No. 1
No. 2
No. 3
No. 4
No. 5
6/9
1.5
1000
1350
1500
1600
700
8/3
1.4
790
1240
1100
1410
490
3-1

-------
4.0 TEST RESULTS
The detailed results of the hexavalent chromium methods development tests
performed at Site 5 on June 9 and August 3, 1989 are presented in this section. Since
sampling train flow rates and sample volumes were not used to assess conversion of
hexavalent chromium, they are presented in Section 5 with the detailed description of the
sampling procedures.
The primary objectives of the testing were to conduct preliminary evaluations of
several available hexavalent chromium sampling and analytical methodologies. The
conversion of hexavalent chromium to other valence states during sampling and sample
storage was of primary concern. Conversion was assessed by determining the recovery of
known spikes of native hexavalent chromium (Cr+6) and radioactively-labeled hexavalent
chromium (slCr+6).
In addition to the test results, variability and outliers in the data are discussed.
The relationship of the process parameters to the results are not discussed.
Results are presented in terms of percent recovery of the native and labeled
hexavalent chromium spikes. Flue gas emission results are presented as measured.
Supporting data for the results presented in this section are included in the appendices.
4.1 TESTING PROGRAM DESIGN
Three sampling techniques were evaluated: 1) a Method 5-type sampling system
that collects the sample in a heated sampling probe and on a heated filter, 2) a reagent
impinger sampling train that collects the sample in an alkaline impinger reagent, and 3)
a dilution train that dilutes the sample gas 15-fold with ambient air and collects the
sample on an unheated glass fiber filter. The Method 5-type samples and the impinger
4-1

-------
train samples were collected by PEI Associates, Inc. The dilution train samples were
collected by Southern Research Institute. The sampling and analytical methods are
described in detail in Sections 5.0 and 6.0, respectively.
4.2 TEST PROGRAM RESULTS
The samples collected during the two methods development tests were analyzed
by two laboratories. EPA has not released the hexavalent chromium data from the TAI
analysis. This section presents only the detailed method evaluation results for the
hexavalent chromium samples analyzed by Entropy Environmentalists Inc..
4.2.1 Flue Gas Conditions
The flue gas conditions measured on the two test days are presented in Table 4-1.
These results are typical for a small multiple hearth incinerator operation. An emission
gas sample was collected in a bag on August 2 and analyzed for carbon monoxide (CO)
by a nondispersive infrared analyzer. The recorded value was 430 ppm CO on a dry
basis.
TABLE 4-1. SUMMARY OF FLUE GAS CONDITIONS FOR SITE 5
Volumetric Flow Rate" Moisture Flue gas Composition (%)
Date


content
temperature


(1989)
acfm
dscfm
(%)
(°F)
o2
C02
6/9
3224
2800
5.5
103
12
6
8/3
3691
3116
6.9
105
16
5
* - acfm (actual cubic feet per min), dscfm (dry standard cubic feet per min)
4-2

-------
4.2.2 Dilution Train Sampling Approach
Five dilution train sample runs were conducted during the first day of testing.
The sample run times (see Table 4-2) were 15, 30, 60, 120, and 60 minutes. The
recoveries seen for both the native and labeled hexavalent chromium spikes were similar
with averages of 76.6% and 77.4%, respectively. Thus, approximately 25% of the native
hexavalent chromium and labeled hexavalent chromium were converted to trivalent
chromium. The recoveries were fairly consistent from run-to-run and did not appear to
be related to the length of the sample run.
However, the dilution train is expensive to purchase, difficult to operate including
the move from sampling point to point, and requires an extremely large filter for sample
collection. This large filter greatly increases the background contamination levels of
total chromium and requires a large volume of extraction reagent.
The dilution train was eliminated from further evaluations because of the cost,
operating difficulties, potential for filter contamination, and the conversion of 25% of the
native and labeled hexavalent chromium during sampling and sample storage.
4.2.3 Method 5-Tvpe Sampling Approach
The Method 5-type sampling runs were conducted on both June 9 and August 3,
1989. Portions of the Method 5-type samples collected on June 9 and August 3 were
analyzed by TAI under contract to EPA's EMSL in Cincinnati, Ohio. The data from
these trains have not been released for publication and are not reported.
On June 3, one Method 5-type quad-train run was conducted for 60 minutes and
the samples analyzed by Entropy. The filters used in all four trains were spiked with
native and radioactively-labeled chromium prior to testing. Four control samples were
included in the test program. The control samples were digested after the test as
follows: 2-with water, 1-with 0.2 M phosphate buffer, and 1-with 0.1 N NaOH. As shown
in Table 4-3, the best recovery of 83% of the native chromium was obtained using the
0.1 N NaOH. Each sample filter was then digested with 250 mL of 0.1 N NaOH.
4-3

-------
TABLE 4-2. HEXAVALENT CHROMIUM RECOVERY USING DILUTION TRAIN
FROM JUNE 9, 1989
Sample
Sample Time % Native nCr+< Recovered
ID
(min)
from Filter
Native Hexavalent Chromium Analytical Data
F-l
15
73.7
F-2
30
67.0
F-3
60 (interrupted)
64.5
F-4
120
99.2
F-5
60
76.8
Average

76.6

Sample
Sample Time
% nCr+< Recovered
ID
(min)
from Filter
Radioactively-Labeled Chromium Analytical Data
F-l
15
67.1
F-2
30
76.5
F-3
60 (interrupted)
75.3
F-4
120
87.3
F-5
60
80.8
Average

77.4
4-4

-------
TABLE 4-3. HEXAVALENT CHROMIUM RECOVERY USING
METHOD 5-TYPE TRAIN FROM JUNE 9, 1989
Sample
Extraction Sample Time
% native Cr+6 Recovered
ID
Reagent (min)
from Filter
Native Hexavalent Chromium Analytical Data
Control 6
Water
55.1
Control 7
Phosphate
ND(0)
Control 8
Water
79.3
Control 9
NaOH
83.0
F-A
NaOH 60
19.1
F-B
60
34.3
F-C
60
38.5
F-D
60
25.9
Average

29.5
Sample
Sample Time
% 51Cr+< Recovered
ID
(min)
from Filter
Radioactively-Labeled Chromium Analytical Data
F-A
60
43.0
F-B
60
57.8
F-C
60
63.2
F-D
60
55.0
Average

54.8
4-5

-------
As shown in Table 4-3, the recoveries of the native and labeled hexavalent chromium
spikes averaged 29.5% and 54.8%, respectively. This demonstrates that approximately
70% of the native hexavalent chromium and 45% of the labeled hexavalent chromium
were converted to trivalent chromium during sampling and/or sample recovery.
•
Entropy encountered similar conversion results for hexavalent chromium in
previous laboratory research using a Method 5-type train to sample emissions containing
organics, acid gases, and sulfur dioxide. Because there were problems in recovering the
native and labeled chromium from the control samples from the first Method 5-type
sample run (June 9), the second set of Method 5-type test runs (August 3) included
posttest as well as pretest spiking of the filters. In addition, the analytical procedures for
separating and measuring labeled hexavalent chromium were improved between the first
and second test dates as described in Section 4.2.4.
On August 3, two quad-train runs were conducted, each over a 2-hr sampling
period. The results for the Method 5-type filter samples are presented in Table 4-4. For
the first quad-train run, four unspiked filters were used to sample the flue gas for 2 hr to
obtain a representative particulate loading. Two of the exposed filters were then
extracted and analyzed for Cr+6 yielding catches of 2.8 and 2.9 fig of hexavalent
chromium in the emissions. The other two filters plus two control filters were spiked
with a mixture of 5,Cr+s and native Cr+S, and extracted. 'Spike recoveries of the native
Cr+6 were 87.0% and 82.4% for the control filters; based on ion chromatographic
separation of the hexavalent and trivalent chromium about 98% of this recovered 5lCr
was in the hexavalent state. For the flue-gas exposed, posttest spiked filters, the
recoveries were 66.1% and 67.3% for the native Cr+S with 89.9% to 91% of the
recovered nCr in the hexavalent state. The expected amount of native Cr (13.7 fig) used
to calculate the spike recoveries was the sum of the native chromium spike (10.8 fig) and
the native chromium in the emissions (2.87 fig) based on the values measured for the
unspiked exposed filters.
The results for the filters spiked with native Cr+6 and "Cr*6 prior to sampling are
also presented in Table 4-4. The recoveries for" three pretest spiked filter samples, based
on an expected value of 13.7 fig, ranged from 47.0% to 48.3%. These recoveries
4-6

-------
TABLE 4.4. RECOVERY OF HEXAVALENT CHROMIUM FOR METHOD 5-TYPE TRAIN
FOR AUGUST 3, 1989
Sainple Identity
Native Hexavalent Chroniun
Spiked Hexavalent "chromium (total counts)
Expected Found X of X of
(ug) (ug) Expected Control
Total Residue Soliile
IC Separation
Percent of Total
6,Cr'3 "cr"
b'Cr° b,Cr"
Post Test Spiked Quartz Method 5-Type Filters Exposed to 2 Hours of Flue Gas
Spiked Control
10.8
9.4
87.0
NA
867869
1619
866250
18261
847989
2.3
97.7
Spiked Control
10.8
8.9
82.4
NA
858487
3962
854525
8804
845721
1.5
98.5
Post Test Spike (B 2)
13.7
9.0
66.1
98.8
866047
17997
848050
81345
766705
11.5
88.5
Post Test Spike (B-4)
13.7
9.2
67.3
100.5
872866
21841
851025
56920
794105
9.0
91.0
Exposed Filter (B-1)
0
2.9
NA
NA
394
9
385
NA
NA
NA
NA
Exposed Filter (B-3)
0
2.8
NA
NA
292
12
280
NA
NA
NA
NA
Pretest Spiked Quartz Method 5-Type Filters Exposed to 2 Hours of Flue Gas
Spiked Filt.
(1-1)
13.7
6.4
47.0
70.3
458210
113495
344715
63106
281609
38.5
61.5
Spiked Filt.
(1-2)
13.7
6.4
47.0
70.2
441934
126759
315175
38554
276621
37.4
62.6
Spiked Filt.
(I -A)
13.7
6.6
48.3
72.1
444449
91019
353430
27489
325941
26.7
73.3
Spiked Filt.
FB
10.8
0.6
5.3
6.2
430161
176761
253400
118253
135147
68.6
31.4

-------
constituted 70% to 72% of the spiked control filter recoveries from the previous quad-
train run. Of the recovered SICr, 61.5% to 73.3% was found to be in the hexavalent state
using the IC/PCR separation.
An improved analytical technique for separation and measurement of hexavalent
chromium including radioactively-labeled species was developed between the first and
second test days (see Section 6.1.3 for details). This technique required that the sample
discharge from the IC/PCR which had been separated by the analytical column be
collected at 30 sec intervals. The radioactively-labeled chromium content of each aliquot
was then measured by scintillation counting. Figure 4-1 contains two radio-
chromatograms of the radioactive content versus the aliquot time in min. Trivalent
chromium is separated by the IC column in the range of 2 to 4 min. The hexavalent
chromium is separated by the IC column in the 5 to 6.5 min range. The radio-
chromatograms show that the soluble radioactively-labeled chromium extracted from
both the pretest spiked and posttest spiked filters was principally in the hexavalent state
87% and 92%, respectively.
Overall, the Method 5-type filter experiments indicated that conversion during
sampling and/or matrix effects on recovery by extraction do occur, and can be measured,
semi-quantitatively, with slCr+s and/or native Cr+S spikes.
4.2.4 Impinger Train Approach
Entropy had determined in previous studies that the recirculating reagent
impinger train, which continuously recirculates the impinger reagents to the inlet of the
sampling probe, reduces the conversion of hexavalent chromium especially in the
sampling probe. Since, PEI Associates, Inc. did not have a "recirculating reagent"
impinger train, the preliminary evaluation testing was conducted using the only impinger
portion of the recirculating reagent train.
On June 9, two quad-train runs were conducted over a 60-min sampling period.
The impinger reagent used was isopropyl alcohol and NaOH (IPA/NaOH). As shown in
Table 4-5, the recovery of the native and labeled hexavalent chromium was 65.4% and
4-8

-------
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
o Spiked Filter 1 x Spiked Filter 4 a Spiked Filter 2 ^ Spiked Filter Field Blank

260

240

220

200

180
*oT
¦o
160
c

a
Ui
140
o

-C
b.
120
in
100
c

3
o
60
(J


60

40

20

0

-20
—i	1	i	1			1	1	:	1	1	1	1	¦		
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
Control Filter 1
Control Filter 2 ° Post Spiked Filter 2
a Post Spiked Filter 4
Figure 4-1. Radiochromatograms for the pretest and posttest spiked Method 5-type
samples.
4-9

-------
TABLE 4-5. HEXAVALENT CHROMIUM RECOVERY USING IMPINGER TRAIN
FROM THE FIRST DEVELOPMENT TEST
Sample
Sample Time
% nCr+< Recovered
ID
(min)
from Solution
Native Hexavalent Chromium Analytical Data
Control sample
NA
95.5
1-1 (A-l)
60
87.0
1-2 (A-2)
60
59.0
1-3 (B-l)
60
24.5
1-4 (B-2)
60
73.7
1-5 (C-l)
60
69.2
1-6 (C-2)
60
74.4
1-7 (D-l)
60
74.0
1-8 (D-2)*
60
21.1
Average

65.4

Sample
Sample Time
% 5ICr+s Recovered
ID
(min)
from Solution
Radioactively-Labeled Chromium Analytical Data
1-1 (A-l)
60
99.8
1-2 (A-2)
60
99.8
1-3 (B-l)
60
99.9
1-4 (B-2)
60
99.8
1-5 (C-l)
60
99.6
1-6 (C-2)
60
99.6
1-7 (D-l)
60
99.9
1-8 (D-2)*
60
(13.7)
Average

99.7
* The sample from Run 1-8 had silica gel collected in the impinger solution;
results not included in average.
4-10

-------
99.7%, respectively. These recovery averages imply that approximately 35% of the native
hexavalent chromium and none of the labeled hexavalent chromium were converted to
trivalent chromium. The radioactively-labeled chromium was measured using a
scintillation counter assuming that all labeled chromium in the sample filtrate must be in
the hexavalent state. The poor agreement between the recoveries for native and labeled
chromium clearly demonstrated that this assumption was wrong and additional laboratory
studies were conducted. These studies revealed that there was soluble labeled trivalent
chromium in the filtrate. It was determined that improved separation of the labeled
hexavalent and trivalent chromium could be accomplished through the use of the
analytical column on the IC/PCR. However, by the time the modified procedure was
developed, the radioactivity of the labeled chromium in the samples collected on June 9
was too low to conduct an additional analysis. In addition, the organic content of the
IPA/NaOH reagent was creating problems with the analytical column.
The results for the spiked impinger sampling conducted on August 3, 1989 are
summarized in Table 4-6. The improved analytical technique for separation of the
hexavalent and trivalent chromium prior to scintillation counting was used by Entropy in
the analysis of these samples. Recoveries of a native chromium spike (10.8 fig) from two
IPA/NaOH impinger reagent samples exposed to the flue gas for 2 hr were 81.7% and
74.8%. Relative to a spiked IPA/NaOH control sample where 8.9 fig was recovered,
the recoveries were 99.0% and 90.6% of the control value. The recoveries of spiked
51Cr+s in the two samples were in agreement with the native Cr recoveries, at 75.7% and
81.1%; the recovery of 51Cr+s in the control sample was 89.8%.
The analysis for 51Cr+s involves filtering the entire impinger sample followed by
IC separation of the resulting filtrate. In using this protocol, it is assumed that soluble
radioactive species in the filtrate that coelute with the native hexavalent chromium will
be hexavalent 5ICr. The radiochromatogram for the IPA/NaOH samples is shown at the
top of Figure 4-2. Note that the majority of the radioactivity eluted from the IC between
4.5 and 6 min, which is the same time period in which the native hexavalent chromium
4-11

-------
TABLE 4.6. RECOVERY OF HEXAVALENT CHROMIUM FOR IMPINGER TRAIN
FOR AUGUST 3, 1989


Native Hexavalent Chromiin

Spiked Hexavalent "
Chromim
(total counts)



Expected
(ug)
Found
(ug)
X of
Expected
X of
Control



IC Separation
Percent
of Total
Sample Identity

Total
Residue
Soluble
"Cr*1
"Cr"
"Cr"
"Cr"

Spiked BOX IPA/20X 1.0 N NaOH
Impinger Reagent Exposed to 2 Hours of Flue Gas



IPA/NaOH Control

10.B
B.9
82.5
NA
536845
450
536395
54571
481824
10.2
89.8
IPA/MaOH Sample (1
-2)
10.8
B.B
81.7
99.0
414468
26458
388010
74057
313953
24.3
75.7
IPA/MaOH Sample (1
-3)
10.8
8.1
74. B
90.6
398358
15750
382608
59632
322976
18.9
81.1
IPA/NbOH Field Blank
10.8
4.9
45.4
55.0
230848
284
230564
47979
182585
20.9
79.1
Spiked 0.5 M Phosphate Buffer Inpinger Reagent Exposed to 2 Hours of Flue Gas
0.5 M P04 Control

10.8
11.2
103.5
NA
NC
NC
NC
11100
151300
6.8
93.2
P04 Sample CP-1J

10.8
2.B
26.2
25.3
563977
5579
558398
487509
70889
87.4
12.6
P04 Sample (P-2)

10.8
3.1
28.5
27.5
541406
6654
534 752
453351
81401
85.0
15.0
P04 Field Blank

10.8
9.2
84.7
81.8
435957
908
435049
41561
393488
9.7
90.3

-------
in
TJ
c
as
in
~
o
100
90
80
70
60
50
40
30
20
10
0
-10
IPA control
IPA Sample (I-3)
a IPA Sample (I-2)
"O
c
3)
in
~
o
-C
t.
100
90
80
70
60
50
40
30
20
10
0
-10

—		1	1	1			1	1	1	1	1	1	:—
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
Tlma
8.0
¦ P04 Control	• P04 Sample (P-1)	a P04 Sample (P-2)
Figure 4-2. Radiochromatograms for the impinger train samples.
4-13

-------
elutes. Of the soluble radioactive chromium, 83% coeluted from the IC/PCR with
native hexavalent chromium.
A field blank sample was also collected with the IPA/NaOH set. Inexplicably, the
recovery of the spike native chromium was only about 50%, and the recovery of the
radioactively-labeled chromium was low also.
Use of a 0.5 M phosphate buffer in the impinger reagent resulted in very low
recoveries, about 25% for the native Cr+6 and about 15% for the 5ICr+6. Both a control
sample and field blank sample showed good recoveries for the native Cr+6 and 5ICr+6.
While the phosphate buffer did not prevent conversion during sampling, the
radiochromatogram (see the bottom of Figure 4-2) for the control sample and the field
samples (P-l and P-2) graphically demonstrates the ability of the improved IC technique
to separate the soluble slCr+6 and 51Cr+3. Of the soluble 51Cr, only 14% coeluted from
the IC/PCR with the native hexavalent chromium.
In summary, the IPA/NaOH proved to be the best collection media, with the
native and radioactively labeled chromium results agreeing quite well. The precision of
the impinger train measurements was also good. However, the IPA did cause long-term
problems with the analytical columns on the IC/PCR.
4.3 CONCLUSIONS
This preliminary method evaluation testing demonstrated that all three candidate
methods showed some conversion of hexavalent chromium during sampling and/or
sample storage prior to analysis. Consequently, it was decided that the Method 5-type
sampling train and the "recirculating reagent" impinger train would undergo further
evaluation testing at Site 6. Although the recirculating reagent impinger train was not
used at Site 5 because PEI did not have the equipment, studies previously conducted by
Entropy indicated that the conversion of hexavalent chromium during sampling was
significantly reduced by continuously recirculating the impinger reagent to the inlet of
the sampling probe. The dilution train was eliminated from further evaluations because
of the cost, operating difficulties, potential for filter contamination, and the conversion of
4-14

-------
25% of the native and labeled hexavalent chromium during sampling and sample storage.
The data collected in this field evaluation was not emissions data and was not
intended to support the OW regulations. The study was conducted to evaluate the
conversion of internal spiked standards of hexavalent chromium. Therefore, none of the
data should be used for any standard setting purposes.
The cause of the conversion of the hexavalent chromium during sample collection
and/or recovery could not be determined from the method evaluation test. Additional
work was conducted on the hexavalent chromium method by Entropy under a contract to
EPA's Quality Assurance Division (QAD) in the Research Triangle Park, North
Carolina. QAD plans to publish a report on the hexavalent chromium method
development at the conclusion of their work on the method. Therefore, a description of
the additional method development work is not discussed in this report.
4-15

-------
5.0 SAMPLING PROCEDURES
Sampling procedures used during the Site 5 program are briefly described in this
section. Standard EPA methods or draft EPA procedures were used when available.
5.1 DILUTION TRAIN
The stack gas dilution sampling system (dilution train) developed by SRI was
evaluated only on the first test day (June 9, 1989). It was thought that the 15-fold
dilution of the emissions using ambient air provided by this train might reduce the effects
of the emissions matrix species in converting the hexavalent chromium to other valence
states. As previously mentioned, prior to this method evaluation testing, no sampling
system yet evaluated had demonstrated the ability to prevent the conversion of
hexavalent chromium during the sampling. The gases contributing to the conversion
were thought to include organics, acid gases, and sulfur dioxide.
The dilution train system used (see Figure 5-1) withdraws the emissions from the
stack through a sampling nozzle and probe. The emission gases then enter a dilution
chamber where they are diluted about 15-fold with ambient air. This approximates
collecting the emissions as they would exist in ambient air.
The dilution train sampling rate from the stack is approximately 17 L/min. The
total diluted flow rate of 425 L/min is monitored by using an orifice in the exhaust line.
Gas flow rates are adjusted using the two blowers shown in Figure 5-1. These blowers
are controlled by variable transformers.
The dilution air is forced through an ice bath condenser, reheated to the desired
temperature, filtered, and introduced tangentially in the inlet assembly at the bottom of
the dilution chamber. Dilution air flow is directed upward in the annulus border on the
5-1

-------
L/l
i
K)
EXHAUST BLOWER
HI-VOL IMPACTOR
AND/OR FILTER
PROCESS STREAM
DILUTION
CHAMBER
SAMPLING
CYCLONE
TO ULTRAFINE
PARTICLE SIZING
SYSTEM (OPTIONAL!
,DILUTION AIR
HEATER
FILTER
PROBE
CONDENSER
lilMiiriin g'li'ii
ICE BATH
GAS FLOW
TO HEATERS, BLOWERS
TEMPERATURE SENSORS
TO ORIFICE
PRESSURE TAPS
FLOW, PRESSURE
MONITORS
MAIN CONTROL
© © ©
© ©
¦ DILUTION AIR
BLOWER
Figure 5-1. Dilution
train sampling system.

-------
outside of the dilution chamber and collected on a glass fiber filter.
The dilution train uses 8.5 in by 11 in glass fiber filters for sample collection.
Prior to sampling, these filters were spiked by Entropy with 10.8 ng of native hexavalent
chromium (Cr+6) and less than 0.1 millicurie of radioactively-labeled hexavalent
chromium (nCr+s). The sampling train conditions including flow rates for the five test
runs conducted by SRI using the dilution train are shown in Table 5-1.
5.2. METHOD 5-TYPE TRAIN
The Method 5-type train was evaluated on both test days. Sampling followed the
procedures of the draft EPA method, "Determination of Hexavalent Chromium from
Stationary Sources," dated December 13, 1984. A diagram of the sampling train is
shown in Figure 5-2; the draft method is not reproduced in this report since the DCr+s
data were not released by EPA Briefly, the procedure involves the use of the EPA
Method 5 sampling train with the following modifications:
•	A glass nozzle and probe liner are used, and
•	the glassware is cleaned according to the procedure in Table 5-2.
The sampling train nozzle, probe liner, and filter holder were made of
borosilicate glass. Both the probe and filter holder were heated to 248°F _+ 25°F to
prevent moisture condensation. High purity quartz fiber filters without organic binder
and with a 99.95% collection efficiency for 0.3 fim dioctyl phthalate (DOP) smoke
particles were used.
The filters spiked with a stable hexavalent chromium isotope ("Cr*"6) were
analyzed by TAI under contract to EPA's EMSL in Cincinnati, Ohio. The results of the
Method 5-type train testing have not been released by EPA and are not presented in this
report.
On June 9, 1989, a single quad-train run was conducted yielding four filters.
These filters were spiked by Entropy prior to testing with 13.18 fig of native chromium
5-3

-------
TABLE 5-1. DILUTION TRAIN SAMPLING CONDITIONS
Sample
ID
Start
Time
24 hr
Run
Time
min.
Sample
Flow Rate
(scfm)
Dilution
Flow Rate
(scfm)
Total
Sample Vol.
(dscf)
Total
Dil. Vol.
(dscf)
Stack
Temp.
(°F)
Filter
Temp.
(°F)
F1
1050
15
0.89
13 .8
12 . 6
207
91
82
F2
1118
30
0.90
13 . 8
25.5
414
93
82
F3 *
1156
60|
0. 90
13.8
51
1450
95
88
F4
1557
120
0.90
14 . 0
102
1680
95
86
F5
1813
60
0.90
14 . 0
51
840
93
84
* Run interrupted by plant operation problems. Sample flow was discontinued and
dilution air flow was reduced during upset period,
t Does not include upset period.

-------
i
L/1
j/
Thermo oouple
Glaii NolzIb j^T"V
Glass Probe U ^^3
Reverse-Type]
Pilot Tube
Glass
Rtler Holder
Thermocouple Check
Valve
Thermometer
Quartz
Rlter
Heated Area
Pilot
Manometer
Dl Water-
ImpJngers
c« Bath
Silica Gel
Empty
Bypass	Vacuum
valve
vacuum
Gauge
Thermocouples
Line
Orlllce
Main
Valve
Alr-Tlght
Pump
Dry Gas
L Meter i
Figure 5-2. Method 5-type sampling train.

-------
TABLE 5-2. Cr/Cr+S GLASSWARE CLEANING PROCEDURES
NOTE: Use disposable gloves and adequate ventilation.
1.	Soak all glassware in hot, soapy water (Alconox).
2.	Rinse with tap water, three times.
3.	Rinse with deionized water, three times.
4.	Soak in 10% HN03 for 10 hours.
5.	Rinse with deionized water, three times.
6.	Cap glassware with Teflon tape.
7.	Leave cleaned glassware remained capped until field assembly.
Cr+6 and less than 0.1 millicurie of radioactive labeled hexavalent chromium (51Cr+s).
Four control filter samples were set aside for digestion with water, 0.2 M phosphate
buffer, and 0.1 N NaOH.
On August 3, 1989, two quad-train runs were conducted yielding eight filters for
Entropy analysis. One of the quad-train runs utilized four filters spiked prior to
sampling (referred to as a pretest spike); and the other run utilized four unspiked filters.
One of the spiked filters was lost during testing. Two of the four unspiked filters used
for the second quad-train run were spiked following sampling just prior to digestion
(posttest spike).
The spiking was done by Entropy using 10.8 /xg of native chromium and less than
0.1 millicurie of radioactive labeled hexavalent chromium (51Cr). Two posttest spiked
control filter samples and one pretest spiked control filter sample were set aside for
analysis with the field samples.
The sampling train conditions for both test days are presented in Table 5-3.
5-6

-------
TABLE 5-3. METHOD 5-TYPE TRAIN SAMPLE CONDITIONS
Sample
Start Time Run Time
Sample Volume
ID.
(24 hr)
(min)
(dscf)


First Test Day
- June 9, 1989
F-A

60

F-B

60

F-C

60

F-D

60



Second Test Day
- August 3, 1989
7-A
0828
120
85.5
7-B
0828
120
84.7
7-C
0828
120
92.1
7-D
0828
120
89.1
8-A
1050
120
86.8
8-B
1050
120
85.5
8-C
1050
120
90.6
8-D
1050
120
89.3
5.3 I MP INGE R TRAIN
The irapinger train was evaluated on both test days. It was originally anticipated
" that the method evaluation would be conducted using the recirculating reagent impinger
developed for hexavalent chromium testing. However, since PEI Associates, Inc. did not
own a recirculating impinger reagent train, an impinger train configured by removing the
front filter from a Method 5 train was used (see Figure 5-3). Because the analysis only
examined the recovery of native and labeled hexavalent chromium spiked into the
impinger reagents, the material collected in the nozzle and probe was not recovered with
the sample. Previous research had shown that glass components in the sampling train
could contribute to the background levels of trivalent chromium in the acid rinse, but,
5-7

-------

Thermocouple
Glass Nozzle /—j~\
Glass Probe	L
Reverse-Type
Pltot Tube
Pilot
Manometer
Orifice
Thermocouple!
Thermocouple
Check
Valve
80V. IPA / 20% 2N NaOH
or
0.5 M Phosphate Buffer
Bypass
Valve
Main
Valve
Implngers
Ice Bath
Sllca
Vacuum
Une
Vacuum
Game
Figure 5-3. Impinger sampling train

-------
since the evaluation did not include a nitric rinse of the sampling train, the use of glass
components was not considered a problem.
Sampling and analysis followed the procedures in the draft EPA method,
"Determination of Hexavalent Chromium from Stationary Sources," except as noted
above. A diagram of the full recirculating reagent sampling train is shown in Figure 5-4;
the draft method is reproduced in Appendix B. This procedure is based on EPA
Method 5 with the following modifications:
•	80% IP A/20% 2 N NaOH replaces water in the impingers (0.5 M
phosphate buffer was used in the impingers for one quad-train run on the
second test day);
•	the entire surface exposed to sample is constructed of Teflon and/or glass;
•	the Teflon and/or glass components are cleaned according to the
procedure in Table 5-4;
•	the train does not have a filter section; and
•	the reagents are continuously recirculated from the first impinger back to
the nozzle to provide a flow of reagents through the probe, and thus
preventing hexavalent chromium in the probe drying out and possibly
converting to another valence state (not done for this testing).
TABLE 5-4. Cr+VCr TEFLON/GLASS COMPONENTS CLEANING PROCEDURES
1.	Soak all components in hot, soapy water (Alconox).
2.	Rinse with tap water, three times.
3.	Rinse with deionized water, three times.
4.	Soak in 10% HNOa for 10 hours.
5.	Rinse with deionized water, three times.
6.	Cap Teflon/glassware with Teflon tape.
7.	Leave cleaned Teflon/glassware capped until field assembly.
5-9

-------
GLASS
IMPINGERS-
Lft

TEFLON
IMPINGERS
\
TEFLON
LINES
ASPIRATOR
150 ml
0.1N NaOH
100 ml
0.1N NaOH
EMPTY
100 ml
0.1N HNOg
SILICA
GEL
RECIRCULATING
LIQUID
ICE BATH
TO
METHOD 5-TYPE
METERBOX
NOZZLE
Figure 5-4.
Recirculating reagent sampling train.

-------
On June 9, two quad-train runs were conducted yielding eight impinger samples.
A 2-L volume of the impinger reagent (80% IPA/20% 2 N NaOH) was spiked by
Entropy prior to testing with 18.83 fig per 200 mL of native hexavalent chromium and
less than 1 millicurie per 200 mL of radioactively-labeled hexavalent chromium (s,Cr+6).
Each sampling train was charged with a 200 mL of the spiked impinger reagent. To
verify the spike concentration and as a check for contamination, control samples of the
spiked impinger reagent were set aside and analyzed with the field samples.
On August 3, two quad-train runs were conducted yielding eight samples. For the
first run, a 2-L volume of impinger reagent (80% IPA/20% 2 N NaOH) was spiked by
Entropy prior to testing with 10.8 fig per 200 mL of native hexavalent chromium and less
than 1 millicurie per 200 mL of radioactively-labeled hexavalent chromium (s,Cr+6).
Each train was charged with a 200 mL spiked impinger reagent. For the second run, a
2-L volume of a phosphate buffer impinger reagent (0.5 M) was spiked by Entropy prior
to testing with 10.8 fig per 200 mL of native hexavalent chromium and less than 1
millicurie per 200 mL of 51Cr+6. Each train was charged with a 200 mL spiked impinger
reagent. One control sample and the concentrated spiking solution were analyzed for
each impinger reagent. The impinger train run conditions for both quad-train runs are
shown in Table 5-5.
5.4 EPA METHODS 1, 2, 3, AND 4
EPA Methods 1, 2, 3, and 4 were used to determine the volumetric flow rate, the
molecular weight, and the flue gas moisture content.
The volumetric gas flow rate was determined during this program using
procedures described in EPA Methods 1 and 2. Based on these methods, the volumetric
gas flow rate is determined by measuring the cross-sectional area of the stack and the
average velocity of the flue gas. The average flue gas velocity is calculated from the
average pitot tube pressure (AP), the average flue gas temperature, the wet molecular
weight, and the absolute static pressure measured at a single point in the stack.
5-11

-------
TABLE 5-5. IMPINGER TRAIN SAMPLING CONDITIONS
Sample Start Time Run Time Sample Volume
ID.	(24 hr) (min)	(dscf)
First Test Day - June 9, 1989
1-1

60

1-2

60

1-3

60

1-4

60

1-1

60

1-2

60

1-3

60

1-4

60



Second Test Day
- August 3, 1989
9-A
1318
120
88.6
9-B
1318
120
88.1
9-C
1318
120
92.8
9-D
1318
120
90.6
The integrated sampling technique described in EPA Method 3 was used to
obtain composite flue gas samples for fixed gas (02, C02) analysis. A small diaphragm
pump and a stainless steel probe were used to extract a single-point flue gas sample
which was collected in a Tedlar bag. Moisture was removed from the gas sample by a
water-cooled condenser so that the fixed gas analysis is on a dry basis. The composition
of the gas sample was determined using an Orsat analyzer.
The moisture content of the flue gas was determined using the methodology
described in EPA Method 4. Based on this method, a known volume of particulate-free
gas was pulled through a chilled impinger train. The quantity of condensed water was
determined gravimetrically and then related to the volume of gas sampled to determine
the moisture content.
5-12

-------
6.0 ANALYTICAL PROCEDURES
Three analytical procedures were employed to speciate chromium compounds in
the samples to determine the recovery of native and labeled hexavalent chromium:
IC/PCR, ICP/MS, and gamma or scintillation counting. The dilution train, Method 5-
type train, and impinger samples collected for Entropy were analyzed using ion
chromatography with a Cr+<-specific post column reaction (IC/PCR). Entropy also
performed gamma counting measurements of labeled hexavalent chromium (5lCr+6)
spiked into samples to determine the recovery and, consequently, conversion of native
and labeled hexavalent chromium that may occur during sampling, sample handling, and
sample preparation. Since the results for the hexavalent chromium filter train analyses
conducted by TAI using IC/PCR and ICP/MS were not released, their analytical
techniques will not be discussed.
The IC/PCR system was calibrated daily with a series of three freshly prepared
Cr+< standard solutions at concentrations ranging from 23-to-108 ng/ml.
6.1 FIRST TEST DAY - JUNE 9, 1989
6.1.1 Dilution Train
The dilution train uses an 8.5 in. by 11 in. glass fiber filter to collect the sample.
Prior to testing, filters were spiked by Entropy with 10.8 ng of native hexavalent
chromium and less than 0.1 millicurie of radioactively-labeled hexavalent chromium
(slCr+6). The mass of radioactively-labeled hexavalent chromium was well below the
detection limit of the IC/PCR, and therefore, there was no interference of the labeled
chromium in the native chromium analysis. If the mass of the radioactively-labeled
6-1

-------
hexavalent chromium was above the detection limit of the IC/PCR, it would measured
as native hexavalent chromium. The radioactive chromium used (51Cr+s) has a half life
of 28 days. Native chromium does not contain any radioactive hexavalent chromium and
would therefore not interfere with the gamma counting analysis.
Each of the five dilution train filter samples was extracted using 500 mL of 0.1 N
NaOH. The extraction included sonication for 50 min followed by vacuum filtering
through a 0.45 fim Teflon filter. The filtrate was then analyzed for native hexavalent
chromium by IC/PCR and radioactive content with a scintillation (gamma) counter. The
glass fiber filter and 0.45 ^m Teflon filter were also analyzed for radioactivity with the
scintillation counter. It was assumed that all the radioactive chromium in the filtrate was
in the hexavalent state.
The recovery of native hexavalent chromium was calculated by comparing the
amount of hexavalent chromium spiked onto the glass fiber filter with the amount of
hexavalent chromium recovered in the filtrate.
At the end of the first test day, there were no spiked filters remaining to serve as
a control sample and, therefore, a control sample could not be analyzed.
6.1.2 Method 5-Type Train
A single Method 5-type quad-train run on the first test day was conducted and
four 82 mm glass fiber filter samples were collected. The filters were spiked by Entropy
prior to sampling with 13.18 fig of native Cr+S and less than 0.1 millicurie of ("Cr*6).
Four control filter samples were spiked and extracted as described for the dilution train
filters. Two with water, 1-with 0.2 M phosphate buffer, and 1-with 0.1 N NaOH. The
best recovery of 83% was obtained using the 0.1 N NaOH to extract the filter. Each
field sample filter was then digested with 250 mL of 0.1 N NaOH using the same
procedure. The filtrate was analyzed for native hexavalent chromium by IC/PCR and
radioactivity with a scintillation counter. The glass fiber filter and 0.45 fim Teflon filter
were analyzed for radioactivity with a scintillation counter. It was assumed that all the
radioactive chromium in the filtrate was in the hexavalent state.

-------
6.1.3 Impinger Train
Two quad-train runs yielded eight impinger samples for analysis. Each impinger
train was charged with 200 mL of spiked impinger reagent (80% IPA/20% 2 N NaOH).
Two liters of impinger reagent were spiked by Entropy prior to testing with 18.83 ng per
200 mL of native Cr+6 and less than 1 millicurie per 200 mL of slCr+6. One control
reagent sample and three dilutions of the concentrated spiking solution were also
analyzed. Each impinger sample was vacuum filtered through a 0.45 fim Teflon filter.
The filtrate was boiled to reduce the concentration of IP A, and then analyzed for native
hexavalent chromium by IC/PCR and radioactivity with a scintillation counter. The
Teflon filter was also analyzed for radioactivity with the scintillation counter. It was
assumed that all the radioactive chromium in the filtrate was in the hexavalent state.
As mentioned in Section 4.0, the poor agreement between the recovery of native
chromium and labeled chromium indicated that all the soluble chromium in the filtrate
must not be in the hexavalent state. Laboratory experiments were then conducted under
a separate work assignment with EPA's Quality Assurance Division, Research Triangle
Park, NC. A new analytical procedure was developed to separate the hexavalent and
trivalent chromium prior to scintillation counting. In this procedure, the discharge from
the IC/PCR (sample separated by analytical column) is collected every 30 sec over the
typical analytical run time of 8 min. These individual aliquots are analyzed using the
scintillation counter. By the time the new analytical technique had been developed, the
radioactivity levels from the first test day samples were too low to detect.
6.2 SECOND TEST DAY - AUGUST 3, 1989
The new analytical technique developed following the first field test was used to
analyze the labeled hexavalent chromium in the samples from the second test. Because
of poor recoveries seen for the control samples for the first test, it was decided to use a
pretest spike and a posttest spike in the second test.
6-3

-------
6.2.1 Dilution Train
No dilution train samples were collected during the second test day.
6.2.2	Method 5-Type Train
Two quad-train runs conducted on August 3 resulted in eight glass fiber filters
samples. The filters for one of the quad-train runs were spiked prior to testing; the
filters for the other quad-train run were unspiked prior testing and two of these were
spiked prior to extraction. One of the spiked filters was lost during sampling. The
filters were spiked by Entropy prior to sampling (pretest spike) or just prior to extraction
(posttest spike) with 10.8 fig of native Cr+S and less than 0.1 millicurie of MCr+<. Two
posttest spiked control samples and one pretest spiked control sample were prepared and
analyzed. Each filter was extracted with 250 mL of 0.1 N NaOH including sonication
and vacuum filtering through a 0.45 fim Teflon filter. The filtrate was then analyzed for
native hexavalent chromium by IC/PCR. The IC/PCR sample discharge was collected at
30-sec intervals and these aliquots were analyzed for radioactivity with a scintillation
counter. The filters were analyzed for radioactivity by scintillation counting.
6.2.3	Impinger Train
Two quad-train runs generated eight impinger samples for analysis. Each
impinger train was charged with 200 mL of spiked impinger reagent. For the first quad
train run, 2 L of impinger reagent (80% IPA/20% 2 N NaOH) were spiked by Entropy
prior to testing with 10.8 ng per 200 mL of native Cr+S and less than 1 millicurie per 200
mL of 51Cr+6. For the second quad-train run, 2 L of impinger reagent (0.5 M phosphate
buffer) were spiked by Entropy in a similar manner. One control sample of spiked
impinger reagent and one sample of concentrated spiking solution were analyzed for
each impinger solution. Each impinger sample was vacuum filtered through a 0.45 /im
Teflon filter. The filtrate was boiled to reduce the concentration of IP A, and then
6-4

-------
analyzed for native hexavalent chromium by IC/PCR and radioactivity with a scintillation
counter. The Teflon filter used to remove the solids for each sample was analyzed for
radioactivity with a scintillation counter.
The resulting filtrates were analyzed by the new IC/PCR analytical technique. To
determine the ratio of the soluble slCr+3 and s,Cr+s species where there was a spike, 30-
sec aliquots was collected during the IC/PCR analysis, and the gamma counts measured
for each fraction.
6-5

-------
7.0 QUALITY ASSURANCE AND QUALITY CONTROL
This section discusses the quality assurance and quality control (QA/QC) program
and the QA/QC results for the sewage sludge incineration test program at Site 5. For
this site, the test program was designed for methods development and evaluation
purposes only and, therefore, no emissions data was collected. This eliminated the need
to perform isokinetic sampling, since primarily quadruplicate sampling trains were
employed and only the relative differences between the trains was of interest. The
objectives of and basic activities for the QA/QC program are briefly discussed in the
section below.
7.1	QA/QC PROGRAM OBJECTIVES
For any environmental measurement, a degree of uncertainty exists in the data
generated due inherent limitations of the measurement systems employed. To assess the
quality of the data and to establish limitations on the ultimate use of the data, a limited
QA/QC program was implemented for this test effort. The objective of the QA/QC
program was to produce complete, representative, and comparable data of known
quality. All elements of the QA/QC program were implemented during the sampling
and analytical phases of the sewage sludge incinerator test program for Site 5.
7.2	FLUE GAS SAMPLING AND ANALYSIS QC RESULTS
Quality control activities for flue gas sampling include: (1) equipment calibrations,
(2) glassware and equipment cleaning, (3) procedural checks during sampling and sample
recovery, (4) sample custody procedures, (5) procedural checks during sample analysis,
7-1

-------
and (6) the use of labeled surrogates, field blanks, laboratory blanks, QC check samples,
matrix spikes, and duplicate analyses. The QC results for these activities are discussed in
this section for each of the three sampling procedures evaluated at Site 5 and for the
analytical procedures employed.
7.2.1	Dilution Train Sampling
During sampling, the dilution rates were kept constant for each of the five test
runs. The dilution ratio of total sample volume to gas sample volume ranged from 15.3
to 15.6. The temperature of the filters where the diluted sample mass was collected
ranged from 82 to 88 °F during the five sample runs. The relatively consistent dilution
train sampling conditions established for the five runs allowed for meaningful
comparison of the main variable, the sample collection time.
Each of the dilution train filters were spiked with 10.8 /ig of native hexavalent
chromium and less than 0.1 millicurie of radioactively-labeled hexavalent chromium
(51Cr+s) to assess conversion and recovery of hexavalent chromium using the dilution
sampling train. Due to process problems, additional sampling runs were required, and
the filters sent to the field designated as controls had to be used. Therefore, no control
samples were available for analysis.
There was good agreement between the recovery of the spiked native hexavalent
chromium and the spiked 51Cr+s that averaged 76.6% and 77.4%, respectively.
Even though the dilution train sampling approach provided comparable recoveries
of hexavalent chromium as the impinger method, the dilution train was eliminated from
further consideration due to practical and technical considerations.
7.2.2	Method 5-Type Sampling
All sampling train glassware and Teflon components and sample containers were
precleaned initially with soap and water followed by a DI water rinse, an 0.1 N nitric
7-2

-------
acid rinse, and a final DI water rinse. During on-site testing, all sampling train glassware
was capped with Parafilm or Teflon tape prior to and immediately after each test run. A
clean, dust-free environment was maintained on-site for sampling train assembly and
sample recovery.
Method 5-type 82 mm quartz fiber filters were used to evaluate this sampling
procedure. The filters were spiked with 13.18 ng of native hexavalent chromium and less
than 0.1 millicurie of 51Cr+< to assess conversion and recovery of hexavalent chromium.
For the first test series, four filters were spiked and then extracted with different
reagents to determine which reagent gave the best recovery. It was determined that 0.1
N NaOH was superior to water and phosphate buffer. The remaining filters were
exposed to flue gas for 60 minutes at the nominal Method 5 sampling rate. The filters
were extracted with 0.1 N NaOH and the native and "Cr*6 recoveries determined. The
radiochromatography procedure (discussed in Section 7.2.4) was not employed on these
samples.
Of the extraction reagents evaluated on the control filters, 0.1 N NaOH gave the
highest recovery (83.0%) of native hexavalent chromium. There was not good agreement
between the recoveries of the spiked hexavalent chromium and the spiked 5,Cr+< that
averaged 29.5% and 54.8%, respectively. This difference may have been due to the
assumption that, after 0.45 /im filtration of the extraction solution, all of the soluble
radioactivity was 51Cr+< (see Section 7.2.4).
For the second test series, filters were spiked with 10.8 ng of native hexavalent
chromium and less than 0.1 millicurie of 51Cr+<. A set of filters were spiked prior to
exposure to flue gas (pretest spike) and a second set were spiked after flue gas exposure
but prior to extraction (posttest spike). The pretest spikes were used to assess native
and 51Cr+< recovery after exposure to flue gas. The posttest spikes were used to assess
matrix effects on the recovery of native and SICr+< from the filter and particulate matter
on the filter. Posttest spiked controls were prepared and analyzed to determine native
and 51Cr+6 from the filter matrix. Unspiked filters were exposed to flue gas to provide a
measure of native hexavalent chromium collected from the source during sampling to
7-3

-------
allow for correction of the spiked native hexavalent chromium recovery. A pretest
spiked filter not exposed to flue gas was used as a field blank.
All recoveries of 51Cr+6 for the second test series were determined by
radiochromatography (see Section 7.2.4).
Posttest spiked controls had 82.4 to 87.0% recovery of native hexavalent
chromium and 97.7 to 98.5% recovery of 51Cr+6. The posttest spikes of exposed filters
had 98.8 to 100.5% and 88.5 to 91.0% recovery of the native and 51Cr+<, respectively,
relative to the posttest spiked controls. The pretest spiked field blank had only 6.5%
and 31.4% recovery of the native and 51Cr+<, respectively, relative to the posttest spiked
controls. The field blank had lower recoveries than the recoveries determined for the
pretest spiked filters exposed to flue gas for 2 hr, and appears to be an outlier.
7.2.3 Impinger Train Sampling
All sampling train glassware and Teflon components and sample containers were
precleaned initially with soap and water followed by a DI water rinse, an 0.1 N nitric
acid rinse, and a final DI water rinse. During on-site testing, all sampling train glassware
was capped with Parafilm or Teflon tape prior to and immediately after each test run. A
clean, dust-free environment was maintained on-site for sampling train assembly and
sample recovery.
For the first test series, the impinger reagent (80% IPA/20% 2 N NaOH) was
spiked with native and 51Cr+6, respectively, at a concentration of 18.83 ng and less than 1
millicurie per 200 mL. Samples of the spiked reagent were collected as field blanks for
analysis with the field samples.
From the control sample, 95.5% and 99.9% of the native and SICr+6, respectively,
were recovered. The recoveries of spiked hexavalent chromium and 51Cr+< from the field
samples averaged 65.4% and 99.7%, respectively. This difference may have been due to
the assumption that, after 0.45 /im filtration of the extraction solution, all of the soluble
radioactivity was nCr+6 (see Section 7.2.4).
7-4

-------
For the second test series, the impinger reagents (80% IP A/20% 1.0 N NaOH
and 0.5 M phosphate buffer) were spiked with native and 51Cr+s, respectively, at a
concentration of 10.8 fig and less than 1 millicurie per 200 mL. Control samples and
field blanks of each reagent were collected and analyzed with the field samples.
Radiochromatography was employed to measure SICr+6 in the sample filtrate.
The IPA/NaOH control sample had recoveries of 82.5% and 89.8% of the native
and SICr+6 spikes. The IPA/NaOH field blank had recoveries of 45.4% and 79.1% of
the native and 5ICr+s spikes. The exposed samples had average recoveries of 78.3% and
78.4%, respectively, of the native and 51Cr+s spikes.
The phosphate buffer control sample had recoveries of 103.5% and 93.2% of the
native and 51Cr+6 spikes. The phosphate buffer field blank had recoveries of 84.7% and
90.3% of the native and 5ICr+6 spikes. The exposed samples had average recoveries of
27.4% and 13.8%, respectively, of the native and 5,Cr+s spikes.
7.2.4 Native Hexavalent Chromium and 51Cr*6 Analysis
The analysis of filter extracts and impinger sample for native hexavalent
chromium was performed by IC/PCR following the procedures in the draft EPA method,
"Determination of Hexavalent Chromium from Stationary Sources." A three-point
calibration curve from 23-to-108 ppb was established. The deviation of the standard
responses from the linear regression line ranged from -8.6% to 5.7%. The largest
deviation calculated, -8.6%, was for the 23 ppb standard where the predicted
concentration was 21 ppb, a difference of 2 ppb.
The discrepancy between the recoveries of the spiked hexavalent chromium and
SICr+s for the first test series led to the development of the radiochromatography
technique for analysis of the samples from the second test series. Upon analysis of the
samples from the second test series, the assumption that all soluble radioactivity was
present as 51Cr+s was found to be false. Of the soluble radioactivity in the NaOH
extracts of pretest spiked Method 5-type filters exposed to flue gas, an average of 87%
7-5

-------
was found to elute from the IC/PCR with native hexavalent chromium. For the
IPA/NaOH impinger samples, an average of 83% of soluble radioactivity was found to
elute from the IC/PCR with native hexavalent chromium. For the phosphate buffer
impinger samples, an average of 14% of soluble radioactivity was found to elute from the
IC/PCR with native hexavalent chromium. In each case the radiochromatography
determinations of slCr+s were in good agreement with the recoveries of native hexavalent
chromium spikes.
7-6

-------
REFERENCES
1.	Drees, L.M. Effect of Lime and Other Precipitants or Sludge Conditioners on
Conversion of Chromium to the Hexavalent State When Sludge is Incinerated. Final
Report. EPA Contract No. 68-03-3346, WA 05. 1988.
2.	Majiam, T.T. Kasakura, N. Naruse and M. Hiraoka. 1977. Studies of Pyrolvsis
Process of Sewage Sludge. Prog. Wat. Tech. Vol. 9, 381-396. Great Britain: Pergamon
Press.
3.	Umashima, T., M. Naruse and T. Nasakura. 1975. Behavior of Cr + 6 in Incinerator
Process of Sewage Sludge. Paper presented at the 12th Annual Meeting of the
Association of Japan Sewage Works.
8-1

-------
Appendix A
Analytical Data
A-l

-------
Chromiun Analytical Data Sheet
Job Nerne:Chrominn Impingers PEI Job Nun.	4181
Analyst: E. Coppedge/A. Carver	Date:	June 19, 1989
Chromiun Standard Calibration Curve 6
Peak Area	Percent Predicted	Percent
Cone. 		 Deviation Cone.	Deviation
(ppb)	Run 1 Run 2 Average	(ppb)	from Actual
23 2296400 2190500 2243450 2.4X 23.72	4.6X
54 4389800 4377300 4383550 0.1X 52.35	-3.OX
108 8626100 8550400 8588250 0.4X 108.60	0.6X
Standard Curve	Slope 74748.28 Intercept 470246
Run 1 Run 2	Chromiun	Total
Sample



Percent
Cone.
Vol une
Catch
Spiked
R
ID
Area
Area
Average Deviation
PPb
(ml)
(ug)
(ug)
(X)
C-108
7710700
7396700
7553700
2.1X
94.76
100
9.48


C-216
14534000
14552000
14543000
0.1X
188.27
100
18.03


C-324
23593000
25001000
24297000
2.9X
318.76
100
31.88


C-I-9
16992000
17162000
17077000
0.5X
222.17
81
18.00
18.83
95.6
1-1
7925400
8034000
7979700
0.7X
100.46
163
16.38
18.83
87.0
1-2
4376200
4320500
4348350
0.6X
51.88
214
11.10
18.83
59.0
1-3
2229700
2235900
2232800
0.1X
23.58
196
4.62
18.83
24.5
1-4
5355600
5364800
5360200
0.1X
65.42
212
13.87
18.83
73.7
1-5
559B700
5492500
5545600
1.0X
67.90
192
13.04
18.83
69.2
1-6
6123800
5960700
6042250
1.3X
74.54
188
14.01
18.83
74.4
1-7
6450300
6393100
6421700
0.4X
79.62
175
13.93
18.83
74.0
1-8*
1962100
2201400
2081750
5.7X
21.56
184
3.97
18.83
21.1
F-A
1255300
1206900
1231100
2.OX
10.18
247
2.51
13.18
19.1
F-B
1798400
1814400
1806400
0.4X
17.88
253
4.52
13.18
34.3
F-C
1977900
1957400
1967650
0.5X
20.03
253
5.07
13.18
38.5
F-D
1684300
1683700
1684000
o.ox
16.24
210
3.41
13.18
25.9
MOTES:
C-108--C-324 are dilutions of the solution used to spike the inpinger
solutions. The actual spiking amount is equal to c-108.
C-1 -9 was a control impinger solution sent to the field.
All impinger solutions were vaccun filtered through 0.45un Teflon
filter paper. The samples ucre then boiled to remove the IPA for analysis,
and a cation resin was used to remove the Na+.
Stack gas was drawn through the solution for indicated minutes.
1-8 contained some silica gel in the recovered solution.
F-A--F-D filters were spiked with 13.18 ug. Stack gas was drawn through the filter for
60 min. Filters were digested with 0.1N NaOH arid vaccun filtered.

-------
Chromium Analytical Data Sheet
Job Name:Chromiun Filters PEI/SRI Job Nun.	A181
Analyst: E. Coppedge/A. Carver Date:	June 14, 1989
Chromium Standard Calibration Curve	6
Peak Area Percent Predicted	Percent
Cone. 			Deviation Cone.	Deviation
(ppb) Run 1 Run 2 Average (ppb)	from Actual
23 1124200 1150800 1137500 1.2X 21.92	-4.7X
54 3598100 3688000 3643050 1.2X 55.70	3.1X
108 7429600 7524200 7476900 0.6X 107.38	-0.6X
Standard Curve	Slope 74180.23	Intercept	-488631
Run 1 Run 2	Chromium	Total
Sample 		 Percent	Cone.	Volume Catch Spiked R
ID Area Area Average	Deviation ppb	(ml) (ug) (ug) (X)
C-6
C-7
1173100 1027300 1100200
NO PEAK NO PEAK
6.6X 21.42
278
5.95 10.8 55.1
C-8
C-9
2419600 2376300 2397950
4383900 4422400 4403150
0.9X
0.4X
38.91
65.94
220 8.56 10.8 79.3
136 8.97 10.8 83.0
F-1
F-2
688600 672130 680365
815170 722990 769080
1.2X
6.OX
15.76
16.95
505 7.96 10.8 73.7
427 7.24 10.8 67.0
NOTES:
Instrument Setting AU: 0.5
C-6 was sent to the field as a control sample.
It was digested with water.
C-7 was sent to the field as a control sample,
a 0.2M phosphate buffer was used for digestion.
C-8 was a laboratory control digested uith water.
C-9 was a laboratory control digested uith the
0.1 N NeOH
F-1 was an actual field sample. Stack gas uas
pulled for 15 min. The filter was digested
in 0.1 N NeOH.
F-2 was an actual field sample uith gas pulled for
30 min, also digested in 0.1N NaOH. The ash
on this filter was a yellow chalky color and very
light compared to the other filters.

-------
Chromiun Analytical Data Sheet
Job NameiChrofliiun Filters SRI	Job Nun.	4181
Analyst: E. Coppedge/A. Carver Date:	June 15, 1989


Chromi im
Standard
Calibration Curve 6




Peak Area

Percent
Predicted

Percent

Cone.



Deviation
Cone.

Deviation

(PPb)
Run 1
Run 2
Average

(PPb)

from Actual

23
1945300
1977100
1961200
0.8X
21.03

-8.6X

54
5481100
5212600
5346850
2.5X
57.10

5.7X

108
9967400
10069000
10018200
0.5X
106.87

-1.0X

Standard Curve
Slope 93861.74
Intercept
-12724




Run 1
Run 2


Chromiun

Total

Sanple



Percent
Cone.
Volune
Catch Spiked
R
ID
Area
Area
Average Deviation
PPb
(ml)
-------
Radioactive Chromiun Analytical Data
PE! Filter & Impinger	Job. Nunber: 4181
Anna C. Carver	Date: June 19, 1989

CT


Sample

R
Sample ID
#
Count 1 Count 2
Average
Vol (ml) Vol Cor.
(X)
C-I-9:Filter after digestion
117
2431
2431
81
2431

C-l-9: 5 ml filtrate
118
199650
199650
5
3234330
99.9
1-1: Filter after digestion
101
4736
4736
163
4736

1-1: 5 ml filtrate
102
89463
89463
5
2916494
99.8
1-2: Filter after digestion
103
3972
3972
214
3972

1-2: 5 ml fiItrate
104
58276
58276
5
2494213
99.8
1-3: Filter after digestion
105
4376
4376
196
4376

1-3: 5 ml filtrate
106
78579
78579
5
30BO297
99.9
1-4: Filter after digestion
107
5940
5940
214
5940

1-4: 5 ml fiItrate
108
69401
69401
5
2970363
99.8
1-5: Filter after digestion
109
7949
7949
192
7949

1-5: 5 ml fiItrate
110
58276
58276
5
2237798
99.6
1-6: Filter after digestion
111
10850
10850
188
10850

1-6: 5 ml filtrate
112
75480
75480
5
2838048
99.6
1-7: Filter after digestion
113
4234
4234
175
4234

1-7: 5 ml fiItrate
114
83306
83306
5
2915710
99.9
1-8: Filter after digestion
115
5011824
5011824
184
5011824

1-8: 5 ml fiItrate
116
21542
21542
5
792746
13.7
F-A: Filter after digestion
145
107705
107705
247
107705

F-A: 15 ml fiItrate
146
66428
66428
15
1093848
91.0
F-B: Filter after digestion
147
929795
929795
253
929795

F-B: 15 ml f11trate
148
75442
75442
15
1272455
57.8
F-C: Filter after digestion
149
809712
809712
253
809712

F-C: 15 ml fiItrate
150
82366
82366
15
1389240
63.2
F-D: Filter after digestion
151
114B05
114805
210
114805

F-D: 15 ml fiItrate
152
84082
84082
15
1177148
91.1
-------
Radioactive Chromiun Analytical Data
SRI Filter Analysis	Job. Nurtoer: 4181
Anna C. Carver	Date: June 15, 1989

CT


Sample

R
Sample ID
#
Count 1 Count 2
Average
Vol (ml)
Vol Cor.
(X)
F-1: Filter after digestion
47-50
681038
681038
504
681038

F-1: 9 ml f iItrate
51
24752
24752
9
1386112
67.1
F-2: Filter after digestion
52-54
496081
496081
427
496081

F-2: 9 ml fiItrate
55
33961
33961
9
1611261
76.5
F-3:Filter after digestion
69-72
410934
410934
513
410934

F-3: 8 ml fiItrate
73
19553
19553
8
1253836
75.3
F-4: Filter after digestion
74-76
236544
236544
500
236544

F-4: 8 ml fiItrate
77
25906
25906
8
1619125
87.3
F-5: Filter after digestion
78-80
390074
390074
462
390074

F-5: 8 ml fiItrate
81
284 97
28497
8
1645702
80.8
-------
TO:	W. G. DeWees
FROM:	Scott Steinsberger
DATE:	August 20, 1989
SUBJECT:	Results From Second Methods Development Effort
This memorandum is to summarize the results from the second methods
development effort conducted by PEI and Entropy. The results of the spiked impinger
experiments are summarized in Table 1. Recoveries of spiked native chromium (10.8) from
two 80% IPA/20% 1.0 N NaOH impinger reagent samples exposed to 2 hours of flue gas
were 81.7% and 74.8%. For the IPA/NaOH reagent relative to a spiked IPA/NaOH
control sample where 8.9 ug was recovered, the recoveries were 99.0% and 90.6% for the
two samples. The recovery of spiked slCr"*6 in the two samples were in agreement with the
native Cr recovery, at 75.7% and 81.1%, with recovery in the control sample being 89.8%.
The 51Cr+6 analysis protocol involves filtering the whole sample followed by IC separation
of the filtrate. With this protocol, only the soluble radioactivity in the filtrate that coelutes
with the native hexavalent chromium is considered to be hexavalent MCr. The
radiochromatograms for the IPA/NaOH samples are shown at the top of Figure 1. Note
that the majority of the radioactivity elutes from the IC between 4.5 and 6 minutes, where
the native hexavalent chromium elutes.
A field blank sample was also collected with the IPA/NaOH set. Unexplainably, the
field recovery was about 50%, and the total recovered radioactivity was low also.
The 0.5 M phosphate buffer resulted in very low recoveries; about 25% for native
Cr+S and about 15% for 5ICr+<. Both control and filed blank samples should good
recoveries of both Cr+6 and slCr*"6. While the phosphate buffer did not prevent conversion
during sampling, the radiochromatograms (see the bottom of Figure 1) for the control
sample and the field samples (P-l and P-2) graphically demonstrate the ability of the IC to
separate soluble 51Cr species.
The results for the Method 5-type filters sample are presented in Table 2. A set of
four unspiked filters were exposed to flue gas for 2 hours to obtain a representative
particulate loading. Two of the samples were extracted and analyzed for Cr""6. Note that
2.8 and 2.9 ug were found on the unspiked filters. The remaining two samples and two
blank filters were spiked with a mixture of s,Cr+s and native Cr"6, and the four filters were
extracted. Recoveries for the blank filters were 87 % and 82.4 % for native Cr+<, with 98%
of the recovered MCr being Cr+<. For the flue-gas exposed, spiked filters, the recoveries
were 66.1% and 67.3% for native Cr+< with 91% of the recovered 51Cr being Cr+<. The
expected amount of native Cr (13.7 ug) was calculated as the sum of the spike amount (10.8
ug) and the amount found on the unspiked exposed filters (2.87 ug).
A-7
-------
The results for the filter spiked with native Cr*' and 5ICr+< prior to exposure for 2
hours to flue gas are also presented in Table 2. The recoveries for three exposed filter
samples, relative to the expected value of 13.7 ug, were 47% to 48.3%. Relative to the
spiked blank filter control samples, the recoveries were 70% to 72%. Of the recovered ^Cr,
61.5% to 73.3% was found to "Cr*4.
In summary, the IPA/NaOH appeared to be the best collection media for this test
series, with both the native and radioactive Cr numbers agreeing quite well. The precisions
of the measurements were also good. The filter experiments indicate that conversion during
sampling and/or matrix effects on recovery by extraction do occur, and can be measured,
semi-quantitatively, with "Cr*4 and/or native Cr+< spikes.
A-8
-------
TABLE 1 RECOVERY OF NATIVE AND LABELED HFXAVALENT CHROMIUM SPIICE IN IMPINGER REAGENT
Sample Identity
Native Hexavalent Chromiun
Spiked Hexavalent "chromiun (total counts)
Expected Found X of X of
(ug) (ug) Expected Control
Total Residue Soluble
IC Separat ion
Percent o1 Total
bl,. • 3 bl~. • 6
Cr Cr
"Cr" slCr'*
Spiked BOX IPA/20X 1.0 N NaOH Inpirvger Reagent Exposed to 2 Hours of Flue Gas
IPA/NbOH Control
1PA/NaOH Sample (1-2)
IPA/NbOH Sample (1-3)
IPA/NaOH Field Blank
10.8 8.9 82.5X NA
10.8 8.8 81.7X 99.OX
10.8 8.1 74.8X 90.6X
10.8 4.9 45.4X 55.OX
536845 450 536395
414468 26458 388010
39B35B 15750 382608
230848 2B4 230564
54571 4B1824
74057 313953
5963? 322976
47V7V 1B25B5
10.21 89.BX
24.3X 75.7X
18.91 81.11
20.9X 79.1X
Spiked 0.5 H Phosphate Buffer Impinger Reagent Exposed to 2 Hours of Flue Cas
0.5
H P04 Control
10.6
11.2
103.5X
NA
NC
NC
NC
11100
151300
6.BX
93.2X
P04
Sanple (P-1)
10.8
2.8
26.2X
25.3X
563977
5579
558398
48 75 09
70889
B7.4X
12.61
P04
Sanple (P-2)
10.8
3.1
28.51
27.5X
541406
6654
534752
453351
B1401
85. OX
15.OX
P04
Field Blank
10.8
9.2
B4.7X
81.81
435957
908
435049
41561
393488
9.7X
90.3X
-------
TABLE 2 RECOVERY OF NATIVE AND LABELED HEXAVALENT CHROMIUM SPIKE ON GLASS FIBER FILTERS


Nat ive
Hexavalent Chromiun

Spiked Hexavalent '
Chromiun (total counts)



Expected
(ug)
Found
(ug)
X of
Expec ted
X of
Control



IC Separation
Percent
of Total
Sample Identity
Total
Residue
Soluble
blCr '3
"Cr "'
"Cr'J
L
"Cr"


Post
Test
Spiked Quartz Method Slype Filters Exposed to 2 Hours of
Flue Gas


Spiked Control
10.8
9.4
87.OX
NA
867869
1619
866250
18261
84 7989
2.3X
97. 7X
Spiked Control
10.8
8.9
82. 4X
NA
858487
3962
854525
8804
845721
1.5X
98.5X
Post Test Spike (B 2)
13.7
9.0
66.1X
98. BX
866047
17997
848050
81345
766705
11.5X
88.5X
Post Test Spike (B-4)
13.7
9.2
67.3X
100.5X
872866
21841
851025
56920
794105
9. OX
91. OX
Exposed FiIter (B-1)
0
2.9
NA
NA
394
9
385
NA
NA
NA
NA
Exposed F i I ter (B-3)
0
2.8
NA
NA
292
12
280
NA
NA
NA
NA

Pretest
Spiked Quartz Method 5-Type
FiIters Exposed to
2 Hours of
Flue Gas




Spi ked Fill.
(1-1)
13.7
6.4
47.OX
70.3X
458210
113495
344715
63106
281609
38.5X
61.5X
Spiked Filt.
(1-2)
13.7
6.4
47.OX
70.2X
441934
126759
315175
38554
276621
37.4X
62.6X
Spi ked Filt.
(1-4)
13.7
6.6
48.3X
72. IX
444449
91019
353430
27489
325941
26. 7X
73.3X
Spi ked Filt.
FB
10.8
0.6
5 .31
6.2X
430161
176761
253400
118253
135147
68.6%
31.4X
-------
FIGURE 1. 51Cr RAD^OCHROMATOGRAMS OF SPIKED [\-PNGER
SOLUTIONS EXPOSED TO FLUE OAS FOR 2 HOURS
100 -
i
90 -
1	1 1	1	1	1	1	1	1— 1			1	1	1	1	1	
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
Time
IPA control
"• IPA Sample (1-3) & IPA Sample (1-2)
T3
C
a
e
~
O
-C
c
3
o

~i	1	1	1	1	1	1	:	1	i	1	1	1	1	1	1
0.5 1.0 1.5 2.0 2.3 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
Time
P04 Control
P04 Sample (P-1)
a P04 Sample (P-2)
A-ll
-------
FIGURE 2. 51Cr RADOCHROMATCGRAMS OF SPKED J^THOO 5
FLTERS EXPOSED TO FLUE OAS FOR 2 HOURS
"O
c
o
a
3
O
(E
C
3
O
(J
o
Spiked i

260

240

220

200

180
n
TJ
c
160
~

n
3
140
o

c.
~—
120
n
100
c

3
Q
80
CJ


60

40

20

0

-20
& Spiked Filler 2 7 Spiked Tiller Field Blank
4-
-i	1	1	1	1	1	1	1	1	r	i i r
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0	5.5 6.0 6.5 7.0
Control Filter 1 + Control Filter 2 ° Post Spiked Filter 2 ^	Post Spikea Filer 4
A-12
-------
Appendix B
Method for Determination of Hexavalent Chromium
using Recirculating Reagent Impinger Train
B-l
-------
DRAFT METHOD - May 1989
METHOD Cr*6 - DETERMINATION OF HEXAVALENT CHROMIUM FROM STATIONARY SOURCES
1.	Applicability and Principle
1.1	Applicability. This method applies to the determination of hexavaient
chromium (Cr*6) emissions from cooling towers, hazardous waste incinerators,
municipal waste combustors, and sewage sludge incinerators. The method, while
not designed to capture insoluble forms of chromium, may, with the approval of
the Administrator, be used to measure total chromium.
1.2	Principle. For cooling towers, the Cr*6 emissions are collected
isokinetically from the exit plane of the tower. For incinerators and
combustors, the Cr*6 emissions are collected isokinetically from the source. The
emission samples are collected with a recirculatory train where the impinger
reagent is continuously recirculated to the nozzle. The impinger train samples
are analyzed for Cr*6 on an ion chromatograph equipped with a post-column reactor
and a visible wavelength detector (IC/PCR). The IC/PCR separates the Cr*6 as
chromate (CrO^*) from other components in the sample matrix that may interfere
with the Cr*6-specific diphenylcarbazide reaction that occurs in the post-column
reactor.
2.	Range, Sensitivity, Precision, and Interference
2.1	Range. Employing a preconcentration procedure, the lower limit of the
range can be extended to 7-5 nanograms per liter (ng/1) of impinger sample. With
sample dilution, there is no upper limit.
2.2	Sensitivity. A minimum detection limit of 2.4 ng/1 of impinger sample
can be achieved by preconcentration.
2.3	Precision. The overall precision of sample collection and analysis
for a cooling tower containing 250 micrograms per liter (ug/1) of Cr*6 in the
cooling water and emitting 2.5 micrograms per dry standard cubic meter (ug/dscm)
is 31 percent. The precision of the IC/PCR with sample preconcentration is 9
percent. No precision measurements have been made for cooling towers emitting
less Cr*6 or for incinerators or combustors.
2.4	Interference. Components in the sample matrix may cause Cr*6 to
convert to trivalent chromium (Cr*3) or cause Cr*3 to convert to Cr*6.
Conversion of Cr*6 to Cr*3 is more likely to occur. Radioactive Cr*6 (51Cr*6)
can employed to monitor conversion within the train. For the IC/PCR analysis,
only compounds that coelute with Cr*6 and affect the diphenylcarbazide reaction
will cause interference. The known interferents to the diphenylcarbazide
reaction will not coelute with Cr*6 on a properly operated ion chromatograph.
3.	Apparatus
3-1 Sampling Train. A schematic of the recirculatory sampling train
employed in this method is shown in Figure Cr*6-1. The recirculatory train is
readily assembled from commercially available components. All portions of the
train that will come into contact with the sample are either glass or plastic,
and are to be cleaned with 0.1 N nitric acid (HN03) and rinsed thoroughly with
distilled, deionized water (DI H20) before and after each sampling run.
The metering system is identical to Method 5 with the exception of the use
of a propeller anemometer for cooling tower testing. The sampling train consists
of the following components:
3.1.1 Probe Nozzle. Glass or Teflon with a sharp, tapered leading edge.
The angle of taper shall be <30° and the taper shall be on the outside to Figure
B-2
-------
TEFLON
IMP1NGERS-
td
i
u>
ASPIRATOR
NOZZLE
TEFLON
LINES
RECIRCULATING
LIQUID
TO
METHODS-TYPE
METERBOX
2SO ml
0.1 N KOM/
BOY.IPA
100 ml
0.1 N KOH/
B0V.1PA
100 ml
0.1 N KOH/
BO% IPA
ICE BATH
Schematic of recirculatory sampling train.
-------
preserve a constant internal diameter. The probe nozzle shall be of the button-
hook or elbow design, unless otherwise specified by the Administrator.
A range of nozzle sizes suitable for isokinetic sampling should be
available, e.g., 0-32 to 1.27 cm (1/8 to 1/2 in.) (larger if higher volume
sample trains are used) inside diameter (ID) nozzles in increments of 0.16 cm
(1/16 in.). Each nozzle shall be calibrated according to the procedures
outlined in Section 6.
3-1.2 Teflon Aspirator. Teflon aspirator capable recirculating absorbing
reagent at 50 ml/min while operating at 0.75 cfm. Teflon fittings, Teflon
ferrules, and stainless steel nuts to connect glass nozzle, Teflon
recirculation, and Teflon sample line to aspirator.
3-1-3 Teflon Sample Line. Teflon, 3/8" outside diameter (O.D.) and 1/4"
inside diameter (I.D.), of suitable length to connect aspirator to first Teflon
impinger.
3-1.4 Teflon Recirculation Line. Teflon tubing, 1/4" O.D. and 1/8" I.D.,
of suitable length to connect first impinger to aspirator. The recirculation
line should be fitted with a non-restricting shut-off device constructed out of
Teflon or equivalent inert material.
3.1.5 Recirculation Control Valve. Inert valve system with 1/8" I.D. flow
path constructed of plastic, Teflon, and/or glass installed at a convenient poir.t
in the Teflon recirculation line.
3-1.6 Teflon Impingers. Three Teflon impingers, 2" diameter by 8", with
vacuum-tight 3/8" O.D. Teflon compression fittings. Inlet fittings on impinger
top to be bored through to accept 3/8" O.D. tubing as impinger stem. Impinger
stem to extend to 1/2" from impinger bottom. First impinger to have bottom cap
with 1/4" O.D. Teflon compression fitting for recirculation line. An optional
knockout impinger may be used for high moisture sources.
3.1.7 Silica Gel Impinger and Thermometer. Vacuum-tight impinger, capable
of containing 200 g of silica gel, with compatible fittings. Thermometer, at the
outlet of the silica gel impinger, to monitor the exit temperature of the gas.
3-1.8 Metering System, Barometer, and Gas Density Determination Equipner.t.
Same as Method 5. Section 2.1.8 through 2.1.10, respectively.
3-2 Sample Recovery. Clean all items for sample handling or storage with
8 N HNOj solution by soaking, where possible, and rinse thoroughly with DI H?3
before use. THe following items are needed:
3-2.1 Wash Bottles. Two polyethylene wash bottles, for DI H20 and HNOj
rinse solution.
3-2.2 Sample Storage Containers. Polyethylene, with leak-free screw cap,
5C0-ml or 1000-ml size.
3.2.3	Graduated Cylinder and/or Balance.
3.2.4	Plastic Storage Containers. Air tight containers to store silica
gel.
3.2.5	Funnel and Rubber Policeman. To aid in transfer of silica gel from
impinger to storage container; not necessary if silica gel is weighed directly
in the impinger.
3-3 Sample Preparation for Analysis. Sample preparation prior to analysis
includes pretreatment with phosphate buffer and aluminum sulphate, if total Cr is
to be determined, followed by filtration to remove particulate matter and
precipitate caused by the phosphate buffer and aluminum sulphate treatment.
Additional sample pretreatment to remove hydroxide ion will be necessary if
sample preconcentration is required.
B-4
-------
3.3-1 Beakers, Funnels, Volumetric Flasks, Volumetric Pipets, and
Graduated Cylinders. Assorted sizes, plastic or glass, for preparation of
samples, sample dilution, and preparation of calibration standards. Prepare
initially following procedure described in Section 5-1-3 rinse between use
with 0.1 N HNO, and DI H20.
3.3.2	Filtration Apparatus. Teflon, or equivalent, for filtering samples.
Teflon impinger components have been found to be satisfactory as a sample
reservoir for pressure filtration using nitrogen.
3.4 Analysis.
3.4.1 IC/PCR System. High performance liquid chromatograph pump, sample
injection valve, post column reagent delivery and mixing system, and a visible
detector, capable of operating at 530 nm, all with a non-metallic flow path. An
electronic recording integrator operating in the peak ^rea mode is recommended,
but other recording devices and integration techniques are acceptable provided
the repeatability criteria and the linearity criteria for the calibration curve
described in Section 5-5 can be satisfied. A sample loading system will be
required if preconcentration is employed.
3-4.2 Analytical Column. A Dionex AS-7, or equivalent non-metallic column
with anion separation characteristics provided resolution described in Section
A non-metallic guard column with the same ion-exchange material is
recommended.
3.4.3	Preconcentration Column. A Dionex AG-7, or equivalent non-metallic
column with acceptable anion retention characteristics and sample loading rates
as described in Section
4. Reagents
All reagents should, at a minimum, conform to the specifications
established by the Committee on Analytical Reagents of the American Chemical
Society, where such specifications are available. All prepared reagents should
be checked by IC/PCR analysis for Cr*6 to assure that contamination is below the
analytical detection limit for direct injection or, if selected,
preconcentration. If total Cr is also to be determined, the reagents should
also be checked by the analytical technique selected to assure that
contamination, is below the analytical detection limit.
4.1	Sampling
4.1.1	Water. Deionized and/or distilled water. It is recommended that
water blanks be checked prior to preparing reagents sampling to ensure that the
Cr*6 content of is less than the analytical detection limit.
4.1.2	Potassium Hydroxide Solution, 1.0 N. Dissolve 56.1 g of potassium
hydroxide (KOH) in 1000 ml of water.
4.1.3	90# Alkaline Isopropanol (IPA/K0H). Mix 8C0 ml of 100* iscprcpanol
(IPA) with 200 ml of 1.0 N KOH. It is recommended that 80# IPA/K0H reagent
blanks be checked prior to sampling to ensure that the Cr*6 content of is less
than the analytical detection limit.
4.2	Sample Recovery. The reagents used in sample recovery are as follow:
4.2.1	Water. Approximately 300 to 400 ml of water for rinsing the
sampling train; significant levels of Cr*6 must not be present in the water.
(See Section 4.1.1.)
4.2.2	Nitric Acid, 0.1 N. Add 6.3 ml of concentrated HN03 (70 percent) to
a graduated cylinder containing approximately 900 ml of water. Dilute to 1000 ml
with water, and mix well.
4.3	Sample Preparation
4.3.1 Water. Same as Section 4.1.1.
B-5
-------
4.3-2 Phosphate Buffer, 2 M at pH 9 (for total Cr determination).
Dissolve 34.84 g of anhydrous dipotassium hydrogen phosphate (K2HP0ij) in 60 xl or
45.65 g of dipotassium hydrogen phosphate hydrate (K2 HPO^ ¦ 3H2 0) in 50 ml. Ad jus*'
volume to 95 with DI water. Adjust the pH to 9 with concentrated phosphoric
acid, and make a final volume adjustment to 100 ml.
4.3-3 Aluminum Sulphate, 0-74 M (for total Cr determination). Dissolve
49-32 g of hydrated aluminum sulphate (Al2(SO^)3•l8H20) in 50 ml of water, and
adjust the final volume to 100 ml.
4.3-4 Filters. Teflon membrane, or equivalent, filters with 0.45-micron or
smaller pore size to remove insoluble material.
4.4 Analysis
4.4.1	Chromatographic Eluent. An effective eluent for use with the Dionex
AS-7 analytical column is a pyridine dicarboxylic acid (PDCA)-based solution.
Heat 3-34 g of pyridine-2,6-dicarboxylic acid (PDCA) in 500 ml of
degassed, deionized water. Continue heating until the PDCA is completely
dissolved. Add 5-36 g of disodium hydrogen phosphate heptahydrate. Add 15-0 g
of sodium iodide. Add 38-5 S of ammonium acetate. Add 1.10 g of lithium
hydroxide monohydrate. Dilute to 1 liter with degassed, deionized water. Pipet
100 ml of this stock eluent into 500 ml of degassed, deionized water and dilute
to 1 liter. Other combinations of eluents and/or columns may also be employed
provided peak resolution, as described in Section	repeatability and
linearity, as described in Section	and analytical sensitivity are
acceptable.
4.4.2	Post Column Reagent. An effective post column reagent for use with
the PDCA-based chromatographic eluent described in Section 4.3-1 is a
diphenylcarbazide (DPC)-based system. Dissolve 0.5 g of 1,5~diphenylcarbazide
(DPC) in 100 ml of HPLC-grade methanol. Add to 500 ml of degassed, deionized
water containing 25 ml of $6% spectrophotometry grade sulfuric acid. Allow
30 minutes for the above to come to solution. Dilute to 1 liter with degassed,
deionized water.
4.4.3	Cation Exchange Resin. Bio-Rad AG 5OW-X8, 200-400 mesh, or
equivalent, in the hydrogen ion (H*) form for sample treatment prior to
preconcencration. Sample pretreatment is necessary remove hydroxide anions that
interfere with the anion retention capacity of the preconcentration column.
Prepacked, disposable resin cartridges can be employed to treat up to 10 ml cf
sample. Larger sample aliquots can be treated by slurrying with the cation
exchange resin.
4.4.4	Cr*6 Calibration Standard. Prepare Cr*6 standards from hydrated
sodium chromate (Na2C0a) that has been desiccated to dryness prior to use. To
prepare a 1000 ug/cl Cr*6 stock solution, dissolve 3-461 g of Na2CrO^¦4H20 in
1 liter of 18 K-ohm deionized water. To prepare working standards, dilute the
stock solution, to the chosen standard concentrations for instrument calibration
with IPA/KCH reagent and water to achieve a matrix similar to the actual field
samples.
4.4.5	Performance Audit Sample. A performance audit sample shall be
obtained from the Quality Assurance Division of EPA and analyzed with the field
samples. The mailing address to request samples is:
U. S. Environmental Protection Agency
Atmospheric Research And Exposure Assessment Laboratory
Quality Assurance Division
Source Branch, Mail Drop 77~A
Research Triangle Park, North Carolina 27711
B-6
-------
The audit sample should be prepared in a suitable sample matrix at a concentra-
tion similar to the actual field samples.
5. Procedure
5.1 Sampling. The complexity of this method is such that to obtain
reliable results, testers should be trained and experienced with test procedures.
5.1.1	Pretest Preparation. All components shall be maintained and
calibrated according to the procedures described in APTD-0576. unless otherwise
specified herein.
Rinse all sample train components from the glass nozzle up to the silica gel
impinger and sample containers with hot tap water followed by washing with hot
soapy water. Next, rinse the train components and sample containers three times
with tap water followed by three rinses with distilled or deionized water. All
the components and containers should then be soaked overnight, or a minimum of -
hours, ir. a 10 percent (v/v) nitric acid solution, then rinsed three times with
distilled or deionized water. Allow the components to air dry prior to covering
all openings with Parafilm, or equivalent.
5.1.2	Preliminary Determinations. Same as Method 5. Section 4.1.2.
5.1.3	Preparation of Sampling Train. Place 250 ml of IPA/KOH reagent in
the first impinger. If a knockout impinger is to be used for high moisture
sources (collection of more than 150 ml), the impinger should be sized to hold
the entire amount of flue gas condensate plus an additional 50# to contain the
recirculation liquid. The volume of IPA/KOH added to the knockout impinger will
be equal to half of the anticipated volume of flue gas condensate. Place 50 ml
of IPA/KOH in the second impinger and, if the knockout impinger is used, place 50
ml of IPA/KOH in the third impinger. The next Teflon impinger is left dry.
Place a preweighed 200- to 300"E portion of indicating silica gel in the final
impinger.
Retain reagent blanks of the IPA/KOH equal to the volumes used with the
field samples.
5.1.4	Leak-Check Procedures. Follow the leak-check procedures given in
Method 5. Section 4.1.4.1 (Pretest Leak-Check), Section 4.1.4.2 (Leak-Checks
During the Sample Run), and Section 4.1.4.3 (Post-Test Leak-Checks).
5.1.5	Sampling Train Operation. Follow the procedures given in Method 5.
Section 4.1.5. At the start of the sampling, open the recirculation line shut
off device. The shut-off device must be closed prior to any interruption in the
flow of sample gas through the aspirator to the first impinger to prevent loss of
sample. Approximately one minute before the completion of the sampling run,
close the shut-off device to permit the recirculating liquid to be cleared from
the sample line.
For each run, record the data required on a data sheet such as the one shown
in Figure 5~2 of Method 5-
5-1.6 Calculation of Percent Isokinetic. Same as Method 5. Section 4.1.6.
5.2 Sample Recovery. Begin cleanup procedures as soon as the probe assembly is
removed from the stack at the end of the of the sampling run.
The probe assembly should be allowed to cool prior to sample recovery.
Disconnect the umbilical cord from the sampling train. When the probe assembly
can be safely handled, wipe off all external particulate matter near the tip of
the nozzle, and cap the nozzle prior to transporting the sample train to a clear,
up area that is clean and protected from the wind and other potential causes of
contamination or loss of sample. Inspect the train before and during disassembly
and note any abnormal conditions.
B-7
-------
5.2.1	Container No. 1 (Impingers 1 through 3. or 4, if knockout is used)
Disconnect the first impinger from the second impinger and disconnect the
recirculation line from the aspirator. Through the recirculation line, drain the
impinger contents into a precleaned graduated cylinder by opening the shut-off
device. Add the contents of the second, third (if used), and dry impinger to the
graduated cylinder, and measure the volume of the liquid to within 0.5 nl•
Record the volume of liquid present as this information is required to calculate
the moisture content of the flue gas sample. Transfer the sample from the
graduated cylinder to a precleaned polyethylene sample container. With water,
rinse the insides of the glass nozzle, the aspirator, the sample and
recirculation lines, the impingers, and the connecting tubing three times, and
combine the rinses with the impinger solution in the sample container. Mark the
height of the fluid level on the container or, alternatively if a balance is
available, weigh the container and record the weight to permit determination if
leakage occurs during transport. Label the container clearly to identify its
contents.
5.2.2	Container No. 2 (HNOj rinse, optional for total Cr). Repeat the
rinse procedure with 0.1 N HN03 , and combine the rinses in a separate precleaned
polyethylene container for possible total Cr analysis, or discard the HN03 rinse.
Repeat the rinse procedure a final time with water, and discard the water rinses.
Mark the height of the fluid level on the container or, alternatively if a
balance is available, weigh the container and record the weight to permit
determination if leakage occurs during transport. Label the container clearly to
identify its contents.
5.2.3	Container No. 3 (Silica Gel). Note the color of the indicating
silica gel to determine if it has been completely spent, and make a notation of
its condition. Quantitatively transfer the silica gel from its impinger to the
original container, and seal the container. A funnel and a rubber policeman may
be used to aid in the transfer. The small amount of particulate that may adhere
to the impinger wall need not be removed. Do not use water or other liquids to
transfer the silica gel. Alternatively, if a balance is available in the field,
record the weight of the spent silica gel (or the silica gel plus impinger) to
the nearest 0.5 g-
5.2.4	Container No. 4 (IPA/KOH Blank). Once during each field test, place
a volume of sampling reagent equal to the volume placed in the sampling train
into a precleaned polyethylene sample container, and seal the container. Mark
the height of the fluid level on the container or, alternatively if a balance is
available, weigh the container and record the weight to permit determination if
leakage occurs during transport. Label the container clearly to identify its
contents.
5.2.5	Container No. 5 (Water 31ank). Once during each field test, place a
volume of water equal to the volume employed to rinse the sampling train into a
precleaned polyethylene sample container, and seal the container. Mark the
height of the fluid level on the container or, alternatively if a balance is
available, weigh the container and record the weight to permit determination if
leakage occurs during transport. Label the container clearly to identify its
contents.
5-2.6 Container No. 5 (0.1 N HN03). Once during each field test, if total
Cr is to be determined, place a volume of 0.1 N HNO^ reagent equal to the volume
employed to rinse the sampling train into a precleaned polyethylene sample
container, and seal the container. Mark the height of the fluid level on the
container or, alternatively if a balance is available, weigh the container
B-8
-------
and record the weight to permit determination if leakage occurs during transport.
Label the container clearly to identify its contents.
5.3 Sample Preparation. For determination of Cr*6, the sample should be
filtered to remove any insoluble matter. For total Cr determination, the sample
is first treated with phosphate buffer, pH adjusted to 9 with phosphoric acid,
and treated with aluminum sulphate prior to filtration.
5.3-1 Container 1 (Impinger Sample) For determining total Cr, add 1 ml of
phosphate buffer per 100 ml of sample. Check the pH and adjust to 9 with
concentrated phosphoric acid, if necessary. Add 0.1 ml of aluminum sulphate
solution per 100 ml of sample.
For determining either Cr'6 or total Cr, filter the entire impinger sample
through the 0.45-um filter, and collect the filtrate. Rinse the sample
container with water three separate times and pass these rinses through the
filter, and add the rinses to the sample filtrate. Determine the final volume of
the filtrate and rinses prior to IC/PCR analysis
Quantitatively recovery the filter and residue, and place them in a vial if
total Cr is to be determined.
5-3-2 Container 2 (HNO^ rinse, optional for total Cr). This sample shall
be analyzed in accordance with the selected procedure for total Cr analysis. At
a minimum, the sample should be subjected to a digestion procedure sufficient to
solubilize all chromium present.
5-3-3 Container 3 (Silica Gel). Weigh the spent silica gel to the nearest
0.5 g using a balance. (This step may be conducted in the field.)
5-4 Sample Analysis. The Cr*6 content of the sample filtrate is
determined by ion chromatography using a post column reactor (IC/PCR). For trace
levels of Cr, a preconcentration system is also used in conjunction with the
IC/PCR. After separation from other sample components, Cr*° forms a specific
complex in the post column reactor with a sym-diphenylcarbazide, and the complex
is then detected by ultraviolet absorbance at a wavelength of 520 nm. The amount
of absorbance measured is proportional to the concentration of the Cr*6 complex
formed. The IC retention time and absorbance of the Cr*6 complex is compared
with known Cr*6 standards analyzed under identical conditions to provide both
qualitative and quantitative analyses.
Prior to sample analysis, establish a stable baseline with the detector set
at the required attenuation by setting the eluent and post column reagent flow
rates to the values recommended by the manufacturer. Inject a sample of water,
and determine if any Cr*6 appears in the chromatogram. If Cr*6 is present,
repeat the water injection until no Cr*6 appears. At this point, the instrument
is ready for use.
First, inject the calibration standards prepared, as described in Section
6.2, to cover the appropriate concentration range, starting with the lowest
standard first. Next, inject, in duplicate, the performance audit sample,
followed by the IPA/K0H field blank, and the field samples. Finally, repeat the
injection of the calibration standards to allow for compensation of instrument
drift. Measure areas or heights of the Cr*6/DPC complex chromatogram peak. The
response for replicate, consecutive injections of samples must be within
5 percent of the average response, or the injection should be repeated until the
5 percent criterion can be met. Use the average response (peak areas or heights)
from the duplicate injections of calibration standards to generate a linear
calibration curve. From the calibration curve, determine the concentration of
the field samples employing the average response from the duplicate injections.
The results for the analysis of the performance audit sample must be within
10 percent for the field sample analysis to be valid.
B-9
-------
6. Calibration
Maintain a written log of all calibration activities.
6.1	Sampling Train Calibration. Calibrate the sampling train components
according to the indicated sections of Method 5- Probe Nozzle (Section 5-1);
Pitot Tube (Section 5-2); Metering System (Section 5-3); Temperature Gauges
(Section 5-5)! Leak-Check of the Metering System (Section 5-6); and Barometer
(Section 5-7)•
6.2	Calibration Curve for the IC/PCR. Prepare working standards from the
stock solution described in Section 4.4.4 by dilution with a IPA/KOH solution
diluted with water to approximate the field sample matrix. Prepare at least
four standards to cover one order of magnitude that bracket the field sample
concentrations. Analyze the standards with the field samples as described in
Section for each standard, determine the peak areas (recommended) or the
peak heights, calculate the average response from the duplicate injections, and
plot the average response against the Cr*6 concentration in ug/1. The individual
responses for each calibration standard determined before and after field sample
analysis must be within 5 percent of the average response for the analysis to be
valid. If the 5 percent criteria is exceeded, excessive drift and/or instrument
degradation may have occurred, and oust be brought under control before further
analyses are performed.
Use linear regression to calculate the formula describing the linear curve.
Employing the regression equation, calculate a predicted value for each
calibration standard with the average response for the duplicate injections.
Each predicted value must be within 7 percent of the actual value for the
calibration curve to be considered acceptable. Remake and/or reanalyze the
calibration standards. If the calibration curve is still unacceptable, reduce
the range of the curve.
7. Calculations
7.1 Dry Gas Volume. Using the data from the test, calculate V , , the
\ 5 C G )
dry gas sample volume at standard conditions as outlined in Section 6.3 of Method
5 •
7-2 Volume of Water Vapor and Moisture Content. Using the date from the
test, calculate V, . . „ . , and B , , the volume of water vapor and the moisture
W^ftCGJ	WS
content of the stack gas, respectively, using Equations 5"2 and 5~3 of Method 3-
7.3	Stack Gas Velocity. Using the data from the test and Equation 2-9 of
Method 2, calculate the average stack gas velocity.
7-4 Total ug Cr*6 Per Sample. Calculate as described below:
m = (S-B) x V
11
Vvr.ere:
m = Mass of Cr*6 in the sample, ug,
S = Analysis of sample, ug Cr*6/ml,
B = Analysis of blank, ug Cr*6/ml,
= Volume of sample after filtration, ml, and,
d = Dilution factor (1 if not diluted).
8. Bibliography
B-10
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before comple'
1. REPORT NO.
EPA/600/R-92/003b
2.
PB92-151562
4. TITLE AND SUBTITLE
EMISSIONS OF METALS, CHROMIUM AND NICKEL SPECIES, AND
ORGANICS FROM MUNICIPAL WASTEWATER SLUDGE INCINERATORS
VOLUME II: SITE 5 EMISSION TEST REPORT - HEXAVALENT
CHROMIUM METHOD EVALUATION	:	
5. REPORT DATE
Marrh 199?
6. PERFORMING ORGANIZATION CODE
7. AUTHORIS)
William G. DeWees, Robin R. Segall
B. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Entropy Environmentalists, Inc.
Research Triangle Park
North Carolina, 27709
10. PROGRAM ELEMENT NO.
B101
1 1. CONTRACT/GRANT NO.
Contract No. 68-CO-0027
Work Assignment No. 0-5
2. SPONSORING AGENCY NAME AND ADDRESS
Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final Report 1989 - 91
14. SPONSORING AGENCY CODE
EPA/600/14
6. SUPPLEMENTARY NOTES
EPA Technical Contact: Dr. Harry E. Boatian, (513) 569-7619, FTS: 684-7619
16. ABSTRACT
At Site 5, three candidate sampling methods and two candidate analytical methods for
hexavalent chromium were assessed. The conversion of hexavalent chromium (Cr*s) to
other valence states of chromium during sampling and sample storage was of primary
concern. Method 5-type samples and impinger train samples were collected by PEI
Associates, Inc. Dilution train samples were collected by Southern Research
Institute. Method 5-type train samples were analyzed by Technology Applications,
Inc.'a (TAI) staff under contract to EPA's Environmental Monitoring Systems
Laboratory (EMSL) in Cincinnati, Ohio. Dilution train samples, Method 5-type
samples, and impinger train samples were analyzed by Entropy Environmentalists, Inc.
TM used an ion chromatograph with post column reaction (IC/PCR) and inductively
coupled argon plaemography/mass spectrometry (ICP/MS) to analyze the Method 5-type
samples. A stable chromium isotope ("Cr*',) spiked onto^ the Method 5 filter prior to
sample collection was used to assess conversion of Cr*?. The BampleB analyzed by
Entropy were collected using a dilution train on an 8.5 in X 11 in glasB fiber
filter, a Method 5-type Bampling train on an 82 mm quartz fiber filter, and an
impinger sampling train wi-th an alkaline impinger reagent, and spiked with native
hexavalent chromium and a radioactively-labeled chromium isotope (B,1Cr*?) . The
samples were analyzed for hexavalent chromium by IC/PCR and for the radioactive
isotopes by scintillation (gamma) counting. EMSL has not released the TAI data. The
preliminary method evaluation testing demonstrated that all sampling methods had
problems with conversion of hexavalent chromium.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Wastewater, sludge disposal,
incinerators, combustion products
Emissions
chromium compounds

18. OIST Rl BUTI ON STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS /Tills Report)
UNCLASSIFIED
21. NO. OF PAGES
88
20. SECURITY CLASS /This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (R«v. 4-77) prkvioui edition ii poiolete
-------