United States
Environmental Protection
Agency
Great Lakes
National Program Office
77 West Jackson Boulevard
Chicago, Illinois 60604
EPA 905-R94-023 (? , /
October 1994
&EPA Assessment and
Remediation
Of Contaminated Sediments
(ARCS) Program
BENCH-SCALE EVALUATION
OF SEDIMENT TREATMENT
TECHNOLOGIES SUMMARY REPORT
United States Areas of Concern
ARCS Priority Areas of Concern
printed on recycled paper
-------
Bench-Scale Evaluation of Sediment
Treatment Technologies
Summary Report
t j>v Prepared by
Cs^
Evelyn Meagher-Hartzell and Clyde 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 Boulevatd, 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 (USACE) 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 USACE
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). Evelyn Meagher-Hartzell and Clyde Dial of
SAIC were the principal authors 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 Sediment Treatment
Technologies Summary Report," EPA 905-R94-011, 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 was
responsible for undertaking a 5-year study and demonstration program for the remediation of contami-
nated sediments. GLNPO initiated the 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 were conducted
as part of the ARCS Program. A series of bench-scale studies using SoilTech's Anaerobic Thermal
Process Technology, Resources Conservation Company's B.E.S.T.* Solvent Extraction Process,
ReTeC's Thermal Desorption Technology, and Zimpro's Wet Air Oxidation Process are the subject of
this report. The specific objectives for this effort were to determine process destruction or 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 SoilTech ATP* technology was tested using sediment samples obtained from the Buffalo
and Grand Calumet Rivers. The concentration of the contaminants of concern in the sediments were
0.34 and 10.7 mg/kg PCBs respectively and 8.7 and 235 mg/kg PAHs respectively. The PCB removal
from the Grand Calumet River was 72 percent. Because of the very low concentration of PCBs in the
Buffalo River sediment it was not possible to make an effective assessment of the PCB removal. The
PAH removal from both sediments was 99 percent.
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 sediments were 0.3 to 22 mg/kg PCBs and 3 to 220 mg/kg PAHs. 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.
The ReTeC Holo-Flite™ Screw Processor was tested using a sediment sample obtained from
the Ashtabula River. The concentrations of the contaminants of concern in the sediment were 14.6
mg/kg PCBs and 6.1 mg/kg PAHs. PCB and PAH concentrations of <0.6 and <2.4 mg/kg, respectively,
were found in the treated solids. This corresponds to PCB and PAH removals of >96 and >60 percent,
respectively.
The Zimpro Wet Air Oxidation Process was tested using a sediment obtained from the Grand
Calumet River. The concentrations of the contaminants of concern in the sediment were 11.9 mg/kg
PCBs and 266 mg/kg PAHs. PCB and PAH concentrations of 8.5 and <2.84 mg/kg, respectively, were
found in the treated solids. This corresponds to PCB and PAH removals of 29 percent and >98.9
percent, respectively.
-------
CONTENTS
Section Page
w^^^_^^«_ ^^^_^^^
Disclaimer i
Foreword ii
Abstract jjj
Figures v
Tables vi
1.0 Introduction 1
1.1 Background 1
1.2 Purpose and Scope 2
1.3 Approach 4
2.0 Experimental Design 19
2.1 Description of the Phased Approach 19
2.2 Characterization of the Various Sediments and Residuals 20
2.3 Sampling and Analysis 21
3.0 Results and Discussion 21
3.1 Summary of Phase I Results 21
3.2 Summary Phase II Results 24
3.3 Quality Assurance/Quality Control 28
Appendix - Data Verification Report for the ARCS Program
IV
-------
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 Ashtabula River Sample Location 9
6 ReTeC Process Flow Diagram 11
7 Flow Diagram for the B.E.S.T.* Process 13
8 Flow Diagram for the Zimpro Wet Air Oxidation Process 15
9 Simplified Anaerobic Thermal Process Flow Diagram 18
-------
TABLES
Number
1 Sediment Characterization Data 3
2 Sediment Treated by Each Treatment Technology 19
3 Parameters for Analysis of ARCS Program Technologies 20
4 Analytical Matrix and Sample Identification 22
5 Removal Efficiencies for Total PCBs 25
6 Removal Efficiencies for Total PAHs 25
7 Battelle Data - Removal Efficiencies for Other Parameters 26
8 PAH and PCB Concentrations in Treated Waters and Produced Oils 27
9 Mass Balances 28
VI
-------
1.0 INTRODUCTION
1.1 Background
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 contami-
nated sediments. Five areas were specified for priority consideration in locating and conducting
demonstration projects: Ashtabula River, Ohio; Buffalo River, New York; Grand Calumet River/Indiana
Harbor Canal, Indiana; Saginaw River, Michigan; and Sheboygan River, Wisconsin. 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 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 was
also utilized. The Engineering/Technology (ET) Work Group was charged with overseeing the develop-
ment and application of the bench-scale and pilot-scale tests.
For purposes of the technology evaluations conducted under the ARCS Program, the term
"bench-scale" refers to laboratory-based tests that utilize glassware simulations of the central reactions
of a particular process. Bench-scale tests are typically conducted on several kilograms or less of
material. The term "pilot-scale" refers to a field-based demonstration utilizing a scaled-down version of
a treatment technology that more closely represents the physical and operational characteristics of a full
size processor. Under the ARCS Program, bench-scale tests were conducted on different treatment
technologies in order to demonstrate if the fundamental chemical or physical reactions could be
successfully applied to contaminated sediments from the Great Lakes, and to identify technologies that
were feasible for demonstration at the pilot scale.
Science Applications International Corporation (SAIC) was contracted to provide technical
support to the ET Work Group. As part of this effort, SAIC performed seven treatability studies on
sediments obtained from four locations in and around the Great Lakes to evaluate the ability of the
technologies to remove organic contaminants, specifically polychlorinated biphenyls (PCBs) and
polynuclear aromatic hydrocarbons (PAHs). PCB and PAH process extraction or destruction efficien-
-------
cies; mass balance closures for solids, water, oil, PCBs, and PAHs; and changes in the concentration
of metals, oil and grease, and several other parameters were evaluated.
1.2 Purpose and Scope
SAIC and its subcontractors conducted seven treatability tests for the ARCS Program on four
different sediments using four treatment technologies: Low Temperature Thermal Desorption Process
(Remediation Technologies, Inc. (ReTeC)), Anaerobic Thermal Process (ATP*) Technology (SoiFTech),
Wet Air Oxidation (Zimpro Passavant), and the Basic Extractive Sludge Treatment (B.E.S.T.*) Solvent
Extraction Process, (Resource Conservation Company (RCC)). The four sediments used during the
treatability studies were obtained from the Ashtabula River, Buffalo River, Grand Calumet River, and
Saginaw River. The contaminants found in these four sediment samples are commonly encountered in
contaminated sediments throughout the Great Lakes, as well as in other areas of the United States and
Canada. These contaminants include oil and grease, heavy metals, pesticides, PCBs and PAHs. The
four sediment samples represented a wide range of contaminant concentrations, as shown in Table 1.
The purpose of this study was to evaluate the feasibility, cost, and effectiveness of various
technologies for treating and removing organic contaminants (PCBs and PAHs) from the designated
sediments. The effects of the technologies on other parameters, including oil and grease, heavy metals
and cyanide, were assessed. The specific objectives of the study were as follows:
• To record observations and data to predict a technology's full-scale performance
• To take samples during the tests and conduct analyses sufficient to allow for calculation of
mass balances for oil, water, solids, and other compounds of interest
• To calculate the destruction or removal efficiencies of target compounds
• To obtain treated solids (300 g dry basis), water, and oil for independent analysis.
This report summarizes the approach used and results obtained during bench-scale (laboratory-
based) testing of the different technologies. According to the various vendors, the results obtained
during testing mirror results achievable during full-scale (field-based) operations and thus an accurate
estimate of the performance associated with a full-scale application of each technology is possible.
Since none of these technologies have been implemented on a full scale with these sediments, this
correlation between laboratory and field operations is unproven. Also, since this study is only one part
of a much larger program and is not intended to evaluate the treatment of the sediments completely,
data interpretations are limited to comparisons of technology performance between the four
-------
Table 1. Sediment Characterization Data
(mg/kg, dry, unless specified)
Total PCBs
Total PAHs
Arsenic
Barium
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Zinc
Total Cyanide
Total Phosphorus
Moisture, %, as received
Oil & Grease
TOC, % weight
Total Volatile Solids, %
pH, S.U., as received
Buffalo River1
0.33
9.31
13.0
405
2.0
118
72
44200
105
680
0.51
44
0.70
0.30
185
5.6
1120
55.1
5980
1.88
4.57
7.72
Saginaw Bay
21.9
2.70
2.2
322
4.1
107
59
7870
46
165
0.17
58
<0.03
0.84
140
4.1
740
24.0
1350
0.83
2.09
7.30
Grand Calumet
River2
12.9
244
<21
296
7.9
1480
230
182000
706
2390
1.5
74
<3.7
4.9
2840
22
2650
51.2
45700
18.4
15.0
7.61
Ashtabula
River
14.6
6.1
21
903
3.1
591
34
42600
59
559
0.19
53
0.91
0.19
234
1.5
1290
35.6
1004
2.00
7.64
7.88
i - Average or two analyses
2 - Average of three analyses
technologies studied. Material balances estimating mass distributions were performed when possible.
This report presents a summary of the various technology tests conducted by SAIC for the ARCS
Program. Detailed reports on each individual technology are available from GLNPO upon request.
Section 3.4 provides the detailed references for the technology-specific reports.
-------
1.3 Approach
1.3.1 Site Names and Locations for Each Sediment Sample
GLNPO collected sediments for study from the following areas around the Great Lakes:
Ashtabula River, Ohio; Buffalo River, New York; Grand Calumet River/Indiana Harbor Canal, Indiana;
Saginaw River, Michigan; and Sheboygan River, Wisconsin. SAIC was contracted to treat sediments
from the Grand Calumet River, Buffalo River, Ashtabula River, and Saginaw River using four technolo-
gies: RCC's B.E.S.T.* Solvent Extraction Technology, Zimpro's Wet Air Oxidation Process, Soil Tech's
ATP® Technology and ReTeC's Low Temperature Thermal Desorption Process. A map provided in
Figure 1 shows the ARCS Priority Areas of Concern. Specifics of the sample location for the Buffalo
River, Saginaw River, Grand Calumet River and Ashtabula River are shown in Figures 2, 3, 4 and 5,
respectively.
1.3.2 Sediment Acquisition and Homogenization
Prior to conducting the treatability study using each technology, the sediment was homogenized
and stored under refrigeration by the U.S. EPA Environmental Research Laboratory in Duluth,
Minnesota. Samples of the homogenized sediments were sent to SAIC by the Duluth laboratory.
Sediments were then transferred by SAIC to the appropriate technology vendor. The technology
vendors performed a series of standard tests on these sediments (Phase I) to determine if the
sediments they were scheduled to treat were compatible with their processes and to determine
optimum testing conditions and procedures for the treatability study (Phase II). Additional sediment was
later forwarded to the vendors by SAIC for the Phase II testing.
1.3.3 Sediment Characterization
SAIC was responsible for the physical and chemical characterization of the raw sediments used
during the tests. Under SAIC's direction, the sediment and residuals were analyzed by Battelle Marine
Sciences Laboratory in Sequim, WA. Table 1 provides characterization data pertaining to the four
sediments.
-------
Ol
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
US ENVHCMUENTM. PHOTIC TON AOENCY
GREAT LAKE* NATKMN. PfVXMMI OFFCt
* Assessment and Remediation of Contaminated Sediments
Figure 1. ARCS Priority Areas of Concern.
-------
Ufa
Erit
o>
Buffalo River
Sediment sample point
t-t-n
Figure 2. Buffalo River Sample Locations.
-------
Saginaw River and Bay
Sediment sample point
Figure 3. Saginaw River Sample Location.
-------
Sediment sample point
GRAND CALUMET RIVER
t«
Figure 4. Grand Calumet River Sample Location
-------
Ashtabula River
The sediment sample was a composite comprised of
subsamples taken throughout the Ashtabula River system
Figure 5. Ashtabula River Sample Location.
-------
1.3.4 Technology Descriptions
1.3.4.1 ReTeC's Low Temperature Thermal Desorption Technology
The ReTeC Low Temperature Thermal Desorption Technology separates the contaminants
present in a solid matrix through volatilization. This technology can be used independently or as part of
a multi-stage treatment train. Volatilized contaminants are subsequently condensed to yield an oily
liquid which can then be further treated through an incineration or other destructive process.
The Holo-Flite™ Screw Processor is the primary component of the ReTeC Thermal Desorption
Technology. The Holo-Flite™ Screw Processor is an indirect heat exchanger used to heat, cool, or dry
bulk solids/slurries. It consists of a jacketed trough housing a double-screw mechanism. Heated fluid
continuously circulates through the hollow flights of the screw augers to elevate the temperature of the
soils. This fluid travels the length of the screws and returns to the heater through the center of each
shaft. The rotation of the screws promotes the movement of the material forward through the
processor.
Volatilized organics are removed from the treatment chamber by means of an induced draft fan
and routed to an off-gas control system. The atmosphere in the treatment chamber is controlled during
treatment to ensure that oxidation of the volatilized materials does not occur. A three-stage approach is
used to control these off-gases. Initially, gas-entrained particulate matter is collected using a series of
cyclones. The volatilized moisture and organics are then removed using a water-cooled condenser.
The remaining non-condensible gas is then passed through a canister containing activated carbon for
VOC control.
A process flow diagram of the 1000 pound-per-hour (460 kilogram-per-hour) thermal desorption
system which was used for Phase II testing is provided in Figure 6. ReTeC conducted Phase II testing
at processing conditions [temperature (525°C), screw rotation rate (0.75 rpm), and residence time (75
min.)] determined following Phase I. Approximately 225 kg of Ashtabula River sediment were used
during the single Phase II run. Because a significant amount of the feed would have been lost by
utilizing the bucket elevator (i.e., relative to the total amount of material being processed), the sediment
was hand-fed into the unit.
To prevent liquids present in the sediment from passing though the system with less than
optimal retention times, decanted sediment solids were initially fed into the unit to produce a "dam"
10
-------
TO
A
lilt HI
GtMHARiH
coinuiiions
MONITORING
02
CONTINUOUS
iRINC
THC
MOHfTORINC —T>
FtEO
HOPPER
0
imiiiL
A (5>
SA1T
TREATED
SOIL
Figure 6. ReTeC Process Flow Diagram.
(Source: ReTeC, Inc.)
-------
effect in the screw processor. During actual operation, dry sand may be used to create the dam effect
in the screw processor. This will keep more liquid feeds from flowing too quickly through the screw
conveyor. After the first three and a half pails of solids were added to the unit, operators began
introducing water with each scoop of solids fed into the unit. All the water associated with the original
sediment was fed through the unit. Approximately 80 kg of dry treated solids were generated.
1.3.4.2 RCC's B.E.S.T* Solvent Extraction Process
The B.E.S.T.* Solvent Extraction Process employs triethylamine (a solvent) to extract organic
contaminants from contaminated media. Triethylamine is an aliphatic amine produced by the reaction
of ethyl alcohol and ammonia. When employed, a single-phase, homogenous extraction solution
containing triethylamine, water, and oil is produced. Any organic contaminants present in the feed
material are solvated in the water and oil portion of the extraction solution. A flow diagram for the
B.E.S.T.* process is shown in Figure 7.
Because triethylamine is inversely miscible (i.e., at temperatures below 18°C, triethylamine is
completely miscible with water, while at temperatures above 18°C, triethylamine is only partially miscible
with water) and can simultaneously solvate oil and water, triethylamine can be used to treat wastes
containing both contaminated oils and water.
Since triethylamine is soluble in water at temperatures below 18°C, the first extraction is
conducted near 4°C. The extract obtained from this reaction will contain most of the water present in
the feed material. Since the solubility of oil in triethylamine increases at temperatures above 54°C,
subsequent extractions conducted at these temperatures enhance the removal of oil from the contami-
nated solids. The extracts from these later reactions contain mostly oil and very little water.
The extracts generated undergo additional treatment to separate them into oil, water, and
triethylamine. If sufficient water is present in the extract generated during the initial "cold" extraction,
the solution can be separated into two phases, a triethylamine/oil phase and a water phase, by heating
the liquid to a temperature above the miscibility limit (54°C). The triethylamine/oil phase is combined
with the extract produced from the subsequent extractions.
In a full-scale unit, the triethylamine is recovered from the triethylamine/oil phase by steam
stripping and is recycled directly to the extraction vessels for the solvent recovery portion of the
12
-------
Solvent Recovery
Recycled
Solvent
Waste
i
Extraction
40°F(5°C)
B.E.S.T.
Solvent / Oil
1
Separation
130°F(55°C)
i
Solvent / Water
Solvent Recycle
Evaporator
on
Solvent Recycle
Water I Water
Stripper
Solvent Recycle
Solids
Solids
Dryer
Figure 7. Flow Diagram for the B.E.S.T.* Process.
(Source: RCC, Inc.)
-------
process. Residual triethylamine in the water and oil products is usually low. Triethylamine remaining in
the treated solids may be removed by indirect heating with steam. Typically, residual triethylamine
within the treated solids biodegrades readily. Since the full-scale unit operates in a closed loop with
one small vent for removal of non-condensing gases, air emissions are minimal. However, RCC
typically uses a water scrubber and activated carbon on this vent to minimize triethylamine releases.
Phase II procedures and associated equipment used are described in the following paragraphs.
With slight variations, the same procedure was used to treat each of the sediments.
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 1°C. Each sample's pH was adjusted
using sodium hydroxide (NaOH) and 2.7 L of chilled triethylamine (2 percent H2O). While immersed in
the cooling bath, the sample was mixed with the NaOH and the chilled triethylamine using a air-driven
prop mixer. 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.
For the second extraction, the samples were heated to 53 to 60°C. The mixture was kept
heated while mixing was in progress. Mixing was conducted with a pneumatic mixer for approximately
20 minutes. The solvent/oil was poured off and held for later combination with the solvent/oil portion
from the decantation procedure. 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 104°C in a forced-draft oven.
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 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. 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. The elevated pH is necessary to ensure that
the majority of the triethylamine remains in the volatile molecular form. The bulk of triethylamine was
14
-------
removed 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.
7.3.4.3 Zimpro's Wet Air Oxidation
Zimpro's Wet Air Oxidation Process employs elevated temperatures and pressures to oxidize
inorganic and organic contaminants under aqueous conditions. Compressed air or pure oxygen
generally serve as the oxidizing agent in the wet air oxidation process. Temperatures ranging from 175
to 320°C are usually employed. System pressures of 2.0 MPa or greater are common. A flow diagram
for the Zimpro wet air oxidation process is shown in Figure 8.
AIR
HIGH
PRESSURE
PUMP
FEED
EXCHANGER
MOT WATER
OR STEAM
REACTOR
AIR COMPRESSOR
VENT
GAS
SEPARATOR
OXIDIZED
LIQUOR
Figure 8. Flow Diagram for the Zimpro Wet Air Oxidation Process.
(Source: Zimpro Passavant Environmental Systems, Inc.)
In processing an aqueous waste, the waste stream containing the oxidizable material is first
pumped to the system using a positive displacement, high pressure pump. The pressurized discharge
15
-------
from the high-pressure pump is combined with the air stream from the air compressor, forming a two-
phase stream. Next the air/waste stream passes through the feed/effluent heat exchanger, recovering
heat from the hot, oxidized effluent. The heated mixture is then routed through an auxiliary heat
exchanger, if needed. A vertical bubble-column is commonly used as the reactor to provide the
required hydraulic detention time to effect the desired reaction. The reactor contents are mixed by the
action of the gas phase rising through the liquid. As the gas phase rises and mixes with the liquid,
oxygen is dissolved into the liquid. The reactor is sized to allow the oxidation reactions to proceed to
the desired level. The desired reaction may range from a mild oxidation, which requires a few minutes,
to total waste destruction, which requires an hour or more of detention time.
The oxidized liquid, oxidation product gases, and spent air leave the reactor and are routed
through the shell side of the feed/effluent heat exchanger. A cooler can achieve additional cooling, if
necessary. The cooled reactor effluent is throttled through a pressure control value into the process
separator where the reactor effluent is separated into a gaseous stream and a liquid stream. The
gaseous stream from the process separator is routed through an off-gas cooler. The liquid stream is
pumped beyond the treatment system's boundary limits. Further treatment of these oxidized liquids by
a biological system may be required prior to discharge into the final receiving system (publicly-owned
treatment work, river, lake, etc.).
The Phase II wet air oxidation tests were performed in titanium-stirred laboratory autoclaves,
each having a capacity of 3.78 L. The autoclaves were equipped with a magnetic stirring device to
help the oxygen diffuse into the liquid and keep the solids in suspension. The stirrer remained on
throughout the oxidation.
The as-received feed samples were removed from their jars and placed in a stainless-steel
mixing bowl. A continuous mixer was used to stir the samples to obtain homogeneity. The feed
material was divided into seven portions for testing and two samples for the analysis of the raw feed.
Separate stirred autoclave oxidations were performed using six of the seven samples. The samples
were diluted, using HPLC grade water, to produce an autoclave feed sample with a suspended solids
concentration of approximately 10 percent. Ten percent suspended solids was used to simulate the 10
to 20 percent concentrations that would be used in a commercial unit to allow the sediment to be
pumped at pressure. This does not mean that all the additional water needs to be supplied as feed
water; some can be recycled from the filtrate after treatment. Based on the Phase I test, a reactor
temperature of 280°C and a hydraulic detention time of 90 minutes was selected for the Phase II tests.
These conditions were selected to provide a balance between PAH destruction and process economics.
16
-------
The autoclaves were charged with the sediment slurry and sufficient compressed air to result in
excess oxygen remaining following oxidation. The charged autoclaves were then heated to the desired
oxidation temperature by electrical heating bands and held at that temperature for the specified reaction
time. Immediately following the oxidation, the autoclaves were cooled to room temperature by internal
water cooling coils and then depressurized.
1.3.4.4 Soil Tech's /ATP® Technology
SoilTech's ATP* Technology employs a modified rotary kiln to thermally separate oil and
hazardous organic waste from contaminated soils and sludges. As shown in Figure 9, the full-scale
ATP* has four treatment zones: the preheat zone, the retort zone, the combustion zone, and the
cooling zone. Organic matter introduced to the ATP* unit undergoes distillation and thermal cracking
within the retort zone. Partially treated solid product entering the rotating, cylindrical combustion zone
receives further treatment at temperatures near 675°C to oxidize the remaining coke present on the
inorganic soil fraction. The volatiles released during treatment are vented into a condensing system,
where aqueous and liquid hydrocarbon (oily) streams are collected. Gaseous combustion products
from the combustion zone of the full-scale ATP* are treated in a flue gas handling system and vented
to the atmosphere. A baghouse, wet scrubber, or combination of the two may be used to remove dust,
trace particulates, or acid gases from flue gases exiting the cooling zone before venting to the
atmosphere. Activated carbon treatment of system off-gases may be required. Tailings exiting the
ATP*'s cooling zone are cooled by water addition.
During Phase II testing, up to 3 kg of material was placed into a steel cylindrical rotating retort
chamber (Batch Pyrolysis Unit). The chamber was rotated at 4 (rpm) and heated by electrical heat
tracing to temperatures up to 700°C. During operation, a vacuum of 1 inch of H2O was maintained in
order to extract distillate vapors from the chamber continuously.
The cooled solids were placed in the aerobic Batch Combustor where they were showered in
an air stream at temperatures of approximately 675°C. The rotational speed was maintained at 4 rpm
and the offgases were continuously monitored for O2, CO2, NO2, and CO using gas analyzers. The run
was terminated when the O2 concentrations achieved levels representative of ambient air, indicating
combustion was complete.
17
-------
00
to
9
SIUPUHED ANACROBIC FHCRUAL
PROCESS fio* UMCRAM
Figure 9. Simplified Anaerobic Thermal Process Flow Diagram.
(Source: SoilTech, Inc.)
-------
1.3.5 Technologies Applied to Each Sediment
Of the seven treatability tests conducted, four different technologies were tested and four types
of sediments were treated. Table 2 differentiates the seven tests according to the technology tested
and sediment treated.
Table 2. Sediment Treated By Each Treatment Technology
Sediment Tested
Ashtabula
Treatment Technology River
RCC's B.E.S.T.® Solvent Extrac-
tion Process
SoilTech's ATP* Technology
Zimpro's Wet Air Oxidation Pro-
cess
ReTeC's Low Temperature Ther- X
mal Desorption Process
Buffalo Grand Calumet
River River
X X
X X
X
Saginaw
River
X
1.3.6 Analytical Laboratory
In order to limit interlaboratory variation, the different sediments and their residuals were
analyzed by a single subcontractor to SAIC, Battelle Marine Sciences Laboratory (Battelle) in Sequim,
Washington.
2.0 EXPERIMENTAL DESIGN
2.1 Description of the Phased Approach
The treatability tests were conducted in two phases. During Phase I, SAIC sent samples of the
untreated sediments to the technology vendors. These samples underwent a series of initial tests to
determine the optimum conditions to be used during the actual treatability (Phase II) bench-scale tests.
As part of the Phase I effort, a limited number of parameters (temperature, time, pressure, etc.) were
varied. Variations were limited to conditions that could realistically be achieved under the single optimal
set of Phase II test operations. Multiple laboratory simulations may have been performed in support of
this effort; however no analysis of these Phase I tests were performed by SAIC. Where available,
analyses performed by the vendor during this phase of testing have been included in the individual
technology reports.
19
-------
During Phase II, raw sediments were treated in order to evaluate technology performance.
Phase II's experimental designs were based on the results obtained during Phase I testing. Samples of
the untreated sediments and end products generated during testing were sent to Battelle for analysis.
The data generated by Battelle were used to determine treatment efficiencies and to evaluate
performance. Although vendor- or subcontractor-generated data were reported in the individual
technology reports, this information was not used when comparing technology testing results in this
report. Representatives of the ET Work Group were present during the conduct of Phase II testing.
2.2 Characterization of the Various Sediments and Residuals
In order to ensure that the data obtained from this study can be objectively compared with data
generated from the other Phase II studies performed in support of the ARCS Program, Battelle was
subcontracted to perform all of the analyses for the seven treatability studies. In the early stages of
design of this test series, the intent was to analyze, for each bench-scale treatability test, the set of
parameters listed in Table 3 to characterize the raw sediments and the end products produced.
Because of the cost of performing some of the treatability tests the quantities of sediments processed
were smaller than that needed for the full range of analyses. Therefore, obtaining the needed solids
(300 g) to complete the critical analyses of the target compounds (PCBs and PAHs) controlled the
quantities tested. Technology performance and treatment efficiencies were determined relative to raw
sediment characterization data.
Table 3. Parameters For Analysis of ARCS Program 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 Manganese
pH Mercury
BOD Nickel
Total Suspended Solids Selenium
Conductivity
20
-------
2.3 Sampling and Analysis
2.3.1 Test Sample Preparation
The contaminated sediments contained free-standing water. Since homogenizing samples with
free-standing water is difficult, the water was decanted and weighed. Before adding the sediment to
the treatment systems for Phase II testing, the water and sediment were proportionally recombined.
2.3.2 Sampling
At the beginning of each Phase II treatability test, SAIC personnel observing Phase II packed
and shipped a sample of the untreated sediments to SAIC's subcontract laboratory, Battelle, in
accordance with written detailed instructions supplied to the SAIC on-site representative. The sample
was representative of the material treated using the specific technology tested.
Except for the Zimpro process, which produced only a treated solid and aqueous residual, the
evaluated technologies produced an oil residual, an aqueous residual, and a solid residual. After
treatment, samples of these residuals were collected by SAIC for shipment to the laboratory. As
specified in the Quality Assurance Project Plan (QAPP), a minimum of 300 g (dry basis) of solid
material were required in order for Battelle to be able to complete the necessary analyses on the raw
sediments and treated solids. Up to 8 L of the water and 25 mL of the oil residual were required in
order to perform a complete set of analyses on these variables. The actual quantities of water and oil
produced were dependent on the raw sediment and the technology.
2.3.3 Analysis
Analyses were conducted by SAIC's subcontracted laboratory, Battelle, on the raw sediments
and the process products produced during Phase II. Table 4 is an example (from the B.E.S.T.* test) of
the number of analyses conducted on these sediments and their residuals. Descriptions of the
analytical methods employed can be found in the individual technology reports.
3.0 RESULTS AND DISCUSSION
3.1 Summary of Phase I Results
The vendors performed a series of initial tests on the raw sediments to determine specific
operating parameters which would optimize the performance of their respective technologies during
Phase II testing. Following analyses of the raw sediment and residuals produced during Phase I testing,
21
-------
Table 4. SAIC's Analysis Schedule for the Phase II Solvent Extraction Evaluation of
Buffalo River, Grand Calumet River, and Saginaw River Sediments
Parameters
Total Solids
(Moisture)
Volatile Solids
O&G
Metals
PCBs
PAHs
TOC
Total Cyanide
Total Phosphorus
PH
BOD
Total Suspended
Solids
Conductivity
QC Sample ()
and
Method Blank
0)
YES
(1)
YES
(1)
YI-S
(«)
YKS
(1)
YI-S**
(1)
YES**
(0)
YES
(«)
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
1 (3)
B,G,S
NA
NA
NA
NA
NA
NA
NA
MS
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
u ••;
MS
(1)
S
(1)
S
"';
Tripli-
cate
(2)
S
(2)
S |
* 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
-------
parameters were evaluated relative to their effect on performance. The depth of Phase I testing varied
significantly among vendors.
Zimpro set test parameters which would optimize the performance of their technology. A
factorial wet oxidation test was performed on the sediment to give an indication of the importance of the
two experimental variables: residence time and oxidation temperature. The Phase I tests were
performed in stainless-steel laboratory autoclaves.
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.
After observing the Phase I tests RCC, using their experience in selecting operating parameters
for the technology, determined the number of extraction stages, solvent-to-feed ratios, and the
proposed temperature and mixing time for each stage of the bench-scale tests.
During Phase I SoilTech placed between 1.4 and 2.4 kg of the sediment/sand feed in their
Batch Pyrolysis Unit (retort unit) and steadily heated the retort in a controlled manner to approximately
650°C (i.e., during the ramp runs). By observing the characteristics of the feed and the temperatures at
which the water and oil were driven off, data were acquired which were used to determine the boiling
point range of the hydrocarbons in the feed. Three high temperature pyrolyses were conducted on
each sediment at varying temperatures, with all other parameters held constant.
The coked solid resulting from these tests represents the product of the full-scale ATP*'s retort
zone and was tested for organic content to determine the degree of decontamination achieved by
thermal separation. During the high-temperature runs, water and oil were collected in the condensing
circuit and non-condensible gases were stored in a gas bag. Samples of the water, oil, and solids were
collected for analysis. Decontamination achieved by thermal separation was determined by analyzing
for organic content in the retort solids.
ReTeC used their 100 pound-per-hour (45 kilogram-per-hour) system during Phase I testing.
Five-gallon (19-L) sediment samples were processed to determine waste-specific processing conditions
for Phase II. The process operates at temperatures ranging from 260 to 450°C with a solids content of
23
-------
20 percent or greater required. Residence times of 30 and 60 minutes were employed during Phase I
testing. The data obtained by analyzing the raw sediments and treated solids for PCBs, PAHs, and
moisture were used to determine optimum percent solids, processing temperatures, and residence
times to be employed during Phase II.
3.2 Summary of Phase II Results
The concentrations of PAHs, PCBs, metals, total solids, total 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. The following sections briefly address the analytical results pertaining to contaminant
concentrations in the raw sediment and the process residuals (i.e., treated solids, water, and oil) as well
as applicable removal efficiencies. The discussion of Phase II results concludes with an analysis of the
mass balance of the media and contaminants. Data generated by Battelle for Phase II of the study can
be found in each of the individual technology reports.
3.2.1 Sediments/Treated Solids
The following sections address the quality of the sediments before and after treatment. Each
section expands upon a different contaminant type and the reductions experienced following treatment.
3.2.1.1 PCBs
Samples of the feed material and the treated solid produced using the various technologies
were analyzed for PCB contamination. The data from these analyses are presented in Table 5. For
the results of both technologies that treated Buffalo River sediments, the reduction or destruction
efficiencies are not meaningful, because in both cases the concentrations of PCBs were so small that
accurate analytical results were not possible. From interpreting the results of their treatment of other
sediments, both technologies showed that they can treat PCBs.
The only technology of the four tested that did not achieve reasonable removal or destruction of
PCBs was Zimpro's Wet Air Oxidation Technology. This technology was not expected to treat PCBs
because they are too refractory for effective treatment by wet oxidation under the test conditions
utilized.
24
-------
Table 5. Removal Efficiencies for Total PCBs
ro
Ol
Feed
(mg/kg)
B.E.S.T.* 0.32
SoilTech 0.34
Zimpro
ReTeC
NC - Not Calculated
All concentrations are mg/kg
Technology Tested
B.E.S.T.*
SoilTech
Zimpro
ReTeC
Buffalo River Saqinaw River Grand Calumet River
Removal/ Removal/ Removal/
Treated Destruction Treated Destruction Treated Destruction
Solids Efficiencies Feed Solids Efficiencies Foed Solids Efficiencies
(mg/kg) (%) (mg/kg) (mg/kg) (%) (mg/kg) (mg/kg) (%)
<0.3 >6 21.9 0.24 99 15.0 . 0.44 97
<5 NC 10.7 <3 >72
11.9 8.5 29
dry weight
Table 6. Removal Efficiencies for Total PAHs
Buffalo River Saqinaw River Grand Calumet River
Removal/ Removal/ Removal/
Treated Destruction Treated Destruction Treated Destruction
Feed Solids Efficiencies Feed Solids Efficiencies Feed Solids Efficiencies
(mg/kg) (mg/kg) (%) (mg/kg) (mg/kg) (%) (mg/kg) (mg/kg) (%)
9.9 0.37 96 2.7 0.95 65 230 37.1 84
8.72 0.11 99 235 1.61 99
266 <2.84 >98.9
Ashtabula River
Removal/
Treated Destruction
Feed Solids Efficiencies
(mg/kg) (mg/kg) (%)
14.6 <0.6 >96
Ashtabula River
Removal/
Treated Destruction
Feed Solids Efficiencies
(mg/kg) (mg/kg) (%)
6.1 <2.4 >60
NC - Not Calculated
All concentrations are
mg/kg dry weight
-------
3.2.1.2 PAHS
Feed material and treated solids were also analyzed for PAHs. As shown in Table 6, all four
technologies were effective in removing or destroying total PAHs.
3.2.1.3 Total Metals
The data for specific metals are contained in the individual technology reports and are
summarized qualitatively here. The Zimpro Wet Air Oxidation Process does not effectively remove
metals. As demonstrated by low or negative removal percentages, in general, the B.E.S.T.* Solvent
Extraction Process does not effectively remove metals. Except for the removal associated with
mercury, the elevated percent removals experienced on the Grand Calumet River sediment should not
in theory be attributed to treatment by the SoilTech ATP* Technology. With the exception of mercury,
there is no indication that the ReTeC technology effectively removes metals.
3.2.1.4 Other Analyses
The feed sediment and treated solids were analyzed for oil and grease, TOG, and total volatile
solids, as shown in Table 7. In the case of the B.E.S.T.* Solvent Extraction Process and the SoilTech
ATP* Technology, oil and grease reduction could possibly be used as a low-cost indicator of technology
efficiency. It does not appear this correlation exists for the other two technologies.
Table 7. Battelle Data - Removal Efficiencies for Other Parameters
Buffalo River
Contaminant
PCBs
PAHs
Oil and Greece
TOG
Total Volatile
Solids
SoilTech
NC
99
96
>99
96
B.E.S.T.*
>6
96
90
39
3
Saginaw River
B.E.S.T*
99
65
80
30
17
Grand Calumet River Ashtabula River
SoilTech
>72
99
99
92
92
Zimpro
29
>98.9
90
52
51
B.E.S.T.*
97
84
99
21
36
ReTeC
>96
>60
57
-14
44
NC - Not calculated
3.2.2 Treated Water and Processed Oils
The concentrations of PAHs and PCBs in the treated waters and the oils that were produced by
the technologies are provided in Table 8. Three of the four technologies tested (B.E.S.T.*, SoilTech,
ReTeC) extracted or removed the PAHs and PCBs from the sediments. Previous sections discussed
26
-------
Table 8. PAH and PCB Concentrations in Treated Waters and Produced Oils
ug/L Water, mg/kg oil dry weight
B.E.S.T.*
Buffalo River Saqinaw River
Residual Residual
Water Oil Water Oil
PAH 1.41 1260 2.3 1610
PCB <0.6 62.3 <0.6 5012
ReTeC SoilTech
Grand Calumet River Ashtabula River Buffalo River Grand Calumet River
Residual Residual Residual Residual
Water Oil Water Oil Water Oil Water Oil
83.2 4310 450 3800 3310 8100 2060 11300
4.8 268 <20 <7 66 <800 64 77.8
ro
-g
-------
the removal efficiencies of these technologies. Analytical results provided in Table 8 show that the
removed PAHs and PCBs transferred to the produced oil and did not enter the residual waters to any
significant amount. These technologies performed as expected by concentrating the contaminants in
the oil, thereby creating significantly less mass of material requiring destructive treatment. The Zimpro
wet air oxidation process produced only a wastewater stream; oily hydrocarbons were largely destroyed
by the process. The filtrate sample, which represents the water effluent which would result from this
process, did not contain detectable PCBs. For all of the technologies evaluated it appears that further
treatment of the wastewater effluents generated by the processes would be necessary prior to
discharge.
3.2.3 Mass Balance
Except for the Zimpro technology, where destruction of the contaminants occurred, mass
balances were calculated for solids, water, oil, PCBs, and PAHs. All closures were satisfactory, or,
where there was a significant variance, a reasonable explanation was available. A summary of the
mass balances is given in Table 9, with the detailed discussions of the mass balance calculations given
in the individual technology reports.
Table 9. Mass Balances (percent recovered)
Parameter
Solids
Water
Oil
PCBs
PAHs
Buffalo
98
68
163
129
90
RCC
Saginaw
99
78
192
80
240
SoilTech
Grand Calumet
92
74
69
94
111
Buffalo
79
59
187
29
839
Grand Calumet
NC
139
29
16
69
ReTeC
Ashtabula
40
89
3500
(D
3170
NC - Not Calculated
(1) Sample was lost due to container breakage.
3.3 Quality Assurance/Quality Control
Upon review of all sampled data and associated QC results, the data generated for all four
technology studies have been determined to be of acceptable quality. The Data Verification Report,
28
-------
prepared by the ARCS Program QA Manager, is attached to this report as the appendix. 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.
The complete analyses for each technology test, related to Quality Assurance/Quality Control,
are contained in the individual technology reports.
3.4 Individual Technology Report References
The reports detailing each individual technology evaluation are available from GLNPO under
the following titles:
Bench-Scale Evaluation ofRCC's Basic Extractive Sludge Treatment (B.E.S.T.9) Process on
Contaminated Sediments from the Buffalo, Grand Calumet and Saginaw Rivers, EPA 905-R94-010
Bench-Scale Evaluation of ReTeC's Thermal Desorption Technology on Contaminated
Sediments from the Ashtabula River, EPA 905-R94-008
Bench-Scale Evaluation of SoilTech's Anaerobic Thermal Process Technology on Contaminated
Sediments from the Buffalo and Grand Calumet Rivers, EPA 905-R94-009
Bench-Scale Evaluation ofZimpro's Wet Air Oxidation Process on Contaminated Sediments
from the Grand Calumet River, EPA 905-R94-007.
29
-------
Appendix
Data Verification Report for the ARCS Program
-------
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 conventionals (% 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 -200° 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 (Bartelle-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) conventional, including percent moisture, pH, percent total volatile, oil and grease, total
organic carbon (TOC), total cyanide, and total phosphorus. Analyses of metals and
conventional 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 conventional
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 calculation^ 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 SO
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-QDo
0-A, B, Q D, P0 S,
0-A, B, Co D, P0 S,
6-A,CoD9S,
15-A,C.
12-C6 P0 S,
14-AoP0
14-AoPo
17-Bj D0
17-D0 S,
ZIMPRO
12-CoD0
3-A, B, Co D, S,
0-A, B, Co D, P0 S,
3-A, BO Q D, S,
6-A, Bj C6 D, S,
12-C6 P0S,
14-AoPo
14-Ao P0
14-A, Bj D0
11-BjDoS.Sj
Soil Tech
12-CoDo
0-A,B,C0D,P0S,
0-A,B,CoD,P0S,
6-A, Q D, S,
6-A, Bj C6 D, So
12-C6 P0 S,
ll-AoP0So
14-Ao P0
14-A, B, DO
17-Do S,
RETEC
12-CoD0
3-A,B,CoD,S,
3-A,B,CoD,S,
6-A, Q D, S,
9-A, DO C, So
9-C6D0P0S,
8-AoD0P,So
ll-AoD0So
11-A.BjDoS,
20-D0
Treated
Sediments
Metals
% Moisture
PH
%TVS
Oil and grease
TOC
Total cyanide
Total
phosphorus
PCBs
PAHs
12-CoDo
0-A, B, Co D, Po S,
0-A, B, Q D, P0 S,
6-A,CoD,S,
15-A,C.
12-C6P0S,
14-AoP.
14-Ao P0
14-6,00?,
14-DoP.S,
12-QDo
0-A,B,CoD,P0S,
3-A,B,CoD,S,
3-A, BO Q D, S,
6-A, Bj C6 D, S,
12-C6P0S,
14-Ao P0
14-Ao PO
11-A.BjDoP,
17-DoS,
12-QDo
3-A, B, Co D, S,
0-A, B, Q D, Po S,
6-A,CoD,S,
9-A, Bj C6 D,
12-C6P0S,
14-AaPo
14-Ao Po
14-BjDoP,
14-DoP.S,
12-QDo
3-A,B,C,D9S,
S-A.B.Q.D.S,
6-A, Q D, S,
6-A, C, Do P, So
12-C, D0 S,
ll-AoD0P,
14-A, D0
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
H-AoDoP.S,
**
**
**
**
««
**
**
**
**
**
14-B2 D0 P0
17-D0 S,
*«
**
*•
**
««
**
**
**
»*
**
S-A.^DoPoS,
s,
17-Do P0
20
**
3-A,B,CoD,S9
6-A,CoD, S«,
6-A,CoD9S,
6-A, Co D, S,
12-A, C. D0
9-AoQD.S,
14-AoD,
14-AoDo
9^0.0,5,
S-AoBjDoPoS,
s.
H-A.DoP.S,
Oil residue
PCBs
PAHs
11-A, RjDoS,
11-AoBjDoSj
*
*
17-Bj D0
14-BjD.S,
ll^DoP.S,
17-BjD,
* 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 up! 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
Ii.&.o* I •
20
8
8
6
17
17
20
20
20-B,
20-S,
ZIMPRO
20
8
8
6
8-Bj D, S,
17
20
20
17-A.B,
14-Bj S, Sj
Soil Tech
20
8
8
6
11-BjD,
17
20
20
17-A.B,
20-S,
RETEC
20
8
8
6
17
17
17-P,
20
17-A.B,
23
Treated
Sediments
Metals
% Moisture
PH
%TVS
Oil and grease
TOC
Total cyanide
Total phosphorus
PCBs
PAHs
20
8
8
6
17
17
20
20
n-BjF,
17-P, S,
20
8
8
6
8-Bj D, S,
17
20
20
14-A.BjP,
20- S,
20
8
8
6
11-8,0,
17
20
20
n-B,p,
20-S,
20
8
8
6
9-P,
17
20
20
17-A, B,
23
-------
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-B,
17-P.S,
**
**
**
»*
**
«*
«*
**
«*
**
**
20-B,
20-S2
**
«»
«*
**
«»
**
**
•«
**
*»
««
14-A, B, S,
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-Bj
17-8,5,
20-Bj
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. PCS 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)
fluoranthene 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, the
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 %moisrure, 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 benzoQc)
-------
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 oU 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 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
conventionals, 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.
-------
NOTE
Appendix A - Laboratory Submitted Data Summary Sheets
and
Appendix D - ARCS Data Verification Templates by Parameter
are not included with 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 = 3
Analytical Replicate = 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
BO = no information available
Bj = reagent blank value exceeded MDL
69 = blanks are not applicable
C = Calibration Problem
Co = no information available
C, = initial calibration problem
C, = on-going calibration problem
Cs = no information on initial calibration
C6 — no information on on-going calibration
€9 = on-going calibration is not applicable
D = Detection Limit Problem
D0 = no information available
D, = detection limit exceeded
D, = 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
Pj = 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
S5 = no information available on matrix spike recovery
S6 = no information available on surrogate spike recovery
S, = spike recovery not applicable
------- |