&EPA
        United States
        Environmental Protection
        Agency
           Great Lakes
           National Program Office
           77 West Jackson Boulevard
           Chicago, Illinois 60604
EPA 905-R94-003
January 1994
Assessment and
Remediation of
Contaminated Sediments
(ARCS) Program
        PILOT-SCALE DEMONSTRATION
        #OF SOLVENT EXTRACTION
        FOR THE TREATMENT OF
        GRAND CALUMET RIVER SEDIMENTS
                           United States Areas of Concern

                           ARCS Priority Areas of Concern

-------
     PILOT-SCALE DEMONSTRATION OF
SOLVENT EXTRACTION FOR THE TREATMENT
  OF GRAND CALUMET RIVER SEDIMENTS
               Prepared for the
      U.S. Environmental Protection Agency
       Great Lakes National Program Office
                   by the
         U.S. Army Corps of Engineers
               Chicago District
 U.S. Environmental Protection /^e
 Region 5, Library .'PL-12.T
 77 West Jackson Boulevard, 12tri Moor
 Chicago, IL  60604-3590

-------
                                              ABSTRACT
                   This report describes the Basic Extractive Sludge Treatment (BEST) process
             (a solvent-extraction treatment technology patented by Resources Conservation
             Company) and its application to the treatment of contaminated sediments. The process
             design and the properties of the solvent triethylamine are presented. Bench-scale and
             pilot-scale tests conducted at the Grand Calumet River in Gary, Indiana as part of the
             Assessment and Remediation of Contaminated Sediments (ARCS) program are described.

                   Two separate tests were conducted on sediment collected from different locations
             in the Grand Calumet River. Five test runs for each sediment type (Sediment A and
             Sediment B) were conducted. The averages of the three optimum runs were used to
             evaluate the technology's performance. Sediment A contained 12 mg/kg PCBs, 550
\           mg/kg PAHs and 6,900 mg/kg O&G. The process removed more than 99 percent of the
{            PCBs, 96 percent of the PAHs, and more than 98 percent of the O&G.  Sediment B
C           contained 430 mg/kg PCBs, 71,000 mg/kg PAHs, and 127,000 mg/kg O&G.  The process
t*-           removed more than 99 percent of the PCBs and PAHs, and more than 98 percent of the
             O&G.

                    Cost estimates for a large-scale treatment of Grand Calumet River sediment using
             a 184 cubic yards per day unit range from $361 to $139 per cubic yard for 5,000 cubic
             yards and 100,000 cubic yards of sediment treated, respectively.

-------
                                  DISCLAIMER

The information in this document has been funded wholly or in part by the U.S. Environmen-
tal Protection Agency (EPA) under Interagency Agreements No. DW96934688-0,
DW96947555-0, DW96947581-0, and DW96947595-0, with the U.S. Army Corps of
Engineers. 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.

-------
                               CONTENTS

 Figures	iii
 Tables	iii
 Plates	iv
 List of Abbreviations and Symbols	v

 Introduction	1
       Objectives	2
       Description of Area of Concern	3
              Watershed	..3
              Status of RAP	5
              Sediment Characteristics and Quality	6

 Demonstration Approach	8
       Planning Activities	8
              Technology Selection	„	8
              Solvent Extraction	9
              Planning Report	9
              Coordination	10
              Site Selection	10
              Sample Collection....	11
              Contracting	11
              Schedule	12
       Technology Description	13
              Process Theory	13
              Pilot-Scale Process Equipment	19
       Sediment Sample Collection	24
       Sediment Pre-treatment	25
              Prescreening	25
              Homogenization	29
      Bench Scale Treatability Tests	29
              Purpose	29
      Pilot Unit Operation	31
              Unit Mobilization	31
              Feed Analysis	31
              Batch Processing	33
      Residuals Management	38

Monitoring Program	40
      Introduction	40
      Sampling Locations	42

-------
                                 CONTENTS (Continued)

       Sampling Procedures	42
              Untreated Sediment (Raw Feed)	42
              Product Solids	42
              Product Water	44
              Product Oil	44
              Oil/Solvent Mix	......44
              Recycled Solvent	44
              Vent Emissions	....44

Results and Discussion	45
       Process Operation	.....45
       Chemical Concentration Data	....45
       Mass Balances	55
              Total Material Balance	 55
              Solids Balance	55
              PCB Balance.	 55
              PAH Balance	.......60
              Oil and Grease Balance	........60
              Water Balance	61
              Solvent Balance	 61
       Discussion and Interpretation	 61
              Introduction	61
              Assessing Data Quality	62
              Removal ofOrganics	 64
              Triethylamine Residual in Products	 66
              TCLP Leachability of Treated Solids	67
              Biodegradation Tests	 67
       Estimated Costs	.68
              Introduction	68
              Assumptions	69
              Basis of Economic Analysis	69
              Cost Conclusions	75
       Conclusions and Lessons Learned	75
              Effects of TSCA on Pilot-Scale Demonstration	76

References	79

-------
                                  FIGURES

Number                                                                 Page
    1      Grand Calumet River/Indiana Harbor Canal Area of Concern	4
    2      Inverse Miscibility of Triethylamine and Water	13
    3      BEST Solvent Extraction Process	16
    4      BEST Pilot Plant	21
    5      Sediment Sample Collection Locations	27
    6      Demonstration Site	28
    7      Solid and Liquid Sample Locations	43
Number
                                  TABLES
Page
   1      Volume-Weighted Mean Concentrations of IHC Federal Navigation
             Project Sediments	„	7
   2      Chronology of Field-Related Activities for the
             Pilot-Scale Demonstration	12
   3      Summary of Bench-Scale Testing in Support of the Pilot
           Demonstration	32
   4      Analysis Performed on Feed Sediment	33
   5      Feed Loading Summary	35
   6      Extraction Sequence used for Sediment A	36
   7      Extraction Sequence used for Sediment B	36
   8      Primary Control Parameter Summary	37
   9      Summary of Analyses	41
   10     PAH Concentrations and Removal EfiBciencies	46
   11     PCB Concentrations and Removal EfiBciencies	47
   12     O&G Concentrations and Removal Efficiencies	48
   13     Total Metals in Test Sediments	49
   14     TCLP Test Results in Test Sediments	50
   15     Noncritical Analyses Results - Sediment A and B	51
   16     PAH and PCB Concentrations of Product Water	52
   17     Total Metals in Product Water	52
   18     Supplemental Analyses Results - Sediments A and B Product Water	53
   19     PAH and PCB Concentrations of Sediment B Product Oil	53
   20     Triethylamine Concentrations in Treated Solids,  Product Water
            and Oil Phases	54
   21     Mass Balances - Sediment A Inputs	56
   22     Mass Balances - Sediment A Outputs	57
   23     Mass Balances - Sediment B  Inputs	58
                                     iu

-------
                               TABLES (Continued)
Number                                                                   Page

   24      Mass Balances - Sediment B Outputs	59
   25      Mass Balance Results, Percent Recovery	 60
   26      Summary of Conclusions for Contaminant Removal	64
   27      Comparison of Organic Analyses between SITE and RCC	65
   28      Residual Triethylamine Concentrations	67
   29      Treatment Costs for 184-cypd BEST System Treating
             25,000 Cubic Yards of Contaminated Soil, Sediment
             or Sludge	73
   30      Treatment Costs as Percentages of Total Costs for
             184-cypd BEST System Treating 25,000 Cubic Yards of
             Contaminated Soil, Sediment, or Sludge	 73
   31      Treatment Costs for 184-cypd BEST System Operating
             with an 80-Percent Online Factor	 74
   32      Treatment Costs as Percentages of Total Costs for
             184-cypd BEST System Operating with an 80-Percent Online Factor..... 74
                                     PLATES
Number
    1      Collection of Sediment Sample B	26
    2      Homogenization of Screened Sediment using a Rotary Mixer	26
    3      BEST Solvent Extraction Pilot-Scale Unit	32
    4      Samples of Process Residuals - Treated Sediment, Water and Oil	32
                                       IV

-------
                          LIST OF ABBREVIATIONS
 AAR        Applications Analysis Report
 AOC        area of concern
 ARCS       Assessment and Remediation of Contaminated Sediments
 ASTM       American Society of Testing Materials
 BEST       Basic Extractive Sludge Treatment
 BOD        biochemical oxygen demand
 COD        chemical oxygen demand
 CFR        Code of Federal Regulations
 COE        U.S. Army Corps of Engineers
 CWA       Clean Water Act
 cypd        cubic yards per day
 ETWG       Engineering Technology Work Group
 OCR        Grand Calumet River
 GLNPO      Great Lakes National Program Office
 HDPE       high density polyethylene
 IDEM       Indiana Department of Environmental Management
 IHC         Indiana Harbor Canal
 IJC          International Joint Commission
 mg/kg       milligrams per kilogram
 mg/1         milligrams per liter
 ml/min       milliliters per minute
 MS          matrix spike
 MSD        matrix spike duplicate
 NIOSH      National Institute for Occupational Safety and Health
 O&G        oil and grease
 PAH        polynuclear aromatic hydrocarbon
 PCB         polychlorinated biphenyl
 pH          - log JIT4" concentration]
 PPE         personal protective equipment
 ppm         part per million
 QAPP       Quality Assurance Project Plan
 QC          quality control
 RAP         Remedial Action Plan
 RCC         Resources Conservation Company
 RCRA       Resource Conservation and Recovery Act
 SAIC        Science Applications International Corporation
 SITE        Superfund Innovative Technology Evaluation
 SW-846      Solid Waste Publication 846 - Test Methods for
             Evaluating Solid Waste
TCLP        Toxicity Characteristic Leaching Procedure
TDS         total dissolved solids
TER         Technical Evaluation Report

-------
                    LIST OF ABBREVIATIONS (Continued)

TIC          total inorganic carbon
TKN         total Kjedahl nitrogen
TOC         total organic carbon
TRPH        total recoverable petroleum hydrocarbons
TSCA        Toxic Substances Control Act
TSS          total suspended solids
USEPA      U.S. Environmental Protection Agency
                                     VI

-------
                     PILOT-SCALE DEMONSTRATION OF
                SOLVENT EXTRACTION FOR THE TREATMENT
                  OF GRAND CALUMET RIVER SEDIMENTS

                               INTRODUCTION

       The International Joint Commission has identified the Indiana Harbor Canal, all of
the Grand Calumet River of Indiana, and the nearshore of Lake Michigan as one of 43
areas of concern (AOC) around the Great Lakes which do not meet one or more of the
objectives of the 1978 Great Lakes Water Quality Agreement or other standards, criteria
or guidelines. The 1987 amendments to the Clean Water Act (CWA),  Section 118(c)(3),
authorized the U.S. Environmental Protection Agency's (USEPA) Great Lakes National
Program Office (GLNPO) to conduct a 5-year study and demonstration project on the
control and removal of toxic pollutants in the Great Lakes, with emphasis on the removal
of toxic pollutants from bottom sediments.  The Clean Water Act (CWA) specified five
areas, including the Grand Calumet River/Indiana Harbor and Canal (GCR/IHC) AOC, as
requiring priority consideration in locating and conducting on-site demonstration projects.

       In response, GLNPO initiated the Assessment and Remediation of Contaminated
Sediments (ARCS) Program to assess the nature and extent of sediment contamination at
the priority AOCs, to evaluate and demonstrate remedial options, and to provide guidance
on the assessment of contaminated sediment and the selection and implementation of
remedial actions in the AOCs and other locations in the Great Lakes. The ARCS Program
created an Engineering and Technology Workgroup (ETWG) to  select promising
technologies and to carry out the on-site pilot-scale demonstration projects.  Other
agencies involved in the cooperative effort included USEPA's Superfund Innovative
Technology Evaluation (SITE) Program, USEPA Region 5 and the U.S. Army Corps of
Engineers (COE), Chicago District.

      The SITE Program promotes the development and use of innovative technologies
to implement federal and state cleanup standards at Superfund sites.  SITE'S
Demonstration Program provides an assessment of an innovative technology's
performance, reliability and cost. For this pilot-scale demonstration, SITE assisted in
preparation of the demonstration site, collection and analysis of samples, and disposal of

-------
treatment residuals. SITE also evaluated the effectiveness of the Basic Extractive Solvent
Technology (BEST) process and published two reports: the Technology Evaluation
Report (TER, EPA/540/R-92/079a) and the Applications Analysis Report (AAR)
(TER, EPA/540/AR-92/079).

       The TER contains a comprehensive description of this demonstration project and
its results. It provides detailed descriptions of the BEST process (a solvent extraction
technology), the sediment used in the project, sampling and analyses, data generated, and
the Quality Assurance Program. The AAR, more general than the TER, includes
estimated costs for the technology and summarizes the results of the demonstration. The
AAR discusses the advantages, disadvantages, and limitations of the BEST process.

       The sampling was conducted through EPA's SITE program via Science
Applications International Corporation (SAIC).  Analytical tests were conducted by SITE
through Maxwell/S-cubed Division, SAIC, Triangle Laboratories, Commercial Testing
and IT Air Quality Services. In addition, RCC independently collected and analyzed its
own set of samples.

       USEPA Region 5 helped procure a location for the demonstration project, acted as
an interface between U.S. Steel Gary Works (who provided use of the demonstration site
and access to utilities), and assisted in ensuring compliance with applicable Federal laws
and regulations, particularly the Resources Conservation and Recovery Act (RCRA) and
the Toxic Substances Control Act (TSCA).

       The COE,  Chicago District, served with the ETWG, awarded and managed a
contract with the developer of the BEST process, Resources Conservation Company, Inc.
(RCC) and prepared this report. This report describes the ARCS pilot-scale
demonstration project conducted at the GCR/EHC AOC.

OBJECTIVES

       The goals of the ARCS program for the pilot-scale demonstration and subsequent
analyses were to evaluate the efficiency of the BEST technology at reducing organic
contamination in OCR sediment with varying levels of contamination and to develop cost
estimates for a full-scale application of the process.  The following objectives for the pilot-

-------
scale demonstration were established by the SITE Program: assessing the quality of the
treated solids, the residual product water, and the concentrated oil product, developing
capital and operating costs for the technology, developing an overall mass balance for
organic contaminants around the BEST pilot plant, evaluating the technology's effect on
metals found in the sediment, and assessing the biodegradation of residual triethylamine in
the product solids.

DESCRIPTION OF AREA OF CONCERN

Watershed

       The GCR/IHC AOC is located about 25 kilometers south of Chicago, Illinois, and
is in the northwestern part of the State of Indiana. The area includes the IHC, all of the
OCR of Indiana, and the nearshore of Lake Michigan as shown in Figure 1.  The AOC is
heavily industrialized, with large steel production and processing facilities, petroleum
refineries and chemical plants in close proximity to each other and the river.

       The drainage basin encompasses nearly 17,500 hectares, according to the USEPA
Grand Calumet Master Plan (1985).  This estimate includes the area's sanitary and storm
sewer systems and the divergence of flow in the west branch of the OCR. The slope of
the drainage basin is very gentle, with a 40 meter change in elevation over a distance of 30
to 40 kilometers.

        The GCR discharges to southwestern Lake Michigan via IHC. Over 90 percent of
dry-weather flows are from municipal and industrial discharges. The State of Indiana has
designated the GCR and IHC as industrial water supply, partial body contact, limited
aquatic life waters.  The levels of ammonia nitrogen exceed state standards throughout the
entire year. Fecal coliforms exceed standards mainly during winter months, and dissolved
oxygen levels in portions of the GCR have dropped to levels low enough in recent years to
cause noticeable odor problems (IDEM, 1990).

      The flow system of the GCR consists of two branches oriented east-west and a canal
section oriented north-south. The east branch of the GCR always flows west to the

-------
  Chicago
     Calumet River
     Indiana
     Harbor
                                                   Lake Michigan
                                                          ARCS DEMO

                                                   SEDIMENT A  \    SEDIMENT B
Lake George
  Branch v
                                     Indiana
                                     Harbor Canal
                                                 Gary
                                               Harbor
                                      Grand Calumet
                                      River Branch
                      HYDRAULIC DIVIDE
                                       TRIPLE JUNCTION
                                                             (MILES, SCALE APPROXIMATE)
                       west Branch  East Branch
Little Calumet River
FIGURE 1.      Grand Calumet River/Indiana Harbor Canal Area of Concern (Source: IDEM, 1990).

-------
junction of the east branch with the west branch and the IHC. The confluence of these
channels is referred to as the triple junction.  The IHC begins at the triple junction and flows
directly north toward Lake Michigan. About halfway to Lake Michigan, however, it turns
to the northeast. At that point, another arm of the Canal goes due west, and is known as
the Lake George Branch of the IHC.

       The west branch of the OCR has a hydraulic divide whose location varies depending
on the amount of water flowing from the east branch, the elevation of Lake Michigan, the
wind direction and speed, and the amount  of discharge from the Hammond and East
Chicago wastewater treatment plants. At times, no divide on the branch exists, and all flow
in the west branch is toward the west. Most frequently, a divide between easterly and
westerly flow exists near the boundary between the East Chicago and Hammond
wastewater treatment plants. Further complicating the hydraulic picture is estuarine flow in
the harbor itself. This flow generally consists of relatively cold Lake Michigan water
entering into the harbor and canal along the bottom, eventually mixing upwards with the
warm waters of the canal which return into Lake Michigan.

       It is estimated that over 60 million  kilograms of sediment enter Lake Michigan
each year from the IHC on these sediment  particles. It is estimated that approximately
170 kilograms of PCBs, 930 kilograms of cadmium and 45,000 kilograms of lead are
transported to the lake.  This material will  continue to move to the lake for many years if
the existing sediment is not removed (COE, in preparation).

Status of RAP

       The International Joint Commission (UC) has identified the GCR/JHC as an  "area
of concern" for the Great Lakes.  In January 1991, the Indiana Department of
Environmental Management released a draft Stage I Remedial Action Plan (RAP) for the
GCR/JHC and the Nearshore Lake Michigan to address water quality, aquatic habitat, and
use impairment issues related to this area of concern. All fourteen of the beneficial uses
(as identified in the Great Lakes Water Quality Agreement) are considered impaired in the
GCR/IHC AOC (IDEM, 1990).

-------
       IDEM is currently preparing the Stage n RAP, which will identify remedial
measures for the above problems. The Stage n RAP will include several components
which will address water quality, sediment and dredging, and habitat. The draft water
quality component has been distributed for public comment.  The draft Stage n RAP is
scheduled for completion in 1994.

Sediment Characteristics and Quality

       Because of the highly urbanized and industrial nature of the GCR/THC watershed,
the bottom sediments are contaminated with a variety of pollutants. The Stage I RAP
identified in-place sediment contamination in the GCR/IHC as a significant environmental
problem that adversely impacts water quality and aquatic life in the waterway and in Lake
Michigan.

       The GCR/IHC is, for the most part, a man-made waterway. The IHC began as a
series of drainage ditches which by 1914 were deepened to navigation channels by local
interests. Much of the GCR has also been modified by municipalities and private industry.
As a result, it is difficult to say there is a "natural" stream bed in the waterway. The bottom
of the IHC was originally excavated to sand or clay till.

       Nearly all of the GCR/IHC has deposits of fine-grained sediment which have
accumulated over the years. The physical and chemical characteristics of these sediments
reflect the land and water uses of northwest Indiana. Significant sources of sediment and
sediment contamination in the GCR/IHC include municipal and industrial discharges,
combined sewer overflows, and urban runoff. It is estimated that about 74 million kilograms
of sediment are discharged to the GCR/IHC from these sources (COE, in preparation).

Indiana Harbor and Canal

       The U.S. Army Corps of Engineers (COE), is responsible for the maintenance of a
deep-draft navigation project in the IHC.  Due to the contaminated nature of the bottom
sediments, and the subsequent lack of an appropriate and acceptable disposal location, the
Federal channel has not been dredged since 1972. The COE and the USEPA have
extensively sampled the bottom sediments in the IHC to determine the appropriate disposal
methods for dredged materials.  Samples were collected from the Federal navigation project

-------
 in 1977, 1979, 1980, 1983, 1984, 1985, 1987, and twice in 1988. The ARCS program
 conducted additional sampling activities in 1989 and 1990.

       The bulk chemistry from the discrete sampling events consistently shows high
 levels of most metals, nutrients, oil & grease, and volatile solids.  Using the results of
 USEPA and COE sediment sampling from the Federal navigation channel, the volume-
 weighted concentrations of measured contaminant parameters were calculated.  The
 concentrations determined from each sampling event were weighted against a
 representative volume of sediment in order to determine the average concentration
 (Table 1). Portions of the IHC contain levels of PCBs in excess of 50 ppm, subjecting the
 sediment to regulation under the Toxic Substances Control  Act (TSCA) of 1976.

 Table 1:  Volume-Weighted Mean Concentrations of IHC Federal Navigation Project Sediments
Volume-Weighted
Parameter Mean Concentration
Volatile Solids
COD
Oil & Grease
TKN
Ammonia
Cyanide
Manganese
Phosphorus
Mercury
PCBs
14%
208,000
64,000
3,000
850
1.4
2,000
2,600
0.7
8.9
Volume-Weighted
Parameter Mean Concentration
Arsenic
Barium
Cadmium
Chromium
Copper
Iron
Nickel
Lead
Zinc

50
50
10
370
160
145,000
100
840
3,700

          All concentrations are rag/kg dry weight, unless otherwise noted.

Grand Calumet River

       Additional sampling events in the GCR have been recently completed or are
scheduled for completion as the result of enforcement actions brought by USEPA Region 5
against area industries and municipalities.  Hammond Sanitary District completed a
sediment study covering two miles of the west branch of the GCR (from Indianapolis
Boulevard to the state line), and a sediment study covering the Illinois portion of the river.
U.S. Steel Gary Works conducted a sediment study on the entire east branch, the portion of
the west branch east of Indianapolis Boulevard, and the IHC from the Triple Junction to
Columbus Drive in 1991.

-------
                         DEMONSTRATION APPROACH

PLANNING ACTIVITIES

Technology Selection

       A literature review of treatment technologies was performed for the ARCS
Program by the Corps of Engineers Waterways Experiment Station (WES) and was used
to screen process options for biological, chemical, extraction, immobilization, radiant
energy, and thermal technologies (Averett, et. al., 1990). Each process option was
assessed  on the basis of effectiveness, implementability, and cost. A number of the higher
cost thermal processes were eliminated from consideration due to the expense of these
processes while numerous other processes were eliminated from further consideration
because of the lack of research and development for application to a specific sediment and
associated contaminant matrix.  The availability of a mobile pilot-scale unit was essential
for implementing an on-site pilot demonstration. Based on these criteria, a list of those
processes that should be  retained for demonstration consideration was developed.

       A matrix was developed for the processes recommended for consideration for the
pilot-scale demonstrations, the principal contaminants treatable by each process, and the
Areas of Concern where  such contaminants are present and the processes are applicable
(Averett  et. al., 1990). A list of potential pilot projects was then prepared and these
alternatives were ranked  for consideration based on factors affecting then* selection.

       All five priority sites, Ashtabula River, Ohio; Buffalo River, New York; Grand
Calumet River, Indiana; Saginaw River, Michigan; and Sheboygan Harbor, Wisconsin are
contaminated by organic compounds. Most of these sites have areas of elevated
contamination that could be used for a demonstration project. Rather than strictly
following the numeric ranking of the potential pilot scale demonstration, the ARCS
ETWG, responsible for recommending and implementing the demonstrations determined
that a variety of the technology groups (biological, chemical, extraction, immobilization,
thermal)  should be selected for demonstration.

-------
 Solvent Extraction

       Solvent extraction is a treatment technology that uses a solvent (a fluid that can
 dissolve another substance) to separate or remove hazardous organic contaminants from
 sediment. Solvent extraction does not destroy contaminants, but rather concentrates them
 so they can be recycled or destroyed. It may be used in combination with other
 technologies to destroy the separated concentrated contaminants (USEPA, 1992).

       Solvent extraction is effective in treating primarily organic contaminants, such as
 PCBs, volatile organic compounds (VOCs), halogenated solvents and petroleum wastes.
 Typical solvents used include liquid carbon dioxide, propane, butane, triethylamine,
 acetone and methanol. Solvent extraction is generally not used to remove inorganics since
 these materials do not readily dissolve in most solvents.

       The BEST process separates contaminated sediments into three fractions:  a solid
 fraction that contains the inorganic contaminants (such as heavy metals); an oil fraction
 that contains the organic contaminants, such as polychlorinated biphenyls (PCBs) and
 polynuclear aromatic hydrocarbons (PAHs); and a water fraction that may contain residual
 amounts of the original sediment contaminants.  This process may significantly reduce the
 volume of material requiring further treatment or regulated disposal.

       The BEST process was selected for this pilot-scale demonstration because:

       • a mobile pilot unit was available which had been employed at several locations;
       • previous bench-scale tests showed successful results; and
       • the technology was compatible with the nature of the GCR/IHC sediment
         (i.e., the sediment has a high oil content and heavy organic contamination).

Planning Report

       A report titled "Project Plan Pilot-Scale Demonstration of a Solvent Extraction
Technology at the Grand Calumet River/Indiana Harbor and Canal Area of Concern"
(COE, 1991) described sediment quality, the GCR/IHC AOC, the selection of the
treatment technology and the planned pilot-scale demonstration.  The report also

-------
contained an estimate of all costs associated with the development and implementation of
the pilot-scale demonstration. The document was distributed to the ETWG, USEPA
Region 5, and state and local agencies for review and comment.

       SITE prepared a "Demonstration and Quality Assurance Project Plan" which
outlined the activities to be conducted by the SITE program for the pilot-scale
demonstration project, including support of the site preparation and mobilization, pre-
screening and homogenization. This report also described SITE'S sampling and analysis
program, and the disposal of all of the treatment residuals (excluding non-contact water)
and excess untreated sediment.

Coordination

       The Chicago District began planning for the pilot-scale demonstration in the fall of
1990. Coordination involved meetings, correspondence and telephone conversations with
the ETWG, with the USEPA's SITE program, Region 5 PCS Control Section, Region 5
Water Division, IDEMs Office of Solid and Hazardous Waste Management and U.S.
Steel Gary Works.

Site Selection

       Two overriding criteria determined the selection of a site for the pilot-scale
demonstration.  The first criterion was to perform the demonstration in close proximity,
if not directly on the bank, of the GCR/IHC.  As has been described previously, the
GCR/EHC area is highly industrialized; access to the areas along the river is controlled
by the industrial landowners. Second, USEPA Region 5 has undertaken a major cleanup
initiative that has resulted in the filing of complaints against nearly every area industry
for violations of environmental statutes. The involvement of USEPA in litigation with
most of the riverside property owners made locating a  demonstration site problematic.

       The U.S. Steel Gary Works entered into a consent decree with USEPA Region  5
to perform sediment characterization and recyclability  studies in  the GCR.  Through this
consent decree, U.S. Steel Gary Works examined the  feasibility of recycling
contaminated sediments from the GCR in their steelmaking process. U.S. Steel Gary
                                       10

-------
Works provided the use of a plot of riverbank property within their Gary Works for the
ARCS pilot demonstration. U.S. Steel Gary Works provided the site because of their
interest in the use of the BEST technology as part of their sediment remediation studies
on the OCR.

       The Chicago District Corps of Engineers and U.S. Steel Gary Works executed a
right of entry agreement in which U.S. Steel Gary Works agreed to provide access to the
property, to provide assistance in traffic and site control during  mobilization and
demobilization, to help set up a temporary sediment storage area and to provide access to
an electrical supply.

Sample Collection

       Planning for sampling involved several considerations. First, it was desired to
test sediment with varying degrees and types of contamination, so that the efficiency of
the BEST process could be evaluated  for sediment  with varying  characteristics. Second,
a sufficient quantity of sediment had to be collected to ensure that each sample contained
enough solids for operation of the pilot unit.  Third, the samples were to be
representative of contamination found in the GCR/IHC.

Contracting

       Several contracts associated with the pilot-scale demonstration were awarded.
These included a contract for the removal of the sediment for pilot-scale tests, a contract
for the shipment of the TSCA-regulated sediment to RCC's laboratory, and a sole-source
contract for the pilot-scale demonstration. Tasks involved in the federal government
contracting process include:  preparation of a scope of work, preparation of a request for
proposals, advertisement of a contract, vendor selection, negotiation and award of
contract; and management of the contract  during the pilot-scale demonstration project.
                                       11

-------
Schedule
       A timetable of activities associated with the operation of the pilot-scale unit is
shown in Table 2 (USEPA, 1993).

            Table 2: Chronology of Activities for the Pilot-Scale Demonstration
Date
Event
Parties Involved
October, 1991
April 14, 1992
May 20, 1992
June 15, 1992
June 16, 1992

June 17, 1992
June 18, 1992
June 23, 1992
June 23, 1992
June 24, 1992
June 25, 1992
June 26, 1992
June 29, 1992
July 1, 1992
July 7, 1992

July  9,  1992
July 10,  1992

July 21,  1992
July 23,  1992
July 25,  1992

July 27,  1992
July 28,  1992
August 3, 1992
July, 1993
September, 1993
1994
1994
Project Plan finalized                           COE
Sediment characterization sampling              SITE
Treatment Contract awarded                    COE, RCC
Transect 28 (Sediment A) collection              COE
Sediment A Screening/homogenization            SITE
Transect 6 (Sediment B) collection               COE
Sediment B Screening/homogenization            SITE
Pilot Unit shipped from Bellevue, Washington     RCC
Pilot Unit arrived in Gary, Indiana                RCC
RCC setup crew arrived onsite                   RCC
Mechanical and electrical hookup of unit began    RCC
System checkout began/bench-scale testing initiated RCC
Pilot unit loaded with solvent
System checkout completed
Sediment A testing started
Visitors' Day

Completed testing on Sediment A
Pilot unit decontamination completed and
Sediment B testing started
Completed testing on Sediment B
Final pilot unit decontamination started
Pilot unit decontamination completed,
 demobilization started
Pilot unit demobilization completed
Pilot unit departed from Gary, Indiana
 Pilot unit arrived in Bellevue, Washington
 Publication of Technology Evaluation Report
Publication of Applications Analysis Report
Publication of Pilot-Scale Demonstration Report
Publication of GCR/IHC Concept Plan
RCC
RCC
RCC, SITE
GLNPO, COE,
RCC, SITE
RCC, SITE
RCC, SITE

RCC, SITE
RCC

RCC
RCC
RCC
RCC
SITE
SITE
GLNPO, COE
GLNPO, COE
Note: COE - US Army Corps of Engineers, RCC - Resources Conservation Company, SITE - Superfund
Innovative Technology Evaluation Program, GLNPO - Great Lakes National Program Office
                                      12

-------
TECHNOLOGY DESCRIPTION
Process Theory

Introduction--

       Resources Conservation Company of Seattle, Washington holds several patents for
the BEST solvent extraction technology. This technology uses triethylamine to remove
organic contaminants from soils, sludges, or sediments. The tetrahedral triethylamine
molecule forms when ethyl alcohol reacts with ammonia.  A nitrogen atom with its
electron cloud occupies one point of the three-sided pyramid, while the other three points
of the structure are occupied by three ethyl functional groups. Because ethyl groups are
essentially nonpolar,  and the electron cloud is polar, triethylamine exhibits dual polarity
(RCC, 1992).

       Triethylamine is successful at extracting organics because it is inversely miscible
(Figure 2). At temperatures below the critical solution temperature (about 16°C)
(USEPA, 1988), triethylamine and water are soluble in each other. Above this
temperature, triethylamine and water are only partially miscible.  The BEST process
exploits this physical property to separate the oil and the water from the feed by mixing
cold triethylamine (chilled below 16°C) with the feed to create an extraction liquid.
                          Inverse Miscibility
                          //  Wller and Trlethyl*mln«  /
                                 lmml»clW«
Temperature
Degrees C
SO

40

30
20

10
0
                   % Water 0   20  40  60  80 100
                         100 80  60  40  20  0 % Triethylamine
 Figure 2: Inverse Miscibility of Triethylamine and Water (Source: RCC, 1991)
                                        13

-------
       This liquid is a homogenous solution of triethylamine and water present in the
feedstock.  The solution solvates the organic contaminants in the feedstock and enables
the triethylamine to achieve intimate contact with solutes at nearly ambient temperatures
and pressures. The BEST process can therefore maintain efficient extraction when
handling feed mixtures with low solids content, and can dewater solids while
simultaneously extracting a portion of the organic contaminants. Afterwards, the
remaining organic contaminants are removed at temperatures above 16°C, where the
solubility of organic compounds in triethylamine  increases.

       Additionally, triethylamine's high vapor pressure allows the solvent to be  recovered
from the extract through steam stripping. By forming a low-boiling azeotrope with water,
triethylamine can be recovered from the extract to very low residual levels. Triethylamine
can be recovered from the treated solids with minimal energy because it has a heat of
vaporization one-seventh that of water. Extractions are carried out in an alkaline
environment (pH >10) to maintain triethylamine in its unionized form.  The high pH
promotes the precipitation of heavy metals as hydroxides, which in turn largely exit the
process with the treated solids.

Pilot Unit

       The BEST pilot plant is a solvent extraction system which demonstrates the
effectiveness of the BEST  process design by performing  onsite pilot-scale testing. Four
basic operations are involved in the system: extraction, solvent recovery  and oil polishing,
solids drying, and water stripping.  The extraction operation for materials with relatively
low solids content is additionally broken down into two types of extraction cycles. The
initial primary extraction cycles are termed "cold extractions," and secondary extractions
are termed "warm" or "hot extractions" depending on the temperature range at which they
are conducted.  Cold extractions were performed prior to the warm and hot extractions
during this demonstration because the sediment samples had low (when compared to soils
or hazardous wastes) solids contents (less than 60 percent by weight).  The cold
extractions dewater the solids since the water is miscible with triethylamine at colder
temperatures (RCC, 1993).
                                        14

-------
       The following sections discuss the four basic operations of the BEST pilot system.
 The extraction, solids drying, solvent recovery, and water stripping operations overlapped
 during the demonstration to increase the efficiency of processing the low-solids-content
 sediments. Oil polishing was only conducted at the end of a test when a sufficient volume
 of oil had been accumulated.

 Extraction —

       Figure 3 illustrates the BEST extraction process. Primary cold extraction cycles
 are conducted on high-water-content materials in a Premix Tank.  The tank is purged with
 nitrogen gas to displace oxygen after the feedstock and caustic are added to the tank, but
 before the addition of chilled solvent. Usually 40 liters of feed material and  120 liters of
 chilled solvent are used. A 50 percent sodium hydroxide solution is added to elevate the
 pH of the solution to a target of pH 11 and maintain the solvent's extractive  ability.  The
 sodium hydroxide is added in 50 ml increments.  A total of 2.5 kilograms of sodium
 hydroxide was added during extraction of sediment  sample A, and 3.8 kilograms were
 added during extraction of sediment sample B.  Approximately 341 and 405  kilograms of
 triethylamine were used during the extraction of Sediment Samples A and B, respectively.
 A paddle impeller mixes the feedstock/chilled solvent solution for 5 to 30 minutes, and
 then the solids in the solution are allowed to settle.  The minimum  settling time is
 established during the preliminary bench-scale laboratory testing.  After the majority of
 solids have settled, the triethylamine/water/oil solution is decanted through decant ports
 and transferred to a centrifuge that separates any fine solids remaining in the solution.

       This process of filling, mixing, settling and decanting is repeated for low solids
 content feeds prior to conducting higher temperature extractions in order to  accumulate
 sufficient solids in the Premix Tank.  For the dryer to achieve efficient drying following
the final extraction, about 30 liters of dewatered solids should accumulate in the Premix
Tank after the decanting.

       At this point, all the dewatered solids are pumped from the Premix Tank to the
Extractor/Dryer vessel for the higher temperature extractions. The fine solids that had
been separated from the decant solution in the centrifuge are added to the larger mass of
                                        15

-------
       Primary Extraction
         ADewatering
Secondary Extraction
  & Solids Drying
Solvent
Storage
Solvent Separation
Solvent Recovery
Sediment
                                                                                                                          Oil Product
                                                                                                                          Water Product
                              FigureS: BEST Solvent Extraction Process (Source: RCC, 1992)

-------
 solids in the Extractor/Dryer, so that nearly all the solids originally loaded into the system
 as feed are present.  The filtrate portion of the decant is pumped through a filter and into
 the Solvent Evaporator tank.

        At the start of the warm and hot extraction cycles, the Extractor/Dryer is filled
 with solvent, heated to temperatures of up to 77°C, and agitated for 5 to 15 minutes
 before settling.  The triethylamine/water/oil solution is decanted and sent to the Premix
 Tank for holding.  The decanted liquid is then centrifuged and the filtrate routed to the
 Solvent Evaporator.  The centrifuge solids accumulate in the solids chute. As in the cold
 extraction process, the steps of filling, mixing, settling, and decanting can be repeated at
 warm (greater than about 38°C) and hot (about 43°C) temperatures.  The treatability test
 results and the initial pilot plant runs help determine the total number of extraction cycles
 required for each sample processed during the demonstration.

 Solvent Recovery and Oil Polishing —

       During the solvent recovery process, the triethylamine/oil/water solution is routed
 to the Solvent Evaporator for heating to its boiling point.  Further heating evaporates an
 azeotrope of solvent and water, leaving the oil behind. This evaporation process
 continues until the water is depleted from the triethylamine/oil/water mixture. At that
 point,  the temperature of the boiling liquid rises until it reaches the boiling point of pure
 triethylamine.

       Evaporation continues at the higher temperature until nearly all of the water or
 solvent is removed. Using cooling water at about 38°C, heat exchangers condense the
 triethylamine/water vapor from the Solvent Evaporator to form a heterogeneous
 condensate (consisting of a solvent phase and a heavier water phase) at a temperature of
 about 43 °C.  The Solvent Decanter separates condensate into water and solvent phases.
 The Solvent Decanter is maintained at about 38°C so that the water and triethylamine are
 only partially miscible. The lighter solvent phase contains about two percent water, while
the heavier water phase contains about two percent solvent.

        During this demonstration project, a distillation stage was added to limit the
carryover of semivolatile compounds from the evaporator into the recycled solvent.
Reflux of recycled solvent to the rectifier was adjusted to provide the necessary
                                         17

-------
knockdown of semivolatiles. The recovered solvent is recycled to the Solvent Storage
Tank, while the recovered water drains into the water storage tank for storage until water
stripping operations remove residual triethylamine.

        When possible, oil polishing (further concentration of the oil) was also conducted
in the Solvent Evaporator after completion of all test runs. Most of the residual solvent
was freed by injecting a small amount of water into the oil to form an azeotrope with the
residual solvent which might otherwise never dislodge from the oil. When virtually all the
triethylamine was removed from the Solvent Evaporator, oil polishing was initiated. With
this polishing, the volume of hazardous material requiring further treatment was reduced.
Occasionally the amount of oil available in the Solvent Evaporator is not sufficient to
warrant oil polishing.  This was the case during the demonstration when treating Sediment
A, which contained less than one percent oil and grease.  In this instance the oil was left in
solution with the solvent and water. Sediment B, however,  did produce sufficient oil to
warrant polishing.

Solids Drying —

       Solids are dried in the Extractor/Dryer (the same vessel used for conducting warm
and hot extraction cycles). The Extractor/Dryer is covered with a steam jacket and
contains direct steam injection ports.  Steam is first supplied only to the steam jacket to
heat the Extractor/Dryer and its contents indirectly to about 77°C.  After removing the
bulk of the solvent by decanting and indirect heating,  steam is injected directly into the
vessel to reduce the triethylamine concentration to less than 1,000 mg/kg.  Rotating
paddle impellers within the Extractor/Dryer aid the drying process by increasing heat
transfer and reducing the drying time. The  triethylamine and direct steam form an
azeotrope which is directed to a dryer condenser. After all the triethylamine is removed,
the temperature of the vapor rises to the boiling point of water, and drying is continued for
a short time to ensure that residual triethylamine is removed. About five percent water
content by weight is left in the solids to keep the solids from being too dusty. After
drying, product solids are removed through a discharge port on the bottom of the
Extractor/Dryer and are collected in a lined receptacle.

       The vapor which had been directed to the dryer condenser during the drying cycle
is then condensed and drained into the Premix Tank.  The triethylamine/water mixture,
                                        18

-------
 and any carryover dust, are directed to the centrifuge for solids removal. The centrate is
 then pumped through the centrate filter and into the Solvent Evaporator, and combined
 with the triethylamine/oil/water solution already in the Evaporator.

 Water Stripping--

        The solvent is recovered from the water stream via stripping.  Decanted water
 contains about two percent triethylamine by weight, which is removed from the water by
 direct-contact steam stripping.  A caustic solution (50 percent sodium hydroxide) is
 added to the water receiver tank and thoroughly mixed with the solution to raise the pH of
 the triethylamine/water mixture. This ensures that the triethylamine is in the more volatile
 molecular unionized form rather than in the ionized form.

        When the desired pH of 11 is reached, steam is injected directly into the bottom of
 the steam stripping column. After heating the column, feed water is pumped through a
 feed preheater and heated to slightly above the solvent/water azeotropic boiling point.
 This hot feedwater enters the top of the column at a rate of approximately 500 ml/min.
 The feedwater runs at constant rate to the column's top tray, then down to all lower trays
 where it is stripped of residual solvent by upflowing steam.  Water stripper bottoms are
 returned to the water receiver tank during startup. When steady state is reached, the
 water stripper bottoms can be rerouted and discharged as product water. The solvent
 azeotrope vapors are directed to the water stripper condenser and the recovered solvent is
 recycled into the system for reuse.

 Pilot-Scale Process Equipment

       The BEST pilot unit is a scaled-down version of a full-scale treatment unit. Using
 smaller  versions of equipment components, the mobile pilot plant enables onsite testing to
be performed at a pilot-scale, thus demonstrating the effectiveness of the BEST Process
design.  Whereas a full-scale treatment system will be able to treat 140 cubic meters of
sediment per day, the pilot-scale unit treated 0.03  cubic meters of contaminated sediment
per day. In the full scale system, the washer/dryer operations are still batch, but all of the
other operations are continuous systems which can occur simultaneously.
                                        19

-------
       The pilot plant consists of two portable skids mounted on a lowboy trailer used to
transport the unit.  The process skid has two levels and contains most of the BEST
process equipment including the Premix Tank, the Extractor/Dryer, the Solvent
Evaporator, the Centrifilge, storage tanks, pumps, and heat exchangers. A second smaller
utility skid contains several utility systems that support the operation of the process skid,
including a refrigeration unit and a cooling water system.  The pilot plant requires 480
volts of three-phase power at 225 amps.  A support trailer is used to transport auxiliary
equipment. Figure 4 contains a schematic of the pilot unit. Each primary component is
described below.

Feed Hopper—

       A 45-liter Feed Hopper is used to add the feed (in this case the sediment) and
caustic to the Premix Tank at the beginning of a test (batch). The hopper is attached to
the top of the Premix Tank.  Feed is loaded into the top of the Feed Hopper by opening
the seal on top and pouring in the feed. The top seal is then shut and the bottom valve
opened, allowing the feed to drop into the Premix Tank. The hopper method of adding
feed minimizes the release of solvent vapor during loading.

Premix Tank--

       Cold extractions are conducted in the Premix Tank. As described in the Extraction
section, the initial cold extractions separate water and oil from solids when treating
materials with high oil and/or low solids content. The Premix Tank is a wide upright
cylinder with a total capacity of approximately 227 liters.  Within the tank, along the
cylindrical axis, there is a rotating shaft with mixing paddles attached and a bottom
scraper.  There are also cooling coils on the inside wall where chilled triethylamine is
constantly recirculated to maintain the tank contents below 16°C. On the outside of the
tank there are two decant ports on the side wall and a suction valve at the tank's bottom
for transferring the solids/solvent mixture to the Extractor/Dryer.
                                        20

-------
K>
                         A.PremixTank
                         B. Extractor/Dryer (Jacketed)
                         C. Solvent Decanter
                         D. Solvent Storage Tank
E. Solvent Evaporator (jacketed)
F. Water Receiver Tank
Q. Process Monitors (lace opposite side of unit)
H. Oil Decanter (not used during demonstration)
I. Treated Solids Receptacle
J. Polished Oil Collection Tap
K. CentrHuge
L Water Stripper Column
M. Primary Carbon Filter
N. Backup Carbon Fitter
O.Vent
P. Decani Pump
Q. Radiator/RadBtor Fan
Ft Non Contact Water Tanks
S. Refrigeration Unte
  (on opposite side of unit)
                                                       Figure 4: BEST PUot Plant (Source:  USEPA,  1993)

-------
Extractor/Dryer--

       The Extractor/Dryer vessel is also used for conducting extractions. It is also used
to dry the solids. This cylindrical vessel is a commercially available 34-gallon (129-liter)
batch blender. When the solids have a low oil and/or high solids content, all the
extractions are performed in this vessel. When the solids have a high oil and/or low solids
content (as was the case in this demonstration) the Extractor/Dryer is used to perform the
secondary warm and hot extractions that were first dewatered in the Premix Tank. The
Extractor/Dryer is  equipped with horizontally aligned mixing blades and is surrounded on
the outside by a steam jacket. Steam injected into the jacket provides heat during solids
drying and removal of residual triethylamine.

Solvent Evaporator--

       The Solvent Evaporator is used in the solvent recovery and oil polishing processes.
The solvent/water azeotrope formed during heating is evaporated from the oil in this tank.
Inside the Solvent Evaporator tank there are coils of stainless steel tubing through which
steam is circulated  to supply heat to the tank contents.  At the bottom of the tank a
recirculation pump maintains the contents as a homogeneous mixture. Vapors leaving the
top of the Solvent Evaporator pass through a set of condensers, which condense the
solvent/water mixture before it is transferred to the Solvent Decanter. The oil product,
which is not evaporated, is collected for disposal.

 Solvent Decanter--

       The Solvent Decanter is a small cylindrical tank where condensed solvent/water
vapors are received from the Solvent Evaporator. Within this tank, triethylamine solvent
and water are separated from one another since they are no longer miscible at their
elevated temperatures. The triethylamine is routed to the Solvent Storage Tank for
eventual reuse in the  extraction process. The water is directed to the Water Receiver
Tank.
                                        22

-------
Solvent Storage Tank—

       The Solvent Storage Tank is a large cylindrical vessel (capacity of approximately
285 liters) that supplies triethylamine for the entire pilot plant. Triethylamine is initially
loaded into the tank and can be pumped to the Premix Tank and Extractor/Dryer for
extractions. Recycled triethylamine is returned to the Solvent Storage Tank from the
Solvent Decanter.  Special precautions are observed when loading triethylamine into this
tank, since triethylamine is flammable.

Water Receiver Tank—

       The Water Receiver Tank is a stainless-steel barrel-shaped vessel with a capacity
of approximately 190 liters. This tank receives water that has been separated from solvent
in the Solvent Decanter and serves as the storage vessel for all contact water used in the
treatment process. A pump recirculates the water within the tank.

Refrigeration System—

       The Refrigeration System is located on the utility skid and is used to cool the
triethylamine solvent that is continuously circulated through the pilot plant. Components
of the Refrigeration System include a compressor hermetically sealed in a stainless-steel
box, a condenser, a noncontact cooling water storage tank, and a radiator that serves as a
heat exchanger.

Water Stripping Column—

       The Water Stripping Column removes residual triethylamine and other volatiles
from product water. As water recirculates, it is heated in a heat exchanger with steam and
injected at the top of the stripping column.  As water flows down through the system of
baffles within the column, steam injected at the bottom of the column rises up through the
column to remove volatiles in the water.  The water is then collected in a container at the
bottom of the column and transferred to a 55-gallon (208-liter) drum for disposal. The
solvent and other volatile vapors received at the top of the column are routed to the vent
system, condensed, and returned to the extraction  system.
                                        23

-------
Vent System--

       There is an atmospheric vent discharge from the BEST pilot plant Those gases
not recovered in the condenser system are vented into the carbon adsorber. Normally this
vent gas consists primarily of nitrogen purge gas, with traces of oxygen and other
atmospheric gases. Solvent gases present are normally condensed by the refrigerated vent
condenser.  However, to ensure that all organic vapors, including the solvent
triethylamine, are recovered with an efficiency of 95 percent or greater, as specified by 40
CFR Section 61.242-11, an activated carbon absorber is installed on the vent outlet
system.  The outlet of the primary carbon canister is monitored for triethylamine daily with
Draeger tubes. RCC performs this monitoring as part of routine unit operations. A
secondary carbon canister is operated in series in case breakthrough occurs in the primary
canister.

       In addition, RCC monitors the ambient air daily for ionizable organic vapors and
triethylamine using a photoionization detector and reaction tubes, respectively.  During
this demonstration, SITE also collected air samples vented to the atmosphere from the
pilot plant and analyzed them for triethylamine to evaluate the emission control
effectiveness of the BEST pilot plant.

SEDIMENT SAMPLE COLLECTION

       Sediment samples treated in the pilot-scale demonstration of the B.E.S.T process
were collected from two separate locations within the OCR (Figure 5).  Transect 6 is
downstream of a wastewater discharge from a coke plant. Sediment from this location
(Sediment B) was expected to contain high levels of organic contaminants (PCBs and
PAHs) and moderate levels of metals. Sediment was also collected from Transect 28
(Sediment A), downstream of a wastewater discharge from an oU-skimming lagoon and
primary bar plate mills.  Sediment near Transect 28 was expected to contain higher levels of
metals and lower levels of organic compounds. The efficiency  of the BEST process for
sediment samples with varying characteristics could therefore be evaluated.

       Each composite sample was representative of the top 1.5 meters or deeper of
sediment. The portion of the OCR from which the sediment  was collected has a water
                                       24

-------
depth of about 0.6 to 1.5 meters and a width of about 18 to 21 meters.  A small
pontoon boat was used during sediment collection. The sample collectors pushed a
hollow, four-inch (10-centimeter) diameter plastic tube down into the sediment, and
pulled up the tube containing the sediment (Plate 1).  Approximately 575 liters
(0.6 cubic meters) of sediment were collected at each site in this manner and placed
into 5-gallon (19-liter) buckets for shipment. Each bucket was labeled with the sample
location, and date and time of collection.  Each container was marked with warning
labels stating that the buckets contain sediment contaminated with heavy metals and
TSCA levels of PCBs. The samples were then brought to the U. S. Steel Gary Works
Buchanan Street site (Figure 6) for the treatment demonstration.  A secure metal roll-off
box was used to store the feed material during the demonstration project.

SEDIMENT PRETREATMENT

       Sediment pretreatment consisted of prescreening and homogenization prior to the
demonstration tests.  Bench-scale treatability testing was performed to provide guidelines
for the operation of the pilot unit.  SITE'S contractor performed the screening and
homogenization and RCC performed the treatability testing. Pretreatment required two
days to complete.

Prescreening

      Prescreening separated oversize material that could interfere with pilot plant
operation from the sediment. The pilot unit can process material up to 1/4 inch (12.7 mm)
in diameter.  However, to minimize abrasion on the equipment, the sediment was screened
with an  18 inch (46 cm) diameter vibrating screen equipped with a 1/8 inch (6.4 mm)
screen.  A lined staging area was set up at the site and each of the two sediment samples
were screened separately by pouring the sediment into the top of the Kason separator.
The buckets were weighed before and after screening so that the mass of feed screened
could be recorded. The material greater than l/8th of an inch (6.4 mm) was directed into
a large trough on the opposite side of the separator. This material was also weighed. The
less contaminated sample, sediment A (Transect 28), was screened first to minimize the
possibility of cross-contamination.
                                       25

-------
           Plate 1: Collection of Sediment Sample B
Plate 2:  Homogenization of Screened Sediment using a Rotary Mixer
                             26

-------
                                           0 KILOMETERS
          EXPLANATION

    O Industrial effluent outfall

    0 Municipal effluent outfall

       Direction of flow
                                                                   Collection Location for
                                                                        Sediment B
                                                                        (Transect 6)
                                        Collection Location for
                                             Sediment A
                                            (Transect 28)
                                                             ARCS DEMO
Gaiy wutawatar TrMtnMrt
        Ptart(GWTP)
JNO EAST-WEST  T
            FigureS:  Sediment Sample Collection Locations (Source:  USEPA, 1993)
                                              27

-------
               ^-^.
                       ;-*-*-«.
to
00
                                                                          Tank for Steam
                                                                      Condensate/Contalnment
                                                                      Area Rainwater (5000 gal.)
                                                                 Electrical
                                                                 attribution   Sieging Aree
                                                                   Panel
vinnn'~v
                                                                                                                                          Legend

                                                                                                                                          Temporary Fence
                                                                                                                                          Chain Link Fence
                                                                                                                                          with Barbed Wire
                                                                                                                                          Chain Link Fence

                                                                                                                                          Permanent Steel Posts

                                                                                                                                          Bumper from Old
                                                                                                                                          Parking Lot
                                                              Figure 6:  Demonstration Site (Source.  USEPA,

-------
       A total of 820 kilograms of Sediment A were screened.  About 6.4 kilograms (0.8
 percent of the material) were rejected as oversized. A total of 663 kilograms of Sediment
 B were screened, and about 3.9 kilograms (or 0.6 percent) of the material were rejected as
 oversized.  The rejected material was placed in drums for disposal.

 Homogenization

       Homogenization was necessary to ensure that the entire volume of screened
 sediment was consistent in terms of sediment type, solids content and contaminant
 concentration. Each sediment type was formed into a discrete sample that would be
 consistent during bench-scale treatability testing and during each test run of the pilot-scale
 unit.  Sample consistency helped ensure that the subsamples collected for the bench-scale
 tests were representative of the larger volume of sediment to be treated in the pilot-scale
 unit, and that the different runs could be evaluated based on differing operating conditions
 rather than differing sample consistency.

       In order to thoroughly mix each sample, each sample was homogenized separately.
 All of the screened sediment from Sediment A was poured into a lined 220-gallon
 (835-liter) trough.  An industrial mixer was mounted on a wooden support placed on the
 trough walls. The mixer was moved  back and fourth along the entire length of the trough
 (Plate 2). After mixing, the material was transferred back in the original 5-gallon (19-liter)
 metal containers. The process was then repeated for Sediment B.

 BENCH-SCALE TREATABILTY TESTS

 Purpose

       The ARCS program performed bench-scale tests on sediment from several AOCs
 to support the decision of which technologies should be evaluated on a pilot-scale basis at
 the various AOCs.  RCC performed bench-scale treatability tests on the homogenized
 sediment prior to beginning the pilot-scale tests because the bench-scale tests provide
guidelines for the operation of the pilot unit. Data collected during bench-scale tests
included settling data, compositional data (oil & grease (O&G), solids content, water
                                       29

-------
content, bulk density, PAH and PCB concentration), pH values and observations on
centrifugation and agglomeration characteristics.

       The bench-scale tests were designed to provide data which simulate performance
of the pilot-scale and the foil-scale system. To simulate the pilot-scale process on the
smaller bench-scale, laboratory equipment resembling the pilot plant components is used.
For instance, a resin kettle is used as an extraction vessel. This was immersed in a cooling
bath for cold extractions and in a heated bath for warm and hot extractions. Gradations
that correspond to decant port levels on the premix tank or extractor/dryer vessel were
marked on the kettle, depending on the type of extraction being simulated.  An air-driven
prop was placed in the kettle to mix feed material and a floor mounted centrifuge was
employed to remove fines from the decant following extraction. A forced-draft oven was
used to dry product solids.

       Settling data was used to predict the level to which material will settle in the pilot
unit's extractor/dryer vessel and the time required to reach that level.  RCC used the
compositional data to determine the amount of sediment to load into the pilot-scale unit
per batch.  Each batch should ideally contain at least 30 liters of solids, so that the
extractor/dryer vessel is filled to at least the one-quarter level. This level is based on
keeping enough solids in the dryer during drying to allow proper heat transfer and mixing.
The weight of sediment to add to the unit was determined by dividing the weight of the
dry solids by the solids fraction of the feed.  The volume of feed to be charged to the
extractor/dryer was determined by dividing the weight of feed by the feed bulk density.

       The pH data was used to determine the amount of caustic to be added to the feed
during testing. Caustic was added to elevate the pH of the feed in order to enable
recovery of the triethylamine from the separated feed products. Compatibility tests
determined that no adverse reactions would occur between the sediment and the
triethylamine.  A summary of RCC's bench-scale testing is listed in Table 3.
                                        30

-------
    Table 3: Summary of Bench-Scale Testing in Support of the Pilot Demonstration

     PCB Content (mg/kg, dry)                 PAH Content (mg/kg, dry)

 Sample      Feed    Product Solids      Feed   Product Solids  Extractions

 Sediment A     6         <1              890        43          6

 Sediment B   700         <1           76,900      346          7


PILOT UNIT OPERATION

Unit Mobilization

       The pilot plant was brought to the demonstration site during bench-scale testing
and placed on the bermed containment area. The setup phase was completed in two days
by installing all process components and operating each piece of equipment individually.
All components removed for shipping were assembled and all interconnections were made.
Utilities were then connected.  Unit checkout began after all equipment had been properly
installed. After testing all the functions individually, the solvent was loaded into the unit
and checked.  The BEST pilot-scale plant is shown in Plate 3.

Feed Analysis

       Five batches of each sediment type were processed through the BEST unit.  Prior
to processing each batch, each sediment type was sampled. These samples were analyzed
for the parameters listed in Table 4. Volume H of the TER (EPA/540/R-92/079a)
contains detailed results of the analyses conducted by SITE. As a condition of RCC's
TSCA treatability permit, RCC analyzed the fifth batch of each sediment type for TCLP
chlorobenzene, TCLP 1-4 dichlorobenzene and TCLP hexachlorobenzene.
                                      31

-------
          Plate 3: BEST Solvent Extraction Pilot-Scale Unit
Plate 4: Samples of Process Residuals - Treated Sediment, Water and Oil
                             32

-------
                    Table 4: Analysis Performed on Feed Sediment

                PCBs                            volatile solids
                PAHs                            total cyanide
                oil & grease                      reactive cyanide
                moisture                         reactive sulfide
                TCLP metals1                    particle size
                proximate/ultimate2               total phosphorus
                total metals                      pH
                TRPH3
             iToxicity Characteristic Leaching Procedure
              ^Note: Proximate/ultimate analysis included a determination of
              the moisture, volatile matter, total carbon, fixed carbon, hydrogen,
              nitrogen, sulfur and ash content of a sample as weight percentages.
              Btu per pound is also determined to indicate a material's fuel
              contribution during incineration.
             3Total Recoverable Petroleum Hydrocarbons
Batch Processing

Introduction--

       Five batches of each sediment were processed. To minimize the possibility of
cross contamination between samples, the less-contaminated Sediment A was treated first,
and the pilot unit was decontaminated and the solvent was replaced with unused solvent
prior to treatment of Sediment B. To determine the operating parameters that would
achieve maximum contaminant removal, the first three batches (phase I) were run with
varying operating parameters.  The operating parameters resulting in the greatest percent
contaminant removal were repeated for the fourth and fifth batches of sediment.  The
results of the three optimal batches were used to evaluate the BEST process.

Feed Loading and Initial Extractions--

       Sediment A (Transect 28) was added to the premix tank in one step. Sediment B
(Transect 6) was added incrementally to the Premix Tank, yielding a total feed load of 98
liters.  Sediment B had a lower solids content, which required that a larger volume of feed
be used to achieve the required solids loading. An incremental process was used for
Sediment B to prevent the heat of solution from rising above 10°C and to accumulate
                                         33

-------
sufficient solids to allow proper drying in the Extractor/Dryer vessel. Table 5 reflects the
incremental additions of feed and solvent for Sediment B treatment. Standing water
accumulated above the sediment in the Sample B buckets. The water was not removed
from all batches of Sample B because it was not part of the approved test plan.  However,
water was removed from one of the "optimization" batches in order to determine if water
removal would be beneficial during foil scale operations.  Testing of the dewatered
sediment was successful.

       For each of the remaining batches of each sediment sample, the separated water
and oil were reincorporated with the solids by mixing the contents. After mixing each
bucket, an equal volume was taken from each, except for the last bucket from which a
proportional amount of the smaller volume used was collected.  These aliquot samples
were composited in a stainless-steel pail for individual weighing on a platform scale. After
the sediment was loaded into the pilot unit, the emptied buckets were reweighed to
acquire the mass of sediment treated.  This measurement was critical for subsequent mass
balance determinations.

       Table  5 contains a summary of the amount of feed used in each batch (or run), the
number of incremental feed portions, and the number of extractions performed on each
batch.  Before each batch began, the sediment was weighed and the amount of the caustic
was calculated based on the sediment weight. Both the sediment and initial caustic dose
were fed to the Premix Tank through the Feed Hopper.

       The release of solvent vapor during loading was minimized by closing the seal on
the top of the hopper after feeding the sediment, and opening the seal on the bottom of the
hopper when the sediment dropped into the Premix Tank.  Chilled solvent was added to
the Premix Tank following the sediment addition and the contents were mixed for five
minutes.

       After mixing, the contents of the premix tank were allowed to settle.  The liquid
triethylamine/oil/water mixture was drained off the top, making room for another
incremental portion of the feed to be added along with the caustic and  solvent. This
procedure was repeated until the entire feed  load had been added to the Premix Tank.
                                       34

-------
                          Table 5: Feed Loading Summary
Total
Feed
(liters)
Sediment A
Batch 1
Batch 2
Batch 3
Batch 4
Batch 5
Sediment B
Batch 1
Batch 2
Batch 3
Batch 4
Batch 5

45
57
53
57
53

98
87
78
87
85
Number of
Incremental Feed
Portions

1
1
1
1
1

3
4
4
4
4
Number
of
Extractions

6
7
7
7
7

7
6
7
6
6
           Source:  RCC, 1993

After decanting all the liquid, the concentrated solids remaining in the bottom of the
Premix Tank were pumped into the Extractor/Dryer vessel for further extractions. All of
the extractions conducted in the Premix Tank were performed cold (below the
temperature where water and triethylamine are miscible.) All of the subsequent
extractions conducted in the Extractor/Dryer vessel were performed at elevated
temperatures (greater than 32°C).  Because most of the water had already been removed
during the cold extractions, it was no longer necessary to continue operating in the cold
temperature range.

       Tables 6 and 7 present the sequence of extractions cycles for each of the five runs,
including the total number of extractions and the temperature at which each was
conducted. The optimum conditions chosen for each sediment type were significantly
different. The best conditions  for treating Sediment A were two initial cold extractions
below 16°C followed by one warm, three hot, and a warm extraction. The water  content
of the solids product contributed to the large variation in extraction temperatures.
                                        35

-------
                  Table 6: Extraction Sequence used for Sediment A




                            Extraction Temperature (°C)




                           Phase 1                            Phase
Extraction
Cycle
1
2
3
4
5
6
7
Runl

cold (17)
warm (41)
warm (35)
warm (35)
warm (39)
hot (77)
•™
Run 2

cold (10)
cold (4.4)
cold (3.3)
warm (37)
warm (52)
hot (71)
hot (71)
Run 3

cold (12)
cold (7.2)
warm (38)
hot (68)
hot (74)
hot (74)
warm (49)
Run 4

cold (8.9)
cold (5.6)
warm (43)
hot (68)
hot (73)
hot (73)
warm (48)
Run 5

cold (11)
cold (7.8)
warm (36)
hot (67)
hot (75)
hot (71)
warm (49)
Source: USEPA, 1993
                  Table 7: Extraction Sequence used for Sediment B




                            Extraction Temperature (°C)




                     Phase 1                            Phase II
Extraction
Cycle
1A1
1A2
1A3
IB 1
IB 2
IB 3
2
3
4
5
6
7
Runl
cold (9.4)
cold (8.3)
(nc)
cold (5)
cold (12)
cold (11)
hot (63)
hot (67)
hot (72)
hot (64)
hot (69)
hot (62)
Run 2
cold (-2.2)
cold (5.6)
cold (3.3)
cold (3.9)
cold (8.3)
cold (2.2)
hot (67)
hot (69)
hot (66)
hot (67)
hot (66)

Run 3
cold (0)
cold (4.4)
cold (4.4)
cold (-1.7)
cold (3.3)
cold (7.8)
hot (66)
hot (66)
hot (67)
hot (66)
hot (63)
hot (66)
Run 4
cold (-2.2)
cold (8.9)
cold (3.9)
cold (11)
cold (12)
cold (7.8)
hot (64)
hot (69)
hot (77)
hot (68)
hot (70)

Run 5
cold (11)
cold (5)
cold (3.9)
cold (3.9)
cold (7.2)
cold (6.7)
hot (63)
hot (71)
hot (67)
hot (68)
hot (67)

Source: USEPA, 1993




Note: Bold columns indicated the three optimum runs, NC = Not Conducted.
                                        36

-------
       Treated solids generated during Run 1 were wet and soupy, likely due to
inadequate dewatering by the one cold wash at 17°C, which is near the triethylamine-

water miscibility threshold. When two cold extractions were conducted during run 2 at

temperatures well below 16°C, the treated solids contained much less moisture, but still

formed clumps of solids. With the extraction scenario shown in Table 6, the treated solids

were dry and essentially cohesive, so this extraction scenario was repeated twice in Phase

II testing for Sediment A.


       Before testing of a batch began, RCC assembled a list of primary control

parameters. Many of the potentially manipulated variables were changed during testing.

Most changes in variables occurred during the Phase I optimization runs. The variables

that affect extraction efficiency include the number of extraction stages, solvent/feed ratio,

extraction temperature, and extraction time (Table 8).


                        Table 8: Primary Control Parameter Summary


              Parameter              Potential Manipulated Variables


        Extraction Efficiency           1) # of Extraction Stages
        (Extractor/dryer operation)      2) Solvent/Feed Ration
                                    3) Extraction Temperature
                                    4) Extraction Time

        Extractor/dryer Capacity 1) Volume feed charged to E/d
        (For full scale estimate)  2) Length of Fill/Extraction/Decant Cycle
                                    3) Length of Drying Cycle

        Solvent Residual in Solids       1) NaOH addition
        (<150 ppm pilot test goal)       2) Drying Duration

        Centrifuge Capture            1) Centrifuge RPM
        (fines removal step)            2) Bowl/Scroll Differential
                                    3) Centrifuge Pool Depth
                                    4) Feed Rate to Centrifuge

        Decanter Performance          None

        Solvent Residual in Water       1) NaOH addition to feed water to
                                    stripper column
                                    2) Steam to stripper

       Source: RCC, 1993
                                         37

-------
       During treatment of each batch, process operating information was collected,
including the volume of caustic added to the Premix Tank, Water Stripper Column and
Extractor/Dryer vessel; the volume of solvent and water used for extractions; and the
volume of water used for steam generation. The volume of caustic added during a test
was determined by incrementally pouring the solution from a graduated cylinder into the
desired vessel. Solvent and water volumes were determined by visually monitoring tank
levels that are equipped with sight glass gauges.  The water added to the water stripper
column was measured by monitoring the pressure drop across an orifice in the steam line.
Following testing of each batch, the mass of process product streams was determined by
weighing the contained products on a platform scale. These volumes were used in the
mass balance calculations.

RESIDUALS MANAGEMENT

       Operation of the pilot plant produced product streams of oil,  water (contact and
non-contact) and solids. Samples of theses product streams are  shown in Plate 4.  Other
non-product wastes included steam condensate from steam traps, boiler blowdown,
stormwater, and vent  gas.  In addition, other residuals resulting from the pilot-scale
demonstration included: untreated sediment, waste solvent, and miscellaneous items
(contaminated PPE, etc.).

       During the pilot-scale demonstration, residual product solids, product water,
product oil and decontamination water were collected in 5-gallon (19-liter) buckets and
stored hi a secure, lined and bermed metal roll-offbox. Logs were kept by both USEPA
SITE and RCC for material transferred to RCC at the beginning of the  demonstration and
to USEPA SITE at the end of the demonstration. All TSCA and/or RCRA regulated
material was manifested.  The entire sediment B sample and its treatment residuals (even if
less than 50 ppm PCBs) were regulated under TSCA since the untreated sediment
contained PCBs in excess of 50 ppm.  The product oil from sediment A was regulated
under TSCA since it contained PCBs in excess of 50 ppm. The only RCRA regulated
material used in the demonstration was triethylamine, which is a hazardous waste due to
its flammability. No RCRA material was generated during the pilot-scale demonstration.
                                       38

-------
       USEPA SITE disposed of the untreated sediment, the treated sediment solids,
product water, product oil and miscellaneous items at Aptus, Inc. in Lakeville, Minnesota,
a permitted treatment, storage and disposal facility. RCC transported the used
triethylamine to Washington via a hazardous waste transporter for disposal at Burlington
Environmental, Inc.

       The non-product wastes were handled as follows. The condensate from the steam
traps is clean water from the pilot plant auxiliary boiler.  This water was not in contact
with any process waste stream, and was collected as non-contact water. The boiler
blowdown water was used to prevent scaling on the heating elements.  This water
normally was generated at a rate of less than five gallons per day and was processed with
the steam condensate.

       Rainwater was collected in both the secondary containment pan and on the tertiary
containment (a raised edge membrane surface). As required by RCC's TSCA permit, the
secondary water was treated with carbon filters.  The filtered water was then added to the
tertiary containment water. The water collected in the tertiary containment was sampled
and analyzed for PCBs. The water was then disposed as non-hazardous.  Lack of PCB
contamination was  used to show that no sediment was spilled on the tertiary containment,
which eliminated the need to test for PAHs and heavy metals.

       The atmospheric vent discharge from the system was necessary to eliminate non-
condensable gases from the various condenser systems in order to prevent reduction of
heat transfer efficiency. The vent gas usually consisted primarily of nitrogen purge gas
with traces of oxygen and other atmospheric gases. To  assure that organic vapors were
removed with an efficiency of 95% or greater (40 CFR Section 61.242-11), an activated
carbon scavenger was located on the vent outlet. SITE disposed of the activated carbon
at Aptus, Inc., hi Lakeville, Minnesota.
                                       39

-------
                           MONITORING PROGRAM
INTRODUCTION
       The goals of the ARCS program for the pilot-scale demonstration and subsequent
analyses included evaluating the efficiency of the BEST technology at reducing organic
contamination in GCR sediment with varying levels of contamination and developing cost
estimates for a full-scale application of the process. The goals of the SITE Program for
this demonstration project were to:

       • assess quality of the treated solids, residual product water
         and concentrated oil product;
       • develop capital and operating costs for the technology;
       • develop an overall mass balance for organic contaminants
         around the BEST pilot plant;
       • evaluate the technology's effect on metals found in the sediment; and
       • assess the biodegradation of residual triethylamine in the product solids.

       SITE designed and carried out a sampling and analysis strategy to evaluate the
performance of the BEST pilot plant. The sampling and analysis program was based on
the following critical test objectives established by RCC and the  SITE program:
a) removing 96-99 percent of PAH and PCB contaminants from river sediments, b)
achieving a mass balance in the range of 80 to  130 percent for feed material mass into the
pilot unit versus the product solid, oil and water masses, and c) producing product streams
of water, solids, and oil having specific triethylamine concentrations of less than 80 ppm,
150 ppm and 1,000 ppm, respectively. These three objectives were critical for the
assessment of the ability of the solvent extraction technology to meet the above objectives.
The noncritical objectives were to determine the technology's general applicability and to
document process performing the analyses listed in Table 9.

       The Demonstration Plan and Quality Assurance Project Plan (QAPP) detailed the
procedures which were used to collect and analyze the samples.  Since PAHs, PCBs and
O&G were known to be present in both sediment samples, and were targeted for removal
by the BEST process, these analyses were critical for all media except for the vent gas.
Moisture and TCLP metals analyses were critical because both sediment samples had low
solids contents and contained significant concentrations of heavy metals.
                                       40

-------
           Table 9: Summary of Analyses Conducted for the GCR/IHC Pilot Scale Demonstration
Parameter

Critical
PAHs
PCBs
Oil & Grease
Moisture
Triethylamine
TCLP metals

Noncritical
TSS
Proximate/Ultimate
Total Metals
TRPH
Volatile Solids
Total Cyanide
Reactive Cyanide
Reactive Sulfide
Particle Size
Total Phosphorus
PH
IDS
TOC/TIC
BOD
Conductivity

Special Studies
Biodegradation
Untreated
Sediment

*
*
*
*

*



*
*
*
*
*
*
*
*
*
*







Treated
Sediment

*
*
*
*
*
*



*
*
*
*
*
*
*
+
*
it






*
Water
Phase

*
*
*

*



*

*
*
*
*



*
*
*
*
*
*



Decant
Water

*
*
*





*

*
*







*






Product
Oil

*
*

*
*





*















Intermediate
Solvent/Oil

*
*
























Solvent Feed and
Recycled Solvent

*
*






* '

















Vent Gas






*





















Source: USEPA, 1993

-------
SAMPLING LOCATIONS

       The process streams sampled and analyzed included: the untreated sediments (raw
feed), product solids, product water, product oil or oil/solvent mix, recycled solvent, and
vent emissions.  In addition to these process streams, the purity of the product
triethylamine was tested, and the decant water collected from Sediment B buckets holding
the feed material was sampled. The decant water was analyzed to determine the quality of
the sediment water phase that would have to be treated separately if pretreatment
decanting of sediments was to be considered an option. Figure 7  illustrates the various
sampling locations.

SAMPLING PROCEDURES

       Samples were generally collected in accordance with the Sampling Plan described
in the Demonstration and Quality Assurance Project Plan (SITE,  1992). Modifications to
and deviations from those plans occurred during the course of the project involving both
field and laboratory activities.  All modifications were documented and submitted to the
project management. Table 9 provides a summary of analyses performed.

Untreated Sediment (Raw Feed)

       The sediments were stored in 5-gallon (19-liter) buckets prior to treatment. Each
bucket was individually and equally mixed to incorporate oil and water phases back in
with the settled solids. Aliquots from each remixed bucket were proportioned out and
composited in a container.  Sediment from the composite sample was placed into sample
jars.

Product Solids

       Product solids were released by way of a port in the bottom of the Extractor/Dryer
and fell into a large plastic-lined can. Solids were scooped out and placed into sample
jars. These samples were considered representative of the  entire batch of solids since the
contents of the Extractor/Dryer are thoroughly mixed with impellers during solids drying.
The sampling personnel wore respirators during this step to prevent inhalation of dust.
                                        42

-------
 Untreated
 Sediment
     i

                          Feed Buckets
                               Product
                            |Triethylamine
Triethytamine
   Drum
Primary Extraction/ j    Secondary Extraction/
   Dewaterlng   I        Solid* Drying
               I
                                            Solvent Storage     I   Solvent Separation I Solvent Recovery


                                      I                       I                   I
   Tank     1
VT
  Cold Wash Solvent |
                      ±
                                    i      I™
                                    I      I
                               i
                         Extractor/Dryer
                   dean
                  Solvent
                       T
                           Steam
 I
  Solid
 Sample      '
      Liquid      '
     I Sample     |
                          Clean Solids
                              Product
                                                              I
                                               Centrifuge
                                                 Clean
                                                Solvent
                         Solvent
                         Makeup
  i '

 I
Product
 Solids
                                                   Recycled
                                                 Triethylamine
                                                                                    Water      Product
                                                                                    Product       Oil
                        Figure 7: Solid and Liquid Sample Locations (Source: USEPA, 1993)

-------
Product Water

       Product water was sampled at two locations. Product water that was to be
analyzed for triethylamine was collected directly into 40 ml vials from a hose from the
Water Stripper Column once it was determined that the stripper temperature and water pH
had stabilized. The vials were then immediately sealed and cooled.  The majority of the
sample product water was sampled from 55-gallon (208-liter) drums which contained
stripped product water.

Product Oil

       Product oil is the concentrated organic oily liquid left in the solvent evaporator
tank after oil polishing had removed practically all triethylamine. Product oil was
produced from sediment B.  Sediment A did not originally contain enough O&G for oil
polishing.  Product oil was collected directly in sample containers from a tap in the line
extending from the Solvent Evaporator tank.  Samples of the unpolished oil/solvent
mixture produced from Sediment A treatment were also collected in this manner.

Oil/Solvent Mix

       After each test run, a sample of the oil/solvent mix accumulating in the solvent
evaporator tank was collected to provide an estimate of the buildup of contaminants
within the Solvent Evaporator as they were being extracted from the solids.

Recycled Solvent

       The separated solvent from the contaminated phases following extraction was
recycled back into the extraction process for reuse.  This recycled solvent was sampled
prior to the start of each test run to determine its purity at  the time of use. The samples
were collected directly into vials from a valve tap located on the cooling loop.

Vent Emissions
       Samples of vent gas were collected to determine if the scrubbers were efficiently
removing triethylamine.  The samples were taken at a constant rate throughout each run.
                                        44

-------
However, the collection times and rates varied from one run to another due to
interruptions in the runs and differences in the lengths of individual runs.  Samples were
collected using a sampling port in the vent pipe located downstream from an elbow and
upstream from the flame arrestor. The tip of the sample probe was positioned near the
center of the vent cross section. The sampling rate for each run was held constant
depending on the operation time needed to collect a minimum 100 L sample volume.

                         RESULTS AND DISCUSSION

PROCESS OPERATION

       Optimal extraction efficiencies were attained by conducting a minimum of one cold
extraction initially to dewater the feed, followed by warm and/or hot extractions to
remove any remaining organic contaminants.  The optimum extraction sequence for
Sediment A was: 2 cold extractions, 1 warm extraction, 3 hot extractions and 1 warm
extraction. The optimum extraction for Sediment B was 2 cold extractions followed by 5
hot extractions.  The approximate temperature ranges for the various extractions were as
follows:  1) cold (-2°C - 17°C),  2) warm (35°C - 52°C), and 3) hot (62°C - 77°C).

CHEMICAL CONCENTRATION DATA

       Unreduced analytical data is contained in Volume n of the Technology Evaluation
Report (TER), which summarizes the activities and results of the pilot-scale demonstration
testing of the BEST system. The TER contains detailed summaries of the following:

       •  PCB, PAH and O&G Removal from the sediment
       •  Triethylamine residual testing on the treated solids,  product water and oil phases
       •  Triethylamine air emission testing on the vent gas
       •  Triethylamine biodegradation testing on the treated solids
       •  Organic analyses of product water, recovered solvent and product oil
       •  Moisture contents on the raw feed and treated solids
       •  Total and Leachable Metals on the raw feed and treated solids
       •  Supplemental analyses (TRPH, volatile solids, total and reactive cyanide,
         reactive sulfide, total phosphorus and pH).
                                      45

-------
       A brief summary of the chemical concentration data is presented here. Tables 10
through 15 contain concentrations and removal efficiencies for the feed and treated solids
for PAHs, PCBs, O&G, total metals, TCLP metals and supplemental analyses. Tables 16
through 18 contain concentrations of PAHs, PCBs, metals and supplemental analyses of the
water product. Table 19 contains PAH and PCB concentrations in the solvent.  Table 20
contains triethylamine concentrations in the treated solids, product water and oil phases for
each sample.

                   Table 10: PAH Concentrations and Removal Efficiencies
PAHAnalyte
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(ghi)perylene
Chrysene
Dibenz(a,h)anthracene
Fluoranthene
Fluorene
Indeno( 1,2,3 -cd)pyrene
2-Methylnaphthalene
Naphthalene
Phenanthrene
Pyrene
Total PAHs
Feeda
68
<16
22
25
24
23
17
15
25
<18
76
51
15
25
<18
92
67
548
SEDIMENT A
Treated Percent
Solids3 Removal15
1.3
<0.8
1.3
0.52
0.34
0.36
0.22
0.20
0.52
<0.76
1.4
1.9
0.18
3.7
5.1
3.6
1.0
22
98.1
—
94.1
97.9
98.6
98.4
98.7
98.6
97.9
—
98.2
96.3
98.8
85.2
—
96.1
98.5
96.0
SEDIMENT B
Treated Percent
Feeda Solids3 Removalb
12800
210
2370
1050
810
857
533
457
937
140
4280
7290
547
6410
18700
10800
2810
70920
42
6.6
16
4.7
4.6
4.1
3.6
2.3
4.7
<2.9
16
35
2.2
83
230
41
12
510
99.7
96.9
99.3
99.6
99.4
99.5
99.3
99.5
99.5
>97.9
99.6
99.5
99.6
98.7
98.8
99.6
99.6
99.3
Source: USEPA, 1993

a Concentrations reported in mg/kg (dry weight basis) and are the average of the three optimum runs each
 sediment. (Sediment A= Runs 3, 4 and 5; Sediment B = Runs 2, 4, 5).
b Percent Removals = Feed Concentration - Treated Solids Concentration  x  100
                            Feed Concentration
                                         46

-------
               Table 11: PCB Concentrations and Removal Efficiencies
Parameter
Sediment A
Total PCBs-Feed
Total PCBs-Treated Solids
Percent Removal
Sediment B
Total PCBs-Feed
Total PCBs-Treated Solids
Percent Removal
Rl

7.33
<0.07
>99

364
1.5
99.6
Test Run
R2 R3

6.41
0.20
96.9

316
2.1
99.3

8.01
0.05
99.4

495
1.2
99.8
R4a

11.8
0.04
99.7

462
1.8
99.6
R5

16.4
0.04
99.8

497
1.4
99.7
Avg.b

10.0/12.1
0.08/0.04
99.2/99.7

427/425
1.6/1.8
99.2/99.6
Std.Dev.b

4.1/4.2
0.07/0.006
—

82/96
0.35/0.35
—
Source: USEPA, 1993
a Concentrations reported for Run 4 are the average of three replicate measurements.
b Two values are given; the first pertains to all five runs and the second pertains to the three optimum runs (Sediment A =
   Runs 3, 4, 5 and Sediment B = Runs 2, 4, and 5).
c All PCB concentrations are in mg/kg dry weight.

-------
                                              Table 12: O & G Concentrations and Removal Efficiencies
oo
Parameter
Sediment A
Total O&G-Feedc
Total O & G-Treated Solids
Percent Removal
Sediment B
Total O & G-Feed
Total O & G-Treated Solids
Percent Removal
Rl

9,400
195
97.9

66,400
1,800
97.3
R2

7,800
169
97.8

116,000
1,330
98.9
Test Run
R3

7,400
203
97.3

67,300
1,490
97.8
R4a

6,600
66
99.0

167,000
1,230
99.3
R5

6,700
65
99.0

99,100
1,810
98.2
Avg.b

7580/6900
140/111 '
98.2/98.4

103,000/127,000
1,530/1,460
98.5/98.9
Std. Dev. b

1130/436
69/79
—

41,600/35,300
266/310
—
               Source:  USEPA, 1993
               a Concentrations reported for Run 4 are the average of three replicate measurements.
               b Two values are given; the first pertains to all five runs and the second pertains to the three optimum runs (Sediment A = Runs 3,4,5 and
                  Sediment B = Runs 2,4, and 5).
               c All O & G concentrations are in mg/kg dry weight.

-------
                        Table 13: Total Metals in Test Sediments
                                      Sediment A                  Sediment B
Analyte                       Feeda     Treated Solids3    Feed3     Treated Solids3
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Selenium
Silver
Thallium
Vanadium
Zinc
82.2
8
60.5
2.2
13.7
402
172
363,000
1,030
6,260
1.4
130
<1.1
NA
<110
47.6
9,510
35.5
10.5
52.9
<0.62
8.1
302
161
353,000
877
5,640
2.1
70.9
<0.93
NA
<95.2
42.9
8,210
<25.2
32.1
126
1.1
3.2
568
131
192,000
484
1,800
14. lb
119
8.3
9
107
33.7
1,690
<10.9
12
133
1.4
2.7
526
131
162,000
433
1,700
0.58b
106
1.9
<1.1
<34.9
28.5
1,690
Source: USEPA, 1993

a Total metals analysis was conducted on only one run for each sediment type. All results are reported
  in mg/kg- dry weight basis.

b No conclusions should be drawn from this single analysis; vendor data indicated no significant
  reduction in concentration.

NA  Not analyzed.
                                            49

-------
                   Table 14: TCLP Test Results in Test Sediments
Regulated Metals

Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver

Non-Regulated Metals
                                   Sediment A                   Sediment B
                         RCRA                Treated                   Treated
                       Reg. Levelb    Feeda     Solidsa       Feed        Solidsa
                         (mg/L)       (mg/L)    (mg/L)       (mg/L)      (mg/L)
5
100
1
5
5
0.2
1
5
<0.03
0.66
<0.03
<0.04
<0.21
<0.002
<0.03
<0.05
<0.03
0.48
<0.03
<0.04
O.21
<0.002
<0.03
<0.05
<0.03
0.76
0.03
O.04
<0.21
<0.002
<0.03
<0.05
<0.03
1.03
<0.03
<0.04
<0.21
0.005
<0.03
<0.05
                                     <0.48
<0.48
NA
NA
Antimony
Beryllium
Copper
Iron
Manganese
Nickel
Thallium
Vanadium
Zinc

0.01
<0.03
4.43
9.67
<0.11
<1.53
<0.07
0.52

<0.01
0.03
0.03
2.48
<0.11
<1.53
<0.07
0.46

NA
NA
NA
NA
NA
NA
NA
NA

NA
NA
NA
NA
NA
NA
NA
NA
Source: USEPA, 1993

a Results are the average of all five runs for each sediment, with the fourth run value being the average of
  laboratory triplicate analysis.

b Levels presented are RCRA regulatory thresholds for hazardous material.

NA Not analyzed.
                                        50

-------
                 Table 15:  Noncritical Analyses Results - Sediment A and B
Parameter
TRPHb
Volatile Solidsb
Total Cyanide0
Reactive Cyanideb
Reactive Sulfideb
Total Phosphorus0
pHd
Proximate/Ultimate Tests6
Moisture
Carbon
Ash
Sulfur
Oxygen
Hydrogen
Nitrogen
Btu/lb
Sediment A
Feeda Treated Solids3 Feeda
4,730
8.5
17.7
<0.5
152
68
8.6 - 8.9

52.14
11.85
87.31
0.16
0.33
0.25
0.10
1,892
<20
7.1
12.2
<0.4
<39
7.4
10.8 - 12.4

7.30
11.40
79.49
0.17
8.26
0.58
0.09
1,318
15,900
31.8
43.2
<0.8
<77
44.6
8.3-8.5

56.09
39.45
50.13
0.88
6.18
2.26
1.11
6,675
Sediment B
Treated Solids1
<20
17.1
96.1
<0.3
53
165
10.5 - 10.8

9.24
23.53
65.40
0.79
8.09
0.93
1.26
3,274
Source: USEPA, 1993
a All concentrations are mg/kg dry weight.
b Results are the average of all five runs.
c Analysis was conducted for one run only.  ..
d Results are the range of all five runs.
e Proximate/Ultimate test results are given as percentages on a dry weight basis.
                                             51

-------
                    Table 16: PAH and PCB Concentrations of Product Water
        Parameter
                           Test Run
            Rl     R2     R3      R4a
                        R5
Sediment A
  Total PAHs (ng/1)
  Total PCBs (ngfl)

 Sediment B
  Total PAHs (ng/1)
  Total PCBs (ng/1)
                                      <3
                    <3
        <3
<3
<3
Source:  USEPA, 1993

a Concentrations reported for Run 4 are the average of three replicate measurements


                            Table 17:  Total Metals in Product Water
                 Reg. Levela (mg/1)       Sed. A Product Water1*
                                         Sed. B Product Water0
Regulated Metals
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
  5
100
  1
  5
  5
  0.2
  1
  5
O.03
<0.06
<0.03
<0.04
<0.21
O.002
<0.03
 NA
Source: USEPA, 1993

a Levels presented are RCRA regulatory thresholds for hazardous materials.
0 Results are reported for one optimal run in mg/1.
NA  Not analyzed.
      <0.03
      0.012
      <0.006
       0.02
      <0.042
       0.001
      <0.03
      O.01
Non-Regulated Metals
Antimony
Beryllium
Copper
Iron
Manganese
Nickel
Thallium
Vanadium
Zinc

<0.48
<0.01
<0.04
0.63
< 0.010
<0.11
<1.5
<0.07
0.13

<0.096
<0.002
0.042
0.56
0.01
<0022
<0.31
0.014
0.07
                                            52

-------
       Table 18: Supplemental Analyses Results - Sediments A and B Product Water
 Parameter
Sediment Aa
Sediment Ba
TRPRfc
Total Volatile Solids0
Total Cyanide (jig/1)0
Total Phosphorus0
pH1*
Total Suspended Solidsb
Total Dissolved Solidsb
Total Organic Carbonb
Total Inorganic Carbonb
Biochemical Oxygen Demand0
Conductivity (|imhos/cm)c
<0.44
116
<5
0.33
12.2
11
1380
16.9
17.1
8.3
2990
17.5d
7350e
<5
0.39
11.9
6
1670
49.7
13.2
9.6
1520
 Source: USEPA, 1993

 a  All measurements are reported in mg/1 except where indicated otherwise.

 b  Results are the average of all five runs.

 c  Analysis was conducted only for one run.

 d  The value given is misleading; all runs had values of < 0.44 mg/1 except for Run 5 where a value of
   87.5 mg/1 was reported for TRPH.

 e  This value is suspected to be incorrect because total volatile solids cannot exceed the sum of total
   suspended solids and total dissolved solids.
              Table 19: PAH and PCB Concentrations of Sediment B Product Oil
Parameter
1
2
Aliquotsa
3
4
5
Avg.
Std. Dev.
Total PAHs (mg/kg) 498,000  438,000  299,000   297,000  436,000  394,000    90,600
Total PCBs (mg/kg)    2,030     1,750    2,520     2,150    2,180    2,130      278

Source: USEPA, 1993

a The aliquots 1 through 5 were collected incrementally as the product oil in the Solvent Evaporator
  tank was drained by a hose into a drum.
                                            53

-------
Table 20: Triethylamine Concentrations in Treated Solids, Product Water, and Oil Phases
Parameter
Sediment A
Triethylamine in Treated
Solids (mg/kg)
Triethylamine in Product
Water (mg/L)
Triethylamine in Oil
Phase (%)
Sediment B
Triethylamine in Treated
Solids (mg/kg)
Triethylamine in Product
Water (mg/L)
Triethylamine in Product
Oil (mg/kg)
Objective Rl


< 150 61.7

<80 <1

<1000 —


< 150 106

<80 <1

< 1000 —
Test Runa
R2 R3 R4b R5 Avg.c


93.1 27.8 28.0 79.6 58/45

<1 <1 <1 2.2 <2/<2

	 	 . 	 	 <« sd
	 ™~™ '""•"" \J*J . U


88.7 55.0 130 89.3 94/103

1.0 <1 <1 <1 
-------
 MASS BALANCES

        Mass balances were completed for individual physical matrices such as oil, solids,
 PCBs, PAHs, and O&G entering and exiting the system during operation.  These balances
 were obtained by comparing the weights and volumes of feed and process additives into
 the system with the various product fractions and samples recovered during testing.
 Analytical data regarding contaminant concentrations within the various fractions, as well
 as percent solids and water, were used along with measurements taken during the
 demonstration. The balances were performed cumulatively over five runs rather than for
 each run because of material buildup within the system that might distort the individual
 mass balances. Cumulative balances for net mass entering and exiting the system were
 also performed (Tables 21 - 24). A demonstration objective was to obtain mass balance
 closures within 80 to 130 percent. Table 25 summarizes the mass balance results.

 Total Material Balance

       Mass balances for all material entering and exiting the process were calculated.
 The closures for Sediment A and B were 99.3 and 99.6 percent, respectively (Table 21).
 These tight closures indicate that, even though individual balances varied because of the
 large number of analyses involved, very little material was lost for either Sediment A or B.

 Solids Balance

       The solids balance was computed by comparing the amount of solids entering the
 system to the process solids recovered. Solid balance closures of 88.5 and 107.5 were
 obtained for Sediment A and B, respectively (Table 21). These closures are consistent
 with the demonstration test objectives of between 80 and 130 percent.

PCS Balance

       The closures for the PCB mass balances were approximately 95 and 112 percent
for Sediments A and B, respectively (Table 21) which are consistent with the
demonstration test objectives. The amount of PCBs entering the system was calculated by
multiplying the PCB concentration within the feed by the weight of the feed entering the
system, and was compared to the cumulative amount of PCBs deposited in the products.
                                       55

-------
                                  Table 21:  Mass Balances (kg) - Sediment A Inputs
Input Sources
Feed
Runl
Run 2
Run3
Run 4
Run 5
Steam
Drying Stages b
Stripping Stages b
Water
Test Startup b
Decanting
Caustic
Extractions
Drying Stages
Stripping Stages
Triethylamine
Totals
Water

31.20
33.44
29.55
30.81
29.99

17.18
37.68
27.77
12.00

1.23
1.36
0.27
—
252.50
Solids

38.65
46.84
43.55
41.32
42.37

—
WH_

1.23
1.36
0.27
—
215.60
0&Ga Triethylamine Total PAHsc

0.26 — 0.0174
0.29 — 0.0239
0.41 --- 0.0201
0.37 --• 0.0237
0.25 — 0.0264

—
«• »• vw_

—
—
341.36
1.58 341.36 0.1115
Total PCBsc

0.000285
0.000302
0.000352
0.000492
0.000699

««™»
r

—
—
—
0.002131
Total Mass

70.11
80.57
73.52
72.50
72.61

17.18
37.68
27.77
12.00

2.45
2.73
0.55
341.36
811.05
Source: USEPA, 1993

a Values are derived from the BEST test method for determination of O & G.
b Values were obtained from RCC.
c These masses are presumed to be included in the mass of O & G.

-------
                                  Table 22: Mass Balances (kg) - Sediment A Outputs
Input Sources
Treated Solids
Runl
Run 2
Run 3
Run 4
Run 5
Product Water
Runl
Run 2
Run3
Run 4
Run5
Product Oil
(65.8% solvent)
Solvent drained
Solvent left in unit b
Decon. Residual
Vent Filter
RCC Samples b
Solvent Decanter b
Filtered)
Totals
Water

13.28
8.90
1.30
5.83
0.65

51.81
42.35
62.34
53.48
57.68

0.04
8.86
—
—

0.01
8.00
—
314.53
Solids

22.50
43.13
30.27
40.40
32.28

0.13
0.05
0.06
0.05
0.07

0.20
...
—
20.73
___
0.19
—
0.73
190.79
0&Ga

0.010
0.012
0.026
0.025
0.020

0.0003
0.0001
0.0002
0,0002
0.0002

3.38
—
—
—
--_
0.03
—
—
3.5
Triethylamine

0.002
0.005
0.001
0.001
0.004

—
—
—
—
—

6.96
256.82
20.00

3.86
1.37
6.59
0.73
296.35
Total PAHsc

0.0006
0.0012
0.0006
0.0011
0.0005

<9.1E-06
<9.1E-06
<9.1E-06
<9.1E-06
<9.1E-06

0.1244
~.
___

•.•_
___
__-
—
0.1286
Total PCBsc

<1.8E-06
8.6E-06
<1.5E-06
1.6E-06
1.3E-06

<1.4E-07
<1.4E-07
<1.4E-07
<1.4E-07
<1.4E-07

0.00200
....
.»_
___
«_
_•_
___
—
0.00202
Total Mass

35.80
52.05
31.59
46.25
32.95
i

51.95
42.40
62.40
53.53
57.75

10.58
265.68
20.23
20.73
3.86
1.60
14.59
1.45
805.16
Source:  USEPA, 1993
a Values are derived from the BEST test method for determination of O & G.
b Values were obtained from RCC.
c These masses are presumed to be included in the mass of O & G.
d The triethylamine and solids masses presented for the filter are estimated to each represent half of the total measured mass of filtered material.

-------
                                                 Table 23:  Mass Balances (kg) - Sediment B Inputs
Ul
m
Input Sources
Feed
Run 1
Run 2
Run 3
Run 4
Run 5
Steam
Drying Stages b
Stripping Stages b
Water
Test Startupb
Decanting
Oil Polishing
Caustic
Extractions
Drying Stages
Stripping Stages
Triethylamine
Solvent added
Solvent left in unit b
Totals
Water

61.13
50.97
42.05
51.35
51.49
21.05
34.18
12.00
20.18

2.11
2.03
0.24
—
348.59
Solids

25.35
25.97
25.59
25.76
22.36
—
—

2.11
2.03
0.24
:
129.40
O&Ga Triethylamine Total PAHsc Total PCBsc Total Mass

7.28 — 1.83 0.01188 93.75
4.87 — 1.98 0.00975 81.82
6.00 — 2.88 0.01564 73.64
5.85 — 2.00 0.01460 82.95
6.60 — 2.47 0.01440 80.45
21.05
34.18
0.00
12.00 ,
20.18

— 4.23
4 05
0.47
385.00 »- — 385.00
20.00 -- — 20.00
30.60 405.00 11.16 0.07 913.60
               Source: USEPA, 1993
               a Values are derived from the BEST test method for determination of O & G.
               b Values were obtained from RCC.
               c These masses are presumed to be included in the mass of O & G.

-------
                                           Table 24: Mass Balances (kg) - Sediment B Outputs
\o
Input Sources
Treated Solids
Run 1
Run 2
Run 3
Run 4
Run5
Product Water
Run 1
Run 2
Run 3
Run 4
Run5
Product Oil
Triethylamine
Solvent drained
Solvent in evaporator
Solvent left in unitb
Decon. Residual
Vent Filter
RCC Samples b
Solvent Decanter**
Filter b
Totals
Water

2.63
3.34
1.94
3.74
3.00

66.50
68.06
46.18
72.66
70.20
0.28

58.18
— -
—
—
0.05
7.00
—
403.76
Solids

18.75
20.94
25.80
26.59
27.37

0.034
0.066
0.127
0.262
0.037
0.53

—
—
15.32
—
0.13
3.18
139.94
O&Ga

0.21
0.27
0.33
0.35
0.31

0.0018
0.0001
0.0002
0.0002
0.0002
33.15

—
1.64
—
—
0.10
—
36.35
Triethylamine

0.002
0.002
0.001
0.004
0.003

—
—
—
—
—
0.03

268.64
23.36
20.00
—
6.36
2.45
6.59
3.18
330.62
Total PAHs c

0.01
0.01
0.01
0.01
0.02

<9.1E-07
<9.1E-07
<4.6E-07
<9.1E-07
<9.1E-07
13.38

—
0.66
—
...
—
—
14.10
Total PCBsc

2.9E-05
4.5E-05
3. IE-OS
4.9E-05
<3.9E-05

<9.1E-07
<4.6E-08
<4.6E-08
<9.1E-08
<9.1E-08
0.07

—
0.00
—
...
—
—
0.07
Total Mass

21.59
24.77
28.07
30.68
30.68

66.54
68.13
46.31
72.93
70.24
33.98

326.82
25.00
20.00
15.32
6.36
2.73
13.59
6.36
909.87
           Source: USEPA, 1993
           a Values are derived from the BEST test method for determination of O & G.
           b Values were obtained from RCC.
           c These masses are presumed to be included in the mass of O & G.

-------
                  Table 25:  Mass Balance Results, Percent Recovery
Test Sediment
Sediment A
Input Total (Ibs)
Output Total (Ibs)
Recovery5 (%)
Sediment B
Input Total (Ibs)
Output Total (Ibs)
Recovery1' (%)
Water

252.50
314.53
124.6

348.59
403.76
115.8
Solids

215.61
190.79
88.5

129.40
139.94
107.5
O&G3

1.58
3.50
222

30.60
36.35
118.8
Triethyl-
amine

341.36
296.35
86.8

405.00
330.62
81.6
Total
PAHs

0.1115
0.1286
115.4

11.16
14.10
126.3
Total
PCBs

0.002131
0.002018
94.7

0.06626
0.07418
112.0
Total
Mass

811.05
805.16
99.3

913.60
909.87
99.6
Source: USEPA, 1993
a Values are derived from BEST test method for O&G
b Percent recoveries = output total/input total x 100

PAH Balance

       The closures for the PAH balances were approximately 115 and 126 percent for
Sediment A and B, respectively (Table 25). The majority of the PAHs, like the PCBs,
entered the system in the raw feed and exited in the product oil.

Oil and Grease Balance

       Closures of approximately 222 and 118.8 percent were obtained for Sediment A
and B, respectively (Table 25). The mass of O&G entering the system was calculated by
multiplying the concentration of O&G in the feed by the weight of the feed entering the
system, while values for the exiting O&G were calculated using O&G concentrations in
the oil product.

       The closure for Sediment B was within the demonstration objectives.  However,
the closure for Sediment A was not within the test objectives.  This is partially attributed
to the low oil content of the feed sediment which resulted in inaccurate analytical values.
Also, calculating the mass of the O&G in the product oil/solvent mixture is difficult due to
the high percentage of solvent remaining hi the mixture.  An additional contributing factor
                                        60

-------
 is that the input O&G concentration was calculated from a methylene chloride extraction
 method, while the pilot unit used triethylamine as the extraction solvent.  It is possible that
 some triethylamine was extracted out of the treated solids by the methylene chloride and
 was thus inadvertently included in the mass value for the O&G.

 Water Balance

       The water balance closures were approximately 125 and 116 percent for Sediment
 A and B, respectively (Table 25), which were within the demonstration objectives.  The
 water balances were performed by comparing the amount of process water entering the
 system to the mass of product water exiting the system.  A majority of the process water
 enters the system as part of the feed. A smaller portion enters the system in the
 Extractor/Dryer vessel and during steam stripping of the product water.

 Solvent Balance

       The triethylamine mass balance closures were 87 and 82 percent for Sediment A
 and B, respectively (Table 25). The used triethylamine is reused within the system,
 although small amounts remain in the treated solids, water product, and oil product. The
 lower recovery for Sediment B was due to the difficulty of allocating mass between water
 and solvent at the end of a series of runs for a given sediment. Various portions of the
 process yield liquid containing separate solvent and water phases.
DISCUSSION AND INTERPRETATION

Introduction

       This section discusses the performance of the RCC BEST pilot plant during the
SITE demonstration testing performed in Gary, Indiana based on interpretations of the
data results. A summary of the Quality Assurance/Quality Control procedures and the
validity of the data are discussed below.

       A QAPP was prepared to obtain data of known quality to be used in evaluating the
BEST process for the sediment samples.  The QAPP specified the guidelines to be used to
                                       61

-------
ensure that each measurement system was in control, and identified that parameters that
were critical to assessing the effectiveness of the technology.  A field audit and one
laboratory audit were performed for each of the laboratories to verify compliance with the
QAPP.  Although there were several minor modifications and deviations from the QAPP,
all of the data was deemed usable (USEPA, 1993).

Assessing Data Quality

       The indicators used to assess the quality of the data generated for the
demonstration are accuracy, precision completeness, representativeness, and
comparability.

       Accuracy is the degree of agreement of a measured value with the true or expected
value, and was expressed as a percent recovery using matrix spikes and/or laboratory
control samples.  Matrix spikes are aliquots of sample spiked with a known concentration
of target analyte(s) used to document the accuracy of a method in the sample matrix.  A
laboratory control sample is a blank matrix spiked with representative target analytes used
to document laboratory performance. In addition, for the organic analyses, surrogates
were added to all  samples and blanks to monitor extraction efficiencies. Surrogates are
compounds similar to target analytes in chemical composition and behavior.

       Precision is the agreement among a set of replicate measurements without
assumption of knowledge of the true value.  The precision of the demonstration project
data was determined using matrix duplicate and matrix triplicate analyses or matrix
spike/matrix spike duplicate/matrix spike triplicate analyses.

       Completeness is a measure of the amount of valid data produced compared to  the
total amount of data planned for the project.  For this demonstration, no samples were lost
due to field or analytical problems. Though all guidelines for QA objectives were not  met,
all data generated were deemed reasonable.

       Representativeness refers to the degree with which analytical results  accurately and
precisely represent actual conditions present at locations chosen for sample collection.
Sediment samples were collected prior to this demonstration and were typical of the area
of concern. These sediment samples were thoroughly homogenized prior to testing.
                                        62

-------
Samples were shipped under chain-of-custody to the laboratories.  Therefore, the data are
representative of material actually treated.

       Comparability expresses the extent to which one data set can be compared to
another.  Since standard USEPA, American Society for Testing and Materials (ASTM) or
National Institute for Occupational Safety and Health (NIOSH) procedures were used in
nearly all cases, results should be comparable to data generated for other similar projects.
Two exceptions include the non-standard procedures used for triethylamine in soil and
water matrices and the BEST method for O&G. These analyses were performed to verify
RCC's results.

       USEPA performed one field audit and one laboratory audit, for each of the two
laboratories performing critical analysis for this demonstration. During the field audit of
July 7,  1992, the auditor noted one concern and some minor issues regarding onsite
sample custody and storage. In each case, corrective action was implemented
immediately. The laboratory audits for SAIC and S-Cubed Laboratory were conducted on
August 5, 1992 and August 6, 1992, respectively.  Several conditions regarding sample
receipt, sample log-in, sample storage and analyses were noted and corrected. Data
quality and project conclusions did not appear to be affected by the noted conditions.

       Upon review of all sample data and associated QC results,  the data generated for
the BEST demonstration was determined to be of acceptable quality. In general, excellent
QC results were obtained for accuracy and precision which can be used to support
removal efficiency results.  Field replicate results and replicate process run results also
serve to document the quality of the data. Overall precision was quite good, indicating
thorough homogenization of the treated sediment and good sampling techniques.

       Evaluation of the demonstration test data indicates that the RCC BEST pilot plant
was very effective in removing organic contaminants from test  sediments collected from
the OCR.  All vendor claims made prior to the demonstration regarding percent removal
of organics, solvent residual in process products, and system mass balance were met.
Specific conclusions are discussed in the following subsections.
                                       63

-------
Removal of Organics

       Removal of organic compounds from GCR sediments ranged from 96 percent (for
PAHs in Sediment A feed) to greater than 99 percent (for PCBs in both sediment feeds).
Table 26 summarizes the conclusions reached in the pilot-scale demonstration. These
organic removal efficiencies are based on analytical results generated by the primary SITE
project laboratory, Maxwell/S-Cubed Division.  Complying with their agreement with the
COE, Chicago District, RCC conducted independent analyses of split samples collected
during the demonstration. Table 27 provides a comparison of SITE'S organic analysis
results to those of RCC. The similarity in data results, generated independently by two
independent laboratories, serves to further substantiate the demonstration results.

             Table 26: Summary of Conclusions for Contaminant Removal
                Contaminant             Results
                 PAHs           96% removal in Sediment A
                                 >99% removal in Sediment B
                 PCBs           >99% removal in both sediments
                 O&G           >98% removal in both sediments
Removal of PAHs-

       The three optimal runs for Sediment A had an average PAH removal efficiency of
96 percent. Using the same optimum run average, the PAH removal efficiency for
Sediment B was 99.3 percent. The increased removal percentage for Sediment B is
attributed to the higher PAH concentration in Sediment B feed (71,000 ppm of total PAHs
versus 548 ppm for Sediment A feed). Field optimization increased removal efficiency
only by about two percent for Sediment A, and did not have much effect on Sediment B.
Sediment B was treated in an apparently optimized system.  These observations confirm
the importance of bench-scale tests, since they had a greater impact on efficiency than did
the field optimization.
                                       64

-------
          Table 27: Comparison of Organic Analyses between SITE and RCC
Analyte
Total PAHs
SITE
RCC
Total PCBs
SITE
RCC
O&G
SITE
RCC
Sediment A
Feeda Treated
Solids*

548
782

12.1
5.5

7,400
18,500

22
34

0.04
0.07

570
710
Percent
Removalb

96.0
95.7

99.7
98.7

92.3
96.2
Feeda

70,920
87,500

425
580

197,000
219,000
Sediment B
Treated
Solids3

510
716

1.8
1.1

12,000
15,400
Percent
Removalb

99.3
99.2

99.6
99.8

93.9
93.0
Source: USEPA, 1993
a Concentrations reported in mg/kg (dry weight basis) and are the average of the three optimum runs for
  each sediment. (Sediment A = Runs 3, 4 and 5; Sediment B = Runs 2, 4 and 5).

b Percent Removals = Feed concentration -Treated Solids Concentration  x 100
                             Feed concentration
       Of the fourteen specific PAH compounds measured above detection limits in
Sediment A feed, eight were removed at efficiencies greater than 98 percent. Only two
PAH compounds were removed by less than 96 percent (anthracene at 94 percent and
2-Methylnaphthalene at 85 percent).  Of the 17 specific PAH compounds detected in
Sediment B Feed solids, the lowest calculated removal efficiency was 97 percent (for
acenaphthylene). Thirteen of the PAH compounds were removed by efficiencies of
greater than 99 percent in the Sediment B sample.

Removal of PCBs-

       The efficiency for removing total PCBs from both sediment feed solids exceeded
99 percent when averaging either the three optimum runs or all five runs, although the
average removals for the optimized runs were higher.  Treatment of Sediments A and B
                                       65

-------
under optimized conditions resulted in average PCB removals of 99.7 and greater than
99.6 percent, respectively.  This easily exceeded the objective of achieving organic
removals in the range of 96 to 99 percent.

      On a per run basis, treatment of Sediment A resulted in total PCB removals
exceeding 99 percent for all test runs except for Run 2, where the removal was calculated
at 96.9 percent.  Under optimized conditions, total PCB removals ranged from 99.3 to
99.8 percent.  On a per run basis, treatment of Sediment B resulted in total PCB removals
exceeding 99 percent for all test runs and under optimized conditions the total PCB
removals ranged from 99.3 to 99.8 percent.

       The PCB data from the two demonstration tests may have been the most important
information acquired from the demonstration as a whole. This is because the total PCBs
were at such contrasting levels in the two sediment types.  Sediment A feed contained, on
the average, PCBs at 12 mg/kg, while Sediment B feed contained an average PCB
concentration more than an order of magnitude higher at 425 mg/kg.  By achieving high
removal efficiencies for both sediment types, the BEST Process was found effective in
removing PCBs to very  low levels (i.e., less than the commonly required regulatory
cleanup level of 2 mg/kg) regardless of the initial concentrations in untreated material.

Removal of Oil and Grease-

       The O&G was removed from  both sediment samples with an efficiency of over 98
percent for both the  three optimum runs and all five runs.  The optimal runs resulted in
average removals of 98.4 and 98.9 percent for sediments A and B, respectively.

Triethylamine Residual in Products

       Residual solvent concentrations in both product solids were less than 150 ppm.
The optimum run average for residual solvent in sediment A and B treated solids was 45
ppm and 103 ppm respectively. The pilot plant was therefore effective at removing
triethylamine from the treated solids after extraction is complete.
                                       66

-------
                  Table 28: Residual Triethylamine Concentrations
                                        Sediment A         Sediment B
       Treated Solids (optimum runs)      45.1 mg/kg           103 mg/kg
       Water                           1.0 mg/1 or less         2.2 mg/1 or less
       Oil product                           -                733 mg/kg
       The residual solvent in product water was less than 2 ppm for the sediment A
water phase and less than 1 ppm for the sediment B water phase.  The objective for the
demonstration was to achieve levels of less than 80 ppm. Most of the triethylamine was
stripped from the product water by adding caustic to adjust the pH.  However, the
addition of caustic to enhance removal of triethylamine is a sensitive operation that may
result in the production of RCRA-regulated corrosive water (pH >12.5). During this
project, the product water for the first batch of Sediment A had a pH of 12.4, while
product water for batch 5 had a pH of 12.2. The addition of caustic must be carefully
monitored to avoid creating a RCRA-regulated product water.

TCLP Leachability Solids

       The teachability of regulated heavy metals from the treated solids from both
sediment samples could not be evaluated since neither of the sediment feeds leached
RCRA-regulated metals during TCLP tests. In general, heavy metals did not leach from
the feed or treated solids. Most metals were at or below the limits of detection in the
TCLP leachate, despite high concentrations in the solids. However, iron and manganese
were less teachable in the treated solids from Sediment A (which had the much higher
metals content), than in the raw feed.

Biodegradation Tests

       An EPA report (EPA-600/2/82-OOla) showed that triethylamine was completely
degraded in water from 200 ppm in 11 hours using aerobacter, a common soil bacteria.
Simple biodegradation testing was performed for this demonstration test to determine if
                                       67

-------
triethylamine would biodegrade in the treated solids. Viable potting soil was mixed with
an equal portion of treated solids collected from three of the demonstration test runs.  For
each of the three sediment samples, two sets of twelve tests vials were prepared; the first
set contained unaltered mixtures and the second "control" set contained the same mixture
spiked with mercuric chloride to prevent native bacteria from degrading the solvent.  The
samples were stored in the dark at room temperature. A pair of samples (one test sample
and one control sample) was analyzed at Day 0, and at the end of 2 weeks, 4 weeks and 8
weeks. No pH adjustments or nutrient additions were performed.

       No evidence was generated indicating that triethylamine present at 25 to  100 ppm
was biodegraded in soil within 2 months of application.  The most likely reasons for lack of
biodegrading activity may be amine binding to the humic fraction and not optimizing pH and
nutrients in the test. The triethylamine concentrations remained relatively level over the
testing period for the test and the control soils, so the solvent apparently did not volatilize.
This study does not, however, preclude the possibility that triethylamine could be
biodegraded after pH adjustment and further nutrient or organism enrichment. Further
details of the biodegradation study are contained in Volume H of the TER  (USEPA, 1993).

ESTIMATED COSTS

Introduction

       USEPA presented an economic analysis of treatment costs for a commercial
treatment  system utilizing the BEST system in the AAR (USEPA, 1993). Rather than
including site-specific costs in the unit treatment costs, the final AAR presented certain
site-specific costs as separate line-item costs to ensure comparability to unit treatment
costs for other innovative technologies. In contrast, the cost estimate in this report
includes several of the site-specific costs in the per cubic yard treatment cost, as developed
for the draft AAR.  The cost analysis was based on the results of this demonstration and
on information from previous tests, including a full-scale test. In this  section on cost
estimates, English units are used for processing equipment and throughput rates, since
these units are commonly used in engineering applications in the United States. The pilot
unit utilized during this demonstration operated at an average feed rate of approximately
90 pounds of contaminated sediment per day. It is projected that a commercial unit would
                                        68

-------
be capable of treating up to 184 cubic yards of contaminated sediment per day. This
report presents estimated costs for treatment of 5,000 cubic yards (cy), 25,000 cy, 50,000
cy and 100,000 cy of contaminated sediment using cost data developed for the AAR.

Assumptions

       Increasing either the feed rate or the online percentage can reduce the unit
treatment cost. An online factor adjusts the unit treatment cost to compensate for the fact
that the system is not online constantly because of maintenance requirements, breakdowns,
and unforeseen delays.  Online percentages of 60 percent, 70 percent and 80 percent are
compared. RCC estimated that approximately 13 percent of the total operating time
available in one year is estimated to be downtime, so use of the 80 percent online factor
appears justified.  Costs incurred while the system is not operating are thus incorporated
into the unit cost.

Basis of Economic Analysis

       The costs associated with the operation of the BEST system, as presented in this
economic analysis, are defined by twelve cost categories that reflect typical activities
encountered on remedial sites. Each category is defined and discussed, forming the basis
for the cost analysis in Tables 29 - 32.  The percentage of the total cost contributed by
each of the cost categories is shown in Tables 30 and 32.

Excluded Costs--

       Many assumptions regarding operating conditions and task responsibilities affect
the cost estimate. These estimates do not contain the following items: the vendor's profit,
costs for preliminary  site preparation, obtaining permits, determining regulatory
requirements, monitoring programs, and post-treatment site restoration.  Also  excluded
are the dredging and  storage of sediment, and final disposal of residuals.  These costs tend
to be site-specific and should be determined on a case-by-case basis. The excluded costs
may significantly increase the total project cost.
                                        69

-------
Site Preparation--

       Site preparation responsibilities can include site design and layout, surveys and site
logistics, legal searches, access rights and roads, preparations for support and
decontamination facilities, and auxiliary buildings. These costs were not included in the
site preparation costs in this estimate since they are site-specific. However, certain site
preparation activities are common for all sites. RCC estimates a site preparation cost of
$100,000 for foundations, electrical power and water.

       Removal of contaminated sediment is also a part of site preparation. This cost
estimate is based on excavation of contaminated soil. The cost of using the
BEST system to treat sediment can be estimated by replacing excavation costs with
dredging costs. Dredging costs for harbor maintenance dredging typically range from $3
to $7 per cy depending on the dredge type and size, and type of sediment.  The adjustment
for environmental dredging production rates may be as much as two to three-fold (or
more) for specific applications.

Utilities-

       Potable water and electrical power are necessary for the operation of the
BEST system. RCC estimates that the full-scale unit will require 40 to 70 kilowatt-hours
per ton of feed. This cost estimate assumed that the average electricity consumption rate
is 55 kilowatt-hours per ton of feed. The cost estimate includes costs associated with
connecting the BEST system to the onsite water and electrical supplies, but not for
providing the utilities to the site.

Operating Times and Labor-

       It was assumed that the unit would be operated 24 hours per day, 7 days per week.
Treatment operations are assumed to require  18 onsite personnel: 12 system operators at
$30 per hour each (4 per shift), 3 operations supervisors at $40 per hour each (1 per shift),
and 3 safety personnel at $40 per hour each (1 per shift). Three administrative and clerical
personnel at $20 per hour each will also be required.  Labor costs consist of wages and
                                        70

-------
living expenses. Labor rates include benefits and overhead costs. Living expenses depend
on several factors:  the duration of the project, the number of local workers hired, and the
geographical location of the project.

Capital, Equipment and Fixed Costs-

       It was assumed that RCC owned and operated the full-scale system, and that RCC
distributed the capital costs across the 10-year useful life of the system. Capital costs are
estimated for all major equipment included in the BEST system. Straight-line depreciation
over a ten-year period yields an annualized equipment cost, which was then prorated for
the period of time that the equipment remained on site.  The portions of this cost that is
accrued during treatment is normalized relative to the amount of sediment treated and
incorporated into the per cubic yard treatment cost. August, 1992 dollars were used for
most of the costs in the estimate.

Supplies--

        Supplies include chemicals and spare parts. A net of two pounds or less of
triethylamine is required per cubic yard of sediment processed. A much larger amount of
triethylamine is recycled within the system and may be reused from project to project.
Triethylamine is purchased hi 330 pound drums for about $1.40 per pound. During this
demonstration project, about 3 gallons of 50 percent sodium hydroxide were consumed
per ton of sediment treated. A sodium hydroxide cost of $2.33 per gallon is used for this
analysis, although sodium hydroxide usage will be site specific. Nitrogen gas used hi the
BEST system ranges from about $1.00 to $1.50 per cubic yard of sediment processed.
The more conservative figure was used in this estimate.  The annual costs for spare parts
is assumed to be 5 percent of the total purchased equipment cost.

Consumables--

       Consumables include compressed air, electricity and water. Costs for compressed
air are included in the electricity and equipment costs since compressed instrument air will
be produced by an air compressor/dryer system transported to the site with the BEST
system.  This analysis used an average electricity consumption rate of 55 kilowatts per ton
                                       71

-------
of feed and an average price of electricity of $0.066 per kilowatt-hour. Operation of a
full-scale BEST system is assumed to use about 1,020 gpd of water at a cost of $0.0011
per gallon.

Facility Modification, Repair, and Replacement Costs—

       For estimating purposes, total annual maintenance costs (labor and materials) are
assumed to be ten percent of annual equipment costs. Maintenance labor typically
accounts for two-thirds of the total maintenance costs and has previously been accounted
for in the labor estimate. Maintenance material costs are estimated at one third of the total
maintenance cost and are prorated over the entire period of treatment.  Costs for design
adjustments, facility modifications, and equipment replacements are included in the
maintenance costs.

Site Demobilization Costs--

       Demobilization costs only include costs associated with the disassembly and
decontamination of the BEST system and auxiliary equipment. Transportation costs are
accounted for under mobilization activities. Disassembly consists of taking the system
apart and loading it and all auxiliary equipment onto eleven trailers for transportation. It
requires the use of an operated 50-ton crane, available at $6,360 per week, for 4 weeks.
Additionally,  disassembly requires an eight-person crew working 10 hours per day, 7 days
per week, for 30 days. Labor costs consist of wages ($40 per hour for the supervisor and
$30 per hour for each of the other workers) and living expenses.

       Site cleanup and restoration are limited to the removal of all equipment from the
site; this is included in the cost of disassembly and decontamination. Requirements
regarding the filling, grading, or recompaction of the soil will vary depending on the
future use of the site and are assumed to be the obligation of the responsible party (or site
owner).
                                        72

-------
                   Table 29:  Treatment Costs for 184-cypd BEST System Treating
                        25,000 Cubic Yards of Contaminated Soil, Sediment, or Sludge
 Item
       Cost  ($/cubicvard)
  60%             70%
online            online
     80%
     online
 Site Preparation                 39.69
 Permitting and Regulatory Costs
 Equipment Cost Incurred         14.52
 Startup and Fixed Costs          40.12
 Labor                          48.62
 Supplies                        16.85
 Consumables                    28.77
 Effluent Treatment and Disposal
 Residuals Shipping,
  Handling, and Transport
 Analytical Costs
 Facility Modification,             0.49
 Repair, and Replacement
 Site Demobilization               4.75
 Total Operating Costs          193.81
                   37.29

                   12.97
                   39.74
                   41.68
                   16.27
                   28.77
                   0.43

                   4.75
                 181.90
      36.27

      11.80
      39.69
      36.47
      15.88
      28.77
      0.39

      4.75
      174.02
                     Table 30: Treatment Costs as Percentages of Total Costs for
                           184-cypd BEST System Treating 25,000 Cubic Yards of
                           Contaminated Soil, Sediment, or Sludge
Item
      Cost (as % of total cost)

  60%         70%
online        online
  80%
online
Site Preparation                 20.5
Permitting and Regulatory
 Costs
Equipment Cost Incurred          7.5
Startup and Fixed Costs           20.7
Labor                           25.1
Supplies                         8.7
Consumables                    14.8
Effluent Treatment and
 Disposal
Residuals Shipping,
 Handling, and Transport
Analytical Costs
Facility Modification,              0.3
 Repair, and Replacement
Site Demobilization               2.4
               20.5
                7.1
               21.8
               22.9
                8.9
               15.8
                0.2

                2.6
 20.8
  6.8
 22.8
 21.0
  9.1
 16.5
  0.2

  2.7
                                              73

-------
                  Table 31: Treatment Costs for 184-cypd BEST System Operating
                             with an 80-Percent Online Factor
Cost ($/cv)
5,000 25,000 50,000
Item cy cy cy
Site Preparation
Permitting and
Regulatory Costs
Equipment Cost Incurred
Startup and Fixed
Costs
Labor
Supplies
Consumables
Effluent Treatment
and Disposal
Residuals Shipping,
Handling, and Transport
Analytical Costs
Facility Modification,
Repair, and Replacement
Site Demobilization
Total Operating Costs
52.27
-

26.26
171.65

36.47
20.70
28.77

-

-
-
0.88

23.75
360.75
36.27
-

11.80
39.69

36.47
15.88
28.77

-

-
-
0.39

4.75
174.02
34.27
-

9.99
23.2

36.47
15.28
28.77

-

-
-
0.33

2.37
150.68
100,000
cy
33.28
-

9.10
15.04

36.47
14.98
28.77

.

-
-
0.30

1.20
139.14
                Table 32: Treatment Costs as Percentages of Total Costs for 184-cypd
                           BEST System Operating with an 80-Percent Online Factor
                                     Cost (as % of total cost)
Item
5,000
   cy
25,000
    cy
50,000
    cy
100,000
   cy
Site Preparation                  14.5
Permitting and
 Regulatory Costs
Equipment Cost Incurred          7.3
Startup and Fixed Costs           47.6
Labor                           10.1
Supplies                         5.7
Consumables                     8.0
Effluent Treatment and Disposal
Residuals Shipping,
 Handling, and Transport
Analytical Costs
Facility  Modification,             0.2
 Repair, and Replacement
Site Demobilization               6.6
              20.8
               6.8
              22.8
              21.0
               9.1
              16.5
              22.7
               6.6
              15.4
              24.2
              10.1
              19.1
               23.9
                6.5
               10.8
               26.2
               10.8
               20.7
               0.2

               2.7
               0.2

               1.6
                0.2

                0.9
                                              74

-------
Cost Conclusions

       RCC's proposed commercial scale BEST system is designed to remediate sediment
and other media contaminated with organics such as PCBs and PAHs.  The cost to
remediate 25,000 cubic yards of contaminated sediment using the 184 cubic yards per day
BEST solvent extraction system is estimated at $194 per cubic yard if the system is online
60 percent of the time and $174 per cubic yard if the system is online 80 percent of the
time.  Projected unit costs for less than 25,000 cubic yards of contaminated sediment are
slightly higher and projected unit costs for more than 25,000 cubic yards of contaminated
sediment are slightly lower.

       These costs are considered "order-of-magnitude" estimates as defined by the
American Association of Cost Engineers. The actual cost is expected to fall between 70
percent and 150 percent of these estimates.  Since these cost estimates are based on a
preliminary design, the range may be wider.  In addition, since this estimate does not
include values for four of the twelve cost categories (permitting and regulatory costs,
effluent treatment and disposal, residuals shipping, handling and transport, and analytical
costs), the actual treatment costs may be significantly higher than those shown in this
estimate.

CONCLUSIONS AND LESSONS LEARNED

       The BEST process proved effective at removing organic contaminants from OCR
sediment.  Sediment A initially contained high concentrations of metals and low
concentrations of organic compounds relative to Sediment B.  Sediment B contained high
concentrations of organic contaminants such as PAHs, PCBs, and O&G.

       The analytical results for Sediment A showed that the process removed greater
than 99 percent of the PCBs, 96 percent of the PAHs, and 98 percent of the O&G. The
residual solvent in the product solids and product water generated from Sediment A was
45 ppm and less than 2 ppm, respectively. An oil product was not generated for Sediment
A because of a low amount of oil (less than  1 percent) in Sediment A feed.
                                       75

-------
       The BEST process removed greater than 98 percent of the O&G for Sediment B
and greater than 99 percent of the PCBs and PAHs. The residual solvent in the product
solids, product water and product oil was 103 ppm, 1 ppm and 730 ppm, respectively.

       Mass balances for total materials in the BEST system achieved closures of 99.3
percent and 99.6 percent for Sediment A and B respectively. Mass balances comparing
feed and product streams excluding triethylamine achieved closures of 108 percent and
114 percent for Sediment A and Sediment B, respectively.

       The Sediment A oil product was not processed to reduce its triethylamine
concentration since very little oil product was generated.

       The BEST process had no measurable effect on the leachability of most heavy
metals; most metals did not leach appreciably from either the feed or treated solids.  The
exceptions were iron and manganese, as measured in Sediment A.  The leachability of iron
in Sediment A was reduced by two orders of magnitude, and the leachability of manganese
was reduced by a factor of four.

       No evidence was generated indicating that triethylamine present at 25 to 100 ppm
is biodegraded in soil during a two-month test.

Effects of TSCA on Pilot-Scale Demonstration

General—

       This section discusses the impacts of treating a sediment regulated by the Toxic
Substances Control Act of 1976 (TSCA). Other pilot-scale demonstration projects and
remediation projects dealing with sediment containing PCBs greater than 50 ppm may
need to meet similar requirements.

       By collecting TSCA-regulated sediment (containing PCBs greater than 50 ppm)
from the OCR, the COE was considered a generator of a PCB waste. USEPA SITE, who
provided for storage of the sediment and TSCA-regulated residuals during the
demonstration project, obtained an EPA ID number.  The vendor, RCC, obtained TSCA
                                       76

-------
Research and Development permits from USEPA Headquarters for both the bench-scale
treatability tests and the operation of the pilot-scale unit. The permits entail substantial
reporting and analytical requirements largely designed to protect human health and the
environment.

Transportation--

       Shipment of the sediment samples from the treatment site in Gary, Indiana to
RCC's laboratory in Bellevue, Washington was performed by a registered PCB transporter
who had filed a "Notification of PCB Activity" and received an EPA identification
number, in accordance with current PCB regulations. Each container was marked in
accordance with CFR 761.40(a)(l)  (Marking Requirements for PCB Containers).

Storage and Disposal —

       USEPA Region 5 has requirements for long-term storage of a generator's own
TSCA material. These requirements included double metal containment for the PCB
waste, a high density polyethylene liner underneath the containers and a berm surrounding
the containers, and a cover over the containers to prevent precipitation from contacting
the containers.  A full-scale treatment project may need to  meet similar storage
requirements.  Additional regulations apply to the commercial storage of PCB wastes.
Both SITE and RCC disposed of PCB-containing wastes at permitted treatment, storage
and disposal facilities.

Recordkeeping and Reporting—

       As required by 40 CFR Part 7661.180(a) and (b), the participants in this pilot-scale
demonstration developed annual records and annual document logs. Manifests were used
during the generation (sediment collection), storage, transportation and disposal stages of
the project. An annual record contains all signed manifests generated during the calendar
year and all Certificates of Disposal received by the facility during the calendar year. The
annual document log contains information including the name, address, and EPA
identification number of the facility and the unique manifest number of every manifest
generated by the facility during the calendar year.  In addition, the annual document log
contains a unique number identifying each container, a description of the contents of each
                                       77

-------
container including total weight in kilograms of the contents of each container, the date
removed from service for disposal, the date placed in transport for off-site storage or
disposal, and the date of disposal, if known.
Costs—
       The cost of the following items associated with the project increased due to
treatment of TSCA-regulated sediment: the contract with the treatment vendor, shipment
of sediment to the laboratory for bench-scale tests, sampling and analysis, disposal of
residuals, reporting, and coordination with regulatory agencies. Compliance with TSCA
had a large impact on the cost of the demonstration project.

State Requirements—

       The Indiana Department of Environmental Management requires a Research and
Development approval (according to 329IAC 4-l-5(g)) to perform a demonstration
project for an alternative PCB disposal technology. A similar approval may be required
for a full-scale cleanup in the GCR/IHC Area of Concern. The Indiana Department of
Environmental Management would also regulate the cleanup under its solid and hazardous
waste regulations.
                                       78

-------
                                 REFERENCES
Averett, Daniel E., Perry, Brett D., Torrey, Elizabeth J., and Miller, Jan A. 1990. Review
of Removal Containment, and Treatment Technologies for Remediation of Contaminated
Sediment in the Great Lakes.  Miscellaneous Paper EL-90-25. US Army Engineer
Waterway Experiment Station, Vicksburg, Mississippi.

Indiana Department of Environmental Management.  1990.  The Remedial Action Plan for
the Indiana Harbor Canal, the Grand Calumet River and the Nearshore Lake Michigan:
Stage One, Final Draft.

Resources Conservation Company. 1993.  Final Scale Test Report for Pilot-scale
Demonstration of the BEST Solvent Extraction Technology at the Grand Calumet
River/Indiana Harbor and Canal Area of Concern. Bellevue, Washington.

Resources Conservation Company. 1992 Test Plan for BEST Solvent Extraction Process
to Conduct Pilot-scale Testing at the Grand Calumet River/Indiana Harbor and Canal Area
of Concern.  Bellevue, Washington.

Risk Reduction Engineering Laboratory, Office of Research and Development, U.S.
Environmental Protection Agency. 1993. Resources Conservation Company B.E.S.T
Solvent Extraction Technology: Applications Analysis Report. EPA/540/AR-92/079.
Cincinnati, Ohio.

Risk Reduction Engineering Laboratory, Office of Research and Development, U.S.
Environmental Protection Agency. 1993. Technology Evaluation Report. SITE Program
Demonstration. Resources Conservation Company Basic Extractive Sludge Treatment
(BEST) Grand Calumet River. Gary. IN: Volume 1. EPA/540/R-92/079a.  Cincinnati,
Ohio.

Simmers, J.W., Lee, C.R., Brandon, D.L., Tatem, H.E., and Skogerboe, J.G. 1991.
Information Summary. Area of Concern: Grand Calumet River. Indiana. Miscellaneous
Paper EL-90-10. US Army Engineer Waterways Experiment Station, Vicksburg,
Mississippi.

Technology Innovation Office, Office of Solid Waste and Emergency Response.  1992. A
Citizen's Guide to Solvent Extraction. EPA/542/F-92/004. US Environmental Protection
Agency.

U.S. Army Corps of Engineers, Chicago District.  1991. Project Plan Pilot-Scale
Demonstration of a Solvent Extraction Technology at the Grand Calumet River/Indiana
Harbor and Canal Area of Concern. Chicago, Illinois.
                                      79

-------
U.S. Army Engineer Waterways Experiment Station, Environmental Laboratory.  1987.
Disposal Alternatives for PCB-Contaminated Sediments from Indiana Harbor. Indiana:
Vol. I: Main Report. Miscellaneous Paper EL-87-9. Vicksburg, Mississippi.
                                       80

-------