EPA/540/5-90/002
January 1990
VOLUME I
TECHNOLOGY EVALUATION REPORT
CF SYSTEMS ORGANICS EXTRACTION SYSTEM
NEW BEDFORD, MASSACHUSETTS
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
NOTICE
The information in this document has been funded by the U.S. Environmental
Protection Agency under Contract No. 68-03-3485 and the Superfund Innovative
Technology Evaluation (SITE) Program. It has been subjected to the Agency's
peer review and administrative review and it has been approved for publication
as a USEPA document. Mention of trade names or commercial products does not
constitute an endorsement or recommendation for use.
ii
-------
FOREWORD
The Superfund Innovative Technology Evaluation (SITE) program was
authorized in the 1986 Superfund amendments. The program is a joint effort
between EPA's Office of Research and Development and Office of Solid Waste and
Emergency Response. The purpose of the program is to assist the development
of hazardous waste treatment technologies necessary to implement new cleanup
standards that require greater reliance on permanent remedies. This is
accomplished through technology demonstrations that are designed to provide
engineering and cost data on selected technologies.
This project consists of an analysis, of CF Systems' proprietary organics
extraction process. The technology demonstration took place at the New
Bedford Harbor Superfund site, where harbor sediments are contaminated with
polychlorinated biphenyls and other toxics. The demonstration effort was
directed at obtaining information on the performance and cost of the process
for use in assessments at other sites. Documentation will consist of two
reports. This Technology Evaluation Report describes the field activities and
laboratory results. An Applications Analysis will follow and provide an
interpretation of the data and conclusions on the results and potential
applicability of the technology.
Additional copies of this report may be obtained at no charge from EPA's
Center for Environmental Research Information, 26 West Martin Luther King
Drive, Cincinnati, Ohio 45268, using the EPA document number found on the
front cover of the report. Once this supply is exhausted, copies can be
purchased from the National Technical Information Service, Ravensworth Bldq
Springfield, VA 22161, (703) 487-4600. Reference copies will be available at
EPA libraries in their Hazardous Waste Collection. You can also call the SITE
Clearinghouse hotline at 1-800-424-9346 or 382-3000 in Washington DC to
inquire about the availability of other reports.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
-------
ABSTRACT
The SITE Program demonstration of CF Systems' organics extraction
technology was conducted to obtain specific operating and cost information
that could be used in evaluating the potential applicability of the technology
to Superfund sites. The demonstration was conducted concurrently with pilot
dredging studies managed by the U.S. Army Corps of Engineers at the New
Bedford Harbor Superfund site in Massachusetts. Contaminated sediments were
treated by CF Systems' Pit Cleanup Unit (PCU) that used liquefied propane/
butane as the extraction solvent. The PCU was a trailer-mounted system with a
design capacity of 1.5 gpm (20 bbl/day). CF Systems claimed that the PCU
would extract organics from contaminated soils based on solubility of organics
1n liquefied propane/butane.
The objectives included an evaluation of (1) the unit's performance, (2)
system operating conditions, (3) health and safety considerations, and (4)
equipment and system materials handling problems. Extensive sampling and
analyses were performed showing that polychlorinated .biphenyl (PCB) extraction
efficiencies of 90 percent were achieved for sediments containing PCBs ranging
from 350 to 2,575 ppm. In Test 2, sediments containing 350 ppm were reduced
to 40 ppm after 10 passes, or recycles, through the PCU. In Test 3, a 288 ppm
feed was reduced to 82 ppm after 3 passes. In Test 4, a 2,575 ppm feed was
reduced to 200 ppm after 6 passes. Some operating problems occurred, such as
the intermittent retention of solids in system hardware and foaming in the
treated sediment collection tanks. These problems did not affect extraction
efficiency but could affect operation of a full-scale unit. Corrective
measures will be addressed by the developer and EPA. A mass balance
established over the entire demonstration showed excellent accountability for
96 percent of the total mass. Operation of the unit did not present any
threats to the health and safety of operators or the local community.
iv
-------
VOLUME I CONTENTS*
' Page
Foreword . . . . . ..... iii
Abstract ...... iv
Figures . . ....;!.' . . '''' v_
Tables . . .; . . . .... . ._. ..'.'.'. viii
Acknowledgements ix
Abbreviations and Symbols. . . . . .^. . . . . x
1. Introduction i
1.1 Background i
1.2 Program Objectives 2
1.3 Technology Evaluation Criteria 2
1.4 Description of Operations . . . 3
1.5 Project Organization. '.'.'.'.'. 4
2. Summary of Results 6
2.1 System Performance. 6
2.2 Operating Conditions . . 7
2.3 Health and Safety Considerations. . . 8
2.4 Equipment and Material Handling Problems . . 9
2.5 Lessons Learned ..... 10
\
3. Proces^ Design 13
\ \ "' .
3.1 Process Description 13
3.2 Equipment Specifications . . . 13
3.3 Process Flow Diagram ......... IB
4. Demonstration Site Description . . . .... . . .... 22
4.1 Site Characteristics 22
4.2 Predemonstration Samples -.',-'.'.'.'.'.'.'. 24
5. Field Activities 29
5.1 Bench-Scale Tests '....".." . . 29
5.2 Operations Summary 31
5.3 Operations Chronology ......... 34
6. Sampling and Analytical Program - 41
6.1 Sampling Locations 41
6.2 Sampling Schedule ............. 43
6.3 Analytical Methods and Physical Tests . . . . 45
6.4 Process Control and Field Measurement Devices 48
*Volume II contains Sampling and Analytical Reports and Operating Log Data.
-------
VOLUME I CONTENTS* (Continued)
7. Results and Discussion
7.1 System Performance
7.2 Operating Conditions
7.3 Developer's Goals
7.4 Health and Safety Monitoring
7.5 Equipment and Material Handling Problems.
7.6 Data Quality Assurance
8. References
APPENDIX A - MASS INVENTORIES FOR TESTS 2, 3, AND 4
APPENDIX B - MASS BALANCES FOR DECONTAMINATION EFFLUENTS
Page
, 51
51
63
70
71
72
73
81
82
92
*Volume II contains Sampling and Analytical Reports and Operating Log Data.
vi
-------
FIGURES
Number
1.1 SIMPLIFIED FLOW CHART
3.1 PIT CLEANUP UNIT
3.2 CF SYSTEMS PROCESS SCHEMATIC
4.1 DEMONSTRATION LOCATION
4.2 PREDEMONSTRATION SAMPLES
5.1 PLAN SKETCH FOR NEW BEDFORD HARBOR DEMONSTRATION
6.1 SYSTEM FLOW DIAGRAM
7.1 TEST 2 PCB REDUCTION
7.2 TEST 3 PCB REDUCTION
7.3 TEST 4 PCB REDUCTION
7.4 POTENTIAL PIT CLEANUP UNIT PCB REDUCTION
7.5 ILLUSTRATIVE INVENTORY SHEET :
7.6 SOLIDS THROUGHPUT PER PASS, TEST 2
7.7 SOLIDS THROUGHPUT PER PASS, TEST 3
7.8 SOLIDS THROUGHPUT PER PASS, TEST 4
7.9 MEAN SOLVENT FLOWRATE, TEST 2
7.10 MEAN SOLVENT/FEED RATIO, TEST 2
7.11 MEAN SOLVENT FLOWRATE, TEST 3
7.12 MEAN SOLVENT/FEED RATIO, TEST 3
7.13 MEAN SOLVENT FLOWRATE PLUS ALTERNATE SOLVENT FLOWRATE, TEST 4
7.14 MEAN SOLVENT/FEED RATIO PLUS ALTERNATE SOLVENT/FEED RATIO
TEST 4
7.15 FEED/TREATED SEDIMENTS SOLIDS CONCENTRATION, TEST 2
7.16 FEED/TREATED SEDIMENTS SOLIDS CONCENTRATION, TEST 3
7.17 FEED/TREATED SEDIMENTS SOLIDS CONCENTRATION, TEST 4
5
14
19
23
25
32
41
53
53
53
55
57
61
61
61
67
67
67
67
67
67
69
69
69
-------
TABLES
Number
3.1
3.2
PROCESS EQUIPMENT DESCRIPTION
RANGE OF OPERATING CONDITIONS FOR TESTING
4.1 PREDEMONSTRATION RESULTS OF TOTAL SOLIDS, OIL AND GREASE, PCS,
AND pH ANALYSES
4.2 PREDEMONSTRATION RESULTS OF METALS ANALYSES
5.1 BENCH-SCALE TEST DATA
6.1 SAMPLES FOR CF SYSTEMS NEW BEDFORD TESTS 2, 3, AND 4
6.2 METHOD 680 ANALYTICAL RESULTS FOR TEST 4
6.3 PROCESS CONTROL AND FIELD MEASUREMENTS
7.1 PASS-BY-PASS PCB CONCENTRATIONS AND REDUCTION EFFICIENCIES
7.2 MASS ACCUMULATION AND LOSS IN THE SYSTEM
7.3 METALS CONTENT OF FEED, TREATED SEDIMENT, AND EXTRACT
7.4 EP TOXICITY CHARACTERISTIC OF TREATED AND UNTREATED SEDIMENTS
Page
15
16
27
28
30
44
47
49
52
58
64
65
vili
-------
ACKNOWLEDGMENTS
This report was prepared under the direction and coordination of Richard
Valentinetti, EPA SITE Program Manager in! the Risk Reduction Engineering
Laboratory, Cincinnati, Ohio. Contributors and reviewers for this, report were
Frank Ciavattieri of EPA Region I, Remedial Project Manager for the New
Bedford Harbor Superfund site; Jim Cummings and Linda Galer from the Office of
Solid Waste and Emergency Response; Paul Desrosiers, Diana Guzman, Paul de
Percin, and Laurel Staley from the Office of Research and Development;
Christopher Shallice and Thomas Cody, Jr. from CF Systems Corporation.
Logistics at the site was coordinated by Mark Otis, U.S. Army Corps of
Engineers; and Alan Fowler and Siegfried Stockinger of EBASCO Services, Inc.
This report was prepared for EPA's Superfund Innovative Technology
Evaluation (SITE) Program by Science Applications International Corporation
(SAIC), McLean, VA for the U.S. Environmental Protection Agency under Contract
No. 68-03-3485, by Don Davidson, Richard Hergenroeder, Kim Gotwals, Jorge
McPherson ,and Fernando Padilla. Laboratory analyses were conducted by
E.C. Jordan, Inc., Portland, ME, and Radian Corporation, Austin, TX.
ix
-------
ABBREVIATIONS AND SYMBOLS
amps amperes
ASTM American Society for Testing and Materials
bbl/day barrels per day
BNAs base/neutral and acid extractable compounds
Cd cadmium
COE U.S. Army Corps of Engineers
cP centipoise
Cr chromium
CR column reboiler
Cu copper
CWA Clean Water Act
dPa.s decapascal.seconds
ECD electron capture detector
EPA Environmental Protection Agency
EPT extract product tank
EP Tox Extraction Procedure Toxicity Test - leach test
F Fahrenheit
FK feed kettle
g grams
GC gas chromotography
gpd gallons per day
gpm gallons per minute
kw-hr kilowatt hours
Ibs pounds
Ib/gal pounds per gallon
Ib/min pounds per minute
max maximum
MBAS methylene blue active substances
mg milligrams
mg/kg milligrams per kilogram
min minimum
ms mass spectrometry
MSA method of standard additions
MS/MSD matrix spike/matrix spike duplicate
-------
ND
NIOSH
NR
ORD
OSWER
OVA
OZ
PAHs
Pb
PCBs
PCU
ppm
psig
QA
QC
RCRA
RPD
RREL
RSD
SARA
SBT
SITE
SRC
TDS
TS
TSS
VAC
VOAs
Zn
not detected
National Institute of Occupational Safety and Health
not reported
Office of Research and Development
Office of Solid Waste and Emergency Response
organic vapor analyzer
ounces
polyaromatic hydrocarbons
lead
polychlorinated biphenyls
Pit Cleanup Unit
parts per million
pounds per square inch gauge
quality assurance
quality control
Resource Conservation and Recovery Act of 1976
relative percent difference
Risk Reduction Engineering Laboratory
relative standard deviation
Superfund Amendments and Reauthorization Act of 1986
still bottoms tank
Superfund Innovative Technology Evaluation Program
solvent recovery column
total dissolved solids
total solids
total suspended solids
volts, alternating current
volatile organic analytes
zinc
less than
-------
-------
SECTION 1
INTRODUCTION
1.1 BACKGROUND
In response to the Superfund Amendments and Reauthorization Act of 1986
(SARA), the Environmental Protection Agency's Offices of Research and Develop-
ment (ORD) and Solid Waste and Emergency Response (OSWER) have established a
formal program to accelerate the development, demonstration, and use of new or
innovative technologies as alternatives to current containment systems for
hazardous wates. This program is called Superfund Innovative Technology
Evaluation or SITE.
The major objective of a Demonstration Program is to develop reliable
cost and performance information on innovative alternative technologies so
that they can be adequately considered in; Superfund decision making. Common
measurement, monitoring, and evaluation guidelines and protocols were
developed by ORD and used to collect the data and information from the
demonstration.
CF Systems Corp., developer of an organics extraction technology, was
selected to demonstrate their system at the New Bedford Harbor, Massachusetts
Superfund site. The system demonstrated was CF Systems' Pit Cleanup Unit
(PCU), a trailer-mounted system with a design capacity of 1.5 gpm (20
bbl/day). Successful application of the technology depends on the ability of
organic pollutants to solubilize in the process solvent, a liquefied gas. The
process used a mixture of liquefied propane and butane, at 240 psi and 69
degrees F, as a solvent for extracting organics from soils. As liquefied
solvent was mixed with the waste, organics were extracted into the solvent.
The solvent-organics mixture was then decanted from the separated solids and
water. The pressure of the solvent-organics mixture was reduced slightly to
vaporize the solvent which allowed separation from the organics. ''The solvent
was recovered by the system and compressed to a liquid for reuse.
The site is located on the Acushnet River Estuary north of Buzzard's Bay
in the city of New Bedford, Massachusetts, where harbor sediments contain
-------
pollutants discharged to the harbor from various industrial sources. The
pollutants include polychlorinated biphenyls (PCBs), polynuclear aromatic
hydrocarbons, copper, chromium, zinc, and lead. PCBs present the greatest
toxic threat and concentrations as high as 30,000 ppm have been observed. The
estimated volumes of harbor sediments containing various concentrations of
PCBs are:
PCB Concentration (ppm)
0-50
50-500
500-5,000
>5,000
Volume (cubic yards)
878,000
236,000
91,000
16,000
Samples of harbor sediments were dredged by the U.S. Army Corps of Engineers
and stored in 55-gallon drums for use in this demonstration.
1.2 PROGRAM OBJECTIVES
The objectives of this SITE demonstration of the CF Systems organics
extraction technology at the New Bedford Superfund site were to evaluate the
following:
1. Performance of the process in terms of PCB extraction efficiency and
a mass balance.
2. Variations in process operating conditions and possible effects on
performance.
3. Potential health and safety impacts resulting from system operation.
4. Equipment and material handling problems.
5. Projected system economics.
1.3 TECHNOLOGY EVALUATION CRITERIA
The following technical criteria were used to evaluate the effectiveness
of the CF Systems process for extracting PCBs from New Bedford Harbor
sediments:
1. System Performance
a. Evaluate PCB concentration in sediments before and after
treatment.
-------
b. Evaluate PCB extraction efficiency with each pass, or recycle of
sediments through the unit.
c. Evaluate mass balances established for total mass, solids, and
PCBs.
2. Operating Conditions
a. Compare operating conditions to operating specifications for
flow, temperature, pressure, and physical sediment character-
istics of the sediment and assess the effect on extraction rate.
3. Health and Safety Considerations
a. Determine if significant amounts of propane/butane or PCBs are
emitted to the air by the process.
b. Determine if staging area soils are contaminated by system spills
or malfunctions.
c. Decontaminate the unit with toluene to levels less than 50 ppm in
decontamination residues. H
4. Equipment and Material Handling Problems
a. Observe equipment and material handling problems that would
affect the performance of a full-scale site cleanup.
1.4 DESCRIPTION OF OPERATIONS
Contaminated sediments from five harbor locations were processed by the
PCU. The Corps of Engineers dredged sediments from the harbor and stored them
in 55-gallon drums for processing by the PCU. , Sediments were obtained from
locations H-20, H-21, H-22, H-23, and I-ll .shown in Figure 4.2. Drummed
sediments were sieved to remove particles greater than one-eighth inch that
could damage system valves. Water was also added to produce a pumpable
slurry. The drummed sediments were blended to provide feedstocks for four
tests as follows:
Test No.
1
2
3
Feed Stock
A 50-gallon mixture of sediments taken from locations H-20
H-21, and H-23. PCB concentration was 360 ppm. '
A 50-gallon mixture of sediments taken from locations H-20
H-21, and H-23. PCB concentration was 350 ppm. '
A 50-gallon mixture of sediments taken from locations H-20
H-21, and H-23. PCB concentration was 288 ppm.
-------
4 A 50-gallon mixture of sediments taken from locations H-22 and
1-11. PCB concentration was 2,575 ppm.
Test 1 was a system shakedown run to set flow rates and operating pressures
and to provide samples for laboratory evaluation of sample matrices. Samples
were collected during Tests 2, 3, and 4 to provide data for evaluating the
system's performance. A fifth test was run with toluene used as a feedstock
for decontaminating the PCU.
The process steps included extraction, phase separation and solvent
recovery. A simplified flowchart is shown in Figure 1.1. In step one,
sediments were fed into the top of an extractor at a rate of 0,9 gpm. In step
two, solvent was compressed to a liquid state and allowed to flow through the
same extractor. In the extractor, the solvent was thoroughly mixed with the
waste at a pressure of 240 psig. Following this extraction procedure, the
residual mixture of water/solids was removed from the base of the extractor
(step three). In step four, the mixture of solvent and organics left the top
of the extractor and was expanded across a valve prior to passing to a
separator. The reduction 1n pressure caused the solvent to vaporize through
the top of the separator. It was then collected and recycled through the
compressor as fresh solvent (step five). The liquid organics left behind were
drawn off from the separator and pumped to storage (step six).
About 1 to 2 hours were required to run a feedstock through the PCU.
Test 2 Involved passing, or recycling, the feedstock 10 times. Test 3
Involved three passes and Test 4 involved six passes. Samples were taken from
the feed kettle, extract collection tank, and treated sediment tank.
1.5 PROJECT ORGANIZATION
Through a Cooperative Agreement between EPA and CF Systems Corp., CF
Systems was responsible for operating their equipment and EPA prepared the
demonstration plan, prepared the test site, arranged for the sampling plan
analyses, conducted sampling, evaluated the data, and prepared the Technology
Evaluation Report. The evaluation included the following activities:
o Preparation of staging area to support testing, equipment setup, and
health and safety orientation of field staff
-------
Pre- and post-demonstration sampling and analysis of staginq area
soils
Public information meetings held to review CF Systems technoloav and
tests on New Bedford Harbor sediments
t A shakedown test, three extraction tests, and equipment
decontamination
Staging area closure and disposal of waste materials.
Simplified Flow Chart
Here
from
1. Solid or liquid waste fed Into
top of extractor.
_ _, unit operating cycle, for extracting
or solid waste:
4. Mixture of solvent gas and
organles leaves extractor, '
passes to separator through ,
valve where pressure Is
partially reduced.
Extractor
1.
Wastewater
or Sludge
2. Condensed by compression
at 70° F, solvent gas flows
upwards through extractor,
making non-reactive contact
with waste. Solvent typically
dissolves out up to 99+* of
organlcs.
3. Clean water or water/solids
mixture then removed from
extractor.
3.
Water
and/or Solids
OrganScs
Compressor
5. In separator, extraction gas
vaporized and recycled as fresh
solvent.
6. Organlcs drawn off from
separator, recovered for
disposal or recycling as
feedstocks or fuel.
Figure 1.1 Simplified Flow Chart
-------
SECTION 2
SUMMARY OF RESULTS
The program obtained a large amount of analytical and operating data and
was able to evaluate the four criteria stated in Section 1.3. A summary of
the results, which correspond to the program objectives, is presented below.
2.1 SYSTEM PERFORMANCE
The performance of the treatment unit was evaluated in terms of extrac-
tion efficiency and a mass balance. Extraction efficiency per pass is defined
as the input PCB concentration minus the output PCB concentration divided by
the input PCB concentration (multiplied by 100 percent). An inventory of
system inputs and outputs was established and evaluated for total mass, total
solids, and the total mass of PCBs. Results of these evaluations are
summarized as follows:
PCB removal efficiencies of 90 percent were achieved for sediments
containing PCBs ranging from 350 to 2,575 ppm. A high removal
efficiency was achieved after several passes, or recycles, of treated
sediments through the unit.
Extraction efficiencies greater than 60 percent were achieved on the
first pass of each test. Later passes of treated sediments through
the unit resulted in efficiencies that ranged from zero to 84 percent.
This wide range was due to solids retention in the system. Solids
retained in the system cross-contaminated treated sediments that were
recycled. Recycling was necessary to simulate the peformarice of a
full-scale commercial system. CF Systems' full-scale designs do not
include recycling, since additional extraction stages and longer
processing times are involved. In addition, only 50 to 150 gpd were
run through the unit, which was designed to handle up to 2,160 gpd.
Therefore, some solids may have been retained in equipment dead spaces
and intermittently discharged during subsequent passes.
A mass balance was not established for PCBs. A total of 157 grams was
fed to the unit during system shakedown and Tests 2, 3, and 4. Of the
total, 80 grams were accounted for in system effluents. Decontamina-
tion washes produced an additional 169 grams. The sum of effluents
and decontamination washes was, therefore, 101 grams greater than that
fed to the unit. This imbalance may be the result of limitations of
the analytical method. PCB analytical Method 8080 precision criteria
established for this project were plus or minus 20 percent and
accuracy criteria were plus or minus 50 percent. In addition, the
mass balance calculation was dominated by the Test 4 feed concen-
tration Therefore, error associated with the Test 4 feed sample
-------
could also be a source of the PCB mass imbalance. Another possibility
is contamination of the PCU from prior use at other sites. CF Systems
did not decontaminate the unit with toluene prior to the tests at New
Bedford. CF Systems' standard operating procedures now incorporate
decontamination with toluene.
A good mass balance was established for total mass and solids through
the system. A total of 3-1/2 tons of solids and water were fed to the
unit during Tests 2, 3, and 4. Of the total, 96 percent was accounted
for in effluent streams. A total of 789 pounds of solids were
processed during Tests 2, 3, and 4. Of the total, 93 percent was
accounted for in effluent streams. The slight imbalances, 4 and 7
percent, are attributed to the inaccuracy of the weighing device (1
percent), sample error, and accumulation of mass in system hardware.
Metals were not expected to be removed from the sediments, and were
not removed during the extraction. EP Tox test results indicate that
metals did not leach from either treated or untreated sediments.
Characteristics of the sediments, with respect to the EP Tox test,
were not changed by the treatment process, although high concentra-
tions of metals were present. Copper and zinc typically exceeded
1,000 ppm. Chromium and lead concentrations ranged from 500 to
1,000 ppm.
The decontamination procedure showed that PCBs were separated from the
sediment during the tests since nearly all of the PCBs were contained
in extract subsystem hardware. Of the 81 grams of PCB fed to the unit
during Tests 2, 3, and 4, only 4 grams remained in the final treated
sediments. Subsequent decontamination of the PCU with a toluene wash
showed that some PCB had accumulated in system hardware. However, 91
percent of the PCBs contained in decontamination residues were found
in extract subsystem hardware.
e A QA/QC review showed that analysis data of PCBs in sediments for
Tests 1 through 5 were sufficiently accurate and precise for an
engineering assessment of the efficiency of this demonstration.
2.2 OPERATING CONDITIONS
Operating conditions essential to the efficient performance of the PCU
were manually controlled and monitored during Tests 2, 3, and 4. Operating
conditions included feed temperature, particle size, flow rate, pH, and solids
content; solvent flow rate and solvent/feed mass ratio; and extractor pressure
and temperature. The unit generally performed as predicted by the developer,
although some deviations from the planned specifications occurred. An
evaluation of operating conditions is summarized as follows:
t Feed flow rates and extractor pressures were controlled throughout the
tests within specified ranges. Feed,flow rates ranged from 0.6 to 1 4
gpm. Extractor pressures ranged from 190 to 290 psig.
-------
During Test 2, feed temperatures for the last 4 passes were 10 degrees
F lower than the minimum specification, 60 degrees F. Decreased
extraction efficiency, which was apparent during this test, may have
been related to low feed temperatures. Sustained low temperatures
could have the effect of seriously reducing extraction efficiency in a
full-scale commercial system.
t Solvent flow fluctuated as much as 75 percent above and below the
nominal flow rate, 12 Ib/min. In Test 2, Pass 1 this caused the
sol vent-to-feed ratio to fall below specifications. This could affect
the extraction efficiency in full-scale system, since 1 ess solvent
would be available to extract organic pollutants from the feed soil.
Specifications for maximum particle size, one-eighth inch, were met by
sieving sediments through a screen. This was necessary to prevent
damage to system valves. Less than 1 percent of the sediment
particles were greater than one-eighth inch.
specifications for maximum viscosity, 1,000 centipoise, were met by
adding water to form a pumpable feed mixture. Feed viscositiestranged
from 25 to 180 centipoise. The mass of waste increased by about 33
percent, because unit operators arbitrarily added water.
Solids contents ranged from 6 to 23 percent and fell below the minimum
specification, 10 percent, after the fourth pass of Tests 2 and 4. A
10-percent minimum spec was set merely to ensure that the technology
would be demonstrated for high solids content feeds.
EPA and the developer will address corrective measures for operational
controls and material handling issues. However, these measures are
not the subject of this report.
2.3 HEALTH AND SAFETY CONSIDERATIONS
The Health and Safety Plan established procedures and policies to protect
workers and the public from potential hazards during the demonstration. Some
observations, based on an evaluation of health and safety monitoring conducted
during the demonstration, were:
Combustible gas meters indicated that the unit did not leak signifi-
cant amounts of propane. Therefore, operation of the unit does not
present an explosion threat much different than that associated with
domestic propane usage. Background air sampling and personnel
monitoring results indicate that organic vapors and PCB levels were
present at levels below the detection limit for the analytical
methods.
The unit did not cause a sudden release of propane and butane or
liquids. Only minor leaks occurred and staging area soils were not
affected.
*8
-------
The unit was not completely decontaminated before it left the site.
Toluene was fed to the unit as a decontamination wash. Toluene wash
was collected at the extract product tank and the two treated sediment
product tanks. The final toluene wash from the treated sediment tank
contained less than 34 ppm of PCBs. The final wash from the extract
tank contained 60 ppm of PCBs.
2.4 EQUIPMENT AND MATERIAL HANDLING PROBLEMS
Equipment and system material handling problems occurred during the
tests, although some problems were anticipated. Problems included:
Internal surfaces of extractor hardware and piping collected PCBs as
evidenced by mass balances for PCBs and subsequent washes of the unit
with a refined naphtha fuel and later with toluene. The washes
recovered accumulated PCBs as well as oil and grease. These accumu-
lations of organics are believed to be the result of the short dura-
tion of the tests and the small volume of organics contained in the
feed sediment, relative to the volume of the extraction system
hardware. PCBs are soluble in oil and grease, which is believed to
coat the internal surfaces of system hardware. Continuous operation
of the unit has resulted in continuous discharge of extracted organics
during other demonstrations of the technology.
The unit intermittently retained and discharged feed material solids.
This is the result of the relatively small volumes that were batch fed
to the unit. The unit was designed for continuous operation, not
short-term tests. In addition, only 50 to 150 gpd were run through
the PCU, which was designed to handle up to 2,160 gpd. Therefore,
some solids may have been retained in equipment dead spaces and
intermittently discharged during subsequent passes.
Solids were observed in extract samples, which were expected to be
free of solids. This indicates poor performance or failure of the
cartridge filter. An alternative type of filter should be
investigated by the developer.
Extractor and treated sediment hardware:contained organic sludge from
prior use of the unit at a petroleum refinery. Presence of the
petroleum residuals prevented complete interpretation of data
collected for oil and grease and semivojatile organics.
Low-pressure dissolved propane caused foaming in the treated sediment
product tanks. This hindered sample collection and caused frequent
overflow of treated sediment to a secondary treated sediment product
tank. CF Systems states that design of a commercial-scale unit will
allow release of propane entrained in the treated sediment and
eliminate the foaming.
-------
2.5 LESSONS LEARNED
Lessons learned that could be useful in planning future technology
demonstrations are discussed in this section. The lessons learned center on
demonstration preparation and logistics, treatment goals, and sampling and
analysis for PCBs.
Demonstration Preparation and Logistics
The developer's equipment must be decontaminated prior to being
brought on site. Quality Assurance Project Plans (QAPPs) should
include predemonstration sampling and analysis to ensure that the
technology is free of contaminants. The PCU contained residue from a
prior demonstration at a petroleum refinery. In addition, the system
carbon canister, which receives vented low pressure gas, had not been
changed.
t The developer must commit time needed to decontaminate the unit before
and after the demonstration, to conduct preventive maintenance, and to
respond to required sampling protocol. No equipment failures occurred
at New Bedford; however, a week was added to the schedule while
preventive maintenance was conducted.
The generation of trash (e.g., gloves and Tyvek suits) and process
residue can be substantial. Adequate provisions should be made for
waste residue and trash management. Program personnel and the
developer should develop realistic projections of the amount of
process residues and trash that will be produced as a result of the
demonstration. At New Bedford, 87 drums of waste were produced,
compared with 3 drums of harbor sediments fed to the unit. The large
volume of waste resulted from water added to the feed; water used to
decontaminate personnel and sampling gear; and trash, such as gloves
and Tyvek suits, that were not compacted.
Developer's Input
The developer should provide specific protocol for conducting shake-
down tests. This protocol should distinguish process control from
onsite process optimization.
The developer should provide a basis or a design procedure for scaling
up bench- or pilot-scale test results to a commercial-scale system and
for correlating batch test results to operation of a continuously fed
unit.
Sampling and Analysis Methods
t Interpretation of results from any PCB treatability study should
include a discussion of the precision of the analysis method.
10
-------
A portable gas chromatograph GC and a chemist should be available
onsite to allow a rapid response to changes in feed composition or
operational control. The Spittler Method was used at New Bedford as a
more timely alternative to EPA methods. However, even with this
method, 24 hours were required for sample shipment and subsequent
analysis.
e The unit did not selectively extract one class of PCBs. Reviewers
suggested the use of EPA Method 680, since the CF Systems technology
could have selectively extracted higher molecular weight PCB
congeners. Method 680 would reveal any selective extraction, since
this method is used to analyze individual PCB congeners. Method 8080,
a less expensive analysis method, would not reveal selective
extraction, since it is used to analyze mixtures of PCBs called
Aroclors, instead of individual congeners. EPA Method 8080 was chosen
over Method 680, since selective extraction was minor and it analyzes
for the classes of congeners that constitute the majority of PCB
contaminants (Aroclors 1242 and 1254) in the harbor sediments.
s Methods 680 and .8080 produced similar relative results, but very
different absolute results. Use of Method 680 in Test 4 showed a PCB
extraction efficiency of 96 percent and Method 8080 showed a similar
efficiency, 87 percent.; However, Method 680 showed an untreated sedi-
ment PCB concentration of 8,700 ppm; while Method 8080 showed 2,575
ppm. Data quality objectives were met for each measurement. There-
fore, regulatory or engineering interpretation of future PCB analyses
should include consideration of the,analysis methods used.
Conclusions and Recommendations
Even though solids retention caused cross-contamination of treated
sediments, significant PCB removals occurred.
System decontamination procedures showed that PCBs Were separated from
the sediment since nearly all of the PCBs were contained in extract
subsystem hardware. Of the 81 grams of PCB fed to the unit during
Tests 2, 3, and 4, only 4 grams remained in the final treated
sediments.
Benctvscale tests are useful for determining whether organics
contained in a soil will be extracted by a liquefied solvent such as a
propane-butane mixture. Bench-scale tests may also be used to
determine if a liquefied solvent selectively extracts specific classes
of organics, such as high or low molecular weight PCBs. Bench-scale
tests, however, do not yield information relating to operational and
material handling issues, such as pumpability, foaming, and
temperature.
Commercial-scale designs for application of the technology should
ensure that operating specifications are maintained. Wide fluctua-
tions in the feed-to-solvent ratio should be minimized, since
extraction efficiency is directly related to the amount of solvent
available for solubilizing organics contained in the feed. However,
the technology does accommodate wide ranges in operating conditions.
Solid wastes with viscosities up to 1,000 centipoise and solids
contents of 60 percent can be fed to the unit.
11
-------
* Feed materials are likely to be well below 60 degrees F throughout
winter months, which could affect system performance. Therefore, heat
must be added to sediments fed to a commercial-scale unit.
Pretreatment technology will be required to condition feed materials.
Coarse solids removal will be required to maintain feed sediment
particle sizes below one-eighth inch and water must be added to ensure
pumpability.
* Health and safety monitoring showed that OSHA Level B protection will
be necessary for personnel handling system input and output. However,
only OSHA Level C protection will be adequate for unit operators.
Operations, materials handling and health and safety issues are
addressed in the Application Analysis Report. Costs are estimated for
several case studies involving the New Bedford Harbor Superfund site.
A significant cost element for a full-scale system is extraction
process equipment, which must be scaled to handle much higher through-
puts (60 gpm) than the PCU (0.9 gpm). Full-scale extractors have 4 to
6 foot diameters as compared with the 18 inch diameter of the PCU
extractors. Recommended pretreatment technology includes conveyors,
screening, heat and water addition, and mixing and holding tanks.
Post treatment technology includes treated sediment dewatering,
wastewater treatment and reuse, holding tanks, conveyors, and disposal
of treated sediments and extracted organics. Onsite analytical
capabilities and health and safety program Implementation are
additional cost elements.
12
-------
SECTION;3
PROCESS DESIGN
3.1 PROCESS DESCRIPTION
CF Systems Pit Cleanup Unit (PCU), shown in Figure 3-1, is a continuous
processing unit that used a liquefied propane/butane mix as the extraction
solvent. The solvent mix was 70 percent propane and 30 percent butane. For
each of the 3 demonstration tests, a batch of approximately 50 gallons of
sediments was fed to the unit at a nominal, rate of 0.9 gpm. Feed viscosity
was maintained below 1,000 cP, by adding water to produce a pumpable slurry.
Particles greater than one-eighth inch were screened from the feed to prevent
damage to valves. Sediments were pumped to the extractors, which were
typically operated at 240 psig and 70 degrees F. Liquefied solvent was also
pumped to the extractors at a rate of 2.3 gpm (10 Ib/min) and mixed with the
sediments. Organics, such as PCBs that are soluble in the liquefied solvent,
were extracted. After extraction, treated sediments were decanted and
separated from the liquefied solvent and organics mixture. The mixture flowed
from the extractor and passed to a separator through a valve that partially
reduced the pressure. The pressure reduction caused the solvent to vaporize
and separate from the extracted organics. The solvent was recycled and
compressed to a liquid for reuse in the system.
The PCU was not designed for large-scale remedial actions. Therefore,
treated sediments were recycled, or passed through the unit to simulate
operation of a commercial-scale unit. CF Systems' commercial-scale designs do
not include recycling. These designs feature 60 gpm flowrates, several
extraction stages, and longer processing times'."*
3.2 EQUIPMENT SPECIFICATIONS
The major pieces of equipment and their function are described in Table
3-1. Process equipment that contacted the solvent or feed materials were
constructed of 316 stainless steel. All process pumps were constructed of
stainless steel, and both compressors were made of carbon steel. All of the
process equipment was designed to withstand'temperatures and pressures that
13
-------
Figure 3-1. Pit Cleanup Unit
14
-------
TABLE 3-1. PROCESS EQUIPMENT DESCRIPTION
Process Equipment Designation
Function in System
Feed Kettle
Basket Strainer
Extractor l
Decanter 1
FK
S-l
E-l
D-l
Extractor 2
Decanter 2
Cartridge Filter
Solvent Recovery
Column
Column Reboiler
E-2
D-2
F-2
SRC
CR
Treated Sediment RPT-l
(Raffinate) Product Tank RPT-2
Extract Product Tank
EPT
Holds approximately 100 gallons of
strained, slurried feed. Counter-rotating
agitators homogenize feed.
Prevents oversized (>l/8 inch) feed
material from entering the system.
Extracts organics from water-solids feed
mixture with sol vent from D-2.
Allows! separation of solvent-organic
mixture from water-solids layer. Sends
water-solids layer to Extractor 2 (E-2)
and sblvent-organics layer to the solvent
recovery system.
Extracts organics from water-solids
mixture with fresh propane from the
solvent recovery process.
Allows separation of solvent-organics
layer from water-solids mixture.
Filters residual solid fines from solvent-
organics stream leaving Decanter 1 (D-l).
Separates propane solvent from organics
via pressure reduction and heat from the
Column Reboiler (CR). Solvent vapor flows
out the overhead while organics are
deposited in the CR.
Provides both holdup for the recovered
organics and heat for the Solvent Recovery
Column (SRC) via a tube bundle heat
exchanger.
Receives treated sediments (raffinate)
from Decanter 2 (D-2). Recovers residual
propane via flash pressure reduction and
heat from water jacket. RPT-2 receives
RPT-l overflow.
Receives extracted organics effluent from
the Column Reboiler (CR). Recovers
residual propane via flash pressure
reduction and heat from the water jacket.
15
-------
TABLE 3-1. PROCESS EQUIPMENT DESCRIPTION (Continued)
Process Equipment Designation
Function in System
Main Compressor
C-l
Low Pressure Solvent
Compressor
C-2
Compresses both Low Pressure Solvent
Compressor (C-2) outlet solvent and
Solvent Recovery Column (SRC) overhead
solvent. Outlet sent to Column Reboiler
(CR) for heat exchange before returning
to Extractor 2 (E-2).
Compresses scavenged propane from Extract
and Raffinate Product Tanks (EPT, RPT-1,
and RPT-2). Sends compressed solvent to
Main Compressor (C-l).
TABLE 3-2. RANGE OF OPERATING CONDITIONS FOR TESTING
Extractor Pressure (PSIG)
Extractor Temp, (degrees F)
Feed Temp, (degrees F)
Solvent Flow (Ib/min)
Feed Flowrate (GPM)
Solvent/Feed Ratio
Feed Solids (percent by weight)
Solids Size (maximum)
pH (standard units)
Viscosity (cP)
Minimum
180
60
60
8
0.2
1
10
6
0.5
Nominal
240
100-110
70
12
0.2-0.5
1.5
30
7
10
Maximum
300
120
100
15
1.5
2
60
1/8 inch
12
1,000
16
-------
exceed normal operating conditions. To guard against sudden overpressure,
each vessel had a relief valve that vented to a header system that discharged
to the pollution control system.
The utility and process materials requirements that were necessary to
operate the PCU at New Bedford Harbor were:
t
Electricity-480 VAC 3 Phase, 100 amps
Process Water5 GPM, 60-80 degrees F inlet, 30-90 psi
Potable WaterAvailable
Propanefour, 100 gallon bullets, 95-97 percent purity
ButaneAs needed, for Propane/Butane (70/30) solvent mix
Nitrogen (for pressure testing during shakedown period)(2) 1A size
cylinders.
Utility usage for a commercial-scale unit cannot be easily compared with the
PCU because pilot-scale equipment consumed much more energy per gallon of
throughput.
The operating conditions listed in Table 3-2 are essential to the
efficient operation of CF Systems' pilot-scale unit. Failure to operate the
unit within the specified operating ranges can result in decreased extraction
performance. The operating parameters were set during the shakedown portion
of the demonstration. CF Systems claimed that minor fluctuations would not
affect performance.
The feed temperature is that of the material piped into the feed kettle.
The feed must be maintained above 60 degrees F to avoid freezing, which could
interfere with the extraction process. The feed must be maintained below 120
degrees F to prevent vaporization of the solvent.
The extractor pressure, measured at the gauges on extractors 1 and 2, is
controlled by the main compressor and at the extract discharge from the
extraction segment of the unit.
17
-------
The viscosity and solids content must be such that the feed material is
pumpable. Pretest sampling determines the viscosity of the potential feed.
Any potential feed with a viscosity above the listed range is slurried with
water to yield a pumpable mixture.
In order to prevent damage to the process equipment, the pilot-scale unit
has a maximum limit for solids size. Basket strainers, located between the
feed pump and the first extractor, prevent larger-than-allowable size solids
from entering the system. Oversized solids removed from the feed were hauled
to a RCRA-approved facility.
The feed flow rate represents the rate at which material is pumped from
the feed kettle into the extraction system. Operational flow rates above the
listed maximum can force segments of the system, such as decanters and control
valves, beyond their effective hydraulic capacity. The feed flow rate is
manually controlled through the feed pump controller located beneath the feed
kettle. Average detention time of throughput is about one hour.
3.3 PROCESS FLOW DIAGRAM
The PCU process flow diagram is shown in Figure 3-2. The extraction
portion of this unit consisted of two stages of counter-current extraction
with solid-liquid separation between the extractors. The feed was transferred
from a feed preparation drum to the feed kettle with a pump. In the feed
kettle, slurry solids were kept suspended while in the feed kettle by two
counter-rotating agitators. During this process, feed was pumped from the
feed kettle through a basket strainer, which removed any particles greater
than 1/8 inch in diameter. Then feed flowed to the first extractor, where
feed was mixed with the liquid propane/butane solvent. An agitator (not shown
in the figure) provided mixing action before the solvent-organics mixture
flowed to decanter 1. At decanter 1, the mixture separated into two
iiiiscible layers. The solids and water settled into the underflow to the
second extractor. The decanter overflow, which contained extracted organics,
propane/butane, and fine solids, flowed through a filter and then to a solvent
recovery column.
18
-------
H
Pressure Letdown Valves
Treated
Sediment
Product
Tank
#1
V
Treated
Sediment
Product
Tank
#2
Extract
Product
Tank
T
Cartridge
Filter
Solvent
Recovery
Column
Figure 3-2. CF Systems Process Schematic
Legend
Feed
Propane-Butane Solvent
Propane/Organics Mixture
Extracted Organics
Processed Sediments
-------
The pressure difference between the first decanter and the second
extractor moved the solid-liquid stream into the second extractor for second-
stage extraction. Fresh liquefied solvent (propane/butane mixture) from the
solvent recovery process then mixed with the solids/water stream and further
extracted the organic components. An agitator (not shown in the figure),
Which was located above the second extractor, provided mixing action before
the solvent-organics mixture flowed to decanter 2. At decanter 2, two
Immiscible layers were formed. The organics-solvent layer floated to the top
while the solids sank into the underlying water layer. The lower water-solids
layer flowed from the bottom of the decanter to the treated sediment product
tanks, while the upper organics-solvent layer recycled to the first extractor
for final stage extraction.
The organic-solvent stream from the first stage extractor passed through
a filter cartridge that collected fine solids and went to the solvent recovery
column. In the solvent recovery column, the solvent vaporized and was removed
from the column overhead, while the organics remained as a separate liquid.
The mixture of organics containing dissolved propane gathered in the column
reboiler and subsequently passed to the extract product tank. Solvent from
the column overhead flowed to the main compressor. The compressed solvent
passed through the column reboiler heat exchanger to provide the heat
necessary to boil off residual solvent remaining in the organic mixture. The
condensed solvent left the reboiler and re-entered the extraction system via
the second extractor.
The residual solvent that vaporized off the system products in the
extract or the treated sediment tanks flowed to the low-pressure solvent
compressor. The outlet stream of the low-pressure solvent compressor fed to
the main compressor, where it was compressed along with vapors from the column
overhead.
During system shutdown or if overpressure within a vessel opened a relief
valve, material was vented to a relief header, which directed the material to
a blowdown tank where solids and liquids were removed from the vented stream.
The gases from the blowdown tank passed through a 42-gallon activated carbon
20
-------
filter to remove contaminants in the propane gas. The gas then passed through
a flame arrestor and was vented to the atmosphere. This system was used only
once during the demonstration, at the conclusion of PCU decontamination.
21
-------
SECTION 4. DEMONSTRATION SITE DESCRIPTION
4.1 SITE CHARACTERISTICS
The site selected for demonstration of CF Systems' process was New
Bedford Harbor, located in New Bedford, Massachusetts. During the 1970s, PCBs
and other contaminants were identified in the sediments and marine life of New
Bedford Harbor and parts of Buzzard's Bay. Studies conducted by EPA in 1980
led to New Bedford Harbor being proposed in 1982 for EPA's National Priorities
List. The main areas of New Bedford Harbor under EPA investigation are the
Acushnet River Estuary and the harbor. The estuary is the area of the site
north of the Coggeshall Street Bridge shown in Figure 4-1. Areas of extremely
high PCB contamination are located at the northern tip of the estuary. The
lower harbor includes Buzzard's Bay and the waters below the Coggeshall Street
Bridge. The demonstration took place on a parcel of city-owned property
adjacent to the cove north of the Coggeshall Street Bridge, as shown in Figure
4-1. Area within the dashed line was also the site for a pilot dredging and
disposal study conducted jointly by EPA, the Massachusetts Department of
Environmental Quality Engineering, and the U.S. Army Corps of Engineers (COE).
PCB concentrations in the harbor range from a few ppm to more than 30,000
ppiB and elevated levels of copper, chromium, zinc, and lead also are present
at the site (COE, 1987). Most organic compounds detected in sediments are
co-located with PCBs and occur at concentrations less than or equal to the
observed PCB concentrations (Ebasco, 1987). An evaluation of the relative
toxlcitles of PCBs indicates that the environmental and public health risks
from sediment exposure will be dominated by the PCB constituents. However,
some polynuclear aromatic hydrocarbon (PAH) compounds occur at certain
locations, and at concentrations that may significantly contribute to the
overall environmental and public health risk associated with exposure to the
sediments.
PCBs consist of a mixture of chlorinated biphenyls, which contain a
varying number of substituted chlorine atoms on aromatic rings. The persis-
tence of PCBs in the environment and their toxicity increases as the chlorine
content increases. The commercial products of the complex chlorobiphenyls
22
-------
Confined
Disposal
Facility
(CDF)
(not to scale)
Figure 4-1. Demonstration Location
23
-------
were registered and manufactured under the trademark "Aroclor." Fifty percent
of the congeners in the New Bedford Harbor area of interest have been identi-
fied as Aroclors 1242 and 1254. Therefore, PCBs were measured using the
Aroclor 1242 and 1254 standards.
Based on the available literature (COE 1987, Ebasco 1987), .the pollutants
of concern were determined to be PCBs, semivolatile organics (which includes
PAHs), cadmium, chromium, copper, zinc, and lead. CF Systems requested
analyses for additional parameters to evaluate the overall system performance.
011 and grease was used by CF Systems as a process control parameter since the
majority of organic pollutants are extracted by the process. Total solids,
pH, and viscosity can affect equipment performance. Therefore, these
parameters were chosen for monitoring in addition to the pollutants of
concern.
4.2 PREDEMONSTRATION SAMPLES
COE dredged collected sediment from the Acushnet Estuary for use in this
demonstration project. An attempt was made to gather sediments from areas
with differing levels of contamination so that a range of concentrations would
be available for the demonstration. Collected sediments were stored in drums
that were assigned identification numbers. Drum Numbers H-20, H-21, H-22,
H-23, and 1-11 were later used in the demonstration. The drum identification
numbers correspond to a grid system used by COE in previous harbor characteri-
zation work. This grid system shown in Figure 4-2 enabled COE to estimate
probable concentrations from particular areas to ensure that an appropriate
range of samples was collected.
The drummed sediments were conditioned for predemonstration sampling by
removing solids greater than 1/8 inch, adding water to produce a slurry, and
stirring. Contents of each drum were hand-shoveled and sieved through a
1/8-inch screen into a second drum. More than 99 percent of the solids were
passed through the screen. As the second drum was filled, harbor water was
added to produce a stirrable slurry. Slurried contents were manually stirred
with a shovel and three samples were drawn after 15 minutes of mixing.
Samples from each drum were analyzed for PCBs, oil and grease, pH, total
24
-------
NJ
Ul
D VMM 0-50ppm
c iHI 5o-500ppm
EZjQ 500-5000ppm
Ex^?j >5000ppm
1. a I '' 14'. ' . «. 7 . '.. .» I « . " . *
Demonstration
Site
, ,
14 . 15 | 16 t 17 | IB ( 19 | 20 | 21 22 23 24 25 26
27 28 29 30 31 32 33 34
Note: Depths are 0-12"
Figure 4-2. Predemonstration Samples
-------
solids, moisture content, semivolatile organlcs, and sediment particle size
distribution. The methods used for the sampling and analysis of predemon-
stratlon samples were the same as those used during the tests (see Section 6.3
for a list of analytical methods).
Predemonstration results of total solids, oil and grease, PCS, and pH
analyses are summarized in Table 4-1. Total solids ranged from 29 to 45
percent solids. Oil and grease contents ranged from 1.5 to 7.8 percent. No
correlation was apparent among total solids, oil and grease, and PCB concen-
trations. As expected, the sediments from location I-ll contained PCBs well
over 5,000 ppm. The other samples ranged from 160 to 640 ppm of PCB. The
range of pH was 7.3 to 8.2 standard units among the samples.
Data reported for particle size distribution and semivolatile organics
were similar among the drummed sediments. Particle size distributions were
approximately 37 percent sand, 41 percent silt, and 22 percent clay for each
drum. PAHs were the predominant class of semivolatiles, as expected, based on
the literature (COE 1987, Ebasco 1987). PAHs detected in each of the 5 drums
included napthalene, acenapthalene, dibenzofuran, fluorene, phenanthrene,
anthracene, d1-n-butylphthalate, fluoranthene, pyrene, benzo(a)anthracene,
chrysene, bis(2-ethylhexyl)phthalate, benzo(a)pyrene, 1ndeno(l,2,3-cd)pyrene,
dibenz(a,h)anthracene and benzo(g,h,i)perylene.
A composite sample and a composite sample replicate were obtained for
overall waste characterization, from the drummed sediments selected for the
demonstration. Also included in these composites were aliquots from four
drums not used in the demonstration. The composites were analyzed for 23
metals, cyanide, volatile organics, and extraction procedure (EP) toxicity
metals. Analysis results for 23 metals and cyanide are shown in Table 4-2.
Metals and cyanide analysis results compare well with previous studies that
showed high levels of copper, chromium, lead, and zinc (COE, 1987). EP tox
metals data are discussed in Section 7.1, and were less than RCRA regulatory
limits. Volatile organics analyses were run for 34 parameters and the fol-
lowing were detected at levels greater than 1 mg/Kg; vinylchloride, methylene
chloride, acetone, toluene, 1,2-dichloroethene, 2-Butanone, and ethylbenzene.
The volatile organics data are not critical to this demonstration since the
measured concentrations are low relative to semivolatile organics data.
26
-------
TABLE 4-1. PREDEMONSTRATION RESULTS OF TOTAL SOLIDS, OIL AND GREASE
PCB, AND pH ANALYSES
Sample (a)
H-20
H-21
H-22
H-23
1-11
Concentration
Total Solids
(percent) (b)
29
34
30
43
35
Concentration
Oil and Grease
(percent) (b)
3.3
1.5
7.8
1.2
7.1
Concentration
PCB
ppm (b, c)
402
333
. 640
162
32,333
PH
(Range in
Standard Units)
7.3-8.2
7.6-8.1
7.7-7.8
7.8-8.2
7.3-7.6
C°E
(b) Represents the average of 3 grab samples taken from each drum.
(c) Polychlorinated biphenyl (PCB) reported as the sum of Aroclors 1242 and
27.
-------
Parameter
TABLE 4-2. PREDEMONSTRATION RESULTS OF METALS ANALYSES
Concentration
(mg/kg) (1)
Aluminum
Antimony
Arsenic
Barium
Berylium
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Nickel
Potassium
Selenium
Silver
Sodium
Thallium
Vanadium
Zinc
Cyanide
10,000
8
12
140
4
29
2,550
425
9
900
16,500
590
5,300
185
1.4
130
2,200
ND (2)
4
11,000
ND
70
1,850
Less than 6
Notes: 1. Data reported are the mean of 2 grabs from a composite of
predemonstration samples.
2. ND indicates not detected.
28
-------
SECTIONS. FIELD ACTIVITIES
SITE Program activities were conducted for the CF Systems New Bedford
Harbor demonstration program from June to October, 1988. In July, the U.S.
Army Corps of Engineers (COE) obtained drummed samples of harbor sediments.
CF Systems conducted bench-scale tests on archived sediment samples in June
and later, in August, conducted bench-scale tests using the COE samples.
Public meetings were held and a Fact Sheet was distributed throughout June,
July, and August. The demonstration was conducted in September.
The bench-scale tests are described in Section 5.1. The plans for the
demonstration are described in Section 5.2. A summary of the actual
operations is described in Section 5.3. The few operational deviations from
the Demonstration Plan are also discussed in Section 5.3 along with some
additions to the analytical testing program that were incorporated into the
Program while at the site.
5.1 BENCH-SCALE TESTS
Bench-scale tests were conducted, prior to the demonstration, in order to
determine the best operating conditions for the PCU. Bench Test 1 consisted
of extracting PCBs from three different sediment samples. In Bench Tests 2
and 3, a fourth and fifth sample were run. Each sample was divided into 3
portions and each portion was extracted at different sol vent-to-feed ratios.
The bench-scale extraction apparatus consisted of a stainless steel,
single-stage, counter-current extractor with a mechanical stirrer and a
solvent delivery system. A nominal amount of material (between 180-190g) was
placed in the extractor (about 0.8 liters capacity) at the beginning of a run.
A portion of the remaining volume in the extractor was filled with solvent.
The system was then pressurized to 150-200 ,psig at an ambient temperature of
70-80 degrees F: The stirrer activated and the contents were mixed. After a
suitable residence time, the stirrer was switched off and the contents allowed
to settle. The system was depressurized in order to permit collection of
raffinate and extract samples.
29
-------
Test data are shown in Table 5-1 that include the test number, sample
number, feed PCB concentration, and treated sediment PCB concentration for
respective sol vent-to-feed ratios. The data generally show that higher ratios
of sol vent-to-feed used in the extraction result in higher extraction
efficiencies. However, data for sample 2 (ratio 20:1) and sample 4 (ratio
2:1) do not show this trend. These differences may be attributed to the range
of accuracy associated with the screening method used to analyze PCBs in each
sample.
CF Systems used the data in Table 5-1 to roughly estimate the number of
passes required to treat wastes to specific levels in the demonstration.
For example, the single-stage bench test results for sample 4 showed that 210
ppm could be reduced to 17 ppm with a 20-to-l feed-to-solvent ratio.
Therefore, CF Systems estimated that the two-stage PCU would require 10
passes, or recycles, at a 1.5-to-l feed-to-solvent ratio to achieve a similar
reduction.
Bench-scale tests were useful for confirming the solubility of PCBs in
the liquefied solvent. However, these tests were only of limited use for
setting PCU operating conditions. In addition, the bench-scale tests did not
provide data relating to operational and material handling issues such as
temperature requirements, foaming, and pumpability.
TABLE 5-1. BENCH-SCALE TEST DATA
Test Number
Sample Number
1
1
1
2
1
3
2
4
3
5
Samp!e
Sol vent/Feed
Ratio
PCB Concentration (ppm)
Feed
Treated Sediment
Treated Sediment
Treated Sediment
(Not Applicable) 6,200
2:1
10:1
20:1
4,600
3,800
3,300
210
170
23
68
5,300
4,000
1,600
430
210
340
57
17
11,000
5,800
940
220
30
-------
5.2 OPERATIONS SUMMARY
CF Systems' pilot unit occupied a 60- by 100-foot staging area adjacent
to Sawyer Street, as shown in Figure 5-1. ; The staging area, covered with an
aggregate base of crushed 3/4-inch stone wjth a 6-inch depth, was amenable to
both tractor trailer and hand-cart traffic. An 8-foot high, chain-link fence
topped with 2 feet of barbed wire cordoned off the staging area from the
remainder of the EPA-COE controlled location. An 18-foot wide gate provided
access to the site. Access to the demonstration test staging area was
provided through a hard-packed earth roadway running through the EPA-COE area
This roadway exited onto Sawyer Street. Security for the demonstration test '
staging area was provided 24 hours per day.
The overall procedure started with preparing sediments for feeding to the
PCU for each test. Sediments previously collected by COE and stored in drums
were sieved to remove particles greater than 1/8 inch and other debris, such
as leaves and sticks. The sieving apparatus consisted of a custom-made steel
rim, a removable 1/2-Inch screen, and a removable 1/8-inch screen. Initially
both screens were being used, but after several days of sieving, it was
determined that only the l/2-inch screen was necessary because the 1/8-inch
screen was not catching any additional material not already caught by the
1/2-inch screen. Later, during test l, CF Systems expressed concerns that
oversized feed material was interfering with the unit's pumps. In order to
determine whether oversized materials had not been sieved out of the feed
material, the feed material from the test was sieved through an aluminum
window screen (1/8 inch) placed on top of the l/8-inch sieve originally used
on the drum samples. A small amount of humus material was caught between the
two screens. Sediments were shoveled onto the screens, which were placed over
an empty drum. A trowel was used to push the sediments through the screens
and harbor water was used to wash the sediments into the drums. Feed prepared
for each test was pumped into the feed kettle.
Each of the four tests were run similarly except that number of passe-s
and PCB concentrations were varied for each test. A pass was defined as one
cycle of the feed through the PCU. A pass of feed results in a treated
sediment product and an extract product. Collecting and recycling the treated
31
-------
U)
ro
o
in
I ( I Jl L I I ^ I
O O f« Alternate Drum
Storage Area
Exclusion Zone
Propane Tanks
QQ99
Unit
Tarps
Alternate Drum
Storage Area
Exclusion Zone
Exclusion
Zone Lines
Power
Pole
Do
o
North
Fence - 8' high topped with 2' barbed wire
Gate No. 1
Vehicle Access
0 Fire Hydrant
Figure 5-1. Plan Sketch for New Bedford Harbor Demonstration
Site Demonstration
Layout
CF Systems
Extraction of PCBs
New Bedford Harbor
Plan Sketch
-------
sediment through the PCU constituted an additional pass. Recycling was
conducted to simulate the operation of a full-scale commercial system. The
PCU is only a two-stage system, whereas commercial designs include four or
more stages, longer extractor residence times, and longer phase separation
times. Conditions that varied for each test were:
*' Ihf^,,WaST£UVS/ shakedown test to set pressure and flowrates in
drL nlh ThVond uao,a 5°-9allon composite of sediments taken from
nf 5m ±PS ?h20' H~21' and H'23' The feed had a PCB concentration
hancmn passes were run to gain experience with materials
2.
3.
Test 2 was a 10 pass test. The feed was a 350 ppm, 511 pound
composite of sediments taken from drum numbers H-20, H-21 and H-23
Ten passes were run to simulate a high-Bfficiency process'and to '
achieve treated sediment levels less than 10 ppm. A 350 ppm
concentration was chosen for this test since this represents an
average, or typical, PCB concentration in the harbor
Test 3 was a 3 pass test. The feed was a 288 ppm, 508-pound
composite of sediments taken from drum numbers H-20, H-21 and H-23
The purpose of this test was to reproduce the results of the first
three passes of Test 2.
nnmnn' ", Ttl8 feed- W3S 8 2,575 PPm, 299-pound
composite of sediments taken from drum numbers I-ll and H-22 The
purpose of this test was to reduce a high-level waste to a lower
level waste such as that used in Tests 1, 2, and 3. High-level
wastes are found at several "hot spots" in the harbor.
Decontamination of the system involved running toluene through the PCU as a
solvent wash.
Samples were taken of the feed at the commencement of each test. Treated
sediment products and extracts were planned for sampling at each pass.
Additional samples were taken of system filters and strainers, although the
amount of PCB contained in these miscellaneous samples later proved to be
small. PCU operating pressures, temperatures, and flow-rates were monitored
throughout the tests. Field tests were conducted for feed viscosity, pH, and
temperature.
Information was logged in a notebook to report the sampling operations
and the overall operations at the site. This included the following:
33
-------
Operating conditions, gage reading, mass,and volume determinations
I
Notes on daily preparations of CF Systems and EPA
Problems
Health and Safety related procedures, meetings, and concerns
Chronology and summary of daily activities, including check-in and
check-out of all personnel
Weather conditions.
5.3 OPERATIONS CHRONOLOGY
The demonstration commenced at the New Bedford site on September 6 and
continued until September 29. Test 1, the system shake-down, was conducted on
September 6. Test 2 was conducted from September 8 through 18. Test 3 was
conducted from September 18 through 20. Test 4 was conducted from September
20 through 29. System decontamination took place from September 30 through
October 4. Highlights of the site preparation, operation, and decontamination
are described below.
Samples
Test 1
The planned system shakedown, Test 1, consisted of three passes.
were taken at the following locations for PCB analyses:
t Feed kettle
Treated sediment product tank
Extract product tank.
The feed was a composite of sediments taken from COE drums nos. H-20,
H-21, and H-23. The feed had a PCB concentration of 360 ppm. Field personnel
reported that Pass 1 treated sediments contained some foam and appeared to
have a very low solids content relative to the feed. A layer of foam several
inches thick formed in the treated sediment collection drum during Passes 2
and 3 Since the foam hindered sampling, the treated sediment was allowed to
degas in collection drums overnight. A Pass 1 extract sample was collected
several days after the Pass 1 raffinate sample. This delay ocurred over a
holiday weekend.
34
-------
Test 2 :
Test 2 was a 10-pass test conducted on a low PCB concentration (350 ppm)
sediment. The source of sediments was the same as for test 1 (COE drums nos.
H-20, H-21, and H-23). Samples were taken at the following locations:
Cartridge filter
Basket strainer
Feed kettle
Treated sediment product tank
Extract product tank.
The cartridge filter was sampled as planned. The sampling plan was
modified to include basket strainer sample collection at the end of each pass
rather than at the end of each test. The basket strainer filled with
particulate matter more rapidly than anticipated. This procedure was repeated
for all passes except Pass 9 when the sample volume was too small. Extract
samples were taken only at the end of Passes 1, 3, and 10 because the volumes
generated with each pass were so small. Treated sediment samples were taken
at the end of each of the 10 passes. Grab samples of tap water and harbor
water also were taken. Field measurements included pH, weight, and viscosity.
in addition to changes in the sampling plan mentioned for extract and
basket strainer sampling, other deviations occurred. For example, the feed
material was sampled three times. The second feed sample was obtained after
operating personnel added tap water to the feed slurry. This was done to
determine the effect water content would have on analyses for PCBs. A third
feed sample, "Pass 4 feed" was a mixture of\ Pass 3 treated sediments taken
from product tank no. l and overflow collected from product tank no. 2. The
second tank needed draining to allow continued operation. The drainage was
mostly foam, which was thought to contain material from Passes 1 and 2, as
well as Pass 3. Therefore, treated sediment collected from both tanks'was
composited as a representative sample of the Pass 4 feed.
Sampling logistics were modified at Pass 4 for all future passes because
of the foaming problem. Since foaming in both treated sediment tanks was
35
-------
significant, collection drums were positioned at each. Treated sediments
collected from each tank would be composited in a single drum after degassing
for each pass. Samples then would be taken from the composite of both tanks.
Foaming was a continuous problem throughout Test 2, so treated sediment was
left in collection drums overnight to degas for Passes 1, 2, 3, 5, and 8.
Saraples for Passes 4, 6, 7, 9, and 10 were able to be collected on the same
day the run occurred since several daylight hours were available to allow
degassing to occur before samples were acquired.
Sampling personnel made the following observations on the physical
characteristics of the various wastestreams:
Treated Sediment. Foaming caused two sampling problems: (1) treated
sediment would splatter out of the collection drums and 2) represent-
ative samples of liquids and solids could not be drawn with a samp e
beaker Analyses for methylene blue active substances (MBAS), total
dissolved solids (TSS), and total suspended solids (TSS) subsequently
were requested for the tap water, harbor water, and Test 3 reed.
Results from these analyses indicated no significant levels of MBAS,
TSS, or TOS.
Extract. Only very small volumes were available for sampling.
Laboratory personnel observed parti culates in each extract sample.
« Basket Strainer. With each pass the strainer collected one gallon of
liquid that contained fibrous solids and large particles.
Carbon Canister. The carbon canister had not been changed since prior
use of the unit at a petroleum refinery.
Tap Water and Harbor Water. Field personnel conducted an
field experiment to determine if either the tap or the harbor could be
the source of foam in the raffinate. Aeration of a iquots of tap and
harbor water produced no foaming. Subsequent methylene blue active
substance analyses indicated that neither tap nor harbor water would
be a source of foaming agents.
Cartridge Filter. The cartridge filter was coated with a dark parti-
culate as anticipated.
Test 3
Test 3 was designed to duplicate the first three passes of Test 2 and the
test was conducted as such. However, two events occurred that caused some
concern about sample integrity: (1) communications among the CF Systems and
EPA revealed that the unit had not been decontaminated prior to departing from
36
-------
a prior demonstration at a petroleum refinery; and (2) potentially contami-
nated wash water inadvertently got mixed with Pass 1 treated sediments.
The source of sediments was the same as used for Tests 1 and 2 (COE drums
nos. H-20, H-21, and H-23). Samples were taken at the following locations:
Cartridge filter
Basket strainer
Feed kettle
Raffinate product tank
* Extract product tank.
The cartridge filter was sampled at the end of the test. The basket
strainer and raffinate were sampled after each of the three passes. Extract
was sampled at Passes 1 and 3. Extract sample volumes were small and their
flow was erratic at each pass. A sample of feed was collected for methylene
blue active substances analyses to determine if the harbor sediment, contained
agents that caused foaming. Foaming problems persisted through Test 3, so
samples were degassed overnight and collected the following day. Field
measurements included pH, volume, weight, and viscosity.
Several additional samples were taken to assess the impact of residuals
left in the unit from a prior demonstration. Site personnel attempted to wash
out these residuals after Test 3, Pass 1 with a refined naphtha fuel. Another
sample was taken to determine the effect of the inadvertent addition of wash
water to the Pass 2 feed. Each of these events is discussed below.
Fuel Wash - Test 3 was interrupted to clean the extract product tank
(EPT) and the column reboiler (CR). The CR was opened and drained for
one hour. The drainage, at first, looked like black water, then some
oil appeared before a yellow-orange, frozen material was drained The
mixture was sampled for PCB analysis. The EPT top was then opened and
one gallon of Coleman brand fuel, which is a naptha-based product was
added. The EPT sample valve was opened and a sample of the fuel wash
was taken. Then 18 gallons of fuel,were added to the EPT. The
IB-gallon fuel wash was released from the EPT and mixed with the
initial wash for sampling and PCB analysis. Subsequent sampling of
Pass 3 extract showed that not all of the fuel wash had drained from
the EPT, since biphasic characteristics were observed in the Pass 3
extract. Laboratory analyses of the fuel wash and column reboiler
37
-------
drainage showed that the unit had been accumulating PCBs in the
extract product tank and still bottoms tank. Most of the PCBs fed to
the unit apparently were not being transferred from the reboiler to
the extract tank with each pass. Therefore, closure of a PCB mass
balance for each pass based on feed, raffinate, and extract samples is
not possible. Similarly, a total mass balance closure for each test
is not possible.
Wash Water - During Pass 2 the hose of the feed pump was placed in a
wash water drum and 20 to 30 gallons of wash water were accidentally
pumped into the feed kettle. The wash water contained residue from
previous samples of treated sediment and feed material. The wash
water was mixed with Pass 1 treated sediment and the mixture was
sampled as feed to Pass 2. Laboratory analyses showed that, this
incident had little effect on the demonstration. The Pass 1 treated
sediment and Pass 2 feed PCB analyses differed by less than 35
percent.
Test 4
Test 4 was conducted, as planned, on a high PCB concentration (2575 ppm)
sediment. Six passes were run in order to ensure that the final treated
sediment PCB concentration would be equal to or less than the Tests 2 and 3
feed concentrations. Treatment beyond those concentrations would be inferred
by Test 2 and 3 results. Test 4 feed was a composite of sediments taken from
COE drums nos. 1-11 (one-third) and H-22 (two-thirds). Samples were taken at
the following locations:
Cartridge filter
Basket strainer
Feed kettle
Treated sediment product tank
Extract product tank.
The cartridge filter was sampled at the end of the test. The basket
strainer and treated sediment were sampled at each of the six passes. Extract
was sampled at Passes 4 and 6. Treated sediment foaming problems persisted
throughout the test and extract volumes continued to be small. Additional
samples were shipped as part of the effort to compare analytical results from
analysis methods 8080 and 680. Field measurements included pH, volume,
weight, and viscosity.
38
-------
Test 4 also included air sampling and analysis to determine the release
of PCBs from sample collection containers to the atmosphere. The potential
release of PCBs to the air from the low temperature process was assumed to be
insignificant based on the very low vapor.pressure of PCBs. Personal air
sampling pumps were mounted on the side of the collection containers. During
mixing of collection container contents, the container was covered with an
aluminum foil and gases evolving from the container were pumped with the
personal air sampler through a sorbent tube. This procedure was conducted
during the degassing of Passes 1, 2, 3, arid 4 treated sediment and Pass 4
extract. Sorbent tubes were shipped for PCB analysis.
Decontamination
System decontamination was conducted with toluene used as the feed.
About 100 gallons of toluene were added to the feed kettle, then collected in
the treated sediment and extract product tanks. The developer defined a feed
volume of 30 gallons, or 10 percent of the extraction system capacity, as
sufficient to run a pass through the extractor. Therefore, the equivalent of
three single passes of fresh toluene was run through the system (100 gallons/
30 gallons per pass = 3.3 passes). Toluene wash was collected from the
primary and overflow treated sediment tanks. Extract was collected in two
batches and contained mostly toluene. All toluene wash samples were analyzed
for the purpose of determining the mass of. PCBs that had accumulated in the
treated sediment and extract product tanks and related equipment. At the end
of tank draining, grab samples were taken at the drain valves to establish
that the unit had been decontaminated.
The demonstration staging area shown in Figure 5-1 was not contaminated
with PCBs that may have been released during testing and sampling. On August
11, 1988, soil samples were taken from 10 locations in the staging area.
Samples of soil were taken at zero to six inches of depth. Analytical results
showed PCB levels to range from less than 5 to 37 parts per million in the
soil. The 10 locations were resampled on October 6, 1988, after CF Systems
removed their equipment from the site and SAIC removed debris from the site.
Analytical results showed PCB levels to range from less than 0.5 to 5.4 parts
per million in the soil. At nine of ten locations, post-demonstration values
39
-------
were less than predemonstration values. This difference may be the result of
variability associated with sampling and analysis.
Debris was collected in 55-gallon drums over the course of the
demonstration. The following materials were produced during the demonstration
and removed from the staging area by Clean Harbors, Inc.
Number of Drums
6
6
2
15
8
20
8
22
Total = 87
Drum Contents
Toluene (unit decontamination residue)
Toluene rinsewater
Naphtha-based fuel product and unit residue
Sediments
Sediments and water
Decontamination water
Tyvek suits and water
Clothing and gloves
In addition to these, 14 drums containing harbor sediments obtained from
previous COE activities were also removed.
40
-------
SECTION 6. SAMPLING AND 'ANALYTICAL PROGRAM
6.1 SAMPLING LOCATIONS
A detailed description of major process equipment and the process flow
diagram were presented in Section 3. As discussed, feed sediment is treated
for organics removal, then the treated sediment is discharged. PCBs and other
organics separated from the feed sediments are discharged as extract. Through
a vapor recompression cycle solvent is reused. Minor solvent losses occur due
to venting. Vented emissions pass through a carbon canister after the uni't is
shut down. Other material losses occur as a result of the use of particulate
filters that are used throughout the system to protect operating equipment.
Materials also exit the unit during system shutdown and cleanout. These
materials are residues that have adhered to internal surfaces of tanks,
piping, and equipment and that are removed by use of a washing medium, such as
toluene or another nonpolar solvent. Figure 6-1 is a simplified diagram of
system inputs and outputs. Each system influent, effluent, or loss is defined
and described below.
Feed Sediments
Carbon Canister
Treatment System
1
Treated
Sediments
I
Extract
Figure 6-1. System Flow Diagram
41
-------
Feed Kettle (FK) -- The feed kettle holds a maximum of approximately 80
gallons of slurried sediments. Contents were continuously agitated to provide
a fairly homogenous feed. Feed drum contents were continuously agitated and
pumped into the feed kettle. Grab samples were taken, during agitation, with
a stainless steel ladle. Feed weight was recorded for Tests 2 through 5.
Analyses were conducted during Tests 2 through 4 for PCBs, semivolatiles,
cadmium, chromium, copper, zinc, lead, oil and grease, total solids, pH, and
viscosity. Only PCBs were analyzed during Test 1. Test 5 used only toluene
as feed for purposes of decontamination.
Treated Sediment Product Tank (RPT) The treated sediment, or raffinate,
product tank was accessible at its outlet, a point underneath the trailer.
During the demonstration, it was drained and product was collected in either a
55-gallon or an 85-gallon drum. At the conclusion of Tests 2 through 5, the
following were determined; PCBs, semivolatiles, cadmium, chromium, copper,
zinc, lead, oil and grease, total solids, and pH. Only PCBs were analyzed
during Test 1. During Test 4, gases evolving from the sampling containers
were sampled using National Institute of Occupational Safety and Health
(NIOSH) physical and chemical analytical method 253. Sorbent tubes were used
to collect analytes contained in gases emitted from the unit. Decontamination
residue was also collected at this point.
Extract Product Tank (EPT) -- The extract product tank was accessible at its
outlet, a point underneath the trailer. Extract drained and collected in a
covered 5-gallon, stainless steel pail during the demonstration. At the
conclusion of Tests 2 through 5, the following were determined: PCBs,
semivolatiles, cadmium, chromium, copper, zinc, lead, oil and grease, total
solids, and pH. Only PCBs were analyzed during Test 1. During Test 4, gases
evolving from the sampling containers were sampled using National Institute of
Occupational Safety and Health (NIOSH) physical and chemical analytical method
253. Sorbent tubes were used to collect analytes contained in gases emitted
from the unit. Decontamination residue was also collected at this point.
Basket Strainer (S-l) The basket strainer prevented oversized feed material
from entering the system. For each pass during Tests 2 (except Pass 9)
through 4, CF Systems removed the basket strainer. Test 2, Pass 9, was not
42
-------
sampled because sample volume was not sufficient. Solids were scraped with a
stainless steel spoon or spatula from the basket strainer and drained the
strainer casing into a stainless steel pail for a weight determination.
Solids were shipped for PCB and total solids analyses.
Cartridge Filter (F-2) - The cartridge filter removed fine particles in the
submicron range from the solvent-organics 'mixture. CF Systems installed
cartridge filters/and at the conclusion of each pass removed the filters and
scraped solids for a mass determination.
Carbon Canister (CC) - Propane blowndown from the PCU at the conclusion of
the tests and the decontamination procedure was vented through a 20-gallon
carbon canister. Following decontamination, CF Systems removed the carbon
canister from the treatment system. The carbon canister was emptied with a
shovel. At 16 equally spaced distances along the canister height, aliquots
were obtained with a stainless steel beaker. The beaker was used to mix the
surface of the carbon bed as each aliquot was obtained. The 16 aliquots were
composited in a stainless steel pail and then mixed prior to filling sample
containers. Sample weight, total solids, and PCB concentration were
determined.
Additional planned sampling was conducted as part of the Health and
Safety Plan to ensure worker safety and to determine if releases to the
environment from process equipment, open tanks, and sample collection
containers occurred.
6.2 SAMPLING SCHEDULE
The sampling frequency scheme planned for Tests 2,3, and 4 is shown in
Table 6-1. Sampling was designed to allow determination of: (1) a PCB
extraction efficiency with each pass and (2) a mass balance with each pass.
Sampling planned for decontamination was similar to that planned for Tests 2,
3, and 4, except that no chemical analyses were conducted for the feed that
was composed of toluene.
43
-------
TABLE 6-1. SAMPLES FOR CF SYSTEMS NEW BEDFORD
TESTS 2, 3, AND 4
Sample Location
Sample
Frequency(l)
Cartridge Filter
Basket Strainer
Feed Kettle
Final Pass
Each Pass
First Pass
Treated Sediment Product Tank Each Pass
Final Pass
Extract Product Tank
Each Pass
Final Pass
Parameter
PCBs
Total Solids
PCBs
Total Solids
Volume and Weight
PCBs
Semivolatiles
Cd, Cr, Cu, Zn, Pb
Oil and Grease
Total Solids, pH
Viscosity
PCBs
Volume and Weight
PCBs
Volume and Weight
Semivolatiles
Cd, Cr, Cu, Zn, Pb
Oil and Grease
Total Solids, pH
Volume and Weight.
PCBs
Total Solids
PCBs
Volume and Weight
Semivolatiles
Cd, Cr, Cu, Zn, Pb
Oil and Grease
Total Solids, pH
Note 1- Test 2 involved 10 passes, Test 3 involved 6 passes, and Test 3
involved 3 passes through the PCU.
44
-------
e
a
0
6.3 ANALYTICAL METHODS AND PHYSICAL TESTS
The analytical methods selected for the demonstration at New Bedford
Harbor included:
PCB by EPA Method 3550/8080 (EPA 1986)
PCB by EPA Method 3550/680 (EPA 1983, EPA 1986)
* 1984)Xtra°ti0n Modificat1on by Spittler Screen (Fowler 1987, Spittler
Waste Dilution by EPA Method 3580; (EPA 1986)
Semivolatiles by EPA Method 3550/8270 (EPA 1986)
Trace Metals by EPA Method 3050/6010 (EPA 1986)
Particle Size by ASTM Method C-136-84A (ASTM 1984)
Total Recoverable Oil and Grease by EPA Method 9071 (EPA 1986)
Percent Solids by EPA Method 160.3 (EPA 1983)
EP Toxicity (metals) by EPA Method 1310/6010 (EPA 1986)
pH in Calcareous and Noncalcareous Soils by EPA Method 9045 (EPA 1986)
Total Metals by EPA Method 3010 (EPA 1986).
All methods were intended to be performed as published, in the event
that modifications to the analytical program were required due to matrix
constraints, only approved methods were substituted. Any deviations from
approved protocol for these methods are discussed in Section 7. The Spittler
Screen and a comparison between Analytical Methods 8080 and 680 are discussed
below.
Spittler Screen
The Spittler Screen was used to identify the approximate level of PCBs in
a sample to aid analysts in choosing dilution factors (Spittler, 1984)
Subsequent analyses by EPA Methods were then carried out more efficiently.
The Spittler Screen provided rapid analytical turnaround for large volumes of
samples to optimize those subject to further analyses. Rapid turnaround of
analytical results allowed CF systems to more closely control operating
45
-------
conditions. This screening procedure was not be used for any critical
measurements.
Comparison of Analytical Methods 8080 and 680 for PCBs
Method 8080 is a Resource Conservation and Recovery Act (RCRA) analysis
method for determining PCBs as Aroclors. The analysis is conducted by gas
chromatography with an electron capture detector (GC/ECD), and quantitation is
performed by external standard calibration against one or more Aroclor stan-
dards. Pesticide compounds are used as surrogate standards for estimating
analytical accuracy.
Method 680 is an analysis method for determining PCBs as congeners
(homologous classes based on degree of chlorination) under the Clean Water Act
(CWA). The analysis is conducted by gas chromatography with detection by mass
spectroscopy (GC/MS), and quantitation is-performed by internal standard
calibration. Carbon 13-labeled PCB compounds (one from each congener) are
used as surrogate standards for estimating analytical accuracy.
Method 8080 was favored as the analysis method for the SITE demonstration
at New Bedford Harbor because:
o Method 8080 is a RCRA method and Method 680 is not
o Procedures for extraction of PCBs from sediments have been developed
and validlted for Method 8080, while Method 680 only addresses water
samples
o Because the sediments from New Bedford Harbor were known to contain
PCBs, and the PCB content has been thoroughly character zed by COE,
the added specificity of GC/MS (Method 680) was not required
o Method 8080 costs less than Method 680.
Method 8080 was finally chosen after a review of the analytical results
for Test 4 samples. In Table 6-2, Method 680 analytical results are shown for
Test 4 feed and Pass 4 treated sediment samples, and are reported as congener
groups rather than the 209 individual congeners. The percent of PCB removed
from the feed after the fourth pass is calculated for each congener group.
The removal data show that the extraction technology removes less chlorinated
46
-------
TABLE 6-2. METHOD 680 ANALYTICAL RESULTS FOR TEST 4
PCB Congener Feed
Group (ppm)
Pass 4
Treated
Sediment
(ppm)
Removal
(percent)
Congener Group
as a Percentage
of the Total Feed
(percent)
Mono-
Di-
Tri-
Tetra-
Penta
Hexa-
Hepta-
TOTAL
39
1,150
2,800
3,000
1,400
260
Not detected
8,700
0.58
30
98
130
69
18
12
350
99
97
97
96
95
93
Not applicable
96
Less than 1
13
32
34
16
3
Less than 1
Not applicable
47
-------
PCBs (e.g., mono- and dl-) more efficiently than more chlorinated PCBs (e.g.,
penta- and hexa-). However, this difference between congener groups is not
significant since all calculated removals differ from the mean removal by less
than five percent. Furthermore, the monochlorobiphenyl group composes less
than one percent of the total mass. Method 8080 was used to measure Arodors
1242 and 1254, since these represented nearly 70 percent of total PCBs
contained in the sediments (COE, 1987).
6.4 PROCESS CONTROL AND FIELD MEASUREMENT DEVICES
Process control and field measurement devices used during the demonstra-
tion are listed in Table 6-3. Also shown in Table 6-3 are measurement
locations, rationale for measurement, measurement frequency, measurement
units, and precision and accuracy. The developer's extractor pressure gages
were calibrated against nitrogen gas regulated by an Airco model 0-4000
pressure regulator/gage at 145 and 261 psig. The Airco pressure gage was
subsequently calibrated by Calibration Central, Inc., of Herndon, Virginia.
The extractor temperature gage was a Sybron Corporation, Bi-Therm model. The
feed slurry and cooling water temperature gages were manufactured by Universal
Enterprises. The temperature gages were calibrated in an equilibrated mixture
of ice and water at 32 degrees F and electrically heated water at 100 degrees
F. Feed kettle slurry volume was estimated using a calibration curve
developed by CF Systems. A Micro Motion, high-pressure Model D Meter measured
solvent flow. The mass flowmeter employed magnetic fields in its operation
and was calibrated electronically. Cooling water flow from the system was
measured via the classic stopwatch-and-bucket method.
Feed slurry viscosity was measured with a Haake Model Vt~02 cup-and-
spindle viscometer. Feed slurry pH was measured by a Corning Model pH 106 pH
electrode. Feed, treated sediment, extract, strainer solids, filter residues,
and other process materials were weighed on one of two scales. A Howe-
Richardson HCR international XL 1000 Ibs. x 1/2 Ib. scale was used to weigh
materials greater than 25 Ibs., and a Hanson 25 Ibs. x 1 oz. General Household
Scale was used to weigh materials less than 25 Ibs. The electric power was
metered by standard residential-use unit supplied by the local electric
utility.
48
-------
TABLE 6-3. PROCESS CONTROL AND FIELD MEASUREMENTS
Critical
Measurement
PROCESS CONTROL
MEASUREMENTS
Extractor
Pressure
Extractor
Pressure
Extractor
Temperature
Feed
Slurry Volume
Feed Slurry
Temp.
Solvent Flow
Cooling Water
Viscosity
Feed Slurry pH
Measurement
Location
E-l Press.
Gage
E-2 Press.
Gage
E-l Temp.
Gage
FK
FK
Control
Room Readout
Outlet Line
Feed Drum
Immediately
Prior to Use
FK
Rationale for
Measurement
Verify Operating
Conditions
Verify Operating
Conditions
Verify Operating
Conditions
Verify Operating
Conditions
Verify Operating
Conditions
Verify Operating
Conditions
Verify Operating
Conditions
Verify Operating
Conditions
Verify Operating
Conditions
Measurement
Frequency
10 min.
10 min.
10 min.
10 min.
Start of
Pass
10 min.
10 min.
Start of
Test
Start of
Pass
Measurement
Units Precision Accuracy
PSig 1% (1) 5% (2)
Psig 1% (1) 5% (2)
degrees F 0% (i) n% (2)
gal/min. (3) (4)
degrees F 5% (1) 2% (2)
Ibs/min. (3) o.2% (5)
GPM (3) 0.25% (2)
dPa.S (3) 5% (2)
PH 0.01 units (6) 0.01 units
-------
TABLE 6-3. PROCESS CONTROL AND FIELD MEASUREMENTS (Continued)
Critical
Measurement
MASS INVENTORY
MEASUREMENTS
Feed, Treated
Sediment Weight
Extract Weight
o strainer
Filter Residues
UTILITIES INPUT
MEASUREMENTS
Electric Power
Cooling Water
Temperature
Measurement
Location
Feed Weight
Scale
Low
Weight Scale
Low Weight
Scale
Low Weight
Scale
Input Line
to Pilot Unit
Input and
Output Streams
Rationale for
Measurement
Mass Inventory
Mass Inventory
Mass Inventory
Mass Inventory
Utilities
Measurement
Utilities
Measurement
Measurement
Frequency
Start of
Pass
End of Test
End of Pass
End of Test
During Pass
10 min.
Measurement
Units Precision
Ib 1% (1)
Ib 1% (1)
Ib 1% (1)
Ib 1% (l)
kw-hr (6)
degrees F 5% (1)
Accuracy
1% (2)
1% (2)
1% (2)
1% (2)
(5)
2% (2)
NOTES: (1) Precision as Relative Standard Deviation (RSD),
(2) Accuracy as Relative Percent Difference (RPD).
(3) Precision data not available.
(4) Accuracy data not available.
(5) Accuracy as reported by manufacturer.
(6) Precision as reported by manufacturer.
-------
SECTION 7
RESULTS AND DISCUSSION
7.1 SYSTEM PERFORMANCE
The evaluation criteria established in Section 1.3 for an evaluation of
system performance were:
* PCB concentration in sediments before and after treatment
PCB extraction efficiency with each pass of sediments through the PCU
t Mass balances established for total mass, solids, and PCBs.
These criteria are discussed with respect to analytical results below.
PCB Concentration Reductions
PCB analyses for feed sediments and treated sediment, conducted for
samples collected at each pass, are shown in Table 7.1. The data are
displayed graphically in Figures 7.1, 7.2, and 7.3. The data show that
treated sediment concentrations of 8 ppm are achievable and that as much as 84
percent of the PCB contained in sediment can be removed in a single pass, in
Test 2, feed containing 350 ppm of PCB was reduced to 8 ppm after 9 passes
through the PCU. in Test 3, a 288 ppm feed was reduced to 47 ppm after just
one pass. In Test 4, a 2,575 ppm feed was reduced to 200 ppm after 6 passes.
The percent reductions in PCB concentration, based on a comparison of
untreated feed to the final pass, for each test were:
Test
2
3
4
Percent Reduction
in PCB Concentration
72%
92%
Number of
Passes
10
3
6
The data for each test show general reduction trends based on differences
between initial feed and final treated sediment concentrations. However,
these trends are not consistent on a pass-by-pass basis. For example, PCB
concentrations in treated sediments increase at Test 2, passes 4 and 10, and
51
-------
TABLE 7-1. PASS-BY-PASS PCB CONCENTRATIONS AND REDUCTION EFFICIENCIES
Test
Number
2
2
2
£^
2
o
-------
Mean PCB Concentration, (ppm)
Mean PCB Concentration, (ppm)
Mean PCB Concentration, (ppm)
u>
0
-
"°_
s*-
tu
a
o
3.
T3
D>
en
en
2 0)
O
"
«.
-
_^
o
08t§8§§8§8§8§
i i i i i t i i i i i i j^
"
*>"'""
i
i
i
-
-
_4
S X 4^
n5 3-
i1* 51 "
1" lo,
o" °
3
».
-
.
O-
°OOOOOOOOOOi
1 1 I i I 1 1 1 1 1 1
1
t
t
V
t
V
o
§
o
-
l».
-
(0 § .Ct-
+ 3
U 0
Q.^ Z 0>.
c .
s
o
3
00-
-
/"*
en -L ~+ ro ro co (
3 S O Ol O 01 O i
3 ° 0 0 0 0 0 <
1 1 1 1 I 1 h
^^* * *"
r"""""
/
j£.
/
/
V
\
1
jt
1
1
1
r
/
\
\
J
fl
3
s1
a
Sa
si
?5
8-
o
s.
o
s
-------
at Test 3, passes 2 and 3. These anomalies are not related to the extraction
process. Instead, they reflect cross contamination within system hardware or
limited analytical precision and accuracy. Since the treated sediment
collection tanks were under pressure, it was not possible to clean out
collection hardware and piping. Therefore, a pass-by-pass mass balance could
not be established.
Data for each test can be used to construct a curve that shows the
potential number of passes required to reduce PCBs in harbor sediments to
specific concentrations using the Pit Cleanup Unit (PCU). If data from Test
2, 3, and 4 are displayed side-by-side, such that similar concentrations
coincide, then a PCB reduction curve can be plotted. Data are displayed
below, side-by-side, so that similar concentrations overlap.
Pass-by-Pass PCB Concentrations
Test 4 Test 3 Test 2
2,575
1,000
990
670
325
240
200
288
47
72
82
350
77
52
20
66
59
41
36
29
8
40
Based on the presentation of the data in Figure 7-4, it can be construed that
harbor sediments containing 2,500 ppm of PCB could be reduced to 100 ppm after
6 passes through the PCU. A level less than 10 ppm may be achievable after 13
passes.
Extraction Efficiency
Pass-by-pass PCB concentration extraction efficiencies are shown in Table
7-1 and are calculated as PCB extracted divided by concentration at the
54
-------
2600
2400
2200
2000
1800
^^^^
f[ 1600
a.
c 1400
g
£ 1200
§
c 1000
£j
CO 800
2
600
400
200
0
i Legend
| - - D Testa
1 * Test3
\ - - m Test 4
\
\
\
\
\
\
\
\
V \^
\ -.
\
\
x .'','
X
---s^ * -
a """""IT- » ,
I I | T" ~H" ""
10
12
14
Pass Number
Figure 7.4. Potential Pit Cleanup Unit PCS Reduction
beginning of the pass (multiplied by 100 percent). For each test, the first
pass results in efficiencies greater than 60 percent. However, at later
passes efficiencies range from negative values to 72 percent. This wide range
is the result of cross-contamination of solids retained in the treated
sediment subsystem. . r ,
Data show that the system irregularly retained and discharged treated
sediments. For some passes, as much as 50 percent of the feed was retained in
the system. That feed was treated sediment that clung to internal piping and
tank surfaces. If discharged with a later pass, the combined discharge could
have a higher concentration than feed for the later pass. For example, assume
an extraction efficiency of 60 percent, a feed concentration of 350 ppm, and a
carry-over of solids from the first pass to the second pass of 25 percent.
Then, the treated sediment would contain 77 ppm, instead of 56 ppm if no cross
contamination occurred.
55
-------
The occurrence of cross contamination affects interpretation of each
test, but it does not invalidate the fact that treated sediment concentrations
as low as 8 ppm were produced. Furthermore, the decontamination procedure
showed that PCB, which accumulated in system hardware, was contained in the
extract subsystem, not the treated sediment subsystem.
Mass Balances
Total mass, total solids, and total mass of PCBs were determined for
various system inputs and outputs for the purpose of establishing a mass
balance. Figure 7-5 depicts the inventory sheet used to account for system
input and output. Input included feed material and water, although some feed
was lost to sampling, sieving, spills, and residuals remaining on the surface
of the feed drums. Outputs from the system included samples, spills, con-
tainer residuals, treated sediment, and residue collected on the basket
strainer and cartridge filter. The difference between input and output
resulted in either accumulations within the system or unaccounted-for
discharges of accumulated material from the system. Appendix A shows mass
Inventories for each test. The items listed in Appendix A correspond to the
depicted in Figure 7-5. Total mass and total solids are shown in the tables
as pounds. The mass of PCB is shown as grams.
The amount of material accumulated or the amount of accumulated material
discharged are shown for each pass in Table 7-2, "Accumulation or Losses."
This term was calculated as total feed minus output. The unit irregularly
retained and discharged material throughout Tests 2, 3, and 4. No correlation
could be established between the mass inventories and extraction efficiency
for each pass. The mass balances for PCB and total solids are discussed in
more detail below.
PCB Balance
Table 7-2 illustrates the fate of PCB on a pass-by-pass inventory basis.
The system accumulated 15.15 grams during Test 2, 6.71 grams during Test 3,
and 42.11 grams during Test 4. Only an approximate PCB balance is possible
for Test 1, since Test 1 was a shakedown test only. Approximately 21 grams of
56
-------
Inventory Sheet
Test pass
1. Feed Material
6. Water
2. Sampling
3. Strainer
4. Spills
5. Residuals
S.Treated Sediment Accumulations and Other Losses
Figure 7.5 Illustrative Inventory Sheet
57
-------
TABLE 7-2. MASS ACCUMULATION AND LOSS IN THE SYSTEM
Accumulation (Loss) in the
Test
2
o
£+
2
2
2
«*
2
2
2
2
3
w
3
4
4
~
4
4
*T
4
Pass
1
2
3
4
*T
5
6
7
8
9
10
Subtotal
1
2
3
Subtotal
i
L
2
3
4
5
6
Subtotal
Total Mass
(Pounds)
122
55
(25)
78
22
68
(51)
(7)
(16)
9
254
24
58
29
in
5
(83)
74
(80)
106
(53)
(31)
Total Solids
(Pounds)
39
6
(16)
32
(6)
3
(1)
(11)
(3)
(3)
"40"
(13)
6
6
T
10
(12)
9
4
6
(3)
IT
System
Total PCBs"
(Grams)
14.21
0.70
0.50
(0.22)
(0.07)
0.3
0.04
(0.07)
0.29
(0.54)
15 . 14
6.28
1.42
(0.99)
6 . 71
37.79
(5.25)
8.72
2.55
1.63
(3.33)
42 . 11
Note: Parentheses indicate a loss or discharge from the system.
58
-------
PCB accumulated within the system during Test 1. Thus, total accumulation
within the system from Test 1 through Test 4 was about 85 grams (where 84.96 =
15.14 + 6.71 + 42.11 .+ 21).
Appendix B shows mass balances for the naptha-based fuel and the toluene
wash, respectively. The fuel wash, which occurred immediately after the first
pass of Test 3, flushed 35 grams of PCB from the extract subsystem. Final
system decontamination with toluene wash delivered an additional 151 grams.
Total wash output was 35 plus 151, or 186 grams. Ideally, the amount of PCB
washed from the system should equal amount accumulated, or
Accumulation - Wash = 0
However, in this case,
85 grams - 186 grams = -101 grams
The amount of PCB washed from the system is shown above to be greater than the
amount fed, which raises the possibility that (1) sampling and analytical
errors occurred, or (2) the system was contaminated from a previous CF Systems
demonstration.
Quality control samples collected during the demonstration indicate the
possibility of sampling and analytical error. For example, laboratory
precision and accuracy criteria were 20 and 50 relative percent difference,
respectively. In addition, quadruplicate grab samples were collected of the
Test 3 feed, the Test 4 feed, and the Test 3 treated sediment and the RPD
calculated for each set ranged from 12 to 47 percent. In particular, the Test
4 feed had a mean concentration of 2,575 ppm, which dominates all other
measurements used in the balance, and it had' an RPD of 22. Another possible
source of the PCB imbalance was contamination of the PCU from prior use at
another site. CF Systems did not decontaminate the unit with toluene prior to
this demonstration. CF Systems' standard operating procedures now incorporate
decontamination with toluene.
59
-------
In spite of the calculated PCB imbalance, a positive separation of PCB
from the harbor sediments was accomplished. The mass balances in Appendix A
show that 81 grams of PCB were contained in sediments fed to the PCU in Tests
2, 3, and 4. Resulting treated sediments contained 4 grams of PCB, which
indicates a mass removal efficiency of 95 percent. Decontamination residue
data (Appendix B) show that some PCB accumulated in system hardware. However,
91 percent of the PCBs contained in decontamination residues were contained in
the extract subsystem. The remaining 9 percent was contained in the treated
sediment subsystem hardware.
Basket Strainer, Cartridge Filter, and Carbon Canister
The basket strainer and cartridge filter, which generate residuals that
are normally discarded as a waste stream separate from extract and raffinate,
did not accumulate a significant PCB mass. The mass balances, shown in
Appendix A, show that the accumulation was approximately 2 percent of the PCB
mass fed to the system. When compared to PCB removals of 90 percent, this
Indicates that PCB removal by the basket strainer was not significant. In
addition, chemical analysis of the PCB content of filtered solids indicate
that the concentration of filtered solids associated with each pass roughly
correlated with the treated sediments from the previous pass.
Low pressure propane/butane was vented through the PCU carbon canister at
the conclusion of the decontamination procedures. The 285 pounds of activated
carbon contained in the canister collected less than 1 gram of PCB. This
indicates that air emissions are not significant and PCBs are separated from
the solvent when expanded in the PCU.
Total Mass of Solids
The PCU retained and discharged feed material intermittantly throughout
the tests. This behavior is demonstrated by tracking the sediment solids.
Figures 7-6, 7-7, and 7-8 show the pass-by-pass throughput of solids for Tests
2, 3, and 4. The mass of solids accumulated on a pass-by-pass basis is
significant. The flow of solids per pass ranges from 55 percent accumulated
to 150 percent discharged. There is no consistent correlation between solids
retention and PCB concentration reduction.
60
-------
Percent Throughput, Output/Input
Percent Throughput, Output/Input
Percent Throughput, Output/Input
ro
m
&
3
a
o
.3
TJ
M
(A
O
o -
CO SO O -l
o
3
tn
Z
03 -
o -
oo o o o oo o oo oooooo
I1 I I I i I
£
3
o_
o'
3
03
W
W
w-
tndqBnojm
indu.Bnoju.i spnos
ssej JOd indijBnojm sp
-------
During Tests 2, 3, 4, and 5 the system accumulated 302 pounds total mass
and 53 pounds total solids. Total mass accumulation represents approximately
4 percent of total mass fed to system during Tests 2 through 5, and total
solids accumulation represents about 7 percent of total solids fed to the
system.
A total of 3-1/2 tons of solids and water were fed to the unit, over the
course of 19 passes throughout Test 2, 3, and 4. Of the total, 96 percent was
accounted for in the system outputs. Of 789 pounds of solids fed to the
system, 93 percent was accounted for in system outputs.
Other Data
Semi volatile Organics
System feed, final treated sediment, and final extract were sampled for
base/neutral and acid extractable organics (semivolatiles) during each test
for the purpose of (1) characterizing materials for disposal and (2) observing
any extraction of semivolatiles. Interpretation of the semivolatiles data,
shown in Volume II, is limited for two reasons: (1) the unit contained
sludges from a previous demonstration at a petroleum refinery, and (2) a
naphtha-based fuel product was added to the unit during Test 3 to clean out
the still, extract product tank and related hardware. The following
conclusions can be drawn:
Semivolatiles detected in the toluene wash were also detected in the
feed drums, the source being New Bedford Harbor sludge.
Phenol and 2-methylphenol were found in treated sediments and extracts
but were not measured in feed drums, the feed kettle, or toluene
washes.
Test 4 resulted in a reduction of 1,4-dichlorobenzene and pyrene, but
chrysene and bis(2-ethylhexyl phthalate) were increased. Similar
inconsistencies occur for Test 2 and 3.
2-Chlorophenol, 1,3-dichlorobenzene, and benzo(k)fluoranthene were fed
to the unit but not detected in any system effluents.
62
-------
Fate of Metals
A firm conclusion cannot be drawn concerning the fate of metals after
each test, since the unit tends to accumulate solids. However, the data in
Table 7-3 show that treated sediments metals concentrations generally equal or
exceed feed metals concentrations. The data also show that metals were not
extracted and discharged in the organics effluent. Metals concentrations in
organic extracts were one to two orders of, .magnitude less than treated
sediments.
EP Toxicity
RCRA regulations at 40 CFR 261.24 specify test methods for determining if
a solid waste exhibits the characteristic of EP (extraction procedure)
toxicity. The maximum concentration of contaminants for the characteristic of
EP Toxicity is shown in Table 7-4. Also shown are analytical results for
(1) two samples taken from a composite of drummed harbor sediment collected by
COE during the waste presampling and (2) a :sample of demonstration Test 4
Pass 6 treated sediment. Concentrations for each sample shown are less than
the regulatory maximum for the definition of the EP toxicity characteristic.
7.2 OPERATING CONDITIONS
The system specifications that CF Systems requires for normal operation
were discussed in Section 3. In this section, observed operating conditions
are summarized and operating data are interpreted with respect to treatment
efficiency. In tables throughout this section, mean operating data are shown
as well as the range of data recorded for each mean value. Generally the
technology accommodated wide ranges of operating conditions, although'precise
operational control was limited since all controls were manual rather than
automatic.
Extraction Pressure
Pressures in both extractors used in the system were fairly stable for
all tests. Pressure levels were close to the nominal level of 240 psig The
maximum pressure, 285 psig, was below the 300 psig maximum specification The
mmmum pressure, 190 psig, was above the 180 psig minimum specification
63
-------
TOBLE 7-3. tT«JS CONTENT OF FEED, TREATED SEDIMENT, HO EXTRACT
Parameter Units
Cadmium, ppm
Chromium, ppm
Copper, ppm
Lead, ppm
Zinc, ppm
Total Residue, %
lest 2
Feed
35.7
596
1790
619
2150
23.3
Test 2
PassS
Treated
Sediment
32.5
581
1£50
537
2220
1S.2
Test 2
Pass 4
Feed
44.0
761
1S30
792
2680
15.0
Test 2
Pass ID
Treated
Sediment
42.8
816
1740
892
2610
9.4
Test 2
Pass ID
Extract
NR(1)
3
5(2)
NR(1)
5(2)
NR(3)
Tests
Feed
32.0
525
1320
520
1900
19.4
.11 ii *
Tests
PassS
Treated
Sediment
62.3
1020
2570
1100
3550
10.3
TestS
Pass 3
Extract
6(2)
20
6(2)
NR(1)
8(2)
NR(3)
Test 4
Feed
87.5
1480
2650
1300
5370
16.4
Test 4
Pass 6
Treated
Sediment
120.0
1790
3700
1SCO
7260
5.6
Test 4
Pass 4
Extract
5
26
5
35
15
NR(3)
Test 4
Pass 6
Extract
5
31
4
40
15
NR(3)
Notes: 1. Not reported, severe matrix effects.
2. Matrix effects indicated.
3. Not reported, insufficient sanple volume for analysis method.
-------
TABLE 7-4. EP TOXICITY CHARACTERISTIC OF TREATED AND UNTREATED SEDIMENTS
UNITS (PARTS PER MILLION)
Composite Sample of
Waste Presampling Drums
Sample 1 Sample 2
Treated
Sediment
Test 4, Pass 6
Maximum Concentration
Allowable for
Characteristics of
EP Toxicity
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
..
0.011
0.16
0.11
0.18
0.34
<0.0002
<0.005
<0.015
_
0.008
0.15
0.12
0.098
0.23
<0.0002
<0.005
<0.015
_______
<0.005
0.36
0.30
0.053
0.16
<0.0002
<0.02
0.015
5 0
100 0
1 0
5 0
5 0
0 2
1 0
5.0
Note: < indicates detected less than the detection limit shown.
65
-------
Because pressures were so stable, no relationship between extraction
efficiency and extractor pressure was apparent.
Feed and Extraction Temperature
Feed and extraction temperatures were stable for Tests 3 and 4. Feed
temperatures ranged between 60 and 70 degrees F while extraction temperatures
ranged between 60 and 80 degrees F. However, data for Test 2 indicate that
feed temperatures fell about 15 degrees F below the minimum specification
after pass 5. This caused extraction temperatures to drop, with pass 9
falling 4 degrees F below the minimum specification, 60 degrees F.
The developer attributes much of the fluctuating extraction efficiencies
calculated for Test 2 to the low feed temperatures, although other factors
were probably important. These factors include cross contamination in the
treated sediments collection tank. In addition, reentrainment of solvent in
decanter underflows may have caused disproportionately large effects on low
concentration sediments. Each factor must be addressed by the developer in
the design of a full-scale system.
Feed Flow Rate
The feed flow rate ranged consistently, throughout the tests, from 0.6 to
1.4 gpm. This range compares well with the 0.2 gpm minimum specification and
the 1.5 gpm maximum specification.
Solvent Flow Rate
The solvent flow fluctuated outside the minimum specification, 8 Ib/min,
and the maximum specification, 15 Ib/min throughout Tests 2, 3, and 4 as shown
1n Figures 7-9 through 7-14. Because of this wide variation, it was suspected
the flow meter was malfunctioning. In Test 4, an alternative measuring device
was used and flow measurements continued to show wide variations, as seen in
Figures 7-13 and 7-14.
The variable solvent flows caused the solvent/feed ratio also to
fluctuate widely. This ratio was calculated as solvent (lb/min)/feed
(gpD/feed density (Ib/gal). The minimum solvent-to-feed ratio specification,
66
-------
Figure 7-9
Mean Solvent Flowrate
Test 2
Figure 7-10
Mean Solvent/Feed Ratio
-iIiII i i i i i
5 1.0
23456789 10
Extraction Pass Number
Figure 7-12
Mean Solvent/Feed Ratio
" Tests
Norn
0123
Extraction Pass Number
Notes: Flowmeter manufacturer reports 0.2% accuracy.
Mean reported with range of recorded data.
2.5
2.0
0.5
0.0
I-
1
1
1
Tes
it 2
-
i
1
Max
0123456789 10
Extraction Pass Number
Figure 7-13
Mean Solvent Flowrate Plus
22
20
18
16
| 14
*£ 12
E
I 10
8
6
4
2
0
Test 4
-
-
-
-
i
-
I
12345
Extraction Pass Number
x - Alternate Solvent Flowrate
Figure 7-11
Mean Solvent Flowrate
Tests
1 2 3
Extraction Pass Number
Figure 7-14
Mean Solvent/Feed Ratio Plus
2.5
I
o
1
0,5
0.0
Alternate solvent/Feed Ratio
Test 4
!
i
X
1
1
0123456
Extraction Pass Number
x . Alternate SorventfFeed Ratio
-------
1.0, was not met on Pass 2 of Test 4 based on mean data. Individual readings
frequently exceeded the 1.0 to 2.0 specification range. A pass-by-pass
comparison of solvent/feed ratios to extraction efficiencies was attempted but
no direct correlation or trend was apparent.
Nonetheless, it is believed that the solvent/feed ratio is a significant
factor in process design since the solubility of an organic in liquefied
propane-butane is the fundamental basis for the extraction. With higher
solvent/feed ratios, the feed is exposed to a larger amount of solvent and
extraction efficiency should increase. However, these relationships were not
observed, given the available data.
Feed Solids
Feed solids content steadily declined during each test as shown in
Figures 7-15 through 7-17. Initial feeds had solids contents ranging from 15
to 22 percent. Final treated sediments ranged from 6 to 11 percent solids.
This change is primarily a result of water added to the feed kettle by
operating personnel, during each pass. This unnecessary practice caused waste
volumes to increase by 33 percent over the course of the demonstration
program. Another, but less significant, factor that affected solids content
was accumulation of solids in system hardware. The solids mass balance showed
that 7 percent of the solids accumulated in the system and were not washed out
during decontamination.
Treated sediments that were fed to the unit after Pass 3 of each test,
had solids contents below the minimitn specification, 10 percent. This
dilution of the feed material is believed to affect system performance.
Viscosity and pH
Feed viscosity and pH fell within specifications and did not affect
system performance. Viscosities for untreated feed and recycled sediments
ranged from 20 to 170 centipoise, well below the 1,000 centipoise maximum
specification. This specification was set by the developer only to ensure
that the feed would be pumpable. Untreated and recycled sediments had pH
values that ranged between 7.3 and 8.5 standard units. This narrow band
68
-------
Figure 7.15 Figure 7 16
Feed/Treated Sediments Solids Concentration Feed/Treated Sediments Solids Concentration
Test #3
Test #2
Figure 7.17
Feed/Treated Sediments Solids Concentration
Test #4
X
CO
5
S 5*
3 C
» C
c
: . : 2
i
f
-
I
I 1 1 I I j
1
1
I"
1
1
l"
. :
1
S 8s
3 C
_ w
-
- co
_
o
Z
-*> 9
CO
Q.
C
,0
tj
« 2
4^
111
CM
O
>
>
X
CO
2
§ 1
u
T
1.
1
1
1
1 '
I
1
1 " .
j !
i i i 1
J s£ vp sP «^P v
d^ 0s 0s 0s- o
> O 0 0 O C
5 W * P5 CM T
1
I t
_. \^r
T
-co
o'
Z
- (D 0)
-------
fell within the 6 to 12 specification range. The developer established this
range to prevent corrosion to PCU hardware.
7.3 DEVELOPER'S GOALS
Since a feasibility study had not been completed for the New Bedford
Superfund site, the following treatment goals were set by the developer:
Test 1 - Shakedown test only.
Test 2 - Reduce PCB content of a 350 ppm feed by at least 90 percent
or to a concentration below 5 ppm after the tenth pass.
Test 3 - Reduce PCB content of a 288 ppm feed by at least 50 percent
or to a concentration below 50 ppm after the third pass.
. Test 4 - Reduce PCB content of a 2,575 ppm feed to the J^ed levels
observed in Test 2 and 3 after 6 passes. (288 ppm was the lower feed
of Tests 2 and 3).
. Test 5 - Decontamination process. Achieve effluent concentrations
less than 50 ppm in final toluene washes collected from the extract
and treated sediment subsystems.
Goals for Tests 3 and 4 were achieved. In Test 3, the Pass 3 treated
sediment PCB concentration was 72 percent less than the feed. Thus, the 50
percent reduction goal was met. In Test 4, the pass 6 treated sediment PCB
concentration was 200 ppm. Thus, the 288 ppm concentration goal was met. The
final treated sediment concentration, at pass 10 for Test 2, was 40 ppm, which
meets neither the 5 ppm concentration nor the 90 percent reduction goals.
However, a 20 ppm level was observed at pass 3 and 8 ppm was observed at pass
9. Results for Test 5 also fell slightly short of goals for decontamination
as discussed in Section 7.4.
These goals were achieved even though (1) operational problems such as
foaming and solids retention occurred and (2) the analysis method provided
results with a precision of plus or minus 20 percent (measured as relative
percent difference).
70
-------
7.4 HEALTH AND SAFETY MONITORING
During the demonstration of CF Systems' process unit, personnel were
potentially exposed to the contaminated harbor sediments. A monitoring pro-
gram was conducted to determine potential exposures and provide a basis for
selection of proper personal protective equipment. Several types of portable
monitoring equipment were used during the various phases of the field investi-
gations, including:
Portable Organic Vapor Analyzer (Century OVA)
Portable Photoionization Meter (HNu)
Combustable gas/oxygen/hydrogen sulfide meters (MSA and
Enmet-Tritector)
Detector tubes and sampling pump (Sensidyne-Gastec)
Personal air sampling pumps (Dupont-P200).
It was suspected that some level of organic vapors would be encountered,
particularly when drums containing contaminated sediments were first opened
during the feed preparation phase. Continuous monitoring using both the OVA
and HNu instruments was conducted while the drums were being opened, These
instruments detected a slight elevation above background levels of organic
vapor immediately upon opening the drums. The levels returned to background
levels within a few seconds. No measurable levels of hydrogen sulfide or
combustible gas were encountered while opening the drums or handling the sedi-
ments during the feed preparation phase
During the various test runs of the extraction unit at the New Bedford
site, organic vapors, PCBs, combustible gases, and hydrogen sulfide were
monitored. The OVA and HNu meters were used to monitor for organic vapors at
all work stations on the extraction unit, while CF Systems and SITE personnel
monitored process equipment. The OVA also was used as a survey meter on the
process equipment to search for possible fugitive emissions from the
equipment. All measurements indicated that organic vapor levels remained in
the range of background levels. Two portable combustible gas meters were used
to check for elevated levels of propane during the equipment shakedown period
and for spot testing during the demonstration. The pilot unit also contained
71
-------
two integral combustible gas detectors located on either end of the unit.
During the normal extraction process, combustible gas readings remained at
background levels. However, while treated sediment and extract samples were
collected, the combustible gas meters indicated that levels exceeding only 20
percent of the lower explosive limit for propane were encountered. These
episodes of elevated propane levels generally lasted for less than 60 seconds
and subsided rapidly depending on the length of time sampling occurred and the
strength of the wind at the time.
Sampling was conducted using personal sampling pumps and 150-mg charcoal
tubes and florosil tubes to determine personal exposures to organic vapors and
PCBs, respectively. All air sample results indicated that, if present,
organic vapors and PCB levels were present only at levels below the detection
limits for the analytical methods. No measurable levels of hydrogen sulfide
were detected using either detector tubes or portable monitoring devices.
Treated sediment and extract subsystems were decontaminated with toluene.
The final concentration of PCB contained in the treated sediment subsystem
toluene wash was 34 ppm, which was below the decontamination goal of 50 ppm.
The final concentration of PCB contained in extract subsystem toluene wash was
60 ppm, which slightly exceeded the decontamination goal of 50 ppm. Staging
area soils were not affected by any leaks or emissions that may have occurred
during the duration of the demonstration as discussed in Section 5.3.
7.5 EQUIPMENT AND MATERIAL HANDLING PROBLEMS
Equipment and material handling problems occurred throughout the dem-
onstration, as described in Section 5. While these problems did not impede
achievement of the developer's treatment goals, they could impact the economic
performance of a full-scale commercial system. Some problems were anticipated
since relatively small volumes of sediments were batch-fed to a unit that was
designed for continuous operation. The nominal capacity of the unit is 700
gallons per day, but only 50 to 100 gallons per day were batch-fed during
shakedown on tests 2, 3, and 4. Consequently, the unit irregularly discharged
and retained solids with each pass.
72
-------
Previous use of the unit affected interpretation of semivolatiles data
and may have contributed to imbalance of the, PCB inventory. Internal surfaces
of extract collection hardware collected PCBs as evidenced by mass balances.
In addition, Test 3 was interrupted and viscous oils were found accumulating
in extract subsystem hardware. PCBs are soluble in oil, which coated the
internal surfaces of system hardware. The amount of oil that can coat
internal piping and collection tanks could be significant. For example,
assume (1) a hardware surface area of 10 square meters, (2) a coating
thickness of 0.1 millimeters, and (3) an oil density of 1.0 grams/cubic
centimeter. This is equivalent to 100 grams of oils that cling to the
internal surfaces of extract subsystem hardware. As a result of this
demonstration, CF Systems now requires more rigorous decontamination
procedures for the PCU.
Solids were observed in extract samples that were expected to be solids-
free. This indicates poor performance or failure of the cartridge filter. An
alternative type of filter should be investigated by the developer.
Low-pressure dissolved propane and butane caused foaming to occur in the
treated sediment product tanks. This hindered sample collection and caused
frequent overflow of treated sediment to a secondary treated sediment product
tank. CF Systems states that design of a commercial-scale unit will allow
release of solvent entrained in the treated sediment and elimination of the
foaming problem.
7.6 DATA QUALITY ASSURANCE
All field sampling and laboratory analyses were accompanied by a program
of quality assurance/quality control (QA/QC) checks designed to assess the
validity of the sampling and analysis effort. This program was designed to
ensure that the samples are representative,, and that the analytical data
accurately describes the characteristics and concentrations of constituents in
the samples. This QA/QC program is outlined in the approved Quality Assurance
Project Plan (QAPP).
73
-------
Technical Systems Reviews
EPA conducted a technical systems review of the field activities during
the dress-rehearsal for the demonstration. This field audit verified that the
sampling procedures used would result in samples that met the requirements of
the QAPP, and resulted in a satisfactory rating, the highest rating
achievable.
EPA also conducted a technical systems review of each of the laboratories
Involved in the analysis of samples from the CF Systems Demonstration project,
examining the laboratory's procedures for sample storage, preparation,
analysis, and documentation. While some concerns were identified during each
of these reviews, these concerns were addressed to the satisfaction of the
reviewer before analyses were continued.
Data Review and Validation
In order to assess the validity of the measurement data, the QC data
generated with the environmental data were evaluated with respect to the
project quality assurance objectives defined in the QAPP, as well as the
specific quality control requirements of the analysis methods used. These
laboratory QC data are presented in Volume II. The evaluation of the QC data
was accomplished through review of:
Completeness of analytical reporting
Analytical holding times
Target analyte identification criteria
Tentatively identified compound (TIC) identification criteria
Calibration frequency and acceptance criteria
Presence and contamination of method blanks
i Surrogate compound recovery
Presence and conformance with acceptance criteria of matrix spike,
matrix spike duplicate, and duplicate analyses.
74
-------
Evaluation and Qualification of Measurement Data
The following sections present brief summaries of the findings of the
data review and validation, by analyte class.
Polychlorinated Biphenyls
Method 8080 Analyses
Fifty-three sediment samples were analyzed for PCBs from Tests 1 through
5, either as feed stock, treated sediments, or decontamination residues.
Ninety-five percent of the recoveries calculated from the matrix spike/matrix
spike duplicate (MS/MSD) sample pairs analyzed with these samples were within
the project quality assurance objectives defined in the QAPP indicating
acceptable accuracy for the analysis of these samples.
Forty-eight percent of the surrogate recoveries calculated from these
samples exceeded the acceptance criteria. Because the surrogate compound used
for these samples (dibutylchlorendate) is not a polychlorinated biphenyl
(PCB), the the matrix spike data are more Indicative of the accuracy of
analysis for PCBs in these samples.
Only twenty-one percent of the relative percent, differences (RPDs)
calculated between the MS/MSD sample pairs were within the defined aceptance
criteria, indicating that the established precision criteria (plus or minus 20
percent) were too ambitious. The average RPD calculated between MS/MSD pairs
was 47 percent. EPA analysis method 8080 QC acceptanqe criteria show RPDs
greater than 60 percent for PCB 1242 and PCB 1254 (EPA, 1986).
None of the method blanks associated with these samples were contaminated
with PCBs.
The results of analyses' of PCBs in extract samples were not used as part
of the demonstration evaluation and have nojt been included as part of this QA
evaluation. ,
Collectively, these data show that the;analysis of PCBs in sediments for
Tests 1 through 5 are sufficiently accurate and precise to allow for engineer-
ing assessment of the efficacy of this Demonstration.
75
-------
Method 680 Analyses
The recoveries of congeners spiked into samples analyzed by method 680
were outside of the interim acceptance criteria for monochlorobiphenyl through
tetrachlorobiphenyl isomer groups in one MS/MSD sample pair (Test 1). The RPDs
calculated for this sample pair were outside of the interim acceptance
criteria for mono- and dichlorobiphenyl. The surrogate recoveries for this
Test 1 analysis were also outside of the acceptance criteria for the
monochlorobiphenyl isomer group. In Test 4 analyses, the accuracy and
precision problems were not the case. Therefore, the data presented are
insufficient for evaluation of the accuracy and precision of analysis for the
less chlorinated congeners. Based on this information, the analytical data
for the more highly chlorinated isomers should, however, be of acceptable
accuracy and precision.
Spittler Analyses
The data from the Spittler screening analyses were used for decisions in
the field, but were not the basis for any conclusions drawn for the final
report. These data were, therefore, not included in the data review and
validation process.
Base/Neutral and Acid Extractable Compounds (BNAs)
Both sediment and oil samples were analyzed for BNAs. The recovery of one
or more compounds in the each of the MS/MSD samples analyzed with the sediment
samples were outside of the established acceptance criteria. In addition, each
of the sediment samples was spiked with surrogate compounds. The recovery of
these surrogate compounds were within acceptance criteria for 20 of the 26
samples. While these data are in apparent conflict, they cast doubt over the
acceptability of the analytical accuracy. The relative percent difference
(RPD) calculated between analyses of at least one compound in each of the
MS/MSD pairs exceeded the acceptance criteria specified in the analytical
method, indicating a generally unacceptable degree of analytical precision.
No MS/MSD samples were analyzed with the oil samples analyzed for BNAs,
therefore, the recovery of matrix spike compounds cannot be used to evaluate
the analytical accuracy. While the samples were spiked with surrogate
76
-------
compounds, the recoveries cannot be calculated in most of the samples because
of dilution effects. There is, therefore, an inadequate basis for evaluation
of analytical accuracy and precision.
The only target analytes detected in the method blanks associated with
these analyses are common plasticizers (phthalic acid esters), which may have
been introduced into the samples during field or laboratory activities.
The BNA data should be considered as qualitative, rather than
quantitative, based on conflicting or inadequate information available to
evaluate analytical accuracy, as well as poor precision of analysis.
Volatile Organic Analytes (VOAs)
The recoveries calculated from the M^S/MSD sample analyzed with the
sediment samples are within acceptance criteria, as are the recoveries
calculated from the surrogate compound spikes. These data indicate an
acceptable degree of accuracy for this noncritical measurement.
The RPDs calculated between the spike compounds in the MS/MSD pair are
within the acceptance criteria specified in the analysis method, indicating
acceptable precision for the analysis of VOAs in sediment samples.
The method blank associated with these samples show that the samples may
have been contaminated with low concentrations of methylene chloride, a common
laboratory contaminant. The data from analysis of sediment samples for VOAs
are not qualified in any way, and should be regarded as quantitative.
Metals analyses were performed on sediment grab samples, composite
samples and oil samples.
The recoveries calculated from one of the two MS/MSD sample pairs
analyzed with the sediment grab samples are outside of acceptance criteria for
many of the target analytes; however, those calculated from the other MS/MSD
pair are acceptable. These conflicting data do not provide sufficient
information for the evaluation of analytical accuracy. The RPDs calculated
77
-------
between one MS/MSD sample pair, and duplicate sample pairs, are within
acceptance criteria, indicating an acceptable level of analytical precision
for these analyses. The method blanks associated with these samples
demonstrate that the samples were not contaminated in the laboratory with the
target analytes. Because of the lack of information for evaluation of
accuracy, these data should be regarded as qualitative.
The recoveries calculated from the MS/MSD samples analyzed with the
composite sediment samples, and the RPDs calculated between the MS/MSD sample
pairs and duplicate sample pairs, indicate an acceptable level of analytical
accuracy and precision for these analyses. The method blanks associated with
these samples demonstrate that the samples were not contaminated in the
laboratory with the target analytes.
All oil samples were analyzed for metals by the method of standard
additions (MSA). Significant matrix interferences were reported for the
majority of these samples. The RPDs reported between the analysis of a
duplicate sample pair were within acceptance criteria for two of the five
analytes of interest. The,method blank associated with these samples
demonstrates that the samples were not contaminated in the laboratory with the
target analytes. Together, these data indicate that the results from metals
analyses in oil samples have an unacceptable level of accuracy and precision,
and should be regarded as qualitative.
In addition to the total metals analyses, sediment samples were analyzed
for the characteristic of Extraction Procedure Toxicity (EP Tox). The
recoveries calculated from the MS sample prepared with these samples are
within acceptance criteria for six of the eight target analytes. The RPDs
calculated between two duplicate sample pairs are within acceptance criteria
for some target analytes, outside of acceptance criteria for others, and
cannot be evaluated for the remaining analytes due to nondetectable
concentrations in the samples. The method blanks associated with these
samples demonstrate that the samples were not contaminated in the laboratory
with the target analytes. Together, these data indicate an acceptable degree
of accuracy and an unacceptable degree of precision for these analyses. The
data should, therefore, be regarded as qualitative.
78
-------
Total Residue
The recoveries calculated from control samples analyzed for total
residue, along with the sediment samples, are within acceptance criteria, as
are the RPDs calculated between 99 percent of the duplicate samples analyzed.
These data indicate acceptable accuracy and precision for these analyses.
The data from analysis of sediment samples for Total Residue are not
qualified in any way, and should be regarded as quantitative.
Total Oil and Grease
Twenty-five percent of the recoveries calculated from the analyses of
control samples for total oil and grease, and fifteen percent of the RPDs
calculated between replicate analyses, were outside of the acceptance
criteria. About one third of the method blanks associated with these samples
were found to have detectable concentrations of oil and grease, demonstrating
the potential for low.-level contamination of-the samples in the laboratory.
Cyanide
The recovery calculated from a control sample analyzed along with sedi-
ment samples for cyanide is within acceptance criteria. The recovery of
cyanide from a post-digestion spike sample was, however, outside of acceptance
criteria, providing conflicting information for the evaluation of analytical
accuracy. The RPD calculated from the analysis of a replicate pair is within
acceptance criteria, indicating acceptable analytical precision.
The data from analysis of sediment samples for cyanide should be regarded
as qualitative, based on conflicting evidence for the evaluation of accuracy.
Total Dissolved Solids
The recovery from a spiked control sample analyzed with water sample for
total dissolved solids is within acceptance criteria, indicating acceptable
accuracy. Because no replicate pairs of samples were analyzed for total
dissolved solids, the analytical precision of this noncritical measurement
cannot be evaluated.
79
-------
The IDS data should be considered as qualitative, rather than
quantitative, based on inadequate information to evaluate the analytical
precision.
Total Suspended Solids
No quality control check sample or replicate sample was analyzed with the
water samples analyzed for total suspended solids, therefore the analytical
accuracy and precision of this measurement cannot be evaluated.
The TSS data should be considered as qualitative, rather than
quantitative, based on inadequate information to evaluate the analytical
accuracy and precision.
PH
No quality control check samples were prepared for the measurement of pH
1n sediment samples, therefore the analytical accuracy of this measurement
cannot be evaluated. The RPD calculated between pH measurements made on a
replicate sample pair are within acceptance criteria; however, the frequency
of replicate pair analysis was insufficient.
The pH data should be considered as qualitative, rather than
quantitative, based on inadequate information to evaluate the analytical
accuracy and precision.
80
-------
SECTION 8
REFERENCES
ASTM, 1984. Standard Method for Sieve Analysis of Fine and Coarse Aggregates
ASTM 0-4059. American Society for Testing and Materials. Philadelphia PA
December 1984.
COE, 1987. Pilot Study of Dredging and Dredged Material Disposal Alterna-
tives. U.S. Army Corps of Engineers. New England Division. Waltham MA
September 1987.
Ebasco, 1987. Draft New Bedford Harbor Feasibility Study Test 22, Environ-
mental Evaluation Activity 22.4 - Selection of Additional Contaminants for
Inclusion in the Risk Assessment and Feasibility Study. Boston, MA. July
1987. ;,
Fowler, Bruce A. and Bennett, Joseph T., 1987. Screening for Characterization
of PCB-containing Soils and Sediments. Proceedings of the National Conference
on Hazardous Wastes and Hazardous Materials. Washington, DC. March 1987.
Spittler, T.M., 1984. "Field Measurement of Polychlorinated Biphenyls in Soil
and Sediment Using a Portable Gas Chromatograph," Environmental Sampling for
Hazardous-Wastes. American Chemical.Society. 1984. 37-42.
EPA, 1983. Methods for Chemical Analysis of Water and Wastes. U.S. Environ-
mental Protection Agency. Environmental Monitoring and Support Laboratory,
Cincinnati, OH. 600/4-79-020. Revised March 1983.
EPA, 1986. Test Methods for Evaluating Solid Waste. U.S. Environmental
Protection Agency. SW-846. U.S. Government Printing Office. Washington,
D.C. Third Edition. November 1986.
81
-------
APPENDIX A
MASS INVENTORIES FOR TESTS 2, 3, AND 4
82
-------
TOTAL MASS BALANCE (Lbs)
TEST #2
ITEM
PASS #
1
10
TOTALS
1 Feed Material
2 (Saving,
3 (Strainer)
4 (Spills)
5 (Residuals)
6 Water
7 Total Feed
| 483.50 |
| 0.00 |
I 4-63 |
| 0.00 |
| 88.00 |
| 120.12 |
| 510.99 |
367.15 |
3.09 |
0.00 |
0.00 |
2.88 |
41.75 |
402.93 |
328.46 | 351
0.00 | 1
0.00 | 0
0.00 | 0
1.33 | 2
20.57 | 41
347.70 | 389
.54 |
.13 |
.00 |
.00 |
.86 |
.88 |
.43 |
296.50 |
0.00 |
0.00 |
0.00 |
1.56 |
49.06 |
344.00 |
305.00 |
0.00 |
0.00 |
0.00 |
0.31 |
31.32 |
336.01 |
259.00 |
5.13 |
0.00 |
0.00 |
0.18 |
29.19 |
282.88 |
321.25 |
0.00 |
0.00 |
0.00 |
0.43 |
34.36 |
355.18 |
347.00 |
0.00 |
0.00 |
0.00 |
0.76 |
22.25 |
368.49 |
378.00
0.00
0.00
0.00
0.76
22.31 |
399.55 |
| 3437.40
| 9.35
| 4.63
| 0.00
| 99.07
| 412.81
3737.16
00
8 Raffinate |
9 Sampling |
10 Spills |
11 Residuals |
12 Basket Strainer |
13 (H20) |
14 Cartridge Filter |
15 Extract |
16 Spills |
Output |
367.15 |
0.00 |
0.00 |
0.00 |
22.00 |
0.00 |
0.00 |
0.06 |
0.00 |
389.21 |
328.46 |
2.74 |
0.00 |
0.00 |
17.01 |
0.00 |
0.00 |
0.05 |
0.00 |
348.26 |
351.54 |
9.53 |
0.00 |
2.75 |
8.75 |
0.00 |
0.00 |
0.19 |
0.00 |
372.76 |
296.50 |
5.19 |
0.00 |
1.50 |
8.50 |
0.00 |
0.00 |
0.00 |
0.00 |
311.69 |
305.00 |
6.09 |
0.00 |
2.25 |
8.38 |
0.00 |
0.00 |
0.00 |
0.00 |
321.72 |
259.00 |
0.00 |
0.00 |
0.75 |
8.19 |
0.00 |
0.00 |
0.00 |
0.00 |
267.94 |
321.25 |
4.44 |
0.00 |
0.00 |
8.25 |
0.00 |
0.00 |
0.00 |
0.00 |
333.94 |
347.00 |
5.77 |
0.00 |
1.25 |
8.25 |
0.00 |
0.00 |
0.00 |
0.00 |
362.27 |
378.00 |
5.77 |
0. 00 |
1.00 |
0.00 |
,;0.00 |
0.00 |
0.00 |
0.00 |
384.77 |
346.75 |
31.81 |
0.00 |
0.00 |
8.94 |
0.00 |
2.81 |
0.31 |
0.00 |
390.62 |
3300.65
71.34
0.00
9.50
98.27
0.00
2.81
0.61
0.00
3483.18
Accumulation &
Other Losses
121.78
54.67
-25.06
77.74
22.28
68.07
I I
-51.06 | -7.09 | -16.28
8.93
253.98
-------
TOTAL SOLIDS BALAJtCC (U*>
TEST «Z
ITEM
1 Feed Material |
2 (Sampling) |
3 (Strainer) |
4 (Spills) |
5 (Residuals)
6 Water
7 Total Feed
8 Raffinate
9 Sampling
10 Spills
11 Residuals
12 Basket Strainer
13 (H20)
14 Cartridge Filter
15 Extract
16 Spills
Output
PASS #
1
111.21 |
0.00 |
1.06 |
0.00 |
12.32 |
0.00 |
97.82 |
55.07 |
0.00 |
| 0.00 |
0.00 |
I 4-18 I
| 0.00 |
| 0.00 |
I o.oo j
| 0.00 |
| 59.25 |
2
55.07
0.46
0.00
0.00
0.43
0.00
54.18
45.98
0.38
0.00
0.00
1.36
0.00
0.00
0.00
0.00
47.73
3
| 45.98 |
| 0.00 !
| 0.00 |
| 0.00 |
1 0.19 |
| 0.00 |
| 45.80 I
| 58.00 |
| 1.57 |
| 0.00 |
I 0.45 |
I 1-66 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
| 61 .69 |
4
58.00 |
0.19 |
0.00 |
0.00 I
0.47 |
0.00 |
57.35 |
23.72 |
0.42 |
0.00 |
0.12 |
1.11 |
0.00 |
0.00 |
0.00 |
0.00 |
25.36 |
5
23.72
0.00
0.00
0.00
0.12
0.00
23.60
27.45
0.55
0.00
0.20
1.42
0.00
0.00
0.00
0.00
29.63
6
I 27.45 I
| 0.00 |
| 0.00 |
| 0.00 |
| 0.03 |
| 0.00 |
| 27.42 |
I 23.31 |
| 0.00 I
| 0.00 |
| 0.07 |
| 1.06 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
| 24.44 |
7
23.31 |
0.46 |
0.00 |
0.00 |
0.02 |
0.00 |
22.83 |
==============
22.17 |
0.31 I
0.00 |
0.00 |
1.16 |
0.00 |
0.00 |
0.00 |
0.00 |
23.63 |
8
22.17 |
0.00 |
0.00 |
0.00 |
0.03 I
0.00 |
22.14 |
==========
31.58 |
0.53 |
0.00 |
0.11 |
0.91 |
0.00 |
0.00 |
0.00 |
0.00 |
33.12 |
9
31.58 |
0.00 |
0.00 |
0.00 |
0.07 |
0.00 |
31.51 |
=== ===========
34.02 |
0.52 |
0.00 |
0.09 |
0.00 |
0.00 |
0.00 |
0.00 I
0.00 |
34.63 |
10
34.02
0.00
0.00
0.00
0.07
0.00
33.95
=====:
31.21
2.86
0.00
0.00
1.25
0.00
1.60
0.00
0.00
36.92
TOTALS
| 432.51
| 1.11
| 1.06
| 0.00
| 13.75
| 0.00
| 416.59
| 352.51
I 7.13
| 0.00
| 1.05
| 14.11
| 0.00
| 1.60
| 0.00
| 0.00
| 376.41
Accumulation &
Other Losses
I 38.57 I
6.45
-15.89 | 31.99
====================
-6.03
2.98
-0.80
-10.99
I I
-3.12 I -2.97 | 40.19
-------
TOTAL PCB BALANCE (Grams)
TEST #2
I TEH
1 Feed Material
2 (Sampling)
3 (Strainer)
4 (Spills)
5 (Residuals)
6 Water
7 Total Feed
§} a Raffinate
9 Sailing
10 Spills
11 Residuals
12 tasket Strainer
13 (H20)
14 Cartridge Filter
15 Extract
16 Spills
Output
Accumulation &
Other Losses
PASS #
1
| 17.67 |
| 0.00 |
1 0.17 |
| 0.00 |
1 0.95 |
| 0.00 |
1 16.55 |
1 1.» I
I 0.00 |
| 0.00 |
| 0.00 |
| 0.41 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
1 2.34 |
I SS S S ESS 5 SSS SS JSS S S'.
\ 1
2
1.93 |
0.02 |
0.00 |
0.00 |
0.02 |
0.00 |
1.89 |
mnnsnun*
1.09 |
0.01 |
0.00 |
0.00 |
0.10 |
0.00 |
0.00 |
0.00 |
0.00 |
1.20 |
1
0.70 1
3
1.09
0.00
0.00
0.00
0.00
0.00
1.08
0.53
0.01
0.00
0.00
0.02
0.00
0.00
0.00
0.00
0.57
0.51
4
1 0.53 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
1 0.52 |
1 0.71 |
| 0.01 |
| 0.00 |
| 0.00 |
| 0.01 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
1 0.74 |
==================
1 1
1 -0.22 1
5
0.71 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.71 |
0.74 |
0.01 |
0.00 |
0.01 |
0.02 |
0.00 |
0.00 |
0.00 |
0.00 |
0.77 |
=========1
1
0.07 1
6
0.74 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.73 |
0.43 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.44 |
1
n i
7
0.43
0.01
0.00
0.00
0.00
0.00
0.42
0.36
0.01
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.38
n lu
8
| 0.36 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
j 0.00 |
| 0.36 |
1 0.42 |
1 o.oi |
| 0.00 |
| 0.00 |
1 0.01 |
1 0.00 |
| 0.00 |
1 0.00 |
| 0.00 |
| 0.44 |
1 1
1 .n n? i
9
0.42 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 }
0.41 |
0.12 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.12 |
:============
1
n -3rt i
10
0.12
0.00
0.00
0.00
0.00
0.00
0.12
0.57
0.05
0.00
0.00
0.01
0.00
0.00
0.03
0.00
0.66
TOTALS
| 23.99
| 0.03
| 0.17
| 0.00
| 0.98
| 0.00
| 22.81
| *.»
| 0.12
| 0.00
| 0.02
1 0.61
| 0.00
| 0.00
| 0.03
| 0.00
| 7.66
1
0.2
-0.54
15.15
-------
TOW KASS BALANCE (Lbs)
TESTtt
I TEH
1 Feed Material
Z (Sampling)
3 (Strainer)
4 (Spills)
5 (Residuals)
6 Water
7 Total Feed
g> 8 Raffinate
9 sampling
10 Spills
11 Residuals
12 Basket Strainer
13 (H20)
14 Cartridge Filter
15 Extract
16 Spills
Output
= sis ss SS===S===S=
Accumulation &
PASS*
1
| 460.75
| 40.74
| 0.00
| 0.00
| 0.87
| 88.86
| 508.00
| 465.00
| 5.95
| 0.00
| 2.75
| 9.75
| 0.00
| 0.00
| 0.19
| 0.00
| 483.64
I
I 24 36
Z
| 563.00 |
I 4.64 |
| 0.00 |
| 0.00 |
| 3.26 |
| 0.00 |
| 555.10 |
| 482.30 |
I 5.81 |
| 0.00 |
| 0.80 |
| 8.44 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.00 |
| 497.35 |
1 1
1 57.75 1
3 4 5 6 7 8 9 10
482.30 | | I I I I I
0.00 | | I I I I I
0.00 | | I I I I I
0.00 | | | I I I I
130.55 | | |l I I I
0.00 | | I I I I I
351.75 | | 1 1 1 1 1
310.75 | | I I I' '
0.00 | I I I I I I
0.00 | | I I I I '
0.00 | | I I I I '
9.37 | I I I I I '
0.00 | | I I I I '
1.37 | I I I l-l I
0.94 | | I I I I '
0.00 | I II I ! '
322.43 | I I I I I '
:==========================================^=====^=~=========~========================
i i i i j | '
29.32 | | | I I I '
TOTALS
| 1506.05
| 45.38
| 0.00
| 0.00
| 134.68
| 88.86
| 1414.85
| 1258.05
| 11.76
| 0.00
| 3.55
| 27.56
| 0.00
| 1.37
| 1.13
| 0.00
} 1303.42
:=============
1
| 111.43
.___
=============================:
-------
TOTAL SOLIDS BALANCE (Lbs)
TEST #3
I TEH
1 Feed Material |
2 (Sampling) |
3 (Strainer) |
4 (Spills) |
5 (Residuals) |
6 Water |
7 Total Feed |
S3 8 Raffinate |
T9 Sampling |
10 Spills |
,11 Residuals |
12 Basket Strainer |
13 (H20) |
14 Cartridge Filter |
15 Extract |
16 Spills |
Output |
Accumulation & |
Other Losses |
PASS #
1
72.94
6.45
0.00
0.00
0.14
0.00
66.35
69.75
0.89
0.00
0.41
3.90
0.00
0.00
0.00
0.00
74.96
-8.61
23 4 56 7 8
| 67.56 | 57.88 | III | |
| 0.56 | 0.00 | II | | |
| 0.00 | 0.00 | | | | | |
| 0.00 | 0.00 | | | | | |
| 0.39 [ 15.67 || | | | |
| 0.00 | 0.00 | || || |
| 66.61 | 42.21 | II | | |
| 57.88 | 34.18 | | | II |
| 0.70 | 0.00 | | | '| 'j I
| 0.00 | 0.00 | II | | |
1 0.10 | 0.00 | | | III
I 1-52 I 1.69 | I | | I I
I 0.00 | 0.00 | | | II |
I 0.00 | 0.74 | | || II
I 0.00 | 0.00 || | | II
| 0.00 | 0.00 | | | | | I
| 60.19 | 36.61 | | | | | |
1 1 I I 1 II 1
1 6-42 | 5.60 | | 1 | | |
9 10 TOTALS
| | 198.37
1 1 7.01
| | 0.00
I | o.oo
I | 16.19
| | 0.00
| | 175.17
| | 161.81
| 1 1.59
| | 0.00
1 1 0.51
1 1 7.11
| | 0.00
1 1 0.74
I I o.oo
I | o.oo
| | 171.75
I I
1 1 3.42
-------
=====================
K'JL I I
I 00'I- I Zri I Z6'9 I sasson
I I I I »
=======================
Sl'9 1111
OO'O I | 1 1
U-L | I 1 1
OO'O 1 1 1 1
OO'O ill)
l£'0 1 1 1 1
lO'O 1 1 1 1
OO'O | I 1 I
WTO 1 1 1 1
1.9", I 1 1 1
9>'£l 1 1 1 1
OO'O 1 I I I
SS'O I 1 1 1
OO'O 1 1 1 I
OO'O 1 1 1 1
88-0 i 1 1 I
06-H 1 1 1 1
S1V101 01 6 8
1 1 1 1 8£'Z
| I i i OO'O
i i i i son
| | 1 1 OO'O
| I 1 1 OO'O
| | 1 1 90'0
| I 1 1 OO'O
| 1 1 1 OO'O
| I 1 1 OO'O
1 1 1 I 6Z't
I 1 1 1 8£'L
III! oo'o
1 1 1 1 is'o
1 1 1 1 OO'O
| 1 1 1 OO'O
1 I 1 1 oo-o
{ I 1 I 68'1
/ 9 S * £
1 U'l
\ OO'O
i oo'o
1 OO'O
1 OO'O
1 SO'O
1 OO'O
1 OO'O
1 ZO'O
i 68' I
1 6£'£
i OO'O
1 ZO'O
1 OO'O
i OO'O
1 £0'0
1 WE
Z
| 08' I 1 V*»0
\ OO'O 1 slHdS 91
i SO'O ! J3SJJX3 SI
I OO'O 1 JailH a6piJJJB3 n
| OO'O 1 (OZH) £1
, oz-o 1 WQSWZI
| lO'O 1 sienpisaa u
| OO'O 1 sll!dS 01
| EQ'O 1 Bujvtaes 6
00
| (,^'1 | aiEUl^jea g OO
-- .^-. - * *-B=====sss:si=sss==^T:==gas
1 ZZ'8 1 P33^ lel°l L
\ oo'o i JS*Bn 9
1 lO'O 1 (sienpjsaa) S
| OO'O 1 (si lids) f
| OO'O 1 CJSUIBJJS) £
| 59-0 I (6u$ldws> 2
1 SS'6 I 1BJJ91BHPMJ I
I H31I
#SSVd
(SMJ9> 33WftV8 83d TVIOI
-------
I TEH
PASS #
1
TOTAL MASS BALANCE (Lbs)
TEST #4
8 9 10 TOTALS
1
2
3
4
5
6
7
Feed Material
(Sampling)
(Strainer)
(Spills)
(Residuals)
Water
Total Feed
| 333.25 |
| 33.86 |
| 0.00 |
| 0.00 |
| 0.56 |
| 0.00 |
| 298.83 |
279,
0.
0.
0.
0.
39.
318.
.75 |
.00 |
,00 |
00 |
68 |
32 |
39 |
348.75 |
0.00 |
0.00 |
0.00 |
0.94 |
39.14 |
386.95 |
298.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
298.00 |
360,
0.
0.
0.
1.
16.
375.
.20 |
,00 |
,00 |
00 |
20 |
70 |
70 |
256
0
0
0
0
22,
278,
.00 | |
.00 | |
.00 | |
.00 | |
.26 | |
.42 | |
16 | |
I | | 1875.95
I I | 33.86
I | | 0.00
I | | 0.00
I I | 3.64
I | | 117.58
I | | 1956.03
s=sss==sss=
a Raffinate |
9 Sampling |
10 Spills |
11 Residuals |
12 Basket Strainer |
13 (H20) |
14 Cartridge Filter |
15 Extract |
16 Spills |
Output |
279.75 |
5.48 |
0.00 |
0.00 |
8.25 |
0.00 |
0.00 |
0.00 |
0.00 |
293.48 |
348.75 |
6.57 |
0.00 |
37.75 |
8.32 |
0.00 |
0.00 |
0.00 |
0.00 |
401.39 |
297.50 |
6.76 |
0.00 |
1.00 |
8.06 |
0.00 |
0.00 |
0.00 |
0.00 |
313.32 |
360.20 |
4.76 J
0.00 |
3.75 |
8.25 |
0.00 |
0.00 |
1.00 |
0.00 |
377.96 |
256.00 |
4.23 |
0.00 |
1.75 |
8.00 |
0.00 |
0.00 |
0.00 |
0.00 |
269.98 1
319.25 | II
o.oo | | |
0.00 | | |
0.50 | ||
7.81 | | |
0.00 | | |
2.00 | | |
1.38 | ||
0.00 | | |
330.94 I 1 I
I | 1861.45
| ... |._.. 27.80
|| 0.00
I | 44.75
I | 48.69
| | 0.00
I | 2.00
1 1 2-38
| | 0.00
1 1 1987.07
Accumulation &
Other Losses
5.35
-83.00
73.63
=========
-79.96
105.72
-52.78
-31.04
-------
TOTAL SOLIDS BALANCE (Lbs)
T6ST«
I
ITEM
1 Feed Material |
2 (Sampling) |
3 (Strainer) |
4 (Spills) |
5 (Residuals) |
6 Water ' 1
7 Total Feed |
8 Raff inate |
9 Saapling |
10 Spills |
11 Residuals |
12 Basket Strainer |
13 (H20) |
14 Cartridge Filter |
15 Extract |
16 Spills |
Output |
Accumulation & |
Other Losses |
>ASS*
1
53.99 |
5.49 |
0.00 |
0.00 |
0.09 |
0.00 |
48.41 |
36.37 |
0.71 |
0.00 |
0.00 |
1.40 |
0.00 |
0.00 |
0.00 |
0.00 |
38.48 |
I
9.93 |
2
36.37 |
0.00 |
0.00 |
0.00 |
0.09 |
0.00 |
36.28 |
========:
41.85 |
0.79 |
0.00 |
4.53 |
0.83 |
0.00 |
0.00 |
0.00 |
0.00 |
48.00 |
I
-11.72 |
3
41.85 |
0.00 |
0.00 |
0.00 |
0.11 |
0.00 |
41.74 |
31.24 |
0.71 j
0.00 |
0.11 |
0.81 |
0.00 |
0.00 |
0.00 |
0.00 |
32.86 i
1
8.88 |
4
31.29 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
31.29 |
26.11 |
0.35 |
0.00 |
0.27 |
1.03 |
0.00 |
0.00 |
0.00 |
0.00 |
27.76 |
I
3.53 |
5
26.11
0.00
0.00
0.00
0.09
0.00
26.02
18.43
0.30
0.00
0.13
0.88
0.00
0.00
0.00
0.00
19.74
6.28
6
| 18.43 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.02 |
| 0.00 |
| 18.41 |
I 19.95 |
| 0.00 |
| 0.00 |
| 0.03 |
I 0.55 |
| 0.00 |
I 0.38 |
| 0.00 |
| 0.00 |
| 20.91 |
:========:
1 1
1 -2.50 |
7 8 9 10 TOTALS
, | | | 208.04
| I I I 5M
| | I | 0.00
II I I o.oo
| | I I 0.40
| | I | 0.00
| | I I 202.15
, | | | 173.95
| I 1 | 2.86
| | I | 0.00
| I 1 | 5.06
,||| 5.50
,||| 0.00
,||| 0.38
| | I 1 0.00
, || | 0.00
1 1 i i 1S7-76
1 II j
=====-==================
-------
TOTAL PCB BALANCE (Grams)
TEST #4
1
2
3
4
5
6
7
==;
ITEM
Feed Material
(Sampling)
(Strainer)
(Spills)
(Residuals)
Water
Total Feed
===============
PASS
1
I 62
I 6
I o,
I o.
I o.
I o.
I 56.
#
.!>6
.36
.00
.00
,11
00
10
2
| 16.51 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.04 |
| 0.00 |
| 16.47 |
3
18
0.
0.
0.
0.
0.
18.
.81 |
.00 |
.00 |
,00 |
05 |
00 |
76 |
4
9.54
0.00
0.00
0.00
0.00
0.00
9.54
s======:::
5
| 3.85 |
| 0.00 |
| 0.00 |
| 0.00 |
| 0.0133 |
| 0.00 |
1 3.84 |
=================:
6
2
0,
0.
0.
0.
0.
2.
.01 |
.00 |
.00 |
00 |
00 |
00 |
01 |
7 8 9 10 TOTALS
I I I | 113.28
I I I | 6.36
I I || 0.00
I I I | 0.00
I I I | 0.2l"
I I I | 0.00
jl|| 106.71
-------
APPENDIX B
MASS BALANCES FOR DECONTAMINATION EFFLUENTS
92
-------
Mass Balance for Test 3 Fuel Wash
Analyte Data
CO
Test 3 Pass 1 09-17-88
EPT-W1----- -- Coleman Fuel (gal), est
* Specific Gravity
* Density of Water (Ibs/gal)
Wash Amount (Ibs)
EPT-U20 Coleman Fuel (gal), est
* Specific Gravity
* Density of Water (Ibs/gal)
Wash Amount (Ibs)
Still Bottoms Residuals (gal), est
* Specific Gravity
* Density of Water (Ibs/gal)
Wash Amount (Ibs)
EPT-U20
Still Bottoms
Total Output (Ibs)
Date
I
Raw
Data
1.00
0.72
8.35
6.00
==========
18.00
0.72
8.35
108.07
0.47
0.96
8.35
3.75
6.00
108.07
3.75
117.82
(Correctec
| Data
| Std Dev
(Ibs)
0.0000
0.0000
0.0000
0.0293
========
0.0077
0.0000
0.0000
0.5277
=========
0.0015
0.0000
o.dooo
0.3125
========
0.0293
0.5277
0.3125
0.9325
Cone.
(ppm)
560
570
:
3400
Std Dev
(ppm)
95.2
96.9
578
Mass
(g)
2.12
38.78
8.03
2.12
38.78
8.03
48.93
Std Dev
(9)
0.05
I
0.52
I
I
0.86
0.05
0.52
0.86
1.20
-------
Analyte Data
Hass Balance for Test 5, Toluene Mash
Test 5 Passes 1-3 10-2-88 Date
* Specific Gravity
* Density of Water (Ibs/gal)
Net Feed
- Drum |9/29
Toluene Wash
Toluene Wash
+ Sample Losses
vo
-P* Total Toluene Wash
* -
in
1 - orun 19/29
3 Toluene Wash
1 Toluene Wash
g + Sample Losses
9 Total Toluene Wash
- Drum A 19/29
Toluene Wash
1
Raw |
Data
100.00
0.87
8.35
723.11
=========
226.00
38.50
187.50
187.50
0.52
188.02
74.00
38.75
35.25
35.25
0.52
35.77
323.50
39.50
284.00
1
:orrected|
(Ibs)
100.00
0.87
8.35
723.11
=======
226.00
38.50
187.50
187.50
0.52
188.02
========
74.00
38.75
35.25
35.25
0.52
35.77
323.50
39.50
284.00
/ariance
(lbs)"2
2.08
0.00
0.00
145.25
========
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
=ssssss=
0.00
0.00
0.00
Cone. |<
(ppa)
0.00
185.00
34.00
842.50
PC
/arionce
(ppn»A2
0.00
989.10
33.41
(20513.40
3
Mass |\
(9)
0.00
18.16
18.16
0.05
18.21
0.63
0.63
0.01
0.64
125.26
/arionce
0.00
7.17
0.00
0.00
0.00
=========
0.01
0.01
0.04
0.04
=========
-
341.02
------- |