United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/625/R-92/014
October 1992
Technology Transfer
&EPA RCRA Corrective Action
Stabilization Technologies
Proceedings
-------
-------
EPA/625/R-92/014
October 1992
RCRA CORRECTIVE ACTION
STABILIZATION TECHNOLOGIES
PROCEEDINGS
U.S. ENVIRONMENTAL PROTECTION AGENCY
CENTER FOR ENVIRONMENTAL RESEARCH INFORMATION
OFFICE OF RESEARCH AND DEVELOPMENT
CINCINNATI, OH
Printed on Recycled Paper
-------
NpTICE
This document has been reviewed in accordance with the U.S. Environmental Protection Agency
(EPA) peer and administrative review policies and approved for publication. EPA strives to provide
accurate, complete, and useful information. Neither EPA nor any person contributing to the prepa-
ration of this document, however, makes any;warranty, expressed or implied, with respect to the
usefulness or effectiveness of any information, method, or process disclosed in this material. Nor
does EPA assume any liability for the use of, or for damages arising from the use of, any
information, methods, or process disclosed in this document. The reader should be aware that
some of the technologies discussed may nothave direct application to the stabilization program.
In addition, other technologies, not discussed, may have application to the stabilization concept.
I !
Mention of trade names or commercial products does not constitute endorsement or recommen-
dation for use. :
-------
ACKNOWLEDGMENTS
This proceedings document was developed by the U.S. Environmental Protection Agency's (EPA's)
Center for Environmental Research Information (CERI), Cincinnati, Ohio, from presentations made
at a series of seminars on Resource. Conservation and Recovery Act (RCRA) Corrective Action
Stabilization Technologies and at the National Corrective Action Stabilization Conference held in
Colorado Springs in February 1992. Ed Barth, CERI, provided technical direction and management
for the project. David Bartenfelder, EPA Office of Solid Waste (OW), Washington, DC, coordinated
the participation of authors from .the Colorado Springs workshop and provided review for the
document. Guy Thomasino and Jon Perry, OSW, served as peer reviewers. John Reinhardt and
Susan Richmond of Eastern Research Group, Inc. (ERG), Lexington, Massachusetts, provided
writing and editorial support, and Ivan Rudnicki and Nick Kanaracus, also of ERG, provided
production assistance.
The authors for the document chapters are as follows:
David Bartenfelder, Ph.D., Office of Solid Waste, U.S. EPA, Washington, DC, Chapter 1
Ronald Sims, Ph.D., Division of Environmental Engineering, Utah State University, Chapters 2 and 6
Harry Compton, Office of Emergency and Remedial Response, U.S. EPA, Chapter 3
Walter Grube, Ph.D., Clem Environmental Corporation, Fairmount, Georgia, Chapter 4
Lawrence Murdoch, Ph.D., University of Cincinnati, Cincinnati, Ohio, Chapter 5
Ryan Dupont, Ph.D., Division of Environmental Engineering, Utah State University, Chapter 6
The authors of the case studies in Appendix A are:
Frank Gardner, U.S. EPA Region IX, San Francisco, CA, Section A. 1
Dennis Zielinski, U.S. EPA, Region III, Philadelphia, PA, Section A.2
Glenn Heyman, U.S. EPA, Region IX, San Francisco, CA, Section A.3
in
-------
PREFACE
The main body of this document was developed from technical presentations given at a series of
seminars on the U.S. Environmental Protection Agency's (EPA's) Resource Conservation and
Recovery Act (RCRA) Corrective Action Stabilization Technologies Program, sponsored by the
Center for Environmental Research Information (CERI), Office of Research and Development
(ORD). Seminars in the series were held in all 10 EPA regional offices during Fall 1991 and Winter
1992. This seminar series introduced several technologies that may have application to the RCRA
stabilization concept; however, many other existing technologies also may be applicable.
Chapter 1 of this proceedings provides an introduction to the RCRA corrective action stabilization
strategy. Chapter 2 presents a conceptual approach for characterizing problems, including the use
of the chemical mass balance. Chapter 3 discusses field screening methods including gas chro-
matography, mass spectrometry, x-ray fluorescence spectrometry, chemical sensor systems, sam-
pling and analysis for soil gases and air, and immunoassay methods. Chapter 4 describes
construction methods and considerations for several technologies such as covers, slurry walls,
and grouting for RCRA corrective actions. Chapter 5 discusses innovative in situ delivery and
recovery of liquids to facilitate RCRA correptive actions. Chapter 6 covers vapor extraction and
gas control and bioventing as technologies tor RCRA corrective actions.
Appendix A is a compilation of case studies1 originally presented at the National Corrective Action
Conference. The conference was held February 25 through 27, 1992, in Colorado Springs, Colo-
rado. Conference participants included consultants and representatives from EPA, other federal
agencies, states, and industry. The conference was organized to educate and demonstrate how
and when stabilization would be an effective remedial approach in the corrective action process.
The selected presentations are a sampling pf all presentations that portrayed the effective imple-
mentation of this strategy.
For additional information on stabilization technologies for RCRA Corrective Actions, the reader
can access the RCRA Corrective Action Bulletin on the ORD Bulletin Board System (BBS) at
(513) 569-7610. ;
IV
-------
CONTENTS
CHAPTER 1 Stabilization: A Strategy for RCRA Corrective Action .....,..• 1
1.1 BACKGROUND . . . . . ..'..' 1
1.2 CORRECTIVE ACTION AND STABILIZATION 1
1.2.1 Objectives and Scope of the RCRA Corrective Action Program 1
1.2.2 Objectives and Scope of the Stabilization Initiative 3
1.2.3 Stabilization as Part of the Corrective Action Process 3
1.3 IMPLEMENTING STABILIZATION . . . . 4
1.3.1 Overview 4
1.4 DATA CONSIDERATIONS FOR STABILIZATION 8
1.4.1 Overview , . . . 8
1.4.2 Phased Investigations 8
1.4.3 Integrating Stabilization into RFI Workplans 8
1.5 REFERENCE 8
CHAPTER 2 Conceptual Approach for Characterizing RCRA Contaminated Facilities 11
2,1 BACKGROUND 11
2.2 THE CHEMICAL MASS BALANCE 11
2.2.1 A Hydraulic Mass Balance Model 12
2.2.2 Transformation Potential : . 13
2.2.3 Subsurface Mass Balance and Phases 13
2.2.4 Distribution Coefficients 16
2.2.5 Migration/Transport ; 16
2.2.6 Management 17
2.3 METHODOLOGY 18
2.3.1 Characterization '. 18
2.3.2 Problem Assessment 18
2.3.3 Treatment Train Selection .19
2.3.4 Monitoring Treatment Performance 19
2.4 REFERENCES 20
CHAPTER 3 Field Screening Methods 21
3.1 FIELD SCREENING REQUIREMENTS: SCOPE AND EXPECTATIONS 21
-------
3.2 FIELD TECHNIQUES AND INSTRUMENTATION '•"'." 22
I
3.2.1 Gas Chromatographs 22
3.2.2 Gas Chromatography/Mass Spectrometry 24
3.2.3 Chemical Sensor Systems ..;..... 24
3.2.4 X-Ray Fluorescence Spectrometers • • 25
3.2.5 Sampling and Analysis Equipment for Soil-Gas and Air . 25
3.2.6 Immunoassay Methods 27
3.3 FUTURE NEEDS AND DIRECTIONS i 28
3.4 REFERENCE | .........:... . 28
i
CHAPTER 4 Covers, Slurry Walls, Grouting, and pynamic Compaction
for RCRA Corrective Action Stabilization 29
4.1 COVER SYSTEMS I . - 29
4.1.1 Considerations '. • • 29
4.1.2 Recommended Design . . . j. 29
4.1.3 Soil Components ...... i 30
4.1.4 Geosynthetic Components . ! 32
4.1.5 Accessory Structures . . . . i • • 33
4.1.6 Subsidence . . i 33
4.1.7 Materials Handling ...... i • 35
4.2 SLURRY WALLS ; 35
4.2.1 Equipment [. 35
4.2.2 Slurry Wall Performance Testing 35
4.2.3 Advances in Slurry Walls . . , 37
4.3 GROUTING ; 37
|
4.4 DYNAMIC COMPACTION TECHNOLOGY 38
I
4.5 QUALITY ASSURANCE i 39
4.6 REFERENCES ! 40
CHAPTER 5 In Situ Delivery and Recovery of Liquids
to Facilitate RCRA Corrective Actions 41
i .
5.1 PERFORMANCE • 41
5.2 LIQUID FLOW : • • • • 41
5.2.1 Flow Rate ; 41
5.2.2 Hydraulic Conductivity • . . ., 42
5.2.3 Distribution Coefficients . . j 43
I
5.3 FIELD METHODS . .! 43
5.3.1 Vertical Wells : 43
5.3.2 Hydraulic Fracturing . . . . j 46
5.3.3 Trenches . . .^. . . 48
5.3.4 Directional Drilling ; 49
5.3.5 Soil Flushing 51
VI
-------
5.3.6 Other Techniques ; ............................... 51
5.4 REFERENCES . . . - . - - - 51
CHAPTER 6 Vapor Extraction, Bioventing . . . 53
6.1 INTRODUCTION . . . . 53
6.2 PROCESS DESCRIPTION 53
6.2.1 System Components '. 53
6.2.2 System Variables 55
6.3 APPLICATIONS OF VAPOR EXTRACTION TO REMOVE LIGHT NAPLS 58
6.4 BIOVENTING .......:.... 59
6.4.1 Application of Bioventing 61
6.4.2 Monitoring 62
6.5 REFERENCES , 62
APPENDIX A Case Studies . 65
A.1 GROUND-WATER CONTAMINATION STABILIZATION WITH LNAPL
COLLECTION SYSTEM: JAMESTOWN, NY,
by Frank Gardner, U.S. EPA, Region IX, San Francisco, CA 65
A.1.1 Introduction • • 65
A.1.2 Stabilization Strategies 66
A.1.3 Implementation and Future Actions 68
A.1.4 Conclusions and Discussion 68
A.2 LANDFILL STABILIZATION: BFI SOLLEY ROAD FACILITY, GLEN BURNIE, MD,
by Dennis Zielinski, U.S. EPA, Region III, Philadelphia, PA 69
A.2.1 Introduction 69
A.2.2 Stabilization Strategies 72
A.2.3 Implementation and Future Actions 74
A.2.4 Conclusions and Discussion '. 74
A.2.5 References . . . '. • 75
A.3 GROUND-WATER CONTAMINATION STABILIZATION: ROMIC CHEMICAL CORPORATION
FACILITY, EAST PALO ALTO, CA,
by Glenn Heyman, U.S. EPA, Region IX, San Francisco, CA 75
A.3.1 Introduction 75
A.3.2 Stabilization Strategies 77
A.3.3 Implementation and Future Actions 77
A.3.4 Conclusions and Discussion 77
VII
-------
FIGURES
Figure
Page
1-1 Stabilization strategy flowchart i "'. .'":. 6
2-1 A bathtub model of a mass balance showing water and various chemicals entering a site and their
phase transformations I . ... 12
2-2 Fate of hazardous contaminants in the subsurface (Sims et all, 1989) 13
2-3 Mass balance conceptual approach (U.S. EPA, 1991) . 13
2-4 Composition of organic matter. [ 14
2-5 U.S. Department of Agriculture soil textural classification (texture triangle) (Soil Conservation
Service, 1971) [ 15
2-6 Values for capillary rise (a), porosity (b), and permeability (c) in various soils (Dragun, 1988) 15
2-7 Methodology for integrating site characterization with subsurface remediation (U.S. EPA, 1990) 16
2-8 Treatment train approach ! 17
2-9 Contaminated site with an unsaturated and saturated zone 18
2-10 Treatment train from product removal to pump and treat, soil flushing, and bioremediation. . 19
3-1 Chemical classes of 200 hazardous substances 22
3-2 Selection of portable gas chromatograph detectors based on volatile target compounds 23
3-3 Sampling train schematic for soil-gas survey. !'. 26
4-1 Corrective action stabilization site i 30
4-2 Recommended landfill cover design (U.S. EPA, 1991) 30
4-3 Recommended landfill cover with options (U.Sj. EPA, 1991) 31
4-4 Recommended procedure (U.S. EPA, 1991)..[ 31
4-5 Leachate or precipitation infiltration with a soili(clay) liner and composite liner (U.S. EPA, 1991) 33
4-6 Cover with gas vent outlet and vent layer (U.S. EPA, 1991) .-••-. 33
4-7 Conductive layer barrier with graded slopes. . [ 34
4-8 Cumulative subsidence (U.S. EPA, 1991). . . j 34
4-9 Derivation of geomembrane tensile stress due, to subsidence of material beneath the geomembrane. . . 34
4-10 Simulation processes in the HELP model (U.S. EPA, 1991) .35
4-11 Slurry wall excavation and backfill | ". . 36
4-12 Methods of construction for a slurry wall.. . .; 36
4-13 Uses of grouting to contain hazardous wastes.' 37
4-14 Cross-section of grouted bottom seal beneath ja landfill (U.S. EPA, 1986) 37
4-15 Schematic of grout injection system (U.S. EPA, 1986). . 38
4-16 Injectibility of particulate and chemical grout in'fine and coarse soils (U.S. EPA, 1986) 38
4-17 Monitoring well configuration ' ' 39
5-1 Aqueous phase displacing nonaqueous phase i(NAPL) in soil 42
5-2 Range of hydraulic conductivity where general! methods of delivery and recovery are regarded
as applicable 42
5-3 Vertical well—water-level activated pump. . ., 43
5-4 Vertical well—two-pump system ! ...:.... 44
5-5 Regional flow from top to bottom approaching 'an irregularly shaped contaminated area 44
5-6 Capture of contaminants by a line of extraction wells. 44
5-7 Approach to low regional flow using three extraction wells and four injection wells. 45
5-8 Concentration from the extraction well as a function of time. . . 46
5-9 Creation of hydraulic fracture j .47
5-10 Map of three fractures created at borehole EL6 , 47
5-11 Cross-sections of three fractures created at borehole EL6 showing topographic profile and details
of fracture traces 47
5-12 Suction head as a function of distance from conventional vapor extraction well and well intersecting
two hydraulic fractures. Vapor discharge from 1 hp blower as a function of time 48
VIII
-------
5-f 3 Use of trench to prevent a contaminant from entering a water supply. 48
5-14 Side view of an interceptor trench : 49
5-15 The trajectory of the borehole from the Eastman Christensen rig 50
5-16 Petroiphysics rig with a shallow radial system. 50
5-17 Three situations where directional drilling could be beneficial 50
6-1 Schematic of ventilated soil (Valsaraj and Thibodeaux,1988). 53
6-2 Laboratory flask apparatus used for mass balance measurements (Sims et al., 1989) 54
6-3 Schematic of a gas extraction well (U.S. EPA, 1991a) 54
6-4 Effect of moisture on VOC adsorption and desorption in soil—VOC adsorption with two moisture
regimes (Reible, 1989) 56
6-5 Effect of soil water content on dieldrin vapor pressure (modified from Spenser and Cliath, 1969). .... 56
6-6 Soil texture trilinear diagram (Soil Conservation Service, 1971) 57
6-7 Diffusive release of contaminants from the soil phase into the gas phase (U.S. EPA, 1989) 57
6-8 Concentration vs. time profile showing restart spike (DiGiulio et al., 1990) 58
6-9 Volatilization of a combination of liquids (Johnson, 1989) 58
6-10 Total solute mass in the subsurface vs. time showing the effects of well spacing on vacuum
extraction effectiveness (modified from Wilson et al., 1989). 58
6-11 Total solute mass in the subsurface vs. time showing the effects of well depth on vacuum extraction
effectiveness (Wilson et al., 1989) 59
6-12 Enhancement of bioremediation of gasoline components using vacuum extraction of soil amended
with nutrients and moisture (Hinchee, 1989) 59
6-13 Aerobic biodegradation (Hinchee, 1989) : 60
6-14 Schematic of recommended bioventing system layout 60
6-15 Oxygen concentration in vadose zone before venting (U.S. EPA, 1989) 61
6-16 Oxygen concentration in vadose zone after venting (U.S. EPA, 1989) 61
6-17 The increase in COa and decrease in oxygen as a result of degradation of the jet fuel
components (U.S. EPA, 1989) 62
A-1 Location of case study site, Browning-Ferris Industries (BFI) landfill 69
A-2 BFI landfill site in relation to Baltimore and Glen Burnie, Maryland 69
A-3 Soil geology of the site showing clay, silt and interbedded clay, and silt and sand 70
A-4 Pump and treat with air stripper at Solley Road site 73
A-5 Air stripper system at Solley Road site 74
IX
-------
TABLES
Table
1-1
1-2
1-3
2-1
3-1
3-2
4-1
4-2
4-3
6-1
6-2
6-3
A-1
A-2
A-3
A-4
A-5
A-6
A-7
A-8
A-9
A-10
Pag©
Examples of Stabilization Measures 2
Key Decisions for Determining the Need for Stabilization 5
Stabilization Data Needs i 9
Important Properties and Parameters Needed'to Perform a Soil-Based Waste Characterization 17
Analytical Levels Relative to Data Quality . J 21
Summary of Three Gas Chromatograph Devices 24
Recommended Tests and Observations on Sijibgrade Preparation 32
Recommended Materials Tests for Barrier Layers 32
Customary Primary Functions of Geosynthetics Used in Waste Containment Systems 32
Vapor Pressures of Some Commonly Detected Compounds 55
Henry's Law Constants for Selected Compounds 56
Oxygen Supply '• 60
VOC Concentrations in the Foundation Area .j 66
Impact of Constraints on Remedial Alternatives 67
Contaminants with Concentrations Greater Than the MCL . . 71
Summary of Sampling Data on VOC Concentrations (ppb) June 1989 . . 72
Summary of Carcinogenic Risks | 72
Summary of Noncarcinogenic Risks | 72
Regulatory Chronology j 75
Major Ground-Water Contaminants ' j .76
Major Surface Water Contaminants ..... j 76
Major Slough Sediment Contaminants . . . j 76
-------
CHAPTER 1
Stabilization: A Strategy for RCRA Corrective Action
David Bartenfelder, Ph.D.
Office of Solid Waste
U.S. Environmental Protection Agency
1.1 BACKGROUND
In 1992, EPA began implementing an aggressive program
strategy to increase the number of cleanup activities and
the pace of cleanup and to achieve positive environ-
mental results at Resource Conservation and Recovery
Act (RCRA) treatment, storage, and disposal facilities
(TSDFs) requiring corrective action/While comprehen-
sive facility cleanup is the long-term goal for the RCRA
corrective action program, this new initiative emphasizes
the importance and value of controlling releases and pre-
venting the further spread of contaminants at RCRA
facilities.
Stabilization of RCRA facilities entails EPA and owner/op-
erators taking whatever action is necessary to address
actual and potential exposures to hazardous waste and
to control any further spread of contamination. Stabiliza-
tion activities will utilize available remedial techniques,
such as those in Table 1 -1, at the earliest practicable time.
Implementation of this strategy may require that the fa-
cility-wide RCRA Facility Investigation (RFI) be resched-
uled or redirected so that the interim remedial measures
can be carried out. The RCRA corrective action program
will demonstrate a "bias" for stabilization actions in the
way it manages corrective actions at RCRA facilities. The
near-term goal of the corrective action program will be to
direct investigations and remedial actions to address re-
leases at high priority RCRA facilities.
The RCRA stabilization initiative is a new program phi-
losophy; it should not be confused with corrective meas-
ures that historically relied on solidification/immobilization
technologies. This new initiative stresses the use .of ex-
isting corrective action vehicles (e.g., interim measures
and conditional remedies) to effect environmental bene-
fits through a coherent and coordinated program policy
direction. Although technologies such as solidification, vit-
rification, and other immobilization techniques may be
effective as stabilization measures in certain situations,
EPA also is interested in other source control measures
and measures that will stop or slow any further spread of
contaminant migration. The measures to stabilize re-
leases or other environmental problems could be quite
extensive, including, for example, the installation of a
large-scale pump-and-treat system combined with treat-
ment and/or containment-based source control actions.
In addition, physical and institutional exposure controls,
such as fences, access controls, or provision of alterna-
tive water supplies, also may be required to mitigate ac-
tual or imminent health threats.
Implementing the stabilization initiative will yield substan-
tial benefits for the corrective action program by focusing
limited resources in the near-term on stabilizing environ-
mental problems at more facilities, initially of high priority,
rather than pursuing final, comprehensive remedies at a
few facilities. This should enable the Agency and states
to address the most serious environmental problems at
a larger number of facilities more quickly. Furthermore,
by imposing such expeditious actions, the extent and
incidence of continued environmental degradation from
existing releases should be significantly reduced. Con-
versely, the environmental benefit gained by taking this
early action should impart greater efficiency in final reme-
dies undertaken.
1.2 CORRECTIVE ACTION AND STABILIZATION
1.2.1 Objectives and Scope of the RCRA Corrective
Action Program
The primary objective of the RCRA corrective action pro-
gram is to clean up releases of hazardous waste or haz-
ardous constituents that pose a threat to human health
and the environment. The program applies to all operat-
ing, closed, or closing RCRA hazardous waste facilities
acquiring a permit or under order. RCRA corrective action
authority can be used to address releases to all media
from all solid waste management units (SWMUs) at a
RCRA facility.
To carry out this corrective action program objective, EPA
developed a five-phased corrective action process, which
-------
Table 1-1. Examples of Stabilization Measures
BY MEDIA i
Ground Water
• Interceptor Trench/Sump/Subsurface Drain ;
• Pump-and-Treat System (Source Removal and
Containment) ;
• Physical Barriers (Covers/Slurry Walls) ;
Soils i
• Runoff/Runon Control (Diversion or Collection Devices) !
• Cap/Cover i
• Source Removal (Excavation)
Surface Water Release (Point and Nonpoint) ;
• Overflow/Underflow Dams ;
• Filter Fences
• RunoftfRunon Control (Diversion or Collection Devices) i
• Regrading/Revegetation !
Gas Migration Control ;
• Barriers/Collection/Treatment/Monitoring :
• Evacuation (Buildings)
I
Particulate Emissions
• Truck Wash (Decontamination Unit)
• Revegetation
• Application of Dust Suppressant |
• Cover/Cap |
BY UNIT TYPE
Containers [
• Overpack/Re-drum
• Construct Storage Area/Move to Storage Area '
• Segregation ;
• Temporary Cover :
Tanks
• Overflow/Secondary Containment
• Leak Detection/Repair/Partial or Complete Removal
Surface Impoundments
• Reduce Head
• Remove Free Liquids and/or Highly Mobile Wastes
• Stabilize/Repair Side Walls, Dikes, or Liner(s)
• Provide Temporary Cover
• Runoff/Runon Control (Diversion or Collection Devices)
Landfills
.• Runoff/Runon Control (Diversion or Collection Devices)
• Reduce Head on Liner and/or in Leachate Collection
System
• Repair Leachate Collection/Removal System or French
Drain
• Cap/Cover
• Waste Removal
Waste Piles
• Runoff/Runon Control (Diversion or Collection Devices)
• Cap/Cover
• Waste Removal
OTHER TYPES OF ACTIONS
• Fencing to Prevent Direct Contact
• Alternate Water Supply to Replace Contaminated
Drinking Water
• Temporary Relocation of Exposed Population
• Institutional Restrictions
it described first in the October 1986 National Corrective
Action Strategy (51 FR 37608) and more recently in t|ie
Subpart S proposed rule (55 FR 30798; July 27, 199p).
The first phase is the RCRA Facility Assessment (RFA).
During the RFA, EPA or the state visits the facility, char-
acterizes the solid waste management units, and identi-
fies releases or potential releases requiring further
investigation. The second phase of the process, if Re-
quired, is the RFI. During the RFI, the facility's owner/op-
erator fully characterizes the nature and extent of the
releases. Based on the results of the RFI, a facijity
owner/operator may be required to conduct a Corrective
Measures Study (CMS) to evaluate the potential alterna-
tives for cleanup. In the fourth step, EPA or the state
selects the final remedy, which is then carried out by the
owner/operator during Corrective Measures Implementa-
tion (CMI). The process can take 5 to 8 years from the
start of the RFA to the implementation of the final corr^c-
tive measures. The stabilization initiative is intended to
shorten the time before cleanup can start by encouraging
appropriate interim remedial actions to be identified and
implemented on a short-term basis.
Such interim measures are generally defined as remedial
measures that are undertaken following the RFA and prior
to completion of the CMS. They are usually designed to
address a specific aspect of the overall contamination at
a facility that can be readily isolated. Interim measures
may encompass a broad range of possible actions. In
some cases, interim measures will involve the control of
the source of the release, while in other cases, they will
involve cleanup of contaminated media (i.e., soil, ground
water, surface water, or air), whereas in still other cases,
they may involve limiting exposure to the contamination.
For example, a facility responsible for contamipating a
public drinking water supply may be required to initiate
-------
cleanup of the water supply, as well as to make an alter-
native water supply available until the original water sup-
ply is remediated. Other examples of interim measures
might include initiating ground-water pump-and-treat sys-
tems to contain the flow of contaminated ground water,
fencing an area of contaminated soils to prevent public
access, or overpaying damaged or corroded drums to
prevent possible leakage.
1.2.2 Objectives and Scope of the Stabilization
Initiative
Approximately 4,600 RCRA facilities may be subject to
corrective action requirements. Based on current program
data, EPA estimates that approximately 80 percent, or
3,600, of these facilities, will require varying degrees of
cleanup. Because resources available to address these
facilities are limited, EPA plans to maximize the near-term
environmental benefits of the RCRA corrective action pro-
gram by focusing resources on those facilities and
SWMUs where accelerated cleanup activities can pro-
duce the greatest environmental benefit. One aspect of
this strategy is to stabilize known releases of hazardous
constituents at "high priority" corrective action facilities.
Stabilization means taking whatever action is necessary
to address exposures to hazardous waste releases and
to control or prevent any further migration of contamina-
tion. After actions have been taken to stabilize releases
at high priority facilities, the Agency will focus its attention
on stabilizing and remediating lower priority facilities.
By implementing this strategy, EPA will be able to con-
centrate its efforts at high priority sites where there are
the greatest threats to human health and the environ-
ment, and eliminate or minimize these threats through
either stabilization or remediation, thus achieving the
greatest environmental benefit. Stabilization of a contami-
nant release at a facility will require less time than full
remediation, thereby allowing a greater number of re-
leases to be addressed, at least partially, in a shorter
period of time. Those sites that are stabilized will continue
through the corrective action process toward final reme-
diation, albeit the continued corrective action is projected
to occur at a slower pace and with the potential for a
reduced Lvel of EPA/state oversight.
Stabilization activities at appropriate facilities could pro-
vide a number of benefits to regulatory agencies, the
regulated community, and the general public. As indicated
above, once a facility has entered the stabilization proc-
ess, EPA/state attention and oversight can be refocused
on other facilities, thus maximizing Agency resources.
Facility owners and operators may benefit from stabiliza-
tion by potentially reducing future liability by limiting the
spread of contamination. They also may benefit from re-
duced overall costs and shorter remediation time due to
early installation of the stabilization measure. The public
also benefits from stabilization because environmental
protection measures, where appropriate, are being imple-
mented earlier at more facilities, thereby reducing the
potential threat of exposures to contaminants, and degra-
dation of resources.
By implementing stabilization measures at a facility, the
Agency may be able to lessen active oversight: In other
circumstances, stabilization could simply be a milestone
within a continuous remediation process. There may be
cases where a stabilization measure could be adequate
to serve as a final remedy for a particular release. Con-
sideration of the stabilization measure as a final remedy
would be based upon an evaluation of performance moni-
toring data collected after the measure was implemented.
When selecting a stabilization interim measure, compati-
bility and assimilation with final corrective actions should
be incorporated.
1.2.3 Stabilization as Part of the Corrective Action
Process
It should be understood that to a large extent, the stabi-
lization initiative builds on vehicles already existing/avail-
able in EPA Regions and States. Although stabilization is
a new RCRA initiative, it was not designed to create a
new regulatory or administrative process. Stabilization
measures will be implemented through the existing proc-
ess described above. The corrective action activity most
relevant to this initiative is interim measures. However,
conditional remedies and voluntary corrective action by
the owner/operator also may be helpful in implementing
this initiative.
Regions have already required a large number of facility
owner/operators to undertake interim measures to ad-
dress obvious environmental problems, particularly where
actual or imminent exposure of human or environmental
populations has been identified. Interim measures, as
described in the proposed Subpart S corrective action
rule, may be conducted at a facility whenever the Agency
determines that a release, or potential release, poses a
threat to human health or the environment. Examples of
releases that may be candidates for stabilization include
those that pose threats to human populations, ecosys-
tems, and/or drinking water sources.
Along with interim measures, conditional remedies and
voluntary actions by owner/operators also may be used
under the stabilization initiative. Conditional remedies are
intended to phase in remedies over time and, therefore,
may include stabilization activities to control the future
migration of wastes on site and to expedite cleanup of
releases that have migrated beyond the facility boundary.
Voluntary corrective actions may be conducted by facility
owner/operators who wish to initiate stabilization activities
rather than wait for EPA to begin actively pursuing cor-
rective action at the facility. In addition, owner/operators
may wish to proceed with corrective action in an attempt
to reduce their liability.
-------
Thus, this stabilization initiative will adopt and institution-
alize existing techniques (e.g., interim measures) ancl
management approaches (e.g., voluntary cleanups) to
implement the corrective action program. While this sta-
bilization initiative builds upon ongoing activities, the na-
tional program is adopting the philosophy that, in the near
term, there are increased environmental benefits associ-
ated with taking stabilization actions at more facilitieis
compared to pursuing final, comprehensive remedies at
fewer facilities. This philosophy, however, does not re-
move the goal of final cleanup of the facility. It does bias
remediation to the earliest possible phases of corrective
action.
Procedurally, stabilization will typically involve an initial
evaluation of RFA information to establish the need for
stabilization. Subsequent information gathering during thje
RFI should be focused on supporting technical decisions
regarding the stabilization approach chosen, and imple-
menting the stabilization response, typically under an "in-
terim measure" authority. In addition, information gleaned
from stabilization actions is envisioned to supplement the
ongoing RFI and support additional corrective actions.
Stabilization measures for any media of concern may be
undertaken separately from the final corrective action or
as an incremental step toward a final remedy. However,
all stabilization projects should be designed to be com-
patible with any anticipated long-term corrective action!
1.3 IMPLEMENTING STABILIZATION !
1.3.1 Overview |
A stabilization strategy may be considered or imple-
mented for either a specific SWMU, a group of SWMUs,
or an entire facility. Stabilization activities, while address-
ing releases from one or more SWMUs, are likely to
concentrate on a specific environmental medium, such as
ground water or soil.
If stabilization is imposed by EPA or the authorized state,
it will be implemented through a six-step process. Trie
first step is the assignment of a priority level for corrective
action and stabilization based on the National Corrective
Action Prioritization System (NCAPS). The second step
is the evaluation of the facility and its contaminant re-
leases to determine whether they can be addressed
through stabilization activities. Third, the Agency may
have to amend a compliance schedule in either an en-
forcement order or permit to include clauses requiring
stabilization. If a permit or order has not yet been issued,
the Agency may have to issue one to provide an enforce-
able agreement outlining the upcoming stabilization ac-
tivities. The fourth step is the collection of data needed
to select and design the stabilization measures that wjll
be used. This data collection will normally be carried out
as part of the early phases of the RFI. The fifth step jis
the selection and design of the stabilization measures. In
most cases, the owner/operator will propose technologies
or techniques for approval by EPA or the state. The final
step is the implementation of the stabilization measure.
An overview of this six-step process is shown in Figure
1-1 and described in greater detail below. It should be
noted that in the case of a voluntary stabilization action
(i.e., one that is initiated by the owner/operator without
being required by EPA or the state), several of the steps
do not apply.
Step 1: Priority Setting
NCAPS was developed to assist in prioritizing TSDFs for
corrective action activities. NCAPS is designed to receive
and manipulate data for four migration pathways, each of
which is scored separately: ground-water releases, sur-
face water releases, air releases, and onsite (e.g., soil)
contamination. A total facility score is calculated and fa-
cilities are ranked as high, medium, or low priorities based
on their total and media-specific scores.
NCAPS is simpler and less data intensive than the Super-
fund Hazard Ranking System (HRS). However, NCAPS
and the HRS consider many of the same ranking factors,
including the history of hazardous waste releases, area
hydrogeology, risks of continuing and sudden releases,
waste types and quantities handled, and likelihood of hu-
man and environmental exposure to hazardous waste and
hazardous constituents through direct contact, air, surface
water, and ground water. The data needed to perform the
priority setting should be available after the RFA or its
equivalent (e.g., a Preliminary Assessment [PA] con-
ducted at storage and treatment facilities under the Envi-
ronmental Priorities Initiative) is completed.
Step 2: Evaluation
Stabilization should be implemented at a particular facility
if it offers clear advantages in terms of protecting human
health and the environment in a timely manner. Identifying
these advantages involves a series of policy and technical
decisions, the most important of which are outlined in
Table 1-2. These judgments are difficult to quantify and,
therefore, must be based on the professional judgment
of the federal and state officials responsible for imple-
menting the RCRA corrective action program. The sum
of these judgments, as a group, should form a basis upon
which the relative benefits to be gained through stabiliza-
tion at a particular facility are evaluated. The types of
benefits envisioned from facility stabilization include lim-
ited contaminant migration, reduced volume of contami-
nated media, lowered risk to human health and the
environment, opportunity to pretest remediation technolo-
gies, reduced remediation costs, and a lower level of
EPA/state oversight of owner/operator's corrective action
activities.
-------
Table 1-2. Key Decisions for Determining the Need for
Stabilization
Have interim measures, if required or completed, been suc-
cessful in preventing the further spread of contamination
at the facility?
To what media have contaminant releases from the facility
occurred or been suspected of occurring?
Are contaminant releases migrating off site?
Are humans currently being exposed to contaminants re-
leased from the facility?
Is there a potential for human exposure to the contaminants
released from the facility over the next 5 to 10 years?
Are environmental receptors currently being exposed to con-
taminants released from the facility?
Is there a potential that environmental receptors could be ex-
posed to the contaminants released from the facility over
the next 5 to 10 years?
If already identified or planned, would final corrective meas-
ures be able to be implemented in time to adequately ad-
dress any existing or short-term threat to human health
and the environment?
Could a stabilization initiative at this facility reduce the pre-
sent or near-term (e.g., less than 2 years) risks to human
health and the environment?
If a stabilization activity were not begun, would the threat to
human health and the environment significantly increase
before final corrective measures could be implemented?
Are appropriate stabilization technologies available to prevent
the further spread of contamination, based on contaminant
characteristics and the facility's environmental setting?
Has the RFI, or another environmental investigation, pro-
vided the site characterization and waste release data
needed to design and implement a stabilization activity?
Can stabilization activities be implemented more quickly than
the final corrective measures?
Can stabilization activities be incorporated into the final cor-
rective measures at some point in the future?
Stabilization should be considered an option at a facility
up until the point where it becomes more expedient and
cost effective to implement the final corrective measures.
Generally, the immediate implementation of final cprrec-
tive measures, rather than stabilization measures, be-
comes more efficient after the RCRA CMS is completed,
because the effort and resources that might be used to
plan, design, and construct stabilization structures may
be more effectively spent implementing the corrective
measures. Furthermore, if final corrective measures are
just about to be constructed, it is unlikely that short-term
stabilization measures could be implemented fast enough
to be more effective in reducing threats to human health
and the environment.
If interim measures have been implemented at a facility
and they have been successful in preventing the further
spread of contamination from all known releases, stabili-
zation may, in effect, already be completed.
The implementation of stabilization measures is, of
course, dependent on the availability of appropriate tech-
nologies and techniques. If there are no identified tech-
nologies appropriate for stabilizing contamination at a
facility, stabilization is not a real option and the facility
should continue toward final corrective measures.
Step 3: Developing Enforceable Agreements
When it is determined that stabilization measures can and
should be required, Regions and states should use en-
forceable agreements to document the responsibilities of
all parties involved. Stabilization can be initiated either
through enforcement actions or through permitting proce-
dures. Generally, an order will be used if the facility is not
yet permitted and the permit issuance date is not in the
near future (i.e., within the next 6 months). A RCRA Sec-
tion 3008(h) order is the enforcement authority most likely
to be used, but a Section 7003 order could be used to
require stabilization if there is an imminent and substantial
endangerment. If the Region or state is in the process of
drafting a permit, the permit would be the most appropri-
ate vehicle for stabilization. If a permit has already been
issued, a permit modification to insert stabilization re-
quirements may be necessary.
Step 4; Data Collection for Stabilization
The data required to determine the need for stabilization
should be available after the RFA has been completed.
However, in many situations, data regarding the nature
and extent of the hazardous constituents that have been
released will not be available until the RFI is under way
or completed. Given that the selection of an appropriate
stabilization measure may depend on the collection of
sufficient site/unit characterization data, EPA encourages
the Regions and states to require owner/operators to de-
velop RFI workplans that gather such data during the
early stages of the RFI. This subject is discussed more
fully in Section 1.4.
Step 5: Selection and Design of Stabilization
Measures
After all of the data needed to design stabilization meas-
ures have been collected and evaluated, the owner/op-
erator should propose the technologies and techniques
he or she feels will be most appropriate to contain the
spread of contamination and/or possibly begin remediat-
ing the facility. In most instances where stabilization is
being carried out under a RCRA permit or enforcement
-------
EPA Region
priori&Mal
fauattie*.
Permitting Continues
Priority sites will be
engaged in normal
RFI and CMS activities
Wtvloboippropruie
lor tto tacn/7
Do not pursue
stabilization at
the facility
1) does the value of stabilization merit
the resources necessary to negotiate
with the reluctant otWer/operaior; and
bthe
own*r,'open!or
wiling to undertake
ttabJizi'.ion sctiorn In
addition to permit or
order actions?
2) does slabBizalion (and any required
public comment) prbduce results in a
significantly shorteritime than
standard RFI and CMS activities?
Decision Making
O Data Gathsrlng
Phate
Does
the facility have
adequate information
(RFI or other) to develop
interim measures to
achieve
stabilization?
Request that the
owner/operator
collect more or
better data (may
require permit mod
or order and
inclusion of
stabilization in the
RFI)
Figure 1-1. Stabilization strategy flowchart.
-------
Permittino Continues
Priority sites will be
engaged in normal
RFI and CMS activities
Owner/operator
develops interim
measures to
achieve
stabilization
/5
stabilizat
still worth pursuing
through othe
Do not continue to
pursue stabilization
at the facility
EPA/state may review planned interim
measures and either approve them or
work with the owner/operator to develop
satisfactory interim measures
Voluntary actions may not require
review
Construct and implement
interim measures to
achieve stabilization;
continue the RFI and CMS
on an Agency-approved
schedule
Are these
interim measures
still appropriate despite
the time and resources
necessary to obtain
required permits and
satisfy other
requirements?
Obtain necessary permits
and meet other
requirements
Figure 1-1. Stabilization strategy flowchart (continued).
-------
order, the owner/operator will be required to obtain either
EPA's or the state's approval for the proposed stabiliza-
tion measures. After approval has been granted, final
specifications and design drawings should be prepared
and any required permits should be secured.
Step 6: Implementation \
After stabilization measures have been designed and ap-
proved, they should be carried out in a timely manner po
that contaminant migration can be limited to the extent
possible. When stabilization measures are in place,! a
monitoring program should be used to ensure that the
measures are effective in controlling the contaminant
movement. This monitoring program will vary in sophisti-
cation depending on the nature of the stabilization meas-
ure. For example, in some instances, removal of drums
and contaminated soil may only require that surficial soil
samples be collected and analyzed to show that the con-
taminant source has been removed. However, in some
instances of widespread ground-water contamination,
monitoring wells and regularly collected samples may|be
required to show that contaminated ground water is not
bypassing whatever system has been built to contain it.
1.4 DATA CONSIDERATIONS FOR
STABILIZATION
1.4.1 Overview ,
The amount of information needed to support technical
decisions for stabilization will vary greatly. Obvious waste
removal situations might often be addressed more or less
immediately, without extensive studies, whereas groupd-
water contamination in a complex hydrogeologic setting
could require extensive investigations before an effective
stabilization remedy is chosen. Table 1-3 depicts some
of the key data elements that may be helpful in designing
some of the more common stabilization actions. Factors
that come into play when identifying data needs include:
• Immediacy of exposure threats.
• Types of contaminants and volumes of releases. ;
• Technical complexity of remediation.
• Media-specific characteristics, such as site hydrogeol-
ogy or prevailing wind direction. !
1.4.2 Phased Investigations
The broad spectrum of situations that may exist at i the
beginning of the RFI process, along with the need to
focus on releases that can be stabilized, calls for a flex-
ible, phased approach to site investigations. A phased
approach begins with an evaluation of existing data, the
development of a conceptual model of the extent of con-
tamination, and identification of possible stabilization
technologies to remedy the situation. This is followed by
the collection of additional data, as necessary, to charac-
terize releases and environmental settings, and to narrow
down the possible stabilization alternatives. From such
data, the conceptual model of the releases can be refined
and used to design an appropriate stabilization program.
In a phased investigation for stabilization, the initial
phases should focus on the SWMUs presenting the most
imminent threat to human health and the environment.
After stabilization for the high priority units has been im-
plemented, the focus of the RFI should shift to charac-
terizing any lower priority units and releases at the site.
1.4.3 Integrating Stabilization into RFI Workplans
In some cases, existing RFI work schedules and require-
ments may have to be revised to accommodate data
collection for stabilization. This is not likely to pose a
problem at facilities where an RFI workplan is currently
being negotiated through either a permit or an enforce-
ment order. Ideas and strategies for stabilization can be
readily incorporated before the workplan becomes final.
Where existing RFI workplans and schedules of compli-
ance are not sufficiently flexible to accommodate stabili-
zation needs prior to completion of the RFI, the potential
benefits of stabilization should be weighed against the
time and effort needed to make the modifications to the
agreement.
The RFI workplan should provide flexibility for data col-
lection and a phased approach for completing the facil-
ity-wide RFI. For some facilities, this may mean altering
the speed of the facility-wide RFI so activities can be
focused on the SWMU(s) undergoing stabilization. Al-
though facility-wide RFls are necessary, they should be
the last step in data gathering for a facility and should
occur after stabilization actions for the worst SWMU re-
leases have been assessed and imposed.
After stabilization measures are taken and imminent re-
leases are contained successfully, the RFI can continue
to address additional releases. Because the worst re-
leases will be stabilized, more time will be available to
work toward appropriate corrective measures to address
less environmentally significant releases.
1.5 REFERENCE
U.S. EPA. 1992. U.S. Environmental Protection Agency. Focusing
RCRA facility investigation (RFI) data collection for RCRA
stabilization. Draft. July.
-------
Table 1-3, Stabilization Data Needs
Depth to Water Table
Surface Water/Ground
Water Relationship
Ground-Water Flow Rates
and Direction
Seasonal Changes in
Drains
and Cap-
Trenches ping
• Q
•
B
• Q
Runoff/ Gas
Slurry Runon Vent- Solidifi-
Wall Control ing cation
• — Q •
Q _ _ _
" - - ~
— ' — Q Q
In
Situ
Soil
Flush- Bioreme-
ing diation
Q Q
Q —
a —
*
Q Q
Vacuum Pump
Extrac- and
tion Treat
• , •
•
•
Q •'
Ground-Water Elevation
Hydraulic
Conductivity/Permeability
Climate/Precipitation
Contaminant Characteristics
Contaminant Concentration
Extent of Contamination
Types, Thicknesses, and
Extents of Saturated and
Unsaturated Subsurface
Materials
Soil Characteristics
Soil Water Content
Depth of Air Permeable
Zone
Topography
Depth of Aquitard
•
Q
Q —
I High Priority; Q Medium Priority; — Low Priority
-------
-------
CHAPTER 2
Conceptual Approach for Characterizing
RCRA Contaminated Facilities
Ronald C. Sims
Division of Environmental Engineering
Utah State University
2.1 BACKGROUND
The conceptual approach to characterizing a contami-
nated site refers to the method by which site information
is generated and interpreted. Many times under both the
Resource Conservation and Recovery Act (RCRA), in-
cluding stabilization, and the Comprehensive Environ-
mental Response, Compensation, and Liability Act
(CERCLA) programs, the scope of information gathered
will determine the solution to the problem. In contrast,
stabilization goals are used to direct the extent of infor-
mation gathering and remedial action. In the past, one
major conceptual approach has been the inherent faith
that enough information will inevitably lead to a problem's
solution. The experiences of more than 10 years of the
CERCLA (Superfund) program, and several years of
RCRA, however, have often shown that more information
leads to additional confusion. This approach has in many
cases led to the investigative phase lasting longer than
4 to 6 years, delaying actual remediation of the facility to
even later in the process (more than 8 years). Without
understanding interactions among soil, contaminants,
and ground water, however, it is not always clear what
information to ask for and what to do with it.
Collecting large amounts of information is an approach
also found in case studies. The investigators will some-
times gather empirical evidence until a solution is found
through trial and error. The final conclusions reached can
be very useful, especially in providing a starting point for
an investigation of a similar site. If the study found that
a combination of pumping and treating and bioremedia-
tion cleaned up ground water in one situation, investiga-
tors might try the same thing at another site. The idea
behind the conceptual approach proposed in this chapter,
however, is to tie the problem identification and the rem-
edy selection/implementation phases together based on
the most scientifically appropriate stabilization strategy.
The following proposed conceptual approach offers a
"tool" to characterize a problem in two stages:
• Ask questions that will direct the investigation toward
obtaining the types of information necessary to charac-
terize the problem at a site.
• Select stabilization technologies that will address the
problem as it has been defined.
The mass balance conceptual approach is a way to ask
these questions. To arrive at an appropriate stabilization
solution, investigators and remediators need to be able
to talk across disciplines, for example, chemist to hydro-
geologist, in the context of the qualitative mass balance.
The next section, Section 2.2, describes the chemical
mass balance concept. Section 2.3 presents a method-
ology for applying the mass balance approach.
2.2 THE CHEMICAL MASS BALANCE
The chemical mass balance approach attempts to char-
acterize the chemicals at a site and the site itself, and to
define the problem in terms of where the contamination
actually occurs in the subsurface arid in what phase. After
the removal of surface sources such as drums, rockpiles,
or tanks, the source of contamination to ground water is
often the subsurface. This subsurface might contain the
contaminant in a number of different forms coating the
soil, in the air, or in the water.
It is very important to match the treatment technology to
the appropriate phase. Pump and treat, for example, is
effective for the removal of dissolved contaminants, but
it is less effective for removing chemicals that are sorbed
or chemicals present as a nonaqueous phase held by
capillary tension or other physical phenomena in the ge-
ology matrix. Scientists and engineers often make the
mistake of claiming that a particular technology failed,
whether it is bioremediation, vacuum extraction, vitrifica-
11
-------
tion, or some other method, when, in fact, they were using
the technology in a situation where it was not applicable.
2.2.1 A Hydraulic Mass Balance Model ;
Hydraulic mass balance for site remediation is basedjon
the movement of water onto and off of a site. Figure 2-1
Mass Balance
Residual •
U-^AAA^UUUUUUkJVA/JUl ^^^
D fT V
(a)
LNAPL
(b)
DNAPL
LT
(c)
Dissolved
Mass Balance
Challenges & Problems |
k -.» •.-..••.,,? "fr 1
J| » ^ ** . f-.\ < .II I
^^^^•^ F ^ • • -. - • -u * \ |
(e)
Figure 2-1. A bathtub model of a mass balance showing
water and various chemicals entering a site and their
phase transformations. [
represents a mass balance by showing water and various
chemicals entering a site (using a bathtub model) and
their phase transformations. If oil is added to the waiter,
it floats on top of the ground water. This floating product
is known as a light nonaqueous phase liquid (LNAP,L),
because it is less dense than water. In this example, 95
percent of the chemical associated with the LNAPL is on
top of the water (Figure 2-1 c). As the chemical constitu-
ents slowly dissolve from this LNAPL in very minor
amounts into the water, the contaminant may eventually
exceed the maximum contaminant level (MCL). Th'us,
even though 95 percent of the mass is LNAPL, the ground
water still is contaminated. The same scenario applieslfor
dense nonaqueous phase liquid (DNAPL) that migrates
downward because of its gravity-driven flow properties
(Figure 2-1d). If the contaminant can be addressed at the
subsurface source, it might be the best possible strategy
to stabilize the site, thus preventing the further spread of
contamination into the ground water. Physically contain-
ing the LNAPL or DNAPL, or hydraulically containing the
LNAPL or extracting the free phase through product re-
covery, may serve as a stabilization measure to further
prevent or minimize dissolved phase ground-water con-
tamination and migration.
If the chemical of concern, such as nitrate, methanol,
ethanol, or acetone, completely dissociates in water (Fig-
ure 2-1 e), then removing the water is going to effectively
remove the chemical. For a chemical predominantly in
the water phase, it is appropriate to stabilize the water
keeping in mind that some portion will still be adsorbed
to the geologic matrix. If, however, chemicals exhibit a
greater propensity to sorb to soil and form a residual, as
in Figure 2-1 b, this material will be trapped or sorbed in
the surfaces of the sand, silt, and clay. Cleaning out the
water, or "flailing the water phase" (a term used by sani-
tary engineers to describe measures that appear to ad-
dress a symptom but fail to alleviate the problem), will
not stabilize the site. When the tub is refilled, the chemi-
cals will desorb or migrate from that residual soil just
enough to exceed the MCLs again (Figure 2-1 e). In this
case, the residual as well as the water phase needs to
be stabilized.
An important step in choosing a technology to stabilize a
site is to determine which phase(s) needs stabilization or
can effectively be stabilized. Once this has been estab-
lished, the best technology or combination of technolo-
gies can be selected to stabilize the residual, water,
LNAPL, DNAPL, or various combinations of phases that
might occur.
A real-life example of the situation demonstrated in Figure
2-1 d occurs at sites with creosote contamination. Con-
tamination sinks to the bottom of the aquifer, and water
is contaminated with chemicals such that the concentra-
tions exceed the MCLs, because the source, which is a
DNAPL, dissolves into the water phase. The water phase
contamination is a symptom at most sites of the presence
of other phases that hold, in many cases, more than 90
percent of the chemical mass.
On the other hand, volatile chemicals such as trichlo-
roethylene (TCE) at a site might, to some extent, move
out of the water into the air or gas phase (as in Figure
2-1 f). If they move and are stopped by a horizontal
layer, either manmade or natural, they may condense
and dissolve into the water phase at another location not
previously contaminated, thereby spreading the contami-
nation further.
The goal of stabilizing sites is to control, minimize, or
.abate threats to human health or the environment from
chemical releases. This approach aims to minimize the
further spread of contamination by addressing the phases
of the contaminant. Thus, in characterizing a site, it is
important to ask in what phases the contaminant may be
12
-------
present in the subsurface which may act as sources for
ground-water contamination....
2.2.2 Transformation Potential
Another question to ask during the initial characterization
stage is whether the phases that currently exist in the
subsurface have the potential to be transformed. If trans-
formation does occur, the remedial engineer must deter-
mine into what phase the product would be transformed.
Figure 2-2 shows another way to draw a mass balance.
Chemicals get into both unsaturated and saturated soil
Fluid Phase
Volatilization
Mineralization
Biomass
Hazardous
Contaminants
Soil Interactions
Phases: Solid
Liquid Gas
Intermediate
Products
Leaching
T
Figure 2-2. Fate of hazardous contaminants in the sub-
surface (Sims et al., 1989).
zones, and may be water soluble and leach to the ground
dissolved in water or as free product. Once the chemicals
reach the ground water, they will move through the aquifer
in the dissolved water phase or as free product. Some-
times those chemicals will transform into intermediate
products, such as in the familiar example of TCE degrad-
ing to vinyl chloride under anaerobic conditions. There-
fore, the investigation must determine where this
transformed product, the vinyl chloride, will go in the sub-
surface. Vinyl chloride has a higher tendency to be a gas
than does its parent compound, TCE, and will tend to
volatilize. Vinyl chloride and TCE also are soluble in water
and therefore will be in the leachate phase.
2.2.3 Subsurface Mass Balance and Phases
Figure 2-3 illustrates the mass balance conceptual ap-
proach for the subsurface. In this diagram, the subsurface
is divided into two phases: the solid phase and the fluid
phase. The solid phase has an organic matter compo-
Solid Phase
Figure 2-3. Mass balance conceptual approach (U.S.
EPA, 1991).
nent, naturally occurring humus in the soil, and inorganic
matter composed of sand, silt, and clay. The solid phase
generally does not move relative to the fluid phase. Fluid
phases, including aqueous and nonaqueous gases, will
move under some type of pressure, suction, or force in
the subsurface.
Often, with regard to sources of drinking water, the water
compartment is contaminated because the sol]d and gas
compartments are contaminated. Stabilizing the site, or
preventing the further spread of contamination, may
mean addressing one or more of the other phases in
order to clean up the water phase. This brings into play
a whole different way of looking at cleanup technologies,
in terms of their ability to clean up not just the phase of
immediate concern, but the other phases, such as the
organic, the inorganic, the nonaqueous, or the gas
phases, that may be contributing to the problem.
In this regard, the mass balance approach offers a way
to ask questions, not an answer. A mass balance ap-
proach still will not allow the identification of 100 percent
of the chemicals in different phases in the subsurface but
can be used to characterize which chemicals will tend to
be in which phases. The cleanup technology then can be
evaluated based upon its ability to control whatever phase
is responsible for the contamination. The mass balance
conceptual approach attempts to link the subsurface and
surface systems.
2.2.3.1 Solid Phase
Organic Matter (Humic Material)
The organic matter content can be useful for defining the
problem at a site with regard to transport, and thus for
evaluating treatment options. Figure 2-4 illustrates a theo-
retical composition of generic organic matter: benzene
rings, aromatic structure, and many straight chain carbon
compounds. Organic matter is chemically complex and
serves as a solvent for organic chemicals. The more a
chemical added to the soil structurally looks like this or-
ganic material, the more it tends to act as a solvent for
13
-------
COOH
HOOC
COOH
H H H2 | H H ' H H H2 » H
N-C-C-C-N-C-C-N—C — C-C-N
H2 H2 H2
c-o-c-c-c-
COOH
OCH3
CH3
Figure 2-4. Composition of organic matter.
the contaminant. A rule in organic chemistry is "like dis-
solves like." For example, if one tries to pour gasoline
into a styrofoam cup, the cup dissolves in one's hand,
because both materials are structurally similar. In a sim-
ple system, such as soil material surrounded by water,
the more the chemical resembles organic matter, the less
it will tend to be in the water phase, and the more difficjult
it will be to clean up the site with pump and treat. There-
fore, knowledge of the organic matter compartment ckn
relate directly to the efficacy of a treatment technology.
Inorganic Matter (Texture)
The texture of soils (inorganic material) can be an impor-
tant consideration in selecting a treatment technology.
Texture influences permeability to fluids, and influences
the storage and sorption of contaminants. Figure 2-5, is
a standard soil classification texture triangle. Gravel j is
made up of relatively large pieces of material; sand, silt,
and clay are composed of increasingly smaller particles.
As the size of the particles shrinks, the interstices be-
tween the particles shrink as well, and surface to volume
ratjo increases; this is accompanied by an increase in the
capillary pressure, called capillary rise, and an increase
in porosity. In clay environments, this can result in the
storage of large amounts of water in the vadose zone.
Figure 2-6a shows the values for capillary rise in various
soils. If a chemical in the water phase is stored in the
vadose zone, the vadose zone can become a source of
contamination to the saturated zone.
Texture also can indicate the porosity of a material, or the
extent of internal pores. Figure 2-6b illustrates the per-
centage porosity of different textures (clays, sands, grav-
els, and silts) ranging from 10 to 60. Clay has a very high
porosity, and because of its internal pores can store a lot
of pollutants. Also, because clay has a low saturated
conductivity, it is difficult to push any fluid, or nonaqueous
phase liquid (NAPL), through it to recover product. Per-
meability, or hydraulic conductivity, indicates the relative
ease with which fluids move, including water with nutri-
ents (Figure 2-6c). Bioremediation, vacuum extraction,
and pump and treat have limited success when remediat-
14
-------
Using materials less than
2.0 mm in size. 11 approximately
20% or more of the soil material
is larger than 2.0 mm, the texture
term includes a modifier.
Example: gravelly sandy loam.
Example of use:
A soil material with
35% clay, 30% silt,
and 35% sand is a
clay loam.
30% Silt
100 90 80 70 60 50 40'
•30. 20 10
100
Percent Sand
35% Sand
Figure 2-5. U.S. Department of Agriculture soil textural
classification (texture triangle) (Soil Conservation Service,
1971).
ing clay, because clay has a much lower hydraulic con-
ductivity than gravel, sand, or silt. Because clays have a
higher storage capacity and lower ability to transmit fluids,
their presence in the subsurface at a site will provide a
challenge for site stabilization or site cleanup.
2.2.3.2 Fluid Phase
Nonaqueous Phase Liquids
The characteristics of NAPLs include their free phase
flow, residual saturation, and ability to retain chemicals
that contaminate air, water, and soil through distribution
among compartments in the subsurface. Some examples
of LNAPLs are oil and pentachlorophenol (PCP) in oil;
PCP alone (without an oil carrier) is a DNAPL, as are
creosote, methylene chloride, and many other chlorinated
solvents. Knowing whether a chemical is a DNAPL or an
LNAPL gives an indication of where to look for it in the
subsurface.
Gas
The gas compartment is an essential part of site charac-
terization and can be useful in characterizing transport
from a site. Depending on their density relative to air,
chemicals in the gas compartment may move upward
toward the atmosphere or downward toward ground water
Values tor Capillary Rise
in Various Soils
20—1
15—
10—
Clay
Sand
Clay
i r
10 20
30 40
(b)
50
60
Silt
K crrVsec
Sand
:; Gravel
I I
10'10 1(
I I
1 \^
1 102
(c)
Figure 2-6. Values for capillary rise (a), porosity (b), and
permeability (c) in various soils (Dragun, 1988).
in the subsurface. If the chemicals move upward and
horizontally, they can increase the threat to human health
and the environment. If they move downward, they can
15
-------
further the spread of contamination into the ground water
and beyond a site. Understanding the gas compartmept
Is useful for evaluating an approach to treatment. For
example, vacuum extraction is effective for controlling or
minimizing the gas phase, whereas many other treatment
systems are not. Treatment of the gas phase with biore-
mediation, for example, can be deceptive because vola-
tilization is sometimes mistaken for biodegradation, since
the contaminant seems to have disappeared.
Water !
Capillary forces, as a function of texture, facilitate tlje
ability of soil to store water, contaminate it, and then
return the water back to the ground-water source through
precipitation. The metric potential or suction in the soil
influences the movement of water and chemicals dis-
solved in the water. It also can indicate storage of con-
taminants and sources of contamination to ground 'water.
2.2.4 Distribution Coefficients
i
The distribution coefficient (Kd) is the concentration of a
chemical adsorbed onto the organic and inorganic solid
phase, compared to the concentration of the chemical in
the water phase:
The higher the Kd, the more the chemical tends to be
adsorbed on the solid phase; therefore, the less efficient
pumping and treating the water will be. If a contaminant
has a high Kd, it is going to associate with the soil phase.
The problem then remains whether to destroy the chemi-
cal in the soil phase or move it from the soil phase into
the water phase using a solvent or surfactant. The lower
the Kd, the more the chemical tends to be in the water
phase, and the better pumping and treating will work|to
stabilize contamination. This, however, should not p^e-
Characterization
Site
Soil
— n
— i
Waste
__i
elude the use of pump-and-treat systems in lower Kd
systems to stabilize the facility by mitigating contaminant
migration.
For a NAPL, this equation reads:
with Co being the concentration of the chemical in the
nonaqueous phase, compared to its concentration in the
water phase. The higher the K0 the more likely the chemi-
cal is to be found in the nonaqueous phase. Remediation
technologies for a chemical with a high K0 should focus
on recovery of the product NAPL.
Finally, the concentration of the contaminant in air versus
water is represented by:
Ca/Cw = Kh
If a chemical has a high Kh, it has a greater affinity to
partitioning into the air phase. A chemical that has a low
Kh tends toward being in the water phase. Vacuum ex-
traction or venting would be a technology to consider for
a chemical with a high Kh.
Distribution coefficients indicate the tendency of a system
to move towards equilibrium by showing in what phase
the chemicals tend to be. This tendency toward equilib-
rium is, conceptually, the reciprocal of fugacity, a term
that implies the tendency of a chemical to escape from a
particular phase.
2.2.5 Migration/Transport
Figure 2-7 shows a flow diagram of the methodology for
integrating site characterization with subsurface remedia-
tion. Investigators need to determine how the waste and
soil interact, whether there is partitioning to the soil in the
gas phase or in the nonaqueous phase, and what type
Constituent
Attenuation
Required
j
k
Distribution
Reaction
Migration/Escape
Exposure
w
w
Jreatment
technique
Selection
w
Constituent
Attenuation
fe
Monitoring
Problem Assessment Treatment (train) Monitoring
Figure 2-7. Methodology for integrating site characterization with subsurface remediation (U.S. EPA, 1990).
16
-------
of cffmafe and precipitation are present at the site. This
information will help investigators determine the migra-
tion and transport potential of contaminants in different
phases.
Table 2-1 presents some of the most important properties
and parameters needed to perform a soil-based waste
Table 2-1. Important Properties and Parameters Needed
to Perform a Soil-Based Waste Characterization
Physical Properties
Density
Viscosity
Chemical Properties
Molecular Weight
Melting Point
Specific Gravity
Structure
Water Solubility
Vapor Pressure
Chemical Class
Acid
Base
Polar Neutral
Nonpolar Neutral
Inorganic
Photodegradation
Chemical Reactivity
Oxidation
Reduction
Hydrolysis
Polymerization
Precipitation
Biotic/Abiotic
Soil Degradation
Parameters
Half-life (ti/2)
Rate Constant
Loss of Parent Cpd
Mineralization
Intermediates
Volatilization Parameters
Air/Water (Kh)
Vapor Pressure
Soil Sorption Parameters
Cation Exchange Capacity
Anion Exchange Capacity
Soil/Water (Kd)
Octanol/Water (K0)
Soil Contamination
Parameters
Concentration in Soil
Soil Horizonation
Depth of Contamination
Adapted from: U.S. EPA (1984).
characterization. Most of this information is contained in
the Corrective Action Plan (CAP). Using this information,
a chemist will be able to determine whether a given
chemical oxidizes or reduces, and in what phases and
where the chemical is likely to be found. PCP, for exam-
ple, is an acid with a maximum dissociation at a pH of
4.7. If there is PCP in a soil, the pH will indicate in what
form (salt or free acid) the PCP will be found. If it is in
the ionized form, pH above 4.7, it is going to be more
water soluble. In the western part of this country, where
soil tends to have a higher pH, the PCP is going to be
more ionized, and thus more water soluble. In the eastern
part of the country, however, the soil pH may be as low
as 4 or 5. In that case, the PCP is going to be sorbed to
soil and/or occur in the nonaqueous phase, i.e., in the oil,
if it is present.
2.2.6 Management
Using a knowledge of chemistry, the remedial manager
first can predict in what phases the different chemicals
will occur, and then select a treatment train based on
those phases. All treatment technologies can be simpli-
fied into three categories:
• Containment technologies
• Extraction technologies
• Destruction technologies
Figure 2-8 offers a basic scheme to describe these basic
Contain
Hydraulic
0=C=O
Destroy
Biological
Chemical
Thermal
Figure 2-8. Treatment train approach.
17
-------
categories. Within these categories, however, are hun;-
dreds of individual processes. |
2.2.6.1 Containment
Containment might be the first step to consider in a sta-
bilization action (Rgure 2-8a). Pumping and treating i^
being studied both as an effective method for containing
the aqueous and the nonaqueous plumes and as k
method for cleaning up a site. Containment could be
hydraulic as well as physical, such as constructing a
slurry wall or using a temporary clay or synthetic cap|.
Chapter 4 will discuss capping and clay liners in more
detail. Temporary containment is also an option, espe-
cially if the remediation team is examining longer term
corrective action remedies for eventual site cleanup anjd
restoration. i
Containment systems can get very, large, such as thp
slurry wall in Laramie, Wyoming, which is 30 to 70 feet
deep and 10,000 feet long. Systems also are available
for combining hydraulic with physical containment, whicp
can be followed up with an additional longer term reme-
dial technology.
2.2.6.2 Extraction
Figure 2-8b shows extraction technologies, including vac-
uum extraction (pulling material out of the soil through the
air phase) and product recovery. The vertically downward
arrows could represent leaching, soil flushing, or pumping
and treating where the material is used to move thje
water-soluble phase through the soil. !
I
2.2.6.3 Destruction ,
Many technologies attempt to destroy contaminants by
biological, chemical, and thermal techniques (Figure 2-
8c). Following the extraction phase in Figure 2-8b, the
contaminants extracted could be destroyed above ground
or destroyed in situ. :
2.3 METHODOLOGY
Figure 2-9 shows a schematic of a particular site that has
an unsaturated zone and a saturated zone. To charai>
terize for a stabilization action, it is necessary to put all
of this information together for the CAP. The key ques-
tions to ask are: ;
• What chemicals, and at what volumes, are at the site?
• What phases of contaminants are present at the sfte
(aqueous, nonaqueous, gaseous, sorbed)? ;
• Where are those phases found in the subsurface, both
in the unsaturated and saturated zone? !
• Where are the LNAPL and the DNAPL? (Both may tpe
caught in the unsaturated zone.)
The investigation must determine how to stabilize the
particular phases present for preventing the further
spread of contamination for particular chemicals.
Land Surface
Belt of Soil Water
Intermediate Belt
Capillary Fringe
Water Table
Ground Water
Figure 2-9. Contaminated site with an unsaturated and
saturated zone.
The four steps in the conceptual approach to stabilization
are:
• Characterization of chemical and site information
• Problem definition with regard to risk, risk assessment,
and exposure to populations and the environment
• Treatment train selection using both chemical and site
data to determine which combination of containment,
extraction, and destruction technologies will be most
effective at the site
• Monitoring the stabilization technology for each phase
of the chemical
2.3.1 Characterization
The two main types of data needed to select treatment
are chemical characteristics and site characteristics. Both
of these must work together; often a technology that
would work based on chemical characteristics might be
rejected because of conditions at the site and vice versa.
In terms of chemical characteristics, for example, the in-
vestigation needs to determine whether a contaminant is
an LNAPL or a DNAPL, in the gas or dissolved water
phase, or sorbed to soil. The site investigation should
reveal information about conductivity of the subsurface,
i.e., how the site allows fluids to move in and out.
2.3.2 Problem Assessment
The next major stage is problem assessment, in which
18
-------
remedialors determine which chemicals are in what
phases, and how those phases will move. In addition,
they need to assess how these chemicals in their various
phases pose risks to human health and the environment
(i.e., an exposure assessment). Once the remedial team
determines whether the population is being exposed
through the air, through pure product, through the
leachate, or through some other medium, they can select
a treatment train, the next stage of the process. The
cleanup goals at a site also are part of the problem
definition.
2.3.3 Treatment Train Selection
Generally, more than one phase at a site requires treat-
ment, so a train instead of a single treatment technique
is usually necessary. If a site has only one contaminant
in only one phase, treatment is much less complex. For
example, at one industrial site, formic acid was the only
contaminant. The partition coefficient for formic acid re-
vealed that it tended to occur only in one phase, the water
phase, so treatment was relatively simple.
As mentioned above, three approaches—containment,
extraction, destruction—form the basis of the treatment
train. Technologies are selected that perform one or more
of these functions. An example of a stabilization train
might combine a slurry wall and hydraulic containment,
vacuum extraction, soil flushing, and bioremediation in a
configuration based upon the phases present at the site.
Figure 2-10 shows a treatment train that begins with prod-
uct removal, then uses pump and treat and soil flushing,
Treatment Train
Product Removal
Pump & Treat
Chemical Characteristics
Site Characteristics
Soil Flushing
Bioremediation
i
Chemical Characteristics
Site Characteristics
Chemical Characteristics
Site Characteristics
Figure 2-10. Treatment train from product removal to
pump and treat, soil flushing, and bioremediation.
and finishes with bioremediation. Hydraulic and physical
containment might be established first to stabilize the site,
followed by product recovery. Pump and treat could be
used to pull the chemicals out of the site that are in the
water phase. Soil flushing, with or without water, would
help remove chemicals in the unsaturated zone that are
soluble in the solution used, but are partitioned to soil. A
surfactant or solvent could be used to decrease the par-
tition coefficient (Kd), thus increasing the tendency of a
particular chemical to go into the solution phase so it can
be pumped out. Flushing moves the chemical from the
solid phase to the solution phase. Bioremediation is the
last stage of the treatment train, the "caboose," and not
generally considered as a front-line stabilization tech-
nique. Bioremediation might be best applied to remediate
whatever cannot be extracted from the soil.
2.3.3.1 Pump and Treat
When considering pumping and treating for stabilizing
plume migration and for source (hot spot) removal, the
investigator needs to know the characteristics of the
chemicals present that tend toward being in the water
phase. Site information would be used to determine
whether the site will allow water to be pushed or pulled
through the material. Based on chemical characteristics,
some contaminants are amenable to pump and treat, but
if the site is largely clay, or has a lot of clay stringers, this
technology may be less effective. The reader is referred
to EPA (1990) on pump and treat technologies.
2.3.3.2 So/7 Flushing
Those chemicals that are more tightly bound to or trapped
in the soil might yield better remediation results with soil
flushing. An important consideration is the extent that
solvents and surfactants partition the chemical out of the
solid phase or trap residual saturation into the solution
phase. The site characteristics again relate to the ease
with which water solvent or surfactants move in and out
of the site.
2.3.3.3 Bioremediation
The relative recalcitrance of a chemical to degrade is
important in bioremediation. Dioxins and PCBs, for ex-
ample, are more difficult to degrade than methanol, etha-
nol, and some hydrocarbons. Chemical characteristics
provide information concerning the tendency of the
chemical to degrade under aerobic or anaerobic condi-
tions. Relevant site characteristics relate to the feasibility
of transporting nutrients and oxygen to the site. Bioreme-
diation is usually limited by the supply of nutrients, the
presence of toxic chemicals, and the supply of oxygen.
2.3.4 Monitoring Treatment Performance
After investigators determine the particular treatment
technologies that will be effective for ..specific phases and
begin implementation, then the investigation must monitor
particular phases for particular chemicals in an attempt
to optimize the technology application.
19
-------
For any stabilization measure, delivery and recovery are!
going to influence the success of application. The impact
of stabilization will be restricted or limited if fluids are(
unable to be moved into and out of the subsurface
through methods such as nonaqueous phase product re-j
covery, vacuum extraction in the air phase, or pump and
treat in the water phase.
2.4 REFERENCES
U.S. EPA. 1984. U.S. Environmental Protection Agency. Review of
in-place treatment techniques for contaminated surface soils.
Vol. 1 and 2. Risk Reduction Engineering Laboratory, Cincin-
nati, OH. EPA-540/2-84-003a,b.
U.S. EPA. 1990. U.S. Environmental Protection Agency. Basics of
pump and treat groundwater remediation technology. Labo-
ratory RSKERL Ada, OK. EPA 600/80-90/003.
i 20
-------
CHAPTER 3
Field Screening Methods
Harry Compton
Environmental Response Team
Office of Emergency and Remedial Response
U.S. Environmental Protection Agency
3.1 FIELD SCREENING REQUIREMENTS:
SCOPE AND EXPECTATIONS
Field screening refers to the use of rapid, low-cost test
methods to determine whether a contaminant of interest
at a given site is present or absent, above or below a
predetermined threshold, or in a concentration within a
predetermined range. The goal is to develop simple and
inexpensive field screening methods for organic and in-
organic analyses to support onsite monitoring and char-
acterization activities. Rapid, low-cost methods are
projected to be the mainstream techniques in stabilization
for characterization and monitoring. The rapid and real-
time capabilities of many of these methods allow for sta-
bilization to be implemented in the short term, at a
reasonable cost, and at an appropriate analytical level.
Figure 3-1 shows the chemical classes of 200 hazardous
substances that can occur at a site and be detected by
field screening methods.
Field screening methods would support stabilization by
helping to accelerate cleanup, instill confidence in
cleanup, save money, allow collection of real-time data,
achieve higher sample density, and effectively detect "hot
spots." Field screening aids site characterization by pro-
viding legally admissible quality data and rapid sample
analysis. It also can direct ongoing stabilization and cor-
rective action remedial field investigations (RFIs), help
assign priorities to sampling surveys and excavations,
and reduce the need for more expensive offsite analysis.
When performing field screening exercises, the investiga-
tor should 1) ensure the objectives of the study are clear
and 2) use the right tools to accomplish the stated objec-
tives. Before initiating field screening at a site scheduled
for stabilization, it is important to consider some basic
issues. The first decision the investigator must make is
where to. sample, taking into account the repre-
sentativeness of samples and sample integrity. The in-
vestigation then must define a sampling strategy.
To provide guidance and standardize procedures for field
screening, a Monitoring and Measurement Technologies
Program has been established. Its purpose is to identify
new or existing technologies; evaluate, demonstrate, and
report on performance; transfer technology; and develop
and distribute standard methods when required. The
types of technical support available include field studies,
data interpretation, and technical oversight including on-
site audits, analytical support, and technical review. A
Field Assessment Program also complements Agency
programs by providing quality field data with a minimum
of analytical waste. The EPA Environmental Response
Team (ERT) offers sampling quality assurance (QA) pro-
grams through Bill Coakley, who has developed Repre-
sentative Field Sampling Guidelines for various media
and techniques.
Field screening can provide data for several levels of
analysis depending on the required quality of the data.
Table 3-1 presents the data quality necessary for each
analytical level.
Table 3-1. Analytical Levels Relative to Data Quality
Analytical
Level
Data Quality
Level I
Level II
Level III
Level IV
Level V
If instruments are calibrated and data are
interpreted correctly, it can provide
indication of contamination
Dependent on QA/QC steps employed
Data typically reported in concentration
ranges
Similar detection limits to CLP
Less rigorous QA/QC
Goal is data of known quality
Rigorous QA/QC
Method specific
21
-------
Halogenated
Pesticides/Related
Compounds
(8.5%)
Nitrosoamines/
Ethers/Alcohols
(7.5%)
Reactive
Intermediates
PAHs
(8.5%)
Phenols/
Phenoxy Acids
(10.5%)
Miscellaneous
Benzidines/Aromatic
Amines
Organophosp hates/
Carbamates
Phthalates
(3%),
Inorganic Elements/
Radionuclides
(17.5%)
VOCs-
(26.5%)
Figure 3-1. Chemical classes of 200 hazardous substances.
Finally, QA is essential for effective field screening. At a
minimum, the following QA activities should be performeci
when conducting field screening: '
i
• Analyze blanks frequently. I
[
• Monitor retention time or detector response to identify
and/or compensate for instrument drift. I
• Adapt data quality objectives as well as QA/quality con-
trol (QC) procedures to site-specific conditions. These
cannot be standardized for field and laboratory. i
• Incorporate blind spiked samples into the analytical se-
quence for precision, accuracy, and comparability.
t
• Consider system and site-specific characteristics,
along with personnel experience, in setting study ob-
jectives. , |
• Send a percentage of the field analysis samples to ah
offsite lab for independent verification of results.
and instruments, and future needs and directions in field
screening. Section 3.2 presents analytical equipment
amenable to stabilization investigations, such as gap
chromatographs, gas chromatographs/mass spectrome-
ters, chemical sensor systems, and x-ray fluorescence
spectrometers; sampling methods for soil-gases and air;
and immunoassay methods. Section 3.3 outlines futur^
needs and directions and Section 3.4 is a bibliography.
3.2 FIELD TECHNIQUES AND INSTRUMENTATION
The Hach® kit is one of the oldest utilized field screening
techniques. The advantages of Hach® kits are that they
have a large library of analytes, are easy to use, and are
quick and relatively inexpensive. When using Hach® kits,
however, there may be interferences from other com-
pounds present. While Hach® kits are used largely and
designed for water samples, they have been adapted for
use with soils. Any chemical or laboratory instrument sup-
ply may have a Hach® catalog.
Other types of field screening devices useful for providing
site information for the implementation of site stabilization
include:
• Gas chromatographs
• Gas chromatographs/mass spectrometers
• Chemical sensor systems
• X-ray fluorescence spectrometers
• Sampling and analysis equipment for soil-gases and
The following sections describe actual field techniques air
• Immunoassay methods
Each of these types of devices is described in the pages
that follow.
3.2.1 Gas Chromatographs
Portable gas chromatographs (GCs) are very useful in
the field for analyzing concentrations of known contami-
22
-------
nanfs in soil and water. They are more complex and
expensive than some other equipment that will be dis-
cussed, but are useful for surveying for extensive con-
tamination. GCs have internal gas supplies, detectors,
and integration systems. Integration systems are impor-
tant for evaluating the concentrations of compounds in
soils or water. GCs provide very low detection limits, but
have a limited library of compounds that they can detect.
Many GCs have routine field applications and numerous
vendors; some units have been specifically designed for
the field, and are proving to be more accurate in field
analyses than are offsite lab analyses. There is, however,
a need for improved sensitivity and selectivity of appro-
priate equipment and for instituting standardized field
methods and QA procedures.
Two types of GCs—the purge and trap and the head-
space—are discussed below.
The purge and trap GC can analyze volatile organic
compounds (VOCs) in soil and water, with low Henry's
Law constants (Henry's Law constants are discussed in
detail in Chapter 6). Two examples of purge and trap GCs
are the argon ionization detector (AID) and the thermal
conductivity detector (TCD). The AID has an electron
capture detector (BCD), and is very useful for halogen-
ated compounds, such as trichloroethylene (TCE) and
especially for the chlorinated alkanes, which cannot be
seen by a photoionization detector (PID). A PID is an
instrument that detects and measures airborne, ionizable
gases and vapors. The AID also can detect benzene,
toluene, and other compounds that a PID can detect, and
it can perform analysis on water and soil samples. A
Scentograph® has a library of 16 compounds. The Micro-
monitor , a very quick field screening method, uses a
thermal conductivity detector (TCD). It is easily portable,
but unfortunately requires a concentrator. It has a large
library of compounds.
The Viking gas chromatograph/mass spectrometer
(GC/MS) is another purge and trap device (not in Table 3-2),
which is portable but more expensive than those dis-
cussed above. It detects nonpolar organics in the air, and
provides the highest level of certainty for identification of
different compounds.
The headspace GC has a higher detection limit than
does the purge and trap GC, unless a sensitive or selec-
tive detector is used. The Photovac®, a popular type, has
a high detection limit, and is used only for compounds
with high Henry's Law constants. The Photovac® has a
library of 25 compounds. EPA's ERT uses the Photovac®
in much of its field screening, soil gas, water, and soil
headspace work.
The Photovac® has a PID, which is sensitive to aromatics
and unsaturated aliphatic compounds, such as benzene
and styrenes. PIDs do not, however, detect methane
or alkanes. A PID will detect only substances that are
ionizable at or below the lamp intensity for the instru-
ment, and its built-in integrator can provide the relative
concentration.
Table 3-2 summarizes advantages and disadvantages of
two categories and three types of gas chromatograph
detectors.
When using GC analytical methods, it is helpful to deter-
mine the target compounds first, in order to select the
appropriate analytical tools and procedures. Choosing the
proper gas chromatograph for a specific situation involves
consideration of the following selection criteria:
• Detector Type—Sensitivity requirements and com-
pound class
• Injection System—Sample matrix
• Column Type—Sample components
• Oven Temperature—Sample components
• Data System—QA/QC and archiving requirements
Figure 3-2 provides some guidance on selecting a GC
detector based on volatile target compounds. A TCD has
two columns, and identifies a compound based on study
ing the resolution times on the two columns. It has a very
short retention time in the columns/and can provide re-
sults in 3 minutes. (Other GCs take 15 to 60 minutes,
depending on the number of blanks run between each
hot sample.) In addition to its unique retention time, the
thermal conductivity detector has a library of 100 com-
pounds. It works primarily in parts per million (ppm) lev-
els, but a concentrator can drive the detection limit down
to parts per billion (ppb) levels, which can be useful when
dealing with unknown contaminants. The disadvantage of
Double Bonds Halogenated General
Aromatics Compounds Compounds
Photoionization Electron Capture Argon Ionization
Detector (PID) Detector (ECD) Detector (AID)
Low ppb Low
Permanent Gases
Thermal Conductivity
Detector (TCD)
Low ppm
ppb Low ppb
General
Compounds
Flame Ionization
Detector (FID)
Low ppm
Figure 3-2. Selection of portable gas chromatograph de-
tectors based on volatile target compounds.
23
-------
Table 3-2. Summary of Three Gas Chromatograph Devices
Advantages
Disadvantages
PURGE AND TRAP j
AID • Data archiving via internal personal computer
• Responds to aromatics and chlorinated alkanes
and alkenes !
i
• Rugged; easily decontaminated |
• Internal adsorption trap allows sample i
preconcentration to enhance sensitivity!
Limited linear range (102 to 103)
Limited capillary column capabilities
TCD
• Fast run times '
• Detects broad range of compounds \
• Best detector for permanent gases (Oa, Na, CO,
CO2, and CKU)
• Isothermal heated column (30° to 180°C)
• Large compound library (100+) for screening
unknowns [
• Dual column for compound !
identification/combination ;
• Poor sensitivity (>1 ppm)
• Sophisticated external data system required
• External concentrator required if ppb levels
required , . •
HEADSPACE
PID
• High sensitivity (low ppb range) ;
i
• Self-contained and rugged ,
• Precolumn with backflush :
• Capillary column capability |
• Widespread use, well-developed methods
• Ideal for aromatics (BTXs), double bonds (PCE,
TCE, DCE) !
• Limited temperature control of columns
• Complicated external plumbing
• Prone to electronic interference
• High ambient temperature breakdowns
this device is that it is very expensive and, because it
requires a high vacuum, has the potential for a long down-
time. !
I
Selection of the oven temperature is determined by the
boiling points of the target compound. For compounds
with a stabilized ambient boiling point, within 20°C of the
ambient temperature, the Photovac® is the most apprb-
priate tool. Isothermal compounds, those with constant,
elevated temperatures of up to 300°C (boiling point |is
less than that of chlorobenzene) work best with the Scen-
tograph®. For situations in which there is a wide variation
in target compound boiling points (temperature pro-
grammed), a laboratory or advanced transportable instiju-
ment is recommended.
: I
3.2.2 Gas Chromatography/Mass Spectrometry ,
Gas chromatography/mass spectrometry is particularly
useful for its ability to measure concentrations of a broad
range of unknown contaminants. This equipment is mini-
aturized and rugged. Much research and development
has been done on these devices for the defense and
aerospace industries. EPA has conducted a demonstra-
tion of one available unit, and other instruments show
promise. Quality assurance and standardized field meth-
ods, as well as a standard data reporting format, are still
needed for this technique.
3.2.3 Chemical Sensor Systems
Chemical sensor systems are an emerging technology
that offers much promise in compound identification and
quantification. Current types of systems include surface
acoustic wave devices, fiber optics, spectroelectrochemi-
cal devices, solid phase sorption/spectroscopic devices,
and porous polymer and glass fibers. Some of these
available technologies, however, have limited environ-
mental application.
A few remote or stand-off sensor systems are available
commercially, two of which EPA is examining under the
24
-------
Superfund Innovative Technology Evaluation (SITE) pro-
gram. Two technologies are long path Fourier transform
infrared (FTIR) spectrometry for air and long path ultra-
violet (UV) spectrometry. Spectrometry uses the energy
source of a particular wavelength to activate the mole-
cules of a substance, thereby producing a spectrum. FTIR
spectrometry identifies unknown contaminants by using
infrared light to determine bonding between atoms, and
thus the molecular structures, of all contaminants present.
FTIR spectrometry evaluates and identifies volatile com-
pounds that pass through a beam stretching from the
device to a mirror and back. It can be operated at any
height, the important parameter being the distance be-
tween the source and the mirrors, which can range from
one to several hundred meters, the length of a substantial
fence line. This spectrometer also gathers virtual real-
time data. Because it provides results in minutes, FTIR
spectrometry provides more valuable information than do
discrete fence-line monitors like SUMA canisters or XAD
tubes, which have to be left in a location for a period of
time, and then sent to a laboratory for analysis. UV spec-
trometry uses ultraviolet light to determine molecular
bonding to ascertain if specific known contaminants are
present. Much research is still needed on these systems,
including a definition of their use in air pathway analysis,
standard methods and QA procedures, and comparability
to point samples.
Spectroscopic techniques using fiber optics are another
emerging technology that is the focus of much research
and development. Studies have been done on lumines-
cence, fluorescence, and absorption. Optical fibers have
the potential to be used for in situ monitoring. Prototype
units now exist, and the technology is beginning to be
commercialized. ;
An infrared temperature indicator, which can be in the
form of a gun-like device, can be used to estimate the
liquid volume in above ground tanks by taking a vertical
profile down the side of the tank. The change in tempera-
ture indicates the air-aqueous interface, from which the
volume estimate of product or liquid in the tank can be
estimated.
3.2.4 X-Ray Fluorescence Spectrometers
X-ray fluorescence (XRF) spectrometry works on the ba-
sis of a photoelectric effect, activating atoms of materials
in dry soil and counting the X-rays that are energized.
This device is in widespread use for determining the ex-
tent of soil contamination, selecting samples for offsite
analysis, guiding cleanup activities, and analyzing air
monitoring filters. An XRF spectrometer also can be
used for determining hot spots in "extent of contamina-
tion" surveys.
The XRF spectrometer can detect as many as 23 different
metal compounds. The detection of matrix effects using
this device is almost more important than the detection
of the compound of interest. A lot of sample preparation
is necessary to get a valid reading on the XRF spectrome-
ter. It is much more effective if it can be calibrated using
6 to 10 site samples of the same relative concentrations
analyzed by a standard method such as an atomic ab-
sorption (AA) or inductively coupled plasma (ICP) detec-
tor. After calibration, the XRF spectrometer can provide
reliable data. Newly released models provide better inter-
nal calibration via different ion excitation,, thereby allowing
for elimination of site-specific sample calibration.
This device is easy to use, can be easily transported in
a backpack, and provides nondestructive analysis. It is
also reliable and quick. The XRF spectrometer has a
relatively low sensitivity, reading primarily in ppms, which
is adequate for most metals, especially lead. If there is a
mix of metals at the site, however, there may be some
interference overlaps. At a site with cadmium or where
concentrations are about 3 ppm, an XRF spectrometer
probably would not be appropriate. For most sites, how-
ever, such as lead battery sites, it works quite well.
Data comparability, standards and performance evalu-
ation methods, standard field methods, and improved
sensitivity for portable units are still needed, but new and
improved instrumentation is continually being introduced.
3.2.5 Sampling and Analysis Equipment for Soil-Gas
and Air
3.2.5.1 Soil-Gas Survey
The soil-gas survey can be useful for finding either con-
taminated ground water or soil contaminated with volatile
organics. Figure 3-3 shows a simple schematic of a soil-
gas sampling device. A slam bar creates a hole into which
a collection tube may be slid. The flux gate magnetometer
ensures the slam bar will not punch through any water or
electrical lines. Other important features are the vacuum
pump, Gillian®, and a simple desiccator that creates a
vacuum to collect the soil pore-space sample. The. FID
detector is used to measure relative concentrations of
different volatile organics in the pore space and may be
used to help locate "hot spots" and subsurface extent for
well placement locations.
The soil-gas survey is a quick and easy way to locate
contaminated spots over a very large area. The contami-
nants, however, should be relatively insoluble, highly
volatile, and have a specific gravity of less than 1. Con-
taminants floating on top of the ground-water table will
volatilize into the porous space in the soil-gas survey
equipment so the survey can be performed. The class of
halogenated compounds, such as TCE, which sink, do
not partition as well into that pore space above the ground
water. In addition, the presence of a perched water table
above the contaminant plume can present .an obstacle
with this method because the vapor phase could solu-
bilize into the perched water phase. An overlying clay
25
-------
Sampling Train Schematic
(not to scale)
. O,D. Tygon
DoGslcnlor (top)
Hoso Clamp j I
— 1/2 in. O.D. Tygon
—1/4 in.Teflon Tubing
(runs from probe, thru
tea, to Tedlar Bag)
Rigid Plastic Tee
Modeling Clay.
Dossicntor (bottom)
Dossicator's Vacuum
Fitting
1/4 in. s.s. sampling probe-
Sample well -
1/4 in. I.D. ,
Thin - Wall !
Teflon Tubing:
Figure 3-3, Sampling train schematic for soil-gas survey.
i
zone above the contaminated ground water or soil makes
detection difficult by restricting vapor mobility. i
3.2,5.2 Geoprobe j
The geoprobe, the next step up from the soil-gas survey
in sophistication, is a hydraulically operated sampling tool
that punches down into the soil. It is a rapid site assesjs-
ment tool, very effective for collecting soil, soil-gas, ahd
ground-water samples. It is a rugged system, which can
collect samples at discrete depths to about 40 feet,
whereas most soil-gas surveys can collect only from 4 to
10 feet. !
The geoprobe can work in three different media types:
pore space, soil, or ground water. It does not perform well
in competent clays or cobble. It may, however, be useful
in weathered bedrock if the overburden is well fractured.
It punches the hole much more quickly than does the
slam bar of a soil-gas survey. This feature is especially
useful with soils that have been compacted by heavy
equipment. The geoprobe is also easier to use than the
slam bar, and can be manipulated like a drill rig by push-
ing different levers. Because it is truck mounted, however,
a truck must be able to get in and out of the area. One
disadvantage of this method is that, because of the di-
ameter of the device, the sample volume is small. It also
provides screening only for ground-water samples and
will not penetrate cobble or hard rock. |
i
3.2.5.3 Flux Measurements \
Flux measurements determine mass emission rates using
the concentration and flow rate of gases through a cham-
ber and the surface area isolated by the chamber. Flux
measurements are used primarily for landfill emissions,
treatability studies, and contaminated aquifers, and less
for underground storage tank work. Flux chambers are
becoming popular for studying overall mass emission
rates from excavated soil piles and landfills. Because
landfills have passive emission rates, flux measurements
are used when residents living near a landfill complain
about odors at active or inactive sites. They also are used
for large-scale excavation where the soil is contaminated
and the potential for offsite contamination or migration is
not known. Recently, EPA's ERT used flux measurements
on a treatability study looking at degradation rates of
polyaromatic hydrocarbons (PAHs) in soil to assess con-
taminant volatilization versus biodegradation in a creo-
sote composting project.
Some limitations to this method are that condensation or
moist soil hampers its effectiveness, cross contamination
sometimes occurs, and careful representative sampling
is required.
3.2.5.4 Other Methods
A number of other methods can be used for field screen-
ing sampling and analyses, including colorimetric tubes,
combustible gas indicators, organic vapor analyzers
(OVA), soil texture kits, hydrolabs, and RAMs. These
methods are described below.
The colorimetric tubeis a quick chemical-specific indi-
cator. It can detect inorganics as well as organics, and
cyanides and chlorines in the air. The colorimetric tube is
usually used to detect threshold limit value (TLV) levels
for site entry. The personal protective gear then can be
adapted to the exposure levels that might be indicated by
the Draeger® tubes. This method is very quick but tem-
perature dependent.
A combustible gas indicator detects oxygen as well as
the lower explosive limit for combustible gases, reading
in ppm levels for the combustible gases. This reading is
very important when entering an enclosed space with
drums or any area that might have a low oxygen content.
It has been used to determine the safety of entering
excavation pits, storm drains, and basements that may
have heavily contaminated atmospheres and where there
might be some negatively buoyant gases that can dis-
place ambient oxygen.
An organic vapor analyzer can be used in situations
involving lagoon sludge in drums. This device is particu-
larly useful for detecting methane, and hydrogen sulfide.
A flame ionization detector (FID) uses a hydrogen flame,
and can be used in a GC or scan mode. FIDs detect
methane and hydrocarbons, especially important at wood
treating sites where naphthalene may be present. An FID
is a little more sensitive to naphthalene than PIDs in a
rough screening. FIDs detect the alkanes, including the
carbon-carbon and hydrogen-carbon bonds.
A mercury organic vapor analyzer, which is an elemen-
tal indicator for mercury vapors, is very useful for a variety
26
-------
of mercury-contaminated sites. It is a gold film detector
and measures the higher resistance, as it accumulates
more mercury on the gold film. The elemental mercury
indicator has been useful at mercury wood treating sites
and in soil-gas surveys. It also is useful in the laboratory
where there may have been broken thermometers or at
mercury spill sites.
A basic soil texture kit is useful for doing soil borings,
working with hand augers, or doing attenuation or profiles
of soil depth. It can be used to determine percent sand,
silt, and clay, and is very simple to use. An organic matter
test kit measures from 0 to 16 percent organic matter
when evaluating adsorption or attenuation of organics
and soil.
A hydrolab is used quite a bit in streams and down wells,
as well as other water bodies, that are not too contami-
nated. In a hydrolab, numerous functions are built into
one tool. It can perform all the various tests for tempera-
ture, pH, dissolved oxygen, conductivity, redox potential,
and salinity. A drawback is that because a hydrolab works
as a function of temperature, the temperature must be
stabilized. A hydrolab is very simple to use, however, and
can be backmounted and transported easily.
A RAM, or an aerosol monitor with a data logger attach-
ment, can be used for evaluating offsite migrations while
evaluating test pits or excavations on soil at different
sites. This instrument measures and evaluates airborne
particulates, such as those from a wood treating site. A
RAM is not ion specific, but it can relate the amount of
dust migrating off site from an excavation with real-time
monitors to the amount measured by XAD or filter cas-
sette tubes. (XAD tubes and filter cassettes are com-
monly used in tandem to measure particulate laden with
contaminant versus semi-volatile organic compounds.)
3.2.5.5 Geophysical Methods
Some geophysical methods that can be used for analyses
include EM31s, magnetometry, well logging devices,
open path infrared spectrometers, and Scanex Juniors.
An EM31, an electrical conductivity device, can be
used to evaluate ground water or to detect metals and
different geology in soils. It also is helpful for finding utility
lines and buried transformers. This device is usually used
in tandem with a magnetometer to detect ferric metal as
well as electrical conductivity, or changes in conductivity,,
of the soil. It can be useful in detecting discontinuities,
such as a buried lagoon or impoundment, and is useful
for examining differences in porosity and permeability in
fluids.
EPA's ERT used sophisticated magnetometry at a site
in Aberdeen Proving Ground to look for a white phospho-
rus munitions burial area in a wetland to which there was
no access, either by boat or land. Investigators suspected
that there was a large dump site of white phosphorus
contained in metal drums on the site: The survey was
performed with a helicopter towing a helium magnetome-
ter. This method is very effective for covering a large area
in a short period of time, but can be quite expensive. That
11/2-day survey, for example, cost about $35,000, which
included the cost of 150 survey lines over an area ap-
proximately Vi-mile square as well as data interpretation.
At the Aberdeen Proving Ground, ERT also used mag-
netometers attached to a frame on a zodiac boat. When
the survey team located spikes which could be indicative
of metallic explosives, investigators deployed a buoy to
pinpoint these areas. With real-time magnetometers, it is
possible to collect and map information about potential
magnetic anomalies and their locations that may be
suggestive of buried containers, tanks, or at Aberdeen,
projectiles.
Well logging devices use electrical conductivity, resis-
tivity, or gamma logs. These devices can supply informa-
tion about well construction, water bearing zones, and
rock type. A geophysical well logging device called a
downhole well logger, an EM39, can be used in wells with
different diameters. It is very adaptable and relatively
inexpensive.
The Scanex Junior has been used predominantly at De-
partment of Defense (DOD) sites. It is useful for evaluat-
ing nitrocellulose compounds and explosives in air, but it
cannot detect phosphorus compounds. It also operates
real-time and has an electronic capture detector.
3.2.6 Immunoassay Methods
Immunoassay field screening kits are becoming increas-
ingly popular, because they can detect compounds, like
PAHs and total petroleum hydrocarbons (TPHs), in a
short turn-around time. They also are very useful for cost
control on treatability studies. Field immunoassay tech-
niques are semi-quantitative, limited to numbers in ranges
of concentrations, and work on antibody mechanisms.
There are many vendors for these kits, and most methods
have undergone preliminary field testing. Currently, EPA
is actively looking at immunoassay techniques for PCBs,
PCP, pesticides, njtroaromatics, and metals in soils and
water. Under the SITE program, EPA has already dem-
onstrated one kit and a plate assay. All methods show
promise as field screening or analytical methods.
The advantage of the field immunoassay for PCP is the
ease of handling methods and low cost use for untrained
personnel. This method provides real-time performance
at slightly higher values than gas chromatography or
mass spectrometry. The test for PCP takes about 30
minutes and has a detection limit of 2 ppm in soil. The
cost of performing the test on a single sample is $30 to
$50. The disadvantage of an immunoassay is the poten-
tial for false positives. The ERT has experienced an ac-
curacy of greater than 50 percent with this method.
27
-------
3.3 FUTURE NEEDS AND DIRECTIONS
There are many areas of development open in fieltd
screening methods. Some immediate needs include im-
provements in the following technologies: !
• In situ or standoff chemical sensors
• Soil and ground-water sampling devices
• Soil-gas sampling devices I
• Fluid sampling devices for the vadose zone !
• In situ and field portable analytical methods for thje
analysis of contaminants in all types of environmental
media j
i
• Geophysical methods
• Data analysis and interpretation tools !
• Sampling network design (e.g., expert systems) j
• New or refined chemical or compound-specific analyti-
cal methods i
Many other tools are being evaluated by the Environ-
mental Response Branch and may soon be available for
performing field screening surveys but it is difficult to
present them all. For more information on field screening
devices in general please contact the Chief of the Site
Investigation Section, Environmental Response Branch,
(908) 321-6740, U.S. EPA.
In addition, the future of the field might include such
technologies as toxicity devices to mimic human re-
sponses, remote sensors to detect surface hot spots, light
pipes to determine concentration gradients below ground,
disposable microsamples arid sensors, expert systems
arid robotics, and continued miniaturization.
With existing technologies it is possible to achieve rapid,
analytic appropriate, and cost-effective site charac-
terizations. Field screening and analytical methods still
have to gain acceptance by the chemists and the courts,
however, and recognized field methods and quality con-
trol procedures do not exist for most emerging and alter-
native technologies. The usability of data also needs to
be clearly defined.
3.4 REFERENCE
Kaelin, L 1992. Senior Field Chemist, Roy F. Weston/Response
Emergency and Analytical Contract, Edison, N.J.
28
-------
CHAPTER 4
Covers, Slurry Walls, Grouting, and Dynamic Compaction
for RCRA Corrective Action Stabilization
Walter Grube, Ph.D.
Clem Environmental Corporation
Fairmount, Georgia
Stabilization activities carried out under RCRA corrective
action involve clear engineering designs and some type
of actual field construction, using common construction
equipment and proven operational technologies. The de-
signs for a short-term stabilization cover may be modified
from standard textbook and guidance manual models,
and, in general, the equipment used will be whatever is
locally available. The theories, concepts, approaches,
and schematics in guidance documents often are modi-
fied significantly for actual practice.
The RCRA Corrective Action Stabilization Program is
young, and few case studies are available. This chapter
will examine several structures and processes that are
well documented in published technical literature. These
technologies can be and are being effectively applied as
corrective actions to physically stabilize sites. The struc-
tures and processes discussed include:
• Cover systems and caps (Section 4.1)
• Slurry walls (Section 4.2)
• Grouting (Section 4.3)
• Dynamic compaction (Section 4.4)
Figure 4-1 is a commonly seen drawing that includes
many features and structures associated with stabilization
of a contaminated site.
Section 4.5 discusses quality assurance and Section 4.6
is a bibliography.
4.1 COVER SYSTEMS
4.1.1 Considerations
A cover system includes not merely a single concrete
pad, clay cap, or geomembrane, but a combination of
materials of varying complexity that are constructed in a
logical order, each to perform a particular function. In
designing a cover system, one needs to consider the
characteristics of the waste, the site geology, any acces-
sory structures, extraction wells, monitoring wells, and
other types of "plumbing" associated with biological,
chemical, or hydraulic technologies.
Both the type of waste and the type of site must be
carefully considered. Municipal solid waste is frequently
found at corrective action sites, with industrial waste prod-
ucts buried throughout the site. Industrial wastewater
ponds are also common.
At one example site, an entire pond, used for industrial
process wastewater, was excavated, including the syn-
thetic membrane liner and the underlying compacted clay
liner. It was discovered that the foundation under the
compacted clay consisted of fly ash from an old power
plant on the site. Remedial engineering designers had to
rebuild this foundation to reconstruct the pond according
to modern standards. This example shows the impor-
tance of not only the specific remedial structure, but also
the surrounding soil material. The new pond was built
with a composite liner of both newly compacted soil and
a new geomembrane.
, Also to be considered as an aspect of design is the impact
on construction, timing, and cost of having operational peo-
ple in confining protective clothing. The severe impact of
having to work under the constraints of health protection
has been well documented in the Superfund program.
4.1.2 Recommended Design
There are recommended standards for covers as interim
protective measures at corrective action sites. The sce-
nario of the beautiful grass cover with carefully designed
runoff control drainage structures and a fence around the
perimeter, however, is probably the exception rather than
the rule. In fact, even a cover system that looks fairly
close to the ideal may not be meeting the regulatory
requirements for minimum erosional soil loss.
Environmental engineering technologies continue to de-
velop rapidly. Thus, it is essential for both design engi-
29
-------
\ I / \ •' \
\ } I I ' > ' >
I ' \ t \ I \ I
I I I / 1 I If'
Unsaturated Zone
Slurry Wall
Contaminated Plume
Figure 4-1. Corrective action stabilization site. ;
I
neers and regulatory staffs to continually look at the vari-
ous guidance documents available from EPA and state
agencies to ensure that they are up to date. The Soil
Conservation Service, for example, no longer recorrj-
mends the universal soil loss equation (USLE), so guid-
ance documents that recommend using USLE may havb
been written several years ago. Any local Soil Conserva-
tion Service office should be able to provide the new or
modified soil loss equation. , !
i
Many Agency-published guidance documents have a
number of standardized line drawings showing the ideaf
structure for a cover system. Figure 4-2 shows the rec-
ommended RCRA multi-layer design with vegetational
topsoil, drainage layer, filter layer, and composite infiltra-
Vegetatlon/soil_
top layer
\\// 4 4 \V/ 4
Drainage layer
Low hydraulic
conductivity.
geomembrane/
soil layer
uilic r
:ivityj
ane/ I
wer L
Waste
°«" °«* »« * 0." O. • O • 0
o..o.',o.'.o.'.o.9.o.V.
60
cm
-Filter layer.
0.5-mm !
60 (20-mil) |
cm geomembrane
Figure 4-2.
EPA, 1991).
Recommended landfill cover design (U.S.
tion barrier system. This design has been modified
slightly for use or application in more arid and other re-
gions where this precise design may not be the most
effective. Figure 4-3, for example, includes a biotic barrier,
originally designed to combat penetration by small mam-
mals such as gophers. How important it is to use a biotic
barrier and under what particular site conditions is still
under debate. This figure also shows a gravel armor sur-
face instead of a vegetative surface; an adaptation of this
type of design may be more desirable and more effective
than the vegetative-cap in many areas. Engineers may
want to look at variations of these designs, depending
upon the degree of stabilization necessary. The drainage
layer, for example, may be unnecessary in areas of 5
inches or less of annual precipitation.
A number of standardized test methods for foundation
preparation have been used by the civil engineering com-
munity for many years. EPA guidance documents contain
recommendations for the minimum frequency of testing
for foundations and other structures related to a cover
system that must restrict rainfall infiltration. Many of the
tests carried out during actual construction involve simple
observation by the onsite engineer. The construction
quality assurance plan for a corrective action cover is
likely to place more emphasis on the field inspection as-
pect than on rigorous laboratory testing. This then places
an increased emphasis on the training and experience of
the field inspection and construction staff.
4.1.3 Soil Components
The barrier layer is the 2 to 3 foot thickness of compacted
soil used as either a liner or an infiltration barrier. EPA
has refined and published criteria for barrier layers in the
last year. These criteria ensure that a contractor will con-
30
-------
Cobbles/soil
top layer
Biotic barrier
(cobbles)
Drainage layer
Low hydraulic conductivity
geomembrane/soil layer
Gas vent layer
Waste
Jl-li-1-li'
JL J_ JL J. J. 1 -L JL -L J. JT
*
0
60
crn^,
30
cm
Geosynthetic filter
Geosynthetic filter
—^4 — 0.5-mm (20-mil) geomembrane
60
cm
-5^ — Geosynthetic filter
cm
Figure 4-3. Recommended landfill cover with options (U.S. EPA, 1991).
struct a low permeability soil layer after performing the
specified Proctor density tests, various permeability stud-
ies, and other tests to obtain the required less than 10"7
centimeters per second permeability for a soil liner or
infiltration barrier. This procedure involves performing a
fairly complex set of laboratory studies using different
compaction energies, developing moisture density
curves, correlating those data with hydraulic conductivity
measured in the laboratory at different holding water con-
tents, and then developing an Acceptable Zone of com-
binations of density and moisture that will result in low
permeabilities (see Figure 4-4). This Acceptable Zone
then must be adapted to a region and presented in a
curve so that the contractor can ensure that all lifts of the
soil-based infiltration barrier will have a moisture content
and/or density inside the shaded region shown in Figure
4-4. The frequency of sampling and testing is largely
determined by the characteristics of individual job sites,
although ASTM testing methods are available as a re-
source. Some states, however, have specified these de-
tails of construction quality verification for waste sites. A
number of the guidance documents now available present
specific details aimed at increasing the probability and
assurance that the final structure will meet the required
(a)
A Modified Proctor
o Standard Proctor
n Reduced Proctor
Acceptable Zone
Molding Water Content
Molding Water Content
(b)
Maximum Allowed k
Acceptable Zone
Modified to Account
for Other Factors
Molding Water Content
Molding Water Content
Figure 4-4. Recommended procedure (U.S. EPA, 1991).
(a) Determine compaction curves with three compactive efforts.
(b) Determine hydraulic conductivity of compacted specimens.
(c) Replot compaction curves using solid symbols for samples with adequately low hydraulic conductivity and open
symbols for samples with a hydraulic conductivity that is too large.
(d) Modify Acceptable Zone based on other considerations such as shear strength or local construction practices.
31
-------
design specifications. Tables 4-1 and 4-2 show recomf
mended tests and observations for subgrade preparation
and recommended materials tests for barrier layers.
4.1.4 Geosynthetic Components
A major component of modem waste site construction
includes geosynthetic materials, such as geomembranes,
drainage materials, support, and strengthening products.
(The term flexible membrane liner is used interchange-
ably with geomembrane by many people in the industry.)
Table 4-3, developed for EPA by the Geosynthetic Ref
search Institute at Drexel University, shows the various
uses in waste containment systems of these materials.
Since this table was developed a new category, the geo|-
synthetic clay liner (GCL), has been added. This category
is an additional type of geosynthetic material used prir
marily for the barrier and to some extent for the separa-
tion and reinforcement functions. I
!
A GCL is a prefabricated composite sheet material, corri-
posed of one or more layers of a synthetic polymeric
membrane or textile and a uniform layer of very low per-
meability bentonite clay. This product swells when hyj-
Table4-1. Recommended Tests and Observations on
Subgrade Preparation ;
Parameter
Percent Compaction
Compaction Curve
Preparation of
Previously
Compacted Lift
Test Method
ASTM D2922 or
ASTM D1556 or
ASTM D2937 or
ASTM D2167
ASTM D698
Observation
Minimum
Testing
Frequency
1 per acre
1 per 5 acre:
Full coveragi
Table 4-2. Recommended Materials Tests for Barrier Layers
Parameter
Test Method
Minimum Testing
Frequency
Percent Fines
Percent Gravel
Liquid & Plastic Limits
Water Content
Water Content
Construction
Oversight
ASTM D1140
ASTM D422
ASTM D4318
ASTM D4643
ASTM D2216
Observation
1 per 1,000yd3
1 per 1,000yd3
1 per 1,000yd3
1 per 200 yd3
1 per 1,000yd3
Continuous in
barrier pit on major
projects; continuous
in placement area
on smaller projects.
Table 4-3. Customary Primary Functions of Geosynthet-
ics Used in Waste Containment Systems
Primary Function
Type of
Geosynthetic Separate Reinforce Filter Drain Barrier
Geomembrane II
Geotextile • * • D •
Geonet •
(Geo) pipe . •
Geocomposite •
Geogrid •
Source: Adapted from U.S, EPA (1991).
drated and forms a layer with a permeability much lower
than can be achieved by compaction of soil, The GCL
offers the advantages of quality-controlled manufacturing,
and it can be easily transported in rolls and installed with
minimal field equipment.
EPA (1991) provides significant details about site prepa-
ration, testing of soil materials, and inspection criteria for
the compacted soil component of a cover. In corrective
action designs that may not be planned for long-term
performance, but must serve as effective interim precipi-
tation barriers, the engineered soil is a preferred choice.
It is difficult to achieve a tightly compacted soil liner.
Desiccation cracks, the definition of which is still open to
interpretation, indicate that the compacted soil or the clay
liner underneath the geomembrane may not be properly
built. A smooth roll subgrade is very important, especially
for some of the heavier or stiffer geomembrane materials
that may not be particularly flexible.
The disadvantages of the simple "clay cap," and the real
problems involved in ensuring that it meets a 10"7 cm/sec
permeability specification, must be realized in cover design
and construction. Conversely, a site may have no current
infiltration protection, so that even a poorly constructed "clay
cap" may have significant short-term value. The regulatory
authority should make sure that a rapidly constructed "clay
cap" is not relied upon for long-term protection.
When employing a geomembrane as a primary cover
component in a RCRA corrective action cover system,
there are several considerations. If such an interim meas-
ure has a useful lifetime of 5 to 10 years, and it is uncer-
tain whether it will be part of the final closure cover
system, the remedial engineer must balance the need for
intensive seam quality control and geomembrane quality
with a cost-benefit or risk analysis of gains. Intensive
seaming and quality control covering with carefully de-
signed soil can be quite expensive. A final closure cover
system, for example, requires very careful cleaning of the
seam area followed by immediate covering with soil. A
; 32
-------
significant concern may be the removal or replacement
of a geomembrane when the final site remedy is imple-
mented. The possibility of completely removing and re-
building a site cover at final site closure should be
considered during interim cover design.
The "composite action" of a geomembrane and com-
pacted soil seepage barrier has recently been an area of
technical discussion and construction concern. Texas
A&M University studies have shown that without intimate
contact between the geomembrane and the underlying
compacted soil, any leachate or infiltration from precipi-
tation coming through a fault in the geomembrane will
spread laterally and across the top of the compacted soil
liner. As it spreads, it will find a fault, fracture, or desic-
cation crack, and will go down through the liner (see
Figure 4-5). A number of studies have verified that inter-
face lateral flow and flow times through compacted clay
liners are much faster than predicted by time of travel
models. These studies confirm the importance of intimate
contact between the geomembrane and underlying soil
liner. Achieving this type of contact begins with construc-
tion. Ideally, the top of the soil liner should be graded
smooth as a pool table top, although that is difficult to
achieve in actual field construction situations.
4.1.5 Accessory Structures
Pipe penetrations or other accessory structures need to
be designed as parts of corrective action covers. Figure
4-6 is a simplified view of an ideal multi-layer cover with
a gas vent (it omits the various collars that would prevent
piping or short circuiting along the sides of the pipe). The
remedial manager needs to assess what structures, such
as collars or graded slopes, are necessary at a particular
site, keeping in mind the interim nature of the action. The
. ' Gas
Clay Liner
Leachate
Composite Liner
Leachate
FML
A = Area of entire liner Area < Area of entire liner
Leachate
Leachate
Figure 4-5. Leachate or precipitation infiltration with a
soil (clay) liner and composite liner (U.S. EPA, 1991).
RCRA corrective action program provides a good oppor-
tunity to try alternative designs. Figure 4-7 shows one
example of a novel design to conduct infiltration away
from underlying wastes. Installing interim covers presents
a good opportunity to evaluate promising new ideas,
since a final closure can still be implemented in the future
if the design is not adequate. If the design is successful,
however, real advances in economics or performance
may be achieved.
4.1.6 Subsidence
Subsidence is an aspect of cover design that comes up
in almost every discussion of cover systems. The problem
Vent
Drain Layer
Membrane
Vent Layer -\ °» !."• T.°? J-
L_j—W—r
. Perforated Pipe
Top Layer
Low-Permeability
Geomembrane/Soil Layer
Waste
A
Figure 4-6. Cover with gas vent outlet and vent layer (U.S. EPA, 1991).
33
-------
Figure 4-7. Conductive layer barrier with graded slopes.
Is how to design a cover that will accommodate the varit
ous depressions and peaks that evolve in the buried
waste. Rgure 4-8 shows the effects of cumulative subsH
dence. Very few sites have been intensively mapped^
Investigated, and documented in terms of the various
degrees of subsidence that occur. As with accessory
structures, the RCRA corrective action program offers ar)
opportunity to research the effects of waste composition
and density using benchmarks or elevation markers to
chart movement. As early as is practical, remedial manj-
agers should take baseline measurements upon which
to base future elevation measurements. This step
will allow the manager to document whether subsidence
is a real defect that can be addressed in designing the
cover system. . |
EPA's Office of Research and Development (ORD) has
developed approaches to predicting the strength of geo!-
synthetic materials required-to bridge specific subsidence
voids. Figure 4-9 shows the derivation of geomembrane
tensile stress due to subsidence of material beneath the
geomembrane.
Cover Soil
Unit Weight "Yes1
-------
Rainfall/Snow
Jill
Snow Accumulation
Snow Melt
Runoff
Lateral
Drainai
Interception T
1
Transpiration
Snow
Evaporation
Evaporation
Vertical Percolatio
t Interception
Evaporation
Plant Growth
Depth
of Head
Barrier Soil
Percolation
Figure 4-10. Simulation processes in the HELP model
(U.S. EPA, 1991).
4.1.7 Materials Handling
Materials handling, like scheduling and logistics, are the
kinds of features that the permit writer and the designer
frequently leave up to people in the field. Materials han-
dling issues to be considered when capping a site include
the amount and types of soil material to be moved and
the logistics and machinery required for distribution.
Again, a decision needs to be made about the minimum
that should be implemented in a corrective action.
Usual cover designs include a variety of different materi-
als — topsoil, sand or gravel drainage materials, clayey
soil for infiltration barrier compaction, geosynthetic mate-
rials for various applications, and pipes or other conduits.
Installation of each of these materials may be greatly
hindered or eased by weather conditions, previously in-
stalled structures, availability of trained installers, and
other site-specific characteristics. The experience or ad-
vice of a construction contractor may significantly affect
the speed of installation of some of these materials.
4.2 SLURRY WALLS
The slurry trench cutoff wall, or slurry wall as it is com-
monly called, is a vertical barrier surrounding a site, de-
signed to control the water quality, flow, and level. The
slurry wall is, in its simplest form, a discontinuity in a
porous geologic formation. The slurry wall should be dif-
ferentiated into its two distinct components — the slurry
trench and the cutoff wall backfill material. The slurry
trench is a mechanism that holds a ditch open in a sandy
or porous geological material, so that plywood or steel
sheets are not necessary. A colloidal bentonite clay sus-
pension provides structural support during excavation.
Filling the trench with backfill material provides a ground-
water cutoff or a lower permeability zone. The engineer-
ing design and construction of the slurry trench requires
different practice and testing than does that of the cutoff
wall. The cutoff wall consists of the material put back into
the trench to provide a ground-water control barrier.
Trench backfill is often a soil enriched with fine or ben-
tonite clay but may also be a cement-bentonite mixture,
or even a geomembrane sheet. Numerous engineering
reports and ASTM special publications are readily avail-
able to describe the state-of-the-art of slurry wall backfill
design, construction, and testing. The slurry wall should
be tied into the low conductivity base material, if possible,
or water can be withdrawn from the enclosed material to
form an inward gradient.
4.2.1 Equipment
Several different types of common construction equip-
ment are used to construct a slurry wall. Backhoes, usu-
ally equipped with a long-reach arm, clamshell on a cable,
and various specially modified excavation buckets are
used, based on site-specific geology and depth. The ex-
cavating equipment opens a trench, usually 2 to 3 feet in
width. The trench is continuously filled with a colloidal
clay suspension designed to hold the trench open without
requiring other mechanical devices. The construction pro-
cedure involves digging the trench, filling it with slurry,
and then as a follow-up activity, adding the backfill that
provides the actual barrier or retarding structure to
ground-water flow (see Figure 4-11).
Figure 4-12 depicts several of the different equipment and
excavation approaches applied, depending on depth and
type of site geology.
One of the better primers describing pollution migration
control cutoff walls is an ORD publication (EPA, 1984b).
This guide provides engineering theory, historic applica-
tions, geologic and other studies needed to characterize
a site, elements of design and construction, information
on monitoring and maintaining a slurry wall, various cost
elements (e.g., cost of a slurry trench cutoff wall in dollars
per face, with a vertical face square foot ranging from $5
to $15), and evaluation guidelines. More current case
histories may be found in civil engineering publica-
tions and in some Superfund conference and symposia
proceedings.
4.2.2 Slurry Wall Performance Testing
Testing of slurry wall performance largely involves very
simple laboratory determination of permeability and the
use of permeability as a measure of compatibility with
solvent, leachate, and nonaqueous phase liquids
(NAPLs). The existence or occurrence of NAPLs, either
light or dense, is a significant site feature affecting slurry
wall performance. A number of NAPLs exist as fairly pure
product, either sinking or floating, and have a detrimental
effect in dehydrating soil and clay materials. Their pres-
ence may result in pods or floating layers of various
NAPLs, which can significantly increase the permeability
of backfill material.
35
-------
Sand & Gravel
Figure 4-11. Slurry wall excavation and backfill. ;
Clamshell Bucket ;
Clamshell Bucket
Dragline Bucket
Figure 4-12. Methods of construction for a slurry wall. !
The compatibility test of the backfill must rely on investi-j
gallons that provide information concerning light or dense
NAPLs on the site. The containment structure may need
to be moved outside the zone of contamination to com-;
pletely avoid NAPLs, or all NAPLs may need to be re-;
moved to reduce the possibility of attack and degradation
of the slurry wall backfill. j
In a RCRA corrective action, the site manager may wish!
to determine whether a slurry wall is going to be an
effective structure that can be part of the total corrective
action program. Two common tests are box tests and
pump tests. The box test approach consists of excavation
and construction of a small-scale cutoff wall fully enclosj
ing an area approximately 50 to 100 ft square, to the
depth of a low permeability geologic formation on sitei
This approach requires mobilization of nearly all th4
equipment required for construction of afull-scale ground^
water barrier. Following construction, the ground water
inside the box can be pumped, and ground-water eleva-
tion data from inside and outside the box applied to meas-
ure the permeability of the barrier. This method has been
used by both governmental agency and private site own-
ers to verify effectiveness before a slurry wall thousands
of feet in length has been constructed.
A pump test consists of a series of wells installed on
either side of the cutoff wall. The wells on one side are
drawn down to a predetermined elevation, and the water
levels in wells on the unpumped side of the wall are
measured to determine the extent of lowering. From these
data, wall permeability may be calculated.
Thin tube samplers from a setup wall have been largely"
experimental, and there have been a number of studies
with ground-water monitoring wells on either side of the
I 36
-------
barrier. Backfill retrieved by thin-tube samplers is normally
tested for permeability by laboratory methods. Ground-
water well arrays and the approach to data analysis from
well water levels has not been standardized. This ap-
proach remains a site-specific and company-specific
method.
4.2.3 Advances in Slurry Walls
The geomembrane industry has made continuing ad-
vances in providing lower permeability barriers for the
backfill in a slurry wall. The current geomembrane tech-
nology involves a channel lock structure along the vertical
edges of each sheet, to provide a tongue-and-groove type
of slip joint between the two sheets. High density poly-
ethylene (HOPE) barriers, however, are not commonly
installed in the United States; they seem to be more
widely accepted in Europe.
4.3 GROUTING
Grouting is a different type of containment or site stabili-
zation structure. Classically, the grouting industry in-
cludes several different approaches:
• Top sealing, or stabilization of a soil area.
• Bottom sealing, or construction of a barrier underneath
a waste site.
• Waste solidification and immobilization, which uses dif-
ferent chemicals to permeate and bond with the soil or
the geologic material.
• Excavation of subsurface material followed by filling the
void with a formulated grout to reduce the void space
through which water can infiltrate and leach the waste.
• Grout curtain.
Figure 4-13 shows four of these types of grouting.
Waste solidification and immobilization, the historic inter-
pretation of stabilization, can be used as a corrective
action on a portion of a site to determine its effectiveness.
The remedial manager can then decide whether it will be
useful in stabilizing the whole site. There are a large
number of industries involved and many case studies are
available throughout the United States and a number of
foreign countries that document the effectiveness of so-
lidification and immobilization on different types of waste.
EPA's research laboratory in Cincinnati has performed
much research in solidification, studying the effectiveness
of various types of leaching tests in evaluating the solidi-
fication processes. A number of different tests are being
evaluated beyond the Toxicity Characteristic Leaching
Procedure (TCLP), the EP Toxicity Leaching Procedure,
and some long-term, large-scale exposure tests.
Grouting is considered to be predominantly a subsurface
operation. Figure 4-14 shows small pods of grout under-
lying a site and isolating it from the surrounding environ-
Top Sealing or Stabilization
of Contaminated Soil
h
Solid Waste-
Bottom Sealing
Cover
Solid Waste
Waste Solidification in Place
Waste Removal and Solidification
• Additives
Liquid Waste
Impoundment
Figure 4-13. Uses of grouting to contain hazardous wastes.
4 Hectares (10 acres)
• Grout Tubes
Stream
Min. 1.5m (5 ft) Soil Layer
Unconsolidated Earth Materials
Bedrock
Grouted Material
Not to Scale
Figure 4-14. Cross-section of grouted bottom seal be-
neath a landfill (U.S. EPA, 1986).
37
-------
ment. The degree of bonding and the possibility of seep-
age between adjacent grout pods are questions that re-
quire further research about this technology. Even!
geophysical methods have not yet been successful in
verifying the degree of hydraulic integrity needed in en-
vironmental isolation applications of grouting. Grouting
methods accepted for structural improvement still need
refinement to assure environmental customers and ap-J
proving agencies that adequate containment is achieved.'
Rgure 4-15 shows a schematic of the basic equipment
needed for a grout injection system. (Chapter 5 will
Grout Mixer - Agitator
Bypass Line
Quick Disconnect
Valve
Casing
Grout
Injection
Point
Slip |
Collar'
Grout !
Pipe ;
Figure 4-15. Schematic of grout injection system (U.Si
EPA, 1986).
scribe the different types of hardware and operations int
volved in placing grout in the subsurface.) One of the
problems with grouting, which again can be examined
carefully in a corrective action application, is how to del
termine where the grout goes beneath the ground and
how effective it will be. If possible, the construction engij
neer should place some grout and then excavate it to
examine its continuity and texture visually and with measi
urement devices. '
Rgure 4-16 shows the injectibility of particulate and
chemical grout in fine and coarse soils. This graph comes
from standard engineering documents, and clearly shows
that grout formulation must be developed specifically for
each type of site geology. An additional reason for devel-
oping site-specific grouts is that many promising new
grout formulas include organic components. If these com •
Coarse
Silt
Fine Soil
Silt
(nonplastic)
Organic
Polymers
Chrome-Lignin
I I I
Resins
Silicates
II I
Bentonite (clay)
10.0
1.0
0.1
Grain Size, mm
0.01
0.001
Figure 4-16. Injectibility of particulate and chemical grout
in fine and coarse soils (U.S. EPA, 1986).
ponents do not react with the soil or other grout compo-
nents, they may form an undesirable presence in the
ground water. Laboratory verification is needed of both
the grout's effectiveness and of the absence of additional
contaminants to a site.
Another important consideration is the effect of the grout-
ing process or construction on the surrounding area. In
one example, the soil surface was raised almost a foot
as a result of injecting a grout material into the subsur-
face. In this case, the area was situated away from habi-
tation and adjacent to a Superfund remedial action site,
since it was a test program to evaluate this particular
subsurface placement project. When grouting in an urban
area or with structures that could be affected by ground
movement, the remedial manager must consider this type
of surface shifting.
4.4 DYNAMIC COMPACTION TECHNOLOGY
The goals of dynamic compaction are to increase bearing
capacity and decrease total and differential settlement.
Dynamic compaction has seen limited use and has only
limited documentation. Over the last 10 years, there have
been about six examples of applying dynamic compaction
to increase the density of wastes prior to capping or final
closure of a site. The most recent application was at the
Savannah River Plant in South Carolina. It is expected
that several studies will be presented at the ASCE Spe-
cialty Conference in New Orleans. These proceedings will
be the most recent update on dynamic compaction, its
implementation, effectiveness, and performance in den-
sifying waste sites as a stabilization technology.
38
-------
(n dynamic compaction, large weights of from 5 to 20 tons
free fall from heights of up to 100 feet. The cumulative
applied energies of this process typically range from 30
to 150 ft-ton/sq ft, and may succeed in densifying soil or
waste down to 50 feet. The spatial distribution and the
time sequence of dropping the weights are critical, and
have been established by the industry. A few additional
factors need to be taken into account, however, such as
the effects on structures in the neighborhood, soil condi-
tions, and soil characteristics in transmitting vibration ef-
fects. Seismic studies done many years ago with
explosives can be applied to determine a safe distance
from structures. Other problems requiring the designer's
attention include present waste density, the geologic
foundation underneath the waste site, impact of local or
seasonal weather conditions on equipment operation,
proof of effectiveness on a test section, and methods to
be used to evaluate degree of success.
The effectiveness of densifying the underlying material is
determined by measuring the volume and area of the
various craters created by dropping the weight in a pre-
planned sequence. The data can be used to calculate the
increase in density of the underlying soil or waste. The
density, or the maximum depth of effectiveness of dy-
namic compaction, is calculated by a simplified equation
proven to be quite accurate for field use. Depth of influ-
ence is found using the equation from Mayne et al.
(1984). Some of the historic applications to waste sites
have shown that induced subsidence, or compaction, is
slower in wetter soil, because water is incompressible.
Evaluating the effectiveness of dynamic compaction pri-
marily requires topographic volumetric methods (i.e.,
topographic surveys, including crater measurement) gen-
erally supported by some of the standard penetration
tests, pressure meters, and geophysical approaches to
confirm the calculations from the topographic measure-
ments. Some of these tests include:
• Standard penetration tests (SPT)
• Cone penetration tests (CPT)
• Pressuremeter tests (PMT)
The economic advantages of dynamic compaction in-
clude reduction of differential settlement and subsidence
and increased waste disposal volume. Stabilization of the
site against settlement increases the reliability of a cover
system in preventing water infiltration. Increases in waste
disposal volume have been reported to offset the cost of
dynamic compaction operations.
4.5 QUALITY ASSURANCE
Quality assurance (QA), credibility of data, and quality
control are essential to the engineering construction in a
corrective action. Ground-water monitoring wells are most
commonly used to determine whether a structure, slurry
wall, or cap is working effectively. If a cap is reducing
infiltration, there will be less infiltration down through the
waste and less offsite migration detected by the offsite
well. Figure 4-17 shows a typical configuration for moni-
1 2 3
12312
\
Landfill
yt/
Sand (K = 1x1(T2 cm/sec)
(K = 1 x1 CT7cm/sec)
Layer 2
Sand(K= 1x1 Cr3 cm/sec)
Layer 3
Figure 4-17. Monitoring well configuration.
taring wells. EPA handbooks and guidance documents
present standardized well arrays (see Section 4.6). Hy-
drologic guidance documents and the geologic, geohy-
drological, and hydrogeological literature contain a wealth
of information about constructing and drilling monitoring
wells.
Collecting and measuring what seeps through a site is
the surest method to determine whether there is seepage
or the amount of seepage or leaching. The leachate col-
lection lysimeter is a structure that is readily placed under
newly constructed covers. The environmental engineer-
ing literature adequately discusses the effectiveness of
leachate collection lysimeters and includes actual field
data. Chapter 5 provides more detailed discussion on
structures such as horizontal drilling and other new tech-
nologies applied to waste sites to improve data credibility.
Construction quality documentation, in addition to collec-
tion of data that show ground-water or seepage quantity
and quality, must begin early in the materials charac-
terization phase of a corrective action. EPA (1986) details
the various aspects of construction QA that should be
addressed. These guidelines cover not only data gather-
ing, but also data handling after collection, which should
be according to a preplanned design. Qualified individuals
at all levels and a continuous dedication to quality data
from upper corporate management to field technicians
are essential. Construction quality begins at the design
stage and is continuous through actual field equipment
39
-------
operation and sample collection, at times long after visible,
field work has ceased. ;
4.6 REFERENCES i
ASCE. 1982. Proceedings of specialty conference on grouting in
geotechnical engineering (Wallace Hayward Baker, ed.), New
Orleans. New York, NY: Geotechnical Engineering Division,'
ASCE.
Grube, W.E., Jr. Slurry trench cut-off walls for environmental pollu-
tion control. In: Slurry walls: design/construction/quality con-i
trol, ASTM STP 1129. Philadelphia, PA: ASTM. |
Kessler, J.H., L.R. Dole, and S.M. Robinson. 1983. Radwaste grout-
ing technologies applicable to hazardous waste manage-:
ment. Proc. second conf. on municipal, hazardous, and coa
waste mgmt., Miami Beach, FL Oak Ridge National Labora-f
lory. P.O. Box X, Oak Ridge, TN. !
Mayna, P.W., J.S. Jones Jr., and J.C. Dumas. 1984. Ground rej
sponse to dynamic compaction. J. Geotech. Eng., ASCE, Vol.
110, No. 6, June, 757-774. :
OCE. 1973. Office of the Chief of Engineers. Chemical grouting,
engineering, and design. EM 1110-2-3504. Department of the
Army, Washington, DC. j
USAEWES. 1978. Bibliography on grouting, misc. Paper C-78-8.
U.S. Army Engineer Waterways Experiment Station, P.O. Box
631, Vlcksburg, MS. I
U.S. EPA. 1991. U.S. Environmental Protection Agency. .Seminar
publication. Design and construction of RCRA/CERCLA finaj
covers. Center for Environmental Research Information. Cinr
cinnati, OH. EPA/625/4-91/025.
U.S. EPA. 1988a. U.S. Environmental Protection Agency. Reactivity
of various grouts to hazardous wastes and leachates.
EPA/600/2-88/021. Avail, from NTIS as #PB 88-182936/AS.
U.S. EPA. 1988b. U.S. Environmental Protection Agency. Lining of
waste containment and other impoundment facilities.
EPA/600/2-88/052. Risk Reduction Engineering Laboratory.
Cincinnati, OH.
U.S. EPA. 1988c. U.S. Environmental Protection Agency. Design,
construction, and evaluation of clay liners for waste manage-
ment facilities. EPA/530/SW-86/007F. OSWER, U.S. EPA,
Washington, DC.
U.S. EPA. 1988d. U.S. Environmental Protection Agency. Guide to
technical resources for the design of land disposal facilities.
EPA/625/6-88/018. CERI, U.S. EPA, Cincinnati, OH.
U.S. EPA. 1986a. U.S. Environmental Protection Agency. Technical
guidance document: construction quality assurance for haz-
ardous waste land disposal facilities. EPA 530/SW-86-031.
Avail, from NTIS as #PB 87-132825.
U.S. EPA. 1986b. U.S. Environmental Protection Agency. Grouting
techniques in bottom sealing of hazardous waste sites.
EPA/600/2-86-020. Avail, from NTIS as #PB 86-158664/AS.
53 pp.
U.S. EPA. 1984a. U.S. Environmental Protection Agency. Compati-
bility of grouts with hazardous wastes. EPA-600/2-84-015.
Avail, from NTIS as,#PB 84-139732.
U.S. EPA 1984b. U.S.
540/2-84-001.
Environmental Protection Agency. EPA
40
-------
CHAPTERS
In Situ Delivery and Recovery of Liquids
to Facilitate RCRA Corrective Actions
Lawrence C. Murdoch, Ph.D.
University of Cincinnati
This chapter reviews some conventional methods and
introduces some innovative methods for in situ delivery
and recovery of liquids that lend themselves to stabiliza-
tion activities. The first sections examine the performance
of delivery and recovery (Section 5.1) and review some
processes of liquid flow through porous material (Section
5.2). The majority of the chapter (Section 5.3) focuses on
field implementation, including conventional vertical wells
and interceptor trenches, and some of the newer tech-
niques, such as hydraulic fracturing, directional drilling,
and soil flushing. It is projected that these techniques, still
regarded as innovative, may increase the efficiency of
subsurface remediation in the corrective action-stabiliza-
tion areas. The use of these techniques in stabilization
actions is likely to effect greater contaminant mass re-
moval, and in less time, thus supporting two of the mass
stabilization goals. Section 5.4 presents a list of refer-
ences used as sources and for additional information.
5.1 PERFORMANCE
There are two types of performance: hydrologic perform-
ance, which includes local and field-scale processes, and
economic performance, which includes fixed and operat-
ing costs. Local processes refer to the performance of
equipment—pump pressures and flows, and power con-
sumption. Field processes, however, refer to what is hap-
pening on a larger scale within the aquifer, including
gradient control and remediation. The elements of interest
here are the flow directions of the fluids, flow through
porous material, the distribution of hydraulic head, and
the transport of contaminants. During remediation, other
elements of interest include the zones of contaminant
capture, the paths contaminants take to these wells,
travel time along those paths, and a variety of processes
that occur along the way, such as adsorption, degrada-
tion, and biotransformation.
A remedial system's design is based on hydrologic per-
formance, but economic performance is also critical, be-
cause cost can be an important decision-making criterion.
Capital costs for constructing the system and operating
and maintenance costs often dictate whether a remedial
design can be implemented. Thus, reducing costs is often
the impetus to improve design methods and seek inno-
vative techniques.
5.2 LIQUID FLOW
Figure 5-1 shows what happens when a nonaqueous
phase liquid (NAPL), such as oil, is displaced from soil.
The cross-hatched areas represent soil grains, the white
area oil, and the black area water. The flow is driven from
left to right by a pressure gradient, the difference in pres-
sure or hydraulic head between two points divided by the
distance between the points (Figure 5-1 a). The flow is
retarded, or resisted, by viscous forces due to the fluid
flowing through the channels; these retarding forces are
called hydraulic conductivity.
5.2.1 Flow Rate
For delivery and recovery, a parameter of particular inter-
est is the flow rate through porous materials. Cross-sec-
tional area, hydraulic conductivity, and hydraulic head
gradient are all parts of Darcy's law, which governs flow
through porous media. The hydraulic gradient and the
cross-sectional area can be affected by field techniques.
Hydraulic gradient can be affected by the use of a pump;
area can be affected by technique, such as drilling versus
use of a trench. Hydraulic conductivity and distribution
coefficient (Kd), however, are controlled by conditions at
the site.
Figure 5-1 shows the initiation of NAPL recovery, with oil
draining from some of the pores, and with water entering
them. In Figure 5-1 b, the flow of oil has dropped off,
shown schematically by the length of the white arrow.
Water has replaced the oil in these pores, and oil can no
longer use the pores for flow. Accordingly, the hydraulic
conductivity relative to the oil phase has decreased. The
41
-------
as a liquid. Instead, components of the oil phase will
partition into the water phase and will be transported with
the water. Thus, the distribution coefficients of the oil
phase and the aqueous phase, and the components of
the oil phase that are adsorbed onto the soil grains, be-
come essential to transport during the later stages of
recovery.
5.2.2 Hydraulic Conductivity
This section discusses how various methods of delivery
and recovery relate to their applicability in materials of
various hydraulic conductivities (e.g., tight, permeable).
Figure 5-2 shows hydraulic conductivity along the bottom
Figure 5-1. Aqueous phase displacing nonaqueous
phase (NAPL) in soil. i
i
hydraulic conductivity relative to one phase is thus som|e
funclion of the degree of saturation of that phase; as the
degree of saturation drops from 100 percent, so does the
hydraulic conductivity. As the recovery of oil proceeds,
the hydraulic conductivity relative to the oil phase wjll
eventually become so low that flow is impossible.
In another important method of transport, components of
the oil phase partition into the aqueous phase, and move
as dissolved components. While the oil is still flowing, this
is of minor importance. When the oil content decreases
to a low percentage, it will remain in the porous material
only as isolated blobs or ganglion between soil grains,
and as adsorbed films on the surfaces of soil grains.
There is no continuous oil phase, so the hydraulic con-
ductivity is essentially zero; that is, there is no flow of oil
Sand
Silt
Clay
I
0
I II I I
-2 -4
log K (cm/sec)
—i — r— i
-6 -8
Hydraulic Conductivity
Figure 5-2. Range of hydraulic conductivity where gen-
eral methods of delivery and recovery are regarded as
applicable.
band in terms of log hydraulic conductivity in centimeters
per second. The band above it shows some soil materi-
als, ranging from permeable sands to clays, and their
conductivities. The top four bands show various methods
of delivery and recovery and the types of soil materials
and conductivities for which they ara generally regarded
as applicable. The lower limit of applicability is cross-
hatched, because the lower limits of these techniques
with respect to hydraulic conductivity involve some un-
certainty. The gravity method, for example, which uses
infiltration galleries or sprinkler systems for delivering liq-
uids to unsaturated zones, relies on gravity head to pro-
duce hydraulic gradient. Gravity delivery methods are
limited by the hydraulic gradient achievable by gravity, so
it is only applicable in relatively permeable materials.
A pump will increase the hydraulic gradient, and surface
area can change depending on the surface exposed with
the chosen method. There is more surface area or more
screened length available for a horizontal well than a
vertical well, and more area exposed in a trench than in
a horizontal well. So it is possible to use these methods
to a greater degree in areas of lower hydraulic conduc-
42
-------
fi'vify. The arrow in Figure 5-2 represents a potential re-
duction in applicable hydraulic conductivity for vertical
wells, using various well stimulation techniques, such as
hydraulic fracturing.
5.2.3 Distribution Coefficients
There are a variety of distribution coefficients: the distri-
bution between components adsorbed onto soil grains in
the aqueous phase, between components in the
nonaqueous phase and aqueous phase, and between the
air phase and the water phase.
For a material to be successfully recovered using con-
ventional methods, the distribution coefficient expressed
in terms of the concentration in the mobile phase (the
phase to be recovered) versus the concentration in the
immobile phase should be greater than 10~2 to 10"3. Even
if the concentration in the mobile phase is 100 to 1,000
times less than the concentration in the immobile phase,
some components may still be recoverable. (Distribution
coefficients are discussed in more detail in Chapter 2.)
For additional information concerning the movement of
liquids, and the distribution coefficient, consult Section
5.4, particularly the works by Bear (1979), Chambers et
al. (1991), Repa and Kufs (1985), U.S. EPA (1987), and
Willhite (1986).
5.3 FIELD METHODS
The types of field techniques discussed in this section
include:
• Vertical wells
• Hydraulic fracturing
• Interceptor trenches
• Horizontal wells
• Soil flushing
The following sections will describe the types of wells,
general applications, performance, and advantages and
limitations of these five techniques.
5.3.1 Vertical Wells
5.3.1.1 Types
Vertical wells are the most popular method of delivery
and recovery. They can be categorized by the type of
completion technique and the pump configuration. Com-
pletion techniques refer to the way a borehole is modified
to make a well. Typically, this involves putting a pipe in
the ground to prevent a borehole from collapsing. The
pipe is slotted, and, in many cases, some kind of a filter
material, such as sand or gravel, is used in the annulus
to prevent fine grain sediment from entering the well and
damaging the pumps.
The water supply industry provides many pumps used in
vertical wells, and there are some specialized pumps for
the environmental industry. Turbine pumps and a variety
of centrifugal pumps are the most popular methods of
recovering water (Driscoll, 1986). In some cases, the
pumps are mounted at the ground surface and lift water
out of the well with a vacuum, in much the same way that •
water is sucked out of a glass with a straw. It other cases,
pumps are submerged in the well and push the water up
to the ground surface.
There are a variety of pumps for environmental applica-
tions (Blake and Lewis, 1982; U.S. EPA, 1987). The gas-
driven pump, for example, is a popular device, of which
a bladder pump is a special variety. With a bladder pump,
a chamber, containing check valves and a gas line that
goes down to the pump, is lowered below the water table.
Water or fluid in the well flows into the chamber, the check
valve is closed, and gas is forced into the chamber, push-
ing the fluid up to the ground surface. One of the reasons
for this pump's popularity is its gentleness; if multiple
fluids are in the well, it will not mix them up as would a
turbine pump. When the fluids are lifted to the ground
surface by a gas-driven pump, they are relatively easy to
separate.
A skimmer pump is an attachment to the intake of a
pump, consisting of a flexible tube and float that ride up
and down with the water level. The float is dense enough
to sit on the interface between the water and floating
product and can skim free product off the water table.
Figure 5-3 shows a water-level activated pump with a
level control, which consists of a pump and a switch
device that turns the pump on and off in response to the
level of water in the well, holding the level constant or
within a short tolerance. There are two reasons for level
Vertical Well
Water-Level Activated Pump
Water + NAPL
Level
Control
NAPL Layer
Pump
Figure 5-3. Vertical well—water-level activated pump.
43
-------
control. For some environmental applications, wells are|
put into formations that are poor aquifers, which are un-i
able to supply water to the well as fast as the pump!
removes it. Therefore, the pump has to be turned off|
periodically to prevent the well from being pumped dryl
and damaging the pump. In addition, a level control de-j
vice can hold the water level at the intake, allowing thej
pump to gather floating free product. ''
A well using water level control is fundamentally different!
from water supply wells. The pump in a water supply type
well supplies a constant flow rate, and the drawdown
measurement gives the local performance of the well. Ifi
the drawdown is fixed, however, the parameter of interest!
is the flow rate. The flow rate will vary as a function of
time; under ideal cases, it will decrease with time, thus,
providing fundamentally different performance on a local
scale. .1
A slight variation of the level control pump is the two-;
pump system shown in Figure 5-4. One pump lifts the[
water and creates a cone of depression, down which the,
floating product flows, accumulating in the well. A level.
Vertical Well
Two-Pump System
Level Control
NAPL Pump
Water I
Pump
Figure 5-4. Vertical well—two-pump system.
switch holds the water position constant, and the NAPL
pump lifts out the floating product. This kind of a systerrji
is used to prevent the mixing of free product and water.
Figure 5-4 shows a pump used for light nonaqueous
phase liquids (LNAPLs); a similar version is used fo[-
dense nonaqueous phase liquids (DNAPLs). With
DNAPLs, the water pump is still used to create a cone of
depression, but, as a result, a zone of relatively low pres-
sure is created in the well, and the DNAPL bulges upward
and is more easily recovered.
5.3.1.2 Applications
The remedial manager has several technology options for
contaminated ground water at a site. In some cases, a
site may be effectively stabilized simply by lowering the
ground-water level below the level of the site.
The following are a couple of idealized examples of the
use of vertical wells. (For additional case studies, see
U.S. EPA, 1989a,b,c.) Figure 5-5 shows regional flow
from top to bottom approaching an irregularly shaped
contaminated area. One way of addressing this situation
is shown in Figure 5-6. The fine lines in the figure are
streamlines that go to the three wells. The heavy dotted
line defines the capture zone for these three wells; that
is, water on the upper side of the capture zone will make
it to the wells, while the water on the bottom side of the
line will be lost. This approach, using extraction wells, is
applied when there is relatively high regional flow and
Regional Flow
Figure 5-5. Regional flow from top to bottom approach-
ing an irregularly shaped contaminated area.
Figure 5-6. Capture of contaminants by a line of extrac-
tion wells. .
;,44
-------
relatively low adsorption of contaminants. The strategy is
to place the line of extraction wells on the downgradient
side of the contaminants, and use the regional gradient
to sweep the contaminants into the wells. The design
criteria balance the yield of these wells and the well spac-
ing against the regional flow rate. If the spacing and flow
rate of the wells are insufficient to remove the regional
flow, then some of the contaminants can escape between
the wells.
In cases where the regional flow is low, the line of extrac-
tion wells can be supplemented with injection wells to
increase hydraulic gradient. Figure 5-7 shows the con-
taminated area with the same three extraction wells, and
Figure 5-7. Approach to low regional flow using three ex-
traction wells and four injection wells.
an additional four injection wells. For this simulation, the
rate of pumping at these wells is balanced with the rate
of injection. This results in an efficient sweep into these
wells. There are two important points to consider with this
kind of an approach. First, the injection wells must be
located outside the contaminated area or contaminants
can be driven away from extraction wells, potentially
spreading contamination. The remedial manager must
define the zone of contamination before considering using
injection wells. Second, it is important to consider where
stagnation zones, areas where the flow rate is negligible,
may occur. If there is limited flow in an area, the area will
act as a persistent source, allowing contaminants to
slowly bleed into the extraction wells over time. Stagna-
tion can occur any time there is more than one well,
regardless of the pattern of extraction and injection, so
long as the pumping rate remains constant. The key to
addressing this problem is to vary the pumping rates so
the stagnation point moves around, and the contaminants
can be swept out of these areas, without leading to con-
tamination of uncontaminated areas.
For additional reading about the applications of vertical,
wells, consult references in Section 5.4, particularly U.S.
EPA (1989a,b,c), Haley et al. (1991), Keely and Tsang
(1983), Keely (1984), and Repa and Kufs (1985).
5.3.1.3 Performance
One measure of performance of an extraction well is the
concentration of contaminants from the extraction well
over time. The typical performance is a high initial con-
centration, followed by a rapid dropping off and then a
gradual leveling out (see Figure 5-8). Ideally, the concen-
trations then decrease to negligible values (Figure 5-8a).
Often, however, the performance is not so favorable;
there is a high concentration, followed by a falling off after
the start of pumping. The concentration never really drops
to zero, however, and persists above the regulatory limit
(Figure 5-8b). This type of performance is the result of a
persistent source, either a stagnation point or contami-
nants that have adsorbed onto soil grains and are slowly
desorbing, or some other process. Figure 5-8c shows an
even worse situation. The concentration drops off at first,
but is followed by a minor increase. Then concentration
falls again, until pumping stops, after which the concen-
tration increases markedly. This pattern indicates an ob-
viously persistent source, and is seen often in both
pump-and-treat systems and vapor extraction systems.
In the situation of a persistent source, the remedial man-
ager has the option to shut off the wells prematurely or
to continue pumping at a very low rate, allowing the con-
centrations to increase. The pumping rate is then raised
to recover water with a relatively high concentration of
contaminants. This method is called pulse pumping and
has been the focus of recent attention. Pulse pumping
decreases the volume of- contaminated water to be
treated, and so may decrease remediation costs. It can
cause problems, however; by turning the-pumps down or
off, the remedial manager may lose hydrologic control of
the site.
5.3.1,4 Advantages and Disadvantages
There are several data issues that must be considered
when installing vertical wells, including access, conduc-
tivity, adsorption, depth, NAPL, and heterogeneity (Bear,
1979; Driscoll, 1986; Repa and Kufs, 1985). Obviously,
the location for the potential well site must be accessible
for the drill rig. Hydraulic conductivity at the site must be
greater than 10~5 cm/s for conventional applications. For
adsorption, as mentioned before, the distribution coeffi-
cient must be less than 10"3, and the depth must be 8
meters for suction lift, if the pumps are at the ground
surface. If downhole pumps or submerged pumps are
being used, there is no depth limit. With regard to NAPLs,
the important question is whether their density is greater
or less than water. Heterogeneities may cause problems
for vertical wells or any other method of delivery and
recovery, but such problems may be addressed by cycling
the flow rates (Keely, 1982 and 1983; U.S. EPA, 1990).
45
-------
1.20O
1982 1983 1984 1985 1986 1987 1988
(a)
1.500-
1.000-
u
500-
1982 1983 1984 1985 1986 1987 1988
14000
13500'
5500'
5000'
4500'
4000'
3500'
3000<
2500'
2000
1500
1000
500
0
JJ JO JJ JDJJ JDJJ JDJJ JDJJ JDJJ JDJJ JDJJ JDJJ JD JJ JD
78 78 79 79 80 80 81 81 82 82 83 83 84 84 85 85 86 86 87 87 88 88
Figure 5-8. Concentration from the extraction well as a
function of time. Figure 5-8a represents ideal performance
(U.S. EPA, 1989a). Figure 5-8b shows concentrations
above the regulatory limit (U.S. EPA, 1989b). Figure 5-8c
shows an obviously persistent source (U.S. EPA, 1989c). j
Operation and maintenance costs for vertical wells can
be fairly high. The radial geometry of flow paths caused
by a vertical well also can be a disadvantage. Because
of its radial geometry, the flow paths will converge on the
well, and energy is required to make flow converge. It is
more efficient to move water through porous material if
the flow lines remain parallel. This disadvantage can be
overcome, however, by changing the geometry of the
recovery system by using a trench or a directional well.
5.3.2 Hydraulic Fracturing
Hydraulic fracturing, a technique developed in the oil in-
dustry about 50 years ago, has received quite a bit of
attention as a method of stimulating well yields. Today,
almost every oil well is hydraulically fractured, because
this technique consistently increases the yield to the well.
Hydraulic fracturing has also been used effectively in
water wells, particularly those drilled in hard rock.
One of the potential problems of hydrofracture is intro-
ducing uncontrolled fractures (vertical and horizontal) that
could contaminate new areas. The technology should
work well in highly stratified, over-consolidated deposits
(glacial, loess, and lacustrine).
5.3.2.1 Method
Over the past 5 years, University of Cincinnati scientists
have shown that hydraulic fractures can be made at shal-
low depths, such as 5 feet in silty clay. This technique
increases the effective permeability, the radius of influ-
ence, and the yield of vertical wells in unconsolidated
formations. Hydraulic fractures also can be used to de-
liver solids. The creation of a hydraulic fracture begins
with a casing that is sealed in the ground and open at
the bottom. A pump injects fluid into the casing, causing
the pressure to increase until it exceeds some critical
value and the fracture is nucleated. The fracture grows
out away from the well, and a granular solid, such as
sand or even a slow release nutrient for bioremediation,
is pumped in to fill the fracture. When pumping ceases,
the fracture closes and the solid remains to prop open
the fracture and create a pancake-shaped layer (Figure
5-9).
Equipment for creating fractures at contaminated sites
includes a device to store and meter sand, a tank for the
fluid, a slurry mixer, a pump, and a special gel used to
suspend the sand during pumping. Leveling methods are
used to measure the uplift, or doming, which is usually
about an inch when the fractures are created.
5.3.2.2 Applications
Hydraulic fracturing has applications for pump and treat,
vapor extraction, soil flushing, and bioremediation. It al-
ready has been used for some pump-and-treat operations
in crystalline rock in New York. It also is used in some
cases to increase the area over which a monitoring well
can sample, and it can be used to recover leachate. The
flat-lying pancake-shaped geometry created by fracturing
is ideal for any of these applications.
For bioremediation, this technique can be used to in-
crease the injection of oxygen dissolved in water, or hy-
drogen peroxide. A recent innovation is the pumping of
a solid compound consisting of encapsulated oxygen into
46
-------
Hydrofractures:
How They Form
Inject Sand Slurry to
Prop Open Fractures
Sand-Filled Hydrofracture
UAV't
»_•*. t
Hydrofracuring Occurs When
Injection Pressure > Confining Pressure
Figure 5-9. Creation of hydraulic fracture.
hydraulic fractures. After being pumped into the fracture,
the capsules slowly release oxygen for bioremediation.
One example of an application of hydraulic fracturing is
a Chicago industrial facility contaminated with solvent,
where remedial engineers wanted to use vapor extrac-
tion. The site is. underlain by silty clay with low perme-
ability, and hydraulic fractures were used to increase the
flow rates. The remedial engineers put in three fractures,
one each at 5, 10, and 15 feet. Each fracture was indi-
vidually screened, so a vacuum or air inlet could be in-
stalled and controlled separately.
During one study, fractures were created that resembled
those at the Chicago site. Figure 5-10 shows a map of
three fractures that were made at one borehole, labeled
EL6, in an area that was excavated to reveal the details
of the fractures. The figure shows several trenches, indi-
cated by dotted lines, that were used to expose the hy-
draulic fracture. The heavy lines are the outlines of the
fractures. Figures 5-11 a and b show the view along the
trench, with the lower, middle, and upper fractures. The
hydraulic fracture itself looks like a sand bed roughly 1/2
inch thick, filled with coarse-grained, well-sorted sand,
that is sandwiched between silty clay glacial material.
At another site where fractures were also created in silty
clay, preliminary measurements of flow rate and pressure
were taken (Figure 5-12). Over the first week or so, the
flow was about 5 cfm from the fractured well. The flow
rate fluctuated, decreasing when water infiltrated after
rainfall and increasing when the water was removed from
the well, but typically ranged between 3 and 4 cfm. At a
nearby conventional well, used for control, the yield was
a uniform 0.30 cfm. Thus, the yield from the fractured well
Figure 5-10. Map of three fractures created at borehole
EL6.
1 Meter
EL6
*• 3
Trench B - East Wall
Figure 5-11. Cross-sections of three fractures created at
borehole EL6 showing topographic profile (above) and de-
tails of fracture traces (below).
was roughly an order of magnitude greater than from a
conventional well. Similarly, the radius of influence of the
conventional well was found to be between 2 and 3 feet,
whereas it was roughly 25 feet from the fractured well.
5.3.2.3 Advantages and Disadvantages
The advantages of hydraulic fracturing include the ability
to increase well yield from tight materials, increase the
radius of influence, form a flat-lying planar sink thus
avoiding some geometry problems, and inject solids. The
47
-------
€C
o.
1.5
i
o
-JQ. » D~"HIL^
\ v n
\
1-cA
'01. \
w
0 5
(Meters)
3.0 4.5 6.0 7.5
i i i i
Approx. Limit 1
otFRX |
*^ ji r—i
"~»^ LJ
~^~-^
Q Conventional Well
y Frx Well, above frx
D Frx Well, in frx
i i i
-2.5
f-0.25 £
- 0.025
- 0.0025
10 15 20 25
Radius from Well (feet)
c
(0
DC
"O
to
»!
td Q}
__l£
c
(0
DC
T3
.
-------
5.3.3.1 Construction
A large backhoe is typically used to build interceptor
trenches. One method of keeping the trench open is to
fill it with guar gum, or other gel, in a manner similar to
the filling of cutoff trenches with bentonite slurry. Guar
gum gel will not penetrate into the trench wall, so it will
not plug soil pores as can a bentonite slurry. Guar gum
is also used for hydraulic fracturing and can be purchased
from a number of vendors. Figure 5-14 shows a side view,
Monitoring Well
Clean
Gravel
Extraction Well
100 ft to 250 ft
Typical
Well
Screen
Figure 5-14. Side view of an interceptor trench.
normal to the plane of the trench. The trench has been
backfilled with gravel, and a pipe has been inserted to
house a pump. The data issues that need to be con-
sidered in constructing trenches are access; hydraulic
conductivity, which must be greater than 10"6 cm/sec;
adsorption; and depth, which should not exceed 40 to
65ft.
5.3.3.2 Applications
The general applications for trenches are 1) to intercept
ground water, 2) to lower the water table, 3) to stabilize
contaminant migration, and 4) to use with a barrier. An
example of the fourth application is using a trench on one
side of the contaminants and a barrier on the other to
prevent uncontaminated water from entering the trench.
Uncontaminated water would dilute the contamination,
increasing the volume of contaminant to be treated.
5.3.3.3 Advantages and Disadvantages
Trenches have a better capture effectiveness than vertical
wells and are relatively simple to construct, especially the
shallow ones. Trenches also have a greater yield in tight
soils than wells and, as a result, can have lower operating
and maintenance costs. The disadvantages are that ex-
cavated contaminated material from trenches could re-
quire disposal as hazardous waste, and access can be
difficult, especially with a large backhoe. In addition, the
depth limit for conventional excavation techniques is
about 15 feet, although more specialized trenches can be
built to a depth of 40 to 65 feet. Another disadvantage of
trenches is the difficulty of abandoning them.
For additional information concerning trenches, consult
the references in Section 5.4, particularly Day (1991),
Gilbert and Gress (1987), Meini et al. (1990), Repa and
Kufs (1985), U.S. EPA (1990), and Zheng et al. (1988).
5.3.4 Directional Drilling
If the remedial manager desires the linear geometry of a
trench, but is constrained by some of the disadvantages
mentioned above, one option is directional drilling. Direc-
tional drilling, which installs horizontal wells, will improve
access and will have some geometric advantages such
as greater working surface area. It also can address prob-
lems related to heterogeneities, such as preferred flow
paths induced by vertical fractures. This method of drilling
is gaining in popularity.
Directional drilling, as defined here, requires a rig that is
surface launched and has a drilling head that can be
steered in any direction. It also must include an electronic
device to monitor location. Two systems meet these re-
quirements: one from the oil industry and one from the
utility industry.
The oil industry directional drilling rigs used in remedial
actions are scaled down from oil well drilling applications.
They are state-of-the-art, and the major manufacturers
are Eastman Christensen (Karlson and Bittb, 1990) and
Petrolphysics (Dickinson et al., 1986; 1987; and 1991).
Eastman Christensen has already installed several wells,
some of them at the Savannah River site. They have
conducted demonstrations using vapor extraction with
those wells, and have obtained some promising results.
The trajectory of the borehole from an Eastman Christen-
sen rig starts on an incline and then curves to the hori-
zontal after about 60 feet, at which point it can continue
horizontally for several hundred feet (Figure 5-15). The
Petrolphysics rig also puts in a horizontal well, but the
geometry is somewhat different, perhaps significantly de-
pending upon the application. This equipment starts with
a vertical hole and turns sharply to a horizontal (Figure
5-16). These horizontals can go out at various depths,
even stacked on top of one another or at different angles
like the spokes of a wheel. This configuration has obvious
advantages in tight locations where there is no room to
gradually achieve the horizontal.
The directional drilling rigs from the utility industry (Wem-
ble et al., 1990) are relatively small rigs, about an order
of magnitude smaller than the oil drilling rigs. They are
used to drill holes to install utilities, electric lines, and gas
lines. Manufacturers include Underground Technologies
and Ditchwitch. Utility rigs are smaller, more compact,
49
-------
Horizontal Wellbore System
Depth Capability
Unit Can Reach Horizontal at Any Specified Depth
.Vertical Section
60ft-
18°TVD
K"
End of
Curve
Variable at 15 ft Increments
Figure 5-15. The trajectory of the borehole from the East-
man Christensen rig.
• Coil Tubing Injector
High Pressure Pump
Coil Tubing Reel ;
Sealed Catchment Cistern
Whipstock
Cement Plug
Seal
Advancing Radial Tube
10-200 feet below surface I
This technique currently uses a radio transmitter and re-
ceiver that monitors location and depth from the ground
surface.
5.3.4.2 Advantages and Disadvantages
Figure 5-17 shows three situations where directional drill-
ing can be applied. Directional drilling increases accessi-
bility, because it can go underneath structures, allowing
for stabilization under buildings. This makes it easier to
use in retrofitting landfills with leachate collection systems
or monitoring and sampling from beneath tanks and other
potential sources that cannot be drilled through (Dickin-
son et al., 1986 and 1987; Wemble et al., 1990). Direc-
tional wells also will show improved performance in
situations where contaminants occur as layers, either as
a floating layer or resting along the interface between
bedrock and overburden. There is more screened length
with a directional well than with a vertical well. Where
vertical fractures provide the majority of permeability, di-
rectional wells also are likely to intersect many more of
those vertical fractures (Dickinson et al., 1986 and 1987).
Access
Figure 5-16. Petrolphysics rig with a shallow radipl
system.
more maneuverable, and more easily mobilized than the
oil industry rigs.
Utility industry rigs can have remarkable directional ca-
pabilities. It is possible for them to drill borings that go
down and level off, then take a right-hand turn and curj/e
around, and keep curving until they essentially drill a bore
that looks like a horseshoe, turning 180° and coming up
about 100 feet away. Directional drilling allows the reme-
dial manager to tailor the well to the specific situation.
Contaminant Layer
Vertical Fractures
Figure 5-17. Three situations where directional drilling
could be beneficial.
One of the disadvantages of this technique is cost, which
ranges widely for the two systems. The oil industry rigs
generally cost approximately $100 to $300 per foot; the
utility rigs cost $20 to $50 a foot. Another potential prob-
lem is availability. Eastman Christensen and Petrol-
physics are currently available to install wells, but they
only have a few rigs and mobilization may be a factor.
The utility rigs are widely available, but, so far, they have
only been used to create wells in a few locations.
50
-------
Techniques for sampling with, both types of rigs are still
urvier development, although some techniques are cur-
rently in the prototype stage.
Hole collapse for bores made with utility rigs can be a
problem, depending on the formation. The initial hole that
is drilled is relatively small—2% to 43/4 inches. After drill-
ing the small hole, however, the engineer can attach a
reamer and enlarge the hole to 10 or even 24 inches or
more, depending upon the rig. One of the techniques
under development is the use of a reamer behind the drill
head (about 20 feet back), so that the small 2-inch hole
is enlarged as it is created.
Completion techniques are available for holes that stay
open, but holes that collapse require different techniques.
The most reliable way to accomplish completion in the
latter case is to advance the casing as the hole is drilled.
For additional information concerning directional drilling,
consult the references in Section 5.4, particularly Dickin-
son et al. (1986), Dickinson et al. (1987), Dickinson et ai.
(1991), Kaback et al. (1989), Kaback et al. (1991),
Karlsson and Bitto (1990), Langseth (1990), Looney et
al. (1991), Wemble et al. (1990).
5.3.5 Soil Flushing
The previous sections in this chapter have reviewed vari-
ous geometries of physical systems for delivery and re-
covery. In some situations, the contaminants are tightly
bound to soil grains, so no matter what the geometry of
the physical system, it is necessary to inject fluids to
enhance recovery. Soil flushing is a technique of deliver-
ing fluids to enhance recovery.
5.3.5.1 Applications
Soil flushing has a variety of applications, some in initially
unsaturated and others in saturated ground. In saturated
ground, infiltration techniques that target certain types of
contaminants, such as bioreactive, semi-volatile, and
inorganics, have been used.. With initially saturated
ground, soil flushing can be used on water soluble com-
pounds, low solubility compounds, or NAPLs.
5.3.5.2 Methods
When many people refer to soil flushing, they mean
chemical flooding. Soil flushing, or chemical flooding, in-
volves mixing a compound with water and injecting it to
increase mobility (Amdurer et al., 1986; Begor et al.,
1989; Carter and Knox, 1986; Trost et al., 1989) or deg-
radation (Amdurer et al., 1986; Chambers et al., 1991) of
contaminants. Mobility is enhanced, for example, by add-
ing a surfactant for oily compounds, or a viscosifier to
help push out thick fluids. For some inorganics, mobility
can be enhanced by altering pH and redox potential.
5.3.5.3 Data Issues
Some of the issues at a site that must be considered for
soil flushing include heterogeneities, ground-water and
soil chemistry, adsorption, and biodegradation (Amdurer
et al., 1986). Probably the most important consideration
is achievement of hydrologic control of the site, so that
the effluent can be captured. If not contained and recov-
ered, the flushing agent can mobilize the contaminant to
new areas, exacerbating the problem. In some cases, the
compounds used for chemical flooding are fairly toxic or
hazardous, so they must be recovered. At the Laramie
Tie site, for example, an extensive slurry wall and array
of wells are used to achieve hydrologic control and ensure
recovery of an alkaline surfactant used during chemical
flooding.
With respect to contaminants, an important consideration
is the octanol-water partitioning coefficient. If the K0 is
less than 1, soil flushing can be accomplished with water
alone. If K0 is greater than 3, a surfactant increases re-
moval efficiency (U.S. EPA, 1990a). In the "gray area" in
between, the necessity for a surfactant depends on site
characteristics. Before considering soil flushing at any
site, the remedial manager must perform bench-scale
studies as well as carefully controlled pilot studies (Sale
et al., 1989; Trost et al., 1989).
For additional information concerning soil flushing, con-
sult the references in Section 5.4, particularly Amdurer et
al. (1986), Chambers et al. (1991), Repa and Kufs (1985),
Sale et al. (1989), U.S. EPA (1990), Trost et al. (1989).
. - !
5.3.6 Other Techniques
There are a variety of other techniques for addressing
problems with moving liquids, including heating, freezing,
and electrical techniques. Heating by steam injection, ra-
dio frequency radiation, or electrical resistance can in-
crease mobility of some liquids and increase recovery by
driving liquids into the vapor phase. Ground freezing has
been used to enhance migration of contaminants, which
are excluded from ice crystals and pushed ahead of a
freezing front. Electrical osmosis uses electrical gradients
to drive fluid flow and is especially effective in fine-grained
soils.
5.4 REFERENCES
Amdurer, M., R.T. Fellman, J. Roetzer, and C. Russ. 1986. Systems
to accelerate in situ stabilization of waste deposits. Septem-
ber. EPA/540/2-86/002.
Bear, J. 1979. Hydraulics of groundwater. McGraw-Hill, New York,
569 p.
Begor, K.F., M.A. Miller, and R.W. Sutch. 1989. Creation of an
artificially produced fracture zone to prevent contaminated
ground-water migration. Ground Water. 27(1):57-65.
Blake, S.B. and R.W. Lewis. 1982. Underground oil recovery. In:
Proceedings of the 2nd national symposium on aquifer res-
toration and groundwater monitoring. NWWA. May 26-28. pp.
69-76.
Canter, LW. and R.C. Knox. 1986. Groundwater pollution control.
Lewis Publishers, Chelsea, Ml.
Chambers, C.D., J.W. Willis, S. Giti-Pour, J.L. Zieleniewski, J.F.
Rickenbaugh, M.I. Mecca, B. Pasin, R.C. Sims, D.L Soren-
51
-------
sen, J.L Sims, J.E. McLean, R. Mahmood, R.R. Dupont, and
K. Wagner. 1991. In situ treatment of hazardous-waste con-
taminated soils. Noyes Data Corp. Park Ridge, NJ. 533 pj
Davis-Hoover, W.J., LC. Murdoch, S.J. Vesper, H. Pahren, O.L.
Sprocket, C.L Chang, A. Hussain, and W.A. Ritschel. 199,1.
Hydraulic fracturing to improve nutrient and oxygen delivery
for in situ bioreclamation. In: In situ bioreclamation, R.E.
Hlnchee and R.F. Olfenbuttel (eds.), Butterworth-Heinemann,
Boston, pp. 67-82.
Day, S.R. 1991. Extraction/interception trenches by the bio-polymer
slurry drainage trench technique. Hazardous Materials Cop-
trol. Sept/Oct, 27-31. ;
Dickinson, W.W., R.W. Dickinson, T. Crosby, and H.N. Head. 198J6.
Horizontal drilling beneath Superfund sites. Presented at
conf. on management of uncontrolled hazardous waste sites,
pp. 258-263. :
Dickinson, W.W., R.W. Dickinson, P.A. Mote, and J.S. Nelson. 1987.
Horizontal radials for geophysics and hazardous waste re-
mediation. In: Superfund '87 Hazardous Waste Control Re-
search Institute. Silver Spring, MD, pp. 371-375.
Dickinson, W.W., R.W. Dickinson, J. Nees, and E. Dickinson. 199,1.
Field production results with the ultrashort radius radial sys-
tem In unconsolidated sandstone formations. Presented at
the 5th UNITAR/UNDP international conference on heayy
crude and tar sands, August 4-9, Caracas, Venezuela.
Driscoll, F.G. 1986. Groundwater and wells. Johnson Division, Min-
neapolis, MN, pp. 1089.
GIdley, J.L., S.A. Holditch. D.E. Nierode, and R.W. Veatch. 198^9.
Recent advances in hydraulic fracturing. Soc. Petroleum En-
gineers AIME, New York, 452 p. •
Gilbert, S.G. and J.J. Gress. 1987. Interceptor trenches for positive
ground-water control. Ground-Water Monitoring Reviejw.
Spring, pp. 55-59. >
Halay, J.L, B. Hanson, C. Enfield, and J. Glass. 1991. Evaluating
the effectiveness of ground-water extraction systems.
Ground-Water Monitoring Review, Winter, pp. 119-124.
Hillel, D. 1980. Fundamentals of soil physics. Academic Press, San
Diego, 412 p. I
Howard, G.C. and C.R. Fast. 1970. Hydraulic fracturing. Soc. P|e-
troleum Engineers AIME, New York, 198 p.
Kaback, D.S., B.B. Looney, J.C. Corey, L.M. Wright, and J.L. Steele.
1989. Horizontal wells for in situ remediation of ground water
and soils. In: Proceedings 3rd nat. outdoor action conf. on
aquifer restoration, ground-water monitoring, and physical
methods, NWWA, Dublin, OH, pp. 121-135. j
Kaback, D.S., B.B. Looney, C.A. Eddy, and T.C. Hazen. 1991. In-
novative ground-water and soil remediation: in situ air strip-
ping using horizontal wells. In: Proc 5th nat. outdoor action
conf. on aquifer restoration, ground-water monitoring, and
physical methods, NWWA, Dublin, OH, pp. 47-58. j
Karlsson, H. and R. Bitto. 1990. New horizontal wellbore system
for monitoring and remedial wells. In: Superfund '90 Hazard-
ous Waste Control Research Institute, Silver Spring, MD, pp.
357-302. |
Keety, J.F. and C.F. Tsang. 1983. Velocity plots and capture zonjes
of pumping centers for ground-water investigations. Ground
Water. 21(6):701-714. •
Keely, J.F. 1982. Chemical time series sampling. Ground-Water
Monitoring Review. Fall, pp. 29-38. ;
Keety, J.F. 1983. Field application of chemical time-series sampling.
Ground-Water Monitoring Review. Fall, pp. 26-33.
Keely, J.F. 1984. Optimizing pumping strategies for contaminant
studies and remedial actions. Ground-Water Monitoring Re-
view, Summer 4(3):136-73.
Keely, J.F. Performance evaluations of pump and treat remedia-
tions. Ground-Water Issue, EPA/540/4-89/005. 19 p.
Langseth, D.E. 1990. Hydraulic performance of horizontal wells. In:
Superfund '90 Hazardous Waste Control Research Institute,
Silver Spring, MD, pp. 389-408.
Looney, B.B., D.S. Kaback, and J.C. Corey. 1991. Field demonstra-
tion of environmental restoration using horizontal wells. Pre-
sented at 3rd forum on innovative hazardous waste treatment
technologies: domestic and international, June 11-13.
Meiri, D., M. Ghiasi, R.J. Patterson, N. Ramanujam, and M.P. Ty-
son. 1990. Extraction of TCE-contaminated ground water by
subsurface drains and a pumping well. Ground Water.
28(1): 17-24.
Murdoch, L.C., G. Losonsky, P. Cluxton, B. Patterson, I. Klich, and
B. Braswell. Feasibility of hydraulic fracturing of soil to im-
prove remedial actions. EPA 600/2-91-012. NTIS #PB 91-
181818.
Murdoch, L.C. 1990. A field test of hydraulic fracturing in glacial
till. In: Proceedings 15th annual USEPA research sympo-
sium, Cincinnati, OH. April 10-12, 1989. EPA/600/9-90/006.
February.
Murdoch, L.C., G. Losonsky, I. Klich, and P. Cluxton. 1990. Hydrau-
lic fracturing to increase fluid flow. In: Contaminated soils,
'90. F. Arendt, M. Hinsenveld, and W.J. van den Brink (eds.),
Kluwer Academic Publishers, Netherlands, pp. 1087-1093.
Sale, T., K. Piontek, and M. Pitts. 1989. Chemically enhanced in
situ soil washing. Presented at petroleum hydrocarbons and
organic chemicals in ground water: prevention, detection, and
restoration, NWWA, Nov. 15-17.
Sims, R.C., D.L. Sorensen, J.L. Sims, J,E. McLean, R. Mahmood,
and R.R. Dupont. 1991. Review of in-place treatment tech-
niques for contaminated surface soils. USEPA Report, Con-
tract no. 68-03-3113. Also published as part of Chambers et,
al. (1991).
Smith, S.A. 1989. Manual of hydraulic fracturing for well stimulation
and geologic studies. NWWA. 66 p.
Trost, P.B., G.A. Pouska, and M. Day. 1989. APS remediation of a
shallow aquifer containing viscous oil. Presented at 6th na-
tional RCRA/Superfund conference. New Orleans, LA. April.
U.S. EPA. 1990a. U.S. Environmental Protection Agency. Handbook
on in situ treatment of hazardous waste-contaminated soils.
EPA/540/2-90/002. January.
U.S. EPA. 1990b. U.S. Environmental Protection Agency. Subsur-
face contamination reference guide. EPA/540/2-90/011. Oc-
tober.
U.S. EPA. 1989a,b,c. U.S. Environmental Protection Agency. Evalu-
ation of ground-water extraction remedies, Volumes 1, 2, and
3. EPA/540/2-89/054.
U.S. EPA. 1987. U.S. Environmental Protection Agency. Under-
ground storage tank corrective action technologies.
EPA/625/6-87-015. January.
U.S. EPA. 1985. U.S. Environmental Protection Agency. Leachate
plume management. EPA/540/2-85/004. November.
Wemble, R., P.P. Lysne, and R. Jacobson. 1990. Recent field trials
of directional boring equipment for emplacing a borehole grid
around and beneath a simulated waste site. Presented at
1990 DOE model conference, Oakridge, Tenn. October 29-
Nov 1.
Willhite, G.P. 1986. Waterflooding. Soc. Petroleum Engineers AIME,
New York, 326 p.
Zheng, C., K.R. Bradbury, and M.P. Anderson. 1988. Role of inter-
ceptor ditches in limiting the spread of contaminants in
ground water. Ground Water. 26(6):734-742.
52
-------
CHAPTERS
Vapor Extraction, Bioventing
Ronald Sims, Ph.D.
Division of Environmental Engineering
Utah State University
and
Ryan Dupont, Ph.D.
Utah Water Research Laboratory
Utah State University
6.1 INTRODUCTION
This chapter presents the fundamentals of both vacuum
extraction and bioventing and some of their potential ap-
plications in the realm of stabilization. Vapor extraction,
or gas control, has been referred to as soil vapor extrac-
tion (SVE), forced air venting, and in situ air stripping.
The process involves extraction of air from unsaturated
soil, and is applied in the vadose zone. Moving air through
the subsurface by vacuum extraction and moving water
through the subsurface by pumping and treating are par-
allel types of processes, applied to different media. Sec-
tion 6.2 describes the process of vacuum extraction, and
Section 6.3 describes some of the applications of this
technology. In many situations, SVE systems can be
modified to introduce oxygen to the subsurface to en-
hance degradation. Section 6.4 discusses the use of
bioventing to enhance the biodegradation of soil chemi-
cals by microorganisms. Section 6.5 is a list of references
for further information.
Vapor extraction technology offers several remedial bene-
fits and can be used to stabilize sites by controlling the
migration of subsurface vapors, preventing vapors from
affecting people or the environment, removing explosive
vapors or harmful gases, removing sources of light and
dense nonaqueous phase liquids in the vadose zone, and
enhancing bioremediation. The following is a simplified
discussion of vapor extraction system principles.
6.2 PROCESS DESCRIPTION
6.2.1 System Components
The principle of vapor extraction is based on the Henry's
Law constant, or air/water partition coefficient (Kh); the
concentration of a chemical in air compared to the con-
centration of the chemical in water (Ca/Cw). The higher
the Kh, the greater the tendency for the chemical to par-
tition into the air phase of the subsurface.
Figure 6-1 is a schematic of ventilated soil. The arrows
represent advection, or air that is being pushed or pulled
through the zone under pressure or suction. In the figure,
the air is going through a subsurface system, where the
irregularly shaped material represents soil particles, such
as sand, silt, or clay. Surrounding those particles under
most field conditions, except in arid areas, is a layer of
water. Chemicals in the water around the soil will partition
Ventilated Soil
Figure 6-1. Schematic of ventilated soil (Valsaraj and Thi-
bodeaux, 1988).
53
-------
between the water and the air phases based on Henry's
Law constant (Kh).
I
If the Kh Is high, the chemical has a greater affinity to
partition into the air phase. The driving force, when con-
stantly depleting vapor concentration established through
advection, is similar to hydraulic head or electrical poten-
tial, a flow of mass or charge from an area of high to lo\jv
potential energy. The process motivates the chemical tp
move into the gas phase from a high concentration zone
(aqueous phase) into a low concentration zone (air
phase).
Vacuum extraction works because the system is deliber-
ately kept out of equilibrium. If there was no advection of
air, the chemical would partition from the water phase t<3
the air phase until it reached the Kh value, and then stop.
The air compartment would be filled, the water compart-
ment would still contain chemicals, and the system would
achieve equilibrium. By advecting air through the system,
however, and keeping one phase relatively clean, thp
chemical is continuously drawn from the water to the air
and stripped away in the air phase. Vacuum extraction
actually frustrates an equilibrium system by not letting the
system ever attain a Henry's Law constant value. j
A simple approach to performing a chemical mass bal-
ance in the laboratory is to put the contaminant from the
soil in a flask and collect the vapors in a sorbent tubte
(see Figure 6-2). The tube will contain resin, generally an
Influent
/""" Purge Gas
Effluent
Purge Gas
SdVWaste
Mbcture
Effluent Purge Gas
Figure G-2. Laboratory flask apparatus used for mass
balance measurements (Sims et al., 1989). !
i
organic polymer with high surface area and functional site
density, for the sorption of the chemical. By eluting the
chemical off the sorbent tube, the mass of material that
has moved into the air phase can be calculated. Ths
procedure is often useful with mixtures of chemicals, such
as oil and grease, in the subsurface. Measuring the ten-
dency of the chemical in the mixture to be in the air phase
can indicate the potential usefulness of applying vacuum
extraction under experimental conditions that may or may
not reflect site conditions. This technique is a first step in
remedy selection.
The following is an example of the misinterpretation of
soil vacuum extraction operational data. Figure 6-3 shows
Threaded Joint •
- Ground Elevation
Grout -
Bentonite Seal-
Slip Cap —x
Christy Box at Grade
4 in. Diameter
PVC Screen
Figure 6-3.
1991 a).
Bin. Diameter Borehole
Schematic of a gas extraction well (U.S. EPA,
a gas extraction well, which is designed to remove air
from the subsurface by suction or to measure the extent
and rate of cleanup at a site. If a remedial engineer were
to measure the concentration of the chemicals in the gas
phase coming out of the well, he or she might suppose
that the site was clean because the chemical concentra-
tion in the gas had decreased. Actually, however, the air
advecting through the high permeability sand and gravel
in the subsurface may have been cleaned by the air
relatively quickly. Because sand and gravel have larger
particles than clay or silt, their surface area to volume
ratio is much less; thus, they hold fewer contaminants.
The clay still might be highly contaminated, but because
little air goes through clay-textured materials, these con-
taminants do not show up in the air phase at the gas
extraction well. High volumes of clean air from the sand
and gravel mix with low volumes of very contaminated air-
from the clay, resulting in a total volume of air that is
relatively clean. If the remedial manager were to shut off
the well, later the air would be contaminated again. In the
interim, the chemical would have desorbed from the clay
into the air phase, and become available. The heteroge-
54
-------
neity in the subsurface and diffusion and desorption
cause chemicals to move from the more highly contami-
nated material back into the less contaminated sand and
gravel.
Two different kinds of wells are needed for vacuum ex-
traction: some for monitoring and some for extraction of
the gas phase. The extraction wells should be used to
extract the contaminated gas. Monitoring wells placed
around areas that are less permeable to air, such as silts
and clays, can be used to monitor the rate and extent of
cleanup. The position of the monitoring wells will depend
on whether the chemicals are associated with the
nonaqueous phase or are in the clays. When the moni-
toring wells indicate that cleanup has been achieved, the
remedial manager may still wish to take soil core samples
to ensure that the clay also has been cleaned up. For the
application of vacuum extraction of mixtures of
nonaqueous chemicals, for example, in gasoline or pe-
troleum (NAPL) phases, Raoult's Law rather than Henry's
Law is applicable. For a discussion of Raoult's Law refer
to Nonaqueous Fluids in Section 6.2.2.2.
6.2.2 System Variables
To explore system variables, this section will examine the
chemical and site characteristics that make vacuum ex-
traction more or less applicable.
6.2.2.1 Chemical Characteristics
To measure the tendency of a chemical to be in the air
phase, vapor pressure is used for the pure-phase chemi-
cal and the Henry's Law constant is used for compounds
dissolved in the water phase. The vapor pressure of a
pure chemical, as a rule of thumb, should be greater than
14 mm of mercury at 20° C, although lower values have
been suggested (U.S. EPA, 1991b). The Henry's Law
constant, as a rule of thumb, should be greater than 0.01
mg/L in air per mg/L in water for vacuum extraction to be
appropriate. In a spill of a pure chemical, such as ace-
tone, the chemical will move down through the unsatu-
rated zone. When acetone traveling through the soil
becomes dissolved in water, the Henry's Law constant is
used, because there is both a water phase and an air
phase.
Acetone, which is used in chemistry laboratories to dry
glassware, has a very high vapor pressure. It is, however,
completely miscible (will completely dissolve) in water. If
a lot of water is present in the soil, the acetone will
dissolve and the Henry's Law constant will be extremely
low, so vacuum extraction will not work.
Table 6-1 presents the vapor pressures of some com-
monly occurring compounds. The vapor pressure of per-
chloroethlyene (PCE) is 14.3 mm Hg. PCE, with its 4
chlorine atoms, is appropriate for vacuum extraction
based on its chemical characteristics. When PCE loses
a chlorine, it may be transformed from 4 to 3 chlorine
Table 6-1. Vapor Pressures of Some Commonly Detected
Compounds
Compound
Vapor Pressure
(mm Hg) at 20°C
Vinyl Chloride*
1,1-Dichlorethylene (1,1-DEC)*
Methyl Ethyl Ketone
Benzene*
Trichloroethylene (TCE)*
Bis(chloromethyl) ether
Toluene
Perchloroethylene (PCE)*
2,660
591
100
95.2
57.9
30
28.4
14.3
'Known or suspected carcinogen.
Source: U.S. EPA (1991a).
atoms and become trichloroethylene (TCE). The vapor
pressure for TCE is almost 60 mm Hg. If the chemical
transforms to two chlorines and becomes dichlo-
roethylene, the vapor pressure goes to 591 mm Hg. If
another chlorine atom is removed, the vapor pressure is
2,660 mm Hg. As this chemical is transformed in the
subsurface, biotically or abiotically, especially under an-
aerobic conditions, each transformed product has a
greater vapor pressure.
Vertical as well as horizontal movement of chemicals with
high vapor pressures may result in the spreading of con-
tamination within the unsaturated zone. Also, runoff or
precipitation events may result in leaching of compounds
that are also soluble in water.
Table 6-2 shows the Henry's Law constants for selected
compounds. The table presents some of the same chemi-
cals as in Table 6-1. In the table, the Henry's Law con-
stant for naphthalene is five times higher than the cutoff
value of 0.01 mg/L in air per mg/L in water. Vacuum
extraction is appropriate for all of the chemicals in this
table, based upon the criterion of 0.01 dimensionless
Henrys Law constant. Again, perchloroethylene has a
dimensionless Henry's Law constant of 0.34, but with one
less chlorine atom (TCE), it goes up to 0.42. Vinyl chloride
with one chlorine atom is 99 mg/L in air per mg/L in water;
it is basically a gas. Dechlorination of these chemicals
results in an increased tendency for a chemical to be in
the air phase.
6.2.2.2 Site Characteristics
The important site characteristics that affect the remedia-
tion process are soil moisture content; soil texture; distri-
bution of contaminants, or heterogeneity; and immiscible
fluids.
55
-------
Table 6-2. Henry's Law Constants for Selected
Compounds
Compound
Henry's Law Constant
(mg/L / mg/L)
Vinyl Chloride*
Mercury (Hg°)
Trichloroethylene (TCE)*
Perchloroethylene (PCE)*
PCB-1260
Benzene
Methylene Chloride
Naphthalene
99.0 ;
0.48
0.42
0.34 [
0.30 |
0.24
I
0.13 '
i
0.05 i
Source: U.S. EPA (1991a).
Soil Moisture Content
Figure 6-4 shows the effect of moisture on VOC adsorp-
tion and desorption in soil at two moisture contents. l|i
the figure, the two bars represent a dry soil and a wet
soil. The benzene rings represent the organic compound
that is being removed with vacuum extraction; the dots
represent water. In the dry soil, the benzene ring, or the
organic chemical, becomes adsorbed to the soil, and the
surface area is occupied by benzene. Under precipitation
or irrigation, the water competes for the surface area an'd
forces the benzene or the organic compound into the
water phase. The chemical will then behave based on the
Nonpolar Organic —v
Vapor ("""l
Phase x/
0
Adsorbed
Layer
Wet
Solid Surfacfe
Figure 6-4. Effect of moisture on VOC adsorption and
desorption In soil—VOC adsorption with two moisture re-
gimes (Reible, 1989). '
principle of the Henry's Law constant (i.e., a water phase
and an air phase).
Water also will vaporize from the subsurface because it
has a vapor pressure of 17 mm Hg at 20°. As water is
removed, the chemical becomes more adsorbed to the
soil surface. A monitoring well may show that the air has
become clean, but the chemical may be even more tightly
adsorbed to the soil. If water is then added to the soil,
the chemical may partition from the soil into the gas
phase and contaminate the gas again. This competition
between water and the organic chemical indicates that
water content of the soil is very important.
Benzene has a very high Kh, but when it is removed from
the soil and the soil becomes dry, Kh is no longer appro-
priate because there is no water phase. When water is
added, the Henry's Law constant is relevant again and
the chemical goes back into the air phase. By simply
adding water to the soil, the remedial engineer can control
the efficiency of vacuum extraction. In using vacuum ex-
traction systems in arid regions, it is important not to dry
the soil out, as the efficiency of vacuum extraction will
actually decrease due to water removal.
2.1% Water
3.94% Water
17% Water
i
40 60 80
Dieldrin in Soil/ppm
100
Figure 6-5. Effect of soil water content on dieldrin vapor
pressure (modified from Spenser and Cliath, 1969).
Figure 6-5 shows the relative vapor density of dieldrin in
soils with three different water contents. At a vapor den-
sity of 1, dieldrin vaporizes into the air. As shown in the
figure, when the soil has 2 percent water by weight (2
grams of water per 100 grams soil dry weight), the dieldrin
is adsorbed to the soil. If water is added to 4 percent, or
4 grams of water per 100 grams of soil dry weight, the
dieldrin changes from the soil phase to the air phase. This
4 percent point is known as the Critical Soil Moisture
Content for Vacuum Extraction. The change from 2 to 4
percent provides just enough water to push the dieldrin
56
-------
off the surface of the soil into the water, where it can
operate by the Henry's Law constant. Increasing the
water content to 17 percent, however, does not effect the
same dramatic increase in the dieldrin in the gas phase.
The range of 2 to 5 percent water content may be a good
low optimum for use with vacuum extraction.
Texture
Clay-textured soil areas, because they are less perme-
able to any fluid—NAPL, water, or air—represent more
of a challenge to the use of vacuum extraction. Chemical
characteristics may indicate that vacuum extraction would
be viable, but if there is appreciable clay at the site it may
be necessary to first increase the permeability by fractur-
ing. If the chemical permeates the soil in fractures, it is
possible to remove it from those fractures, but if the
chemical has permeated by diffusion, it will be more dif-
ficult to remove. Because of the low permeability of clay,
vacuum extraction will be much more successful at a site
with a sandy loam texture than at a site with clay and
clay loam textures.
In the soil texture triangle in Figure 6-6, vacuum extrac-
tion is more applicable for the soils near the bottom left
of the triangle and less applicable for those such as clays
at the top.
Heterogeneity
At a site with different soil textures, heterogeneity can be
used to assess how much of the chemical is in the clay
and how much is in the sand. This information can help
Percent by
Weight Clay
Percent by
Weight Silt
100 90 80 70 60 50 40 30 20 10
Percent by Weight Sand
100
Figure 6-6. Soil texture trilinear diagram (Soil Conserva-
tion Service, 1971).
determine where to install monitoring wells to evaluate
the rate and extent of cleanup, and how relatively fast or
slow to expect the rate of cleanup.
In Figure 6-7, the area labeled molecular diffusion is the
low ..permeability strata, where diffusion dominates the
/
<
\
A A A A A A
Molecular Diffusion
A A A A A A
Low Permeability
Strata
\
/ A A A A A A A \ ^W
XT T T T T T T /
Air Advection
A A A A A A\
Molecular Diffusion /
T T V T T T/
Figure 6-7. Diffusive release of contaminants from the
soil phase into the gas phase (U.S. EPA, 1989).
movement of chemicals within clay. The arrows show air
advection through a more permeable area consisting pri-
marily of silts and sands. Advection will proceed easily
through the silts and sands, cleaning that area very
quickly. The chemical that is in the clay, however, can
take 50 years or more to come out by diffusion, unless
fracturing is used. The chemical diffuses very slowly back
into the area where the sand and silt are being cleaned
up. A gradient for molecular diffusion shows the chemical
starting at a very high concentration, then steadily declin-
ing. But if treatment ceases, after several weeks the con-
centration rises again because of the slow diffusion or
desorption in tighter materials, primarily clays and silts.
Monitoring for Heterogeneity. It is important to place
monitoring wells so as to obtain an accurate assessment
of the cleanup. If the well is placed through the whole
system, the clean air will dilute the contaminated air, al-
lowing potential misinterpretation of monitoring results.
Instead site personnel should locate the clay stringers
and install monitoring wells in those areas, sampling the
air right around the clay and not diluting the air around
the silts and sands. This is called monitoring a rate-lim-
iting flow path.
Figure 6-8 shows the vapor concentration of a chemical,
such as benzene, decreasing as the pump begins to
advect air through the subsurface. The site manager turns
off the pump when a level is obtained that is acceptable,
57
-------
I
Figure 6-8. Concentration vs. time profile showing restart
spike (DIGiulio et at., 1990).
then waits for several weeks before turning the pump
back on. The initial sample shows a higher concentration
of benzene in the advected air than the concentration
when the pump was stopped. This observation is called
a restart spike or yield spike, and is due to the sources
in the subsurface that continue to sponsor chemicals in
the air phase due to diffusion. This phenomenon is com-
mon in pump-and-treat systems and can be observed in
vacuum extraction systems as well. If the pump is turned
on again, the concentration will decrease further, because
there is only a finite mass of material to remove, assum-
ing all the drums, rock piles, etc. have been removed.
The slope of the curve in Figure 6-8 presents a rough
estimate of the rate of cleanup.
Nonaqueous Fluids
The following is a simplified discussion of the application
of Raoult's Law to nonaqueous fluids remediation.
When a site has a combination of liquids such as gasoline
or petroleum in the subsurface, then Henry's Law is no
longer applicable because there is no water phase. Tjie
chemical, however, will tend to volatilize out of the pure
liquid in proportion to its concentration in that liquid. !
Raoult's Law is expressed as the following: a chemical
will volatilize in proportion to its concentration and jits
vapor pressure. Therefore, vacuum extraction can never
remove 100 percent of the hydrocarbons present. It will
only remove the percent that can be volatilized, maybe
40 percent. Fifty or 60 percent of the hydrocarbons re-
main in the subsurface, because they are not volatile!
Figure 6-9 shows the volatilization of a combination, of
liquids. The fewer the number of carbons, the lighter the
molecule and the higher the vapor pressure, and the more
of the chemical that will be removed with vacuum extrac-
tion. In the figure, the lighter components of the mixture
are labeled C4 and C5. When the mass of this material
Is removed via vapor extraction, the residual mass left
over will contain the heavier constituents, C6, C7, and
T—-i 1 1 r
40 60
Percent Volatilized
100
Figure 6-9. Volatilization of a combination of liquids
(Johnson, 1989).
C8, that do not volatilize as readily. The result is an
increase in the concentration of the high molecular weight
compounds.
6.3 APPLICATIONS OF VAPOR EXTRACTION
TO REMOVE LIGHT NAPLS
Vacuum extraction will remove the lighter components of
a contaminant in water systems based upon Henry's Law,
and will remove the lighter components of a contaminant
in nonwater systems based upon Raoult's Law.
The design of extraction wells in the field needs to ad-
dress spacing as well as depth. The wells should be
placed apart according to media properties, and should
be as deep as the contaminant. Where contamination is
deep and permeability is high throughout the profile, the
slotted (screened) interval should be extended to the
maximum depth possible.
Figure 6-10 shows the effect of well spacing on total
solute mass removal from the subsurface over a period
of about 12 days. Wells should be placed more closely
to achieve overlapping cones of influence, as in pump-
T 1 1 T
0 1.0E + 06sec
Figure 6-10. Total solute mass in the subsurface vs. time
showing the effects of well spacing on vacuum extraction
effectiveness (modified from Wilson et al., 1989).
58
-------
arvd-treat systems. If the wells are spaced 20 meters
apart, the site is stabilized more rapidly than if the wells
are spaced 30 meters apart. Then more of the vapor
phase is able to be moved more quickly to clean up the
site. The spacing is also a function of soil texture. In clay
environments, the wells need to be closer together; for
gravel, they can be farther apart.
The remedial manager must be concerned with the pos-
sibility of explosion of collected vapors, which is normally
alleviated with an open air valve on the collection line.
30m
I \ I T
1.0E + 06 sec
Figure 6-11. Total solute mass in the subsurface vs. time
showing the effects of well depth on vacuum extraction ef-
fectiveness (Wilson et al., 1989).
Figure 6-11 shows the effects of well depth on the total
solute mass with the same axes as in Figure 6-10. The
well depth can be considered the distance between the
ground water and the end of the well in the vadose zone.
In these three examples, the vacuum extraction well was
screened 3, 6, and 9 meters above the ground water. The
closer the well screening is to the ground water, the more
rapid the cleanup of the site. This is because it is easier
to pull air downward through the vadose zone to a well
screened near the ground water than to pull air upward
from a location near the ground water to a well screened
near the ground surface. Therefore, when the vadose
zone is contaminated from the ground surface to ground
water, it is more effective to screen the vacuum extraction
system close to the ground water.
6.4 BIOVENTING
Vacuum extraction can enhance biodegradation of vola-
tile and semi-volatile chemicals in the soil by providing
oxygen to the soil for use by microorganisms. Larger
amounts of oxygen can be supplied per volume of air
than per volume of water. This use of vacuum extraction
to enhance biodegration is also known as bioventing.
Hill Air Force Base in Ogden, Utah, had a gasoline spill
that penetrated the soil to a depth of more than 100 feet.
The remedial team decided to apply vacuum extraction,
or bioventing, to remove the gasoline, and then incinerate
the gasoline once at the surface. At this site, 50 to 60
percent of the cost of vacuum extraction is attributed to
aboveground treatment of the vapors. The bioventing sys-
tem lowered the concentration of jet-fuel in the subsur-
face, but the remedial team had to add fuel from an
outside source to burn the residual fuel they were extract-
ing from the subsurface. The remedial team then began
to look at the possibility of degrading the residual vapor
phase in the subsurface with microorganisms that would
biodegrade fuel constituents including benzene, xylene,
and toluene. In this case, they decided to use bioventing
to control the residence time of the gas phase in the
subsurface. Using bioventing to enhance biological proc-
esses significantly reduced the costs of site remediation
using aboveground treatment.
When trying to accomplish subsurface in situ remediation
of the gas phase, the remedial team found that adding
water to the subsurface stimulated COa production and
mineralization of the gasoline components. Moisture is
important for vacuum extraction, whether it is used as a
physical-chemical process or a biological process. Figure
6-12 shows the enhancement of bioremediation of gaso-
line components using bioventing of soil amended with
110-i
100-
80-
70-
60-
50-
4 40-
30-
8
ro
I
-o
-------
/-+
Hydrocarbon /
H- -\
Oxygen \
^->
Nutrients
CO2+
Biomass
(Respiration)
Figure 6-13. Aerobic biodegradation (Hinchee, 1989).
in a contaminated subsurface, the subsurface is pfteijt
anaerobic. The addition of oxygen and nutrients can
stimulate indigenous biomass to degrade hydrocarbons
to COa and water.
Table 6-3. Oxygen Supply
Water
Air Saturated
Pure Qz Saturated
500 mg/L HaOa
Air
Ib carrier/lb Oz
100,000 ;
25,000
10,000 ,
4
Table 6-3 illustrates the amount of water or air requireid
to supply the oxygen necessary to biodegrade hydrocar-
bons. About 100,000 ib of water, saturated with air, are
required to supply 1 Ib of oxygen. Also about 4 Ib of air
are required to biodegrade 1 Ib of benzene. Therefore, to
degrade 1 Ib of benzene, 400,000 Ib of water are requirejd
to supply the needed oxygen. Since it requires only 4 Ib
of air to degrade 1 Ib of benzene in the subsurface, air
is a more efficient way to supply the terminal electron
acceptor .(oxygen molecule) without increasing leaching
potential.
Bioventing systems are composed of hardware identical
to that of conventional soil vacuum extraction (SVE) sys-
tems, with vertical wells and/or lateral trenches, piping
networks, and a blower or vacuum pump for gas extrac-
tion. They differ significantly from conventional systems,
however, in their configuration and philosophy of design
and operation. As indicated above, the primary purpose
of a bioventing system is to employ moving air to transfer
oxygen to the subsurface where indigenous organisms
can utilize it as an electron acceptor to carry out aerobic
metabolism of soil contaminants. As such, bioventing sys-
tem extraction wells are not placed in the center of the
contamination as in conventional SVE systems, but on
the periphery of the site (Figure 6-14), where low flow
rates maximize the residence time of vent gas in the soil
to enhance in situ biodegradation and minimize contami-
nant volatilization.
Because it is a biological treatment approach, however,
bioventing does require the management of environ-
mental conditions to ensure maintenance of bioactivity at
the site. Management of soil moisture and soil nutrient
levels to avoid inhibition of microbial respiration within the
vadose zone can be accomplished fairly easily, and have
been used to optimize contaminant biodegradation at field
sites when other variables, i.e., toxicity, do not limit mi-
crobial activity. '
Cyclic or "surge" pumping of vent systems may minimize
system operating costs. Surge pumping in a bioventing
By-Pass
Off-Gas
Treatment
Injection
Well
Extraction
Well
Knock-Out
Drum
Figure 6-14. Schematic of recommended bioventing system layout,
60
-------
mode entails operating the blower system until soil gas
oxygen levels reach near ambient conditions throughout
the site being remediated. The system then would be shut
off for some period of time during which soil-gas oxygen
concentrations would be routinely monitored until they
reach a level that inhibits aerobic microbial activity. Once
this limiting soil-gas concentration is reached, the vent
system would be restarted, and the on-off cycle would
continue once again. Based on a Henry's Law constant
for oxygen, this oxygen limitation would be expected to
occur at a soil-gas concentration of approximately 2.0
volume percentage, corresponding to soil water oxygen
concentrations of approximately 1 mg/L An inhibition of
soil respiration has been reported at the 2.0 volume per-
centage soil oxygen level in venting systems treating jet
fuel contaminated soils and in vented soil piles contami-
nated with PCP waste (McGinnis, 1992), suggesting that
this value represents a good operating number for field-
scale applications.
The two major design considerations for bioventing sys-
tems are 1) whether the contaminants of concern are
biodegradable under prevailing site conditions, i.e.,
whether inhibition or toxicity is evident at the site, and 2)
whether the required terminal electron acceptor, i.e., oxy-
gen, can be effectively transported within the soil to en-
courage aerobic contaminant biodegradation. The first
question can be answered using soil-gas composition
and in situ respiration measurements, while the second
question is answered from in situ air permeability
measurements.
Bioventing system design today can be carried out by
estimating the equivalent daily oxygen demand and vent
airflow rate as determined from in situ respiration meas-
urements, and by estimating pump operating conditions
at these required air flow rates based on field-determined
in situ air permeability measurements. The feasibility of
surge pumping and vacuum pump/blower scheduling can
be assessed based on required versus maximum oxygen
transfer rates possible under a given set of field and
pump/blower operational constraints.
6.4.1 Application of Bioventing
Figure 6-15 is a diagram of the Hill Air Force Base site
showing the 62 concentration in the air phase with depth
near a vacuum extraction well. Wells were installed to
depths of 50 to 70 feet. The chemicals at Hill Air Force
Base, a jet fuel spill, have a high vapor pressure, and
thus tend toward being in the air phase. Site charac-'
teristics show that the soil texture is a sandy material. As
discussed previously, vacuum extraction was used as a
method of remediation. Vacuum extraction pulled out 80
percent of the benzene, xylene, and toluene. When the
chemicals reached such low concentrations that the in-
cinerator had to be supplemented with an external supply
of fuel, the remedial team shut off the system and began
Distance (feet)
Figure 6-15. Oxygen concentration in vadose zone be-
fore venting (U.S. EPA, 1989).
considering the use of vacuum extraction to supply oxy-
gen to the subsurface.
Before venting was initiated, the oxygen concentration
was 10 percent at 10 feet below the surface, 5 percent
at 20 feet below the surface, and 1 percent at 30 feet
(Figure 6-15). After initiating bioventing, the oxygen went
up to 15 percent at 10 feet and 10 percent at 20 feet
(Figure 6-16). Even at depths of 60 and 70 feet there was
Distance (feet) ...
Figure 6-16. Oxygen concentration in vadose zone after
venting (U.S. EPA, 1989).
10 percent oxygen in the soil. This process of adding
oxygen using vacuum extraction combined with increas-
ing the soil moisture content to 75 percent field capacity
enhanced biological remediation. Bioventing was ob-
served to reduce the volatile organic contaminants in the
vapors off the surface to less than 20 ppm. (The State of
Utah cleanup level was 50 ppm.) So, in this example,
vacuum extraction was part of a treatment train, where it
61
-------
was used first as a physical-chemical process and then
to augment biological processes. In the first case, the
process extracted the chemical and destroyed it abo\)e
ground, and in the second case the chemical was de-
stroyed in situ through biodegradation.
There is very little temperature change below 20 or 30
feet because the soil acts as a temperature buffer. In hot
climates, however, there may be changes in the tempera-
ture in the subsurface at 6 or 8 feet, which will augment
volatilization. Sometimes injecting steam will increase the
temperature for several feet, but, because of the buffering
capacity of the soil, significant and continuous energy
input to the subsurface is required to sustain this effect.
6.4.2 Monitoring
Hill and Tyndall Air Force Bases were monitored by
measuring oxygen concentration with an oxygen probe
and carbon dioxide using a COa probe. Figure 6--) 7
x Oxygen, k = -.00059/mm
o Carbon Dioxide
1000 2000 3000
Time (minutes)
4000 4500
Figure 6-17. The increase in COg and decrease in oxygen
as a result of degradation of the jet fuel components (U.S.
EPA, 1989).
shows the increase in CO2 and decrease in oxygen as a
result of degradation of the jet fuel components. The
vacuum extraction well was shut off, and the parameters
were measured. Oxygen depletion indicates biodegrada-
tion, and an increase in COa production indicates miner-
alization. Ultimately, when the hydrocarbon is removed
by biodegradation, these lines (Figure 6-17) will flatten
out because oxygen is no longer being produced. When
this flattening occurs (meaning that biological activity has
slowed down), it is an appropriate time to sample a soil
core and determine if the hydrocarbons in the core have
been removed. If hydrocarbons are still present, even
after biological activity has slowed or ceased, the process
probably needs more oxygen, moisture, and nutrients.
6.5 REFERENCES |
DiGiulio, D.C., J.S. Cho, R.R. Dupont, and M.W. Kemblowski. 1990.
Conducting field tests for evaluation of soil vacuum extraction
application. In: Proceedings of the 4th nat. outdoor action
conf. on aquifer restoration, ground-water monitoring, and
• geophysical methods, National Water Well Association, Dub-
lin, OH. pp. 587-601.
Dupont, R.R., W.J. Doucette, and R.E. Hinchee. 1991. Assessment
of in situ bioremediation potential and the application of
bioventing at a fuels-contaminated site. In: In situ biorecla-
mation: applications and investigations for hydrocarbon and
contaminated site remediation. Boston, MA: Butterworth-
Heinemann.
Hinchee, R. 1989. Soil vacuum extraction laboratory and physical
methods studies. Presented at workshop on soil vacuum ex-
traction held at U.S. EPA Robert S. Kerr Environmental Re-
search Laboratory, Ada, OK. April 27-28 (Dominic DiGiulio,
Technical Coordinator).
Hinchee, R.E., S.K. Ong, and R. Hoeppel. 1991. A field trealability
test for bioventing. Proceedings of the 84th annual meeting
and exhibition of the air and waste management association,
Vancouver, BC, Canada. Air,and Waste Management Asso-
ciation. Preprint 91.19.4.
Johnson, P.C., M.W. Kemblowski, and J.D. Colthart. 1990. Quanti-
tative analysis for the cleanup of hydrocarbon-contaminated
soils by in-situ soil venting. Ground Water. 28(3): May-June.
Johnson, R.L. 1989. Soil vacuum extraction laboratory and physical
methods studies. Presented at workshop on soil vacuum ex-
traction held at U.S. EPA Robert S. Kerr Environmental Re-
search Laboratory, Ada, OK. April 27-28 (Dominic DiGiulio,
Technical Coordinator).
McGinnis, D., R.R. Dupont, and K. Everhart. 1992. Determination
of respiration rates in soil piles to evaluate aeration efficiency
and biological activity. Proceedings of the 85th annual
meeting and exhibition of the air and waste management
association, Kansas City, MO. Air and Waste Management
Association. Preprint 92-13.05.
Reible, D.D. 1989. Introduction to physicochemical processes influ-
encing enhanced volatilization. Presented at workshop on soil
vacuum extraction held at the U.S. EPA Robert S. Kerr En-
vironmental Research Laboratory, Ada, OK. April 27-28
(Dominic DiGiulio, Technical Coordinator).
Sims, J.L., R.C. Sims, and J.E. Matthews. 1989. Bioremediation of
contaminated surface soils. Robert S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency,
Ada, OK. EPA-600/9-89/073.
Soil Conservation Service. 1971. Handbook of soil survey investi-
gations procedures. Soil Conservation Service, U.S. Depart-
ment of Agriculture, Washington, DC.
Spenser, W.F. and M.M. Cliath. 1969. Vapor density of dieldrin.
Environ. Sci. Technol. 3:670-674.
U.S. EPA. 1991a. U.S. Environmental Protection Agency. Site char-
acterization for subsurface remediation. Chapter 15: reme-
diation techniques for contaminated soils. Office of Research
and Development, Washington, DC. November. EPA/625/4-
91/026.
U.S. EPA. 1991 b. U.S. Environmental Protection Agency. Soil vapor
extraction technology: reference handbook. Cincinnati, OH:
Risk Reduction Engineering Laboratory.
U.S. EPA. 1989. U.S. Environmental Protection Agency. Site char-
acterization for subsurface remediation: speaker slide copy.
Roberts. Kerr Environmental Research Laboratory, Ada, OK,
and the Center for Environmental Research Information, Cin-
cinnati, OH, CERI-89-224, September.
Valsaraj, K.T. and L.J. Thibodeaux. 1988. Equilibrium adsorption of
chemical vapors on surfape soils, landfills, and landfarms —
a review. J. Hazardous Materials. 19:79-99.
Wilson, D.J., R.O. Mutch, Jr., and A.N. Clarke. 1989. S^l vacuum
extraction laboratory and physical methods studies. Pre-
62
-------
senfed af workshop on soil vacuum extraction held at
U.S. EPA Robert S. Kerr Environmental Research Labora-
tory, Ada, OK, April 27-28 (Dominic DiGiulio, Technical
Coordinator).
Wilson, J.T. and C.H. Ward. 1988. Opportunities for bioremediation
of aquifers contaminated with petroleum hydrocarbons. J.
Ind. Micro. 27: 109-116.
63
-------
-------
APPENDIX A
Case Studies
This appendix contains brief summaries of selected case
studies presented at the U.S. EPA's RCRA Corrective
Action Conference, held February 25-27,1992. The case
studies were presented by EPA staff, consulting engi-
neers, and private industrial representatives. They exam-
ined RCRA stabilization efforts that illustrate experience
with key stabilization techniques, technologies, and con-
cepts. Three of the case studies represented were se-
lected for inclusion in this document to elaborate on the
application of stabilization strategies at RCRA Corrective
Action facilities. Unfortunately, it was not possible to in-
clude all of the presentations in this document.
Each case study in Appendix A contains three parts:
• Introduction. The statement of problems; descriptions
of waste management units requiring stabilization; site
descriptions including history, topography, geology, and
hydrology; pre-RCRA activities; and/or contaminant re-
leases including those across media and site bounda-
ries.
• Stabilization Strategies. The stabilization action ra-
tionale, the spectrum of strategies considered, and/or
the selected strategies with their constraints.
• Implementation and Future Actions. Stabilization ac-
tion design criteria, construction, timing, and/or system
operations.
A.1 GROUND-WATER CONTAMINATION
STABILIZATION WITH LNAPL COLLECTION
SYSTEM: JAMESTOWN, NY, by Frank Gardner,
U.S. EPA, Region IX, San Francisco, CA
A.1.1 Introduction
This case study covers the selection, design, construc-
tion, and operation of a light non-aqueous phase liquid
(LNAPL) collection system for the stabilization of contami-
nated ground water at a former industrial site in
Jamestown, New York. This project was conducted from
July 1989 to August 1990. The LNAPL collection system
consists of a subsurface trench with a network of perfo-
rated and corrugated metal pipes installed at the oil-water
interface. LNAPL is periodically or continuously vacu-
umed, pumped, or skimmed from system manholes.
The case study illustrates the use of an interim measure
to stabilize subsurface contamination by removing LNAPL
(the contamination source) from the ground-water con-
tamination plume. The system selected was designed
for integration into the final ground-water contamination
remedy.
The formerly industrial site of this case study is presently
municipally owned. The city discovered subsurface con-
tamination during the initial stages of constructing a park-
ing lot. Upon discovery of the contamination, the city
stopped construction, contacted regulatory agencies, and
conducted a preliminary hydrogeologic investigation. The
investigation confirmed contaminated soils, and ground
water, and #6 fuel oil (the LNAPL) floating on the water
table. The #6 fuel oil was thought to be spilled product,
as opposed to spilled waste, due to the lack of heavy
metals or other contaminants.
The LNAPL is contained by a remnant building founda-
tion. While LNAPL migration has stopped, it continues to
contaminate ground water below, which flows under the
foundation. The several remedies considered include ex-
traction wells with submersible or dual phase pumps,
open trenches, and subsurface trenches. Subsurface
trenches were selected based on:
• Site hydrology
• Nature of the contamination
• Limited financial resources
• Compatibility with a long-term remedy
• Compatibility with the planned use of the site
A. 1.1.1 Site Background
The case study site is in Jamestown, New York, about
80 miles southwest of Buffalo. The site is located in an
industrial/commercial sector of the city about 250 feet
from the north bank of the Chadakoin River, which flows
from Chautauqua Lake. The site is located at the base
of a hill from which the land slopes down towards the
eastwardly flowing river. The Chadakoin River is used
only for recreation, not for drinking water. Ground water
on the site flows towards the river, and, while not pres-
ently used, is protected as a potential drinking water
source.
65
-------
The history of the 1.9-acre site includes the following
major elements: |
• Prior to 1902, the site was probably undeveloped or
agricultural. '
• From 1902 to 1984, the site contained a tool factory
that manufactured adjustable wrenches. !
• After 1984, the site was used for a metal fabricating
operation and then donated to the city. The city prp-
posed to construct a municipal parking lot after demol-
ishing the buildings. '
i
• In 1989, during the construction of a storm sewer fpr
the proposed parking lot, the city found subsurface
petroleum contamination. ;.
Upon discovery of contamination, appropriate regulatory
agencies, including the New York Department of Environ-
mental Conservation (NYDEC), were contacted and an
investigation was initiated.
A. 1.1.2 Site Hydrology i
During the hydrogeologic investigation, five soil bpr-
ing/ground-water monitoring wells (MW-1 through MW^5)
were installed to characterize site hydrology and assejss
the extent and nature of the subsurface contamination.
The depth of the borings ranged from 14 to 16 feet, ahd
ground water was encountered 3 to 10 feet below the soil
surface. The shallow soils beneath the site consisted of
surflcial silt and sand fill overlaying a native gray medium
clay formation with traces of silt and sand. In two of the
borings, peat was found between undisturbed clay and
fill materials. Since peat commonly occurs in the regipn,
it may be found throughout the site.
Ground water flows primarily to the south toward jhe
Chadakoin River; however, this overall flow pattern is
affected by remnant foundations from the tool plant in jthe
site's subsurface. The ground water flows under the foun-
dations. The overall difference in ground-water elevation
across the site is about 9.4 feet, which corresponds to a
gradient of 0.031 ft/ft. Due to the site's proximity to jthe
river, however, it is possible that the direction of ground-
water flow is reversed during flood periods. This phe-
nomenon has been observed in similar sites nearby, [but
has not been confirmed at this site. :
A. 1.1.3 Nature and Extent of Contamination \
During the installation of the soil borings, soils heavily
contaminated with total petroleum hydrocarbons (TPH)
identified as #6 fuel oil were found the entire depth of
MW-2. MW-2 was located within the former tool plant
foundation and near the location where contamination
was first discovered. The other four borings, all outside
the underground foundation, did not detect contaminated
soils. ,
The borings were developed into monitoring wells, after
which ground-water samples were taken to determine the
extent of contamination. TPH ranging from 190 to 920
micrograms per liter (ppb) was detected in all five wells
and was identified as #6 fuel oil. In addition, LNAPL con-
sisting of #6 fuel oil was discovered in MW-2. The quantity
of LNAPL is estimated at 5,000 to 10,000 gallons. The
source of the LNAPL is believed to have been under-
ground fuel oil storage tanks or floor drains at the former
tool plant.
Prior to December 1989, volatile organic analysis of MW-
2 detected ethylbenzene and xylenes in the ground water.
In December 1989, five additional monitoring wells (MW-6
through MW-10) were installed to better define the lateral
extent of the ground-water contamination within the foun-
dation area. In January 1990, all 10 wells were sampled
for VOCs, confirming the presence of ethylbenzene and
xylenes in MW-2, MW-7, and MW-8, all within the foun-
dation area. None of the other well samples contained
VOCs. The concentrations detected are listed in Table
A-1.
Table A-1. VOC Concentrations in the Foundation Area
VOC
Ethylbenzene
Xylenes
Concentration
300|og/L
170|ig/L
MCL*
700 pg/L
10,000 pg/L
SMCL**
30pg/L
* EPA Maximum Contaminant Level
** EPA Secondary Maximum Contaminant Level
As the table indicates, site concentrations of ethylben-
zene and xylenes were below the Maximum Contaminant
Levels (MCLs) but above the Secondary Maximum Con-
taminant Levels (SMCLs). Other analyses, such as
the EP Toxicity test for metals and PCBs, did not reveal
any additional contaminants in the soil or ground-water
samples.
A.1.2 Stabilization Strategies
A. 1.2.1 Remedial Options
While no operational drinking water supplies are near the
site, the NYDEC requires protection of potential ground-
water uses. The NYDEC required the city to install an
interim remedial system to remove the LNAPL from the
water table, predicated on the city's willingness to imple-
ment more complete ground-water remediation if required
in the future. The corrective action was required to contain
and remove the LNAPL. The four interim remedial options
included:
• Extraction wells with submersible pumps
• Extraction wells with dual phase pumps
• Open trenches
• Subsurface trench systems
66
-------
Extraction wells with submersible pumps remove the
LNAPL and the ground water together by creating a gra-
dient towards the well for both liquids. An oil/water sepa-
rator separates the two before disposal. Extraction wells
with dual phase pumps are able to pump LNAPL and
ground water separately and simultaneously. Dual phase
pumps eliminate the need for oil/water separators and the
risk of emulsifying oil In water, but they are more expen-
sive to buy and maintain than are submersible pumps.
Open and subsurface trenches recover oil by providing a
conduit of negligible hydraulic conductivity enabling vacu-
uming, skimming, or pumping of the oil. Both types of
trenches use similar excavations. Subsurface trenches,
however, are backfilled with high permeability materials,
such as gravel or perforated drainage pipes, so they can
be covered with soil. Trenches are technologically sim-
pler, less costly to operate and maintain, and more effec-
tive on soils with low hydraulic conductivity than are
extraction wells. Disposal costs for contaminated exca-
vated soils, however, can dramatically increase total costs
for both options.
A.1.2.2 Interim Measure Selection
A number of physical, financial, and logistical constraints
were considered in selecting the appropriate remedial
option for stabilizing this site, including:
• Site hydrogeology
• Nature of the contamination
• Available financial resources
• Compatibility with a long-term remedy
• Compatibility with the site's long-term use
Based on an analysis of these factors, the subsurface
trench was selected because it was the oniy option to
satisfy all five constraint categories. These factors are
summarized in Table A-2 and the paragraphs below.
Site Hydrogeology
All four options are compatible with the site hydrology.
The native geology of the site consists primarily of a
low-permeability clay formation. The clay overlays fill in
some areas of the site which were probably former build-
ing sites. Trench systems, with larger wetted perimeters,
are more effective than extraction well systems for low-
permeability soils. Properly designed extraction wells,
however, also can be effective for these soils.
Nature of the Contamination
The recovery target LNAPL is #6 fuel oil, which is fairly
viscous and much less mobile than water in the soil me-
dia. The best suited remedy would address a large hori-
zontal collection area to compensate for the relatively
immobile LNAPL. This factor favored the trench options
because trench systems are inherently suited to cover a
Table A-2. Impact of Constraints on Remedial
Alternatives
REMEDIAL ALTERNATIVE
Constraint
Extraction Extraction
Well/Sub- Well/Dual
mersible Phase
Pump Pump
Subsur-
Open face
Trench Trench
Hydrogeology X
Contamination
Cost
Long-Term X
Compatibility
Planned Site X
Use Com-
patibility
X X
X
X
X X
X
X
X
X
X
X
large horizontal area. Extraction wells, in contrast, rely on
drawing contaminants toward the well and would not be
as effective for this LNAPL.
Available Financial Resources
The city is relatively small with limited financial resources.
Clearly, the trench system is less expensive. Estimated
costs for the trench system are from $45,000 to $60,000.
The trench options also allowed for removal of the LNAPL
without the water, which reduced waste volume and as-
sociated disposal costs. These low estimates are partially
due to the presence of tight clay soils, which make sheet-
ing and shoring (a significant component of trench system
costs) largely unnecessary. Also, the soil turned out to be
nonhazardous. If it had been hazardous, the trenches,
which require more voluminous excavations than do
wells, would have incurred much higher costs.
Estimated costs for the extraction well system are from
$98,000 to $160,000. These estimates are based on us-
ing the local publicly owned treatment works (POTW) for
ground-water disposal. If the POTW were not available,
then costs would be much higher.
Compatibility with a Long-Term Remedy
The state indicated that the final corrective action remedy
will include some comprehensive ground-water treat-
ment. Both extraction wells and trench options are com-
patible with long-term remedies, since both are able to
accommodate pumping large volumes of ground water.
Extraction wells are easily upgraded into long-term reme-
dies with the addition of more extraction wells. Similarly,
. well-constructed and -located trenches of sufficient depth
are well suited for extracting large quantities of ground
water.
67
-------
Compatibility with the Site's Long-Term Use
A parking lot is the planned use for the site. The extraction
well and closed trench options are both compatible with
this use, while an open trench cannot coexist with a park-
ing lot. '
A.1.3 Implementation and Future Actions
A.1.3.1 Design and Construction
The trench system consists of a network of perforated
and corrugated pipe (CMP), underlain and backfilled with
crushed stone, with two access manholes. The most im-
portant design parameter to effectively remove LNAPL
from ground water is the burial depth of the CMP. If; the
centerline for the CMP is not at the oil/water interface,
then the trench system could be ineffectual. Other impor-
tant design parameters include pipe layout pattern, pipe
diameter, perforation size, and backfill type. '
The elevation of the oil/water interface at three monitoring
wells varied temporally and spatially between 1,289 and
1,295 feet above sea level. It was assumed that the water
table would equilibrate to a single elevation within!the
pipe. Therefore an intermediate elevation of 1,291.5 feet
above sea level was selected. j
i
The main trunk of the CMP network was configured to
intercept the LNAPL plume along the entire downgradient
(southern) edge. The remnant foundation was left to
serve as a downgradient barrier. Two lengths of CMP
were also installed as branches, perpendicular to'the
trunk, to facilitate LNAPL collection from the central area
of the plume. Manholes, consisting of vertically placed
sections of 6-foot diameter CMP, were installed at the two
intersections for access and LNAPL removal. The man-
hole inverts were set 2.5 feet lower than the CMP invert
to accommodate water table drawdown from possible fu-
ture remediation involving ground-water extraction. Inj ad-
dition, the manholes have locking covers and concrete
aprons at ground level to accommodate parking lot, ve-
hicular traffic loads. '
I
The selected CMP diameter of 3 feet provides a large
surface area and encompasses the range of temporal
water table variation while keeping the size of the exca-
vation within reasonable width. The chosen perforation
size of 0.25 inches, coupled with crushed stone backfill
between 0.5 and 1.0 inches, prevents the backfill from
entering the CMP network. The CMP is underlain with 1
foot of crushed stone and backfilled with 6 feet of the
material. i
The high hydraulic conductivity of the crushed stone [pro-
vides a margin of safety for the system. That is, if the
oil/water interface rises or falls below the CMP due to
natural changes or ground-water extraction, then the sys-
tem will still function. ,
The LNAPL collection system was constructed quickly
during a 3-week period in August 1990. The project was
uneventful except for field modifications in system layout
due to the unexpected discovery of additional subsurface
concrete structures. The additional subsurface structures,
which included an interior wall foundation and a very large
machine pedestal, caused the pipe layout to contain more
zigzags than planned. The interior foundation was
breached to allow the CMP network to pass through and
collect the LNAPL in that area, and the layout was modi-
fied to go around the machine pedestal. Construction
costs, including the field modifications, totaled about
$50,000. c
A large quantity of contaminated soil was excavated dur-
ing construction. Based on the TCLP analysis, the soil
was not hazardous waste and was disposed of in a local
solid waste landfill for $27,000. If the soil had been clas-
sified as RCRA-hazardous waste, then the project would
have incurred an additional $50,000 in disposal costs.
Construction costs totaled $77,000. The city constructed
a parking lot immediately after completion of the LNAPL
collection system. Parking lot paving ensured access to
the two manholes and monitoring wells.
A.1.3.2 System Operation
The system is designed to recover LNAPL either on an
occasional basis by a vacuum truck or on a continuous
basis using skimming technology. The city chose the vac-
uum truck based on the negligible startup costs. The truck
vacuums up a significant amount of water (50 to75 per-
cent water) along with the oil, however, unnecessarily
increasing the volume of waste requiring disposal and the
associated costs. Last year, the truck collected 500 gal-
lons of oil and water, which cost $250 (or $0.50 per
gallon) to dispose of. Consequently, the city is pursuing
installation of a skimming system.
The city is likely to choose an oleotrophic rope skimmer
to remove the LNAPL from the water surface without
removing any water. Collecting dramatically decreases
both the volume and consequently the disposal costs.
This type of skimmer has been used successfully at simi-
lar sites, and the skimmer manufacturer guarantees its
performance.
Ground-water remediation will probably be required as a
long-term solution.
A.1.4 Conclusions and Discussion
Although this facility is not a RCRA or CERCLA site, the
interim measure selection and implementation were con-
sistent with the goals of the RCRA Stabilization Initiative.
Once the contamination was identified, interim measures
were applied to stabilize the subsurface contamination by
removing the LNAPL. This provided time to more com-
pletely characterize the site and to determine the neces-
sity of long-term ground-water remediation. Several
interim measures were analyzed based on technical, 16-
68
-------
gistical, and cost criteria, The subsurface trench system
was selected due to the clay geology, the LNAPL's rela-
tively immobile nature, the low estimated cost, and com-
patibility with potential long-term remedies and planned
use for the site. The system has operated since Septem-
ber 1990 and was constructed for about $77,000.
A.2 LANDFILL STABILIZATION: BFI SOLLEY ROAD
FACILITY, GLEN BURNIE, MD, by Dennis
Zielinski, U.S. EPA, Region III, Philadelphia, PA
A.2.1 Introduction
Browning-Ferris Industries (BFI) owns and maintains a
65-acre closed landfill on Solley Road in Anne Arundel
County in Glen Burnie, Maryland (see Figure A-1). The
site is located approximately 3 miles south of Baltimore
and approximately 2 miles east of Glen Burnie, in the
south-central portion of Marley Neck peninsula (see Fig-
ure A-2). The peninsula is bounded on the west by Marley
Creek, on the east by Stony Creek, and on the north by
Curtis Bay. The areas north and west of the site are
generally considered wetlands. The site was used for the
disposal of both hazardous and nonhazardous municipal
and industrial waste from 1963 until 1982.
The area surrounding the site is sparsely populated and
zoned for light industry. It is estimated that approximately
750 people live within 1 mile of the site. Only 20 of these
people live northwest of the site. This is fortunate be-
cause the contamination plume is moving in that direc-
tion. Residential development within the area has
remained scattered, with very few residences found north
Scale
2000
2000 Feet
Site Location Map
Remedial Engineering Plan Revision 2.0
Solley Road Facility
Figure A-1. Location of case study site, Browning-Ferris
Industries (BFI) landfill.
Location Map
Scale
404 mites
Figure A-2. BFI landfill site in relation to Baltimore and
Glen Burnie, Maryland.
and west of the site. Private drinking water wells are not
located in close proximity to or downgradient of the site.
Stabilization is being performed in lieu of waiting 2 to 3
years for a full-scale corrective action. The stabilization
action, which is a cooperative effort among EPA, the State
of Maryland, and BFI, is still in the preliminary stages of
review for BFI's proposed plans.
The following subsections discuss the site's solid waste
units, geology and topography, ground-water contamina-
tion, regulatory chronology, and risk.
A.2.1.1 Solid Waste Management Units
The Solley Road site was developed as a sanitary landfill
in the early 1960s by the former owner of the site. BFI
purchased the site in 1973.
The site consists of four SWMUs: two closed landfills with
leachate collection systems and two sedimentation
ponds. The landfills are divided by overhead electrical
transmission lines, with a 120-foot wide right-of-way run-
ning in a north-south direction.
The site received mostly municipal refuse, and had an
East Fill and a West Fill. The East Fill, the area first
developed as a landfill, lies in the northern portion of the
area east of the right-of-way. The West Fill, a triangular
shaped fill located west of the right-of-way, was the sec-
ond area to receive waste at the site, and accumulated
the majority of the volume. This case study focuses on
the West Fill, which was closed in late 1977.
A permitted "Active Cell" within the East Fill started op-
erating in July 1980 as a secure hazardous waste dis-
69
-------
posal site. This cell occupies 8 acres in the southern
portion of the East Fill. The rest of the East Fill was closed
In 1980 and did not receive municipal wastes after the
"Active Cell" was opened. The entire site was closedjon
December 31, 1982.
In 1990, BFI installed a leachate collection system in ^he
southernmost section of the East Fill. The leachate is
carried by double-lined pipes on the ground to an under-
ground double-lined tank, which stores the leachate for
less than 90 days. Monitoring systems in the outside
piping and tank walls will detect a release. The material
In the tank is pumped weekly to a tank truck and trans-
ported by a licensed hazardous waste transporter to[ an
approved hazardous waste facility.
The West Fill was not designed or constructed with a
leachate collection system, since operations at the West
Rll predated regulations requiring these systems. The
West Fill, however, does have a well system that collects
leachate. Leachate is pumped directly from the wells to
a tank truck, which takes the leachate to an approved
facility.
Rain water runoff from the site is collected in a sedimen-
tation pond located south of the West Fill. The rainwater
runoff may contain contaminants from the landfills as a
result of erosion of the landfill covers. There is no evi-
dence that hazardous wastes, or hazardous waste con-
stituents, have been released from the permitted
hazardous waste cell to the ground water, surface water,
or the land surface.
A.2.1.2 Geology and Topography
BFI's Solley Road site is situated on an outcrop of the
Patapsco Formation, which is a multilevel aquifer ssystem
consisting of water-bearing sands alternating with confin-
ing beds of clay or interbedded clays, silts, and sands.
The surficial soil is a red clay, silt and interbedded clay,
and silt and sand (see Figure A-3). It is approximately 80
Approximate
Depth _ VP5
North/Northeast
VP2
VP1 VWD
VWM
VWS
VPS
VP4
Shallow
M. Sand
51—
VP5-S
VP2-SB VP1-S
VP3-SI
VP4-S
10' —
15' —
20' —
F-VF Sands
TR. Silt
VP5-MI
VP2-MSVP1-M1
VP3-M
VP4-M
25'
025+
M-C Sands
VP5-D
VP1-M2
VP2-DM VP1-DI
VP3-D
VP4-D
055f
Figure A-3. Soil geology of the site showing clay, silt and interbedded clay, and silt and sand.
70
-------
to 100 feet thick along the eastern boundary of the site,
and thins to less than 20 feet in the western portion of
the site. Contact between the overlying clay and under-
lying sands of the upper portion of the Patapsco Aquifer
is well defined at the eastern portion of the site, but
becomes transitional towards the west. A second semi-
continuous zone of clay or interbedded clay, silt, and sand
separates the upper portion of the aquifer from the mid-
dle. A similar sequence of clay or interbedded clay, silt,
and sand separates the middle and lower portions of the
Patapsco Aquifer.
in general, ground water flows to the north and west at
approximately 8 to 10 feet per year. The Patapsco Aquifer
is the first continuous water-bearing zone underlying the
site. The top of the Patapsco Aquifer beneath the site
varies from 10 feet above mean sea level to 30 feet below
mean sea level.
A.2.1.3 Ground-Water Contamination
The site has two contamination plumes, both located near
the West Fill, one in the upper Patapsco Aquifer and the
other in the middle Patapsco Aquifer. The contaminated
plume in the upper Patapsco Aquifer begins on the west
side of the West Fill and is approximately 1,000 feet along
the fill and 700 feet to the west of the fill. The total VOCs
isoconcentration contours range from 5 to 1,000 ppb. The
plume in the middle Patapsco Aquifer is approximately
1,200 feet along the west side of the West Fill and 200
feet west of the fill. The total VOCs isoconcentration con-
tours range from 5 to 500 ppb. The plumes have not
changed significantly since 1985, but they have migrated.
The water quality analyses of samples taken in June and
December 1989 indicated that the contaminants in the
plume consisted of VOCs, including aromatics and chlo-
rinated aliphatics. Of the 15 chemicals of concern, 5 had
concentrations greater than the MGL (see Table A-3).
One important difference between the June and Decem-
ber samplings was that in December the farthest well in
the northwest location showed VOCs for the first time.
Table A-3. Contaminants with Concentrations Greater
Than the MCL
Contaminant
1,1 -DCE
Trans-1 ,2-DCE
Vinyl Chloride
Benzene
TCE
Number of
Multiples of the
MCL Measured
by the Highest
Sample on 6/89
5
8
20
100
1,300
Number of
Multiples of the
MCL Measured
by the Highest
Sample on 12/89
NA
10
50
65
80
This might be due to continued slow migration of con-
taminants from the site. Table A-4 summarizes the results
of water quality analyses of samples taken in June and
December 1989.
A.2.1.4 Risk — Current/Potential
There is no immediate human health threat at this site.
The contamination plume is moving towards the Marley
Neck Creek, which feeds into the Patapsco Aquifer and
then into the Chesapeake Bay. BFI has performed a pub-
lic health assessment of the Sol ley Road site and as-
sessed the "potential" risk posed by possible future
exposure to the chemicals of concern in ground water,
surface water, and/or sediments. They assessed "poten-
tial" exposure because none of the chemicals of concern
have been detected in any residential well. Human expo-
sure to surface water and sediment is not known to have
occurred.
From past sampling, the 15 chemicals listed in Table A-4
were identified as the chemicals of concern. In addition,
chromium also was considered a chemical of concern in
sediments and surface water. Fifteen downgradient wells
were used to determined the average concentrations of
the chemicals. Then the wells were grouped into three
zones based on their distance from the northern and
western boundaries of the landfill. Zone 1 wells are less
than 100 feet from the landfill; zone 2 wells are 100 to
600 feet from the landfill; and zone 3 wells are greater
than 600 feet from the West Fill.
Of the 15 organic chemicals of concern, only benzene,
chloroform, trichloroethene (TCE), and vinyl chloride are
potential carcinogens. Table A-5 lists lifetime carcinogenic
risks, using conservative concentrations posed by aver-
age intake from all routes of exposure, of these chemicals
in the ground water.
Acceptable Intakes for Chronic Exposure (AlCEs) were
used to assess the noncarcinogenic risks posed by hy-
pothetical intake of the chemicals of concern. The total
intake/AlCE value for a particular chemical was used to
assess the potential for toxic effects. Total intake/AICE
values less than 1 indicate that the level of intake would
not adversely affect human health. Table A-6 summarizes
total intake/AICE values for an adult, using conservative
concentrations.
The noncarcinogenic risks due to intake of chromium
from wading and ingestion were very low. Values for total
chromium intake/hexavalent chromium reference dose
(RfD) and total chromium intake/trivalent chromium RfD
are 3.48 x 10'3 and 1.74 x 10'5, respectively, indicating
that adverse health effects should not result from the
chromium intake level calculated for the average 10-year-
old child for either species of chromium.
In summary, the conservative analysis of the carcinogenic
and noncarcinogenic risks posed by hypothetical expo-
71
-------
Table A-4. Summary of Sampling Data on VOC Concen-
trations (ppb) ;
Table A-5. Summary of Carcinogenic Risks
Compound
June 1989
Acetone
4-Methyl-2-pentanone
(MIBK)
Benzene
Toluene
Ethyl Benzene
Xylenes
Chloroform
Trfchlorofluoromethane
1,1-Dichloroethane
(1,1-DCA)
Vinyl Chloride
1,1-Dichloroethene
(1,1-DCE)
Trans-1,2-DCE
Trichloroethene
(TOE)
Chlorobenzene
1,1,1-TCA
December 1989
Acetone
MiBK
Benzene
Toluene
Ethylbenzene
Xylenes
Chloroform
Trichlorolluoromethane
1,1-DCA
Vinyl Chloride
1,1-DCE
Trans-1,2-DCE
TCE
Chlorobenzene
1,1,1-TCA
Number
of Wells
w/VOCs*
1
2
8
8
2
13
1
0
1
3
2
6
13
0
0
0
0
3
4
1
2
0
0
3
1
1
10
17
0
3
Range
56
35-83
4-470
4-33
10-12
8-37
5
N/A
6
24-37
38-40
3-830
4-640
N/A
N/A
N/A
N/A
15-340
5-23
2
9
N/A
N/A
6-11
, 96
2
2-970
3-380
N/A
2-6
MCL
[
3.5001
1.8001'2
5
1,000
700
10,000
100
1.8001
5801
2
7
100
5
100
200
3.5001
1.8001'2
5
1,0,00
700
io,o!oo
100
1 ,8001
5;801
'. 2
> 7
100
; 5
100
200
1 No MCL, the health-based concentration in drinking water exists.
Using tie health-based concentrations, no adverse health effect ^would
be expected if individuals were to drink 2 liters per day of water con-
taining any single one of the VOCs shown at levels indicated. j
2 Health-based limit for methyl ethyl ketone, not 4-methyl-2-pentanone.
* 37 total Wells ;
sure to the chemicals of concern indicates low lifetime
risk for receptors using ground water in zones 2 and 3.
Lifetime cancer risks posed by hypothetical lifetime expo-
sures to benzene, chloroform, TCE, and vinyl chloride in
zone 1 exceeded the upper bound of the EPA acceptable
risk range. '
Chemical
Zone
Risk
Benzene
Benzene
Chloroform
TCE
TCE
TCE
VC
Total Risk
Total Risk
Total Risk
1
2
1
1
2
3
1
1
2
3
1 X 10"4
6X 10"8
1 X 10"7
7X10"5
3X10'5
1 X 10"5
2X 10"3
2X 10'3
3X10'5
1 X 10"5
Table A-6. Summary of Noncarcinogenic Risks
Chemical
Zone
Total Intake/AICE
Acetone
Chloroform
1,1 -DCA
1,2-DCE
1,2-DCE
1,2-DCE
Ethylbenzene
Isophorone
MEK
MIBK
Naphthalene
Toluene
Toluene
1,1,1-TCA
1,1,1-TCA
Xylene
Sum of intake/AICE
Sum of intake/AICE
Sum of intake/AICE
1
1.
1
1
2
3
\
1
1
1
1
1
2 ,
1
2
1
1
2
3
3.44 X 10"4
5.86 X 10"4
6.44 X10-4
1 .48 X 1 0"1
2.07 X 10'2
1.17X 10'2
1 .35 X 1 0'3
o
5.44 X 10'3
o
2.24 X ID'3
' 9.08 X 10"3
2.70 X 10'5
o
1.39X 10'°
1.24X 10"5
, 1.76X 10"4
1.61 X 10"4
7.00 X 10"5
1.69X 10'1
2.08 X 10'2
1.17X10'2
A.2.2 Stabilization Strategies
Due to the many studies performed over the years, BFI
felt confident that it accurately characterized the ground-
water contamination. BFI wanted to limit or delay further
study and proceed with a pump-and-treat measure. They
requested a permit modification under 40 CFR Section
270.42 to incorporate the stabilization measure into their
permit. The permit requires posting of signs, regular in-
spections of fences, maintenance of covers, verification
investigation, a RCRA Facility Investigation (RFI), a Cor-
rective Measures Study (CMS), and initiation of corrective
measures. If the stabilization effort is effective in reducing
ground-water contamination, then BFI wants it to become
the final corrective measure.
The following subsections discuss various aspects of the
pump-and-treat strategy, including an overview, extraction
72
-------
weff data, air stripper description, reinjection well location,
leachate collection system design, and monitoring system
setup and operation.
A.2.2.1 Pump-and-Treat Stabilization Measure —
Overview
BFI has submitted a Remedial Engineering Plan to the
State of Maryland and EPA for approval. This plan is
currently under review by both regulatory agencies. This
plan is designed to stabilize the ground-water contamina-
tion associated with the Solley Road site using pump-
and-treat technology with a leachate collection system.
BFI will pump contaminated ground water from the Patap-
sco Aquifer to an air stripper (see Figure A-4). The con-
taminants of concern will be removed and the water will
be reinjected, through three reinjection wells, back into
the Patapsco Aquifer downgradient of the plume, thereby
reversing the flow of contaminants in the Patapsco Aqui-
fer in that area. Secondary impacts to other ground-water
users in the area are expected to be negligible. The ex-
traction wells will draw about 120 gallons per minute out
of the middle and upper aquifers. The site will be capped
with a vegetative cover and have a leachate collection
system.
A.2.2.2 Data and Flow of Wells
BFI will install four extraction wells designated as X-1,
X-2, X-3, and X-4. Well X-1 will be located 500 feet south
of the northwest corner of the West Fill. The well will be
screened over the length of the upper Patapsco Aquifer
with a solid section across the clay layer, and pumped al
a rate of 20 gallons per minute (gpm). Well X-2 will be
located 450 feet north of X-1. It will be screened over the
lower portion of the upper Patapsco Aquifer and be
pumped at 35 gpm. Well X-3 will be located 40 feet north
of X-2, screened over the middle Patapsco Aquifer, and
pumped at 30 gpm. Well X-4 will be installed 80 feet east
of X-3, screened over the middle Patapsco Aquifer, and
pumped at 35 gpm.
A.2.2.3 Air Stripper
The air stripper system will be designed to reduce TCE
concentrations in contaminated ground water from 212
ppb to an effluent concentration of less than 5 ppb. The
system will be designed to operate at flow rates between
80 and 160 gpm, with a normal flow rate of 120 gpm (see
Figure A-5).
Extracted water will be treated in a 34-ft tall 2.5-ft diame-
ter packed tower. The tower will contain 24 feet of 2-inch
polypropylene packing and will be equipped with a 1.5
horsepower blower, and will provide air-to-water ratios
ranging from 40:1 to 85:1. The process will effectively
remove 98 percent of the influent VOCs, and the treated
ground water will be reinjected into the aquifer.
To determine influent design concentrations, the concen-
trations of VOCs detected for all sampling events were
averaged for each well. Based on the location of these
wells and the recovery wells, and the design flow rates
of the recovery wells, a weighted average concentration
Control Building
Recovery Well
with Pump (Typ)
^-Recharge Well (Typ)
D-1
Effluent Sump
X-4
Figure A-4. Pump and treat with air stripper at Solley Road site.
73
-------
Air Exhaust
Extracted
Ground-Water Flow
Air Blower
Packed Tower
Air Stripper
Air Contra Flow
To Ground-Water
Recharge Wells
Figure A-5. Air stripper system at Solley Road site.
was calculated. For additional safety, the design concen-
trations were set at double the weighted average concen-
trations. The design effluent concentration was based on
the Safe Drinking Water Act primary MCLs for VOCs.
A.2.2,4 Reinfection Wells
BFI will install three reinjection wells designated as b-1,
D-2, and D-3, which will be located in a line approximately
100 feet east of Marley Neck Road. These reinjection
wells will be screened in all three layers of the Patapsco
Aquifer and will receive a total of 120 gpm: well D-1J will
receive 55 gpm; well D-2,40 gpm; and well D-3, 25 gpm.
The flow rates to wells D-1 and D-2 are greater than trjose
to D-3 because some mounding potential exists near D-3,
due to the enhanced recharge from the stream.
A.2.2.5 Leachate Collection System \
The leachate collection system consists of excavkted
trenches in which filter geotextile, aggregate, and perfo-
rated pipes are placed along the perimeter of the; East
and West Fills. These trenches and pipes gravity drain to
four reinforced concrete standpipe risers, which pro'vide
access for leachate removal via pumps or vacuum trucks.
A clay cap, topsoil, and vegetative layer will be placed
over the trenches.
A.2.2.6 Monitoring Systems
A monitoring system will verify the performance of the
pump-and-treat system's removal efficiency, the hydraulic
response of the aquifers, and changes in ground-water
quality.
The removal efficiency of the air stripper will be measured
to verify that the effluent discharge limits are being met.
A portable gas chromatograph (GC) will be used at the
site to monitor contaminant concentrations in the stripper
influent and effluent. Laboratory analysis also will be per-
formed to verify system performance. The sampling fre-
quency will be as follows: first week — daily; next 6
months — weekly; next 6 months — monthly; next year
— every other month; and the remainder of the project
— quarterly.
BFI will monitor the hydraulic response of the aquifer to
pumping to verify that the capture zone of the system is
sufficient to recover the majority of the ground-water con-
tamination plume. This monitoring will use the same sam-
pling frequency as above.
Ground-water quality monitoring will consist of baseline
monitoring 4 weeks prior to startup to provide a "snap-
shot" of existing conditions, and routine monitoring at
regular intervals. Routine sampling on 19 wells will be
performed quarterly for the first year after startup and
semi-annually thereafter. Routine sampling will verify the
systems' performance relative to the baseline.
A.2.3 Implementation and Future Actions
BFI believes that, having studied the site for years, they
have characterized the ground-water contamination.
They would prefer to limit or delay further investigative
stages for ground water, and start implementing a pump-
and-treat measure to stabilize the ground-water contami-
nation. The site may request a modification to their
corrective action permit under 40 CFR Section 270.42 to
incorporate the stabilization effort.
The modification of EPA's corrective action permit should
take place in March or April 1992. Installation of the
pump-and-treat system should begin in April and the sys-
tem should be operational by August 1992.1
A.2.4 Conclusions and Discussion
In summary, there is no indication that contamination from
the landfills presents an immediate threat to human health
1 Due to EPA and Maryland's concern with media transfer of contami-
nants, BFI also will be installing a carbon adsorption unit to treat the ef-
fluent from the air stripper. EPA now plans to modify BFI's permit by
November 1992. This modification will require BFI to install the new
system in March or April 1993, and begin pumping and treating in Sep-
tember 1993 (Zielinski, 1992).
74
-------
and the environment in the site vicinity. BFI has identified
a localized plume consisting of VOCs in concentrations
as much as 100 times the MCL, migrating in the upper
portions of the Patapsco Aquifer. Therefore, instead of
performing additional studies required in EPA's corrective
action permit, BFI would like to stabilize the site by in-
stalling a pump-and-treat system. BFI also would install
a leachate collection system to eliminate significant re-
leases to the environment. They would continue to study
the impact to the other media (e.g., air and soils), as well
as the ground water, while pursuing the cleanup of
Table A-7. Regulatory Chronology
Date Event
10/72 MDHMH* issues Refuse Disposal Permit for East and
West Fills
1973 BFI buys the site
8/78 MDHMH amends Refuse Disposal Permit to allow
hazardous waste disposal into "Active Cell" in the East
Fill
1.979 MDHMH requires installation of monitoring system with
6 wells
7/82 MDEP** issues BFI a Treatment, Storage, and Disposal
Facility (TSDF) Permit
1982 MDHMH determines that the monitoring system is
damaged and requires installation of 10 more wells
12/82 Entire facility was closed; state permit was amended
with an approved post-closure plan with comprehensive
monitoring
1984 Ground-water quality data collected indicate contamina-
tion extends north and west of the site
4/85 Remedial Investigation starts
8/88 Remedial Investigation agreement is amended to
include studies to prevent further migration of
contamination
10/86 BFI submits Evaluation Reports that indicate that
and ground-water contamination is not emanating from the
3/89 East Fill
10/89 BFI submits a Remedial Engineering Plan, which
includes construction on adjacent sites; BFI had
. difficulties accessing the adjacent land
10/90 MDEP issues Post-Closure Permit
8/91 EPA Region III issues a Corrective Action Permit
consistent with the MDEP Post-Closure Permit
9/91 BFI acquires parcels of land adjacent to the site to
implement remediation plan
11/91 EPA Region III suggests that BFI consider the state's
ground-water remediation plan as a "stabilization"
activity under the EPA permit to avoid the delay involved
by waiting for the full corrective action permit start date
of 1994
*MDHMH = Maryland Department of Health and Mental Hygiene
"MDEP = Maryland Department of Environmental Protection
ground-water contamination. Modeling and pilot studies
performed by BFI indicate that VOCs in the ground water
can be successfully treated using air stripping.
Table A-7 presents a summarized chronology of regula-
tory action.
If this interim measure effectively cleans up the ground-
water contamination, it will become the final corrective
measure.
A.2.5 References
Epsy, Huston & Associates, Inc. 1991. RCRA Corrective Action
Permit, Solid Waste Management Unit Assessment, BFI Sol-
ley Road Landfill. Glen Bumie, MD. October.
Epsy, Huston & Associates, Inc. 1991. RCRA Corrective Action
Permit, Verification Sampling Plan — SWMU 5 & 6, BFI
Solley Road Landfill. Glen Bumie, MD. December.
Nye, Alan C. 1989. Solley Road Site Public Health Assessment,
Volume 4, Supplemental Studies — Remedial Investigation
Report. TERRA, Incorporated. March.
Rizzo, Paul C. Associates, Inc. 1991. Remedial Engineering Plan
Revision 2.0, Ground Water Recovery and Treatment System
Solley Road Facility. Anne Arundel County, Maryland. May
12.
Rotenberg, Samuel L. 1992. U.S. EPA Region III, Memorandum to
Dennis Zielinski, U.S. EPA Region III, MCLs and Health-
Based Water Concentrations for some VOCs. February 6.
Scheline, James R. 1991. BFI, Status of Corrective Action Activities
at Browning-Ferris, Inc. (BFI) Solley Road Facility, Glen
Bumie, Maryland, to Dennis Zielinski. U.S. EPA Region III.
December 6.
U.S. EPA. 1990. U.S. Environmental Protection Agency. BFI Solley
Road Closed Landfill, Fact Sheet. August 7.
U.S. EPA. 1991. U.S. Environmental Protection Agency. Permit for
Corrective Action and Waste Minimization under the Re-
source Conservation and Recovery Act as Amended by the
Hazardous and Solid Waste Amendments of 1984. July 18.
Zielinski, D. 1992. Conversation between Dennis Zielinski, U.S. EPA
Region III, and Susan Richmond, Eastern Research Group,
Inc. September 15.
A.3 GROUND-WATER CONTAMINATION
STABILIZATION: ROMIC CHEMICAL
CORPORATION FACILITY, EAST PALO ALTO,
CA, by Glenn Heyman, U.S. EPA, Region IX, San
Francisco, CA
A.3.1 Introduction
The Romic Chemical Corporation owns and operates a
solvent recycling and wastewater treatment facility in East
Palo Alto, California. Romic has owned the site since
1979 and has operated the facility since 1963. Solvent
recycling operations began on this site in the late 1940s.
The facility currently recycles about 7 million gallons of
halogenated and nonhalogenated solvents annually.
The 10-acre site is slightly less than V2 mile west of the
southern end of San Francisco Bay. It is bordered by
industrial sites and auto dismantling shops on the south
and west, a former salt evaporation pond to the east-
75
-------
northeast, and salt marsh and tidal sloughs on the north
and east. A nearby industrial site to the south is on the
National Priority List of Superfund sites. i
The following subsections discuss the site's hydrogebl-
ogy, contamination, environmental concerns, and regula-
tory activities. ;
A.3.1.1 Hydrogeology
The Romic site is underlain by recent alluvium composed
of approximately 70 feet of interbedded sand and clay.
Three shallow interconnected aquifers occur at depths
from near surface to 15 feet, 30 to 40 feet, and 60 to I70
feet and are informally referred to as the A, B, and C
zones, respectively. Their flow directions are variable and
are connected to the surface water system in the tidal
slough. A deep underlying aquifer at approximately 150
to 170 feet below surface is utilized as a drinking water
supply. The nearest extraction wells for the drinking waiter
are about 1 mile west-southwest of the site. While water
in zone A is affected by tides, ground-water flow in all
zones is generally eastward toward the bay.
A3.7.2 Contamination
VOCs have been detected above regulatory levels in the
ground water and slough sediments. Ground-water con-
tamination extends horizontally beyond site boundaries
and vertically to the aquitard. Romic has conducted a RFI
pursuant to its permit and Section 3008(h) Consent Order.
This investigation has shown that releases from the site
have contaminated surface water, soils, subsurface gas,
and shallow ground water down to at least the base of
zone C. VOC contaminants also have been detected in
the shallow ground water and subsurface gas horizontally
beyond the site boundaries. Analysis of soil and ground-
water samples have detected about 27 organic com-
pounds along with lesser inorganic contamination. Tables
A-8 through A-10 contain contaminant profiles for site
i
Table A-8. Major Ground-Water Contaminants |
Table A-9. Major Surface Water Contaminants
Contaminant
Vinyl Chloride
cis-1 ,2-Dichloro-
ethylene
1 ,2-Dichloroethane
Trichloroethene
Freon 113
Methylene Chloride
Acetone
Methyl Ethyl Ketone
Tetrahydrofuran
Concen-
tration on
8/91 In
ppm
16.0
34.0
3.4
1.9
0.4
64.0
120.0
840.0
94.0
Subpart S
Action
Levels in
ppm
—
—
—
—
—
0.005
4
2
"
MCLs
in
ppm
0.002
0.007
0.005
0.005
—
— -
—
— r
~~
Contaminant
Concentration
on 8/91 in
ppm
Marine
Chronic
Criteria in ppm
Organic Contaminants
cis-1,2-Dichloro- 0.150
ethylene
1,2-Dichloroethane 0.044
Vinyl Chloride 0.104
Tetrahydrofuran 2.300
Inorganic Contaminants
Copper 0.270
Lead 0.399
Mercury 0.005
Nickel 0.039
224
113
0.0029
0.0056
0.00025
0.0083
Table A-10. Major Slough Sediment Contaminants
Contaminant
Concentration
on 11/91 in
ppm
PSDDA
SLin
ppm*
PSDDA
ML in
ppm**
Vinyl Chloride 7.200
1,1-Dichloroethylene 0.240
cis-1,2-Dichloro- 3.500
ethylene
Trichloroethene 0.130
Toluene 1.000
Tetrachloroethane 0.029
1,4-Dichlorobenzene 0.031
Ethylbenzene 0.260
Total Xylene 0.780
Tetrahydrofuran 0.130
0.160
0.014
O.Q26
0.010
0.012
1.600
0.210
0.260
0.050
0.160
*PSDDA SL stands for Puget Sound Disposal Data Analysis Safe Level,
which indicates no observed effect.
"PSDDA ML stands for Puget Sound Disposal Data Analysis Minimum
Level, which indicates an observed adverse effect.
ground water, adjacent surface water, and slough sedi-
ments, respectively.
The primary contamination sources appear to be unpaved
drum storage areas and two former surface ponds in the
northern part of the site. These ponds had received sur-
face runoff, process wastewater, dimethyl sulfoxide salts,
and miscellaneous waste materials between the 1950s
and the late 1970s. Beginning in the late 1970s, the
ponds were filled and drum storage buildings built over
them. These buildings were constructed over concrete
slabs poured over a synthetic liner that covered the filled
ponds. An additional contamination source consists of a
trough that carried wastewater from the process area to
76
-------
the ponds. The contamination covers all three zones A,
B, and C. Because of a high background level of inorganic
contamination in surface water from the proximity of other
industries, assessing Romic's contribution of metals to
surface water is difficult.
A.3.1.3 Environmental Concerns
Ground-water contamination migrating off site presents
potential hazards to nearby water supplies and surface
waters. The closest drinking water supplies are municipal
water wells located about 1 mile southwest of the site.
These drinking water wells, however, are isolated from
the shallow aquifer by a confining clay layer. The wells
draw on an aquifer about 170 feet under the surface, and
the protective clay layer over the aquifer is about 100 feet
thick. Shallow ground water from this site discharges di-
rectly to an adjacent salt marsh, abandoned salt evapo-
rator, and surface water body. In addition, two species
listed on state and federal endangered species lists, the
California clapper rail and the saltmarsh harvest mouse,
inhabit the adjacent wetlands.
A.3.1.4 Regulatory Activities
The State of California and EPA have been involved with
this site since the 1980s. California issued an operating
permit in 1985, and in December 1988 EPA and Romic
signed a RCRA Section 3008 (h) Consent Order to initiate
a RFI, CMS, and a CMI Plan. The RFI is now almost
complete. Sampling conducted as part of the RFI in-
cludes:
• Installation of 34 monitoring wells
• Drilling and sampling of 32 soil borings
• Sampling of slough sediment at 8 locations
• Establishment of 8 surface water sampling locations
« Completion of a soil-gas survey
The permit-required sampling originally discovered the
VOC contamination in the soils and the ground water.
A.3.2 Stabilization Strategies
Romic began stabilizing this site in the 1970s by filling
the settling ponds, covering them with a synthetic liner,
and paving the site. In April 1991, the final Interim Cor-
rective Measures Evaluation and Feasibility study was
completed. Its guiding philosophy is to stop offsite migra-
tion of contaminated ground water and soil gas and to
begin long-term remediation of the ground water beneath
the site. Interim corrective measures will be in place as
a conditional remedy until the facility closes, at which time
a determination of final remediation will be made.
The preliminary stabilization plan is a pump-and-treat op-
tion involving extraction wells and granular activated carb-
on adsorption treatment. Once the preliminary system is
installed and evaluated, additional source controls (such
as slurry cutoff walls) and treatment technologies will be
evaluated. Treatment technologies initially under consid-
eration included ultraviolet oxidation using peroxide, a
biological reactor for ketones, and air stripping for volatile
chlorinated solvents. Ultraviolet oxidation using peroxide
was rejected, however, after a pilot-scale treatability test.
Testing of the preliminary system cannot begin until the
wastewater discharge issue is resolved. The local POTW
has denied Romic a discharge permit. Romic has sub-
mitted a National Pollutant Discharge Elimination System
(NPDES) permit application in the event that the local
POTW denies Romic's permit denial appeal.
A.3.3 Implementation and Future Actions
One of the impediments to stabilizing this site is the need
for gaining access to adjacent parcels of land to complete
evaluation studies. Romic has not been able to gain ac-
cess to an adjacent hiking and biking trail to complete
determination of the eastward extent of contamination.
Another impediment is obtaining the requisite permits to
discharge effluent to the sanitary sewer. Discussions with
the local sanitary district began in June 1991. The dis-
charge permit application was denied in September 1991,
and has been subsequently resubmitted.
Current activities at the site include a RCRA Facility In-
vestigation and stabilization activities. The RFI activities
include:
• Borings and well installations to delineate the horizontal
and vertical extent of the VOC contamination
• Preparation to conduct an air pathway monitoring pro-
gram using surface flux chambers
• Detailed ecological assessment preparation
Stabilization activities include NPDES permit application
revisions requested by the Regional Water Quality Con-
trol Board (RWQCB), additional pump tests, and ground-
water modeling to determine recovery well spacing.
A.3.4 Conclusions and Discussion
The overarching lesson from this case study is the need
for thorough and clear communication. It is important to
first realistically evaluate local politics to assess the most
expeditious investigation, treatment, and disposal op-
tions. In addition, all local agencies must be educated
about their roles in the process, since, as in this case,
the peripheral agencies may present the largest obstacles
to proceeding with stabilization efforts. Another lesson is
the importance of thoroughly exploring reuse options for
treated ground water.
77
-------
-------
-------
EPA/625/R
1
<£>
g
— ••
4^
"§
§g «:
^SJL
3-g
"• to
-05-
g. co
< CO
CD co
CD
c
CO
CD
o o m c
g § i i-
§• ff g g.
ls:|^
g||f
•^ 3 T
en S 2
w — 2.
O) Zl CD
00 CD O
3 Cf«
33 ^
CD CO
CD
------- |