&EPA
United States
Environmental Protection
Agency
Great Lakes
National Program Office
77 West Jackson Boulevard
Chicago, Illinois 60604
EPA905-R94-010
October 1994
Assessment and
Remediation
Of Contaminated Sediments
(ARCS) Program
BENCH-SCALE EVALUATION OF Rcc's
BASIC EXTRACTIVE SLUDGE TREATMENT
(B.E.S.T.)® PROCESS ON CONTAMINATED
SEDIMENTS FROM THE BUFFALO, SAGINAW,
AND GRAND CALUMET RIVERS
United States Areas of Concern
ARCS Priority Areas of Concern
printed on recycled paper
-------
/Bench-Scale Evaluation of RCC's Basic Extractive Sludge
Treatment (B.E.S.T.®) Process on Contaminated Sediments
from the Buffalo, Saginaw, and Grand Calumet Rivers
Prepared by
Clyde J. Dial
Science Applications International Corporation
Cincinnati, Ohio
for the
Assessment and Remediation of Contaminated Sediments (ARCS) Program
Great Lakes National Program Office
U.S. Environmental Protection Agency
Chicago, Illinois
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Chicago, IL 60604-3590 '
-------
DISCLAIMER
The information in this document has been funded wholly or in part by the U.S. Environmental Protection
Agency (EPA) under Contract No. 68-C8-0062, Work Assignment No. 3-52, to Science Applications
International Corporation (SAIC). It has been subjected to the Agency's peer and administrative review and
it has been approved for publication as an EPA document. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-------
ACKNOWLEDGEMENTS
This report was prepared by the Engineering/Technology Work Group (ETWG) as part of the Assessment
and Remediation of Contaminated Sediments (ARCS) program. Dr. Stephen Yaksich, U.S. Army Corps
of Engineers (USAGE) Buffalo District, was chairman of the Engineering/Technology Work Group.
The ARCS Program was managed by the U.S. Environmental Protection Agency (USEPA), Great Lakes
National Program Office (GLNPO). Mr. David Cowgill and Dr. Marc Tuchman of GLNPO were the ARCS
program managers. Mr. Dennis Timberlake of the USEPA Risk Reduction Engineering Laboratory was the
technical project manager for this project. Mr. Stephen Garbaciak of USAGE Chicago District and GLNPO
was the project coordinator.
This report was drafted through Contract No. 68-C8-0062, Work Assignment No. 3-52, to Science
Applications International Corporation (SAIC). Clyde Dial of SAIC was the principal author of the report,
with final editing and revisions made by Mr. Garbaciak prior to publication.
This report should be cited as follows:
U.S. Environmental Protection Agency. 1994. "Bench-Scale Evaluation of RCC's Basic Extractive
Sludge Treatment (B.E.S.T.®) Process on Contaminated Sediments from the Buffalo, Saginaw and Grand
Calumet Rivers," EPA 905-R94-010, Great Lakes National Program Office, Chicago, IL.
-------
ABSTRACT
The Great Lakes National Program Office (GLNPO) leads efforts to carry out the provisions of
Section 118 of the Clean Water Act (CWA) and to fulfill U.S. obligations under the Great Lakes Water
Quality Agreement (GLWQA) with Canada. Under Section 118(c)(3) of the CWA, GLNPO is responsible
for undertaking a 5-year study and demonstration program for the remediation of contaminated sediments.
GLNPO has initiated an Assessment and Remediation of Contaminated Sediments (ARCS) Program to
carry out this responsibility. In order to develop a knowledge base from which informed decisions may be
made, demonstrations of sediment treatment technologies are being conducted as part of the ARCS
Program. Bench-scale studies on the B.E.S.T.® Solvent Extraction Process, which is the subject of this
report, took place at Resources Conservation Company (RCC) in Bellevue, WA on August 5 to 9, 1991.
The specific objectives for this effort were to determine process extraction efficiencies for polychlorinated
biphenyls (PCBs) and polynuclear aromatic hydrocarbons (PAHs); to conduct a mass balance for solids,
water, oil, PCBs and PAHs; and to examine process effects on metals, oil and grease, and several other
parameters.
The B.E.S.T.® Solvent Extraction Process was tested using sediment samples obtained from the
Buffalo River, Saginaw River, and Grand Calumet River. The concentration of the contaminants of concern
in the sediment were 0.3 to 22 mg/kg PCBs and 3 to 220 mg/kg PAHs. The PCB and PAH concentrations
of 0.2 to 0.4 and 0.4 to 37 mg/kg, respectively, were found in the treated solids. This corresponds to PCB
and PAH removals of >95 to 99 percent and 65 to 96 percent, respectively. Metals analyses were
performed on the treated solids and untreated sediments. The data demonstrate that the treatment
process, as expected, had little affect on metal removal from the sediments. The feed sediments and
treated solids were analyzed for percent moisture, oil and grease, total organic carbon (TOC), total volatile
solids, and pH. Reductions in oil and grease concentrations (ranging from 80 to 99 percent) correspond
to sediment PCB and PAH removal. A mass balance was also carried out as part of this study for the
different constituents: solids, oil, water, PCBs, and PAHs.
-------
TABLE OF CONTENTS
Section
Disclaimer
Acknowledgements
Abstract
Figures
Tables
1.0
2.0
Page
3.0
Executive Summary
Introduction
2.1 Background
2.2 Sediment Descriptions
2.2.1 Site Names and Locations for Each Sediment
2.2.2 Sediment Acquisition and Homogenization . . .
2.3 Sediment Characterization
2.4 Technology Description . .
Treatability Study Approach
3.1 Test Objectives and Rationale
3.2 Experimental Design and Procedures
3.2.1 Phase I
3.2.2 Phase II
3.3 Sampling and Analysis
3.3.1 Sampling
3.3.2 Analysis
VI
vii
1
3
4
4
4
9
9
10
12
12
14
14
15
18
19
19
IV
-------
TABLE OF CONTENTS (Continued)
Section
4.0 Results and Discussion 21
4.1 Summary of Phase I Results 21
4.2 Summary of Phase II Results 22
4.2.1 Sediments/Treated Solids 23
4.2.2 Oil 26
4.2.3 Water 28
4.2.4 Mass Balance 28
4.3 Summary of Vendor Results 33
4.4 Quality Assurance/Quality Control 34
Appendix A — B.E.S.T.® Bench-Scale Treatability Test Report 35
Appendix B — B.E.S.T.® Bench-Scale Treatability Test Plan 64
Appendix C — Quality Assurance Project Plan 72
Appendix D — B.E.S.T.® Treatability Study Analytical Matrix and Methods 117
Appendix E — Battelle Data 123
Appendix F — Quality Assurance/Quality Control 143
-------
FIGURES
Number
1 ARCS Priority Areas of Concern 5
2 Buffalo River Sample Location 6
3 Saginaw River Sample Location 7
4 Grand Calumet River Sample Location 8
5 Flow Diagram of the B.E.S.T.® Process 11
VI
-------
TABLES
Number page
1 Battelle and RCC Data - PCB Summary 1
2 Battelle Data - Summary of Total PAHs 1
3 Mass Balance Summary 2
4 Battelle Data - Data Characterization of Feed Sediments 9
5 Parameters for Analysis of ARCS Program Technologies 14
6 Sodium Hydroxide Addition 16
7 SAIC's Analysis Schedule for the Phase II Solvent Extraction of Buffalo River,
Grand Calumet River, and Saginaw River Sediments 20
8 RCC Analyses 21
9 pH Adjustments 22
10 Battelle Data - Total PCBs 23
11 Battelle Data - Feed and Treated Solid PAH Concentrations 24
12 Battelle Data - Metals Concentration in the Feed and Treated Solids 25
13 Battelle Data - Removal Efficiencies for Other Parameters 26
14 Battelle Data - PAH Concentrations in the Treated Solids, Water and Oil 27
15 Battelle Data - PCB Concentrations in the Treated Solids, Water and Oil 28
16 Battelle Data - Solid Mass Balance 29
17 Battelle Data - Water Mass Balance 30
18 Battelle Data - Oil Mass Balance 30
19 Battelle Data - PCB Mass Balance 31
20 Battelle Data - PAH Mass Balance 32
21 RCC Data - PCB Summary 33
22 RCC Data - Mass Balance Summary 33
vii
-------
1.0 EXECUTIVE SUMMARY
The B.E.S.T.® Solvent Extraction Process was tested using sediments obtained from the Buffalo
River, Saginaw River, and Grand Calumet River. The contaminants of concern in the sediments for these
tests were PCBs and PAHs. Samples of the feed material and the treated solids produced using the
B.E.S.T.® Solvent Extraction Process were analyzed by Battelle Marine Sciences Laboratory and RCC for
residual PCB contamination. The data from these analyses are presented in Table 1.
Table 1. Battelle and RCC Data - PCB Summary
Sample
Feed
(mg/kg, dry basis)
Battelle RCC
Treated Solids
(mg/kg, dry basis)
Battelle RCC
Removal Efficiency
Battelle
RCC
Buffalo River
Saginaw River
Grand Calumet River
0.32
21.9
15.0
0.60
21
22
<0.3
0.24
0.44
<0.03
0.18
0.23
>6
99
97
>95
99
99
As these data obtained by RCC and Battelle demonstrate, PCB removal efficiencies for the Grand
Calumet River and Saginaw River sediments complement each other. However, the removal efficiencies
determined by RCC and Battelle for the Buffalo River sediment are substantially different. This can be
attributed to the fact that the contaminant concentration in the raw Buffalo River sediment was close to the
analytical detection limit achievable by Battelle. The potential errors associated with these data undermine
the relevance of the removal efficiency obtained by Battelle for the Buffalo River sediment.
Feed material and treated solids were also analyzed for residual PAH concentrations. Table 2
outlines the analytical results obtained by Battelle.
Table 2. Battelle Data - Summary of Total PAHs
Sample
Buffalo River
Saginaw River
Grand Calumet River
Feed
(mg/kg dry basis)
9.90
2.70
230
Treated Solids Removal Efficiency
(mg/kg dry basis)
0.37
0.95
37.1
%
96
65
84
-------
During the RCC analyses of the Buffalo River and Saginaw River sediments, residual PAH
concentrations of <0.2 mg/kg per compound were found in the treated solids. Treated solids with PAH
concentrations ranging from <1 to <3 mg/kg per compound were obtained for the Grand Calumet River
sediment. Because RCC was unable to report lower detection limits, comparisons between RCC and
Battelle data are not conclusive.
Metal analyses were performed on the treated solids and untreated sediments (see Table 11). The
Battelle data demonstrate that the treatment process, as expected, had little affect on metal removal from
the sediments. The RCC data cannot be compared to the Battelle data because these data were obtained
using different analytical methods than those employed by Battelle. Because of the ashing of the sediment
feed sample (potentially causing metals to be lost by volatilization) and because different methods were
used to analyze the feed sediments and product solids, a reliable comparison of the RCC and Battelle data
is not possible.
The feed sediments and treated solids were analyzed for percent moisture, oil and grease, Total
Organic Carbon (TOC), volatile solids, and pH (see Table 12). As the data in Table 12 shows, the
reductions in oil and grease concentrations (ranging from 80 to 99 percent) correspond to sediment PCB
and PAH removal.
A mass balance was also carried out as part of this study. Table 3 summarizes the results obtained
for the different constituents: solids, oil, water, PCBs, and PAHs.
Table 3. Mass Balance Summary (percent recovered)
Solids
Sample
Buffalo River
Saginaw River
Grand Calumet
River
Battelle
98
99
92
RCC
97
98
86
Oil
Battelle
163
192
69
Water
RCC
112
137
97
Battelle
68
74
78
RCC
70
82
75
PCBs
Battelle
129
80
94
RCC
70
280
64
PAHs
Battelle
90
240
11
Assuming that a full-scale application of this technology occurred and a volume of 500,000 tons of
sediment required treatment, RCC estimated that it would cost approximately $150 to $250/ton to treat the
material. The cost is dependent on the quantity of material processed, the cleanup target and the settling
-------
characteristics of the waste. The waste would be treated at a rate of 200 to 300 tons per day using the
B.E.S.T.® Model 615 Unit operated on a 24-hour-per-day basis. This estimate includes mobiliza-
tion/demobilization costs but does not account for costs associated with site excavation, civil work,
applicable taxes, pre-screening needs, and overall site management and disposition of the product oil.
Small vials of the residuals from the treatability test were retained and given to the EPA Technical
Project Manager for the GLNPO for "show" purposes. All quantities of the test products (water, solids, and
oil residuals) from each treatability test were sent to the analytical laboratory, Battelle Marine Sciences
Laboratory, for analysis. Due to the quantities generated from the tests, none were retained and shipped
to EPA for possible further treatability studies.
2.0 INTRODUCTION
The Great Lakes National Program Office (GLNPO) leads efforts to carry out the provisions of
Section 118 of the Clean Water Act (CWA) and to fulfill U.S. obligations under the Great Lakes Water
Quality Agreement (GLWQA) with Canada. Under Section 118(c)(3) of the CWA, GLNPO was responsible
for undertaking a 5-year study and demonstration program for the remediation of contaminated sediments.
Five areas were specified for priority consideration in locating and conducting demonstration projects:
Saginaw River and Bay, Michigan; Sheboygan River, Wisconsin; Grand Calumet River/Indiana Harbor
Canal, Indiana; Ashtabula River, Ohio; and Buffalo River, New York. In response, GLNPO initiated the
Assessment and Remediation of Contaminated Sediments (ARCS) Program.
In order to develop a knowledge base from which informed decisions may be made, bench- and
pilot-scale demonstrations of sediment treatment technologies were conducted as part of the ARCS
Program. Information from remedial activities supervised by the U.S Army Corps of Engineers and the
Superfund program were also utilized. The Engineering/Technology (ET) Work Group was charged with
overseeing the development and application of the bench- and pilot-scale tests.
Science Applications International Corporation (SAIC) was contracted to provide technical support
to the ET Work Group. The effort consisted of conducting bench-scale treatability studies on designated
sediments to evaluate the removal of specific organic contaminants. The bench-scale studies of the
B.E.S.T.® Solvent Extraction Process, which are the subject of this report, took place at Resources
Conservation Company (RCC) in Bellevue, WA on August 5 to 9, 1991. The specific objectives for this
effort were: to determine process extraction efficiencies for polychlorinated biphenyls (PCBs) and
-------
polynuclear aromatic hydrocarbons (PAHs); to conduct a mass balance for solids, water, oil, PCBs and
PAHs; and to examine process effects on metals, oil and grease, and several other parameters.
2.1 Background
SAIC and its subcontractors have conducted seven bench-scale tests for the ARCS Program on
four different sediments using four treatment technologies: B.E.S.T.® Solvent Extraction Process (RCC),
Low Temperature Thermal Desorption Process (ReTeC), Wet Air Oxidation (Zimpro Passavant), and
Anaerobic Thermal Process Technology (SoilTech). This report summarizes the approach used and results
obtained during treatability testing of the B.E.S.T.® Solvent Extraction Process. The sediments used during
this technology evaluation were obtained from the Buffalo River, Grand Calumet River/Indiana Harbor
Canal, and Saginaw River.
The primary objective of this portion of the study was to determine the feasibility and cost-
effectiveness of the B.E.S.T.® Solvent Extraction Process for treating and removing PCBs and PAHs from
the three sediments. Based upon previous tests performed by RCC, it is their experience that the data
obtained from the bench tests simulate full-scale operation. Thus, data generated by these tests may be
used to estimate treatment costs for full-scale operation and to evaluate process feasibility. The ability to
evaluate process feasibility from these tests was also reported by the U.S. Environmental Protection
Agency (EPA) in their report entitled, " Evaluation of the B.E.S.T.® Solvent Extraction Sludge Treatment
Technology - Twenty-Four Hour Test."
2.2 Sediment Descriptions
The sediments used for these tests are typical of sediments in the Great Lakes and their tributaries.
They are representative of locations where future field demonstration projects may be conducted. For the
purpose of these tests, the primary contaminants in these sediments were PCBs and PAHs.
2.2.1 Site Names and Locations for Each Sediment
GLNPO collected sediments for study from the following areas around the Great Lakes: Saginaw
River, Michigan; Sheboygan River, Wisconsin; Grand Calumet River/Indiana Harbor Canal, Indiana;
Ashtabula River, Ohio; and Buffalo River, New York. SAIC was contracted to treat four of the sediments
(from the Grand Calumet River/Indiana Harbor Canal, Buffalo River, Ashtabula River, and Saginaw River)
using four different technologies. Samples from Grand Calumet River/Indiana Harbor Canal, Buffalo River,
and Saginaw River were treated using the B.E.S.T.® Extraction Process. A map is provided in Figure 1
which shows the ARCS Priority Areas of Concern. Specifics of the sample location for the Buffalo River,
Saginaw River and Grand Calumet River are shown in Figures 2, 3, and 4, respectively.
-------
en
ARCS* PRIORITY
AREAS OF CONCERN
ARCS AREAS OF CONCERN
1. SHEBOYGAN RIVER
2. GRAND CALUMET RIVER / INDIANA HARBOR
3. SAGINAW RIVER/BAY
4. ASHTABULA RIVER
5. BUFFALO RIVER
' Assessment and Remediation of Contaminated Sediments
0 80 100 150 200
KLOMETER9
US ENVIRONMENTAL PROTECTION AGENCY
(MEAT LAKES NATONAL PflOQRAU OFFICE
Rgure 1. ARCS Priority Areas of Concern
-------
Ida
Em
Buffalo River
O)
Sediment sample point
Figure 2. Buffalo River Sample Location
-------
Saginaw River and Bay
Sediment sample point
"»»*
'•ta»
Figure 3. Saginaw River Sample Location
7
-------
Sediment sample point
GRAND CALUMET RIVER
M IJ 2.0
I 1
Figure 4. Grand Calumet River Sample Location
-------
2.2.2 Sediment Acquisition and Homogenization
Prior to conducting the bench-scale treatability study using the B.E.S.T.® technology, the GLNPO
samples were homogenized and stored under refrigeration by the U.S. EPA Environmental Research
Laboratory in Duluth, Minnesota.
The homogenized sediments were sent to SAIC by the Duluth laboratory. Eighty ounces of each
sediment sample were then transferred by SAIC to RCC. RCC used these samples to perform a series
of standard tests to determine if the waste samples were compatible with their process and to determine
optimum testing conditions and procedures for the treatability study (Phase I). The sediments used during
the treatability studies also originated from this stock and were forwarded to RCC by SAIC.
2.3 Sediment Characterization
SAIC was responsible for the physical and chemical characterization of the raw sediment samples
used during the tests. Table 4 provides characterization data of the sediments. In order to limit inter-
laboratory variation, the different sediments and their residuals were analyzed by Battelle Marine Sciences
Laboratory in Sequim, Washington. Raw sediment analyses conducted by RCC are also included in this
report and can be found in Appendix A. The raw sediment samples analyzed by RCC and Battelle were
collected simultaneously.
Table 4. Battelle Data - Characterization of Feed Sediments
(mg/kg, dry basis, unless specified)
Total PCBs
Total PAHs
Moisture, % (as received)
Oil & Grease
TOC, % weight
Total Volatile Solid, %
pH, S.U. (as received)
Buffalo River
0.32
9.90
42.0
2420
1.98
4.03
7.29
Saginaw River
21.9
2.70
24.0
1350
0.83
2.09
7.30
Grand Calumet River
15.0
230
57.0
32200
17.03
14.2
7.35
-------
2.4 Technology Description
The B.E.S.T.® process is a patented solvent extraction technology developed by RCC. This
process employs triethylamine (a solvent) to extract contaminants from wastes. Triethylamine is an
aliphatic amine produced by reacting ethyl alcohol and ammonia. This solvent is distinguished from other
solvents because it is inversely miscible. At temperatures below 65° F, triethylamine is completely miscible
with water, while at temperatures above 65° F, triethylamine and water are only partially miscible. Since
oil and water are similarly soluble in cold triethylamine, it can be utilized to treat wastes containing both
contaminated oil and water.
The B.E.S.T.® process produces a single-phase extraction solution. If water and oil are present
in the feed material, a homogeneous mixture of triethylamine, water, and oil is produced. Any organic
contaminants contaminating the feed material, such as PCBs, PAHs, and volatile organic compounds
(VOCs) are trapped within the water and oil portion of the extraction solution. Since triethylamine achieves
intimate contact with the waste at nearly ambient temperatures and pressures, emulsions (oil containing
the organic contaminants) are not expected to occlude the solute. Thus the extraction efficiency of the
B.E.S.T.® process will not be compromised by feed mixtures with high water content.
RCC utilizes triethylamine because it exhibits several characteristics that enhance its use in a
solvent extraction system. These characteristics, as reported by RCC, include: 1) a high vapor pressure
(therefore, the triethylamine can be recovered from the extract via simple steam stripping); 2) formation of
a low-boiling azeotrope with water (therefore, the solvent can be recovered from the treated solids by heat
with a low energy input); 3) triethylamine has an alkaline pH=10 (therefore, some heavy metals are
converted to the hydroxide form, which precipitate and exit the process with the treated solids); and 4)
triethylamine is only moderately toxic and readily biodegrades (data available in EPA document EPA-600/2-
82-001 a show that a level of 200 ppm triethylamine in water was degraded completely in 11 hours by
Aerobacter, a common soil bacteria).
A block diagram for the B.E.S.T.® Process is presented in Figure 5. Since triethylamine is soluble
in water at temperatures below 65° F, the first extraction of the contaminated material is conducted at
temperatures near 40° F. Therefore, the first extract solution will contain most of the water initially present
in the feed material. If the first stage extract contains sufficient water to allow a phase separation of the
triethylamine and water, the extract is heated to a temperature above the miscibility limit (130° F). At this
temperature, the extract separates into two distinct phases; a triethylamine/oil phase and a water phase.
The two phases are separated by gravity and decanted.
10
-------
Solvent Recovery
Recycled
Solvent
Was
1
Extraction
40°F(5°C)
B.E.S.T.
Solvent Recycle
Solvent / Oil
Evaporator
Separation
130°F(55°C)
I
Solvent Recycle
Water I Water
Stripper
Solvent / Water
Solvent Recycle
Solids
Solids
Dryer
Figure 5. Flow Diagram of the B.E.S.T.® Process (Source: RCC, Inc.)
-------
At 130° F the solubility of oil (organic contaminants) in triethylamine increases. Since this enhances
the removal of oil from the contaminated solids, subsequent extractions are conducted at temperatures
above 130° F. Because these extracts contain mostly oil and very little water, they are combined with the
decanted triethylamine/oil phase from the first extraction stage. In the full-scale unit, the solvent is
recovered from the two phases by way of steam stripping and is recycled directly to the extraction vessels
for the solvent recovery portion of the process. Residual triethylamine in the water and oil products is
usually low.
Triethylamine is removed from the treated solids by indirect heating with steam. A small amount
of steam may be added directly to the dryer vessel to provide the water required to form the low boiling
azeotrope. Typically the residual triethylamine remaining with the treated solids biodegrades readily. Thus,
unless restricted by a contaminant not treated by the process, the dry treated solids may be used on site
as backfill.
The B.E.S.T.® Process operates near ambient pressure and temperature and at a mildly alkaline
pH. Liquid temperatures vary from about 40 to 170° F and high pressures are not required. A low-
pressure nitrogen blanket creates a small positive pressure in tanks and vessels. Since the process
operates in a closed loop with one small vent for removal of non-condensing gases, air emissions are
minimal. RCC typically uses a water scrubber and activated carbon on this vent to minimize triethylamine
releases.
3.0 TREATABILITY STUDY APPROACH
3.1 Test Objectives and Rationale
SAIC was contracted by the ARCS Program to test four technologies for removing organic
contaminants (PCBs and PAHs) from sediments typical of locations around the Great Lakes. This
treatability study was performed to determine the feasibility and cost-effectiveness of the B.E.S.T.® Solvent
Extraction Process for treating and removing PCBs and PAHs from three different sediments. In order to
accomplish this, this bench-scale test had the following objectives:
• To record observations and data to predict full-scale performance of the B.E.S.T.® process
• To take samples during the extraction tests and conduct analyses sufficient to allow for
calculation of mass balances for oil, water, solids and other compounds of interest
• To calculate the extraction efficiency of target compounds
• To obtain treated solids (300 g dry basis), water, and oil for independent analysis
12
-------
Based upon previous tests performed by RCC, it is their experience that the data obtained from
the bench test simulate full-scale operation. Ultimately, this data may be used to estimate both the
feasibility and treatment costs associated with a full-scale application of the technology. The ability to
evaluate process feasibility from these tests was also reported by EPA in their report entitled," Evaluation
of the B.E.S.T.® Solvent Extraction Sludge Treatment Technology - Twenty-Four Hour Test."
A two-phase approach was used during this study. During Phase I, SAIC sent a sample of the
untreated sediments to RCC. These samples underwent a series of initial tests in order to determine the
optimum conditions to be used during the actual treatability tests (Phase II). During Phase II, wet sediment
from each of the three locations (Buffalo River, Grand Calumet River, and SJaginaw River) was sent to
RCC. Samples of raw (untreated) sediments and the various end products generated during the treatability
tests (Phase II) were obtained and analyzed by both SAIC and RCC. The data generated by SAIC were
primarily used to determine treatment extraction efficiencies and mass balances. Vendor- or subcontractor-
generated data are reported and commented on when available.
This study is only one part of a much larger program and is not intended to evaluate the treatment
of the sediments completely. In order to ensure that the data obtained from this study can be objectively
compared with data generated from the other studies performed in support of the ARCS Program, Battelle
Marine Sciences Laboratory was subcontracted to perform all analyses for the different treatability studies
performed by SAIC (seven treatability studies utilizing four technologies on four sediments). The same set
of analyses listed in Table 5 was applied during the characterization of each raw sediment and end
products from the different treatability tests. I n addition, representatives from SAIC observed how all Phase
II treatability tests were conducted.
13
-------
Table 5. Parameters for Analysis of ARCS Technologies
Parameters
TOC/TIC Arsenic
Total Solids Barium
Volatile Solids Cadmium
Oil & Grease Chromium
Total Cyanide Copper
Total Phosphorus Iron (total)
PCBs (total & Aroclors) Lead
PAHs (16) Manganese
pH Mercury
BOD Nickel
Total Suspended Solids Selenium
Conductivity Silver
Zinc
3.2 Experimental Design and Procedures
3.2.1 Phase I
Phase I was designed to allow RCC to explore a range of variables in order to set test parameters
which would optimize the performance of the B.E.S.T.® technology for the bench-scale tests (Phase II).
In order to accomplish this, samples of the different sediments were sent to RCC by SAIC prior to bench-
scale testing. The amount of material sent was approximately 1 kg of each sediment, as specified by RCC.
RCC analyzed the three raw sediment samples to determine whether the sediments were
compatible with triethylamine. During these compatibility tests, the feed samples were individually mixed
with cold triethylamine and then monitored to observe the amount of heat generated as well as any visual
signs that adverse reactions (such as an extremely exothermic chemical reaction) were occurring. All
samples satisfied the compatibility criteria.
Since triethylamine can be ionized at a low pH into unrecoverable triethylammonium salts, the pH
of the sample needed to be adjusted to approximately 11 during Phase II in order to enable RCC to
efficiently recover the triethylamine from the separated phase fraction products. To determine the amount
of caustic needed to increase the pH of the raw feed to the operating pH of the B.E.S.T.® process (pH =
11), RCC slurried 5 g portions of the feed samples with deionized water. Incremental portions of the
caustic (50% sodium hydroxide) were added to bring the pH to 11. The amount of caustic required to
adjust the pH was recorded.
14
-------
After observing these simple mixing tests previously described, RCC applied its past experience
in selecting operating parameters for the technology to determine the number of extraction stages, solvent
to feed ratios, and the proposed temperature and mixing time for each stage for the bench-scale tests.
3.2.2 Phase II
Phase II of the treatability program is referred to as the "B.E.S.T.® Bench-scale Treatability Test
Workup" in RCC's B.E.S.T.® Bench-Scale Treatability Test Plan. This section outlines five major
operations including: 1) Pre-treatment and first wash; 2) Second wash; 3) Third wash and solids drying;
4) Decantation; and 5) Distillation. These procedures and associated equipment used are described in
Appendix B.
3.2.2.1 Procedures
Test Sample Preparation-
The contaminated samples from the Buffalo River, Saginaw River, and Grand Calumet River sites
were gray-colored sediments with very little debris present. Each of the three samples contained free-
standing water. As it was very difficult to homogenize the samples with the free-standing water present,
this water was decanted prior to conducting the bench-scale tests and proportionally recombined with the
portion used for the bench testing. This was done by weighing the entire decanted sediment and the
portion used for the test. The percentage of the portion used to the whole determined the percentage of
the decanted water to recombine with the test sample.
Bench-scale testing requires material greater than 1/4 inch be removed. Full-scale processing
requires that the feeds be screened to remove only material greater than 1 inch in diameter. There was
no material greater than 1/4 inch in any of the three samples received. Therefore, the samples were not
screened.
The following are summaries of the five major operations. With slight variations, the same
procedure was used to treat each of the sediments.
Pre-treatment and First Wash-
During the treatability tests of the different sediments, each sample was placed in a 4-L resin kettle
immersed in a temperature-controlled water bath set at 33° F. Each sample's pH was adjusted using
sodium hydroxide (NaOH) and 2.7 L of chilled triethylamine (2 percent H2O). The 2 percent water was
15
-------
added to offset the amount of water remaining in the triethylamine after the extraction step. The amount
of sodium hydroxide needed to adjust the pH of each sample to 11 was determined during Phase I and
is listed in Table 6, as is the amount of sediment treated during each trial.
Table 6. Sodium Hydroxide Addition
Sediment
Buffalo River
Saginaw River
Grand Calumet River
Sample Weight
(g)
700
700
1400
Caustic Added per
kg of Sediment
(ml 50% NaOH)
6.0
6.0
6.0
While immersed in the cooling bath, the sample was mixed with the NaOH and the chilled
triethylamine using a air-driven prop mixer. Mixing occurred for approximately 20 minutes for the Saginaw
River and Buffalo River sediments with the pneumatic mixer in the chiller bath. Because of the high water
content of the Grand Calumet River sediment, a larger sample size was needed to yield the quantity of
treated solids required. For this sediment, the first extraction was conducted in two steps, using 700 g of
feed for each step. Each sample was mixed for 10 minutes.
At the end of the first mixing stage, the sample was allowed to separate by gravity. The
particulates were then separated from the liquid by centrifuging the extract at 2,100 rpm for 10 minutes.
The solvent/oil/water/centrate were set aside for later decantation. The solids from the centrifuge were
placed back into the resin kettle for additional wash stages.
Second Wash--
Solids recovered from the first extraction were mixed with 2.7 L of fresh triethylamine. Part of the
triethylamine was used to transfer solids from the centrifuge bottles into the mixing container. For the
second extraction, the samples were heated to 127 to 140° F. The mixture was kept heated while mixing
was in progress. Mixing was conducted with a pneumatic mixer for approximately 20 minutes. This sample
settled very quickly. The solvent/oil was poured off and held for later combination with the solvent/oil
portion from the decantation procedure.
16
-------
Third Wash and Solids Drying--
For the third wash, the same procedure that was used in the second wash was repeated. Mixing
for this wash was for 30 minutes.
The treated solids resulting from the third wash were then dried at 220° F in a forced-draft oven.
Occasional mixing in the oven to facilitate triethylamine volatilization was conducted. This mixing was
accomplished by turning the sample in the oven with a clean spatula. After the initial drying, a portion of
de-ionized water was added to wet the solids thoroughly. The solids were then redried in order to reduce
residual triethylamine concentrations further. To ensure that the triethylamine residual in the dried solids
was low, the solids were treated with caustic soda (applied with the de-ionized water) when the pH of these
solids was less than 10. Sufficient caustic soda was added to raise the pH to approximately 10.5. The
required amount of caustic soda added was determined on a small portion of the solids.
Decantation-
The first stage extracts trap nearly all of the water present in the feed sample. Because of this,
only the water from the first stage extracts is recovered. There are two methods to decant water. The
decantation method is chosen depending on the water content of the feed. The following methods were
employed to recover the water from the test series samples.
Method 1: During the decantation of the Buffalo River and Grand Calumet River extracts, a 4-L
separatory funnel immersed in a temperature-controlled water tank was employed. The tank was kept at
140° F by circulating water between the tank and a temperature-controlled water bath set at 140° F.
Supernatant/centrate from the first wash (chilled to this point at 40° F) was heated to above 130° F
with continuous mixing on a hotplate and poured into the separatory funnel. Since above 130° F water is
no longer miscible with the triethylamine, the water settled to the bottom of the mixture. Forty minutes
quiescent residence time in the separatory funnel for the Buffalo River sample and 15 and 90 minutes for
the Grand Calumet River sample were required. The length of time required depended on how long it took
for near-separation of the layers. A sample from the rag layer, which is an emulsion where any solids
present tend to collect and create a region where the triethylamine/oil/water separation is not distinct, was
taken by RCC for later possible analyses. Generally, the smaller the rag layer is in comparison to the
triethylamine/oil and water phases, the better the separation. The rag layers for all three sediments were
relatively small and within expected volumes. SAIC did not collect samples from this rag layer.
17
-------
Method?. Because of its low water content (<25%), the water present in the Saginaw River extract
was separated from the oil by evaporation instead of decantation. When the first extract was evaporated,
the water in the triethylamine/oil/water mixture formed an azeotrope with the distilled triethylamine, leaving
the oil behind. After the triethylamine/water isotope had condensed, the water was decanted by heating
the triethylamine/water mixture above 130° F and pouring the mixture into the 4-L separately funnel. Since
no temperature control system was required, separation occurred immediately. This method produces a
much purer water stream and is preferable for low-water-content feeds where the extra energy cost to
evaporate the water is small.
Distillation-
Water Layer: After recording the water's pH, which should have been >10, and its volume, the
water was stripped by steam at 110° C in a Buchi Rotovapor apparatus to ensure that the triethylamine left
the water. Periodically, the water volume and water pH were checked. When the water pH was <10, the
pH was adjusted to >12 and stripping continued until the pH of the water remained above 10. The pH was
periodically checked, and if found <10, the previous steps were repeated until the water pH remained above
10. At this point, distilling continued for 15 minutes longer before being terminated. The elevated pH is
necessary to ensure that the majority of the triethylamine remains in the volatile molecular form.
Oil/Triethylamine Layer. The bulk of triethylamine was removed from this layer by boiling the
triethylamine/oil mixture at 110° C in the Rotovapor (no steam necessary). The triethylamine condensed
as it evaporated and was collected separately. Normally the oil remaining in the flask would then be steam-
stripped of any residual triethylamine by adding a known quantity of water (typically 5 ml) to the hot oil in
the boiling flask of the rotovapor and then measuring the volume of distillate recovered. When all
triethylamine was removed, the amount of the distillate recovered would equal the amount of water added.
However, because of the low oil content of the feed in this test, the amount of oil recovered was too small
to enable the oil to be effectively stripped. Triethylamine was allowed to remain in the Rotovapor, thereby
allowing the oil from the sample to remain in solution. The homogeneous oil was able to be poured out
of the Rotovapor flask. The final product oil/triethylamine weight was recorded. This procedure is an
artifact of this test due to the small quantities of sediment.
3.3 Sampling and Analysis
The Quality Assurance Project Plan is presented in Appendix C.
18
-------
3.3.1 Sampling
At the beginning of the Phase II treatability test, SAIC personnel observing Phase II packed and
shipped a sample of the untreated Buffalo River, Grand Calumet River and Saginaw River sediments to
SAIC's subcontract laboratory, Battelle, in accordance with written detailed instructions supplied to the SAIC
on-site representative. Each sample contained free-standing water which was decanted prior to conducting
feed analyses and was proportionally recombined prior to any analysis and bench testing. This was done
because it is very difficult to homogenize material when free-standing water is present. These samples
were obtained from separate unopened containers of the sediments sent for Phase II.
Although the samples would normally be screened to remove any material greater than 1/4 inch,
Phase I results indicated that no material of that size or greater was present in the three samples. Thus
the samples were not screened.
After the extractions were complete, samples of the final water, oil, and solids residuals were
distributed to SAIC and RCC. As specified in the Quality Assurance Project Plan (QAPP) a minimum of
300 g (dry basis) of solid material was required in order for Battelle to be able to complete the necessary
analyses of that material. Since the quantity of oil and water was dependent on the sediment and the
technology employed, it was not possible to obtain enough water and oil to perform the full scope of
analyses specified in Table 7.
3.3.2 Analysis
Two separate sets of analyses were conducted by SAIC's subcontracted laboratory, Battelle, and
RCC on the three raw sediments and the process products during Phase II. Battelle's data was used for
the results presented in this report. RCC's data is discussed and commented upon, where possible, to
facilitate interpretation of the results of the treatability test.
3.3.2.1 Battelle Analyses
Following the Phase II treatability test, Battelle conducted analyses on the three raw sediments and
the end products. The number of analyses conducted on these sediments and their residuals are listed
in Table 6. Descriptions of the analytical methods employed can be found in the QA Section of this report.
Since the actual quantities of oil and water produced by the technology during the bench-scale
treatability tests were not sufficient to perform all the analyses in Table 6, only PCB and PAH analyses
were performed on the water and oil.
19
-------
Table 7. SAIC's Analysis Schedule for the Phase II Solvent Extraction Evaluation of
Buffalo River, Grand Calumet River, and Saglnaw River Sediments
Parameters
Total Sol ids
(Moisture)
Volatile Solids
O&G
Metals
PCBs
PAHs
TOC
Total Cyanide
Total Phosphorus
PH
BOD
Total Suspended
Solids
Conductivity
QC Sample ()
and
Method Blank
(1)
YES
(1)
YES
(1)
YES
(0)
YES
(1)
YES**
(1)
YES**
(0)
YES
(0)
YES
(0)
YES
(0)
YES
NA
NA
NA
Untreated
Sediment
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
MS
(1)
S
(1)
S
(1)
S
NA
NA
NA
'"
Tripli-
cate
(2)
S
(2)
S
(2)
S
(2)
S
(2)
S
(2)
S
NA
NA
•NA
NA
Treated
Solids
(3)
B,G,S
(3)
B.G.S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
(3)
B,G,S
-
MS
(1)
S
(1)
S
(1)
S
NA
NA
NA
MSD
(1)
S
(1)
S
NA
Tripli-
cate
(2)
S
(2)
S
(2)
S
(2)
S
NA
NA
NA
Water
•" : • ^
NA*
NA
NA
(3)
B,G,S
(3)
B,G,S
NA
NA
NA
NA
NA
NA
NA
MS
f
'
NA
NA
NA
NA
NA
NA
MSD
"•
•>
NA
NA
%
Tripli-
cate
-
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Oil
(3)
B,G,S
(3)
B,G,S
MS
(1)
S
(1)
S
Tripli-
cate
-
(2)
S
(2)
S
ro
o
* Not Analyzed
** A laboratory pure water spike is required for recovery determination
(3) = Number of Analyses
B = Buffalo River
G = Grand Calumet River
S = Saginaw Bay
MS = Matrix Spike
MSD = Matrix Spike Duplicate
-------
3.3.2.2 RCC Analyses
RCC analyzed the sediment samples for moisture content, oil and grease, ash content, metals,
PAHs, PCBs, and particulate solids content. RCC also conducted their own analyses on the products for
Phase II. Table 8 shows the analyses performed by RCC. Details on the analytical methods used by RCC
are presented in Appendix D. The following section is a summary of the RCC analyses conducted.
Table 8. RCC Analyses
Matrix Sample
Raw Sediment
Treated Solids
Residual Oil
Residual Water
PAHs
yes
yes
no
no
PCBs
yes
yes
no
yes
Total
Metals
yes
yes
no
yes
Oil&
Grease
yes
yes
no
no
Solvent
no
yes
yes
no
Water
yes
no
no
no
Particulate
Content
yes
yes
no
no
TCLP
no
metals
no
no
PH
yes
yes
no
no
TCLP = Toxicity Characteristic Leaching Procedure
4.0 RESULTS AND DISCUSSION
4.1 Summary of Phase I Results
RCC analyzed the three raw sediment samples to determine whether the sediments were
compatible with triethylamine. Since there were no visible signs indicating that adverse reactions were
occurring and the heat of the solution did not exceed normal expectations, the B.E.S.T.® bench-scale
treatability tests proceeded.
The amount of caustic (NaOH) needed to increase the pH of the raw sediments to the operating
pH of the B.E.S.T.® process (pH = 11) was determined. This information, as well as the original pH of the
sample, is summarized in Table 9.
21
-------
Table 9. pH Adjustments
Sediment
Buffalo River
Saginaw River
Grand Calumet River
Initial
PH
7.6
8.1
7.5
Caustic Added per
kg of Sediment
(ml 50% NaOH)
6.0
6.0
6.0
A sieve analysis of the raw sediment was conducted to determine the screening and size reduction
requirements. Since there was no material greater than 1/4-inch in diameter, sample screening was
determined unnecessary.
4.2 Summary of Phase II Results
As stated previously, the concentrations of PAHs, PCBs, metals, total solids, volatile solids, and
oil and grease present in the untreated sediments and treated solids are the critical measurements
associated with this study. Oil and water residuals were analyzed to determine the fate of the contaminants
of concern from the process. Since insufficient water and oil was produced from the quantity of untreated
sediments used with the B.E.S.T.® process to perform all the analyses listed in Table 7, only PCB and PAH
analyses were performed on the water and oil residuals. The following sections briefly address the
analytical results obtained for contaminant concentrations present in the raw sediments and the process
residuals (i.e., treated solids, water, and oil), as well as applicable extraction efficiencies. The discussion
of Phase II results concludes with an analysis of the mass balance of the media and contaminants. The
analytical data received from Battelle can be found in Appendix E.
Individual PAH compounds, PCB Aroclors, and metals were quantitated during sample analyses.
In order to determine overall removal efficiencies for each class, it was necessary to sum these individual
results. In instances where all reported results were less than the analytical detection limits, total
concentrations are reported as less than the sum of the individual detection limits. Where one or more
individual components are above detection limits, total concentrations are reported as the sum of these
detected values.
22
-------
4.2.1 Sediments/Treated Solids
4.2.1.1 PCBs
Samples of the feed material and the treated solids produced using the B.E.S.T.® Solvent
Extraction Process were analyzed for PCB contamination. The data from these analyses are presented
in Table 10.
Table 10. Battelle Data - Total PCBs
Sample
Buffalo River1
Saginaw River2
Grand Calumet River3
Feed
(mg/kg, dry basis)
0.32
21.9
15.0
Treated Solids
(mg/kg, dry basis)
<0.3
0.24
0.44
Removal Efficiency
%
>6
99
97
1 Identified primarily as Aroclor 1248
2 Identified primarily as Aroclor 1242
3 Identified primarily as Aroclor 1248
As demonstrated by these data, PCB concentrations of 0.24 mg/kg and 0.44 mg/kg were found in
the treated solids generated from the Saginaw River and Grand Calumet River sediments, respectively.
This corresponds to PCB removal efficiencies of 99 and 97 percent. At first glance, the Buffalo River
solids achieved a much lower removal efficiency (i.e., >6 percent). This is attributed to the low PCB
concentrations initially present in the untreated Buffalo River sediment and the high analytical detection
limits achieved. As the concentration of a contaminant approaches analytical detection limits, the error
associated with the analytical readings obtained increases. Thus, the relevance of the removal efficiency
achieved for the Buffalo River sediment is undermined by: 1) error associated with the measurement of
the contaminant concentrations near detection limits and 2) the high detection limits obtained for the
samples.
4.2.1.2 PAHs
Feed material and treated solids were also analyzed for PAHs. As shown in Table 11, total PAH
concentrations of 0.37 mg/kg, 0.95 mg/kg, and 37.1 mg/kg were found in the treated solids produced by
treating the Buffalo River, Saginaw River, and Grand Calumet River sediments. These values correspond
to removal efficiencies of 96, 65, and 84 percent, respectively.
23
-------
Table 11. Battelle Data - Feed and Treated Solid PAH Concentrations (mg/kg, dry basis)
Buffalo River
Contaminant
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
ldeno(1 ,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
Total PAH
Feed
0.107
<0.080
<0.200
0.160
1.020
0.547
1.20
1.16
0.861
1.04
0.876
0.733
0.887
0.607
0.205
0.495
9.90
Treated '
0.037
<0.020
<0.030
<0.020
0.068
0.030
0.045
0.039
0.022
0.039
0.027
0.004
0.018
0.018
0.006
0.014
0.37
% Removal
64
NC
NC
>88
93
95
96
97
98
96
96
100
98
97
95
96
96
Feed
0.026
<0.020
<0.030
0.033
0.267
0.066
0.397
0.439
0.186
0.269
0.242
0.179
0.225
0.207
0.043
0.117
2.70
Saginaw River
Treated
0.024
<0.020
<0.020
<0.020
0.099
0.017
0.138
0.120
0.057
0.088
0.097
0.066
0.076
0.090
0.016
0.060
0.95
% Removal
08
NC
NC
>39
62
70
65
73
68
67
60
63
65
56
68
48
65
Grand Calumet River
Feed
4.40
2.32
4.40
4.62
15.2
5.63
32.0
32.0
18.3
24.4
19.2
13.4
20.6
14.7
5.22
13.8
230
Treated
2.25
0.121
0.726
1.08
5.64
1.47
3.11
3.55
3.13
3.99
1.89
1.32
3.22
1.36
1.93
2.35
37.1
% Removal
49
95
84
77
63
74
90
89
83
84
90
90
84
91
63
83
84
NC = Not Calculated
Generally, the low removal efficiencies obtained for the PAHs in the Saginaw River sediment can
be attributed to the low concentration of PAHs initially present in the sediment and errors associated with
evaluating contaminant concentrations close to analytical detection limits.
The removal efficiency of 84 percent for the total PAHs in the Grand Calumet River sediment
resulted in a final concentration of 37.1 mg/kg of PAHs in the treated solids. Additional extractions would
likely reduce PAH concentrations in the treated solids even further.
24
-------
4.2.1.3 Total Metals
The data in Table 12 highlight the removal achieved for the metal contaminants present in the
untreated feed and the treated solids. As demonstrated by the low or negative removal percentages, in
general, the B.E.S.T.® Solvent Extraction Process does not effectively remove metals.
Table 12. Battelle Data - Metals Concentration in the Feed and Treated Solids (mg/kg, dry basis)
Buffalo River
Contaminant
Arsenic
Barium
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Zinc
Feed
12.7
413
2.10
109
70.2
42900
102
667
0.551
43.1
0.74
0.31
180
Treated
14.6
396
2.11
113
61.2
44200
102
684
0.627
42.1
0.87
0.24
190
%
Removal
-15
4
-0
-4
13
-3
0
-3
-14
2
-18
23
-6
Saqinaw River
Feed
2.21
322
4.14
107
58.8
7870
45.5
165
0.167
58.3
<0.3
0.84
140
Treated
2.85
319
4.26
118
64.1
8260
46.6
177
0.335
64.3
<0.3
0.82
169
%
Removal
-29
1
-3
-10
-9
-5
-2
-7
-99
-10
NC
2
-21
Grand Calumet River
Feed
22.8
317
8.56
2270
188
188000
582
3230
1.53
12.9
<0.3
4.84
2380
Treated
29.0
290
6.97
1710
223
82500
656
2540
1.46
<10
4.94
4.34
2810
%
Removal
-27
9
19
25
-19
56
-13
21
4
>22
NC
10
-18
NC = Not Calculated
4.2.1.4 Other Analyses
The feed sediments and treated solids were analyzed for percent moisture, oil and grease, TOC,
total volatile solids, and pH as shown in Table 13. As shown by comparing data in Tables 10 and 11 with
data in Table 13, reductions in oil and grease concentrations correspond to PCB and PAH removal. This
demonstrates that oil and grease analysis could possibly be used as a low-cost indicator for technology
effectiveness for a given sediment.
25
-------
Table 13. Battelle Data - Removal Efficiencies for Other Parameters
(mg/kg, dry basis, unless specified)
Buffalo River
Saginaw River
Grand Calumet River
Contaminant
Total PCBs
Total PAHs
Moisture, %
(as received)
Oil & Grease
TOC, % weight
Total Volatile
Solids, %
pH, S.U.
(as received)
Feed
0.32
9.90
42.0
2420
1.98
4.03
7.29
Treated
<0.3
0.37
3.72
238
1.21
3.91
10.30
Removal
>6
96
90
39
3
Feed
21.9
2.70
24.0
1350
0.83
2.09
7.30
Treated
0.24
0.95
0.16
265
0.58
1.73
10.73
Removal
99
65
80
30
17
Feed
15.0
230
57.0
32200
17.0
14.2
7.35
Treated
0.44
37.1
0.50
470
13.4
9.06
10.25
Removal
97
84
99
21
36
4.2.2 ON
The concentrations of PAHs and PCBs in the oil extracted from the three sediments can be found
in Tables 14 and 15. Final concentrations in the process solids and water have been included as a
comparative measure of performance. Using values for percent oil determined in the samples received by
Battelle (i.e., 9.3 percent for Saginaw River extract, 6.3 percent for the Buffalo River extract, and 60.0
percent for the Grand Calumet River extract), these concentrations have been adjusted to account for the
triethylamine diluent found in the different oil samples received for analysis. The triethylamine was left in
these samples because of the low oil content and small sample size used for these tests.
Please note that the possibility for introducing error to these corrected oil concentrations does exist.
Analytically determined values are adjusted for the amount of oil determined to be present in the
oil/triethylamine solution. When the PAH and PCB concentrations are adjusted, any error in this oil analysis
may be conveyed to the new concentrations. Samples with less oil (i.e., Saginaw River and Buffalo River)
are more likely to be affected.
26
-------
Table 14. Battelle Data - PAH Concentrations in the Treated Solids, Water, and Oil
Buffalo River
Contaminant Solid Water
(ug/kg, dry) (ug/L)
Naphthalene
Acenaphthylene
Acenaphthene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
lndeno(1 ,2,3-cd)pyrene
Dibenzo (a, h) anthracene
Benzo(g,h,i)perylene
Total PAH
37
<20
<30
<20
68
30
45
39
22
39
27
4
18
18
6
14
370
0.998
<0.4
<0.5
<0.5
<0.3
<0.4
<0.3
<0.3
<0.3
0.242
<0.3
<0.2
<0.3
<0.3
<0.3
0.184
1.41
Saginaw River
Oil' Solid Water
(ug/kg) (ug/kg, dry) (ug/L)
<9000
< 10000
<20000
24500
160000
126000
201000
183000
89400
1 20000
86900
68400
84900
67500
13300
39400
1260000
24
<20
<20
<20
99
17
138
120
57
88
97
66
76
90
16
60
950
1.44
<0.6
<0.7
<0.7
<0.5
<0.6
0.402
<0.5
<0.5
<0.4
<0.4
<0.3
<0.4
<0.4
<0.4
0.289
2.13
Grand Calumet
Oil" Solid Water
(ug/kg) (ug/kg, dry) (ug/L)
<9000
<20000
<20000
18000
170000
104000
280000
257000
1 1 6000
143000
125000
92400
1 1 4000
90500
31000
69500
1610000
2250
121
726
1080
5640
1470
3110
3550
3130
3990
1890
1320
3220
1360
1930
2350
37100
0.301
0.495
0.165
0.278
2.72
0.997
17.1
18.0
8.42
10.9
6.80
3.97
6.18
3.24
0.762
2.84
83.2
River
Oil'
(ug/kg)
<30000
48000
<60000
51800
213000
110000
636000
608000
349000
484000
432000
289000
433000
362000
75400
217000
4310000
* Corrected for actual volumes of oil present in oil/triethylamine samples analyzed.
(9.3 percent for Saginaw River, 6.3 percent for Buffalo River, 60.0 percent for Grand Calumet River)
27
-------
Table 15. Battelle Data - PCB Concentrations in the Treated Solids, Water, and Oil
Buffalo River Saginaw River Grand Calumet River
Contami- Solids Water Oil3 Solids Water Oil3 Solids Water Oil8
nant (ug/kg, dry) (ug/L) (ug/kg) (ug/kg, dry) (ug/L) (ug/kg) (ug/kg, dry) (ug/L) (ug/kg)
Total PCBs <300 <0.6 62300 235 <0.6 5012000 440 4.8 268000
• Corrected for actual volumes of oil present in oilAriethylamine samples analyzed.
(9.3 percent for Saginaw River, 6.3 percent for Buffalo River, 60.0 percent for Grand Calumet River)
4.2.3 Water
The concentrations of PAHs and PCBs in the water extracted from the three sediments can also
be found in Tables 14 and 15. As the data demonstrate, individual PAH and PCB concentrations for the
Buffalo River and Saginaw River residual waters were mainly below the detection limits. Please note that,
like the PAH and PCB concentrations associated with the treated solids and untreated sediments, the PAH
and PCB concentrations found in the Grand Calumet River residual waters were substantially higher than
the concentrations found in the Buffalo River and Saginaw River residual waters. Possibly additional
extractions could reduce these concentrations to levels comparable to Buffalo River and Saginaw River
concentrations.
4.2.4 Mass Balance
For the B.E.S.T.® bench-scale treatability tests, good mass balance closures were obtained for the
solids, water, oil, PCBs, and PAHs. The following sections address the different mass balances and
expand on those factors that influenced their closure. Tables are included in these sections which provide
the data used to calculate the mass balance.
During the mass balance discussions, terms are introduced which require definition. These
definitions are provided as follows:
• Input solids include the solids initially present in the sample plus those solids introduced by
caustic addition.
• Output solids are the final product solids and the RCC samples taken during the tests.
• Input water includes the water initially present in the sample and the water contributed by the
addition of caustic.
• Output water consists of the volume of product water obtained after the cold wash extraction.
28
-------
4.2.4.1 Solids
Closure of the solids was very good, ranging from 92 to 99 percent (Table 16). Since the values
used to determine closure were simple weights rather than analytical results, the mass balance was not
compromised by errors associated with analytical methods. Small quantities of solids deposited in the
vessels and containers used during the treatability study and found in the rag layer resulting from the cold
wash were not accounted for in the mass balance.
Table 16. Battelle Data - Solid Mass Balance
Buffalo River Saginaw River Grand Calumet River
Input
Total Feed, g 900 899.7 1399.5
H2O, % 41.96 23.98 57.04
Total Feed Solids, g (dry) 522.4 682.6 601.2
Caustic (d=1.53) 4.1 4.1 12.6
Total Input, g (dry) 526.5 686.7 610.8
Output
RCC Sample Cold Wash, g
RCC Sample 1st Hot Wash, g
RCC Sample 2nd Hot Wash, g
Final Dry Solids, g
Total Final Solids Output, g
(dry)
Recovery, %
11.2
16.3
29.2
461
517.7
98.3
30.6
23.4
21.3
606
681.3
99.2
6.9
7.1
12.3
485
511.1
91.9
4.2.4.2 Water
Closure of the water mass balance ranged from 67.5 to 77.9 percent (Table 17). Water lost to
evaporation and water remaining in the rag layer and in the solids after the product water was decanted
from the reactor vessel following the cold wash (i.e., beyond that directly associated with the triethylamine)
did not contribute to the output water recovered. Because the majority of the water was removed from the
raw sediments during the cold wash, closure calculations are based on data obtained during the cold wash
only. The closed loop of the full-scale system would probably improve closure results.
29
-------
Table 17. Battelle Data - Water Mass Balance
Total Feed, g
Water, %
Total Feed Water, g
Caustic Water, g
Total Input Water, g
Water Recovered, g
Net minus 2% TEA, g
Recovery, %
Buffalo River
900
41.96
377.6
4.1
381.7
262.8
257.5
67.5
Saginaw River
900
23.98
215.8
4.1
219.9
174.9
171.4
77.9
Grand Calumet River
1399.5
57.04
798.3
9.6
807.9
609.8
597.6
74.0
4.2.4.3 Oil
The amount of oil initially present in the raw sediment and the amount of oil produced by the tests
were used to determine the oil mass balance. The amount of oil present in either the treated solids or
water is known from past tests to be insignificant and, therefore, was not accounted for. In order to retrieve
the residual oil from the distillation flask, triethylamine was added to the oil so it could be poured. Battelle
later determined the percentage of oil present in the resulting solutions (6.3 percent for the Buffalo River
solution, 9.2 percent for the Saginaw River solution, and 60.0 percent for the Grand Calumet River solution)
and compensated for these percentages when reporting final data. The need to compensate for these oil
determinations possibly introduced error into the calculation of the oil mass balance. The calculated
closures ranged from 69 to 192 percent as shown in Table 18.
Table 18. Battelle Data - Oil Mass Balance
Feed Input, g
Oil and Grease, %
Input Oil, g
Final Oil & TEA, g
Oil, %
Final Oil, g
Recovery, %
Buffalo River
900
0.24
2.16
55.9
6.3
3.52
163
Saginaw River
899.7
0.14
1.22
24.4
9.2
2.34
192
Grand Calumet River
1399.5
3.22
45.1
51.6
60.0
31.0
68.7
30
-------
4.2.4.4 PCBs
The PCB mass balance was calculated using the amount of PCBs found in the feed, product oil
(with triethylamine), and product solids. The contribution of the PCBs found in the product water was
negligible and therefore not included in the mass balance calculations.
The closures of the PCBs were good, ranging from 80 to 129 percent (Table 19). When calculating
these closures, it was assumed that the concentrations of the PCBs in the samples taken by RCC during
the treatability study were the same as those found in the final products. In reality, these concentrations
should be higher, since they were removed during earlier extraction stages. The need to compensate for
the excess triethylamine present in the oil extract solutions may have introduced errors to the determination
of the PCB concentrations found in the product oils. These errors could have subsequently been conveyed
to the PCB mass balance closures.
Table 19. Battelle Data - PCB Mass Balance
Feed Input, g
Solids, %
Dry Solids, %
PCBs, ug/kg
PCBs Input, mg
Buffalo River
900
58.0
522.4
325
0.17
Saginaw River
900
76.0
684.2
21865
15.0
Grand Calumet River
1399.5
43.0
601.2
15003
9.02
OUTPUT
OH
TEA & Oil wt., gm
PCBs Cone., ug/L
d of TEA & Oil, gm/L
PCBs Output wt., mg
Solids
55.9
2702
690
0.22
25.4
349,109
752
11.8
51.6
117,156
730
8.28
Dry Solids wt., gm
PCBs in Solids, ug/kg
PCBs wt., mg
Total Output PCBs, mg
Recovery, %
522.7
ND
0.0
0.22
129
681.3
235
0.16
12.0
80
511.1
440
0.23
8.51
94
ND = Not Detected
31
-------
4.2.4.5 PAHs
The closures of the PAHs were calculated using the amount of PAHs found in the feed, product
oils (with triethylamine), and product solids. The contribution of the PAHs found in the product water was
negligible and therefore not included in the mass balance calculations.
The closures of the PAHs were good (Table 20), ranging from 90 to 111 percent for Buffalo River
and Grand Calumet River sediments. Saginaw River sediments, however, realized a closure of 240
percent. This may be attributed to the low concentrations of PAHs initially present in the sediment and the
large errors associated with contaminant concentrations close to analytical detection limits.
Table 20. Battelle Data - PAH Mass Balance
Buffalo River Saginaw
River
Feed Input, g
Solids, %
Dry Solids, %
PAHs, ug/kg
PAHs Input, mg
OUTPUT
OH
TEA & Oil wt., g
PAHs Cone., ug/L
d of TEA & Oil, g/L
PAHs Output wt., mg
Solids
Dry Solids wt., g
PAHs in Solids, ug/kg
PAHs wt., mg
900
58.0
522.4
9,900
5.17
55.9
54,864
690
4.44
522.7
370
0.19
900
76.0
684.2
2,700
1.85
25.4
112,070
752
3.79
681.3
950
0.65
Grand Calumet River
1399.5
43.0
601.2
230,000
138
51.6
1,886,715
730
134
511.1
37,100
19
Total Output PAHs, mg
Recovery, %
4.63
90
4.44
240
153
111
32
-------
4.3 Summary of Vendor Results
The analytical results and the extraction efficiencies and mass balances performed by RCC match
well with the analytical results provided by Battelle and the extraction efficiencies and mass balance
calculated by SAIC. The RCC data were used to make comparisons with the results obtained and to help
interpret data. Samples of the feed material and the treated solids produced using the B.E.S.T.® Solvent
Extraction Process were analyzed by RCC for residual PCB contamination. The data from these analyses
are presented in Table 21.
Table 21. RCC Data - PCB Summary
Sample
(dry Basis)
Buffalo River
Saginaw River
Grand Calumet River
Feed PCB
Content, mg/kg
(dry basis)
0.60
21
22
Treated Solids
PCB Content, mg/kg
(dry basis)
<0.03
0.18
0.23
PCB Removal
Efficiency
(%)
>95
99
99
Feed material and treated solids were also analyzed for residual PAH and for metals concentra-
tions. Residual PAH concentrations of <0.2 mg/kg per compound were found in the treated solids produced
by treating the Buffalo River and Saginaw River sediments. Treated solids with PAH concentrations
ranging from <1 to <3 mg/kg per compound were obtained for the Grand Calumet River sediments. In tests
done by RCC, all of the treated solids passed the TCLP Toxicity Test for the leaching of metals.
RCC performed a mass balance of solids, water, oil, and PCBs using their own data. Table 22
summarizes the results of this mass balance.
Table 22. RCC Data - Mass Balance Summary (Recovery %)
Sample
Buffalo River
Saginaw River
Grand Calumet River
Solids
97
98
86
Oil
112
137
97
Water
70
82
75
PCBs
70
280
64
33
-------
4.4 Quality Assurance/Quality Control
The conclusions and the limitations of data obtained during the evaluation of RCC's B.E.S.T.®
Process are summarized in the following paragraphs.
Upon review of all sample data and associated QC results, the data generated for the B.E.S.T.®
treatability study have been determined to be of acceptable quality. In general, QC results for accuracy
and precision were good and can be used to support technology removal efficiency results.
In some cases, the demonstration of removal efficiency for PAHs and PCBs may be limited if
relatively small amounts of these compounds are present in the untreated sediments. If minimal amounts
are present, then detection limits become a factor. Removal efficiency demonstration may be limited by
the sensitivity of the analytical methods.
Refer to Appendix F for the complete analysis related to Quality Assurance/Quality Control.
34
-------
APPENDIX A
B JLS.T.® BENCH-SCALE TREATABILITY
TEST REPORT
Great Lakes National Program Office
Buffalo River, Saginaw Bay and Indiana Harbor Sites
I. INTRODUCTION
SUMMARY
A bench-scale treatability test of the B.E.S.T. solvent extraction process was conducted on three
pclychlorinated biphenyl (PCB) contaminated sediment samples. One sample was received from
each of the sites, Buffalo River, Saginaw Bay and Indiana Harbor. A summary of the bench-scale
treatability test results follows:
BENCH SCALE TREATABILITY TESTS RESULTS
Feed PCB Product Solids PCB Removal
Sample Content, mg/kg PCB Content, ma/kg Efficiency, %
(dry basis) (dry basis)
Buffalo River 0.60 < 0.03 > 95
Saginaw Bay 21 0.18 99
Indiana Harbor 22 023 99
As can be seen from the data above, the PCB residuals of the treated solids (Product Solids) varied
from < 0.03 mgAg to 0.23 mgAg, yielding PCB removal efficiencies of > 95 to 99%. Individual.
residual PAH concentrations in the treated solids were < 0.2 mg/kg for the Buffalo River and Saginaw
Bay samples and ranged from < 1 to < 3 mg/kg for the Indiana Harbor sample.
All of the treated solids readily passed the TCLP Toxicity Test for the leaching of metals.
35
-------
THE B JLS.T. SOLVENT EXTRACTION PROCESS
The B.E.S.T. process is a patented solvent extraction technology using triethylamine as the solvent.
Triethylamine is an aliphatic amine that is produced by reacting ethyl alcohol and ammonia.
Triethylamine is an excellent solvent for treating hazardous wastes because it exhibits several
characteristics that enhance its use in the solvent extraction system. These characteristics include:
• A high vapor pressure; therefore, the solvent can be easily recovered from the extract
solution (oil, water, and solvent) via steam stripping.
• Formation of a low boiling temperature azeotrope with water, allowing the solvent to be
recovered from the oil to very low residual levels (typically less than 100 ppm).
• A low heat of vaporization (1/7 of water), allowing solvent to be recovered from the
treated solids with very low energy input.
• Triethylamine is alkaline (pH=10); therefore, some heavy metals are converted to the
hydroxide form, precipitate and exit the system with the treated solids.
• Triethylamine readily biodegrades. Data available in EPA document EPA Data QRD
USEPA Washington. D.C. 20460. Feb. 1983 (reprint) Manual. Volume 1 600/2-82-001 a.
shows that a level of 200 ppm triethylamine in water was degraded completely within 11
hours by the common soil bacteria aerobacter.
A block diagram of the B.E.S.T. process is presented in Figure 1. The first extraction of the
contaminated feed is conducted at low temperatures (about 40 degrees F). At this temperature,
triethylamine is soluble with water. Therefore, the extract solution contains most of the water in the
feed sample. If the first extract solution contains sufficient water to allow a phase separation of the
solvent and water, the extract is heated to a temperature above the miscibility limit (130 degrees F).
At this temperature, the extract solution separates into two distinct phases, a solvent/oil phase and a
water phase. The two phases are separated by gravity and decanted. The extract solution from the
subsequent stages is combined with the decanted solvent/oil phase from the first extraction stage.
The solvent is recovered by steam stripping and evaporation.
Triethylamine is removed from the treated solids by indirect steam heating. A small amount of steam
may be added directly to the dryer vessel to provide the water required to form the low boiling
temperature azeotrope. Residual solvent biodegrades readily, allowing the treated solids to be used as
backfill at the site in some cases.
The B.E.S.T. process operates near ambient pressure and temperature and at a alkaline pH.
Temperatures of the liquid streams within the the unit vary from about 40 to 170 degrees F, and
elevated pressures are not required. This gives the B.E.S.T. process the advantage that it can use
standard off-the-shelf processing equipment.
36
-------
RCC
CO
Waate
B.E.S.T. PROCESS CONCEPT
Extraction
Recycled
Solvent
Solvent Recovery
I
Solvent (to recycle)
Solvent/OH f
Extraction
Subsequent
Extractlone
First
Extraction
<3>40*F
l
Separation
Steam
Stripping
Oil
Solvent/Water
Solvent (to recycle)
f
Steam
Stripping
Water
Solids
Solvent (to recycle)
t
Solids
Drying
-Solids
-------
AIR EMISSIONS AND ABATEMENT
The B.E.S.T. process uses one vent to the atmosphere. The vent provides pressure equalization for
the nitrogen blanketing system and a purge for noncondensible gases from process condensers. RCC
uses a refrigerated condenser and an auxiliary water scrubber system to reduce solvent emissions from
the vent
During a performance test in February 1987 at the General Refining Superfund Site cleanup, a third
party reported the following emissions from the B.E.S.T. process vent at a time when the auxiliary
water scrubber was not in operation:
Emission Rate. Ib/hr
Benzene 0.00114
Mercury < 0.000000043
Toluene 0.000614
Triethyiamine 0.0954
Xylene 0.000884
RCC expects air emissions from future operations to be similar to these results. The use of the
auxiliary water scrubber will lower the triethylamine release rate even further. RCC now utilizes
activated carbon filters on the single vent line to achieve zero emissions of triethylamine.
EQUIPMENT DESCRIPTION
RCC proposes using a B.E.S.T. Model 615 unit to treat the PCB-contaminated material at this site.
The B.E.S.T. Model 615 unit has a design capacity of approximately 200 - 300 tons of feed per day.
A flow schematic for the B.E.S.T. Model is presented in Figure 2.
The B.E.S.T. Model 615 uses an extractor/dryer vessel to extract and dry the PCB-contaminated
materials. The extractor/dryer is a horizontal, steam-jacketed vessel that allows for solvent
contacting, mixing, solids/solvent separation, solids drying, and solids conditioning in one vessel.
The extractor/dryer vessel is an off-the-shelf assembly that has a long history of reliable performance
in a wide range of process industry applications.
Contaminated materials are excavated from the site and screened to one inch maximum dimension.
The screened material is then loaded into top-loading, bottom-discharge hoppers. An overhead crane
facilitates the positioning and lowering of the loaded hopper onto the loading port of the
extractor/dryer unit. The flow of material through the extractor/dryer system is shown in Figure 3.
Treated solids are discharged into hoppers and transported to a holding area.
Figure 4 provides the standard Site Plan for RCC's B.E.S.T. Model 615.
38
-------
CONDENSER
CONDENSER
(DRYING CYCLE)
SOLVENT
OPTIONAL
CENTRIFUGE
CLEAN
SOLIDS
PRODUCT
SOLVENT
SPENT
SOLVENT
TANK/SETTLER
TANK
n
r i
CHILLER
lANKh
SOLVENT
MAKEUP
CLEAN
SOLVENT
TANK
SOLVENT
EVAPORATOR
SOLVENT
DECANTER
PRODUCT
INOKCOHO
PROCESS FLOW SCHEMATIC
r®
B.E.S.TW MODEL 315/615
SOILS TREATMENT UNIT
DWG NO.
B-221
»«»«
1 OF
•"
A
A
-------
f?CC
Httourttt
Conttrvitlon
Company
B.E.S.T.® PROCESS STEPS
BATTERY LIMITS OPERATION
SOUDS SETTUNG
AND SOLVENT DRAINING
WASHER/DRYER CHARQNO
3
-------
Procession
Storane Tanks Equipment Six Extractor/Dryers
RCSOURttS COM$tBV«nOM COUPtHt
-------
BENCH-SCALE TREATABILITY TEST DATA CORRELATION TO FULL-SCALE
PERFORMANCE
In order to evaluate each potential application for the B.E.S.T. process. RCC has developed a low cost
bench-scale treatability test protocol that provides data that closely simulates full-scale system
performance. The bench-scale treatability test data allows RCC to evaluate the feasibility of the
process on a particular sample and to estimate treatment costs.
The reliability of the bench-scale treatability tests to predict full-scale performance has been verified
by the US EPA report Evaluation of the B.E.S.T. Solvent Extraction Sludge Treatment Technology -
Twenty-Four Hour Test, by Enviresponse. Inc., under EPA Contract 68-03-3255. A quote from this
report evaluating the B.E.S.T. process states:
"Resources Conservation Company has conducted many laboratory tests and developed
correlations to which data from full-scale operations, such as the General Refining site,
can be compared."
Figures 5 and 6 present data from two separate bench-scale treatability tests and full-scale operating
performance data at the General Refining. Inc.. Superrund site, as collected by an EPA contractor.
This data demonstrates a close correlation between bench-scale treatability test data and full-scale
operating data.
Bench-scale treatability testing provides valuable information about the use of the B.E.S.T. process at
full-scale including:
• The PCB removal efficiency from the sample.
• Solids separation requirements for full-scale operation.
• The separation efficiency of water from the water/solvent/oil solution by decapitation.
• General information on the partitioning of metals and organic compounds in the oil, water.
and solids products.
• Full-scale operating parameters to develop treatment costs.
42
-------
RCC
Resources
Conserirjffon
Company
GENERAL REFINING SITE
PCB CONCENTRATIONS IN RAW SLUDGE & PRODUCT FRACTIONS
(ppm)
LAB SCALE TESTING (1986) FULL SCALE PROCESSING
TEST "A" TEST "B" FEB. 26-27,1987
RAW SLUDGE (DRY BASIS) mg/kg 14.
PRODUCT SOLIDS mg/kg 0.02
PRODUCT WATER mg/L <0.01
"/..EXTRACTION EFFICIENCY 99.9
12.
0.14
<0.01
98.8
13.5
<0.13
<0.005
>99.0
ua
3
in
-------
COMPARISON OF BENCH SCALE TO FULL SCALE
Off**™"™'**" PHASE SEPARATION PERFORMANCE
• • ^* ^* Company FOR
GENERAL REFINING SITE SLUDGE
Raw Sludge
Oil % 36
Water % 56
Solids % 8
N/A Not Available
BS&W = 2.8%
Bench Scale
Separated Phase Fractions
Oil Water Solids
>97. .017 5.7
N/A <1.0
N/A >94.
Raw Sludge
27
66
7
Full Scale
Separated Phase
Oil Water
99. 0.0033
0.88 >99.
* 0.81
Fractions
Solids
0.81
<05
>98.
•a
PUOAG4
-------
BENCH-SCALE TREATABILJTY TEST DOCUMENTATION
The documentation of the testing can be separated into three distinct categories. The following
summarizes the procedures used for each step of the trcatability process:
1. When the samples were received in the laboratory, the shipment was checked for
correctness of accompanying paper work, including Chain of Custody. The information
was recorded both in a hardbound sample logbook and on a computer system that has been
specifically designed by RCC for use in tracking samples. The samples were issued a
discrete laboratory sample number, and a test request form was completed. The samples
were kept in a refrigerator under controlled and documented temperature prior to any lab
analysis or the trcatability study. Chain of Custody records and other information received
with the samples are kept as pan of the project file.
2. The bench-scale treatability testing was conducted in accordance with the test plan, and all
records and observations taken during the simulation of the process were recorded in
laboratory notebooks. The laboratory notebooks are the property of RCC, and each analyst
and engineer has been issued a notebook. The notebooks are retained by RCC as
permanent record of raw data collection.
3. Samples that were collected during the bench-scale test, including samples internal to the
process, were submitted to the RCC analytical chemistry laboratory for further analysis.
Each sample collected was issued a discrete laboratory number. An analysis request form
was completed. The samples were analyzed in accordance with RCC's Laboratory Quality
Management Plan and reviewed for correctness prior to issuance. A file is maintained to
permanently store the accumulated test results from completion of the analytical testing.
The bench-scale treatability test plan is provided in Attachment 1. PCB chromatograms are given in
Attachment 2. PAH chromatograms are given in Attachment 3.
H. BENCH-SCALE TREATABILITY TESTING
SAMPLE PREPARATION
The contaminated samples from the Buffalo River, Saginaw Bay and Indiana Harbor Sites arrived at
RCC's laboratory in July, 1991. The samples were labeled B-US-RCC, S-US-RCC and I-US-RCC,
respectively. All samples were gray-colored sediments with very little debris present. Each of the
three samples contained free standing water. This water was decanted prior to conducting feed
analyses and was proportionally recombined prior to any analysis and bench testing. This effort was
taken since it was very difficult to homogenize the samples with the free standing water present
Bench-scale testing requires material greater than 1/4 inch be removed. There was no material greater
than 1/4 inch in any of the three samples received. Therefore, the samples were not screened.
45
-------
FEED COMPOSITIONAL ANALYSIS
The feed was analyzed for percent oil, water, solids and metals per the following methods:
• The oil & grease content was determined as per Standard Methods for the Examination of
Water and Wastewater. 16th Edition, Method 503D, with two exceptions: the extraction
time was extended from 4 to 16 hours, and methylene chloride (MeQ2)was substituted for
Freon based on RCC experience that MeCl2 is a better solvent for oils and greases.
• The water content was determined by weight loss at 70 degrees C.
• The paniculate solids content was determined by rinsing a known quantity of feed through
a Whatman GF/C filter under vacuum with acetone followed by Med^ The ^idue was
then dried and weighed. Since the paniculate solids content was by far the largest
component of all of the feeds, the percent solids values below are by difference.
• The PCS concentration was determined per EPA Publication SW846 Test Methods for
Evaluating Solid Waste. Method 8080. The sample extraction method was by Soxhlet
extraction with 1:1 acetone:hexane for 16 hours. The PCBs were quantitated as Arocior
1242.
• The Polynuciear Aromatic Hydrocarbon (PAH) concentrations were determined per EPA
Publication SW846, Method 8100 (1:1 Acetone:Hexane extraction solvent).
• The metals composition (except for Mercury) was determined by nitric acid digestion after
ashing at 550 degrees C, followed by ICP analysis (EPA SW846, Method 6010).
• Mercury concentration was determined by the Cold Vapor Technique, Method 303F, of
Standard Methods for the Examination of Water and Wastewater.
• Loss on ignition was determined by heating a sample from 105 to 550 degrees C which is a
measure of the total organic content.
46
-------
The results of these analyses on a wet basis were as follows:
Analyte
Feed Compositional Analysis
(wet baste unless noted)
Sample Results
Oil & Grease (by MeCl2),
Water, %
Solids. %
Loss on Ignition, %
PCBs, mg/kg, dry basis
Buffalo River
0.48
41.
59.
4.8
0.6
Saginaw Bay Indiana Harbor
0.36
23.
77.
21
2.4
56.
42.
26.
22
Individual feed PAH concentrations for the Buffalo River sample ranged from <0.5 to 6 mg/kg, for
the Saginaw Bay sample they were <0.4 to <3 mg/kg and for the Indiana Harbor sample they were <8
to <60 mg/kg. These results are presented on page 17 and 18.
The heavy metals composition of each feed was as follows:
Feed Metals Composition, mg/kg
(As received basis)
Anaivte
Antimony
Arsenic
Barium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Silver
Sodium
Zinc
Buffalo River
<30.
34.
<2.
14.
16.
28.
0.38
9.7
<20.
< 1.
115.
52.
Saginaw Bay
< 15.
<40.
32.
<3.
62.
37.
37.
0.10
36.
<25.
< 1.
150.
120.
Indiana Harbor
< 15.
<35.
76.
6.4
156.
96.
290.
0.72
29.
<20.
<0.7
300.
1,340.
47
-------
TRIETHYLAMINE COMPATIBILITY TEST
Tricthylamine is a compound with a unique chemical structure. The geometry of the structure is
tetrahedral, meaning that the nitrogen atom is at the center of a three-sided pyramid. The four points
of the pyramid structure are occupied by three ethyl functional groups and one electron cloud. This
structure gives triethylamine dual polarity characteristics. The ethyl groups are essentially nonpolar.
the electron cloud is polar. Although triethylamine is a very stable solvent, there is a remote
possibility that the electron pair can react with certain types of materials. In order to determine if this
will occur with a sample, a compatibility test is performed. This involves mixing of the sample with
triethylamine and making observations as to the heat of solution and any other visual signs of
reaction.
When each feed sample was mixed with cold triethylamine, no visible sign of adverse reaction was
observed, and the heat of solution was in a normal range. The triethylamine was observed to darken
upon mixing, indicating that extraction of the organic compounds was occurring.
Based on the favorable results of this preliminary test, it was decided that the B.E.S.T. bench-scale
treatability test should proceed.
FEED pH ADJUSTMENT
Triethylamine can be ionized at low pH to triethyiammonium salts that cannot be removed from the
products. The alkaline nature of triethylamine will buffer the pH of the sample to a pH of around 9.
The solvent spent in the pH buffering will be lost. In order to efficiently recover the triethylamine
from the separated phase fraction products, the pH of the sample is adjusted to about 11 with causuc
soda.
A 5-gram portion of each feed sample was siurried with deionized water. The pH of this mixture
indicated that caustic would need to be added to each sample. Incremental portions of causuc soda
(NaOH) were added to bring the pH to 11. The amount of causuc that was required to perform this
pH adjustment and the original sample pH is summarized below:
Sample pH and Caustic Dose
Caustic Dose
pH (mis 50% NaOH per kg)
Buffalo River 7.6 6.0
Saginaw Bay 8.1 6.0
Indiana Harbor 7.5 9.0
48
-------
SAMPLE EXTRACTION/PRODUCT SOLIDS
A portion of the Buffalo River, Saginaw Bay and Indiana Harbor samples was prechilled by placing
each in a 4-liter resin kettle, immersed in a temperature controlled water bath set at 0.5 degree C.
Each sample pH was adjusted by adding caustic soda at the same time that chilled triethylamine was
added. Mixing was performed by an air-driven prop mixer in the same 4-liter resin kettle immersed
in the cooling bath.
As expected, the solvent became colored for all three samples, indicating extraction of organic
compounds was occurring. After mixing, the solvent/oil/water liquid extract was separated from the
solids by centrifugation. The liquid extract was temporarily set aside for testing as discussed later
under DECANTATION OF WATER.
Two more extraction stages were performed on the solids, for a total of three extraction stages. No
additional caustic was added for the subsequent extraction stages. A sample of the Product Solids
was collected for analysis as follows:
• The oil & grease content was determined as per Standard Methods for the Examination of
Water and Wastewater. 16th Edition, Method 503D, with two exceptions: the extraction
lime was extended from 4 to 16 hours, and methylene chloride (MeCty was substituted for
Freon based on RCC experience that MeCl2 is a better solvent for oils and greases.
• Loss on ignition was determined by heating a sample to 550 degrees C for 3 hours.
• The PCB concentration was determined per EPA Publication SW846 Test Methods for
Evaluating Solid Waste. Method 8080. The sample extraction method was by Soxhlet
extraction with 1:1 Acetone:Hexane for 16 hours. The PCBs were quantitated as Aroclor
1242.
• The metals composition was determined by aqua rcgia digestion, followed by ICP analysis
(EPA SW846, Method 6010).
• The triethylamine content was determined by shaker bath water extraction and packed
column gas chromaiography with a flame ionization detector.
• The pH was determined by measuring the pH of a slurry of 5 grams of sample and 50 mis
of deionized water. The slurry was tested by pH probe after mixing overnight.
• The Polynuclear Aromatic Hydrocarbon (PAH) concentrations were determined per EPA
Publication SW846, Method 8100 (1:1 Acetone:Hexane extraction solvent).
• Mercury concentration was determined by the Cold Vapor Technique, Method 303F, of
Standard Methods for the Examination of Waste and Wastewater.
49
-------
PCB analytical results of all of the solid samples were as follows:
PCB Analysis Summary, me/kg
(all data dry basis)
Buffalo River Saginaw Bay Indiana Harbor
Feed 0.60 21 22
Product Solids <0.03 0.18 0.23
Total polynuciear hydrocarbon (PAH) analytical results of the feeds and product solids from the three
samples are given on the following two pages. Due to a large number of interfering compounds eluting
at or near the retention times of the analytes of interest, it was not possible to report lower detection
limits.
50
-------
PAH Summary for Buffalo River
(nig/kg)
Feed
Product
Solids
Naphthalene
2-Methylnaphthalene
Acenaphthylene
Acenaphthene
Dibenzofuran
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
B enzo(k)fluoranthene
Benzo(a)pyrene
Indeno(l,2,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(gji,i)perylene
<0.5
<0.5
<0.5
<0.5
< 1.
<0.5
<0.5
<0.5
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
PAH Summary for Saginaw Bay
(mg/kg)
Feed
Product
Solids
Naphthalene
2-Methylnaphthalene
Acenaphthylene
Acenaphthene
Dibenzofuran
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
Indeno( U,3-cd)pyrene
Dibenzo(a,h)anthracene
Benzo(gji,i)perylene
<0.4
<0.4
<0.4
<0.4
<0.4
<0.4
<0.8
< 1.
<2.
< 1.
<0.4
<0.4
<0.4
<0.4
<0.4
<0.4
51
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
-------
PAH Summary for Indiana Harbor
(mg/kf)
Feed
Product
Solids
Naphthalene
2-Methylnaphthalene
Acenaphthylene
Acenaphthene
Dibenzoniran
Fluorene
Phenanthrcne
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fiuoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
Indeno(1.2,3-cd)pyrene
Dibenzo(aji)anthracene
Benzo(gth,i)perylene
< 6.
< 6.
<4.
< 7.
< 15.
< 12.
< 20.
< 12.
< 45.
< 48.
< 40.
< 38.
< 25.
<28.
< 33.
< 13.
< 10.
< 13.
< 3-
< 3-
< 1-
< 1-
< 3.
< 1-
< 3.
< 1-
< 2.
< 2.
< 2.
< 3.
< 2.
< 1-
< 1-
< 1-
< 1-
< 1-
Additional product solids analyses, all on a dry basis since the product solids were dried in-process,
follows:
Product Solids Analysis
(three extraction stages)
Analyte
Sample Results
Buffalo River Saginaw Bav Indiana Harbor
PCBs,mg/kg
Oil & Grease (by MeCl2), %
Triethylamine, mg/kg
Loss on Ignition, %
<0.03
0.22
37.
4.0
0.18
0.15
20.
2.1
023
0.52
28.
20.
52
-------
PCB removal efficiency is determined by comparing the amount of PCBs in the feed to the amount
remaining in the environment after treatment. The fraction of PCBs remaining in the environment is
calculated by dividing the PCB content of the product solids by the PCB content of the feed, on a dry
basis. An example of the calculation, the Saginaw Bay sample, follows:
Saginaw Bay Sample
PCB Removal Efficiency Calculation
Fraction of PCBs remaining
in environment
Product solids PCB Content (dry basis)
Feed PCB Content (dry basis)
0.18 me/kg
20.5 mg/kg
= 0.00878
% Removal from = 100 • (1 - fraction of PCBs remaining
environment in environment)
100 • (1-0.00878)
99.1 %
The reduction in the PCB content and the corresponding removal efficiency of PCBs from the
environment is summarized below for all of the samples:
Total PCB Removal Summary
Sample
PCBs in
Feed, me/kg
(dry basis)
PCBs in Product
Solids, mg/kg
(dry basis)
Removal
Efficiency. %
Buffalo River
Saginaw Bay
Indiana Harbor
0.60
21
22
<0.03
0.18
0.23
>95
99
99
53
-------
Total heavy metal analysis of the product solids was as follows:
Product Solids
Total Metals Analysis, (me/kg)
Analvte
Buffalo River Saginaw Bav Indiana Harbor
Antimony < 20. < 20. <20.
Arsenic <5Q. <50. <50.
Barium 73, 41 igo.
Cadmium < 3. < 5. 24.
Chromium 51. 93 430.
Copper 62. 60. 270.
Lead 110. 56. 750.
Mercury 0.68 0.20 1.6
Nickel 32. 53. 88.
Selenium < 30. < 30. < 30.
Silver
-------
TOXICITY CHARACTERISTIC LEACHING PROCEDURE ANALYSIS ON PRODUCT
SOLIDS
The product solids from each sample were extracted using the Toxicity Characteristic Leaching
Procedure (TCLP) in accordance with Federal Register, March 29, 1990. Each TCLP leachate was
analyzed for metals content The results from this analysis were as follows:
Product Solids
TCLP Leachate Analysis, mg/I
Analvte
Buffalo River
Saginaw Bav Indiana Harbor
Regulatory
Level, mg/1
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
<0.5
0.11
<0.03
<0.05
<0.1
< 0.002
<0.3
<0.01
<0.5
0.14
<0.03
0.15
<0.1
< 0.002
<0.3
<0.01
<0.5
0.04
<0.03
0.07
<0.1
0.0024
<0.3
0.05
5
100
1
5
5
0.2
1
5
As can be seen from the above data, all product solids readily passed the TCLP test for metals.
55
-------
DECANTATION OF WATER
The solvent recovered from each first extraction stage was separated into its aqueous and organic
components. Only the extract from the first extraction stage had a significant amount of water in
solution, so only the water in the first stage extract is recovered. There are two methods available to
recover the water in the first stage extract. Each method has its advantages. The method chosen
largely depends on the water content of the feed.
In the first method of recovering the water, the triethylamine mixture is heated to about
140 degrees F. The water phase is separated from the triethylamine/oil phase by decamation. At the
elevated temperature, water is no longer miscible in triethylamine and settles to the bottom of the
mixture. The heated mixture is allowed to stand and separate for 30 minutes in a 4-liter separatory
funnel. The separatory runnel is immersed into a clear tank into which water from a temperature
controlled water bath is pumped. Excess water from the tank drains back into the water bath. The
temperature controlled bath is set at 140 degrees F. This low energy method is preferable for high
water content feeds.
Decantation (the first method) was used for both Buffalo River and Indiana Harbor. The separation
does not give only TEA/oil and water phases due to the presence of oil and solids in the mixture. In
between the two phases is a 'rag' layer or emulsion where any solids present tend to collect and create
a region where the TEA/Oil/Water separations is not distinct. The smaller this rag layer is in
comparison to the TEA/OH and water phases, the better the separation. For the Buffalo River sample
the rag layer was only 2.6% of the entire TEA/Oil/Water mixture. For the Indiana Harbor sample the
rag layer was 3.1%. In both cases, the rag layer is a very small fraction of the whole mixture,
therefore, the separation of TEA/Oil from water was good.
In the second method of recovering water, the water from the feed is separated from the oil by
evaporation as opposed to decantation. When the triethylamine/oil/water first stage extract is
evaporated, as described in section SOLVENT EVAPORATION/PRODUCT OIL, the water forms an
azeotrope with the distilled triethylamine, leaving the oil behind. The water is then separated from
the triethylamine of the condensed triethylamine/water azeotrope by decantation. The
triethylamine/water recovered from the Solvent Evaporation/Product Oil step is heated to
140 degrees F, then poured into a 4-liter separatory funnel. Separation occurs immediately, so no
temperature control system is required. This separation is highly effective because there is virtually
no oil or solids in the condensed triethylamine/water that would hinder the separation of triethylamine
from water by decantation. This method is preferable for low water content feeds where the extra
energy cost to evaporate the water is small. This method yields a much purer water stream, usually
avoiding the requirement for possible post-treatment of the water.
Evaporation (the second method) was used for the Saginaw Bay sample because of its low water
content (< 25%). The aqueous phase separated from the organic phase, producing excellent results
for each.
56
-------
PRODUCT WATER
Removal of residual triethyiamine from each decant water was accomplished by heating the water on
a hot plate while maintaining an elevated pH. The elevated pH is necessary to ensure that the
majority of the triethyiamine remains in the volatile molecular form. Triethylamine/waier azeotrope
boils at about 170 degrees F. When the triethyiamine is removed, the water temperature increases to
212 degrees F. Analysis of each stripped water, labeled Product Water, was conducted as follows:
• The triethyiamine content was determined by packed column gas chromatography with a
flame ionization detector.
• There was insufficient sample for Total Petroleum Hydrocarbons, Oil & Grease, PCS or
total metals analyses since the bulk of the water was sent to a third party lab.
Results of these analyses were as follows:
Product Water Analysis, mg/l
Analvte Buffalo River Saginaw Bav Indiana Harbor
Triethyiamine 7. JQ. 13.
PH 10.3 10.4 10.8
57
-------
SOLVENT EVAPORATION/PRODUCT OIL
Recovery of product oil (the organic compounds in the feed) is normally accomplished in two steps.
First, the bulk of the triethylamine is recovered by distillation. This is done by boiling the
triethylamine/oil mixture in a Bucni Rotovapor® apparatus. The oil remains in the boiling flask of
the Rotovapor while the triethylamine is condensed as it evaporates and is collected separately.
Second, any residual triethylamine is stripped from the oil by adding water to the hot oil in the boiling
flask of the Rotovapor. Water is added to form the low boiling triethylamine/water azeotrope. This
second step was not used for these samples due to the low oil content of the feeds. With a low oil
content feed, the amount of oil recovered is so small that effective stripping is not possible. The oil
from the sample was kept in a solution of triethylamine with only a fraction being completely dried of
solvent In this way, the oil remains homogeneous and can be poured out of the rotovapor flask. This
is vital for the integrity of the analyses on the oil, including PCBs from which a PCB mass balance is
determined. The analysis of each product oil follows:
• The metals composition was determined by nitric acid digestion after ashing at
550 degrees C, followed by ICP analysis (EPA SW846, Method 6010).
• The PCB concentration was determined by dilution of the oil in hexane, followed by EPA
SW846 (Test Methods for Evaluating Solid Waste). Method 3620. sulfuric acid and/or
Florisil column cleanup. The prepared sample was then analyzed by EPA SW846, Method
8080.
The oil in a triethylamine diluent was analyzed and the results converted to a pure oil (triethylamine
free) basis. The results were as follows:
Product Oil Analysis, dry basis
Analvte Buffalo River Saginaw Bay Indiana Harbor
PCBs. mg/kg 4.0 1.600 160
Metals, mg/kg
Antimony < 15. < 20. < 4.
Arsenic < 40. <50. < 10.
Barium 3. 15. 2.
Cadmium 3. 4. <0.6
Chromium 5. 58. 50.
Copper 110. 3.800. 12.
Lead 20. 290. 6.
Nickel 20. 1.400. 7.
Selenium < 25. < 30. < 6.
Silver < 0.8 39. 0.2
Sodium 400. 22,000. 44.
Zinc 19. 100. 5.
58
-------
m. MASS BALANCES
The data gathered during the bench-scale treatability test provides the data required to calculate mass
balances. The mass balances have been segregated into four groups: solids, oil, water, and PCBs.
SOLIDS MASS BALANCE
The mass balance for solids is a comparison of the solids input during the test to the solids recovered
after the test The mass of solids input during the test includes the solids portion of the feed extracted
and the solids portion of caustic soda added. The solids portion of the feed extracted was calculated
by multiplying the weight of feed extracted by the solids content as determined by analysis. The
solids portion of the caustic soda added was calculated by multiplying the weight of the 50 percent
NaOH solution added by 0.50.
The mass of the solids recovered from the test is equivalent to the sum of the product solids and
samples taken for stage-by-stage assays. A summary of this data follows:
Solids Mass Balance
Total Feed Extracted. Wei Basis
Solids Portion of Feed
Solids Portion of Caustic
Total Calculated Solids Input
Buffalo
900
530
+ 4
= 534
River
g
g
•1 g
g
Saginaw
900
690
+ 4.1
= 694
Bav
g
g
g
g
Indiana Harbor
1,400 g
584 g
+ 9.5 g
= 594 g
Weight of Product
Solids Recovered
Weight of Solids
Samples Recovered
461 g
+ 57 g
666 g
+ 75 g
485 g
+ 26 g
Total Solids Recovered
= 518 g
= 681 g
= 511 g
Recovery, %
97
59
98
86
-------
OIL MASS BALANCE
The oil mass balance was computed using the same method used in calculating the solids mass
balance. The oil & grease content of each feed was determined by extracting a sample of the feed
with methylene chloride. This oil &. grease content (by MeCl2) was multiplied by the weight of the
feed input to determine the amount of oil input. The mass of oil recovered from the test was
equivalent to the product oil recovered. The residual oil in the product solids and product water was
negligible when calculating an oil mass balance.
The oil mass balances (based on methylene chloride) were as follows:
Oil Mass Balance
Buffalo River
Saginaw Bay
Indiana Harbor
Calculated
Oil Input
4.30 g
3.24 g
33.6 g
Equivalent Product
Oil Recovered
4.83 g
4.44 g
32.5 g
% Recovery
112 %
137 %
97 %
Virtually all of the PCBs from the sample now reside in the product oil. For each sample, the weight
of PCB contaminated material was reduced from 900 grams (1400 grams for the Indiana Harbor
sample) to 3.4-30 grams corresponding to a 47-270 times reduction in mass.
WATER MASS BALANCE
The water mass balance was computed similarly to the method used for solids. The mass of water
input came from the water in the feed, plus the water introduced with the caustic. The water portion
of each feed was calculated by multiplying the weight of the feed by the water content as determined
by analysis. The water portion of the caustic input was calculated by multiplying the weight of the 50
percent NaOH solution by half.
The mass of water recovered was equivalent to the sum of the decant water, residual water in the
decant triethylamine/oil, the water contained in the rag layer, and residual water in the subsequent
extraction extracts. A summary of this data follows:
60
-------
Water Mass Balance
Buffalo River Saginaw Bay Indiana Harbor
Water Portion of Feed 365 g 207 g 784 g
Water Portion of Caustic + 4.0 g + 4.1 g + 9.5 g
Total calculated water input = 369 g = 211 g = 793 g
Water recovered from
decant water 258 g 172 g 598 g
Total water recovered = 258 g = 172 g = 598 g
% Recovery 70 82 75
The recovery of water was low. The temperature tends to increase above the triethylamine/water
miscibility limit when the treated solids arc centrifuged. At these conditions, some water may have
exited the centrifuge with the solids. This water was lost when the solids were dried. In addition, a
portion of the water in the feed was left behind in the resin kettle after decantation of the first
extraction since it is not possible to decant all the solvent from the solids. This water was lost when
the solids were dried. This portion of the water lost in the dryer is not accounted for in the water mass
balance. (In RCC's Pilot Unit, and Full-Scale Unit, all such water is recovered from the dryer.)
PCB MASS BALANCE
The PCB mass balance was computed similarly to the method used for oil. The mass of PCBs input
was calculated by multiplying the weight of each feed by the PCB concentration as determined by
analysis. The PCBs recovered from the test reside in the product oil. The PCBs in the product solids,
product water and recovered triethylamine were negligible when calculating a PCB mass balance.
The mass of PCBs recovered in the oil was calculated by multiplying the weight of oil recovered by
the PCB concentration as determined by analysis. The PCB mass balance for each sample was as
follows:
61
-------
PCB Balance
Calculated
PCBs Input
0.32 mg
14.2 mg
12.8 mg
Calculated
PCBs Recovered
0.22 mg
40.0 mg
8.3 mg
Total PCB
% Recovery
70 %
280 %
64 %
Sample
Buffalo River
Saginaw Bay
Indiana Harbor
SUMMARY OF MASS BALANCE CALCULATIONS
The following table summarizes the mass balance calculations for each of the constituents considered.
The mass balances were based on the amount of the fraction recovered from the simulation divided by
the calculated input amount to the simulation.
Mass Balance Summary, %
Sample
Buffalo River
Saginaw Bay
Indiana Harbor
Solids
97
98
86
on
112
137
97
Water
70
82
75
PCBs
70
280
64
62
-------
IV. CONCLUSIONS
The PCS-contaminated sediment samples finom tbe Buffalo River, Saginaw Bay and Indiana Harbor
Sites are suitable for treatment with the B.E.S.T. solvent extraction process. No problems were
observed during testing of the samples. Consequently, full-scale processing should be
straightforward.
1. The samples were chemically compatible with triethylamine.
2. The total PCB concentrations in the samples tested, 'Buffalo River', 'Saginaw Bay' and
'Indiana Harbor' were 0.60,21 said 22 mg/kg, respectively.
3. After treatment, the PCB residual removal efficiencies were > 95% for the Buffalo River
sample and 99% for the Saginaw Bay and Indiana Harbor samples.
4. The PAH residual concentration in the treated product solids were < 0.2 mg/kg for the
Buffalo River and Saginaw Bay samples and ranged from < 1 to < 3 mgAg for the
Indiana Harbor sample.
5. All three treated solids readily passed the TCLP Toxicity Test for leaching of metals.
6. Virtually all of the PCBs from the samples have been concentrated into the product oils.
For each sample the weight of PCB contaminated material was reduced 47-270 times.
*****
63
-------
APPENDIX B
B.E.S.T.® BENCH SCALE TREATABILITY TEST PLAN
The original Plan contained proprietary information. RCC has removed that
Information from this copy. For this reason, this copy contains blackened-
out areas.
February, 1990
R«qit»f«d in ff» OS. P«»ne Otloi
64
-------
1.0 TEST OBJECTIVES
The test objectives are:
o To achieve good separation of each type of sample's phase components (oil, water,
solids) with low thethylamine residuals in each.
o To record observations and data that will allow us to predict how a full-scale B.E.S.T.
separation of the samples might proceed.
o Take samples during the extraction tests and conduct analysis sufficient to allow for
calculation of mass balances for oil. water, solids and other compounds of interest
o To calculate the extraction efficiency of compounds of interest (i.e., PCB's if present)
achieved during the bench-scale workups in order to determine the number of
stages appropriate for full scale treatment of the site materials.
Evaluation of attainment of these objectives will consist of analysis of the feed and products and
observation of the bench-scale simulation in action.
2.0 TEST PLAN
The following tasks will be performed on the feed sample:
2.1 Characterize the Feed Sample
o Phase compositional analysis (oil. water, solids)
o Raw feed metals composition
o Analysis of other compounds of interest (i.e., PCB's if present)
2.2 Perform Preliminary Tests
o Sample pH adjustment characteristics
o TEA/Feed compatibility study
2.3 Conduct a B.E^.T. Bench-scale Treatabillty Test Wortcup
2.4 Feed Compositional Analysis
2.4.1 Analyze the feed for percent oil, water, and solids.
Determine the total oil per Standard Methods for the Examination of Water and Wastewater.
16th Edition, Method 503D, with two exceptions; extend the extraction time from 4 to 16 hours
and substitute methylene chloride for Freon based on RCC experience that methylene chloride
is a superior solvent for oils and greases.
Determine the total water content by Karl Rsher titration. (The water content is generally not
determined by a simple oven test since most oils are relatively volatile.. The oven test is used if
the ratio of water to oil present in the sample is relatively large.)
65
-------
Determine the paniculate solids content by rinsing a known quantity of feed through a Whatman
GF/C fitter under vacuum with acetone followed by methylene chloride, drying and weighing the
residue.
Normalize the oil + water + solids concentration to 100% if it is 90-110%. Repeat the assays to
determine the source of the error(s) if the oil + water + solids concentration is less than 90% or
greater than 110%.
3.0 TEST PROCEDURE
3.1 Feed Characterization
In conjuction with the performance of a B.E.S.T. laboratory-simulation, a data set must to be
developed for the sample. The data required includes compositional (oil/water/solids), metals
analysis, as well as other compounds of interest.
a. Determine the total oil per Standard Methods for the Examination of Water and
Wastewater. 16th Edition, Method 503D, with two exceptions; extend the extraction
time from 4 to 16 hours and substitute methylene chloride for Freon based on RCC
experience that methylene chloride is a superior solvent for oils and greases.
b. Determine the total water content by Karl Fisher titration. (The water content is not
typically determined by a simple oven test since most oils are relatively volatile. An
oven test will be used to cross-check the results of the Karl Fisher titration if the ratio
of water to oil present in the sample is relatively large.)
c. Determine the paniculate solids content by rinsing a known quantity of feed through
a Whatman GF/C filter under vacuum with acetone followed by methylene chloride,
drying and weighing the residue.
d. Normalize the oil + water * solids concentration to 100% if it is 90-110%. Repeat the
assays to determine the source of the error(s) if the oil + water + solids concentration
is less than 90% or greater than 110%.
3.1.1 Feed Sample Metals Composition
Dry approximately 10-20 gms of the feed in the oven at 105°C. Use a ceramic crucible for this
task. Dry for at least 6 hours. Record the initial sample weight.
After drying at 105°C put the sample into the muffle furnace and ash at 550°C for at least 3
hours. Once finished, let the sample cool to room temperature. Record the final sample (ash)
weight.
Digest the ash with nitric acid as a prelude to analysis of metal concentrations by ICP.
66
-------
3.1.2. TCLP Extraction Metals (SVf 846 Method 1311)
Perform preliminary extraction on a representative portion of each sample to determine the
appropriate extraction fluid. After determining the correct extraction fluid, perform 18 hour
TCLP extraction on a portion of representative sample.
After filtering the leachate, perform metals analysis using SW846 Method 6010.
3.1.2 PCB Analysis
PCB analysis, if present, will be conducted in accordance with SW846 method 8080. Soxhlet
extraction will be used (method 3540). The Aroclor type found in each sample should be
recorded.
3.2 Preliminary Testing
TEA can be ionized at some pH conditions. In the ionic form TEA is non-volatile and will not be
recovered from the process product phase fractions. To determine the proper pH control
requirements for each sample, a pH adjustment test is conducted.
TEA has the potential to react with some rare types of samples. To determine if this will pose a
problem during the study a compatibility study will also be performed.
3.2.1 Adjustment Characteristics
Measure the sample pH using pH paper if the sample is mostly solid, and pH probe if it is
mostly liquid. If neither seem to be providing a good measurement (i.e., the pH paper color
cannot be read or the pH probe readings keep drifting), then add 100 mis of distilled water to
about 5 gms of sample, stir well, and remeasure pH.
Obtain 5-10 gms of sample and adjust pH with 5% NaOH solution. If the pH could not be
measured as discussed above, be sure to add distilled water prior to the pH adjustment
measurements.
After the sample pH has been adjusted to 11 or above, record the amount of caustic spent.
Cover the sample with parafilm and leave mixing overnight. On the next day, check the pH
again and adjust with caustic soda if needed. Record all amounts of caustic spent.
3.2.2 TEA/Feed Compatibility Test
Mix about 5 gms of each type of sample with 50 mis of cold TEA. Make observations about the
ability of the TEA to dissolve the sample.
Use a thermometer to measure the amount of temperature change when the sample is mixed
with TEA.
Note any effervescence that takes place such as the formation of hydrogen. Note any unusual
reactivity. If there is any effervescence or unusual reactivity, notify the Lab Director before
proceeding.
67
-------
3.3 B.E.S.T. Bench-scale Treatablllty Test Workup
3.3.1 Pre-Treatment and First Wash
Adjust feed pH as determined previously.
Chill sample to below 40° F. Chill TEA to below 38°F.
Mix^^grams of the sample with^^tres of chilled TEA _
for ^Pninutes with the pneumatic mixer in the chiller bath (at < 40°F).
At the end of the first mixing stage, remove the particulates with the appropriate method
determined during the optimization testing.
Decant and retain supernatant/centrate/filtrate. Keep chilled at < 40°F until heating and
decantation can be performed.
Place the centrifuge cake or filter cake, if collected, back into the extraction beaker for
additional wash stages.
3.3.2 Second Wash
Mix recovered first extraction stage solids wfflf^iitres of fresh TEA. Fqrthe second and
TEA to transfer solids from the centrifuge botties (if used) into the mixing container. Keep the
mixture heated while mixing is in progress.
Mix forJFminutes with the pneumatic mixer.
Perform particle removal from extraction mixture again as performed for the first stage
extraction. If desired, collect a portion of the solids for later analysis.
3.3.3 Third Wash and Solids Drying
Repeat section 3.32 as a third wash.
Dry the solids at 220°F (105°C) in the forced draft oven. Mix occasionally to facilitate TEA
volatilization.
After the initial drying, add a portion of de-ionized water adequate to thoroughly wet the solids,
then redry in order to further reduce residual TEA concentrations. To insure that the TEA
residual in the dried solids will be low. treat the solids with caustic soda (applied with the de-
ionized water) if the pH of these solids is less than 10. Add sufficient caustic soda to raise the
pH to approximately 10.5. Determine the required amount of caustic soda on a small portion of
the solids.
NOTE: Upon occasion, additional washes may be required to achieve required treatment
standards.
68
-------
3.3.4 Decantatlon
Use the large separator/ funnel that is in the temperature controlled bath. Keep the water bath
at 140°F.
Heat supematant/centrate/filtrate from first wash (chilled to this point at 40°F) up to 140°F with
continuous mixing on a hot plate. Pour into separatory funnel.
Allow 30 minutes quiescent residence time in the separatory funnel. Decant only if the layers
appear to have stopped separating. If separation is still proceeding, wait until a better
separation is achieved, and then decant
Record observations of the speed of separation and measure 'rag' volume. Rag layer should
ultimately be centrifuged and the resulting TEA/oil and water layers should be added to the
appropriate decanted fraction pnor to the stripping operation.
Record the weights of the TEA/oil, rag, and water fractions from the decantation.
3.3.5 Distillations
Water layer
Record initial water pH (should be >11).
Steam strip the water at 110°C (in the rotovap) until no TEA odor is detected in collected
distilling drops. Record the initial water volume.
At this point, check water pH. If the pH is >10. distill for 15 minutes more and then terminate
distillation.
If water pH is <10, adjust to >12 and continue distillation until no TEA is detected in the
collected distillate. Check the latter every 15 minutes, then recneck water pH. If pH is still
below 10, then repeat this section (d.) until water pH is >10.
Oil/TEA laver
Remove the bulk of TEA without steam at 110°C (in rotovap). Record initial and final oilfTEA
volume.
Steam strip the oil at 110°C. Perform this operation by adding a known quantity of water
(typically 5 mis) and then measuring the volume of distillate recovered. When all TEA is
removed, the recovery of the distillate should be equal to the amount of water added.
Perform oil polishing by distilling without steam until the oil temperature reaches 120°C to
facilitate excess water removal. Record the final product oil weight
Be sure to collect some of the final distillate for measurement of the extent of any volatile
organic carryover.
69
-------
4.0 ANALYTICAL REQUIREMENTS
4.1 Feed Analysts
o Composition - % (wt) oil, water and solids by:
105°C total solid
Oil by Soxhlet extraction (by dichloromethane)
Solids by difference
o Total metals
Dryat105°C
Ash at 550°C
Nitric acid digest for ICP heavy metals determination
o Physical properties:
pH
Specific Gravity
o pH Adjustment:
Amount of caustic added to pH adjust the feed to pH of 11
o PCB's by Method 8080, if present
e TCLP Mctola Dtfraetien Analysis
4.2 Product Solids Analysis
o Residual Oil and Grease by Soxhlet extraction
o Aqua regia digest for ICP total trace metals
o TCLP Extraction Metals Analysis
o PCB's. if present
o Residual TEA
4.3 Decant Water Analysis
o PCB's, if present and if sample size permits
o Oil & Grease by freon extraction (IR if volume is limited)
o Total metals
4.4 Oil Analysis
o PCB's. if present
70
-------
5.0 SAFETY CONSIDERATIONS
Because of the unknown reactivity of these samples, extreme care should be taken not to allow
a dangerous situation to develop. Results of the compatfciitty study should be reviewed with
the Lab Director prior to initiation of the full scale bench-scale extraction. The bulk of these
simulations will be done in the laboratory hood to decrease TEA emissions. Prior to testing, the
feed will be subjected to TEA-compatibility tests to verify that it does not react violently with
TEA. Other safety precautions involving personnel conducting B.ES.T. laboratory work that will
be followed are described in RCC's Laboratory Safety Manual. If you have any other safety
related concerns, contact the Lab Director.
NOTE: Be sure to check that the ventilation system is working properly and wear
appropriate protective equipment when handling these samples.
71
-------
APPENDIX C
QUALITY ASSURANCE PROJECT PLAN
FOR
GLNPO - ASSESSMENT AND REMEDIATION OF
CONTAMINATED SEDIMENT TECHNOLOGY
DEMONSTRATION SUPPORT
Revision
February 15, 1991
Submitted to:
U.S. Environmental Protection Agency
Great Lakes National Program Office
230 S. Dearborn
Chicago, Illinois 60604
Submitted by:
Science Applications International Corporation
635 West Seventh Street, Suite 403
Cincinnati, Ohio 45203
EPA Contract No. 68-C8-0061, Work Assignment No. 2-18
SAJC Project No. 1-832-03-207-50
72
-------
GLNPO - QAPjP
Section No.: Q_
Revision No.: 2
Date: Feh 15. 1991
Page: 1 of 2
TABLE OF CONTENTS
SECTION
1.0 INTRODUCTION
2.0 PROJECT DESCRIPTION
3.0 QUALITY ASSURANCE OBJECTIVES . . .
4.0 SAMPLE TRANSFER AND PREPARATION
PROCEDURES
5.0 ANALYTICAL PROCEDURES AND
CALIBRATION
6.0 DATA REDUCTION, VALIDATION AND
REPORTING
7.0 INTERNAL QUALITY CONTROL CHECKS
8.0 PERFORMANCE SYSTEMS AUDITS
9.0 CALCULATION OF DATA QUALITY
IMPUCATORS
10.0 CORRECTIVE ACTION
11.0 QA/QC REPORTS TO MANAGEMENT . . .
APPENDIX A - TECHNOLOGY SUMMARIES
2
12
2
REVISION DATE
1
2
1
2
2
1
1
1
1
1/9/91
2/15/91
2/15/91
1/9/91
2/15/91
1/9/91
2/15/91
2/15/91
1/9/91
1/9/91
1/9/91
1/9/91
73
-------
QUALITY ASSURANCE PROJECT PLAN APPROVALS
QA Project Plan Title: GLNPO Assessment and Remediation of Contaminated
Sediment Technology Demonstration Suooort
Prepared by: Science Applications International Corporation (SAIC)
QA Project Category: II
Revision Dace: January 9. 1990
SAIC's WorK Assignment Manager (print)
Clyde J. Dial
SAlC's QA Manager (print)
Steve Yaks:en
3U;PO '-eric Group Chair (print)
Brian Schumacne"
Af.CS QA Officer (print)
3ene Easterly
£?A. EMSL-LV, NRD QA Officer (print,
Salon Chr-'szensen
I7A Tecnnicai Project Manager (print
Z'ave Cowo:" 1
AX.—> .-rogras Manager .print;
Signature
Signature
Signature
Signature
/Date
Date
Date
Data
-ate
-ate
74
-------
DISTRIBUTION LIST:
Gene Easterly
Brian Schumacher
Tony Kizlauskas
Thomas Wagner
Clyde Dial
Steve Garbaciak
Dennis Timberlake
Steve Yaksich
David Cowgill
Gary Baker
Vic Engleman
U.S. EPA, EMSL (Las Vegas)
LOCKHEED (Las Vegas)
SAIC (Chicago)
SAIC (Cincinnati)
SAIC (Cincinnati)
U.S. COE (Chicago)
U.S. EPA, RREL (Cincinnati)
U.S. COE (Buffalo)
U.S. EPA, GLNPO (Chicago)
SAIC (Cincinnati)
SAIC (San Diego)
GLNPO - QAPjP
Section Nou £
Revision No.: 2
Date:
Page:
Feh IS 1991
2of 2
75
-------
GLNPO - QAPjP
Section Nou J_
Revision No.: 1
Date: Jan. 9. 1991
Page: 1 of 2
1.0 INTRODUCTION
The Great Lakes National Program Office (GLNPO) leads efforts to carry out the
provisions of Section 118 of the Clean Water Act (CWA) and to fulfill U.S. obligations
under the Great Lakes Water Quality Agreement (GLWQA) with Canada. Under Section
118(c)(3) of the CWA, GLNPO is responsible for undertaking a 5-year study and
demonstration program for contaminated sediments. Five areas are specified for priority
consideration in locating and conducting demonstration projects: Saginaw Bay, Michigan;
Sheboygan Harbor, Wisconsin; Grand Calumet River, Indiana (aka: Indiana Harbor);
Ashtabula River, Ohio; and Buffalo River, New York. In response, GLNPO has initiated
an Assessment and Remediation of Contaminated Sediments (ARCS) Program. The ARCS
Program will be carried out through a management structure including a Management
Advisory Committee consisting of public interest, Federal and State agency representatives,
an Activities Integration Committee which is made up of the chairpersons of the technical
work groups, and technical work groups.
In order to obtain the broadest possible information base on which to make
decisions, the ARCS Program will conduct bench-scale and pilot-scale demonstrations and
utilize opportunities afforded by contaminated sediment remedial activities by others, such
as the Corps of Engineers and the Superfund program, to evaluate the effectiveness of those
activities. These bench-scale and pilot-scale tests will be developed and conducted under
the guidance of the Engineering/Technology (ET) Work Group for ARCS.
SAJC has been contracted to supply technical support to the ET Work Group. The
effort consists of conducting bench-scale treatability studies on designated sediments to
evaluate the removal of specific organic contaminants.
76
-------
GLNPO - QAPjP
Section No.: J_
Revision No.: 1_
Date: Jan. 9. 199L
Page: 2 of 2
Sediments have been obtained by GLNPO from various sites and represent the type
of material that would be obtained for onsite treatment. The primary contaminants of these
sediments are polychlorinated biphenyls (PCBs) and polynuclear aromatic hydrocarbons
(PAHs). Analyses to date show PCB concentrations are less than 50 ppm. These sediments
have been homogenized and packaged in smaller containers by EPA.
77
-------
GLNPO - QAPjP
Section Nou 2-
Revision No.: 2-
Date: Feb. 15. 1991
Page: 1 of 12
2.0 PROJECT DESCRIPTION
2.1 Background
SAIC and its subcontractors will conduct seven (7) bench-scale (several liters) tests
on wet contaminated sediments using four treatment technologies.
The seven treatability tests (as currently planned) will utilize sediments from 4 sites
(Saginaw River, Buffalo River, Indiana Harbor Canal, and Ashtabula River). Five
sediments have been collected from these sites by GLNPO. These samples have been
homogenized by the U.S. EPA and are being stored under refrigeration in 5 gallon
containers by EPA in Duluth, MN.
These five sediments are currently being analyzed in the U.S. EPA, Environmental
Research Laboratory in Duluth. The Duluth Laboratory is analyzing the sediments for total
organic carbon/total inorganic carbon (TOC/TIC), particle size, density of dry material,
total sulfur, acid volatile sulfide, oil and grease (O & G), total PCBs, PAHs (10), and metals
including mercury. Table 2-1 is a summary of the data received to date.
A portion (small vial) of each residual of each treatability test may be retained and
sent to the GLNPO office for "show" purposes. If available, sub-regulated quantities of the
solid and oil residuals from each test treatability study may also be retained and shipped to
EPA for possible further treatment studies.
The following is a list of technologies and the proposed number of sediment samples
to be tested by each technology:
a. B.E.S.T.™ Extraction Process on three samples (Buffalo River, Indiana
Harbor, Saginaw TRP 6)
b. Low Temperature Stripping (RETEC) on one sample (Ashtabula River)
c. Wet Air Oxidation (Zimpro Passavant) on one sample (Indiana Harbor)
d. Low Temperature Stripping (Soil Tech) on two samples (Buffalo River and
Indiana Harbor)
Summaries of these technologies are included in Appendix A.
78
-------
TABLE 2-la. Preliminary Analytical Results on ARCS Sediments
Description
Saginaw 221
Saginaw TRP6
Ashlabula River
/
Indiana Harbor
Buffalo River
Concentration (Mg/kgm)(a) Concentration (%)(*)
Total Total
PCB PAH
06 1.2
6.0 3.1
C C
0.2 96
0.4 5.6
c» Cd Ni Fe(%) Cr Zn Pb TOC OAG Moisture (b)
33 0.9 76 1.4 NO 240 30 1.4 O.I 40.3
81 4.7 110 0.9 200 200 47 1.2 0.3 3|.|
55 3.0 96 3.7 550 240 48 2.6 1.7 52.9
320 9.4 150 16 540 3300 780 2t 5.8 61.0
85 1.9 57 3.9 NO 200 94 2.0 0.5 41.5
(D
(a) Concentration In ppm and dry weight basis unless otherwise Indicated.
(b) As received basis.
TABLE 2-lb. Preliminary Particle Size Distribution (%)
Description
Buffalo River
Particle Size (a)
>50u 50-20 u 20-5 u 5-2 u 2-0.2 u 0.2-0.08 u <0.08u
198 12.1 29.0 11.8 24.3 2.4 0.6
Median
Diameter, u
9.3
(a) u mlcarons
-------
GLNPO - QAPjP
Section No J 2.
Revision No.: 2
Date: Feb. IS. 1991
Page: 3 of 12
22. Testing Program for Chemical Characterization
SAIC shall be primarily responsible for the physical and chemical characterization
of both the sediment samples prior to testing and the residuals created during the tests.
Analyses conducted by the vendors or subcontractors will not be depended on, but such data
shall be reported whenever available.
Two different sets of chemical analyses will be conducted during the performance of
the treatability tests: optimization test analyses and performance evaluation analysis. The
Phase I optimization test analyses will be conducted by the subcontractor or vendor during
the series of initial technology tests. The Phase II performance evaluation analyses will be
conducted by SAIC (or its analytical subcontractor) on the raw sediment sample prior to the
treatability test run at optimum conditions and on the end products produced by that
particular test. These tests are described further in this section.
In order to assure objectivity and consistency of data obtained from multiple vendors
running different technology tests, SAIC shall conduct analyses as described in Table 2-2 for
characterization of the sediments and the end products of the treatability tests at optimum
conditions (Phase II).
The analyses described for the solid fraction in Table 2-2 shall be performed by
SAIC's analytical subcontractor once on a subsample taken from each sample sent to each
vendor or subcontractor for treatability tests (Phase II). This subsample will be taken at the
same time that the sample for the Phase II treatablility study is taken by the vendor. This
data will serve as the measure of the raw sediment quality for comparison to analyses of
treated end products from each technology test that may be conducted on sediments from
a particular area of concern.
Each bench-scale technology test may actually involve the performance of multiple
laboratory simulations. During the initial tests (Phase I), any analyses performed by the
80
-------
GLNPO - QAPjP
Section No.: 2.
Revision NOJ 2-
Date: fph, 15 1991
Paige: 4 of 12
vendor or subcontractor shall be reported, as available. For the tests run at optimum
conditions (Phase II), SAIC shall conduct the full suite of analyses, as detailed in Table 2-2,
on the end products if sufficient quantities are produced by the technology. Quotes solicited
for each technology specified that a minimum 300 grams dry basis of treated solid had to
be produced for SAICs analyses. Table 2-3 shows the apportionment of the 300 grams for
the solid analyses. The quantity of water is depended on the sediments and the individual
technologies. To do all the analyses listed in Table 2-2, and associated QC, approximately
10 liters of water are required. Table 2-4 listed specified sample volumes for each analysis,
and gives a priority to each analysis. It is possible that only the PCB and PAH analysis and
associated QC will be performed on the water samples. If any oil residue is produced, it
will be analyzed by dilution with appropriate sample cleanup steps for PCBs and PAHs.
The data generated by SAICs analyses of the untreated sediment and the treated end
products from the test at optimum conditions will be primarily relied upon to determine
treatment efficiencies. Vendor- or subcontractor-generated data will not be relied upon but
shall be reported when available.
23 Required Permits
Because of the small quantities of sediments required for the bench-scale treatabiliry
tests, SAIC anticipates that no formal permits will be required to conduct these tests. If this
is not the case and permits (such as TSCA, RD&D or RCRA permits) are required, the
subcontractor will notify SAIC and the TPM will be notified to obtain approval for
acquisition of the permit(s).
All unused sediment samples requested by SAIC for the treatabiliry test and all
testing residuals, except those requested by the TPM for "show" purposes and those
requested by the TPM for possible further testing, will be properly disposed of per federal
and state regulations.
81
-------
GLNPO - QAPjP
Section No.: 2.
Revision No.:
Date: Feb. 15. 1991
Page: 5 of 12
TABLE 2-2
Parameters and Detection Limits for Analysis of ARCS Technologies
Parameter
TOC/TIC
Total Solids4
Volatile Solids4
Oil & Grease4
Total Cyanide
Total Phosphorus
Arsenic4
Barium4
Cadmium4
Chromium4
Copper4
Iron (total)4
Lead4
Manganese4
Mercury4
Nickel4
Selenium4
Silver4
Zinc4
PCBs (total & Aroclors)4
PAHs (16)4-5
PH
BOD5
Total Suspended Solids4
Conductivity
so*
300
1000
1000
10
0.5
50
0.1
0.2
0.4
0.7
0.6
0.7
5
02
0.1
2
02
0.7
0.2
0.02
0.2
full range
Water
1000
1000
1000
10
10
1
2
4
7
6
7
50
2
0.01
20
1
7
2
0.07
2
full range
1000
1000
full range
Qift
0.1
0.1
NOTES:
1 Detection limits for solids are ppm (mg/kgjgijLweigiit). The D.L.'s for metals should
be obtainable by ICP except for As, Se, and Hg. If GFAA is used, the D.L-'s will be
2 mg/kgm except Hg, Cd, and Ag which will be 0.1 mg/kgm.
- Detection limits for water are ppb (ug/1). The D.LJs for metals should be obtainable
by ICP except for As, Se, Hg. If GFAA is used D.L.'s will be 1 ug/L except Hg
which will be 0.01 ug/L.
3 Detection limits for oil are ppm (mg/1).
4 Parameters tentatively identified for QC analyses.
5 Polynuclear aromatic hydrocarbons to be analyzed are the 16 compounds listed in
Table 5-2.
82
-------
GLNPO - QAPjP
Section No.: 2.
Revision No.:
Date:
Page:
2
Feh 15 1991
6 of 12
TABLE 2-3
Solid Sample Quantities for Analyses
Parameter
TOC/TIC
Total + Volatile Solids
Oil & Grease
Total Cyanide
Total Phosphorous
Metals (except Hg)
Hg
PCBs + PAHs
PH
Subtotals
Reserve
TOTAL
Initial
Sample (g)
15
5
20
10
5
5
1
30
20
111
—
—
OC (tf
10
40
—
_
15
3
90(60)3
—
158(128)
—
—
Total (g)
15
15
60
10
5
20
4
90
20
269(239)
31(61)
300
QC Approach
None1
Triplicate/Control
Triplicate/Control
None2
None2
MS/Triplicate
MS/Triplicate
(3)
None4
1 For sample set II that does not have such a limited quantity of solid, The QC described in
footnote 3 will be implemented.
2 For sample set n, MS/triplicate QC will be implemented.
3 Quality control for untreated solids is Triplicate and spike and for treated solids matrix spike
and matrix spike duplicate.
4 For sample set n, Triplicate/Control sample QC will be implemented. The control sample
may be an EPA QC check sample, an NBS - SRM, a standard laboratory reference solution,
or other certified reference material.
83
-------
GLNPO - QAPjP
Section No.: 2
Revision No.: 2 .
Date: Fftt? 15 199!
Page: 7 ^f 12
TABLE 2-4
Sample Volumes Required and Priority Ranking for Water Analyses
Parameter
TOC/TIC
Volatile Solids
Oil & Grease
Total Cyanide
Total Phosphorus
Arsenic
Barium
Cadmium
Chromium
Copper
Iron (total)
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Zinc
PCBs (total & Aroclors)
PAHs (16)
PH
BOD
Total Suspended Solids
Conductivity
Priority
1
5
6
1
7
4
2
2
2
2
2
2
2
3
">
4
->
9
1
1
7
7
5
7
Analysis
Volume, ml
25
d
1000
500
50
100
100
b
b
b
b
b
b
100
b
c
b
b
1,000
a
25
1,000
200
100
QC
Volume, ml
_
d
2000
—
—
300
300
b
b
b
b
b
b
300
b
c
b
b
2.000
a
—
—
400
QC
Approach
None (e)
Triplicate/Control
Triplicate/Control
None (f)
None (f)
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/Triplicate
MS/MSD
MS/MSD
None (f)
None (f)
Triplicate/Control
None (f)
Note:
a) same aliquot as PCBs
b) same aliquot as Barium
c) same aliquot as Arsenic
d) same aliquot as Total Suspended Solids
e) see footnote 2, Table 2-3
f) see footnote 4, Table 2-3
84
-------
GLNPO - QAPjP
Section NOJ 2_
Revision Nou 2-
Date: Feb. 15. 1991
Page: 8 of 12
2.4 Purpose of Phase I Experimental Design
The purpose of the Phase I technology experimental design is for each subcontractor
to establish a range of variables best suited for feasibly implementing their technology on
a full-scale basis (Phase II). SAIC will send a quantity (specified by the vendor) of each
sediment to the vendor to accomplish this. All data generated by the vendor during Phase
I will be supplied to SAIC for inclusion in the report for that technology. This information
will include the operating conditions/parameters, the input/output data for the contaminants
of interest to show the range of effectiveness associated with various operating conditions,
and the quantities of the input material and the various residuals resulting from the test.
The optimum set of conditions to be used for Phase II will be reported to SAIC along with
appropriate revisions to the Phase I experimental design to make it applicable to Phase II.
2.5 Purpose of Phase n Treatahilitv Test
SAIC will send another container of sediment(s) to the vendor (quantity to be
specified by the vendor). This container will not be opened until a representative of SAIC
arrives for the scheduled treatability test(s). Other observers from U.S. EPA, COE and/or
the GLNPO may also be present during the Phase II treatability test(s).
The new sample will be homogenized and a sample equivalent to a minimum of 300
gm of dry solids will be set aside for characterization analyses (Table 2-2) by SAIC. SAIC
will observe the treatability tests and obtain samples of process residuals for analyses (Table
2-2). The bench-scale test(s) must produce enough solid residual for all vendor
requirements and a quantity equivalent to 300 gm of dry solids for SAIC analyses. SAIC
can utilize up to 10 liters of water for analysis and 25 ml of the oil residual. The actual
quantities of water and oil that will be produced are dependent on the initial sediment and
the technology. All technologies except wet air oxidation are expected to produce an oil
residual. Also, if additional solid and/or oil residue is available, EPA may ask for these
materials to be sent to them for storage for possible future evaluation.
85
-------
GLNPO - QAPjP
Section NOJ 2-
Revision No.: 2.
Dale: Fffhi 15 1991
Page: 9 of 12
All data generated by the vendor during Phase n is to be supplied to SAIC for
inclusion in the report for that technology. The vendor must stipulate in their work plan,
prior to conducting the test(s), the process locations to be sampled, the frequency and the
information being obtained.
All other residuals from both phases of the treatability study, including any untreated
sediment, will be properly disposed of by the vendor.
SAIC shall oversee the treatability test assessment(s) by vendors or subcontractors,
including all QA/QC aspects, monitoring and analysis. SAIC shall ensure compliance with
the specific experimental design during the tests conducted by vendors or subcontractors.
SAIC will make specific notes regarding the equipment being used, any pretreatment of the
sediment(s), the operation of the equipment, and any post treatment of the residuals. SAIC
personnel will pack the untreated sediment sample and the end product samples from the
Phase II test for each technology in an appropriate fashion for shipment from the vendor
or subcontractor to the laboratory SAIC is using for the analysis. Proper chain-of-custody
procedures will be developed in the QAPjP and strictly followed by SAIC personnel.
SAIC plans to take photos of the equipment while at the vendor's location for
inclusion in the report.
SAIC shall perform limited interpretation of technology test results, specifically the
development of material and energy balances. No test of air or fugitive emissions will be
done. For material balances, estimates of the mass distribution of the analytes of interest
(Table 2-2) among the residuals will be made. The term energy balance is interpreted to
mean an estimation by the vendor of the energy input into the process at a pilot- or full-
scale.
86
-------
GLNPO - QAPjP
Section NOJ jj_
Revision No.: 2
Date: Fah, is 1991
Page: 10 nf 12
SAIC shall collect any information available from the vendor or subcontractor
concerning the actual or estimated costs of constructing and operating full-scale versions of
the technology tested.
The purpose of this project is to test five technologies for removing organic
contaminants (PCBs and PAHs) from sediments typical of locations around the Great Lakes.
GLNPO is specifying the technologies and the sediment(s) to be treated by each technology.
This study is only one pan of a much larger program, and it is not necessarily intended to
evaluate the complete treatment of these sediments. Other aspects or treatment options are
being evaluated by a number of agencies, contractors, etc.
Therefore, this study is based on the following assumptions:
• The percent removal of the PCBs and PAHs from the solid residual is the
most important object of this study.
• The untreated sediments and solid residuals are the most important matrices.
• If water and oil residuals are generated by a technology, the existence of an
appropriate treatment or disposal option for these residuals is assumed.
PAHs and PCBs will be determined in these residuals as a cross check of
their fate in treating the solids.
Based on the intents of this study, the critical measurements are PAHs, PCBs, metals,
total solids, volatile solids, and oil and grease in the untreated and treated solids.
2.6 Organization and Responsibilities
A project organization and authority chart is shown in Figure 2-1. The
Environmental Monitoring Systems Laboratory (EMSL) is cooperating with GLNPO and
SAIC on this evaluation. Mr. Thomas Wagner is the SAIC Work Assignment Manager and
is responsible for the technical and budgeting aspects of this work assignment. Mr. Clyde
Dial is QA Manager and is responsible for QA oversight on this work assignment.
87
-------
us i-PAi
I'lUHliCI MANAIililt
II S P.PA QA MANAOnil
-------
GLNPO - QAPjP
Section NOJ i.
Revision NoJ 2
Date: Feb. 15. 1991
Page: 12 of 12
2.7 Schedule
The Phase I experimental designs are scheduled for mid to late February 1990, and
the Phase El Treatability Tests are scheduled for March and April 1991.
89
-------
GLNPO - QAPjP
Section NOJ i.
Revision NOJ 2-
Date: Feb. 15. 1991
Page: 1 of 2
3.0 QUALITY ASSURANCE OBJECTIVES
3.1 Precision. Accuracy. Completeness, and Method Detection Limits
Objectives for accuracy, precision, method detection limits, and completeness for the
critical measurements of solids are listed in Table 3-1. Accuracy (as percent recovery) will
be determined from matrix spike recovery for PAHs, PCBs and metals, and from laboratory
control samples (certified reference material- CRM) for the remaining analyses. Precision
(as relative standard deviation) will be determined from the results of triplicate analyses for
PAHs, PCBs, solids (total, volatile and/or suspended), oil and grease, and metals. Matrix
spike and matrix spike duplicate analyses will be used for treated solids for PCBs and PAHs.
The completeness will be determined from the number of data meeting the criteria in Table
3-1 divided by the number of samples that undergo performance evaluation analyses.
3.2 Representativeness and Comparability
Representativeness and Comparability are qualitative parameters. The sediment
samples have already been collected and have been reported to be representative of the
areas to be remediated. The data obtained in this program will be comparable because all
the methods are taken from a standard EPA reference manual and all the analyses will be
conducted at the same laboratory. Reporting units for each analysis are specified in Section
6 of this document and are consistent with standard reporting units in this program.
3.3 Method Detection Limits
The target detection limits (TDLs) were specified by GLNPO (Table 2-2). Based on
the analytical methods appropriate for the analyses and the amount of samples specified in
the methods, the detection limits listed in Table 3-1 should be achievable. Generally the
instrument detection limits are'defined as 3 times the standard deviation of 15 blanks or
standards with a concentration within a factor of 10 of the IDL.
90
-------
TABLE 3-1. Quality Assurance Objectives Tor Critical Measurements
(Sediments and Treated Solids)
Parameter
Total Solids
Volatile Solids
Oil & Grease
Arsenic
Barium
Cadmium
Chromium
Copper
Iron (total)
lead
Manganese
Mercury
Nickel
Selenium
Silver
Zinc
PCBs (total
ft Aroclort (e)
PAIls (Table 5-2)
Method (a)
160.3
160.4
9071
3050/7060
3050/6010
3050/6010
3050/6010
3050/6010
3050/6010
3050/6010
3050/6010
7471
3050/6010
3050/7740
3050/6010
3050/6010
3540 or
3550/8080
3540 or 3550/
8270 or 8 100
Accuracy (b)
(as % recovery)
80-120
80-120
80-120
85-115
85-115
85-115
85-115
85-115
85-115
85-115
85-115
85-115
85-115
85-115
85-115
85-115
70-130
70-130
Precision (c)
%
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Method
Detection Limit (d)
(mg/kgm)
1000
1000
10
O.I
0.2
0.4
0.7
0.6
0.7
5
0.2
O.I
2
0.2
0.7
0.2
0.02
0.2
Completeness
%
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
90
00
90
90
(a) References are to "Methods for Chemical Analysis of Water and Wastes", EPA/600/4-79/020 or "Test Methods for
Evaluating Solid Waste", SW-846, 3rd. Ed.
(b) Determined from MS or MS/MSD analyses for metals, PAHs, and PCBs; others determined from
laboratory control samples.
(c) Determined as relative percent standard deviation of triplicate analyses, except PAIls and PCBs
in treated solids where MS/MSD will be used.
(d) See Footnotes I and 2 of Table 2-2
(e) Detection limits based on extraction of 30 gram samples.
p
-------
GLNPO - QAPjP
Section No_- 4_
Revision NoJ 1
Date: Jan. 9. 199L
Page: 1 of 4
4.0 SAMPLE TRANSFER AND PREPARATION PROCEDURES
As described in Section 2, SAIC will receive a number of 5 gallon containers of
previously homogenized sediments from the U. S. EPA in Duluth, Minnesota. The number
of containers of each sediment is dependent on the final determination by GLNPO of which
sediments will be tested by the various technologies. Only if smaller portions of sediments
are requested by the vendors will these containers be opened by SAIC. If smaller portions
are required, SAIC will resuspend the solids and water within an individual container by
rolling, tumbling, and stirring of the contents. The final stirring will be in the original
containers using a metal stirrer as would be used to mix a 5 gallon container of paint. The
metal stirrer is appropriate because metals are not the primary constituents of concern in
these treatabiliry tests.
The Chain of Custody Record shown in Figure 4-1 will be completed for each cooler
shipped to the subcontractor or vendor that will conduct the optimization and performance
evaluation tests. The samples obtained from the vendor for analysis will be labeled as
shown in Figure 4-2. The labels will document the sample I.D., time and date of collection,
and the location from where the sample was taken. The amount/type of preservative that
was added will also be recorded.
SAIC personnel will pack and ship the untreated sediment and the end product
samples (residuals) from the optimum conditions test for each technology. The amount of
preservative will be recorded. Samples will be labeled (see Figure 4-2) and shipped by
overnight delivery service to the laboratory in coolers containing ice. If "blue ice" is used
in the coolers, samples will be initially cooled with regular ice prior to being packed in the
coolers with blue ice. The Chain of Custody Record (Figure 4-1) will be completed for each
cooler shipped to the laboratory.
92
-------
GLNPO - QAPjP
Section NOJ ±
Revision No- J
Date:
Page:
Jan. 9 1991
2 of 4
Solid, sediment and oil samples require no preservative other than cooling to 4° C.
The appropriate types of containers (solid and liquids), holding times, and preservatives for
water samples are listed in Table 4-1.
TABLE 4-1. Sample Containers, Preservation and Holding Times
Parameter
TOC
Solids (Total,
Volatile &
Suspended
Oil and Grease
Total Cyanide
Container
P,G
P,G
G
P,G
Preservation of Water Samples
Cool 4° C, H2SO4 to pH < 2
Cool 4° C
Cool 4° C, H2SO4 to pH < 2
Cool 4° C, NaOH to pH > 12
Holding Time
28 days
7 days
28 days
14 davs
Total Phosphorous P,G
P,G
Metals
(except Cr VI)
Cr(VI)
P,G
PAHs & PCBs
BOD5
PH
Conductivity
G teflon
lined cap
P,G
P.G
P,G
0.6g Ascorbic acid
Cool 4° C, H2SO4 to pH < 2
HNO3 to pH < 2
Cool 4° C
Cool 4° C, store in dark
Cool 4° C
Cool 4° C
28 days
6 months except Hg
(Hg 28 days)
24 hours
Extract within 7 days
Analyze within 40 days
48 hours
Performed immediately
28 davs
93
-------
GLNPO - QAPjP
gffpnn NOJ ^
Revisioa No:
Date:
Page:
hn, 9 1991 _
* -
|
w
•Q o
O
ft
£ .
S £
o
M
3
J
O
1
5
j
0
75
C
If
• e r
i ° <2
ctenct Appllci
tomvNonal Co
n Empfcir** Ownr
ly
Z
1 .
jj
zo c
B
g
*
i
1 n
' M
g a
>U-
s
Phone Nunit
i!
UO2>-<_
|
«
riONS. COMMENTS. 1
L INSTRUCTIONS 1
OBSERVA
SPECIA
.ZUC 01
!
1 t
s
I -
t
, i
Sampler (Slgna
l*««wyN» |
1
i
I
1
5
g
£•
1
j
£•
f
1
1
|
*
i
i
i
<
• j
s
f
2
I;
i
i\
\
1
i
!
I
B
?
:
.
I
B
i
S
8J
III
\
\
\l
*\
fl
J
I
i
1
j
|jj
i
;
\l
\\
|
r
1
i
1
fi
it Q
I
; *
M i
: !
f i
-5 «
] {
I
£
c
I
c
fc
1
1
i
1
u!
in
i
i
8
i
|
!
1
i
i
I
M
S
3
i
j
I
<_
;
h
U
* £
l!
j
I
i
I
1
|
s
[
i
!
l
c
o
1
s
f
j
!
n
I
i
•
u
i
T
I
\
\
a
i
1
i
!
i
R
ol/oni In
I
;
j
(ft
94
-------
GLNPO - QAPjP
Section No.: 4 —
Revision NOJ I —_
Date: J«n 9 199L
Page: 4 of 4
535 W. 7th Street. Suite 403, Cincinnati, OH 45203
Sample No.:
Sample Location/Date/Time:
Project Location/No.:
Analysis:
Collection Method: Purge Volume:
Preservative:
Comments:
Collector's Initials
Figure 4-2. Example Sample Label
95
-------
GLNPO - QAFjP
Section NOJ £.
Revision NOJ 2
Date: Feb. 15. 199L
Page: 1 of 3
5.0 ANALYTICAL PROCEDURES AND CALIBRATION
Analytical procedures for all critical measurements are referenced in Table 3-1. The
non-critical measurements are for any residual water and oil remaining after the
performance evaluation tests and some additional analyses on the solid samples. The EPA
procedures are specified in Table 5-1.
The required calibration for all analyses are specified in the methods and will be
followed. All instruments will be calibrated as specified in the methods prior to performing
any analysis of the samples. Internal QC checks, including initial calibration and continuing
calibration checks, for the critical measurements are listed in Table 7-1.
Table 5-2 contains the minimum list of the sixteen PAHs that must be determined
by either analytical method. Additional compounds may be included, but none of these
sixteen may be deleted from the target list.
The laboratory is responsible for maintaining a preventive maintenance program
consistent with manufacturers recommendations for all instruments required for this
program. In addition, they are responsible for having a sufficient supply of routine spare
pans necessary for the operation of the analytical equipment in order to complete the
analysis in a timely fashion.
96
-------
GLNPO - QAPjP
Section No.: £
Revision No.: 2.
Date:
Paige:
Feb. 15. 1991
2 of 3
TABLE 5-1
Analytical Methods for Critical and Non-critical Measurements
Methods?
Parameter
Solid
Water
Oil
TOC
Total Solids
Volatile Solids
Oil and Grease
Total Cyanide
Total Phosphorous
Arsenic
Mercury
Selenium
Other Metals
PCBs
PAHs
pH
BOD
Total Suspended Solids
Conductivity
9060
1603
160.4
9071
9010
3652
3050/7060
7471
3050/7740
3050/6010
3540 or
3550/8080
3540 or 3550/
8270 or 8100b
9045
NA
NA
NA
9060
NA
160.4
413.1
9010
3652
7060
7470
7740
3010/6010 (7760 Ag)
3510 or
3520/8080
3510 or 3520/
8270 or 8100b
9040
405.1
160.2
9050
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
3580/8080
3580/8270
NA
NA
NA
NA
(a) References are to "Methods for Chemical Analysis of Water and Wastes", EPA/600/4-
79/020 or 'Test Methods for Evaluating Solid Waste", SW-846, 3rd. Ed.
(b) Where options for methods are given,-Either is acceptable if the detection limits given
in Table 2-2 can be achieved.
NA - Not analyzed
97
-------
GLNPO - QAPjP
Section No.: JL
Revision No.: 2
Date: Feh 15. 1991_
Page: 3 of 3
TABLE 5-2
List of PAHs'
Acenaphthene Chrysene
Acenaphthylene Dibenzo(a,h)anthracene
Anthracene Fluoranthene
Benzo(a)anthracene Fluorene
Benzo(a)pyrene Inden(lA3-cd)pyrene
Benzo(b)fluoranthene Naphthalene
Benzo(k)fluoranthene Phenanthrene
Benzo(ghi)perylene Pyrene
PAH analyses must determine these 16 compounds at a minimum.
98
-------
GLNPO - QAPjP
Section NOJ £_
Revision NOJ 1
Date: Jan. 9. 199L
Page: 1 of 1
6.0 DATA REDUCTION, VALIDATION AND REPORTING
Data will be reduced by the procedures specified in the methods and reported by the
laboratory in the units also specified in the methods. The work assignment manager or his
designer will review the results and compare the QC results with those listed in Table 3-1.
Any discrepancies will be discussed with the QA Manager.
All data will be reviewed to ensure that the correct codes and units have been
included. All organic and inorganic data for solids will be reported as mg/kgm except TOC,
oil & grease (O&G), moisture and iron that will be reported as percent and pH that will
be reported in standard pH units. All metals and organics in water samples will be reported
as ug/1. TOC, solids (suspended and volatile), O&G, cyanide, phosphorus, and BOD will
be reported as mg/1. Conductivity will be reported as umhos/cm and pH as standard pH
units. After reduction, data will be placed in tables or arrays and reviewed again for
anomalous values. Any inconsistencies discovered will be resolved immediately, if possible,
by seeking clarification from the sample collection personnel responsible for data collection,
and/or the analytical laboratory.
Data Tables in the report will be delivered in hard copy and on discs. The discs will
be either in Lotus files or WordPerfect 5.1 files.
99
-------
GLNPO - QAPjP
Section NOJ 2
Revision No_- 2.
Date:
Page:
peb. 15. 199L
1 of 7 __
7.0 INTERNAL QUALITY CONTROL CHECKS
The internal QC checks appropriate for the measurement methods to be utilized for
this project are summarized in Table 7-1. These items are taken from the methods and the
QC program outlined in Section 3 of this QAPjP.
For the GLNPO program, the following QC measures and limits are employed:
on-going calibration
checks
method blanks
matrix spikes
replicates
beginning, middle, and end of sample set for metals, pH,
TOC/TIC, total cyanide, and total P
mid-calibration range standard
± 10% limit unless otherwise stated
± 0.1 pH unit for pH
± 10 umhos/cm for conductivity at 25° C
beginning, every 12, and end of sample set for PCBs and
PAHs
mid calibration range standard
± 10% limit
one per sample set for PCBs and PAHs
< MDL limit unless otherwise stated
beginning, middle and end for metals, TOC/TIC, total
P, total cyanide, and pH
beginning, middle and end for conductivity with
acceptance limits of < 1 umho/cm
one per sample set
1 to 1.5 times the estimated concentration of sample
± 15% limit for metals; ± 30% for PCBs and PAHs
triplicate analyses
RSD s 20% unless otherwise stated
one per sample set
± 0.1 pH unit for pH
± 2 umhos/cm for conductivity
100
-------
GLNPO - QAPjP
Section No- J
Revision No.: 2
Date: Feb. IS 1991
Page: 2 of 7
QC sample - - minimum of one per sample set
(CRM) - ± 20% of known CRM
- ± 0.1 pH unit for pH
- ± 1 umhos/cm for conductivity
surrogate spikes - added to each sample
(PCBs and PAHs only) - ± 30% recovery
The surrogate for PCB analysis is tetrachlorometaxylene and the internal standard is 1,2,3-
trichlorobenzene.
Table 7-2 shows an analytical matrix that will be completed for each technology
tested. For example, consider the case of a bench scale treatability test of (1 kilogram)
Indiana harbor sediment by low temperature stripping. Based on the data presented in
Table 2-la and assuming complete separation and recovery of oil, water, and solid, a 1
kilogram sample of untreated sediment will produce 58 grams of oil, 610 ml of water, and
332 grams of dry treated solids. For the purpose of this program, this sample set consists
of 1 untreated solid, 1 treated solid, and the water and oil generated by the process. Table
7-3 is a completed analytical matrix for this test. Table 7-3 is based on Tables 2-2 and 2-4
and the QC approach described in this QA plan. The analysis of the water sample in this
example is severely limited by the relatively small amount of sample obtained.
Table 7-4 is a matrix summarizing the anticipated samples to be analyzed for this
project. The sets for each technology (see section 2.1) are:
I B.E.S.T.
D ReTec
III Wet Air Oxidation
IV Soil Tech
The Soil Tech process will process treated soils at two distinct points. Therefore,
four treated solids are produced from the two untreated sediments.
101
-------
TABLE 7-1. Internal QC Checks for Measurements
o
ro
Parameter
Solids
(Total &
Volatile
Oil & Grease
Metals
MelaU
PCBs (b)
PAHs
Method (a)
160.3
160.4
9071
6010
7000
series
8080
8270 or
8100
Initial
Calibration
Balance
(Yearly)
See Above
2 points
4 points
5 points
5 points
Calibration
Checks
Balance
Each Day
See Above
Every 10th
Sample
Every 10th
Sample
Every 10th
Sample
Every 12
Hours
Method
Blank
Yes
Yes
Yes
Ye«
Yes
Yes
MS/MSD
NA
NA
MS only
MS only
Yes (treated)
MS only (untreated)
Yes (treated)
MS only (untreated)
Triplicate
Sample
Analysis
Yes
Yes
Yes
Yes
NA (treated)
Yes (untreated)
NA (treated)
Yes (untreated)
QC
Sample
Yes
Yes
Yes
Yes
Yes
Yes
Surrogate
Spikes
NA
NA
NA
NA
Yes
Yes
(a) References are to "Methods for Chemical Analysis of Water and Wastes*, EPA/600/4-79/020
or "Test Methods for Evaluating Solid Waste*, SW-846, 3rd. Ed.
(b) Second column confirmation of positive results is required.
NA - Not Applicable
R
8
O
-------
TABLE 7-1. Internal QC Checks for Measurements (continued)
Parameter
PH
Conductivity
Cyanide
Phosphorous
TOC/TIC
Method (a)
9045/9040
9050
9010
365.2
9060
Initial
Calibration
1 points
1 point
7 points
9 points
3 points
Calibration
Chech
Every 10th
Sample
Every 15th
Sample
Every 15th
Sample
Every 15th
Sample
Every 15th
Sample
Method
Blank
NA
NA
Yes
Yes
Yes
MS/MSD
NA
NA
NA
NA
NA
Triplicate
Sample
Analysis
NA
NA
NA
NA
NA
QC
Sample
Yes
Yes
Yes
Yes
Yes
Surrogate
Spikes
NA
NA
NA
NA
NA
(a) References are to "Methods for Chemical Analysis of Water and Wastes", EPA/600/4-79/020
or "Test Methods for Evaluating Solid Waste". SW-846, 3rd. Ed.
NA - Not Applicable
S» -4
-------
TABI.K7-2. Analytical Matrix
o
-P-
QC Sample
*nd
lelhod Bltnk
SO!
'
-------
TABLE 7-3. Example
o
en
QC Simple
tnd
Method Blank
8
-------
TABLE 7-4. Analytical and QC Sample Matrix for GLNPO Trealability Studies (numbers of samples)
SAMPLE SET
SET I
Untreated S.
Treated S.
Water
Oil
SETW
Untreated S.
Treated S.
Water
Oil
SETH
Untreated S.
Treated S.
Water
Oil
SETHI
Untreated S.
Treated S.
Water
TOTALS
Solida
Water
Oil
roc/r/c
S<») QC(b)
3
3
2
4
1
I
1
1
I
16
-
1
3
2
3
-
5
3
TOTAL
SOLIDS
S QC
3
2
4
I
1
1
1
16
2
3
2
3
2
3
2
20
VOL
SOLIDS
S QC
3
2
4
1
1
1
I
16
2
3
2
3
2
3
3
2
20
3
OAO
S QC
3
2
4
1
1
1
1
16
2
3
2
3
2
3
3
2
20
3
TOTAL
YANIDB
S QC
3
2
4
1
1
1
1
16
1
-
-
3
3
3
-
6
3
TOTAL
PHOS
S QC
3
2
4
1
1
1
1
16
1
-
-
3
3
3
.
6
3
METALS
S QC
3
2
4
1
1
1
1
1
16
1
3
3
3
3
3
3
3
3
24
PCBt
S QC
3
3
3
2
4
2
2
I
1
1
I
1
1
1
16
6
2
1
3
3
2
I
3
3
2
2
3
3
2
2
20
9
PAH
S QC
3
3
3
2
4
2
2
I
1
1
I
1
1
1
16
6
2
1
3
3
2
1
3
3
2
2
a
3
2
2
20
9
PH
S QC
3
2
4
I
1
1
1
16
1
-
-
3
2
3
-
S
3
BOD
S QC
-
-
1
-
1
-
-
3
-
3
TSS
S QC
-
-
1
-
1
-
-
3
-
3
COND
S QC
-
-
1
-
1
-
-
3
-
3
(a) Number of original lamplea.
(b) Number of quality control (ampler A *3* represent* two additional repllcatei (triplicate determination) and • ipilce or control
lample analysia resulting in an additional three QC analyses. A *2* represents matrix spike/malrix spike duplicate analysis
scheme resulting in an additional two QC inilysei. A * I * indicates a blank spike or other control sample analysis resulting
in one additional QC analysis.
(c) Treated and untreated solidf does not apply, and only one control sample per set will be analysed.
-------
GLNPO - QAPjP
Section NOJ £_
Revision NOJ JJ_
Dare: Feb. 15. 1991
Page: 1 of 1
8.0 PERFORMANCE AND SYSTEM AUDITS
The laboratory will perform internal reviews by the QA officer or a designee. These
reviews should include, as a minimum, periodic checks on the analysts to assess whether they
are aware of and are implementing the QA requirements specified in the ARCS QA
program.
The laboratory will be prepared to participate in a systems audit to be conducted by
the SAIC QA Officer or his designee and/or ARCS QA Officer.
The vendors of the various technologies have all been advised that a number of
representatives from SAIC, GLNPO, and other organizations will be present during
Phase II of the treatability studies. Thus the ARCS QA officer can be present during
Phase II of any or all of the treatability studies.
107
-------
GLNPO - QAPJP
Section NOJ 2_
Revision No- 1
Date: Jan. 9. 1991
Page: 1 of 3
9.0 CALCULATION OF DATA QUALITY INDICATORS
9.1 Accuracy
Accuracy for PAHs, PCB and metals will be determined as the percent recovery of
matrix spike samples. The percent recovery is calculated according to the following
equation:
% R = 100% xC' ~£?
Q
where
%R = percent recovery
C, = measured concentration in spiked sample aliquot
C0 = measured concentration in unspiked sample aliquot
C, = actual concentration for spike added
Accuracy for the other critical measurements will be determined from laboratory
control samples according to the equation:
%R = 100% :2L
Q
where
%R = percent recovery
Cm = measured concentration of standard reference material
C, = actual concentration for standard reference material
9.2 Precision
Precision will be determined from the difference of percent recovery values of MS
and MSDs for PAHs and PCBs or triplicate laboratory analyses. The following equations
will be used for all parameters:
108
-------
GLNPO - QAPjP
Section No.: 5 .
Revision NOJ 1
Date: /an 9. 1991
Page: 2 of 3
When 2 values are available:
[Q - CJ x 100%
[Q + CJ/2
where
RPD = Relative percent difference
Cj = The larger of two observed values
C = The smaller of the two observed values
When more than 2 values are available:
N
I
i = l
N
2 _
x i
N i
N - 1
where
S = standard deviation
X., = individual measurement result
N = number of measurements
Relative standard deviation may also be reported.
will be calculated as follows:
RSD = 100 5
If so, 11
109
-------
GLNPO - QAPjP
Section NOJ JL
Revision No_- 1
Date: Jan. 9. 199L
Page: 3 of 3
where
RSD = relative standard deviation, expressed in percent
.5 = standard deviation
X = arithmetic mean of replicate measurement.
9.3 Completeness
Completeness will be calculated as the percent of valid data points obtained from the
total number of samples obtained.
% Completeness = VDP x 100
TOP
where
VDP = number of valid data points
TOP = total number of samples obtained.
110
-------
GLNPO - QAPJP
Section No.: 10
Revision No_- 1
Date: Jan. 9. 1991
Page: 1 of 2
10.0 CORRECTIVE ACTION
Corrective actions will be initiated whenever quality control limits (e.g., calibration
acceptance criteria) or QA objectives (e.g., precision, as determined by analysis of duplicate
matrix spike samples) for a particular type of critical measurement are not being met.
Corrective actions may result from any of the following functions:
• Data Review
• Performance evaluation audits
• Technical systems audits
• Interlaboratory/interfield comparison studies
AJ1 corrective action procedures consist of six elements:
• Recognition that a Quality Problem exists
• Identification of the cause of the problem
• Determination of the appropriate corrective action
• Implementation of the corrective action
• Verification of the corrective action
• Documentation of the corrective action
For these treatability studies after initial recognition of a data quality problem, the
data calculation will be checked first. If an error is found, the data will be recalculated and
no further action will be taken. If no calculation error is found, further investigation will
be conducted. Depending on the cause and the availability of the appropriate samples.
reanalysis or flagging of the original data will be utilized.
111
-------
GLNPO - QAPjP
Section No_- 10
Revision No.: 1
Date: Jan. 9. 199L
Page: 2 of 2
All corrective action initiations, resolutions, etc. will be implemented immediately and
will be reported in Sections One and Two (Difficulties Encountered and Corrective Actions
Taken, respectively) in the existing monthly progress reporting mechanisms established
between SAIC, EPA-RREL, GLNPO, AND THE ARCS QA officer and in the QA section
of the final report. The QA Manager will determine if a correction action has resolved the
QC problem.
112
-------
GLNPO - QAPjP
Section NOJ 11
Revision
Date: Jan. 9. 1991
Page: 1 of 1
11.0 QA/QC REPORTS TO MANAGEMENT
This section describes the periodic reporting mechanism, reporting frequencies, and
the final project report which will be used to keep project management personnel informed
of sampling and analytical progress, critical measurement systems performance, identified
problem conditions, corrective actions, and up-to-date results of QA/QC assessments. As
a minimum, the reports will include, when applicable:
• Changes to the QA Project Plan, if any.
• Limitations or constraints on the applicability of the data, if any.
• The status of QA/QC programs, accomplishments and corrective actions.
• Assessment of data quality in terms of precision, accuracy, completeness,
method detection limit, representativeness, and comparability.
• The final report shall include a separate QA section that summarizes the data
quality indicators that document the QA/QC activities that lend support to
the credibility of the data and the validity of the conclusions.
For convenience, any QA/QC reporting will be incorporated into the already well-
established monthly progress reporting system between SAIC and EPA-RREL for all TESC
Work Assignments. In addition, copies of monthly reports will be sent to the ARCS QA
officer. Any information pertaining to the above-listed categories will be reported under
Sections One through Three (Difficulties Encountered, Corrective Actions Taken, and
Current Activities, respectively) in the monthly reports.
113
-------
GLNPO - QAPjP
Section NOJ Appendix A
Revision No.: 1
Date: Jan. 9. 1991
Page: 1 of 3
APPENDIX A
TECHNOLOGY SUMMARIES
114
-------
GLNPO - QAPjP
Sf-ftio" No- Appendix A
Revision NOJ 1
Dais: Jan. 9. 1991
Page:: 2 of 3
B.E.S.T.™ Process Description
The B.E.S.T.™ process is a patented solvent extraction technology utilizing triethylamine
as the solvent Triethylamine is an aliphatic amine that is produced by reacting ethyl
alcohol and ammonia. The key to success of the B.E.S.T.™ process is triethylamine's
property of inverse miscibility. At temperatures below 65°F, trietihylamine is completely
soluble with water. Above this temperature, triethylamine and water are only partially
miscible. The property of inverse miscibility can be utilized since cold triethylamine can
simultaneously solvate oil and water.
The B.E.S.T.™ process produces a single phase extraction solution which is a homogeneous
mixture of triethylamine and the water and oil (containing the organic contaminants, such
as PCBs, PNAs, and VOCs) present in the feed material. In cases where the extraction
efficiencies of other solvent extraction systems are hindered by emulsions, which have the
effect of partially occluding the solute (oil containing the organic contaminants),
triethylamine can achieve intimate contact at nearly ambient temperatures and pressures.
This allows the B.E.S.T.™ process to handle feed mixtures with high water content without
penalty in extraction efficiency. This process is expected to yield solid, water, and oil
residuals.
Low Temperature Stripping
Low-temperature stripping (LTS) is a means to physically separate volatile and semivolatile
contaminants from soil, sediments, sludges, and filter cakes. For wastes containing up to
10% organics or less, LTS can be used alone for site remediation.
LTS is applicable to organic wastes and generally is not used for treating inorganics and
metals. The technology heats contaminated media to temperatures between 200-1000°F,
driving off water and volatile contaminants. Offgases may be burned in an afterburner,
condensed to reduce the volume to be disposed, or captured by carbon adsorption beds.
For these treatability studies, only processes that capture the contaminants driven off will
115
-------
GLNPO - QAPjP
$f»/t'f>f No- /Appendix A
Revision No.: 1
Date: Jan. 9. 1991.
Page: rlnf3
be considered. The process (for these treatability studies) is expected to yield solid, water,
and oil residuals.
Wet Air Oxidation
Wet air oxidation is a process that accomplishes an aqueous phase oxidation of organic or
inorganic substances at elevated temperatures and pressures. The usual temperature range
varies from approximately 350 to 600°F (175 to 320°C). System pressures of 300 psig to well
over 300 psig may be required. However, testing has been done at temperatures exceeding
the critical point for water to limit the amount of evaporation of water, depending on the
desired reaction temperature. Compressed air or pure oxygen is the source of oxygen that
serves as the oxidizing agent in the wet air oxidation process. This process is expected to
yield only solid and water residuals.
116
-------
APPENDIX 0
B.E.S.T.® TREATABIUTY STUDY
ANALYTICAL METHODS
Analytical Methods Used In B.E.S.T.® Workuos
Feed Analysis
Moisture:
Oil & Grease:
Ash Content:
Metals Analysis:
Polynudear Aromatic
Hydrocarbons*:
Karl Fisher titration or Gravimetric @ 105°C, 16 hrs
Both EPA/SW846 Method 9071 and Methylene Chloride Soxhiet
gravimetric (16 hr. extraction)
Gravimetric @ 550°C for 16 hrs.
Digestion: EPA SW846/3050 or, alternatively, ash digestion at
550°C, followed by heating with nitric acid
Analysis: EPASW846/6010
EPA SW846/8100 (Methylene Chloride Extraction)
Benzene, Toluene, Xylene*: EPA SW846/8020
pCB*: EPA SW846/8080, Method 3540 extraction (soxhlet
extraction with 1:1 acetonerhexane for 16 hours)
Product Water
Total Solids:
Total Dissolved Solids:
Total Organic Carbon:
Oil & Grease:
Triethylamme:
Metals:
Product Analysis
Standard methods 209A
Standard methods 209B
EPA 600/415.2
EPA 600/413.1 or 413.2 depending on level of sample and
quantity of sample available
See the attached RCC method
Digestion: SW846/3005
Analysis: SW846/6010
• If present
117
-------
GC METHODS FOR TRIETHYLAMINE (TEA) ANALYSIS
IN AQUEOUS SOLUTIONS, SOLIDS AND OILS
I Summary of Methods
Triethyiamine (TEA) can be determined using Gas Chromatography with flame
ionization detection. Aqueous solutions can be injected directly onto a packed
column after pH adjustment and filtration. Solids are extracted into water without pH
adjustment and then are analyzed using the same column and parameters as
aqueous solutions. Oils are dissolved in a solvent (typically methylene chloride) and
analyzed, using a megabore HP-1 Methyl Siiicone column.
II TEA In Aqueous Solutions
A. Equipment and Operating Parameters
1. Gas Chromatograph: Hewlett Packard 5890A with 3392A Integrator
2. Column: 4% Carbowax-20M, 0.8% KOH, 60/80 Carbopack B
3. Injector Temp.: 200°C
4. Detector: Flame Ionization Detector (FID), set at 300°C
5. Oven Temperature and Time:
Initial Temp: 90°C, Initial Time: 0 minutes
Final Temp: 170°C, Final Time: 30 minutes
Rate: 5°/min.
6. Column Flow: ~30ml/min.
7. TEA Peak Retention Time: Approximately 8 minutes
B. Procedure
1. Standardization
a) Inject I microliter of 73 mg/1 TEA. This standard is prepared by
serial dilution from pure TEA (successive 1:100 dilutions from pure
TEA, which is 730,000 mg/l TEA).
b) Use the peak area at approximately 8 min. retention time to
quantitate TEA. Repetitive standards injections should agree to
within 10%.
118
-------
Product Oil
Triethylamine:
Viscosity:
Water:
Suspended Solids:
Metals:
See the attached RCC method
Brookfield
Karl Fisher titration
Filtration/Gravimetric (Whatman GF/C)
Product oil is diluted 1:10 with Xylene and filtered through
GF/C filter, then analyzed with organometallic standards on
ICP or, alternatively, ash digestion at 550°C, followed by
heating with nitric acid
Product Solids
Residual Triethylamine:
Oil & Grease:
Metals:
TCLP:
Polynudear Aromatic
Hydrocarbons':
Benzene, Toluene, Xylene*:
PCS*:
See the attached RCC method
Both SW846/9071 and Methylene Chloride Soxhlet
gravimetric (16 hr extractions)
Digestion: 1 gm sample refluxed with 15 mis Aqua Regia
Analysis: SW846/6010
SW846/1311 followed by 3010 & 6010
EPA SW846/8100 (Methylene Chloride Extraction)
EPA SW846/8020
EPA SW846/8080, Method 3540 extraction (soxhiet
extraction with 1:1 acetone:hexane for 16 hours)
* If present
119
-------
2. Sample Preparation and Analysis
a) If necessary, dilute the sample in deionized. distilled water until the
TEA concentration is at or below 73 ppm and record the dilution
factor.
b) Inject 1 microliter sample.
c) Inject a standard at least once every 10 samples and at the end of
an analytical sequence.
3. Quantitation
Quantitate the TEA using direct comparison of peak area
Peak Area of Standard = (Peak Area of Sample)
mg/l Standard (mg/l of Sample) x (dllutlon
III TEA In Solids
A. Equipment and Operating Parameters
The same equipment and parameters are used as in aqueous solutions.
Additional equipment includes the following:
1. Shaker Bath (example: Forma Scientific model 2564).
2. 50 ml Erienmeyer flasks.
3. Cover with Parafilm.
B. Procedure
1. Standardization: asinll.B.1.
2. Sample preparation and analysis.
a. Weigh out 3-5 g solids into an Erienmeyer flask. Record exact
weight.
b. Add 25 ml distilled water.
c. Cover with 1 layer of parafilm.
120
-------
d. In the shaker bath, shake vigorously at ambient temperature for 1
hour.
e. Let the mixture stand quiescent allowing the solids to settle.
f. Continue as in II.B.2.
3. Quantitation
Quantitation is the same as that for aqueous solutions, correcting for the
extraction of the solids into the water.
Peak Area of Standard _ Peak Area of Extract
TEA in Standard jig/g TEA of Extract
ng/gm TEA in Solids = (jig/ml TEA in Extract) x (mis of Extraction Water)
(g of Solids)
IV TEA In Oil
A. Equipment and Operating Parameters
1. Gas Chromatograph: Hewlett Packard 5890 with 3392A Integrator
2. Column: Megabore 15m x .53mm J&W BD1 Methyl Silicone
3. Injector Temp: 200°C
4. Detector: Flame lonization Detector (FID), set at 300°C
5. Oven Temperature and Time:
Initial Temp: 35°C. Initial Time: 12 minutes
Final Temp: 250°C, Final Time: 20 minutes
Rate: 25°C/min.
6. Column Flow: 2 ml/mm.
7. Makeup Gas: 20 ml/mm.
8. TEA Peak Retention Time: Approximately 7-8 min.
121
-------
B. Procedure
1. Standardization
a) Inject 1 microliter 73 mg/l TEA (dissolved in GC grade methyiene
chloride). This standard is prepared by serial dilution from pure
TEA (successive 1:100 dilutions into methyiene chloride from pure
TEA, which is 730,000 mg/l TEA).
b) Use the peak area at approximately 7-8 minutes retention time to
quantitate TEA. Repetitive standard injections should agree to
within 10%.
2. Sample Preparation and Analysis
a) Dissolve the oil in methyiene chloride such that the TEA
concentration is at or below 73 ppm.
b) Inject 1 microliter sample.
c) Inject a standard at least once every 10 samples.
3. Quantitation
Quantitate the TEA using direct comparison of peak area.
Peak Area of Standard m Peak Area of Sample
mg/l of Standard mg/l of Sample
122
-------
ro
co
SAIC-GLNPO (CF #361)
CONVENTIONALS IN UNTREATED SEDIMENT
B.E.S.T.
REVISED
2/18/92
% Total OH & Grease
MSLCode
MDL
Sponsor ID
% Moisture
0.01%
PH
NA
Volatile Solids
0.001%
(mg/kg)
0.1
TCC
% weight
0.007%
Total Cyanide
(mg/kg)
0.001
Total Phosphorus
(mg P/kg)
0.001
361-5. Rep 1
361-5, Rep 2
361-5. Rep 3
361-6
361-7
Method Blank
S-US-RCC, Rep 1
S-US-RCC. Rep 2
S-US-RCC, Rep 3
BUS-RCC
I-US-RCC
STANDARD REFERENCE MATERIAL
MESS-1 SRM
In-house Concensus Value '
MATRIX SPIKE RESULTS
Amount Spiked
Sample
Sample + Spike
Amount Recovered
Percent Recovery
REPLICATE ANALYSES
361-5, Rep 1
361-5, Rep 2
361-5. Rep 3
S-US-RCC, Rep 1
S-US-RCC. Rep 2
S-US-RCC. Rep 3
RSD%
23.98%
NA
NA
41.06%
57.04%
NA
7.30
NA
NA
7.29
7.35
NA
2.24%
2.03%
2.01%
4.03%
14.2%
0%
1485
1318
1256
2415
32165
0.1
0.83%
NA
NA
1.98%
17.03%
0.0117.
4.05
NA
NA
2.05
25.17
0.001 U
23.98%
NA
NA
41.96%
57.04%
NA
NA
NA
NA
NA
NA
NA
NA
23.98%
NA
NA
NA
7.30
NA
NA
7.29
7.35
NA
NA
NA
NA
NA
NA
NA
NA
7.30
NA
NA
NA
NA
NA
NA
NA
2.45
2.3
NA
NA
23.98%
NA
NA
NA
7.30
NA
NA
NA
2.24%
2.03%
2.01%
6%
1485
1318
1256
9%
0.63%
NA
NA
NA
4.05
NA
NA
NA
NA - Not analyzed
U = Below detection limit
- TOC value lor MESS determined based on past In-house analyses. Not a statistical determination.
NOTE: All Conventional results are reported on a dry weight basis.
742
NA
NA
735
2625
0.089
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
361-5
4305
1485
6383
4698
114%
NA
NA
NA
NA
NA
361-5
267.10
4.05
251.90
247.85
93%
361-6
2419
735
3088
2353
97%
§
m
742
NA
NA
NA
-------
SAIC-GLNPO (CF (361)
METALS IN UNTREATED SEDIMENT
(Concentrations
MSI Code
MX
361-5. Rep 1
361 5, Rep 2
381-5. Rep 3
361-6
361-7
In ug/g dry weight)
Sponsor ID
S-US-RCC. Rep 1
S US RCC. Rep 2
S-US HCC. Rep 3
BUS RCC
1 US RCC
Ag
»•
0007
0.87
0 85
0.80
0 31
4 84
As
MF
2 5
2.5 U
265
25 U
127
22.8
Ba
**
43
322
322
321
413
317
Cd
»A
0006
4.38
4.08
3.96
2.10
8.56
Cr
**
33
03
112
117
100
2270
B.E.S.T.
Cu
WF
5.6
55.8
67.3
63.4
70.2
188
%Fe
WT
0.26
0.780
0.765
0.616
4.200
18.770
HB
CVAA
0.0003
0.162
0.178
0.160
O.SS1
1.528
Mi
XV
66
162
ISO
173
667
3230
Nl
XRF
75
58. 1
55.2
61.5
43.1
12.0
Pb
XT
6 2
43.0
42.5
500
101.0
582 0
REVISED
2/21/02
Se
A*
022
0.22 U
0.22 U
0.22 U
0.74
0.22 U
Zn
Hf
It
125.7
132.6
162.1
180.3
2380
Method Blank
0016
NA
NA 0006
NA
NA
NA 0.00066
NA
NA
NA 0.22 U
NA
STANDARD REFERENCE MATERIAL
1646 SRM
certified
value
MATRIX SPIKE RESULTS
Amount Spiked
361-5*
361-5 + Spike
Amount Recovered
Percent Recovery
REPLICATE ANALYSES
361-5. Rep 1 S US RCC. Rep 1
361-5, Rep 2 S US RCC. Rep 2
381-5. Rep 3 S-US-RCC. Rep 3
RSD%
0.126
rC
rC
2
084
333
240
125%
087
085
080
4%
11.3
11 6
11 3
NS
NS
NS
NS
NS
1.86
265
2 12
18%
J - Values delected below MDL.
U - Below detection limit
NA - Not analyzed/applicable
- Mean ol triplicated sample
NS - Not spiked
NOTE: AH Metals results are 'blank corrected "
406
NC
N3
NS
NS
NS
NS
NS
322
322
321
0%
0.42
0.36
±007
2
4 15
643
228
114%
4.30
408
3.06
5%
78
76
±3
N3
NS
NS
NS
NS
03
112
117
12%
106
18
±3
NS
NS
NS
NS
NS
55.8
57.3
63.4
7%
336
3.35
±01
NS
NS
NS
NS
NS
0.780
0.785
0.816
3%
0.066
0.063
±0012
1.072
0.166
2.207
2.041
103%
0.162
0.178
0.160
e%
339
376
±20
NS
NS
NS
N3
NS
162
150
173
4%
31.1
32
13
NS
NS
NS
NS
NS
68.1
55.2
61.5
B%
28.1
28 2
±18
NS
NS
NS
NS
NS
43.0
42.5
50.0
10%
0.75
NC
tc
2.72
022 U
3.10
3.10
114%
0.22 U
0.22 U
022U
NA
126.3
138
±6
NS
NS
NS
NS
NS
125.7
132.6
162.1
14%
-------
SAIC-GLNPO (CF #361)
CONVENTIONALS IN TREATED SEDIMENT
B.E.S.T.
REVISED
2/18/92
% Total Oil ft Grease
MSLCode
MDL
Sponsor ID
% Moisture
0.01%
PH
NA
Volatile Solids
0.001%
(mg/kg)
0.1
TOO
% weight
0.007%
Total Cyanide
(mg/kg)
0.001
Total Phosphorus
(mg P/kg)
0.001
O1
361-8, Rep 1
361-8, Rep 2
361-8, Rep 3
361-9
361-10
Method Blank
S-TS-RCC, Rep 1
S-TS-RCC, Rep 2
S-TS-RCC. Rep 3
B-TS-RCC
I TS-RCC
STANDARD REFERENCE MATERIAL
MESS-1 SRM
In-house Concensus Value *
0.16%
NA
NA
3.72%
0.50%
NA
10.73
NA
NA
10.30
10.25
NA
1.76%
1.70%
1.74%
3.91%
9.06%
NA
297
293
206
238
470
0.1
0.58%
NA
NA
1.21%
13.36%
0.011%
50.72
NA
NA
8.11
6B.5
0.001 U
NA
NA
NA
NA
NA
NA
NA
NA
2.45
2.3
NA
NA
63
NA
NA
20
7214
0.089
NA
NA
REPLICATE ANALYSES
381-8. Rep 1
361-8, Rep 2
361-8, Rep 3
S-TS-RCC, Rep 1
S-TS-RCC, Rep 2
S-TS-RCC, Rep 3
RSD%
0.16%
NA
NA
NA
10.73
NA
NA
NA
1.76%
1.70%
1.74%
2%
297
293
206
19%
0.58%
NA
NA
NA
50.72
NA
NA
NA
NA • Not analyzed
U - Below detection limit
• TOC value for MESS determined based on past In-house analyses.
* - Mean for replicated sample.
NOTE: All Conventional results are reported on a dry weight basis.
Not a statistical determination.
63
NA
NA
NA
-------
SAIC GLNPO (CF 1361)
METALS IN TREATED SEDIMENT
(Concentrations
MSLCode
MDL
361 -8. Rep 1
361 8. Rep 2
381 8. Rep 3
361 9
361-10
In ug/g dry weight)
Sponsor ID
S-TS-RCC. Rep 1
S TS-RCC. Rep 2
S-TS-RCC. Rep 3
BTSRCC
1- TS-RCC
AQ
AA
0007
1 03
076
067
024
434
As
»T
2.5
2.77
267
292
14 6
290
Ba
*r
43
325
321
310
396
290
Cd
AA
0.006
4.38
4 17
424
2.11
697
Cr
*t
33
130
113
110
113
1706
B.E.S.T.
Cu
XT
5.5
686
66.3
57.4
61 2
223
%Fe
*t
0.26
0.855
0627
0 795
4.42
825
HJ
CVAA
0.0003
0.505
0.290
0 209
0.627
1.458
Ml
HT
56
181
177
172
684
2540
Nl
wr
7.5
71.5
64 1
572
42 1
10 U
Pb
•Hf
6 2
533
450
41.4
101 5
6560
REVISED
2/5/92
Se
AA
0.22
0.22 U
022U
022U
0.87
4 94
Zn
we
78
194.0
185.1
147.0
189.7
2810.0
Method Blank
0020
NA
NA 0.006
NA
NA
NA 000013
NA
NA
NA
0.22 U
NA
to
en
STANDARD REFERENCE MATERIAL
1646 SRM
ctrlllUd
v»lut
MATRIX SPIKE RESULTS
Amount Spiked
361-8*
361-8 * Spike
Amount Recovered
Percent Recovery
REPLICATE ANALYSES
361 -B. Rep 1 S-TS HOC. Rep 1
361-8. Rep 2 S-TS RCC, Rep 2
361-8. Rep 3 S-TS-RCC. Rep 3
RSO%
0 113
rC
N3
2
0 82
3 22
2 4
120%
1 03
0.76
0 67
23%
13 1
11 6
±1.3
NS
NS
NS
NS
NS
2 77
2 87
292
3V.
J . Values delected below MDL.
U - Below detection limit
NA - Not analyzed/applicable
. Mean ol triplicated sample
NS . Not spiked
NOTE. All Metals results aie 'blank corrected.'
411
N3
tc
NS
NS
MS
NS
NS
325
321
310
2%
042
036
±007
2
4.26
6.25
1 99
100%
4.36
4 17
4.24
3%
65
76
±3
NS
NS
NS
NS
NS
130
113
110
m.
22 3
IB
13
NS
NS
NS
NS
NS
68.6
66.3
57.4
B%
347
3.35
±0.1
NS
NS
NS
NS
NS
0.655
0.827
0.795
4%
0.066
0.063
±0012
1.987
0.335
2.282
1.947
98%
O.S05
0.290
0.209
46%
350
375
±20
NS
NS
NS
NS
NS
161
177
172
3%
38.1
32
±3
NS
NS
NS
NS
NS
71.5
64.1
57.2
11%
286
282
tt a
NS
NS
NS
NS
NS
533
450
41.4
13%
074
NO
NO
2 74
022U
2.99
2.99
109%
0.22 U
0.22 U
022U
NA
133.4
136
±6
NS
NS
NS
NS
NS
194.0
165 1
147.0
14%
-------
to
-vl
&/MU ULNh-U (Lit- »3B1)
PAH IN UNTREATED SEDIMENT
Low Molecular Weight PAHs (ng/g dry wl )
B.E.5.T.
REVISED
2/9/92
Naphthalene Acenaphthytene Acenapnttiene Fluorene Phananlhrene Anthraon*
MSL Code Sponsor ID
3615, Rep 1 R S US RCC, Rep 1
3615. Rep 2 R S US RCC. Rep 2
361-5. Rep 3 R S US RCC. Rep 3
361-6 BUS RCC
361-7 1 US RCC
Method Blank 3
Method Blank-R
STANDARD REFERENCE MATERIAL
SRM NIST1941
ccrlllUd v»lu»
MATRIX SPIKE RESULTS
Amount Spiked
361 5 i ~K -
361 5 • Spike R
Amount Recovered
Percent Recovery
26 B
24 B
27B
107B
4402 B
set
1 1
364
NC
933
26
406
380
41%
15
14 U
18
79 U
2322
58 U
1 1 U
54
NS
933
V6- <\%
577
RW 51T
66H loi'/»
19
20 U
22 U
113 U
4396
83 U
16 U
60U
NC
933
£9
-------
to
CO
SAIC GLNPO (CF »361)
PAH IN UNTREATED SEDIMENT
High Molecular Welqhl PAHs (nq/q dry wl )
MSLCode Sponsor ID
381-5. Rep 1 R S US RCC. Rep 1
361-5. Rep 2 R S US RCC, Rep 2
361-5. Rep 3 R S US RCC, Rep 3
361-6 BUS RCC
361-7 1 US RCC
Method Blank 3
Method Blank- R
STANDARD REFERENCE MATERIAL
SRMNIST1041
cerlllled value
Amount Spiked __
381 5 • \ •»
361-5 + Spike R
Amount Recovered
Percent Recovery
Fhjotarv
Ihene
376 B
464 B
351 B
1197 B
32032 B
44
0
1114
1220
933
307
1075
678
73%
Pyrene
425 B
491 B
402 B
1157 B
32022 B
39
9
1034
1080
933
439
1077
637
68%
Benzo(a)-
anlhracene
184
210
163
861
18282
37U
5U
461
550
933
186
019
733
79%
Chrysene
285 B
299 B
242 B
1037
24399
36 U
S
703
NC
033
269
056
686
74%
Beruo(b)-
tuoranlhene
238 B
270 B
217B
878
19187
27U
6
766
760
033
242
1044
802
86%
B.E.S.T.
Benio (k)-
fluorantiene
180 B
200 B
158 B
733
13350
23 U
5
603
444
033
170
648
669
72%
Indeno
Dlbenzo
REVISED
2/9/02
Benio(a)- (l.2.3.c.d) (a.h)amhra- Benio(g.h.l)-
pyrene
224 B
250 B
200 B
887
20581
SOU
5
500
670
033
225
087
762
82%
pyrene
201 B
231 B
186 B
607
14653
27U
5
408
569
933
207
1095
888
9S%
cene
44
47
39
205
5224
35U
4 U
141
NC
033
43
012
660
03%
113B
120 B
100 B
405
13767
26 U
6
421
516
033
117
660
543
56%
R - Re-extracted sample results.
• - Mean ol klpllcaled sample
B - Analyle detected In Blank associated with sample.
U - Below detection limits
NC - Nol collided
* - Value outside ol Internal QC limits (40 120%)
-------
SAIC OLNPO (CF §361)
PAH IN UNTREATED SEDIMENT
B.E.S.T.
REVISED
2/5/03
1 Surrogate Recovery %
MSLCoda
361-5. Rep 1 R
361-5. Rep 2 R
361-5, Rep 3 R
361-6
361-7
Method Blank 3
Method Blank R
Sponsor ID
S US RCC. Rep 1
S US RCC. Rep 2
S US RCC. Rep 3
BUSRCC
1 US RCC
DB Naph-
thalene
37%*
34%*
37%'
25% *
27% '
26% '
51%
D10 Acenaph-
thalene
53%
47%
50%
45%
48%
26% *
62%
D12 Perylene
79%
63%
76%
64%
60%
90%
72%
STANDARD REFERENCE MATERIAL
SRMNIST1041
26%*
47%
74%
CD
MATRIX SPIKE RESULTS
Amount Spiked
391-5 *
361 5 4 Spike R
Amount Recovered
Percent Recovery
R - Re extracted sample results.
* - Value outside ol Internal QC llmlls (40 120%)
NA - Not appllcabie.
NA
38%
40%
NA
NA
NA
50%
54%
NA
NA
NA
79%
88%
NA
NA
-------
CO
o
SAIC GLNPO (CF #361)
PAH IN TREATED SEDIMENT
Low Molecular Weight PAHs (ng/g dry wl )
B.E.S.T.
REVISED
2/18/92
Naphthalene Acenaphthytene Acenaphtiene
MSL Code Sponsor ID
361-8 H S-TS-RCC
381 9 R B-TS RCC
361-10 I-TS RCC
Method Blank-4
Method Blank R
STANDARD REFERENCE MATERIAL
SRMNIST1941
certified vilue
MATRIX SPIKE RESULTS
Amount Spiked
361 8
361-8 + Spike R
Amount Recovered
Percent Recovery
Amount Spiked
361-8 DUP
361-8 + Spike DUP R
Amount Recovered
Percent Recovery
24 B
37B
2253
38 U
11
364
tc
2222
24
723
699
31%'
1799
24
999
975
54%
14 U
15U
121
39 U
11 U
54
tc
2222
14 U
1004
1004
45%
1799
14 U
1262
1249
69%
19 U
21 U
726
56 U
16 U
60 U
NC
2222
19 U
933
933
42%
1799
19 U
1149
1130
63%
Fluorene Phenanlhrene Anthracene
16 U
18 U
1078
49 U
13 U
03
ND
2222
16 U
1165
1185
53%
1799
16 U
1281
1265
70%
99 B
68 B
5642
33 U
9
550
577
2222
99
2052
1954
88%
1799
99
1468
1369
76%
17
30
1474
36 U
9 U
165
202
2222
17
1658
1642
74%
1799
17
1370
1353
75%
R - Re-extracted sample results.
U - Below detection limits
NC - Not certified
• - Value outside ol Internal OC limits (40-120%)
-------
SAIC GLNPO (CF »361)
PAH IN TREATED SEDIMENT
Hlqh Molecular Weigh! PAHs (nq/g dry wl )
MSL Cod* Sponsor ID
361-8 R S-TS-RCC
381-0 R B-TS -RCC
361-10 I-TS-RCC
Method Blank-R
Method Blank R
STANDARD REFERENCE MATERIAL
SRM-NIST1941
certified value
MATRIX SPIKE RESULTS
Amount Spiked
361-8R
381 8 4 Spike R
Amount Recovered
Percent Recovery
Amount Spiked
361 8 R DUP
361-8 + Spike DUP R
Amount Recovered
Percent Recovery
B.E.S.T.
Fluoran-
thene
138 B
45 B
3105
25U
0
1114
1220
2222
138
2655
2717
122% '
1798
136
1698
1561
87%
Pyrene
REVISED
2/18/82
Indeno Dlbenio
Benzo(a)- Chryscne Benzofb)- Beruo(k)- B«nio(B>- (1,2.3.c.d) (a.h)anthra- B«nzo(g,hJ)-
anthracene
120B
38 B
3553
26 U
0
1034
1080
2222
120
2555
2434
110%
1789
120
1584
1473
62%
57
22
3126
26U
5 U
481
650
2222
57
2654
2506
117%
1788
57
1600
1632
81%
luoranlhena ftuorantiene
88 B
38 B
3882
24 U
5
703
rC
2222
88
2688
2502
113%
1788
88
1611
1623
85%
87B
27B
1887
16U
6
766
780
2222
87
2670
2773
125%'
1788
87
1734
1637
81%
88 B
4 B
1316
16 U
5
603
444
2222
66
2367
2301
104%
1788
66
1608
1443
80%
pyrene
76 B
18 B
3220
20 U
6
500
670
2222
76
2584
2508
113%
1788
78
1506
1430
79%
pyrene
00 B
18B
1360
18U
5
488
568
2222
80
2838
2848
128%'
1788
00
1703
1613
00%
cene
18
6
1830
24 U
4 U
141
rC
2222
16
2610
2584
117%
1788
16
1638
1622
80%
perytene
80 B
14 B
2354
18
6
421
616
2222
60
1786
1726
78%
1788
60
106S
1005
66%
R - Re-extracted sample results.
U . Below detection limits
NC - Not certified
* - Value outside ol Internal OC limits (40-120%)
-------
SAIC GLNPO (CF »361)
PAH IN TREATED SEDIMENT
B.E.S.T.
REVISED
2/18/82
( Surrogate Recovery %
MSLCode
Sponsor ID
DB Naph-
thalene
01 0 Acenaph-
thalene
012 Perylene
361-8 R
361-0 R
361-10
S-TS-RCC
B-TS-RCC
I-TS-RCC
41%
31%*
25%'
46%
43%
S4%
67%
73%
85%
Method Blank-R
Method Blank-R
STANDARD REFERENCE MATERIAL
21%*
61%
19%'
62%
60%
72%
SRMNIST1041
28%*
47%
74%
CO
to
MATRIX SPIKE RESULTS
Amount Spiked
36I-8R
361-8 + Spike R
Amount Recovered
Percent Recovery
Amount Spiked
361 8 R DUP
361-8 + Spike DUP R
Amount Recovered
Percent Recovery
R - Re extracted sample results.
• - Values outside of Internal QC limits (40-120%).
NA - Not applicable.
NA
41%
31%'
NA
NA
NA
41%
66%
NA
NA
NA
46%
40%
NA
NA
NA
46%
64%
NA
NA
NA
67%
108%
NA
NA
NA
67%
87%
NA
NA
-------
SAIC GLNPO (CF »361)
PAH IN WATER
B.E.S.T.
REVISED
2/3/92
low Molecular Walghl PAHs (ng/U
MSL Code Sponsor ID
361-1
361-2
381 11
SWHRCC
B WHRCC
1 WR HCC
Naphthalene Acenaphlhytene Acenaphftene
1441 B
988 B
301 B
538 U
342 U
495
679 U
432 U
165
Fluorene Phenanlhrene Anthracene
641 U
408 U
278
421 U
268 U
2719
508 U
323 U
997
CO
00
Method Blank 2
MATRIX SPIKE RESULTS
Amounl Spiked
Blank 2
Blank-2 + Spike
Amounl Recovered
Percent Recovery
1767
5000
1767
2552
785
16%'
963 U
1216 U
B - Analyle present In method blank associated with sample
U - Below detection limits
* - Value outside ol Internal OC limits (40-120%).
1148 U
754 U
910U
5000
963 U
1072
1072
21% '
5000
1216 U
1143 U
1143U
0% '
5000
1148 U
1108
1108
22% '
5000
754 U
1504
1504
30%'
5000
910 U
1420
1420
28% '
-------
SAIC GLNPO (CF
PAH IN WATER
•361)
B.E.S.T.
REVISED
2/S/02
I Surrogate Recovery %
MSLCode
361-1
361-2
361-11
Sponsor ID
SWRRCC
BWRRCC
IWRRCC
08 Naph-
thalene
39%*
35%*
20%'
010 Acenaph-
thalene
44%
37%*
28%
012 Perylene
130%
111%
00%
CO
Method Blank-2
MATRIX SPIKE RESULTS
Amount Spiked
Blank-2
Blank-2 + Spike
Amount Recovered
Percent Recovery
37%
NA
37%'
20%'
NA
NA
38% '
NA
38% '
20% '
NA
NA
83%
NA
83%
109%
NA
NA
* - Value outside ol Internal QC limits (40-120%).
NA - Not applicable.
-------
SAIC GLNPO (CF (361)
PAH IN WATER
High Molecular Weight PAHs (ng/l)
B.E.S.T.
REVISED
MSLCode
361 1
361-2
361-11
Sponsor ID
S WR RCC
B-WR-RCC
I-WR-RCC
Fhrararv
Ihene
402
256 U
17056
Indeno Dlbenzo
Pyrene Beruo(a) Chrysen* Benzo(b)- Benzof»- Bento(i}- (1.2.3.c.d) (a.h)anlhra- Beruo(o,M-
anlhracene tuoranlhene Ruorantiene pyrene pyrane cent pcryton*
405 U 436 U 360 U 361 U 200 U 367 U 348 U 377 U 280
258 U 289 U 242 230 U 18SU 247U 221 U 240 U 184
17998 6418 10870 8708 3068 6181 3235 762 2841
Method Blank 2
MATRIX SPIKE RESULTS
Amount Spiked
Blank 2
Blank 2 + Split*
Amount Recovered
.^Percent Recovery
CO
721 U
726 U
B - Analyle present In method blank associated with sample
U - Below detection limits
* - Value outside ol Internal QC limits (40-120%).
613U
680 U
648 U
620 U
604 U
623 U
676 U
S18U
6000
721 U
3816%
3816%
76%
6000
726 U
3830%
3830%
77%
6000
813U
6076
6076
140%'
6000
680 U
6752
5752
115%
6000
648 U
6620
8620
133%*
6000
620 U
6570
5570
112%
6000
604 U
4704
4704
eo%
6000
623 U
6686
5685
114%
6000
676 U
6011
6011
118%
6000
518
4854
4336
67%
-------
iAIC-GLNPO (CF 11361)
»AH IN OIL
bow Molecular Weight PAHs (no/ml)
: Sample
MSL Code Sponsor ID Density (g'ml)
361-3 S-OR-RCC 0
361-4 R B-OR-RCC 0
361-12. Rep 1 I-OR-RCC. Rep 1 0
361-12. Rep 2 I-OR-RCC. Rep 2 0
361-12. Rep 3 I-OR-RCC. Rep 3 0
Method Blank
OIL CONCENTRATIONS ON % OIL BASIS
Low Molecular Wekftl PAHs (oofcaoH)
MSL Code Sponsor ID
361-3 S-OR-RCC
361-4 R B^OHRCC
361-12. Rep 1 1 OR RCC. Rep 1
361-12. Rep 2 1 OR RCC, Rep 2
361-12. Rep 3 1 OH RCC. Rep 3
MATRIX SPIKE RESULTS
Amount Spiked
361-3
361-3 + Spike
Amount Recovered
Percent Recovery
7525
6003
7301
7301
.7301
%ON
0 25
6 20
5908
5998
59.06
Naphthalene
603 U
369 U
12127 DU
1 1 1 78 DU
11639DU
1774DU
Naphthalene
8664 U
8508 U
27694 DU
25526 DU
26351 DU
SOOOO
603 U
6260
5260
11%*
B.E.S.T.
Acenaphthylene Acenaphtiene
764 U
396 U
20391 D
22779 D
198590
1902U
Acenaphthylene
10977U
9121 U
46565 0
52019 D
45351 0
SOOOO
764 U
16600
16800
34%*
985 U
657 U
18291 DU
221840
18657 D
2675 DU
v*in6pnV)9n*i
13865U
12B32U
41770 DU
50660 0
42377 D
SOOOO
965 U
19300
19300
30%'
Fkiorene Phenanlhrene
12SO
1063
21142D
24980 D
21992 D
2241 DU
Fkiorene 1
17960
24483
482800
57045 D
50222 D
SOOOO
1250
32400
31150
62%
11800
6943
92741 0
97299 0
00220 D
1312 DU
•henanthrene
160540 D
160068
211786 D
222105 D
206020 D
SOOOO
11600
56700
44000
90%
REVISED
2/21/92
Anthncvnt
7210
6465
47165 D
61328 0
45668 D
1489 DU
Anthracene
103592 D
125930
107684 D
1172140
104289 D
SOOOO
7210
61800
54590
109%
R - Re-extracted sample results
D - Samples diluted 1:10 and rerun
U - Below detection limits.
• > Outside of Internal QC limits (40 120%)
-------
CO
SAIC GLNPO (CF «361)
PAH IN OIL
Htoh Molecular Wolghl PAHs (ng/ml)
B.E.S.T.
Sample
Density
MSL Cod* Spomof ID (gyrrt)
361-3 S-OH-RCC 0
361-4 R &ORRCC 0
361-12. Rap 1 I-OR-HCC. Rep 1 0
361-12. Rap 2 1 OR RCC. Rep 2 0
361-12. Rep 3 1 OR RCC. Rep 3 0
Method Blank
7525
6903
.7301
7301
.7301
Fhioran-
Ihene
10500
8715
277651 D
287605 D
270480 D
808 DU
Pyrene
Benzo(a)-
anthracene
17900
7932
266075 D
274465 D
258488 0
895 DU
8070
3870
151300 D
160160 D
146080 D
846 DU
Chryten*
Benzo(b)-
Benzo (k)-
luoranthene duorantfiene
0060
5220
210701 D
220510 D
204500 0
80S DU
8710
3771
1 80630 D
1 00805 D
177767 D
646 DU
6430
2068
126840 D
120943 D
122228 D
520 DU
Benzo(*}-
pyrene
7020
3684
180062 D
108435 D
180074 D
688 DU
Indeno Dlbenro
(1.2,3.c.d) (a.h) anlhra-
pyrene
6300
2028
159145 D
165497 D
1507580
cen*
2160
670
32883 D
35296 D
30003 D
707 DU 604 DJ
REVISED
2/21/02
Benzo(g,h.l)-
perylene
4840
1708.00
06127 D
08204 D
00713 D
1430 O
OIL CONCENTRATIONS ON % OIL BASIS
Low Molecular WakjhlPAHs (ugAflolQ
MSL Cod*
Sponsor ID
361-3 S-OR-RCC
361-4 R B^ORRCC
381-12. Rep 1 I-OR RCC. Rep 1
361-12, Rep 2 I-OR-RCC. Rep 2
361-12, Rep 3 I-OR-RCC, Rep 3
MATRIX SPIKE RESULTS
Amount Spiked
361-3
361-3 * Spike
Amount Recovered
Percent Recovery
%OI
f*l
0.25
6.20
60.08
60.08
60.08
Fhioran-
Ihene
280172
200614
634051 0
656088 D
617606 D
50000
10500
83600
84100
128% '
Pyrene
267184
182760
607616 D
626776 D
500240 0
50000
17000
77500
59600
119%
Beiuofa)-
anthracene
116048
80370
346718 D
36S743D
336647 D
60000
6070
78600
70530
141%'
Ctuytene
B*iuo(b>-
Bemo(k>-
luoranthene luoranlieoe
143301
120480
4813680
603562 D
467228 D
60000
0080
86100
56120
112%
125144
86800
433065 D
4684850
405000 D
50000
8710
75800
67090
134%*
023BS
68386
280865 0
206741 D
270123 D
60000
6430
64300
67870
116%
B«UO(a)-
pyren*
113703
84870
433602 D
453151 D
4132770
50000
7020
72700
64780
130% '
Indeno
(l.2.3.e.d)
pyren*
00517
87465
363428 D
377033 D
3442760
60000
6300
77100
70800
142% '
Dlbenzo
(a.h) anthra-
cene
31034
13336
75002 0
80603 0
70571 D
60000
2160
81400
70240
158%*
Benzo(g,h.l)-
penrtene
60540
30367
210518 D
224467 D
207166 D
50000
4840
64700
50860
120%
R - Re-extracted •ample results.
D - Samples diluted 1.10 and re-run.
U - Below detection limits
• . Outside ol Internal QC limits (40-120%)
-------
co
oo
SAIC-GLNPO (CF 1361)
PAH IN OIL
B.E.S.T.
REVISED
2/21/82
I Surrogate Recovery % I
MSLCode Sponsor 10
381-3 S-OR-RCC
301-4 R B-OR-RCC
381-12. Rep 1 I-OR RCC. Rep 1
361-12. Rep 2 1 OR RCC. Rep 2
361-12. Rep 3 I-OR-RCC. Rep 3
Method Blank
OIL CONCENTRATIONS ON % OIL BASIS
Low Molecular Wclgjil PAHs (uoAooN)
MSLCod* Sponsor ID
361 -3 S-OR-RCC
361-4 R B-OR-RCC
361-12. Rep 1 I-OR-RCC. Rep 1
361-12, Rep 2 I-OR-RCC, Rep 2
361-12. Rep 3 I-OR-RCC, Rep 3
MATRIX SPIKE RESULTS
Amount Spiked
361-3
361-3 + Spike
Amount Recovered
Percent Recovery
D8 Naph-
thalene
6% '
22%'
23% D*
30% 0
19% D*
60% D
DID Acenaph-
Ihalene
46%
39%'
81% D
63% 0
59% 0
118% D
1 Surrogate Recovery %
D8 Naph-
thalene
8%*
22%'
23% D*
30% D
10% 0*
NA
8% '
11% '
NA
NA
01 0 Acenaph-
thalene
46%
39%*
61% D
63% D
59% D
NA
46%
38%'
NA
NA
D12 Perylene
181%
85%
123% D'
108% D
108% D
73% D
1
01 2 Perylene
161%
05%
123% D*
106% D
108% D
NA
161%
130%
NA
NA
R - Re-extracted •ample results.
D - Samples diluted 1:10 and re-run.
• . Outside of Internal OC limits (40 120%).
NA - Not applicable.
-------
CO
CD
RE-PROCESSED RESULTS (1/92)
PCBs IN UNTREATED SEDIMENT
Concentrations In ug/kg dry weight
B.E.S.T
SAIC-GLNPO(CF#381)
2/18/92
MSLCode Sponsor ID
361-5. Rep 1 S-US-RCC, Rep 1
361-5, Rep 2 S-US-RCC. Rep 2
361-5. Rep 3 S-US-RCC. Rep 3
361-6 B-US-RCC
361-7 I-US-RCC
Blank-3
STANDARD REFERENCE MATERIAL
SRM-1 (HS-2)
certified value
MATRIX SPIKE RESULTS
Amount Spiked
361-5*
361-5 + Spike
Amount Recovered
Percent Recovery
Amount Spiked
Blank-3
Blank-3 + Spike
Amount Recovered
Percent Recovery
REPLICATE ANALYSES
361-5, Rep 1 S-US-RCC, Rep 1
361-5, Rep 2 S-US-RCC. Rep 2
361-5. Rep 3 S-US-RCC. Rep 3
Aroclor
1242
25372
16249
17666
200 U
200 U
200 U
40 U
N3
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
25372
16249
17666
RSD% 25%
Aroclor Aroclor Aroclor
1248 1254 1260
200 U
200 U
200 U
325
11257
200 U
40 U
NC
N3
NS
NS
NS
NS
NS
NS
NS
NS
NS
200 U
200 U
200 U
0%
2591 E 100 U
1691 E 100U
2025 E 100 U
100 U 100U
3746 E 100 U
100U 100U
132 127
111 N3
609S NS
2102 NS
6129 NS
4027 NS
66V. NS
3333 NS
100 U NS
2498 NS
2498 NS
75% NS
Tetrachloro- Oclachloro-
m-Xylene naphthalene
60.9% 84.2%
58.0% 94.1%
59.6% 89.5%
63.3% 66.2%
61.7% 57.1%
256.1% ' 122.7% '
39.0% ' 42.4%
N3 N3
NA NA
58.8% 89.3%
61.0% 90.1%
NA NA
NA NA
NA NA
258.1% * 122.7% *
48.2% 55.8%
NA NA
NA NA
2591 100U 60.9% 84.2%
1691 100U 56.0% 94.1%
2025 100U 59.6% 89.5%
22% 0%
4% 6%
U • Below detection limits.
E • Values due to residuals from high Aroclors 1242 and 1248 levels.
' « Value outside of Internal QC limits (40-120%).
NC - Not certified.
# - Mean of 3 replicates.
NS = Not spiked. NA = Not applicable.
-------
RE-PROCESSED RESULTS (1/92)
PCBs IN TREATED SEDIMENT
Concentrations In uq/kq dry weight
B.E.S.T
SAIC-GLNPO(CF#361)
2/18/92
% Surrogate Recovery
MSLCode Sponsor ID
361-8 S-TS-RCC
361-9 B-TS-RCC
361-10 " I-TS-RCC
Blank-4
STANDARD REFERENCE MATERIAL
SRM-2 (HS-2)
certified value
MATRIX SPIKE RESULTS
Amount Spiked
361-8
361-8 -i- Spike
Amount Recovered
Percent Recovery
Amount Spiked
361-8
361 -8 + Spike DUP
Amount Recovered
Percent Recovery
U - Below detection limits.
J - Detected below detection limit.
Aroclor Aroclor Aroclor Aroclor lletrachloro- Octachloro-
1242 1248 1254 1260 | m-Xvlene naphthalene
205
100 U
100 U
200 U
200 U
N3
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
100 U
100 U
440
200 U
200 U
N3
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
30 J
50 U
50 U
100 U
100 U
111
2404
30 J
2060
2030
84%
2273
30 J
1169
1139
50%
SOU
SOU
SOU
100U
100U
ND
NS
NS
NS
NS
NS
NS
NS
NS
NS.
NS
35.2% *
34.5% '
89.7%
21 .0% *
44.8%
N3
MA
35.2% *
72.3%
NA
NA
NA
35.2% *
30.2% *
NA
NA
71.2%
90.0%
50.3%
71.5%
41 .7%
N3
NA
71.2%
89.2%
NA
NA
NA
71.2%
55.4%
NA
NA
* . Value outside of Internal QC limits (40-120%).
" . Early elutlng peaks present that do not match Arolcor pattern.
NC - Not certified.
NS » Not spiked.
NA = Not applicable.
Quantity estimated at -1000-2000 ppb, based on TCX response factor.
-------
RE-PROCESSED RESULTS (1/92)
PCBs IN WATER
Concentrations In ug/L
B.E.S.T
SA1C-GLNPO (CF #381)
2/18/92
MSLCode
361-1
361-2
361-11
Blank-1
Sponsor ID
S-WR-RCC
B-WR-RCC
I-WR-RCC
Aroclor
1242
0.2
0.2
0.2
0.2
U
U
U
U
Aroclor
1248
0.2
0.2
4.8
0.2
U
U
U
Aroclor
1254
0.1
0.1
0.1
0.1
U
U
U
U
Aroclor
1260
0.1
0.1
0.1
0.1
U
U
U
U
Tetrachloro-
m-Xylene
41.2%
35.2% *
43.5%
36.5% '
Octachloro-
naphlhalene
114.1%
98.0%
86.0%
98.0%
MATRIX SPIKE RESULTS
Amount Spiked
Blank-1
Blank-U Spike
Amount Recovered
Percent Recovery
U « Below detection limits.
* - Value outside of Internal QC limits (40-120%).
NC . Not certified.
NS - Not spiked.
NA - Not applicable.
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
50
0.1 U
42.6
42.6
85%
NS
NS
NS
NS
NS
NA
38.5% '
33.6% '
NA
NA
NA
98.0%
123.8%
NA
NA
-------
RE-PROCESSED RESULTS (1/92)
PCBs IN OIL
B.E.S.T
8AIC-GLNPO (CF §361)
2/26/92
% Surrogate Recovery
Sample
MSLCode Sponsor ID Density (g/ml)
361 -3 S OR RCC
361-4 B OR RCC
361-12, Rep 1 1 OR RCC. Rep 1
361-12. Rep 2 I-OR-RCC. Rep 2
361-12, Rep 3 I-OR-RCC, Rep 3
Blank-1 OH
0.7525
0.6903
0.7301
0.7301
0.7301
Aroclor
1242
318320
2000 U
2000 U
2000 U
2000 U
2000 U
Aroclor
1248
2000 U
2702
118616
120743
112109
2000 U
Aroclor
1254
30789
1000 U
1000 U
1000 U
1000 U
1000 U
Aroclor
1260
1000U
1000U
1000U
1000 U
1000 U
1000U
OIL mur«ITR»T10N9 ON % OIL BASIS
Concentrations In ug/Kq oil \
MSLCode Sponsor ID
361 -3 8 OR RCC
381-4 B OR RCC
361-12, Rep 1 I-OR-RCC, Rep 1
361 -1 2. Rep 2 1 OR RCC, Rep 2
381-12, Rep 3 I-OR-RCC, Rep 3
MATRIX SPIKE RESULTS
Amount Spiked
361-3
361-3* Spike
Amount Recovered
Percent Recovery
REPLICATE ANALYSES
381-12, Rep 1 I-OR-RCC. Rep 1
361-12. Rep 2 1 OR-RCC, Rep 2
381-12. Rep 3 IOH-RCC. Rep 3
%OII
(%)
0.25
8.20
50.08
50.08
50.08
RSD%
Aroclor
1242
4573563
46083 U
4587 U
4567 U
4587 U
NS
NS
NS
NS
NS
2000 U
2000 U
2000 U
NA
Aroclor
1248
28738 U
62258
270875
275732
256015
NS
NS
NS
NS
NS
118616
120743
112109
4%
Aroclor
1254
442371
23041 U
2284 U
2284 U
2284 U
60000
30780
84012
34123
68%
1000 U
1000 U
1000 U
NA
Aroclor
1260
Tetrachloro-
m-Xylene
65.0%
71.5%
03.7%
85.5%
85.2%
48.6%
Octachloro-
naphlhalene
00.5%
01.0%
96.0%
83.7%
90.4%
96.7%
% Surrogate Recovery
retrachloro-
m-Xylene
14368U 65.0%
23041 U 71.5%
2284 U 03.7%
2284 U 85.5%
2284 U 85.2%
NS
NS
NS
NS
NS
1000
1000
1000
NA
NA
65.0%
61.2%
NA
NA
U 93.7%
U 85.5%
U 65.2%
5%
Octachloro-
naphlhaUne
00.5%
01.0%
06.0%
83.7%
00.4%
NA
00.5%
00.4%
NA
NA
06.0%
83.7%
00.4%
7%
U - Below detection limits.
NC - Not certified.
NS - Not spiked.
NA - Not applicable.
-------
APPENDIX F
QUALITY ASSURANCE/QUALITY CONTROL
In order to obtain data of known quality to be used in evaluating the different technologies for the
different sediments, a Quality Assurance Project Plan (QAPP) was prepared. The QAPP specified the
guidelines to be used to ensure that each measurement system was in control. In order to show the
effectiveness of the different technologies, the following measurements were identified in the QAPP as critical
- PAHs, PCBs, metals, total solids, volatile solids, and oil and grease in the untreated and treated sediments.
Other parameters analyzed in the sediments included pH, TOO, total cyanide, and total phosphorus. If water
and oil residuals were generated by a technology, then PAHs and PCBs were determined as a check on
their fate in treating the sediments. Each of these measurements and the associated quality control (QC)
data will be discussed in this section.
Also included in this section are a discussion of the modifications and deviations from the QAPP
and the results of a laboratory audit performed. Any possible effects of findings on data quality will be
presented.
PROCEDURES USED FOR ASSESSING DATA QUALITY
The indicators used to assess the quality of the data generated for this project are accuracy,
precision, completeness, representativeness, and comparability. All indicators will be discussed generally
in this section; specific results for accuracy, precision, and completeness will be summarized in later
sections.
Accuracy
Accuracy is the degree of agreement of a measured value with the true or expected value.
Accuracy for this project will be expressed as a percent recovery (%R).
Accuracy was determined during this project using matrix spikes (MS) and/or standard reference
materials (SRMs). Matrix spikes are aliquots of sample spiked with a known concentration of target
analyte(s) used to document the accuracy of a method in a given sample matrix. For matrix spikes,
recovery is calculated as follows:
c,-c0
%R = X 100
Ct
where: C, = measured concentration in spiked sample aliquot
C0 = measured concentration in unspiked sample aliquot
Ct = actual concentration of spike added
143
-------
An SRM is a known matrix spiked with representative target analytes used to document laboratory
performance. For SRMs, recovery is calculated as follows:
Cm
%R = _ X 100
where: Cm = measured concentration of SRM
Ct = actual concentration of SRM
In addition, for the organic analyses, surrogates were added to all samples and blanks to monitor
extraction efficiencies. Surrogates are compounds which are similar to target analytes in chemical
composition and behavior. Surrogate recoveries will be calculated as shown above for SRMs.
Precision
Precision is the agreement among a set of replicate measurements without assumption of
knowledge of the true value. When the number of replicates is two, precision is determined using the
relative percent difference (RPD):
(C, - C2) X 100
RPD = _
(C, + C2)/2
where: C-, = the larger of two observed values
C2 = the smaller of two observed values
When the number of replicates is three or greater, precision is determined using the relative standard
deviation (RSD):
S
RSD = X 100
X
where: S = standard deviation of replicates
X = mean of replicates
Precision was determined during this project using triplicate analyses for those samples suspected
to be high in target analytes (i.e., untreated sediments). Matrix spike and matrix spike duplicate (MSD)
analyses were performed on those samples suspected to be low in target analytes (i.e., treated sediments).
A MSD is a second spiked sample aliquot with a known concentration of target analyte used to document
accuracy and precision in a given sample matrix.
144
-------
Completeness
Completeness is defined as the percentage of valid data points to the total number of data points
obtained.
VDP
% Completeness = X 100
TOP
where: VDP = number of valid data points
TOP = total data points obtained
For this project, completeness was determined for each parameter for each technology evaluated.
(Add more)
Representativeness
Representativeness refers to the degree with which analytical results accurately and precisely
represent actual conditions present at locations chosen for sample collection. Sediment samples were
collected prior to this demonstration and were reported to be representative of the areas to be remediated.
Comparability
Comparability expresses the extent with which one data set can be compared to another. As will
be discussed in more detail in the section MODIFICATIONS AND DEVIATIONS FROM THE QAPP, the data
generated are comparable within this project and within other projects conducted for the ARCS Program.
ANALYTICAL QUALITY CONTROL
The following sections summarize and discuss analytical procedures and the results of the QC
indicators of accuracy and precision for each measurement parameter.
PAH Procedures
Sediments and waters were extracted and analyzed using modified SW-846 procedures as described
in the section MODIFICATIONS AND DEVIATIONS FROM THE QAPP. Oils were diluted 1:10 in hexane.
Three radiolabelled PAH surrogates were added to all samples and blanks prior to extraction. Daily mass
tuning was performed using decafluorotriphenylphosphine (DFTPP) to meet the criteria specified in Method
8270. The instrument was calibrated at five levels for the sixteen polynuclear aromatic hydrocarbons (PAHs).
The RSD of the response factors for each PAH was required to be <25 percent. Calibrations were verified
every 12 hours for each PAH; criteria for % difference from the initial calibration was <25 percent. An
internal standard, hexamethyl benzene, was added prior to cleanup and was used to correct PAH
145
-------
concentrations for loss during cleanup and extract matrix effects. Quantification was performed using
Selective Ion Monitoring (SIM).
PAH QC Results and Discussion
Surrogate recoveries for all PAH samples for the B.E.S.T. demonstration are summarized in Table
QA-1. If more than one of the three surrogates fell outside the control limits used, corrective action
(reanalysis) was necessary. (Insufficient sample remained for reanalysis of water residuals). Surrogate
recoveries were generally low for samples and method blanks, indicating a possible analytical problem rather
than matrix effects. An investigation indicated possible problems with the evaporator used to concentrate
the extracts. This concentration step was not performed for the oil residuals and, as can be seen in Table
QA-1, acceptable surrogate recoveries for the method blank were obtained. In summary, low surrogate
recoveries indicate that PAH target concentrations may be biased somewhat low. Since both the untreated
and treated sediments were affected similarly, relative removal percentages should be valid.
As required by the QAPP, triplicate analyses of the Saginaw River untreated sediment (S-US-RCC)
were performed to assess precision. These results are summarized in Table QA-2. A matrix spike was
performed on this same sample to assess accuracy. These results are included in Table QA-2. All RSD and
spike recoveries fell within the control limits specified. It should be noted that these QC analyses were
reanalyses; the initial analysis yielded unacceptable surrogates and poor precision. The effect of missed
holding times on data quality will be discussed in the section HOLDING TIMES.
As required by the QAPP, a matrix spike and a matrix spike duplicate (MS/MD) analysis was
performed for the treated Saginaw River sediment (S-TS-RCC). These results are presented in Table QA-3.
While recoveries were generally acceptable, RPDs were consistently outside the control limits specified in
the QAPP. These RPDs, however, are generally within acceptance criteria specified in Method 8270; data
should be of sufficient quality to support project results. It should be noted that spike levels averaged up
to one hundred times the target concentrations and that accuracy and precision data for this MS/MSD may
not be indicative of the accuracy and precision obtainable at the target concentrations.
Due to the minimal amount of water generated by the B.E.S.T. process, no QC analyses were
performed on this matrix.
The QAPP specified that triplicate analyses and a matrix spike be performed on the Saginaw River
oil residual (S-OR-RCC). This sample was spiked but triplicate analyses were performed on the Indiana
Harbor oil residual. These results are summarized in Tables QA-4 and QA-5, respectively.
One certified NIST standard reference material (SRM) was extracted and analyzed with the sediment
samples. The recoveries for this standard are summarized in Table QA-6.
Method blanks were extracted and analyzed with each set of samples extracted. Insignificant
146
-------
TABLE QA-1. PAH SURROGATE RECOVERIES, PERCENT
Sample
S-US-RCC, Rep. 1
S-US-RCC, Rep. 2
S-US-RCC, Rep. 3
B-US-RCC
I-US-RCC
Method Blank
Method Blank
S-TS-RCC
B-TS-RCC
I-TS-RCC
Method Blank
S-WR-RCC
B-WR-RCC
I-WR-RCC
Method Blank
S-OR-RCC
B-OR-RCC
I-OR-RCC, Rep. 1
I-OR-RCC, Rep. 2
I-OR-RCC, Rep. 3
Method Blank
d8-Napthalene
37 *
34 *
37 *
25 *
27 *
26 *
25 *
41
31 *
25 *
21 *
39 *
35 »
20 *
37 *
8 *
22 *
23 *
30 *
19 *
60
d 1 0-Acenapthalene
53
47
50
45
48
26 *
24 •
46
43
54
19 *
44
37 *
28 *
38 *
46
39 *
61
63
59
118
d12-Peryiene
79
83
76
64
60
90
90
67
73
85
50
130 *
111
90
83
161 *
95
123 *
106
108
73
Control Limits
40-120
I
I
I
I
I
I
40-120
I
I
I
40- 120
I
I
I
40-120
I
I
I
I
I
* Outside Control Limits
quantities of some PAHs were found in a few blanks; total concentrations are unaffected. No corrections
for method blanks were performed.
147
-------
TABLE QA-2. PAH REPLICATE AND SPIKE RESULTS FOR S-US-RCC
Compound
Napthalene
Acenaphthylene
Acenaphthene
Ruorene
Phenanthrene
Anthracene
Ruoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
lndeno(1 ,2,3,c,d)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
NC
U
Replicate 1
dry ppb
26
15
19
33
249
64
376
425
184
265
238
180
224
201
44
113
Not Calculated
Undetected
Replicate 2
dry ppb
24
20U
20U
32
302
70
464
491
210
299
270
200
250
231
47
129
Replicate 3
dry ppb
27
18
30U
33
249
63
351
402
163
242
217
158
200
188
39
109
Mean
26
NC
NC
33
267
66
397
439
186
269
242
179
225
207
43
117
RSD
6.0
NC
NC
1.8
11
5.7
14
11
13
11
11
12
11
11
9.4
9.1
Precision %
Control Limits Recovery
20 41
| 62
I 57
I 64
I 64
| 63
I 73
| 68
I 79
I 74
| 86
I 72
| 82
| 95
| 93
I 58
Accuracy
Control Limits
40-120%
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
TABLE QA-3. PAH MS/MSD RESULTS FOR S-TS-RCC
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Ruorene
Phenanthrene
Anthracene
Ruoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo (b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
lndeno(1 ,2,3,c,d)pyrene
Dibenzo(a,h)anthracene
Benzo (g,h,i)perylene
MS Recovery
31 *
45
42
53
88
74
122 *
110
117
113
125 *
104
113
128 *
117
78
MSD Recovery
54
70
64
71
76
75
87
82
91
85
91
80
79
90
90
56
Accuracy Precision
RPD Control Limits Control Limits
32 * 40 - 120% 20
43 *
42 *
29 »
15
1.3
33 *
29 *
25 *
28 *
31 *
26 *
35 *
35 *
26 *
33 *
Outside Control Limits
PCBs
PCB Procedures
Sediments and waters were extracted and analyzed using modified SW-846 procedures as described
in the section MODIFICATIONS AND DEVIATIONS FROM THE QAPP. Oil were diluted 1 no in hexane. Two
surrogates, tetrachloro-m-xylene and octachloronaphthalene, were added to all samples and blanks prior
to extraction. The gas chromatograph (GC) employed electron capture detection (ECD) and was calibrated
at three levels for each of four Arociors (1242, 1248, 1254, 1260). The RSD of the response factors for each
Aroclor was required to be <25 percent. Calibrations were verified after every ten samples; criteria for %
difference from the initial calibration was <25 percent. An internal standard, dibromooctafluorobiphenyl, was
added prior to cleanup and was used to correct PCB concentrations for loss during cleanup and extract
matrix effects. Quantification of Arociors was performed on two columns (DB-5, primary and 608,
confirmation) as a confirmation of their presence.
149
-------
TABLE QA-4. PAH MS RESULTS FOR S-OR-RCC
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Ruorene
Phenanthrene
Anthracene
Ruoranthene
Pyrene
Benzo (a)anthracene
Chrysene
Benzo (b)fluoranthene
Benzo (k)fluoranthene
Benzo(a)pyrene
lndeno(1 ,2,3,c,d)pyrene
Dibenzo (a.h)anthracene
Benzo(g,h,i)perylene
MS Recovery % Control Limits
1 1 Not Specified
34
39
62
90
109
128
119
141
112
134
116
130
142
158
120
PCB QC Results and Discussion
Surrogate recoveries for all PCB samples for the B.E.S.T. demonstration are summarized in Table
QA-7. If both recoveries fell outside the control limits used, correction action (reanalysis) was necessary.
All samples were acceptable with respect to the surrogate criteria used.
As required by the QAPP, triplicate analyses of the Saginaw River untreated sediment (S-US-RCC)
were performed to assess precision. These results are summarized in Table QA-8. A matrix spike using
Aroclor 1254 was performed on the same sample to assess accuracy. These results are included in Table
QA-8. RSDs were outside specified control limits but within precision specified in Method 8080; data should
be of sufficient quality to support project results.
As required by the QAPP, a matrix spike and a matrix spike duplicate (MS/MSD) analysis was
performed for the treated Saginaw Rive sediment (S-TS-RCC). These results are presented in Table QA-9.
The RPD was outside control limits; however, no Aroclor 1254 was found in the sample and the data is not
impacted. Matrix spike recoveries were acceptable. Due to the minimal amount of water generated by the
B.E.S.T. process, no QC analyses were performed on this matrix.
The QAPP specified that triplicate analyses and a matrix spike be performed on the Saginaw River
150
-------
oil residual (S-OR-RCC). This sample was spiked but triplicate analyses were performed on the Indiana
Harbor oil residual. These results are summarized in Tables QA-10 and QA-11, respectively.
One standard reference material (SRM) certified by the National Research Council of Canada
(NRCC) for Aroclor 1254 was extracted and analyzed twice with the sediment samples. Recoveries of 82.8%
and 78.6% were obtained. The average of 80.7% fell within the 80-120% criteria specified in the QAPP.
Method blanks were extracted and analyzed with each set of samples extracted. No PCBs were
found in any method blanks.
METALS
Metals Procedure
Sediments were prepared for metals analysis by freeze-drying, blending, and grinding. Sediments
for As, Cl, Hg, and Se were digested using nitric and hydrofluoric acids. The digestates were analyzed for
As, Cd, and Se by graphite furnace atomic absorption (GFAA) by SW-846 Method 7000 series using Zeeman
Background correction. The digestates were analyzed for mercury by cold vapor AA (CVAA) using SW-846
Method 7470.
Sediments for As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn were analyzed by energy-diffusive X-Ray
fluorescence (XRF) following the method of Sanders (1987). The XRF analysis was performed on a 0.5g
aliquot of dried, ground sediment pressed into a pellet with a diameter of 2 cm.
Metals QC Results and Discussion
Triplicate analyses of the Saginaw River untreated sediment (S-US-RCC) and treated sediment (S-TS-
RCC) were performed to assess precision. Matrix spikes were analyzed for the same samples to assess
accuracy. Results are summarized in Tables QA-12 and QA-13. It should be noted that the sediments were
not spiked for XRF analysis.
Accuracy and precision results for metals were acceptable with only a few minor exceptions, as
shown in Tables QA-12 and QA-13. These exceptions have little, if any, impact on data quality and project
results.
One NIST certified standard reference material (SRM) was digested and/or analyzed twice with the
sediment samples for XRF, GFAA, and CVAA analyses. These results are presented in Table QA-14.
Method blanks were digested and analyzed for the metals analyzed by GFAA and CVAA. (Method
blanks are not applicable to XRF analysis). If analyte was detected in the method blank, blank correction
was performed. Minimal amounts of some metals were detected; data quality is not affected.
151
-------
TABLE QA-5. PAH REPLICATE"0 RESULTS FOR I-OR-RCC ng/ml
en
ro
Compound
Napthalene
Acenaphthylene
Acenaphthene
Ruorene
Phenanthrene
Anthracene
Ruoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
lndeno(1 ,2,3,c,d)pyrene
Dibenzo(a,h)anthracene
Benzo(g,h,i)perylene
Replicate 1
ppb
30000 U
46600
50000 U
48300
212000
108000
634000
608000
346000
481000
433000
290000
434000
363000
75100
220000
(a) Replicate results represent values
NC
U
Not Calculated
Undetected
Replicate 2
ppb
30000 U
52000
50100
57000
222000
117000
657000
627000
366000
504000
456000
297000
453000
378000
80600
224000
after correction for percent oil concentration.
Replicate 3
ppb
30000 U
45400
42400
50200
206000
104000
618000
590000
336000
467000
406000
279000
413000
344000
70600
207000
RSD Control Limits
NC Not Specified
7.4
NC
8.9
3.8
6.1
3.1
3.0
4.4
3.8
5.8
3.1
4.6
4.7
6.7
4.1
-------
TABLE QA-6. PAH SRM RESULTS
Compound
Naphthalene
Acenaphthylene
Acenaphthene
Ruorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo (a)anthracene
Chrysene
Benzo (b)f luoranthene
Benzo(k)fluoranthene
Benzo(a)pyrene
lndeno(1 ,2,3,c.d)pyrene
Dibenzo(a,h)anthracene
Benzo (g,h,i)pery1ene
Ftecovery, % Control Limits
NC 80 - 120%
NC
NC
NC
95
81
91
96
87
NC
98
136 *
75 *
88
NC
82
NC = Not Certified
Outside Control Limits
TABLE QA-7. PCB SURROGATE RECOVERIES, PERCENT
Sample
S-US-RCC, Rep. 1
S-US-RCC, Rep. 2
S-US-RCC, Rep. 3
B-US-RCC
I-US-RCC
Method Blank
S-TS-RCC
B-TS-RCC
I-TS-RCC
Method Blank
S-WR-RCC
B-WR-RCC
I-WR-RCC
Method Blank
S-OR-RCC
B-OR-RCC
I-OR-RCC, Rep. 1
I-OR-RCC, Rep. 2
I-OR-RCC, Rep. 3
Method Blank
Tetrachloro-m-xylene
53
56
51
63
62
25 *
35 *
34 *
89
21 *
41
35 *
44
36 *
65
71
92
86
84
48
Octachloronaphthalene Control Limits
85 40-120
94
92
70
61
130 *
76 40-
96
54
76
124 * 40 -
109
98
107
100 40-
103
107
97
102
105
120
120
120
Outside Control Limits
153
-------
TABLE QA-8. PCB REPLICATE AND SPIKE RESULTS FOR S-US-RCC
Replicate 1 Replicate 2 Replicate 3
Aroclor ppb dry ppb dry ppb dry Mean
1242/1248
1254
1260
Precision
RSD Control Limits
20
20
20
Accuracy
% Recovery Control Limits
„
40 - 120
—
01
-p.
TABLE QA-9. PCB MS/MSD RESULTS FOR S-TS-RCC
PCB
MS Recovery
MSD Recovery
PRO
Accuracy Control
Limits
Precision
Control Limits
Aroclor 1254
40-120%
20
-------
TABLE QA-10. PCB MS RESULT FOR S-OR-RCC
PCB
Aroclor 1254
MS Recovery
Control Limits
Not Specified
OIL AND GREASE
Oil and Grease Procedures
Sediment samples were extracted with freon using Soxhlet extraction according to SW-846 Method
9071. The extract was analyzed for oil and grease by infra-red (IR) as outlined in Method 418.1 (Methods
for Chemical Analysis of Water and Wastes, 1983).
Oil and Grease QC Results and Discussion
Both the untreated and treated Saginaw River sediments (S-UC-RCC and S-TS-RCC) were analyzed
for oil and grease in triplicate. In addition, matrix spike was performed for S-TS-RCC. These results are
summarized in Table QA-15. All QC results fell within specified control limits.
TOTAL VOLATILE SOLIDS
Total Volatile Solid Procedures
Sediments were analyzed for total volatile solids (TVS) following the procedures in Method 160.4.
Methods for Chemical Analysis of Water and Waste, 1983) modified for sediments. An aliquot of sediment
was dried and then ignited at 550°C. The loss of weight on ignition was then determined.
Total Volatile Solid QC Results and Discussion
Both the untreated and treated Saginaw River sediments (S-US-RCC and S-TS-RCC) were analyzed
for TVS in triplicate. Results are summarized in Table QA-16. Both RSDs fell within specified control limits.
AUDIT FINDINGS
An audit of the Battelle-Marine Sciences Laboratory was conducted on September 25 and 26,1991.
Participants included EPA, GLNPO, and SAIC personnel. The path of a sample from receipt to reporting
was observed specifically for samples from these bench-scale treatability tests. Two concerns were
identified in the organic laboratory: 1) the preparation, storage, record-keeping, and replacement of
155
-------
TABLE QA-11. PCB REPLICATE RESULTS FOR I-OR-RCC
Aroclor
1242/1248
1254
1260
Replicate 1 , ppb
Replicate 2, ppb
Replicate 3, ppb
Mean
RSD
Control Limits
20
20
20
TABLE QA-12. METALS REPLICATE AND SPIKE RESULTS FOR S-US-RCC
Metal
Ag
As
Ba
Cd
Cr
Cu
Fe(1)
Hg
Mn
Ni
Pb
Se
Zn
Method
GFAA
XRF
XRF
GFAA
XRF
XRF
XRF
CVAA
XRF
XRF
XRF
GFAA
XRF
Replicate 1,
ppm dry
0.87
1.86
322
4.39
93
55.8
0.780
0.162
162
58.1
43.0
0.3 U
126
Replicate 2,
ppm dry
0.85
2.65
322
4.08
112
57.3
0.765
0.178
159
55.2
42.5
0.3 U
133
Replicate 3,
ppm dry
0.80
2.12
321
3.96
117
63.4
0.816
0.160
173
61.5
50.9
0.3 U
162
Mean
RSD
Precision
Control
Limits
%
Recovery
Accuracy
Control
Limits
* Outside Control Limits NS
(1) Results in Percent for Fe U
Not Spiked NC
Undetected
Not Calculated
-------
TABLE QA-13. METALS REPLICATE AND SPIKE RESULTS FOR S-TS-RCC
Metal
Ag
As
Ba
Cd
Cr
Cu
Fe(1)
Hg
Mn
Ni
Pb
Se
Zn
Method
GFAA
XRF
XRF
GFAA
XRF
XRF
XRF
CVAA
XRF
XRF
XRF
GFAA
XRF
Replicate 1,
ppm dry
1.03
2.77
325
4.38
130
68.6
0.855
0.505
181
71.5
53.3
0.3 U
194
Replicate 2,
ppm dry
0.76
2.87
321
4.17
113
66.3
0.827
0.290
177
64.1
45.0
0.3 U
165
Replicate 3,
ppm dry
0.67
2.92
310
4.24
110
57.4
0.795
0.209
172
57.2
41.4
0.3 U
147
Mean
0.82
2.85
319
4.26
118
64.1
0.826
0.335
177
64.3
46.6
0.3 U
169
RSD
23
2.7
2.4
2.5
9.2
9.2
3.6
46*
2.6
11
13
NC
14
Precision
Control
Limits
20
I
I
I
I
I
I
I
I
I
I
I
I
%
Recovery
120*
NS
NS
100
NS
NS
NS
98
NS
NS
NS
101
NS
Accuracy
Control
Limits
85-115
—
—
85-115
—
—
—
85-115
—
—
—
85-115
™"~
en
* Outside Control Limits
(1) Results in Percent for Fe
NS = Not Spiked
U = Undetected
NC = Not Calculated
-------
TABLE QA-14. METALS SRM RESULTS, % RECOVERY
Metal
Ag
As
Ba
Cd
Cr
Cu
Fe
Hg
Mn
Ni
Pb
Se
Zn
SRM-1
NC
113
NC
117
85.5
124*
104
105
93.3
113
101
NC
96.7
SRM-2
NC
97.4
NC
117
103
110
100
105
90.4
97.2
99.6
NC
93.0
Control Limits
80-120%
* Outside control limits.
NC = not certified.
standards is not well-documented; and 2) the nonstandard procedures used to extract, clean up and analyze
samples needs to be documented with reported data.
During the audit, the use of nonstandard procedures was discussed. It was concluded that data
comparability within this project and within the ARCS program should not be an issue, as the Battelle
laboratory has performed all analyses to date. However, comparability to data generated outside the ARCS
program is not possible.
MODIFICATIONS AND DEVIATIONS FROM THE QAPP
Laboratory activities significantly deviated from the approved QAPP in two areas-analytical
procedures and quality assurance (QA) objectives. Specific deviations and their effect on data quality are
discussed in this section.
158
-------
en
CO
TABLE QA-15. OIL AND GREASE REPLICATES AND SPIKE RESULTS FOR S-US-RCC AND S-TS-RCC
Sample
S-US-RCC
S-TS-RCC
Replicate 1,
ppm dry
1480
297
Replicate 2,
ppm dry
1320
293
Replicate 3,
ppm dry
1260
206
Mean
1350
265
RSD
8.4
19
Precision
Control Limits
20
20
% Recovery
114
NS
Accuracy
Control Limits
80-120%
NS
Not Spiked
TABLE QA-16. TVS REPLICATES FOR S-US-RCC AND S-TS-RCC
Sample
S-US-RCC
S-TS-RCC
Replicate 1, % dry
2.24
1.76
Replicate 2, % dry
2.03
1.70
Replicate 3, % dry
2,01
1.74
Mean
2.09
1.73
RSD
6.1
1.8
Control Limits
20
20
-------
ANALYTICAL PROCEDURES
The Assessment and Remediation of Contaminated Sediments (ARCS) Program was initiated by the
Great Lakes National Program Office (GLNPO) to conduct bench-scale and pilot-scale demonstrations for
contaminated sediments. To date, all laboratory analyses performed in support of the ARCS Program have
been done at the Battelle-Marine Sciences Laboratory (MSL) in Sequim, Washington. Standard procedures
used by Battelle-MSL often do not follow those procedures identified in SW-846 and the QAPP. While these
nonstandard procedures yield results of acceptable quality, comparability with analyses performed outside
the ARCS Program is not possible.
PAH Analysis
• Samples were co-extracted with PCB samples using a modified SW-846 extraction procedure which
entailed rolling of the sample in methylene chloride and an additional clean-up step using high
pressure liquid chromatography (HPLC). An internal standard, hexamethyl benzene, was added
prior to this clean-up step to monitor losses through the HPLC. Final results were corrected for the
recovery of this internal standard. A second internal standard, d12-phenanthrene, was added prior
to analysis; however, no corrections were made based on its recovery. Neither of these internal
standards are specified in Method 8270.
• SW-846 Method 8270 was modified to quantify the samples using Selective Ion Monitoring (SIM)
Gas Chromatography/Mass Spectrometry (GC/MS). This modification results in improved detection
limits.
• Three radiolabelled PAH compounds were used as surrogates rather than those recommended in
Method 8270., Recoveries of these compounds should better represent the recoveries of target
PAHs.
PCB Analysis
• Samples were extracted using the modified extraction procedures as described for the PAH
analysis. An internal standard, dibromooctafluorobiphenyl, was added prior to the HPLC clean-up
to monitor losses. Final results were corrected for the recovery of this standard. A second internal
standard, 1,2,3-trichlorobenzene (required by QAPP) was added prior to analysis; however, no
corrections were made based on its recovery.
• Quantification of PCBs was not done on a total basis as required by SW-846 Method 8080 but by
quantifying four peaks for each Aroclor and averaging these results.
-------
• The surrogate required by the QAPP, tetrachloro-m-xylene, was used. A second surrogate,
octochloronaphthalene, was also added to monitor extraction efficiency.
Metals Analysis
• Nine of the 13 metals analyzed for sediment samples were measured by energy-diffusive X-Ray
fluorescence (XRF) - As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn. This procedure yields a total metals
concentration instead of the recoverable metals determined by SW-846 methods.
• Sediments for Ag, Cd, Hg, and Se were subjected to an acid digestion using nitric and hydrofluoric
acids. This digestion again yields total rather than recoverable metals.
Oil and Grease
• Oil and grease extracts for sediments were analyzed using infrared (IR) detection rather than the
gravimetric procedures specified in the QAPP. This should have no effect on data quality.
QUALITY ASSURANCE OBJECTIVES
Many of the QA objectives and internal QC checks criteria specified in the QAPP (particularly for
organic analyses) are not routinely achievable by standard or nonstandard methods. To avoid excessive
reanalyses (both costly and time-consuming), acceptance criteria established internally by Battelle were used
for this project. These internal limits are adequate for use in determining whether or not project results are
valid.
PAH Analysis
• Both surrogate and matrix spike objectives for PAHs were specified in the QAPP to be 70-130%.
For surrogates, Battelle actually used internal limits of 40-120%, with one of the three surrogates out
of limits being acceptable. If more than one surrogate did not fall within 40-120%, reanalysis was
required. For matrix spikes, internal limits of 40-120% were also used; no reanalyses however, were
performed based on exceedences of these limits.
• Limits for continuing calibration checks were specified as ±10% in the QAPP; limits of ± 25% were
used.
PCS Analysis
• Both surrogate and matrix spike objectives for PCBs were specified in the QAPP to be 70-130%.
For surrogates, Battelle actually used internal limits of 40-120%. If both surrogates exceeded these
limits, re-extraction was performed. For matrix spikes, internal limits of 40-120% were also used;
no reanalyses, however, were performed based on exceedences of these limits.
161
-------
Limits for continuing calibration checks were specified as ±10% in the QAPP; limrts of ±25% were
used.
Metals Analysis
Samples analyzed by XRF cannot be spiked. Therefore, no measure of sample accuracy was
obtained for those metals previously identified as being analyzed by XRF. An SRM was analyzed,
providing a means to measure method accuracy for eight of the nine metals determined by XRF (all
but Ba).
e.qarep.bgr
162
-------
Data Verification Report For Assessment and
Remediation of Contaminated Sediment Program
Report Number 8
(SAIC, Bench-Scale Tests)
By
M. J. Miah, M. T. Dillon, and N. F. D. O'Leary
Lockheed Environmental Systems and Technologies Company
980 Kelly Johnson Drive
Las Vegas, Nevada 89119
Version 1.0
Work Assignment Manager
Brian A. Schumacher
Exposure Assessment Research Division
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89193
Environmental Monitoring Systems Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Las Vegas, Nevada 89193
-------
ABSTRACT
Data submitted by the Science Applications International Corporation
(SAIC) of Cincinnati, Ohio, have been verified for compliance of the QA/QC
requirements of the Assessment and Remediation of Contaminated Sediment
(ARCS) program. This data set includes results from bench-scale technology
demonstration tests on wet contaminated sediments using four treatment
technologies, namely, B.E.S.T. (extraction process), RETEC (low temperature
stripping), ZIMPRO (wet air oxidation), and Soil Tech (low temperature
stripping). The primary contaminants in these sediments were polychlorinated
biphenyls (PCBs) and polynuclear aromatic hydrocarbons (PAHs). In addition,
metal contents and conventional (% moisture, pH, % total volatile solids, oil and
grease, total organic carbon (TOC), total cyanide, and total phosphorus) in these
sediments were also considered for this project. The objective of the bench-scale
technology demonstration study was to evaluate four different treatment
techniques for removing different organic contaminants from sediments. Both
treated and untreated sediment samples were analyzed to determine treatment
efficiencies.
A total of seven sediment samples from four different areas of concerns
(Buffalo River, Ashtabula River, Indiana Harbor, and Saginaw River) were
analyzed under the bench-scale technology demonstration project. The samples
from these areas of concern (AOCs) were collected by the Great Lakes National
Program Office (GLNPO) in Chicago, IL, and sample homogenization was
performed by the U. S. EPA in Duluth, MN. SAIC was primarily responsible
for the characterization of the sediment samples prior to testing and for the
residues created during the test. The solid fraction analyses were performed by
SAIC's analytical subcontractor Battelle-Marine Sciences Laboratory of Sequim,
Washington, and Analytical Resources Incorporated of Seattle, Washington.
The submitted data sets represent analyses of untreated sediments, as well
as solid, water, and oil residues obtained by using different treatments. The
verified data set is divided into several parameter groups by sampled media. The
data verifications are presented in parameter groups that include: metals, PCBs,
conventionals, and PAHs.
The results of the verified data are presented as a combination of an
evaluation (or rating) number and any appropriate data flags that may be
applicable. The templates used to assess each individual analyte are attached in
case the data user needs the verified data of a single parameter instead of a
parameter group.
-------
INTRODUCTION
The bench-scale technology demonstration project was undertaken to evaluate the
efficiencies of four techniques used for the removal of specific contaminants from wet sediments
collected from designated Great Lakes areas of concern. Four different sediment treatment
techniques, namely, B.E.S.T (Basic Extraction Sludge Technology), RETEC, ZIMPRO, and Soil
Tech were considered for evaluation. B.E.S.T. is a solvent extraction process, RETEC and Soil
Tech are low temperature stripping techniques, and ZIMPRO is a wet air oxidation technique.
Wet sediments were collected by the Great Lakes National Program Office (GLNPO) from four
Great Lakes sites, namely, the Buffalo River in New York, the Saginaw River/Bay (referred to
as Saginaw River throughout the following discussions) in Michigan, the Grand Calumet
River/Indiana Harbor (referred to as Indiana Harbor throughout the following discussions) in
Indiana, and the Ashtabula River in Ohio. The four techniques were used to treat the sediment
samples from these four sites. The sediment samples represent the sediment that would be
obtained for on-site treatment.
The B.E.S.T. process is a patented solvent extraction technology that uses the inverse
miscibility of triethylamine as a solvent. At 65° F, triethylamine is completely soluble in water
and above this temperature, triethylamine and water are partially miscible. This property of
inverse miscibility is used since cold triethylamine can simultaneously solvate oil and water.
RETEC and the Soil Tech (low temperature stripping) are techniques to separate volatile and
semivolatile contaminants from soils, sediments, sludges and filter cakes. The low temperature
stripping (LTS) technology heats contaminated media to temperatures between 100 -2Qtf F,
evaporating off water and volatile organic contaminants. The resultant gas may be burned in
an afterburner and condensed to a reduced volume for disposal or can be captured by carbon
absorption beds. For these treatability studies, only the processes that capture the driven off
contaminants were considered. The ZIMPRO (wet air oxidation) process accomplishes an
aqueous phase oxidation of organic and inorganic compounds at elevated temperatures and
pressures. The temperature range for this process is between 350 to 600" F (175 to 320* Q.
System pressure of 300 psi to well over 300 psi may be required. In this process, air or pure
oxygen is used as an oxidizing agent.
Samples for the technology demonstration projects were obtained by GLNPO (Chicago,
Illinois) and were analyzed by Battelle-Marine Sciences Laboratory (Battclle-MSL, Sequim, WA)
and by Analytical Resources Incorporated (Seattle, WA). To evaluate the bench-scale
technologies, the sample analyses were divided into four parts: (1) raw untreated sediment
samples, (2) treated sediments, (3) water residues, and (4) oil residues. The amount of residues
available for the analyses depended upon the corresponding sediment samples and on the
individual technology used to treat those sediment samples.
The analyses of sediment and residue parameters for these projects were divided into four
different categories: (1) metals, including Ag, As, Ba, Cd, Cr, Cu, Fe, Hg, Mn, Ni, Pb, Se,
and Zn; (2) polychlorinated biphenyls (PCBs); (3) polynuclear aromatic hydrocarbons (PAHs);
-------
and (4) conventionals, including percent moisture, pH, percent total volatile, oil and grease, total
organic carbon (TOC), total cyanide, and total phosphorus. Analyses of metals and
conventionals were performed on treated and untreated sediment samples only for B.E.S.T.,
ZIMPRO, and Soil Tech, while for the RETEC process, analyses of metals and conventionals
were performed on treated and untreated sediment samples as well as water residue samples.
No oil residues were produced by the ZIMPRO technique (wet air oxidation treatment
technique), while in the other three techniques, oil residues were analyzed after appropriate
sample cleanup steps for PCBs and PAHs.
QUALITY ASSURANCE AND QUALITY CONTROL REQUIREMENTS
The objective behind all quality assurance and quality control (QA/QC) requirements is
to ensure that all data satisfy predetermined data quality objectives. These requirements are
dependent on the data collection process itself. Under the bench-scale technology demonstration
project, QA/QC requirements were established for:
1. Detection limits,
2. Precision,
3. Accuracy,
4. Blank analyses,
5. Surrogate and matrix spike analyses, and
6. Calibration
a) initial
b) ongoing.
Four parameter groups analyzed in the sediment and water residue phases were of interest
in the bench-scale technology demonstration project. These groups included: (a) metals, (b)
PCBs, (c) PAHs, and (d) conventionals. The conventionals included: percent moisture, pH,
percent total volatile, oil and grease, TOC, total cyanide, and total phosphorus. In addition,
total solids, total suspended solids, and conductivity were included in the conventionals group
for RETEC conventional analyses. The analyses for metals and conventionals were performed
for solids only, except for RETEC, where metals and conventionals were analyzed in solid and
water residue phases. Parameter groups analyzed in the oil residue phase are PCBs and PAHs.
The objective of these analyses was to characterize samples both before and after each treatment
was applied.
The detection limits for metals, PCBs, PAHs, and conventionals (where appropriate)
were defined as, three times the standard deviation for 15 replicate analyses of a sample with
an analyte concentration within a factor of 10 above the expected or required limit of detection.
Individual parameter detection limits are presented in the approved quality assurance project plan
for SAIC on file at the Great Lakes National Program Office in Chicago, IL.
-------
Precision requirements were based on analytical triplicate analyses for all parameters of
sediment samples and treated residues, at the rate of 1 per 20 samples. The results of the
triplicate analyses provided the precision for the analytical laboratory. An acceptable limit was
the coefficient of variation less than or equal to 20 percent. The precision requirement was
established for all variable types in this project. For treated sediments, the relative percent
difference (RPD) between the matrix spike and matrix spike duplicate was used as a measure
of precision with an acceptance limit of less than 20% .
Accuracy was defined as the difference between the expected value of the experimental
observation and its "true" value. Accuracy in this project was required to be assessed for each
variable type using analysis of certified reference materials, where available, at the rate of 1 per
20 samples. Acceptable results must agree within 20 percent of the certified range. Since no
PCBs and PAHs were expected to be detected in the treated sediment, matrix spikes and matrix
spike duplicate analyses were required during the analyses of treated sediment for the organic
parameters. Matrix spike analyses were used as a measure of accuracy for treated sediment
analyses, with an acceptance limit of ±30% from the known value.
Matrix spikes were required to be used at a rate of 1 per 20 samples and to be within
plus or minus 15 percent of the spiking value for metals and 70 to 130 percent of the spiking
value for organics (PCBs and PAHs).
Surrogate spike analyses were only required for each sample in organic analyses. The
acceptable limits for the surrogate recovery was between 70 and 130 percent of the known
concentration.
The observed values should have been less than the method detection limit for each
parameter for method blanks (run at the beginning, middle, and end of each analytical run).
The ongoing calibration checks were required at the beginning, middle, and end of a set
of sample analyses for all variable types. The maximum acceptable difference was ±10% of
the known concentration value in the mid-calibration range. Initial calibration acceptance limits,
for metals, was the _> 0.97 coefficient of determination for the calibration curve, while a %RSD
of the response factors of less than or equal to 25% was required for organics.
RESULTS AND DISCUSSION
The ARCS QA program was formally adopted for use when SAIC received final approval
from the GLNPO on May 31, 1991. An evaluation scale, based upon the QA program
developed for the ARCS program, was developed to evaluate the success of the data collection
process in meeting the QA/QC requirements of the ARCS program. The following section
discusses how to interpret the data verification results.
-------
The Verification Process and Evaluation Scale
For verification purposes, the data set from each technology was divided into 4 different
sample media as follows:
1. Untreated sediment,
2. Treated sediment,
3. Water residue, and
4. Oil residue.
The verification process included QA/QC compliance checking for accuracy, precision,
matrix spike analysis, surrogate spike analysis, blank analysis, detection limits, initial and
ongoing calibration checks, and holding times as well as checks on calculational correctness and
validity on a per parameter/analyte basis. Compliance checks were performed to ensure that the
QA/QC measurements and samples: (a) met their specified acceptance limits; (b) had reported
results that were supported by the raw data; and (c) were analyzed following good laboratory
practices, where checking was possible. Upon completion of the verification process, a final
rating was assigned for each of the individual categories. The final ratings are presented as a
combination of a number value and a flag list.
The numerical value for the rating of a given parameter was assigned based upon the
successful completion of each required QA/QC sample or measurement. The QA/QC samples
were broken down into four different sample groups, namely, accuracy, precision, blanks, and
spike recoveries. A fifth category was included for QA/QC measurements to address the
successful completion of instrument calibrations (both initial and ongoing) and the determination
of method detection limits. If the laboratory successfully met the acceptance criteria of 50
percent or more of the parameters in a given QA/QC sample group, then the laboratory received
the full value for that category. For example, if 50 percent or more of the reagent blanks for
the metals in sediment analyses had measured values below the method detection limit, then
three points were awarded for that category, assuming reagent blanks were the only blank
samples analyzed by the laboratory. The individual point values for each QA/QC sample type
or measurement and the minimum acceptance levels for each category are presented in Appendix
B. The final numerical rating presented for each parameter category is the summation of the
point values from each of the five categories.
Along with each numerical rating, a list of appropriate flags has been attached to the final
rating value (Appendix C). The flag indicates where discrepancies exist between the laboratory
data and the acceptance limits of the required QA program. Different flags are presented for
each category of QA sample (accuracy, precision, blanks, and spike recoveries) and for the
QA/QC measurements (instrument calibration and detection limit determination). The flags have
a letter and subscript configuration, such as A,. The letter of the flag represents the category
of the discrepancy while the subscript designates the form of the discrepancy. For example, the
A flags indicate discrepancies in the use of accuracy checking samples, such as reference
materials or standards. A flag with a subscript of 1 indicates that the laboratory failed to meet
-------
the acceptance criteria. Using the example of the A, flag, this flag would then indicate a failure
of the laboratory to meet the QA/QC requirements for the use of reference materials in their
appraisal of accuracy. A flag with the subscript 0 indicates that no information was received
(or no standards were available in the case of accuracy) from the analytical laboratory, and
therefore, no points could be allotted towards the final calculated rating value for that particular
category. It should be noted that the 0 flag does not necessarily indicate that the analytical
laboratory did not perform the QA/QC analyses, only that no information was received from the
laboratory.
The subscript 9 flag indicates that the sample category or QA/QC measurement is not
applicable to that particular parameter or parameter group (Appendix C). For example, an S,
flag indicates that a matrix spike for that given parameter or analyte is not applicable, such as
was the case for percent moisture. Where subscript 9 flags occur, an adjustment to the passing
and maximum scores (to be discussed) for a parameter group was made and will be reported in
the appropriate tables.
A complete presentation of the QA/QC rating factors (point values by sample type) and
the various data flags and their subscripts are presented in Appendices B and C, respectively.
A more complete discussion of the rating scale can be found in the report submitted to the
RA/M workgroup by Schumacher and Conkling entitled, "User's Guide to the Quality
Assurance/Quality Control Evaluation Scale of Historical Data Sets."
Individual parameter flags are presented in the templates found in Appendix D. The
objective of the presentation of the individual flag templates is to help the data user make a
determination regarding the useability of the data set for any given purpose and to provide the
data user with a means to assess any individual parameter that may be of specific interest
The Interpretation and Use of the Final Verified Data Rating Values
The data verification scale was developed to allow for the proper rating of the verified
data and the subsequent interpretation and evaluation of the ratings. Two different
interpretations can be made using the ratings provided in this report, namely, the actual or "true"
rating and the potential rating. The first interpretation is based upon the formal ARCS QA
program, while the second interpretation scale is based upon the "full potential" value of the
submitted data set. In the following sections, each interpretation of the results will be discussed.
Data Interpretation Based upon the Formal ARCS OA Program
For each of the four parameter categories, the data were initially verified for QA/QC
compliance following the requirements specified in the signed QAPP submitted by SAIC and the
ARCS QAMP on file at the GLNPO in Chicago, Illinois.
Table 1 provides the verified data ratings for each variable class for the four different
technologies studied based on the current ARCS QA program. The ratings of these variable
-------
classes are presented to provide the data user with a means for comparing the ARCS QA
program-based verified results with other data sets, using the same or similar parameters, that
were generated prior to and after the initiation of the formal ARCS QA program.
Table 2 provides the data user with the full compliance and acceptable scores presented
for each parameter group based upon the current ARCS QA program. The full compliance score
represents the numerical rating value if all required QA/QC samples and measurements were
performed by the analytical laboratory and successfully met all the QA/QC requirements of the
ARCS QA program. An acceptable score is lower than the full compliance score and accounts
for laboratory error that can be reasonably expected during an analysis of multiple samples.
Any final rating value less than the acceptable score indicates that problems were identified in
the data that could adversely effect the quality of the data. The acceptable score was set at 60
percent of the full compliance score. To determine the percentage of QA/QC samples and
measurements successfully analyzed for a given parameter versus the number analyzed following
the complete ARCS QA protocols, divide the numerical rating received by the full compliance
score. An acceptable data set, in this case, has a rating of 60 percent or greater.
In some cases, all the QA/QC requirements may not be applicable (e.g., matrix spikes
for percent solids are not applicable). If this is the case, a flag with the subscript 9 was used,
and the full compliance and acceptable scores were adjusted by lowering the score on appropriate
number of points for nonrequired sample type, as identified in Appendix B. An example of this
situation is % moisture, as indicated in Table 1, the subscript 9 flag has been applied to
accuracy, blank, detection limit, and spike samples. Therefore, the full compliance and
acceptable scores (Table 2) are only based upon the possible points for the successful completion
of the remaining QA/QC samples that have cumulative points value of 8 (Appendix B).
Data Interpretation Based upon the" Potential' Value of the Data Set
A second interpretation scale has been presented to allow the data user to establish the
"full potential" value of the submitted data set. The numerical value and associated flags
presented in the first interpretation can be considered as an absolute rating for that data set or
parameter. These ratings were based upon all the data submitted to Environmental Monitoring
Systems Laboratory - Las Vegas (EMSL-LV) and to Lockheed for review by the analytical
laboratory. If one or more parameter or parameter groups qualifying flags had the subscript of
5, 6, 9, or 0 (Appendix C), the required information was not available or not applicable at the
time of sample analysis, and consequently was not included during the data verification and
review process. The equivalent point value(s) for each individual sample type may be added to
the reported point sum to give the data user the full potential value of the data set. This process
assumes that if the "missing" QA/QC samples or measurements were performed, the results
would fall within the ARCS QA program specified acceptance limits. For example, if the point
value (including qualifying flags) for the metals was 6-Bo Q, D0 So, then the data user could
potentially add 14 points to the score since the blank analyses, spike information, detection limit,
-------
and calibration (initial and ongoing) information was not available for verification. The resulting
data would then have a rating of 20.
TABLE 1. Verified Data Ratings Based on the Current ARCS QA Program
Untreated
Sediments
Metals
% Moisture
pH
%TVS
Oil and grease
TOC
Total cyanide
Total
phosphorus
PCBs
PAHs
B.E.S.T.
12-CoD0
0-A,B,CoD,P0S,
0-A,B,CoD,P0S,
6-A,CoD,S,
15-A, C,
12-C6P0S9
14-AoPo
14-AoP,
17-B,D0
17-D0 S,
ZBVfPRO
12-CoD0
3-A,B,CoD,S,
0-A,B,CoD,PoS,
3-A,BoCoD,S,
6-A, B, C« D, S,
12-C6P0S,
14-AoP0
14-Ao P0
14-A, B2 D0
11-8,005,5,
Soil Tech
12-Q D0
0-A,B,QD,P0S,
0-A,B,C0D,P0S,
6-A, Co 0,5,
6-A, B, C. D, S0
12-C6 P0 5,
H-AoP0S0
14-Ao P0
14-A, B, DO
17-D0 5,
RETEC
12-QDo
3-A,B,CaD,S,
3-A, B, Co D, 5,
6-A, C, D, S,
9-A, De C, So
9-C6D0P,S,
8-AoD0P,So
11-AoD.So
H-A.BjDoS,
20-D0
Treated
Sediments
Metals
% Moisture
pH
%TVS
Oil and grease
TOC
Total cyanide
Total
phosphorus
PCBs
PAHs
12-CoD,
0-A9B9C0D,P0S9
0-A,B,CoD,PoS,
6-A, 0, 0,5,
15-A,C«
12-C6P0S,
14-AoPo
14-AoPo
14-BjDoP,
14-D0P,S,
12-CoDo
0-A,B,CoD,P0S,
3-A,B,CoD,S,
3-A,BoCoD,S,
6-A, B, C6 D, 5,
12-C, P0 5,
14-Ao P0
14-Ao P0
H-A.BjDoP,
17-D0S,
12-C, D0
3-A, B, Co D, 5,
0-A,B,CoD,PoS,
6-A,CoD,S,
9-A, B, C. D,
12-C4P0S,
14-Ao P0
14-Ao P0
14-B,D0P,
14-DoP.S,
12-C.D.
3-A,B,C,D,S,
3-A,B,C.D,S,
6-A,C,D,S,
6-A,C,D,P,So
12-C.D.S,
11-AoD.Po
14-Ao D.
14-A, B, D0
20-D,
-------
TABLE 1. Verified Data Rating Based on the Current ARCS Program
(Continued)
Water
residue
Metals
% Moisture
PH
Total
Suspended
Solids
%TVS
Total Solids
Oil and grease
TOC
Total cyanide
Total
phosphorus
Conductivity
PCBs
PAHs
**
**
»*
**
**
**
**
**
**
**
14-B2 D0 P0
11-AoDoP.S,
«*
**
««
**
**
**
**
*«
**
**
14-Bj D0 P0
17-D0 S,
**
**
«»
**
**
**
*«
**
**
**
S-A^DoPoS,
s.
17-D0 P0
20
»»
3-A,B,Q,D,S9
e-A.CoD.S,
6- A, Co D9 S,
6-A,CoD,S,
12-A, C, D0
9-AoC4D0S,
14-A,D0
14-AoD0
9-AoC.D,S,
5-AoBjDoPoS,
s.
ll-A.DjP.S,
Oil residue
PCBs
PAHs
H-A.BjDoS,
H-AoBjD0S,
*
*
17-B, Do
14-Bj Do Sj
11-B2D0P0S3
17-BjDo
* No oil residue was produced by this treatment
** Analyses were not conducted for this treatment
-------
TABLE 2. Full Compliance and Acceptable Scores Based on the Current ARCS QA Program
Variable Class
Metals in Treated Sediment
Metals in Untreated Sediment
% Moisture
pH
%TVS
Oil and grease
TOC
Total cyanide
Total phosphorus
Conductivity
Suspended Solids
Total Solids
PAHs
PCBs
Full Compliance
20
20
8
8
9
17
17
20
20
14
9
9
23
23
Acceptable
12
12
5
5
6
11
11
12
12
9
6
6
14
14
Table 3 presents the verified data ratings for each variable class in the four technologies
based on their full potential value. All data qualifying flags with the subscripts 5, 6, 9, or 0
have been removed. The appropriate point values for each of the 5, 6, or 0 flags (Appendices
B and C) were added to the final rating scores for each parameter or parameter group. In
contrast, the removal of the subscript 9 flags resulted in an adjustment to the full compliance and
acceptable scores, and noj in an addition to the calculated point scores since these analyses were
not applicable to the methodologies used by the laboratory (Table 2).
-------
10
TABLE 3. Verified Data Ratings Based on the Full Potential of the Data set
Untreated
Sediments
Metals
% Moisture
pH
%TVS
Oil and grease
TOC
Total cyanide
Total phosphorus
PCBs
PAHs
B.E.S
-------
11
TABLE 3. Verified Data Ratings Based on the Full Potential of the Data set
(continued)
Water
residue
Metals
% Moisture
pH
%TVS
Oil and grease
TOC
Total cyanide
Total phosphorus
Conductivity
Suspended Solids
Total Solids
PCBs
PAHs
**
**
**
**
*«
**
**
**
**
**
«*
20-Bj
17-P.S,
**
**
**
**
*«
**
**
««
»*
**
**
20-Bj
20-83
**
**
*«
**
«*
**
**
**
«*
«*
**
14-A.BjS,
23
20
8
8
6
17
17
20
20
14
6
6
20-Bj
14-A, P, S,
Oil residue
PCBs
PAHs
14-A, Bj S,
17-8,5,
*
*
20-B,
17-8,5,
20-82
20-Bj
* No oil residue was produced by this treatment
** Analyses were not conducted for this treatment
To evaluate the data using the values presented in Table 3, the final ratings should be
compared to the full compliance and acceptable scores presented in Table 2. The data user
should bear in mind that these values are only the potential values of the data set and assumes
that the "missing" QA/QC data could have been or were performed successfully by the
laboratory. Any value falling below the acceptable value presented in Table 2 clearly indicates
that major QA/QC violations were identified and the data should be used with a great deal of
caution by the data user.
-------
12
Data Verification Results for Bench-scale Technology Demonstration Project
B.E.S.T.
The B.E.S.T. technology was evaluated by analyzing sediment samples and their treated
residues (treated sediments, water residues, and oil residues) for metals, conventionals, PCBs
and PAHs. PCB and PAH analyses were performed for sediments, water, and oil residues. The
metals and conventional analyses were performed for the sediment samples only.
In the majority of the cases studied, the accuracy objective was satisfactory for the metal
analyses in treated and untreated sediments. Of the thirteen metals analyzed, accuracy
information was not available for Ba, Se, and Ag. In both treated and untreated sediments, ten
of the thirteen metal analyses (As, Cd, Cr, Cu, Fe, Pb, Mn, Hg, Ni, Pb, and Zn) satisfied
ARCS specified QA/QC requirements for accuracy. Four of the thirteen metal analyses (Cd,
Hg, Se, and Ag) satisfied QA/QC requirements for blank analyses, while the remaining nine
metals (As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn) were analyzed by XRF techniques. In all of
the XRF analyses, results from blank sample analyses were not applicable. Both initial and
ongoing calibration for Cd, Hg, Se, and Ag analyses met the ARCS QA/QC specifications for
both treated and untreated sediments, while for the remaining nine metals (As, Ba, Cr, Cu, Fe,
Mn, Ni, Pb, and Zn) calibration information was not available. Detection limits information for
metal analyses in treated and untreated sediments were not available for verification except for
Cd, Hg, Se, and Ag where detection limits were satisfactory. The precision information for the
metal analyses in treated sediment was not available for Se, but was satisfactory for the
remaining elements, with the exception of Hg, where precision information did not satisfy
QA/QC requirements. The precision information for the metal analyses in untreated sediment
was not available for Se, but was satisfactory for the remaining twelve metal (Ag, As, Ba, Cd,
Cr, Cu, Fe, Hg, Mn, Ni, Pb, and Zn) analyses. The matrix spike information for both treated
and untreated sediment analyses were satisfactory for Cd, Hg, and Se, were unsatisfactory for
Ag, while the remaining nine metals (As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn) were analyzed
by XRF techniques. In all of the XRF analyses, results from matrix spike analyses were not
applicable.
Of the seven conventional analyses, the accuracy information in both treated and
untreated sediments was satisfactory for TOC and was not available for total cyanide, and total
phosphorus. In the remaining four conventional analyses, accuracy was not applicable. In both
sediments, five of the seven conventionals (%TVS, oil and grease, TOC, total cyanide, and total
phosphorus) satisfied QA/QC requirements for blank analyses, and the blank information was
not applicable for moisture, pH, and TVS. Both initial and ongoing calibration information was
satisfactory for all conventional analyses in both treated and untreated sediments except for
moisture and pH where calibration information was not available and for TOC and oil and grease
where ongoing calibration information was not available. Detection limits were satisfactory for
four (oil and grease, TOC, total cyanide, and total phosphorus) of the seven conventional
-------
13
analyses in treated and untreated sediments, and were not applicable for moisture, pH, and TVS.
The precision information was satisfactory for two (%TVS, oil and grease) of the seven
conventional analyses in treated and untreated sediments. No precision information was
available for the remaining five conventional analyses in treated or untreated sediments. The
matrix spike information for both treated and untreated sediment analyses were satisfactory for
oil and grease, total cyanide, and total phosphorus, while for the remaining four conventional
analyses the matrix spike information was not applicable.
In treated sediments, untreated sediments, and water residues, the accuracy objective for
PCBs was satisfactory for Aroclor 1254 analyses only and could be used to represent the whole
PCB group. No accuracy information was available for the remaining three Aroclor analyses.
In oil residues, accuracy information was not satisfactory for PCB analyses. In both sediments
and in both residues, PCB analyses did not satisfy ARCS specified QA/QC requirements for
blank analyses indicating potential contamination at the laboratory. Initial and ongoing
calibration was satisfactory for all PCB analyses in both treated and untreated sediments as well
as in water and oil residues. Detection limit information were not available for PCB analyses
in treated and untreated sediments and for water and oil residues. In the untreated sediments,
the precision information was satisfactory for Aroclors 1242 and 1254, and no precision
information was available for Aroclors 1248 and 1260. In the treated sediments, the precision
information was not satisfactory for Aroclor 1254, and no precision information was available
for Aroclors 1242, 1248, and 1260. In water residues, no precision information was available
for any of the Aroclors. In oil residues, the precision information was satisfactory for Aroclor
1248, and no precision information was available for Aroclors 1242, 1254, and 1260. The
matrix spike for Aroclor 1254 was satisfactory for both sediment and water residue analyses and
could be used to represent the whole PCB group. The matrix spike for Aroclor 1254 was
unsatisfactory for the analyses of oil residue. In both sediment or residue analyses, no matrix
spike information was available for Aroclors 1242, 1248, and 1260. The surrogate spike
recoveries were satisfactory for PCB analyses in both sediments and residues.
In eight of sixteen PAH analyses of treated and untreated sediments, the accuracy
objective was satisfactory. No accuracy information was available for six PAHs (naphthalene,
acenaphthylene, acenaphthene, fluorene, chrysene, and dibenzo(a,h)anthracene) analyses in both
treated and untreated sediments. The accuracy objective was not satisfactory for benzo(k)
fiuoranthene and benzo(a)pyrene in treated or untreated sediments. No accuracy information was
available for any of the PAH analyses in water and oil residues. In treated and untreated
sediments, and in water residues, PAH analyses satisfied ARCS specified QA/QC requirements
for blank analyses. In all cases of oil residues, the blank analyses exceeded the MDL indicating
potential contamination at the laboratory. Initial and ongoing calibration limits for PAH analyses
met the ARCS QA/QC specifications for both treated and untreated sediments and water and oil
residue analyses. Detection limit information was not available for PAH analyses in treated and
untreated sediments, nor for water and oil residues. In untreated sediments and oil residues, die
precision information was satisfactory for all PAH analyses, except for acenaphthene in untreated
sediment, and naphthalene in oil residues where no precision information was available. In
treated sediments, the precision information was satisfactory for fluorene, phenanthrene, and
-------
14
anthracene but was unsatisfactory for the remaining PAH analyses. In water residues, no
precision information was available for PAH analyses except for benzo(g,h,i)pyrene where
precision was unsatisfactory. The matrix spike information was satisfactory for twelve of sixteen
PAH analyses in treated sediment and for eight of the sixteen analyses in untreated sediment and
in water and oil residues. Surrogate recoveries were not satisfactory for PAHs in either
sediment and residue analyses.
ZIMPRO
The ZIMPRO technology was evaluated by analyzing sediment samples, treated
sediments, and water residues for metals, conventionals, PCBs, and PAHs. PCB and PAH
analyses were performed for both sediment and water residues. The metals and conventional
analyses were performed for the both sediment samples only.
In the majority of the cases studied, the accuracy objective was satisfactory for the metal
analyses in treated and untreated sediments. Of the thirteen metals analyzed, accuracy
information was not available for Ba, Se, and Ag. In both treated and untreated sediments, ten
of the thirteen metal analyses (As, Cd, Cr, Cu, Fe, Pb, Mn, Hg, Ni, and Zn) satisfied ARCS
specified QA/QC requirements for accuracy. Four of the thirteen metal analyses (Cd, Hg, Se,
and Ag) satisfied QA/QC requirements for blank analyses, while the remaining nine metals (As,
Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn) were analyzed by XRF techniques. In all of the XRF
analyses, blank sample analyses are not applicable. Both initial and ongoing calibration for Cd,
Hg, Se, and Ag analyses met the ARCS QA/QC specifications for both treated and untreated
sediments while for the remaining nine metals (As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn),
calibration information was not available. Detection limit information for metal analyses in
treated and untreated sediments was not available for verification except for Cd, Hg, Se, and
Ag where the detection limits were satisfactory. The precision for the metal analyses in treated
sediment was not satisfactory for As, but was satisfactory for the remaining elements. The
precision information for the metal analyses in untreated .sediment was satisfactory for all
elements. The matrix spike information for both treated and untreated sediment analyses were
satisfactory for four (Cd, Hg, Se, and Ag) of the thirteen elements while the remaining nine
metals (As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn) were analyzed by XRF techniques. In all of
the XRF analyses, results from matrix spike analyses were not applicable.
Of the seven conventional analyses, the accuracy information in both treated and
untreated sediments was satisfactory for TOC and was not available for total cyanide, and total
phosphorus. In the remaining four conventional analyses, accuracy was not applicable. In both
sediments, three of the seven conventionals (TOC, total cyanide, and total phosphorus) satisfied
QA/QC requirements for blank analyses. The blank information was unsatisfactory for oil and
grease, was not available for %TVS, and the blank information was not applicable for moisture
and pH. Both initial and ongoing calibration information was satisfactory for all conventional
analyses in both treated and untreated sediments except for % moisture, pH, and TVS where
calibration information was not available, and for TOC and oil and grease, where ongoing
-------
15
calibration information was not available. Detection limits were satisfactory for three (TOC,
total cyanide, and total phosphorus) of the seven conventional analyses in treated and untreated
sediments. Detection limits were unsatisfactory for oil and grease analyses in treated and
untreated sediments and were not applicable for % moisture, pH, and %TVS. The precision
information was satisfactory for pH, %TVS, and oil and grease analyses in treated, and for
%moisture, %TVS, and oil and grease analyses in untreated sediment. No precision information
was available for % moisture, TOC, total cyanide, and total phosphorus analyses in treated
sediment and for pH, TOC, total cyanide, and total phosphorus analyses in untreated sediments.
The matrix spike information for both treated and untreated sediment analyses were satisfactory
for total cyanide and total phosphorus, were unsatisfactory for oil and grease while for the
remaining four conventional analyses the matrix spike information was not applicable.
The accuracy objective was unsatisfactory for the PCB analyses in treated and untreated
sediments for Aroclor 1254. No accuracy information was available for the remaining three
Aroclor analyses in treated and untreated sediments. In water residue, the accuracy objective
for PCBs was satisfactory for Aroclor 1254 analyses only and could be used to represent the
whole PCB group. No accuracy information was available for the remaining three Aroclor
analyses in water residues. In water residues and in both treated and untreated sediments, the
blank analyses exceeded the detection limits specified in the QAPP indicating potential
contamination at the laboratory. Initial and ongoing calibration was satisfactory for all PCB
analyses in both treated and untreated sediments as well as in water residues. Detection limits
information were not available for PCB analyses in treated and untreated sediments, nor in the
water residues. In untreated sediment analyses, most PCB observations were below the
instrument detection limits, therefore it was not possible to calculate meaningful precision
information for PCB Aroclors, with the exception of Aroclor 1248 analyses, where precision
information satisfied QA/QC requirements. No precision information was available for PCB
analyses in treated sediments, except for Aroclor 1254 in treated sediment where it did not
satisfy QA/QC requirements. In the water residue, no PCB precision information was available.
The matrix spike for Aroclor 1254 was satisfactory for both sediments, and the water residue
analyses and could be used to represent the whole PCB group. The matrix spike information
for sediments and water residue analyses for Aroclor 1242, 1248, and 1260 were not available
for verification. The surrogate recoveries were satisfactory for PCB analyses in sediment and
residue analyses.
In ten of the sixteen PAH analyses in treated sediment and nine of the sixteen PAH
analyses in untreated sediments, the accuracy objective was satisfactory. No accuracy
information was available for six PAHs (naphthalene, acenaphthylene, acenaphthene, fluorene,
chrysene, and dibenzo(a,h)anthracene) analyses in treated and untreated sediment. The accuracy
objective was not satisfactory for benzo(k)fluoranthene in untreated sediment. Accuracy
information in water residue was unsatisfactory for naphthalene, acenaphthylene, acenaphthene,
phenanthrene, and benzo(a)pyrene. Accuracy was satisfactory for the rest of the PAH analyses
in water residues. In treated sediments and water residues, PAH analyses satisfied ARCS
specified QA/QC requirements for blank analyses. In all cases of untreated sediment analyses,
the blank analyses exceeded the detection limit specified in the QAPP. Calibration limits for
-------
16
PAH analyses met the ARCS QA/QC specifications for both treated and untreated sediments,
and also for water residue analyses. Detection limits information were not available for PAH
analyses in treated and untreated sediments, nor for the water residues. The precision
information was satisfactory for PAH analyses in both sediments except for naphthalene,
acenaphthylene, acenaphthene, fluorene, and benzo(a)pyrene analyses in treated sediment and
for naphthalene, acenaphthene, phenanthrene, and benzo(a)pyrene in water residue, where
precision was unsatisfactory. The matrix spike information was satisfactory for fifteen of the
sixteen PAH analyses in treated sediment, for five of the sixteen analyses in untreated sediment
and for eleven of the sixteen analyses in water residues. Surrogate recoveries were not
satisfactory for PAHs in the sediment and residue analyses.
SOIL TECH
The Soil Tech technology was evaluated by analyzing sediment samples and their treated
residues (treated sediments, water residues, and oil residues) for metals, conventionals, PCBs,
and PAHs. PCB and PAH analyses were performed for sediment and residues. The metals and
conventional analyses were performed for the sediment samples only.
In the majority of the cases studied, the accuracy objective was satisfactory for the metal
analyses in treated and untreated sediments. Of the thirteen metals analyzed, accuracy
information was not available for Ba, Se, and Ag. Four of the thirteen metal analyses (Cd, Hg,
Se, and Ag) satisfied QA/QC requirements for blank analyses, while the remaining nine metals
(As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and Zn) were analyzed by XRF techniques. In all of the
XRF analyses, blank sample analyses are not applicable. Both initial and ongoing calibration
for Cd, Hg, Se, and Ag analyses met the ARCS QA/QC specifications for both treated and
untreated sediments while for the remaining nine metals (As, Ba, Cr, Cu, Fe, Mn, Ni, Pb, and
Zn), calibration information was not available. Detection limits information for metal analyses
in treated and untreated sediments were not available for verification except for Cd, Hg, Se, and
Ag where detection limits were satisfactory. The precision information for the metal analyses
in treated sediment was not available for Se and Hg but was satisfactory for the remaining
elements with the exception of Cr, where precision information did not satisfy the QA/QC
requirements. The precision information for the metal analyses in untreated sediment was
satisfactory for all metal analyses. The matrix spike information were satisfactory for four (Cd,
Hg, Se, and Ag) of the thirteen elements for treated sediments and two (Cd, Hg) of the thirteen
elements for untreated sediments. The matrix spike information were unsatisfactory for Se and
Ag analyses in untreated sediments. The remaining nine metals (As, Ba, Cr, Cu, Fe, Mn, Ni,
Pb, and Zn) were analyzed by XRF techniques. In all of the XRF analyses, results from matrix
spike analyses were not applicable.
Of the seven conventional analyses, the accuracy information in both treated and
untreated sediments was satisfactory for TOC and was not available for total cyanide, and total
phosphorus. In the remaining four conventional analyses, accuracy was not applicable. In both
sediments, four of the seven conventionals (%TVS, TOC, total cyanide, and total phosphorus)
-------
17
satisfied QA/QC requirements for blank analyses, and the blank information was not applicable
for moisture and pH, while blank analyses was not satisfactory for oil and grease. Both initial
and ongoing calibration information was satisfactory for all conventional analyses in both treated
and untreated sediments, except for % moisture, pH, and %TVS where calibration information
was not available. Ongoing calibration information was not available for TOC and oil and
grease. Detection limits were satisfactory for three (TOC, total cyanide, and total phosphorus)
of the seven conventional analyses in treated and untreated sediments. Detection limits were
unsatisfactory for oil and grease and were not applicable for %moisture, pH, and %TVS. The
precision information was satisfactory for % moisture, %TVS, and oil and grease in treated
sediments. The precision information was satisfactory for %TVS, and oil and grease in treated
sediments. No precision information was available for the remaining conventional analyses in
treated or untreated sediments. The matrix spike information were satisfactory for oil and
grease, total phosphorus, and total cyanide in treated sediment analyses and for total phosphorus
in untreated sediment analyses. The matrix spike information were not available for oil and
grease and total cyanide in untreated sediment analyses. While for the remaining four
conventional analyses, the matrix spike information was not applicable.
The accuracy objective was satisfactory for the PCB analyses in treated sediments and
in oil residue analyses for Aroclor 1254 only and could be used to represent the whole PCB
group. The accuracy objective was unsatisfactory for the PCB analyses in untreated sediments
and in water residue analyses for Aroclor 1254. No accuracy information was available for the
remaining three Aroclor analyses in sediment or residue analyses. In both residues and in both
treated and untreated sediments, the blank analyses exceeded the detection limits specified in the
QAPP, except for Aroclor 1260 in oil residue. Initial and ongoing calibration was satisfactory
for all PCB analyses in both treated and untreated sediments, as well as in both water and oil
residues. Detection limit information was not available for PCB analyses in both sediments and
residues. In untreated sediment analyses, most PCB observations were below the instrument
detection limits, therefore, it was not possible to calculate meaningful precision information for
PCB Aroclors, with the exception of Aroclor 1248 analyses, where precision information
satisfied QA/QC requirements. No precision information was available for PCB analyses in
treated sediment, except for Aroclor 1254, where it did not satisfy QA/QC requirements. No
precision information was available for PCB analyses in oil and water residues, except for
Aroclor 1248 in oil residue, where precision was satisfactory. The matrix spike for Aroclor
1254 was satisfactory for both sediments and the oil residue analyses and could be used to
represent the whole PCB group. The matrix spike for Aroclor 1254 was unsatisfactory for the
water residue analyses, and the matrix spike information for both sediment and residue analyses
for Aroclor 1242,1248, and 1260 were not available for verification. The surrogate recoveries
were satisfactory for PCB analyses in sediment and residue analyses, except for water residue
where surrogate information was not available.
In eight of sixteen PAH analyses in treated and untreated sediments, the accuracy
objective was satisfactory. No accuracy information was available for six PAHs (naphthalene,
acenaphthylene, acenaphthene, fluorene, chrysene, and dibenzo(a,h)anthracene) analyses in both
treated and untreated sediments. The accuracy objective was not satisfactory for benzo(k)
-------
18
fluoranthene in treated or untreated sediments nor for benzo(g,h,i)perylene in untreated
sediment. Accuracy information was satisfactory for the PAH analyses in water and oil
residues. In treated and untreated sediments and water residues, PAH analyses satisfied ARCS
specified QA/QC requirements for blank analyses. In all cases of oil residues, the blank
analyses exceeded the MDL. Calibration limits for PAH analyses met the ARCS QA/QC
specifications for both treated and untreated sediments as well as water and oil residue analyses.
Detection limit information was not available for PAH analyses in treated and untreated
sediments nor for water and oil residues. In untreated sediment and oil residues, the precision
information was satisfactory for all PAH analyses, except for acenaphthene and acenaphthene
in untreated sediment, and naphthalene in oil residues, where no precision information was
available. In treated sediments, the precision information was satisfactory for naphthalene,
acenaphthylene acenaphthene, fluorene, phenanthrene, and anthracene, and was unsatisfactory
for the remaining PAH analyses. In water residues, no precision information was available for
any of the PAH analyses. The matrix spike information was satisfactory for twelve of sixteen
PAH analyses in treated sediment, and for thirteen of the sixteen analyses in untreated sediment
and ten of the sixteen analyses in water and all analyses in oil residues. Surrogate recoveries
were unsatisfactory for PAHs in either sediment and oil residue analyses but were satisfactory
in water residue.
RETEC
The RETEC technology was evaluated by analyzing sediment samples and their treated
residues (water residues and oil residues) for metals, conventional, PCBs and PAHs. PCB and
PAH analyses were performed for sediment and residues. The metals and conventional analyses
were performed for both sediment samples and water residues.
In a majority of the cases studied, the accuracy objective was satisfactory for the metal
analyses in treated and untreated sediments. Of thirteen metals analyzed, accuracy information
was not available for Ba, Se, and Ag. In both treated and untreated sediments, ten of the
thirteen metal analyses (As, Cd, Cr, Cu, Fe, Pb, Mn, Ni, Hg, and Zn) satisfied ARCS specified
QA/QC requirements for accuracy. The accuracy objective was satisfactory for all metal
analyses in water, except for Se, where accuracy did not satisfy QA/QC requirements. Four of
the thirteen metal analyses (Cd, Hg, Se, and Ag) satisfied QA/QC requirements for blank
analyses. The remaining nine metal analyses (As, Ba, Cr, Cu, Fe, Pb, Mn, Ni, and Zn) were
analyzed by XRF techniques. In all of the XRF analyses, blank sample analyses are not
applicable. In water residue, blank analyses were satisfactory for all metals except for Fe, Mn,
and Se, where blank analyses exceeded the detection limits specified in the QAPP, and for Ba,
where no information regarding blank analyses was available. Both initial and ongoing
calibration met the ARCS QA/QC specifications for Cd, Hg, Se, and Ag for both treated and
untreated sediments, and for all metals in water residue analyses. While in both treated and
untreated sediments the remaining nine metals (As, Ba, Cr, Cu, Fe, Pb, Mn, Ni, and Zn),
calibration information were not available. Detection limits information for metal analyses in
treated and untreated sediments were not available for verification, except for Cd, Hg, Se, and
-------
19
Ag, where detection limits were satisfactory. Detection limits for metal analyses in water
residue were satisfactory, except for Mn, Se, and Zn, where detection limits exceeded the
QA/QC requirements. The precision information for the metal analyses in treated and untreated
sediments, and in water residue was satisfactory for all elements, except for Hg in treated
sediment, and Se and Hg in water residue analyses, where precision information did not satisfy
QA/QC requirements. The matrix spike information for treated sediment analyses were
satisfactory for Cd, Hg, and Ag, and was not satisfactory for Se. The matrix spike information
for untreated sediment analyses were satisfactory for Cd and Hg, and was not satisfactory for
Se and Ag. The remaining nine metals (As, Ba, Cr, Cu, Fe, Pb, Mn, Ni, and Zn) were
analyzed by XRF techniques for treated and untreated sediment. In all of the XRF analyses,
matrix spike analyses are not applicable. The matrix spike information for water residue
analyses was satisfactory for all metals except for Ag where matrix spike information did not
satisfy QA/QC requirement.
Of the seven conventional analyses in both treated and untreated sediments, accuracy
information was satisfactory for TOC, and was not available for total cyanide, or total
phosphorus. In the remaining four conventional analyses accuracy was not applicable. Of ten
conventional analyses in water residue, accuracy information was not available for TOC, total
cyanide, total phosphorus, and conductivity. In the remaining seven conventional analyses
accuracy was not applicable. In both treated and untreated sediments and in water residue
analyses, %TVS, oil and grease, TOC, total cyanide, and total phosphorus satisfied QA/QC
requirements for blanks. Also, the blank information was satisfactory for total solids and total
suspended solids in water residue analyses. The blank information was not applicable for the
remaining conventional analyses in sediment and water residue analyses. Both initial and
ongoing calibration information was satisfactory for all conventional analyses in both sediment
and water residue, except for % moisture (in sediment), pH, and TVS, TSS, TS where
calibration information was not available, and for TOC and oil and grease, where ongoing
calibration information was not available. Detection limit information was not available in both
treated and untreated sediments and in water residue for oil and grease, TOC, total cyanide, and
total phosphorus, and was not applicable for the remaining conventional analyses. In treated
sediment, the precision information was not satisfactory for oil and grease and no precision
information was available for total cyanide. In untreated sediment, the precision information
was not satisfactory for total cyanide, and no precision information was available for TOC. The
precision information was satisfactory for the remaining five conventional analyses in treated and
untreated sediments. In water residue, the precision information was; satisfactory for all the
conventional*, except for moisture, where no precision information was available. The matrix
spike information was not available for oil and grease, and was satisfactory for total cyanide and
total phosphorus in treated sediment analyses. The matrix spike information was not available
for oil and grease, total cyanide, and total phosphorus in untreated sediment analyses. The
matrix spike information was satisfactory for oil and grease, total cyanide, and total phosphorus
in water residue analyses. The matrix spike information for the remaining conventional analyses
was not applicable for sediment and water residue analyses.
-------
20
The accuracy objective was unsatisfactory for the PCB analyses in treated sediments,
untreated sediments, and oil residue for Aroclor 1254 and could be used to represent the whole
PCB group. No accuracy information was available for the remaining three Aroclor analyses
in treated and untreated sediments. No accuracy information was available for PCB analyses
in water residues. In both sediments and residues, the blank analyses exceeded the detection
limits specified in the QAPP. Both initial and ongoing calibration for PCB analyses met the
ARCS QA/QC specifications for both treated and untreated sediments, as well as for water and
oil residues. Detection limit information was not available for PCB in either sediments or
residue analyses. The precision information for the PCB analyses in treated and untreated
sediment was satisfactory for Aroclor 1254. In all remaining analyses, precision information
was not available. The matrix spike was satisfactory for Aroclor 1254 in treated sediment and
in oil residue analyses, and could be used to represent the whole PCB group. The matrix spike
information was not available for the remaining Aroclors in treated sediment and oil residues.
The matrix spike information was not available for PCB analyses in untreated sediment and in
water residues. The surrogate recoveries were satisfactory for PCB analyses in sediment and
residue analyses.
In ten of the sixteen PAH analyses in treated sediments and in seven of the sixteen PAH
analyses in untreated sediments, the accuracy objective was satisfactory. No accuracy
information was available for six PAHs (naphthalene, acenaphthylene acenaphthene, fluorene,
chrysene, dibenzo(a,h)anthracene) analyses in treated and untreated sediment. The accuracy
objective was not satisfactory for benzo(k)fluoranthene, benzo(a)pyrene, and benzo(g,h,i)
perylene in untreated sediment. Accuracy information was satisfactory for fourteen of the
sixteen PAH analytes in oil residue. Accuracy information was unsatisfactory for PAH analyses
in water residue, except for benzo(k)fluoranthene, indeno(l,2,3,c,d)pyrene,
dibenzo(a,h)anthracene. The blank analyses for the PAHs in treated and untreated sediment was
satisfactory in all cases except for acenaphthylene, acenaphthene, fluorene, phenanthrene, and
anthracene. In water residues, all PAH analyses satisfied ARCS specified QA/QC requirements
for blank analyses. In all oil residues, the blank analyses exceeded the detection limit specified
in the QAPP. Both initial and ongoing calibration information for PAH analyses met the ARCS
QA/QC specifications for both treated and untreated sediments, and also for water and oil
residue analyses. Detection limit information was not available for PAH analyses in either
sediments or residues. The precision information was satisfactory for PAH analyses in treated
sediments, except for benzo(k)fluoranthene, where precision did not satisfy QA/QC
requirements. The precision information was satisfactory for PAH analyses in untreated
sediments except for acenaphthylene and acenaphthene, where precision information was not
available, and for benzo(k)fluoranthene, where precision did not satisfy QA/QC requirements.
The precision information was satisfactory for PAH analyses in oil residue, except for
benzo(k)fluoranthene, where precision information did not satisfy QA/QC requirements. In
water residue, precision was unsatisfactory for PAH analyses except for benzo(k)fluoranthene,
indeno(l,2,3,c,d)pyrene, and dibenzo(a,h)anthracene, where precision was satisfactory. The
matrix spike information was satisfactory for ten of the sixteen PAH analytes in treated
sediment, for fourteen of the analytes in untreated sediment, for thirteen of the analytes in oil
residues, and for three of the analytes in water residues. Surrogate recoveries were satisfactory
-------
21
for PAHs in both treated and untreated sediments as well as for oil and water residue analyses.
Summary
Based on the compliance with the ARCS QA/QC requirements, SAIC was capable of
supplying acceptable results for metals, conventionals, PCBs, and PAHs. The results received
for all four technologies satisfied ARCS QA/QC requirements.
An examination of results of the bench scale technology demonstration data set indicates,
that SAIC could have successfully provided acceptable data for all parameters. The data user
should be aware that some QA/QC discrepancies were identified, as indicated by subscript 1 and
2 flags in Table 3.
-------
APPENDICES A and D
are not included in this report.
Copies are available from GLNPO upon request.
-------
APPENDIX B
QA/QC Sample Rating Factors
-------
CATEGORY
RATING FACTORS
CATEGORY
SCORE ACCEPTABILITY LEVEL
Accuracy
Precision
Certified Reference Material
Analytical Replicate
3
3
Acceptable = 3
Acceptable = 3
Spike Recovery
Blanks
Miscellaneous
Matrix Spike = 3
Surrogate Spike (organics) = 3
Blanks = 3
Instrument Calibration (initial) = 3
Instrument Calibration (on going) = 2
Instrument Detection Limit = 3
Acceptable = 3
(organics) = 6
Acceptable = 3
Acceptable = 3
-------
APPENDIX C
Data Verification Flags
-------
A = Accuracy Problem
AO = no standard available/no information available
A, = accuracy limit for the reference materials exceeded
A, = accuracy is not applicable
B = Blank Problem
BQ = no information available
B; = reagent blank value exceeded MDL
B, = blanks are not applicable
C = Calibration Problem
CQ = no information available
C| = initial calibration problem
Cj = on-going calibration problem
Q = no information on initial calibration
C6 = no information on on-going calibration
C, = on-going calibration is not applicable
D = Detection Limit Problem
D0 = no information available
D, = detection limit exceeded
D9 = detection limit is not applicable
-------
H = Holding Times Exceeded
P = Precision Problem
P0 = no information available
P, = precision limit for analytical replicate exceeded the QA/QC
requirements
P3 = MSD exceeded the QA/QC requirement
P9 = precision is not applicable
S = Spike Recovery Problem
S0 = no information available on spike
S, = limit of matrix spike recovery exceeded
S2 = limit of surrogate spike recovery exceeded
Si = no information available on matrix spike recovery
S6 = no information available on surrogate spike recovery
S, = spike recovery not applicable
------- |