United States
Environmental Protection
Agency
Solid Waste and Emergency Response (5102W)
Innovative Site
Remediation
Technology
Vacuum Vapor Extraction
Volume 8
-------
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, Uin rioor
Chicago, IL 60604-3590
-------
INNOVATIVE SITE
REMEDIATION TECHNOLOGY
VACUUM VAPOR
EXTRACTION
One of an Eight-Volume Series
Edited by
William C. Anderson, P.E., DEE
Executive Director, American Academy of Environmental Engineers
1994
Prepared by WASTECH®, a multiorganization cooperative project managed
by the American Academy of Environmental Engineers® with grant assistance
from the U.S. Environmental Protection Agency, the U.S. Department of
Defense, and the U.S. JDepartment of Energy.
The following organizations participated in the preparation and review of
this volume:
Air & Waste Management
Association
P.O. Box 2861
Pittsburgh, PA 15230
, Hazardous Waste Action
Coalition
1015 15th Street, N.W., Suite 802
Washington, D.C. 20005
American Academy of
Environmental Engineers®
130 Holiday Court, Suite 100
Annapolis, MD 21401
f S \
\ W J Society for Industrial
^ = * Microbiology
3929 Old Lee Highway, Suite 92A
Fairfax, VA 22030
lASOEl
Yify American Society of
\/4> Civil Engineers
345 East 47th Street
New York, NY 10017
Water Environment
' Federation
601 Wythe Street
Alexandria, VA 22314
Published under license from the American Academy of Environmental Engineers11.
© Copyright 1994 by the American Academy of Environmental Engineers®.
-------
Library of Congress Cataloging in Publication Data
Innovative site remediation technology/ edited by William C. Anderson
224 p. 15.24 x 22.86cm.
Includes bibliographic references.
Contents: — [2] Chemical treatment - [3] Soil washing/soil flushing
— [4] Stabilization/solidification -- [6] Thermal desorption - [7] Thermal destruction
[8] Vacuum vapor extraction
1. Soil remediation. I. Anderson. William, C., 1943- .
II. American Academy of Environmental Engineers. III. Johnson, Paul C., 1961- .
TD878.I55 1994 628.5'5 93-20786
ISBN 1-883767-02-4 (v. 2) ISBN 1-883767-06-7 (v. 6)
ISBN 1-883767-03-2 (v. 3) ISBN 1 -883767-07-5 (v. 7)
ISBN 1-883767-04-0 (v. 4) ISBN 1-883767-08-3 (v. 8)
Copyright 1994 by American Academy of Environmental Engineers. All Rights Reserved.
Printed in the United States of America. Except as permitted under the United States
Copyright Act of 1976, no part of this publication may be reproduced or distributed in any
form or means, or stored in a database or retrieval system, without the prior written
permission of the American Academy of Environmental Engineers.
The material presented in this publication has been prepared in accordance with
generally recognized engineering principles and practices and is for general
information only. This information should not be used without first securing
competent advice with respect to its suitability for any general or specific applica-
tion.
The contents of this publication are not intended to be and should not be
construed as a standard of the American Academy of Environmental Engineers or of
any of the associated organizations mentioned in this publication and are not
intended for use as a reference in purchase specifications, contracts, regulations,
statutes, or any other legal document.
No reference made in this publication to any specific method, product, process,
or service constitutes or implies an endorsement, recommendation, or warranty
thereof by the American Academy of Environmental Engineers or any such
associated organization.
Neither the American Academy of Environmental Engineers nor any of such
associated organizations or authors makes any representation or warranty of any
kind, whether express or implied, concerning the accuracy, suitability, or utility of
any information published herein and neither the American Academy of Environ-
mental Engineers nor any such associated organization or author shall be responsible
for any errors, omissions, or damages arising out of use of this information.
Book design by Lori Imhoff
Printed in the United States of America
WASTECH and the American Academy of Environmental Engineers are trademarks of the American
Academy of Environmental Engineers registered with the U.S. Patent and Trademark Office.
-------
CONTRIBUTORS
I
This monograph was prepared under the supervision of the WASTECH®
Steering Committee. The manuscript for the monograph was written by a task
group of experts in vacuum vapor extraction and was, in turn, subjected to two peer
reviews. One review was conducted under the auspices of the Steering Committee
and the second by professional and technical organizations having substantial
interest in the subject.
PRINCIPAL AUTHORS
Paul Johnson, Ph.D., Task Group Chair
Associate Professor
Department of Civil and Environmental Engineering
Arizona State University
Arthur Baehr, Ph.D. Robert Hinchee, Ph.D.
-\ Water Resources Division Research Leader
SJ U.S. Geological Survey Battelle Memorial Institute
{] Richard A. Brown, Ph.D. George Hoag, Ph.D.
>j Director, Remediation Technology Director
Groundwater Technology Environmental Research Institute
University of Connecticut
REVIEWERS
The panel that reviewed the monograph under the auspices of the Project
Steering Committee was composed of:
Peter W. Tunnicliffe, P.E., DEE Chair Richard L. Johnson, Ph.D.
Senior Vice President Oregon Graduate Institute of Science and
Camp Dresser & McKee Inc. Technology
Deptartment of Environmental Science
R. Ryan Dupont, Ph.D. and Engineering
Department of Civil and Environmental
Engineering Brian B. Looney, Ph.D.
Utah State University Savannah River Laboratory
John Eisenbeis, Ph.D. Michael C. Marley
Camp Dresser & McKee Inc. ENVIROGEN
Chi-Yuan Fan, P.E., DEE Donald Mohr
U.S. Environmental Protection Agency Chevron Research & Technology
Company
-------
r
k-
STEERING COMMITTEE
Frederick G. Pohland, Ph.D., P.E., DEE
Chair
Weidlein Professor of Environmental
Engineering
University of Pittsburgh
Domy Adriano, Ph.D.
Professor and Head
Biogeochemical Ecology Division
The University of Georgia
Representing, Soil Science Society of
America
William C. Anderson, P.E., DEE
Project Manager
Executive Director
American Academy of Environmental
Engineers
Colonel Frederick Boecher
Director of Risk Management
Chemical and Biological Defense
Command
U.S. Army
Representing, American Society of Civil
Engineers
Paul L. Busch, Ph.D., P.E., DEE
President and CEO
Malcolm Pirnie, Inc.
Representing, American Academy of
Environmental Engineers
Richard A. Conway, P.E., DEE
Senior Corporate Fellow
Union Carbide Corporation
Chair, Environmental Engineering
Committee
EPA Science Advisory Board
George Coyle
Division of Technical Innovation
Office of Technical Integration
Environmental Education Development
U.S. Department of Energy
Timothy B. Holbrook, P.E.
Engineering Manager
Groundwater Technology, Inc.
Representing, Air & Waste Management
Association
Walter W. Kovalick, Jr., Ph.D.
Director, Technology Innovation Office
Office of Solid Waste and Emergency
Response
U.S. Environmental Protection Agency
Joseph F. Lagnese, Jr., P.E., DEE
Private Consultant
Representing, Water Environment
Federation
Peter B. Lederman, Ph.D., P.E., DEE, P.P.
Center for Env. Engineering & Science
New Jersey Institute of Technology
Representing, American Institute of
Chemical Engineers
Raymond C. Loehr, Ph.D., P.E., DEE
H.M. Alharthy Centennial Chair and
Professor
Civil Engineering Department
University of Texas
Timothy Oppelt
Director, Risk Reduction Engineering
Laboratory
U.S. Environmental Protection Agency
David Patterson
Senior Technical Analyst
Waste Policy Institute
Representing, U.S. Department of Defense
George Pierce, Ph.D.
Editor-in-Chief
Journal of Microbiology
Manager, Bioremediation Technology
Development
Cytec Industries
Representing, Society of Industrial
Microbiology
Peter W. Tunnicliffe, P.E., DEE
Senior Vice President
Camp Dresser & McKee, Incorporated
Representing, Hazardous Waste Action
Coalition
Charles O. Velzy, P.E., DEE
Private Consultant
Representing, American Society of
Mechanical Engineers
Walter J. Weber, Jr., Ph.D., P.E., DEE
Earnest Boyce Distinguished Professor
University of Michigan
Representing, Hazardous Waste Research
Center
IV
-------
REVIEWING ORGANIZATIONS
The following organizations contributed to the monograph's review and
acceptance by the professional community. The review process employed by each
organization is described in its acceptance statement. Individual reviewers are, or
are not, listed according to the instructions of each organization.
Air & Waste Management
Association
The Air & Waste Management
Association is a nonprofit technical and
educational organization with more than
14,000 members in more than fifty
countries. Founded in 1907, the
Association provides a neutral forum
where all viewpoints of an environmen-
tal management issue (technical,
scientific, economic, social, political,
and public health) receive equal
consideration.
This worldwide network represents
many disciplines: physical and social
sciences, health and medicine, engineer-
ing, law, and management. The
Association serves its membership by
promoting environmental responsibility
and providing technical and managerial
leadership in the fields of air and waste
management. Dedication to these
objectives enables the Association to
work towards its goal: a cleaner
environment.
Qualified reviewers were recruited
from the Waste Group of the Technical
Council. It was determined that the
monograph is technically sound and
publication is endorsed.
The lead reviewer was:
Paul Lear
OH Materials, Inc.
American Society of Civil
Engineers
Qualified reviewers were recruited
from the Environmental Engineering
Division of ASCE and formed a Sub-
committee on WASTECH®. The mem-
bers of the Subcommittee have
reviewed the monograph and have
determined that it is acceptable for
publication.
Hazardous Waste Action
Coalition
The Hazardous Waste Action Coali-
tion (HWAC) is an association dedi-
cated to promoting an understanding of
the state of the hazardous waste practice
and related business issues. Our mem-
ber firms are engineering and science
firms that employ nearly 75,000 of this
country's engineers, scientists, geolo-
gists, hydrogeologists, toxicologists,
chemists, biologists, and others who
solve hazardous waste problems as a
professional service. HWAC is pleased
to endorse the monograph as technically
sound.
The lead reviewer was:
James D. Knauss, Ph.D.
Shield Environmental
Associates, Inc.
-------
r
*
Water Environment
Federation
The Water Environment Federa-
tion is a nonprofit educational orga-
nization composed of member and
affiliated associations throughout the
world. Since 1928, the Federation
has represented water quality spe-
cialists including engineers, scien-
tists, government officials, industrial
and municipal treatment plant opera-
tors, chemists, students, academic
and equipment manufacturers, and
distributors.
Qualified reviewers were re-
cruited from the Federation's Hazard-
ous Wastes, Industrial Wastes, and
Groundwater Committees. A list of
their names, titles, and business
affiliations can be found below.
It has been determined that the
document is technically sound and
publication is endorsed.
The reviewers were:
Terry E. Baxter
Assistant Professor
Civil & Environmental Engineering
Department
Northern Arizona University
S. Bijoy Ghosh
Principal Engineer
Engineering Science, Inc.
Jim Hartley
CH2M Hill
William Holt
Marion Environmental, Inc.
Byung R. Kim
Principal Staff Engineer
Ford Research Laboratory
Paul D. Kuhlmeier
Southern Pacific Lines
Kenneth S. Stoller
Vice President
Sadat Associates, Inc.
George M. Wong-Chong*
Manager, Remediation Processes
ICF Kaiser Engineers, Inc.
*WEF lead reviewer
Society for Industrial
Microbiology
The Society for Industrial Microbiol-
ogy (SIM) is a nonprofit professional
association dedicated to the advance-
ment of microbiological sciences,
especially as they apply to industrial
products, biotechnology, materials, and
processes. Founded in 1949, SIM
promotes the exchange of scientific
information through its meetings and
publications, and serves as liaison
among the specialized fields of microbi-
ology. Membership in the Society is
extended to all scientists in the general
field of microbiology. Corporate
membership is available to industrial
companies interested in the aims of the
Society.
It has been determined that the
document is technically sound and
publication is endorsed.
vi
-------
ACKNOWLEDGMENTS
The WASTECH® project was conducted under a cooperative agreement
between the American Academy of Environmental Engineers® and the Office
of Solid Waste and Emergency Response, U.S. Environmental Protection
Agency. The substantial assistance of the staff of the Technology Innovation
Office was invaluable.
Financial support was provided by the U.S. Environmental Protection
Agency, Department of Defense, Department of Energy, and the American
Academy of Environmental Engineers®.
This multiorganization effort involving a large number of diverse profes-
sionals and substantial effort in coordinating meetings, facilitating communica-
tions, and editing and preparing multiple drafts was made possible by a
dedicated staff provided by the American Academy of Environmental Engi-
neers® consisting of:
Paul F. Peters
Assistant Project Manager & Managing Editor
Karen M. Tiemens
Editor
Susan C. Zarriello
Production Manager
J. Sammi Olmo
Project Administrative Manager
Yolanda Y. Moulden
Staff Assistant
I. Patricia Violette
Staff Assistant
vii
-------
viii
-------
TABLE OF CONTENTS
CONTRIBUTORS III
ACKNOWLEDGMENTS vii
LIST OF TABLES xv
LIST OF FIGURES xvi
1.0 INTRODUCTION 1.1
1.1 Vacuum Vapor Extraction 1.1
1.2 Development of the Monograph 1.2
1.2.1 Background 1.2
1.2.2 Process 1.3
1.3 Purpose 1.4
1.4 Objectives 1.4
1.5 Scope 1.4
1.6 Limitations 1.5
1.7 Organization 1.6
2.0 PROCESS SUMMARY 2.1
2.1 A History of Soil Vapor Extraction 2.1
2.1.1 Soil Vapor Extraction 2.3
2.1.2 Air Sparging 2.4
2.1.3 Bioventing 2.5
2.1.4 Thermal Enhancements of Soil Vapor Extraction 2.6
2.2 Fundamentals and Basic Phenomena 2.7
ix
-------
Table of Contents
2.3 System Design 2.8 ~
2.4 Costs 2.9*
2.5 Performance Monitoring 2.9
2.6 Potential Applications 2.9
2.7 Process Evaluation 2.10
2.8 Prognosis 2.11
3.0 PROCESS IDENTIFICATION AND DESCRIPTION 3.1
3.1 Description of the Technologies 3.2
3.1.1 Soil Vapor Extraction 3.2
3.1.2 Air Sparging 3.8
3.1.3 Bioventing 3.12
3.1.4 Thermal Enhancements of Soil Vapor Extraction 3.14
3.2 Fundamentals and Basic Phenomena 3.16
3.2.1 Qualitative Description of the Subsurface Distribution
of an Immiscible Liquid 3.17
3.2.2 Quantitative Description of Multiphase Partitioning 3.19
3.2.3 Basis for Mathematical Models of Induced Air Flow
in the Unsaturated Zone 3.28
3.2.4 One-Dimensional Vapor Flow Scenarios 3.30
3.2.5 Two-Dimensional Vapor Flow Scenarios 3.36
3.2.6 General Air-Flow Modeling Considerations 3.38
3.2.7 Transport Considerations and the Estimation of
Removal Rates and Residuals 3.39
3.2.8 Nondimensional Solutions - Constant Flow/Constant
Vapor Concentration 3.42
3.2.9 One-Dimensional Models - Equilibrium-Based
Well-Mixed Systems 3.43
3.2.10 Two-Dimensional Model - Cleanup Along
Streamlines 3.45
-------
Table of Contents
3.2.11 Mass-Transfer Limitations, Nonideal Scenarios,
and Transient Effects 3.46
3.2.12 Microbiological Processes (Bioventing) 3.50
3.2.13 Air Injection Below the Groundwater Table 3.52
3.3 Characterization Activities 3.58
3.3.1 Contaminant Assessment 3.59
3.3.2 Geologic/Hydrogeologic Assessment 3.66
3.3.3 Laboratory Soil Column Tests 3.67
3.3.4 Field Pilot-Scale Activities 3.72
3.3.4.1 Vapor Flow vs. Applied Vacuum/Pressure
Test 3.76
3.3.4.2 Extracted Vapor Characterization vs. Time 3.77
3.3.4.3 Subsurface Vapor-Phase Pressure Distribution 3.80
3.3.4.4 Subsurface Vapor Concentration Distribution
and In Situ Respirometry 3.83
3.3.4.5 Groundwater Elevation Changes 3.84
3.3.4.6 Groundwater Monitoring 3.85
3.3.4.7 Tracer-Gas Tests 3.87
3.4 System Design 3.88
3.4.1 Extraction/Injection Well Construction 3.91
3.4.2 Vapor Treatment 3.93
3.4.3 Extraction Pump/Blower and Injection Blower 3.95
3.4.4 Instrumentation 3.96
3.4.5 Manifolds 3.96
3.4.6 Surface Seals and Passive Inlet Wells 3.97
3.4.7 Design Approaches 3.98
3.4.7.1 Empirical Approach 3.98
3.4.7.2 Matching System Design to Existing
Equipment 3.99
xi
-------
Table of Contents
3.4.7.3 Radius-of-Influence Approaches 3.99
3.4.7.4 Screening-Level Model Approaches 3.101"
3.4.7.5 Detailed Numerical Modeling and
Optimization-Based Approaches 3.104
3.4.8 Comparison of Various Design Approaches 3.104
3.4.9 Other Design Considerations 3.109
3.5 Costs 3.109
3.6 Performance Monitoring 3.110
3.6.1 Primary Process Variables (Vapor Flow Rates,
Pressure, Extracted Gas Characterization) 3.110
3.6.2 Respiratory Gas Monitoring 3.118
3.6.3 Cost 3.121
3.6.4 Environmental Factors 3.122
3.6.5 Subsurface Monitoring 3.124
3.6.5.1 In Situ Soil-Gas Monitoring 3.124
3.6.5.2 Subsurface Vapor Phase Pressure
Distribution 3.128
3.6.5.3 Subsurface Respiratory Gas Monitoring 3.128
3.6.5.4 Soil Borings and Site Sampling 3.130
3.6.5.5 Groundwater Sampling 3.131
3.6.5.6 Water Table Fluctuations 3.131
3.6.5.7 Tracer Gas Tests 3.133
3.7 Technology Variations: Combined In Situ Soil Heating
and Vapor Extraction 3.133
3.7.1 Steam Stripping 3.137
3.7.1.1 Steam Stripping Technology 3.138
3.7.1.2 Advantages and Disadvantages of Steam
Stripping 3.141
3.7.2 Radio Frequency Heating 3.142
xii
-------
Table of Contents
3.7.2.1 Status of RF Heating Technology 3.143
3.7.2.2 Advantages and Disadvantages of RF
Heating 3.143
3.7.3 Joule Resistance Heating 3.144
3.7.3.1 Status of the Joule Resistance Heating
Technology 3.145
3.7.3.2 Advantages and Disadvantages of Joule
Resistance Heating 3.145
3.7.4 Conductive Heating 3.145
3.7.5 Modifying Soil Surface 3.146
3.7.5.1 Status of Soil Surface Modification
Technology 3.147
3.7.5.2 Advantages and Disadvantages of Soil
Surface Modification 3.148
3.7.6 Hot Air or Hot Gas Injection 3.149
3.7.7 Fiber Optic Heating 3.150
3.7.8 Warm Water Injection 3.151
3.8 Summary of Good Practices 3.151
3.8.1 Site Characterization 3.152
3.8.2 Defining Remedial Objectives 3.154
3.8.3 Screening Treatment Alternatives 3.154
3.8.4 Field Tests 3.155
3.8.5 System Design 3.155
3.8.6 Operation and Monitoring 3.155
3.8.7 Engineering Analysis 3.156
4.0 POTENTIAL APPLICATIONS 4.1
4.1 Review of Reported Applications 4.1
4.2 Review of Recommendations 4.2
4.3 Quantifying Applicability 4.3
XIII
-------
Table of Contents
5.0 PROCESS EVALUATION 5.1
6.0 LIMITATIONS 6.1 *
7.0 TECHNOLOGY PROGNOSIS 7.1
7.1 Soil Vapor Extraction 7.1
7.2 Air Sparging 7.3
7.3 Bioventing 7.4
Appendices
A. List of References A.I
B. Suggested Reading List B.I
xiv
-------
LIST OF TABLES
Table Title Page
3.1 Thermodynamic data for selected hydrocarbons (T = 20°C) 3.25
3.2 Gasoline hydrocarbons grouped in constituent classes 3.27
3.3 Site characterization activities 3.59
3.4 Methods for hydrocarbon delineation 3.64
3.5 Soil vapor extraction-based processes design approaches 3.90
3.6 Results of example design problem 3.107
3.7 Process monitoring options and data interpretation 3.111
3.8 Vadose zone monitoring well vapor concentrations and
compositions measured during a vapor extraction
application 3.126
3.9 Thermal properties of selected materials 3.136
6.1 Limitations of vapor extraction-based processes 6.2
xv
-------
List of Figures
LIST OF FIGURES
Figure Title Page
2.1 Groundwater and free-product elevations prior to and
during a vapor extraction application 2.2
3.1 Simplified soil vapor extraction schematics: (a) in situ 3.3
(b) ex situ 3.4
3.2 Typical soil vapor extraction performance curves 3.5
3.3 Effect of gas-phase diffusion limitations on performance 3.7
3.4 Simplified air sparging/vapor extraction schematic 3.9
3.5 Extracted vapor concentrations for a soil venting/air
sparging application 3.10
3.6 Comparison of removal rates attributed to volatilization
and biodegradation during soil venting at the Tyndall
AFB site 3.14
3.7 Simplistic bioventing process schematics 3.15
3.8 Illustration of multi-phase hydrocarbon distribution
resulting from an underground storage tank leak 3.20
3.9 Cross-section illustration of concentration of total volatile
hydrocarbons detected in soil gas at USGS study site near
Bemidji, MN 3.21
3.10 Areal distribution of (a) oxygen and (b) carbon dioxide as
a percentage of the soil gas at a depth of 6 ft BGS at the
Galloway Township, NJ, gasoline-spill site, December
1989 3.22
3.11 Comparison of vapor concentration prediction models 3.26
3.12 One-dimensional vapor flow scenarios: (a) linear flow,
(b) radial flow 3.32
3.13 Predicted vapor flowlines to a vapor extraction trench 3.33
XVI
-------
List of Figures
Figure Title Page
3.14 Predicted and measured steady-state radial pressure
distributions 3.35
3.15 Predicted steady-state flowrates (per unit well screen
depth) for a range of soil permeabilities and applied
vacuums (Pw) 3.37
3.16 Comparison of total hydrocarbon fluxes from steady-flow
venting experiments to predictions obtained from the
mathematical model 3.44
3.17 Limiting model scenarios for removal rate estimates 3.47
3.18 Estimated maximum removal rates and residual hydrocarbon
reduction for a venting operation limited by diffusion 3.50
3.19 Biodegradation of JP-4 jet fuel in soils collected near
Fairbanks, Alaska 3.53
3.20 Biodegradation of JP-4 jet fuel at 20°C in various soils
collected near Fairbanks, Alaska; Fallon, Nevada; and
Panama City, Florida 3.55
3.21 Observed air channel pattern in uniform mixture of 0.75 and
0.3 mm glass beads 3.56
3.22 Observed air channel pattern in stratified medium 3.57
3.23 Recommended presentation of total hydrocarbon distribution
and subsurface geology 3.60
3.24 Boiling point distribution curves for samples of "fresh" and
"weathered" gasolines 3.62
3.25 Basic laboratory soil column treatability test schematic 3.68
3.26 Hypothetical laboratory soil column test results 3.73
xvii
-------
List of Figures
Figure Title Page
3.27 Simplified field pilot test schematic for vapor extraction-
based technologies 3.75
3.28 Presentation of (a) extraction and (b) injection test data 3.77
3.29 Schematic of sampling apparatus used when sampling
vapors under vacuum conditions 3.79
3.30 Presentation of extracted vapor analyses from pilot test 3.80
3.31 Vadose zone monitoring installation 3.81
3.32 Presentation of subsurface pressure monitoring results from
pilot test a) transient results, b) steady-state results 3.82
3.33 Oxygen utilization and carbon dioxide production in various
phases of a bioventing project at Tyndall AFB, Florida 3.84
3.34 Groundwater elevation monitoring approaches: a) direct
measurement b) indirect measurement 3.85
3.35 Measured water table upwelling 3.86
3.36 Groundwater monitoring results during application of air
sparging 3.87
3.37 Presentation of tracer test results 3.88
3.38 Schematic of standard extraction/injection wells 3.92
3.39 Flowchart for radius-of-influence design approach 3.100
3.40 Sample worksheet for a screening-level model design
approach 3.103
3.41 Total hydrocarbon distribution and subsurface geology for
sample design problem 3.105
3.42 Summary of design data for sample design problem 3.106
3.43 Presentation of "minimum" data collection needs for vapor
extraction systems 3.113
xviii
-------
List of Figures
Page
3.44 Generalized concentration or mass of contaminant observed
in gas extracted from a vacuum extraction system 3.116
3.45 Tetrachloroethylene (PCE) concentration in extracted gas
from a site near Milan, Italy 3.117
3.46 Oxygen and carbon dioxide concentrations in the venting of
gas versus time at the Hill AFB, Utah, soil venting site, from
December 18, 1988 to April 1, 1989 3.119
3.47 Cumulative hydrocarbon removal (volatilized and bio-
degraded) at Hill AFB, Utah, soil venting site (from
December 18, 1988 to November 14, 1990) 3.123
3.48 Schematic of subsurface flow and relation between extracted
gas concentration and in situ soil gas concentrations in
stratified flow system 3.125
3.49 Typical vapor extraction monitoring point construction detail 3.127
3.50 Oxygen utilization and carbon dioxide production in various
phases of a bioventing project at Tyndall AFB, Florida 3.129
3.51 (a) Vapor flowrate and groundwater upwelling dependence
on applied vacuum, (b) transient water table and subsurface
vacuum response at a vapor extraction application 3.132
3.52 Vapor pressure for various organic compounds as a function
of temperature 3.135
3.53 In situ soil venting system design process 3.153
4.1 Soil vapor extraction applicability nomograph 4.3
4.2 Applicability decision process 4.5
4.3 Results of applicability screening analysis 4.8
xix
-------
-------
Chapter 1
INTRODUCTION
This monograph on vacuum vapor extraction is one of a series of eight
on innovative site and waste remediation technologies that are the culmina-
tion of a multiorganization effort involving more than 100 experts over a
two-year period. It provides the experienced, practicing professional guid-
ance on the application of innovative processes considered ready for full-
scale application. Other monographs in this series address bioremediation,
chemical treatment, solvent/chemical extraction, soil washing/soil flushing,
stabilization/solidification, thermal desorption, and thermal destruction.
7.7 Vacuum Vapor Extraction
A basic in situ vapor (or soil vapor) extraction system couples vapor
extraction wells with blowers or vacuum pumps to remove contaminant
vapors from zones permeable to vapor flow, thereby enhancing the volatil-
ization and removal of contaminants from the subsurface for treatment, as
appropriate. The vacuum developed in the extraction well boring results in
air being drawn from the atmosphere through the soil to the well. More
complex soil vapor extraction systems incorporate trenches, horizontal
wells, forced-air injection wells, passive air inlet wells, groundwater recov-
ery systems, impermeable surface seals, multiple vapor extraction wells in
single boreholes, and various thermal enhancements. Soil vapor extraction
can also be practiced ex situ with excavated soils; perforated pipes are
placed within soil piles to produce air movement through the pile.
In addition to vapor extraction, the monograph addresses complementary
technologies - air sparging (injecting air under pressure below the water
table) and bioventing (applying air in the course of vapor extraction into the
subsurface to accelerate aerobic biodegradation).
1.1
-------
Introduction
1.2 Development of the Monograph
1.2.1 Background
Acting upon its commitment to develop innovative treatment technolo-
gies for the remediation of hazardous waste sites and contaminated soils
and groundwater, the U.S. Environmental Protection Agency (US EPA)
established the Technology Innovation Office (TIO) in the Office of Solid
Waste and Emergency Response in March, 1990. The mission assigned TIO
was to foster greater use of innovative technologies.
In October of that same year, TIO, in conjunction with the National Ad-
visory Council on Environmental Policy and Technology, convened a work-
shop for representatives of consulting engineering firms, professional soci-
eties, research organizations, and state agencies involved in site
remediation. The workshop focused on defining the barriers that were im-
peding the application of innovative technologies in site remediation
projects. One of the major impediments identified was the lack of reliable
data on the performance, design parameters, and costs of innovative pro-
cesses.
The need for reliable information led TIO to approach the American
Academy of Environmental Engineers®. The Academy is a long-standing,
multidisciplinary environmental engineering professional society with
wide-ranging affiliations with the remediation and waste treatment profes-
sional communities. By June 1991, an agreement in principle (later formal-
ized as a Cooperative Agreement) was reached. The Academy would man-
age a project to develop monographs describing the state of available inno-
vative remediation technologies. Financial support would be provided by
the EPA, U.S. Department of Defense (DOD), U.S. Department of Energy
(DOE), and the Academy. The goal of both TIO and the Academy was to
develop monographs providing reliable data that would be broadly recog-
nized and accepted by the professional community, thereby, eliminating or,
at least, minimizing this impediment to the use of innovative technologies.
The Academy's strategy for achieving the goal was founded on a
multiorganization effort, WASTECH®(pronounced Waste Tech), which
joined in partnership the Air and Waste Management Association, the
American Institute of Chemical Engineers, the American Society of Civil
Engineers, the American Society of Mechanical Engineers, the Hazardous
1.2
-------
Chapter 1
Waste Action Coalition, the Society for Industrial Microbiology, and the
Water Environment Federation, together with the Academy, EPA, DOD,
and DOE. A Steering Committee composed of highly respected representa-
tives of these organizations having expertise in remediation technology
formulated the specific project objectives and process for developing the
monographs (see page iv for a listing of Steering Committee members).
By the end of 1991, the Steering Committee had organized the Project.
Preparation of the monograph began in earnest in January, 1992.
1.2.2 Process
The Steering Committee decided upon the technologies, or technological
areas, to be covered by each monograph, the monographs' general scope,
and the process for their development and appointed a task group composed
of five or more experts to write a manuscript for each monograph. The task
groups were appointed with a view to balancing the interests of the groups
principally concerned with the application of innovative site and waste
remediation technologies — industry, consulting engineers, research, aca-
deme, and government (see page iii for a listing of members of the Vacuum
Vapor Extraction Task Group).
The Steering Committee called upon the task groups to examine and
analyze all pertinent information available, within the Project's financial
and time constraints. This included, but was not limited to, the comprehen-
sive data on remediation technologies compiled by EPA, the store of infor-
mation possessed by the task groups' members, that of other experts willing
to voluntarily contribute their knowledge, and information supplied by pro-
cess vendors.
To develop broad, consensus-based monographs, the Steering Committee
prescribed a twofold peer review of the first drafts. One review was con-
ducted by the Steering Committee itself, employing panels consisting of
two members of the Committee supplemented by at least four other experts
(see Reviewers, page iii, for the panel that reviewed this monograph). Si-
multaneous with the Steering Committee's review, each of the professional
and technical organizations represented in the Project reviewed those mono-
graphs addressing technologies in which it has substantial interest and com-
petence. Aided by a Symposium sponsored by the Academy in October
1992, persons having interest in the technologies were encouraged to par-
ticipate in the organizations' review.
1.3
-------
Introduction
Comments resulting from both reviews were considered by the Task
Group, appropriate adjustments were made, and a second draft published.
The second draft was accepted by the Steering committee and participating
organizations. The statements of the organizations that formally reviewed
this monograph are presented under Reviewing Organizations on page v.
1.3 Purpose
The purpose of this monograph is to further the use of innovative site
remediation technologies, i.e., technologies not commonly applied, where
their use can provide better, more cost-effective performance than conven-
tional methods. To this end, the monograph documents the current state of
vacuum vapor extraction and of the complementary technologies, air
sparging and bioventing, as they relate to it.
1.4 Objectives
The monograph's principal objective is to furnish guidance for experi-
enced, practicing professionals and users' project managers. The mono-
graph is intended, therefore, not to be prescriptive, but supportive. It is in-
tended to aid experienced professionals in applying their judgment in decid-
ing whether and how to apply the technologies addressed under the particu-
lar circumstances confronted.
In addition, the monograph is intended to inform regulatory agency per-
sonnel and the public about the conditions under which the processes it
addresses are potentially applicable.
7.5 Scope
The monograph addresses vacuum vapor extraction and other comple-
mentary innovative technologies. It addresses all aspects of the technolo-
1.4
-------
Chapter 1
gies for which sufficient relevant data was available to the Vacuum Vapor
Extraction Task Group to describe and explain the technologies and assess
their effectiveness, limitations, and potential applications. Laboratory- and
pilot-scale technologies were addressed, as appropriate.
Application of site remediation and waste treatment technology is site
specific and involves consideration of a number of matters besides alterna-
tive technologies. Among them are the following that are addressed only to
the extent essential to understand the applications and limitations of the
technologies described:
• site investigations and assessments;
• planning, management, specifications, and procurement; and
• regulatory requirements.
1.6 Limitations
The information presented in this monograph has been prepared in accor-
dance with generally recognized engineering principles and practices and is
for general information only. This information should not be used without
first securing competent advice with respect to its suitability for any general
or specific application.
Readers are cautioned that the information presented is that which was
generally available during the period when the monograph was prepared.
Development of innovative site remediation and waste treatment technolo-
gies is ongoing. Accordingly, postpublication information may amplify,
alter, or render obsolete the information about the processes addressed.
This monograph is not intended to be and should not be construed as a
standard of any of the organizations associated with the WASTECH®
Project; nor does reference in this publication to any specific method, prod-
uct, process, or service constitute or imply an endorsement, recommenda-
tion, or warranty thereof.
1.5
-------
Introduction
7.7 Organization
This monograph and others in the series are organized under a uniform
outline intended to facilitate cross reference among them and comparison
of the technologies they address. Chapter 2.0, Process Summary, provides
an overview of all material presented. Chapter 3.0, Process Identification,
provides comprehensive information on the process addressed and fully
analyzes it. The analysis includes, to the extent information and data are
available, a description of the process (what it does and how it does it), its
scientific basis, status of development, environmental effects, pre- and post-
treatment requirements, health and safety considerations, design data, op-
erational considerations, and comparative cost data. Also addressed are
process-unique planning and management requirements and process varia-
tions. Chapter 3.0 and subsequent chapters focus upon the principal tech-
nology, vapor extraction, and address the complementary technologies, air
sparging and bioventing, where appropriate.
Chapter 4.0, Potential Applications, Chapter 5.0, Process Evaluation, and
Chapter 6.0, Limitations, provide syntheses of available information and
informed judgments on the process. Each of these chapters addresses the
process in the same order as it is described in Chapter 3.0. Technology
Prognosis, Chapter 7.0, identifies aspects of the process needing further
research and demonstration before full-scale application can be considered.
1.6
-------
Chapter 2
2
PROCESS SUMMARY
2.7 A History of Soil Vapor Extraction
Although articles relating to soil vapor extraction began to appear in
journals and conference proceedings in the early 1980s, soil vapor extrac-
tion is no different in principle from the vapor abatement processes prac-
ticed much earlier for the control of volatile organic compound (VOC)
migration into buildings. The majority of early articles are anecdotal re-
ports of field applications, many of which have been summarized by
Hutzler, Murphy, and Gierke (1989). Many of the sites treated were gaso-
line or solvent spills in relatively pervious sands. The process was nearly
always reported to be successful, but appropriate monitoring and system
controls were frequently absent and final cleanup levels were rarely re-
ported, if determined.
The transition from the use of soil vapor extraction as a vapor abatement
technology to the process now used to remove nonaqueous phase liquids
(NAPL's) is clearly documented in the literature through the contributions
of the Texas Research Institute (1980) and Thornton and Wootan (1982).
They introduced the concept of vertical vapor extraction and injection wells
for the removal of residual hydrocarbon (gasoline product, as well as conse-
quential gasoline compounds in the gas phase) and the use of vapor probes
for sampling and qualitative and quantitative analysis of the diffused gaso-
line hydrocarbon vapors. A further enhancement of this research entailed
hypothesizing of various venting geometries and subsequent air-flow paths
and their testing in a pilot-scale soil tank (Texas Research Institute 1984).
In the Institute's first study (1980), 50% removal of gasoline was achieved
by means of soil vapor extraction. In its second study (1984), 84% of the
2.1
-------
Process Summary
gasoline product was removed; however, it was believed that up to 30% of
the gasoline may have been removed through stimulated biological activity.
Encouraged by results of the Texas Research Institute's research, Marley
and Hoag (1984) and Marley (1985) conducted laboratory soil column ex-
periments and demonstrated that all of the gasoline at residual saturation
could be removed using soil vapor extraction sands in the range from 0.224
mm to 2.189 mm (0.009 to 0.086 in.) average diameter at air-flow rates
from 16.1 cm3/cm2-min to 112.5 cmVcm2-min (0.53 to 3.7 ftW-min).
This was followed by the field study of Hoag and Cliff (1988), who re-
ported that an in situ soil vapor extraction system was able to remove
1,330 L (350 gal) of gasoline at residual saturation and at the capillary
fringe in 100 days. Cleanup levels of below 3 ppmv were reported using
GC/FID analysis of soil-gas samples. Groundwater elevation and product
thickness were logged for a substantial period before and after the applica-
tion of soil vapor extraction. Figure 2.1 is a well log of one of the monitor-
ing wells at the site. One-to-two feet of gasoline product persisted in the
Figure 2.1
Groundwater and Free-Product Elevations Prior to and
During a Vapor Extraction Application
93.0-
I 89.(H
S 870-
c/l
I 85.0H
g 83.0-
81.0-
"J
79.0-
Product
Thickness
Product Elevation
40
80 120 160 200 240 260 320
* DEC -»K JAN -»|«.FEB*|<-MAR->)<-APR-«+«-MAy-»|«-JUNB*j*JULY*H-AUO*)*- SEP -«+«- OCT-H
Time (Days)
Source- Hoag and Cliff 1988
2.2
-------
Chapter 2
well for a period of approximately 210 days. By day 250 (40 days of soil
vapor extraction), the product thickness was not measurable and only a
skim was detected on the water table. After day 290 (80 days of soil vapor
extraction), no skim of gasoline was detected. Thus, soil vapor extraction
was effective in removing both gasoline at residual saturation and as free
product on the water table.
In one of the first attempts to model the fundamental processes govern-
ing soil vapor extraction performance, Marley and Hoag (1984), using
Raoult's Law to predict the concentration and composition of extracted air
and 52 compounds in gasoline as an approximate composition, developed
an equilibrium-based model describing residual hydrocarbon-gas phase
interaction during the soil vapor extraction column tests. This work was
further advanced by Baehr and Hoag (1988) who developed a coupled one-
dimensional compressible air flow, three-phase local-equilibrium based
model to describe the results of the Marley (1985) soil vapor extraction
column tests. These equilibrium-based modeling approaches have also
been used successfully by Johnson, Kemblowski, and Colthart (1988),
Johnson, Kemblowski, and Colthart (1990), and Johnson et al. (1991).
Other screening level approaches have been used to assess vapor-flow dy-
namics and mass transport considerations (Massmann 1989; Johnson,
Kemblowski, and Colthart 1988; Johnson et al. 1990, 1991; Shan, Falta, and
Javandel 1992). More sophisticated levels of modeling have been recently
advanced by others; these are discussed in Section 3.2.
2.1.1 Soil Vapor Extraction
A typical in situ soil vapor extraction system couples vapor extraction
wells with blowers or vacuum pumps to remove contaminant vapors from
zones permeable to vapor flow, thereby enhancing the volatilization and
efficiency of removal of contaminants from the subsurface. The compo-
nents of soil vapor extraction systems are usually available off-the-shelf.
Qualified engineering firms can install the necessary wells and trenches.
Aboveground equipment usually consists of:
• vacuum blower and controls;
• control valves to adjust air flow;
• pressure gauges and flow meters at wellheads;
2.3
-------
Process Summary
• air-liquid separator (for removing moisture from the extracted
gases);
• pressure gauge and flow meter at the pump; and
• vapor treatment unit.
If air treatment is required, vapor treatment systems, such as catalytic or
thermal destruction systems, activated carbon adsorbers, or biological gas
treatment systems are employed.
Soil vapor extraction can be employed ex situ for treating excavated
soils. In this scenario, perforated pipes are placed within soil piles to draw
air through the pile. The equipment is the same as that used for in situ ap-
plications.
Although the equipment and process are relatively simple, the design,
operation, and monitoring of soil vapor extraction systems are not. The
central questions of cost-effectiveness and practicability may be quite com-
plex and subject to a high degree of uncertainty. Determining optimal de-
sign criteria - the number, location, spacing, and construction of extraction/
injection wells, vapor treatment systems, degree of operational control, and
requisite monitoring - is frequently difficult. In practice, soil vapor extrac-
tion is often limited by such factors as the equipment available, the
contaminant(s) of concern, site conditions (geology, location and construc-
tion of buildings, etc.) and regulatory cleanup criteria.
2.1.2 Air Sparging
Air sparging, a more recent innovation, may extend the application of
soil vapor extraction to water-saturated soils. The technology, which is still
emerging, has been applied at sites contaminated with chlorinated solvents,
gasoline, and aviation fuel.
Air sparging is accomplished by injecting air under pressure below the
water table. It is expected that contaminants located within air-flow path-
ways will volatilize or biodegrade and there is the potential that dissolved-
phase contaminants that contact the air-flow field could potentially be
volatilized or biodegraded.
Two theories have been advanced to explain the effect of air sparging:
• injected air strips contaminants from the soil and groundwater
into the vapor flow; and
2.4
-------
Chapter 2
• injected air increases the oxygen content of the groundwater,
resulting in increased aerobic biodegradation.
It is likely that both contribute to remediation at all sites. Improperly
controlled air sparging systems can pose significant health and safety risks.
The pressurized air can accelerate the uncontrolled migration of contami-
nant vapors and the consequent accumulation in buildings or other vapor
receptors. It has been suggested that there may also be the potential for
enhanced spreading of dissolved contaminant plumes as the injected air
initially displaces groundwater. In addition, it has been suggested that the
air injection may result in increased mixing and, therefore, increased mass
transfer of contaminants into groundwater. To minimize the risk of uncon-
trolled vapor or groundwater migration components, the following measures
should be considered for effective and safe operation:
• concurrent installation of a soil vapor extraction system to cap-
ture the entire volume of contaminant vapors; and
• containment of groundwater in the air injection zone to prevent
off-site migration of dissolved contaminants.
In addition to the health and safety risks, another concern is that air
sparging may lead to modified aquifer conditions such as aquifer plugging
because of iron precipitation stimulated by increased oxygen levels.
2.1.3 Bioventing
Numerous investigations have shown that many contaminants of concern
are biodegradable and indigenous microorganism populations can carry out
the remediation task if subsurface conditions are amenable. The microbes
utilize oxygen to break down hydrocarbons into carbon dioxide, water, and
biomass (cells) and, under natural conditions, aerobic biodegradation rates
are typically limited by oxygen supply rates in the subsurface. But in the
course of vapor extraction, air is drawn from the atmosphere into the sub-
surface, and accelerates the resupply of oxygen. Several researchers have
shown that this enhanced oxygen delivery will affect an increase in the rate
of aerobic biodegradation of contaminants.
Bioventing is still under development, but has been used to remediate
sites where various fuels and nonchlorinated compounds have been re-
leased. Various bioventing configurations have been proposed, two of
which are illustrated in figure 3.7 on page 3.15.
2.5
-------
Process Summary
By monitoring oxygen utilization and carbon dioxide production aerobic
biodegradation rates can be estimated. In bioventing, the object is to opti-
mize biodegradation rates. This often requires maintaining the oxygen
content of the soil at some set value (2 to 4% v/v, note: % v/v is % volume
to volume and is the equivalent of % by volume). Bioventing is also af-
fected by other site conditions. Low soil moisture content and low tempera-
ture can slow microbial degradation, while gentle heating of the soil to 30°
to 40°C (86° to 104°F) and increasing moisture content can enhance bio-
logical activity.
Engineered bioventing systems can offer capital equipment and operating
and maintenance cost reductions by utilizing lower air flows and eliminat-
ing the need for vapor treatment equipment.
In a bioventing application, lower airflow rates frequently reduce opera-
tion, monitoring, and maintenance costs on a daily basis, but they can in-
crease the time required for cleanup. The net impact on cost depends on
site-specific factors.
It should be noted that bioventing has recently evolved into a more
widely applied technology in its own right. This manuscript was prepared
initially in 1992, and at that time the decision was made to discuss
bioventing as it applies to soil vacuum extraction applications. The manu-
script subsequently has undergone review and limited modification, but the
emphasis has not changed. More detailed manuals specifically addressing
bioventing are in preparation.
2.1.4 Thermal Enhancements of Soil Vapor Extraction
It is generally accepted that soil vapor extraction becomes less effective
where compounds have vapor pressures less than 0.1 mm Hg to 1.0 mm Hg
at ambient temperatures. It is believed, however, that the range of applica-
bility can be extended by heating the subsurface because contaminant vapor
pressures increase with temperature. Also, biodegradation can be increased
by warming the subsurface to temperatures in the range of 30° to 40°C (86°
to 104°F).
There are a number of in situ heating processes under study. These in-
clude steam injection, hot air injection, electrical resistance heating, radio-
frequency heating, thermal conduction heating, warm water injection, and
solar heating. Each combines conventional vapor extraction equipment
2.6
-------
Chapter 2
with a means to elevate the subsurface temperature. Most of these tech-
nologies are still under development, although steam injection has been
used extensively by the petroleum industry.
There are several limitations in applying thermal enhancement technolo-
gies. Generation of excessive soil temperatures inhibits biodegradation and
affects soil structure. Some thermal enhancement processes entail large
capital equipment and energy costs.
Pertinent to steam injection, contaminants in the soil may vaporize and
become dissolved in the condensate front or be displaced. Heterogeneities
in the geological formations and in the contamination can decrease the
process's effectiveness. The process must be properly controlled in order to
minimize possible detrimental effects, such as contaminant "smearing" or
enhanced vapor transport away from the source area.
2.2 Fundamentals and Basic Phenomena
The basic phenomena governing the performance of soil vapor extraction
systems are easily described. Applying a vacuum to an extraction well
creates an air-flow field that originates at the ground surface and proceeds
along the path of least resistance (i.e., highest permeability to air flow) to
the screened interval of the extraction well. Contaminants volatilize within
the air-flow field, are swept into the vapor extraction well, and removed
from the soil. Once these vapors are removed, additional contaminants
outside the flow field volatilize, diffuse to the air-flow field, and are re-
moved from the soil. Ideally, residual contaminant volatilization and re-
moval occur continuously until all of the residual contaminant in the
unsaturated zone and above the capillary fringe is removed from the soil.
The performance of all vapor extraction-based processes can be related
to three main factors:
• equilibrium partitioning into the vapor space of a porous me-
dium;
• vapor flow characteristics of the porous medium; and
• mass transfer (kinetic) considerations and limitations (e.g., pro-
cesses that prevent equilibrium from being achieved).
2.7
-------
Process Summary
2.3 System Design
Competent characterization of the site and contaminants present is neces-
sary in order to assess the feasibility of vapor extraction-based technologies
and design effective remedial systems. See table 3.3 on page 3.57 for an
outline of essential site characterization activities.
System designs and matching system components with remedial objec-
tives is only the first step in successful application of this technology. Be-
cause of compounding uncertainties and inherent limitations arising from
natural heterogeneities, site characterization data, and predictive capabili-
ties, the design and optimization of vapor extraction-based processes con-
tinues even after the initial system is installed and turned on. As with all
engineered systems, system performance must be monitored correctly, the
results must be interpreted, and system modifications must be made accord-
ingly. Therefore, vapor extraction-based systems should be flexible in de-
sign so that they can handle a wide range of modifications and operating
conditions.
Along with the operating conditions, the following components are typi-
cally specified in a soil vapor extraction-based system design:
• number of vapor extraction wells;
• number of air injection wells (not at all sites);
• well location(s);
• well construction(s) (depth, screened interval, materials, etc.);
• extraction blower(s) or vacuum pump(s);
• injection blower(s) (not at all sites);
• vapor treatment unit(s);
• equipment manifolding & piping;
• instrumentation (flow meters, sampling ports, vapor concentra-
tion monitoring, etc.); and
• monitoring program/target remediation goals.
2.8
-------
Chapter 2
2.4 Costs
Like other in situ technologies, the cost of vapor extraction-based pro-
cesses is not easily quoted per volume of soil treated as is done for many ex
situ processes (incineration, thermal desorption, soil washing, etc.), as use-
ful as that would be, because of large variations in site geology, type of
contaminant, etc. Also, these technologies are very sensitive to depth to
contamination. For example, clean overburden soil must be excavated to
remove contaminated soil and this affects the total cost. However, fairly
educated estimates can be made for major capital equipment costs and these
have been set forth, where appropriate, in subsequent chapters. The major
part of the total process cost associated with vapor extraction-based pro-
cesses is usually the operating expenses for labor, maintenance, and moni-
toring.
2.5 Performance Monitoring
Performance data are monitored in order to assess performance, calibrate
models, and help guide needed system modifications resulting from changes
in operating conditions. There is a wide range of process monitoring op-
tions from which the practitioner can select depending upon the particular
need for data. See table 3.7 on page 3.109, listing options in order of im-
portance.
2.6 Potential Applications
Case studies and guidance documents indicate that vapor extraction-
based technologies are most likely to be successful at sites where volatile
compounds have impacted permeable soils. Typically, sites considered for
vapor extraction-based technologies are those where petroleum products
(e.g., gasoline and other fuels) or chlorinated solvents have spilled or leaked
into the subsurface. This guidance, however, is intended only to roughly
define the range of applicability, rather than to limit it. Remedial goals
2,9
-------
Process Summary
consist of target cleanup levels and an acceptable time frame for
remediation. The goals are based on considerations of risk and of the po-
tential for preserving reasonable beneficial use. Typical constraints include
total project cost, technical feasibility, permitting requirements, physical
boundaries, community-imposed limitations, and equipment availability.
Site-specific data, including those on geological conditions and contaminant
characteristics, are also considered, along with pilot-test data, if available.
2.7 Process Evaluation
For properly designed systems, removal rates are highest when the sys-
tem is first turned on. The rates then decline over time, and often reach
very low levels, often called "asymptotic". It should be noted that the re-
moval rates do not follow the strict definition of asymptotic, as they will
probably continue to decline very slowly. However, for practical engineer-
ing purposes, we can think of them as asymptotic.
The practicable degree of remediation that can be achieved with vapor
extraction-based technologies is limited by site characteristics and contami-
nant properties. The actual degree of remediation that can be achieved,
however, depends also on the skill and knowledge of the practitioner, which
is reflected in system design and operation. Inefficient remedial operations
may appear to be highly successful in the eyes of untrained practitioners.
Against this background, unfortunately, adequately documented case stud-
ies are not available that clearly demonstrate the results of vapor extraction-
based processes. Most case studies fail to report final soil and groundwater
contaminant levels, costs, and original target cleanup levels. In general, the
case studies fail to state what triggered the decision to turn the system off
and whether the cleanup goals were achieved. It is difficult, therefore, to
judge whether the application was "successful."
Process monitoring does not always indicate when cleanup goals have
been reached. Some vendors claim that vapor extraction cleanups occur in
fewer than 90 days at some sites. In the authors' experience, however, most
systems installed at commercial gasoline service stations operate for one to
five years.
2.10
-------
Chapter 2
Ultimately, soil and groundwater sampling are required to verify
cleanup. Soil sampling should be performed when contaminant and respira-
tory gas concentrations approach background levels, no significant restart
spike is observed, and oxygen utilization does not exceed background.
Many closure plans include a minimum one year period of posttreatment
vapor and groundwater monitoring in both vadose and saturated zones to
demonstrate long-term effectiveness. An endpoint can also sometimes be
negotiated based on economic analysis, i.e., when a predetermined cost per
unit contaminant removed is reached and operations are halted.
2.8 Prognosis
Of the vapor extraction-based technologies, soil vapor extraction has
been practiced longest and at the most sites. Laboratory and modeling stud-
ies have contributed a basic understanding of the processes involved in
system performance. Even with this seemingly mature process, however,
there is room for improving system performance and reducing costs. Engi-
neering analysis is required to assess the feasibility of vapor extraction-
based processes, interpret field test results, design systems, and optimize
system performance. Engineering analysis (design equations, computer
programs, etc.) could help reduce uncertainty in predictions of system per-
formance and cost. It could also supplement field experience by allowing
practitioners to predict what will happen at one site based on appropriate
applications of models and observations at other similar sites.
Air sparging and bioventing are promising emerging technologies at
different stages of development. A realistic niche has been identified for
each, but the site remediation industry has little practical experience with
them. Other vapor extraction-based processes, such as in situ soil heating,
are still considered experimental.
2.11
-------
-------
Chapter 3
3
PROCESS IDENTIFICATION AND
DESCRIPTION
Recently, in situ subsurface remediation processes have gained consider-
able attention among practitioners and researchers concerned with the
remediation of sites contaminated with volatile and semivolatile organic
compounds (VOCs and SVOCs). Of the in situ processes available, soil
vapor extraction has the greatest potential for widespread application in the
remediation of these sites. In addition, a number of complementary tech-
nologies such as air sparging, bioventing, subsurface heating, and
aboveground soil treatment using soil piles, are applied using soil vapor
extraction as the central remedy. Their application holds promise of
remediating sites having difficult geology or compounds in the subsurface
not fully removable by soil vapor extraction alone. At the time of writing
this document, the states-of-the-art and practice are both clearly more de-
veloped for in situ soil vapor extraction applications than for the comple-
mentary vapor extraction-based technologies. Therefore, emphasis will be
placed in this monograph on soil vapor extraction, while certain of the
complementary technologies will be addressed to the extent necessary to
reflect both the states-of-the-art and practice of soil vapor extraction.
According to the United States Environmental Protection Agency (US
EPA) the use of soil vapor extraction has dramatically increased in the past
few years. Based on US EPA's Superfund Remedial Action Records of
Decision (ROD) involving innovative treatment technologies as of June
1993, 202 were based on treatment (US EPA 1993). Of these, 107 cited
soil vapor extraction, while the next most frequently cited technologies
were bioremediation with 60 and thermal desorption with 32. The selection
of soil vapor extraction for Superfund sites increased 45% in federal fiscal
year 1991 over 1990. As of February 1992, soil vapor extraction had been
the most frequently selected source control technology under the entire
Superfund Program, representing 40% of all technologies at predesign
3.1
-------
Process Identification and Description
through completion stages. The inherent flexibility of soil vapor extraction
in application, the nature of the technology, and its ability to complement
other source control technologies have resulted in its increasing use as a
remedial technology.
3.1 Description of the Technologies
3.1.1 Soil Vapor Extraction
A basic in situ soil vapor extraction system, such as the one depicted in
figure 3.la (on page 3.3), couples vapor extraction wells with blowers or
vacuum pumps to remove contaminant vapors from zones permeable to
vapor flow, thereby enhancing the volatilization and removal of contami-
nant from the subsurface. The vacuum developed through a screened cas-
ing in the extraction well boring results in air being drawn from the
atmosphere through the soil to the well. Aboveground equipment often
consists of blower(s) or vacuum pump(s), control valves to adjust airflow
(particularly useful in multiple well, single pump applications), pressure
gauges and flow meters at wellheads, an air-liquid separator (for removal of
moisture from the extracted gases), a pressure gauge and flow meter at the
pump, and a vapor treatment unit. Vapor treatment systems such as cata-
lytic and thermal destruction systems, activated carbon adsorbers, and bio-
logical gas treatment systems, are included if air treatment is required. Air
discharge stacks may be adequate where it can be demonstrated that subse-
quent dispersion of the vapors reduces their concentrations to acceptable
levels. Also shown in figure 3. la is a groundwater pump and skim well,
which is one method to access residual contaminants that are trapped below
the normal groundwater level. When this occurs, the groundwater recovery
well is used to depress the water table, thereby exposing previously water-
saturated soils to vapor flow. More complex soil vapor extraction systems
incorporate trenches, horizontal wells, forced-air injection wells, passive air
inlet wells, low permeability or impermeable surface seals, multiple level
vapor extraction wells in single boreholes, and sophisticated unsaturated
zone moisture, contaminant, and temperature monitoring systems.
3.2
-------
Chapter 3
Figure 3. la
Simplified In Situ Soil Vapor Extraction Schematic
n ,,
Pressure flow 7aP°.r-
Gauge Meter
Vapor
Treatment Unit
Groundwater
Treatment
Soil vapor extraction can also be practiced ex situ in excavated soils;
perforated pipes are placed within soil piles to draw air through the pile, as
shown in figure 3.1b (on page 3.4). The equipment (blowers, vapor treat-
ment, etc.) is the same as that used in in situ applications.
The above-ground components of soil vapor extraction systems are typi-
cally off-the-shelf items, and qualified engineering firms can install wells
and trenches. Some large-scale applications may require firms experienced
in managing large construction projects, particularly where there are
trenches or aboveground applications. Despite the seeming simplistic na-
ture of the process and equipment, effective design, operation, monitoring,
and optimization of soil vapor extraction systems requires a thorough un-
3.3
-------
Process Identification and Description
Figure 3. Ib
Simplified Ex Situ Soil Vapor Extraction Schematic
1b Vapor Treatment
System I (-Q\
I I
Leachate Collection
System
derstanding of the governing phenomena. The basic issue of assessing soil
vapor extraction feasibility for any given site is itself attended by a great
degree of uncertainty. And, if one selects soil vapor extraction, design cri-
teria, involving the number, location, spacing, and construction of the ex-
traction/injection wells, vapor treatment systems, appropriate degree of
operational control, and requisite monitoring, can be difficult to determine.
In this monograph, an attempt will be made to compare and contrast the
states-of-the-practice and art of soil vapor extraction feasibility screening,
design, and operation.
Figure 3.2 (on page 3.5) provides a representative performance curve
(actual removal rates and time scales vary from site to site). Removal rates
of properly designed systems are highest when the system is first turned on.
They then decline over time and often reach an apparent asymptotic limit
that may be indicative of either mass reduction, compositional changes in
the residual hydrocarbon, or mass-transfer limitations (see Subsections
3.2.9 and 3.2.11).
As a result of experimental and theoretical investigations, the basic phe-
nomena governing performance of soil vapor extraction systems can be
readily explained. The application of a vacuum to an extraction well cre-
ates an air-flow field that originates at the ground surface and proceeds
preferentially along the path of least resistance (highest permeability to air
flow) to the screened interval of the extraction well. Air still follows other
3.4
-------
Chapter 3
Figure 3.2
Typical Soil Vapor Extraction Performance Curves
1200
1000
M
^
I
800
Source: Johnson et al 1991
paths but to a lesser degree. The air-flow field, shown in figure 3.la (on
page 3.3), has vertical and horizontal velocity components that are the result
of the horizontal and vertical intrinsic permeabilities of the soil, kv and kh,
and the relative permeability of the soil to air, which is a function of re-
sidual contaminant concentrations and soil moisture content. An additional
factor that may influence the air-flow field induced by soil vapor extraction
is the permeability of any surface material, ks. The comparative permeabil-
ity of the soil surface, ks, or the vertical permeability, kv, to that of the hori-
zontal component of the subsurface soil, kh, will determine the amount of
leakage occurring near a vapor extraction well. Contaminant vapors gener-
ated from volatilization within the air-flow field are then swept into the
vapor extraction well and removed from the soil. Additional contaminants
outside the flow field volatilize, diffuse in the air-flow field, and are simi-
larly removed from the soil. Ideally, this process of contaminant volatiliza-
tion and removal from the subsurface is continuous until all of the
3,5
-------
Process Identification and Description
contaminant in the unsaturated zone and above the capillary fringe are re-
moved from the soil.
It can readily be seen that a major goal of any vapor extraction operation
is to induce air to flow through the zone of contamination, thereby creating
ideal contact. When this occurs, vapors emanating from the residual con-
taminant do not need to diffuse a great distance to enter the air-flow field.
On the other hand, if the air-flow field does not pass through the residual
contaminant, but, for example, passes several feet above it, gas phase mo-
lecular diffusion will limit the rate of contaminant removal, as illustrated in
figure 3.3 (on page 3.7). In this situation, the process could take orders of
magnitude more time to remove residual contaminants than in the ideal case
of a vapor flow passing through the contaminant zone. It follows, therefore,
that the objective of a soil vapor extraction system is to minimize the length
of the diffusion path the volatilized contaminants must take to enter the air-
flow field. This can sometimes be accomplished by the addition of more
wells or well screens placed at specific locations within the subsurface. It
should be noted that microscale diffusion resistances (diffusion from within
soil particles) may also limit vapor extraction performance.
Although many complex processes govern the behavior of air flow and
contaminant interactions in the unsaturated zone, the three main factors that
control performance of soil vapor extraction-based systems are the (1)
chemical composition and characteristics of the contaminants to be re-
moved, (2) air-flow rates achievable through the contaminated soil, and (3)
air-flow path relative to the hydrocarbon distribution in the unsaturated
zone. Following are some of the mechanisms controlling vapor removal
from the unsaturated zone:
• gas advection;
• gas phase contaminant diffusion;
• sorption of residual contaminant and dissolved phase hydrocar-
bon on soil surfaces;
• chemical partitioning to the vapor phase;
• biological transformations;
• chemical transformations; and
• mass-transfer resistances controlling transport into the air-flow
field.
3.6
-------
Chapter 3
Figure 3.3
Effect of Gas-Phase Diffusion Llimitations on Performance
Vapor
Treatment Unit
Time
3.7
-------
Process Identification and Description
Although these mechanisms may affect the performance of a soil vapor
extraction system, it may not always be feasible to quantify their effect.
3.1.2 Air Sparging
It should be evident that an inherent limitation of soil vapor extraction is
that it will not remove contaminants that are beneath the unsaturated zone.
In many cases, contaminants are released into the subsurface in such large
quantities that the liquid migrates down to the saturated zone. Contaminant
migration through the vadose zone and toward groundwater occurs through
several mechanisms. Above the capillary fringe, the soil is only partially
saturated with water and contaminants migrate principally by vapor diffu-
sion, free-phase infiltration, and infiltration in solution. The majority of
contaminant mass migrates in free-phase, but mass migrating by this
mechanism is impeded in soil that is nearly saturated. This may occur in
any wet zone in the vadose zone, but always becomes the case within the
capillary fringe, which, depending on the soil and its capillary characteris-
tics, may extend less than a meter to as much as ten meters above the static
water table. Within the capillary fringe the free-phase contaminant accu-
mulates in the non-saturated portions of soil. Monitoring wells within the
zone typically show a layer of "pure" light non-aqueous phase liquid
(LNAPL) floating on the groundwater, but the actual transition from
LNAPL to groundwater in the subsurface is more gradual, with the capillary
fringe representing a zone of partial water- and partial LNAPL-saturation.
When the water table level fluctuates (because of seasonal or groundwater
pumping changes), the contaminant becomes "smeared," or immobilized
below the water table. If a basic soil vapor extraction system is used to
remediate the smear zone, groundwater depression is necessary to expose
the contaminant to vapor flow, as illustrated in figure 3. la (on page 3.3). At
some locations, however, depression of the water table may be impracti-
cable. This is often the case in very permeable (high yield) aquifers, or very
thick smear zones. Where dense nonaqueous phase liquids (DNAPLs) are
present, the penetration to confining layers deep in the saturated zone may
often preclude use of soil vapor extraction as the single treatment technol-
ogy because of the inability to fully dewater the saturated zone.
Air sparging, a recent innovation, may be able to extend the utility of soil
vapor extraction to water-saturated soils. Air is injected under pressure
below the water table as shown in figure 3.4 (on page 3.9). It is commonly
envisioned that injected air in the form of bubbles travels up through the
3.8
-------
Chapter 3
saturated zone, thereby transferring the hydrocarbons from the soil and
groundwater phase into the vapor phase. It is more realistic, however, to
expect continuous air channels to direct air flow in the saturated zone. In
addition to volatilization, it is postulated that the injected air increases the
oxygen resupply rate to the groundwater, resulting in increased aerobic
biodegradation rates. Thus, there is some debate whether air sparging is
primarily a volatilization-based process or simply an engineered
bioremediation process. In figure 3.4, air injection wells are coupled with
vapor extraction wells to assure that hydrocarbon vapors pushed into the
unsaturated zone are recovered and handled appropriately aboveground.
Once air is injected into the saturated zone, its advective flow is gov-
erned largely by the applied pressure, buoyant forces, vertical and horizon-
Figure 3.4
Simplified Air Sparging/Vapor Extraction Schematic
Vapor
Treatment Unit
Pressure Gauge
Blower/Compresser
i
^
3.9
-------
Process Identification and Description
tal permeability distributions in the saturated zone, immiscible fluid dis-
placement phenomena, and capillary properties of the soils. Depending on
the size of the pores within the media, either air bubbles (extremely coarse
materials) or continuous air channels (fine subsurface materials) form, pro-
viding a pathway for the injected air to the unsaturated zone. Significant
channeling of the injected air may result from relatively minor heterogene-
ities in the saturated zone. It is expected that contaminants located within
air-flow pathways will volatilize, and the potential exists for the stripping of
dissolved-phase contaminants that contact the air-flow field. It is intended!
that air be injected to flow primarily through the area of contamination in
the saturated zone as shown in figure 3.4 (on page 3.9); however, in view of
the complex phenomena governing the injected air-flow field and the lim-
ited ability to characterize soil heterogeneities, there is a large degree of
uncertainty a priori as to the true behavior of any air-sparging system de-
sign. Figure 3.5 provides sampling data indicating that operation of the air
sparging system resulted in an initial doubling of the extracted vapor con-
centration from the soil vapor extraction system.
Figure 3.5
Extracted Vapor Concentrations for a Soil
Venting/Air Sparging Application
~r—i—i—i—i 1—r^n—i 1 1 1 1—i 1 1 w i
20 40 60 80 100 120 140 160 180 200 220 240
Days of Operation
Source1 Brown, Henry, and Herman 1992
3.10
-------
Chapter 3
To the authors' knowledge, the first use of air sparging reported in the
literature as a remedial technology was in an EPA office of Solid Waste and
Emergency Response publication (US EPA 1988) for an installation in the
Federal Republic of Germany (see also Bohler et al. 1990; and Middieton
and Hiller 1990), mainly to treat groundwater contaminated with chlori-
nated solvents. It is likely to have been used prior to this time. The tech-
nology was later used in the U.S. to remediate gasoline-contaminated
groundwater and soil in the saturated zone (Ardito and Billings 1990;
Brown and Fraxedas 1991; and Marley 1991).
Although the use of air sparging has increased rapidly, considerable re-
search and demonstration must be conducted before a consistent and reli-
able design approach can be realized and before a full understanding of its
effectiveness and feasibility can be achieved.
As opposed to basic soil vapor extraction applications, improperly con-
trolled air-sparging systems can create significant health and safety risks.
This results from the introduction of pressurized air in the subsurface,
which can accelerate the uncontrolled migration of hydrocarbon vapors, and
the consequent accumulation in buildings, utility trenches, and other vapor
receptors above and adjacent to the remediation area. In addition, the in-
jected air initially displaces groundwater. Thus, there may be the potential
for enhanced spreading of soluble hydrocarbon plumes. Another possible
consequence is that air injection will result in increased mixing and, there-
fore, increased mass transfer of residual hydrocarbon components into dis-
solved and vapor phases. If the increased concentrations of these dissolved
constituents are not captured in a pump-and-treat system or stripped by the
injected air, it is possible that increases in groundwater concentrations could
result. Safeguards, two of which are described below, should be considered
to insure that these risks do not materialize:
• concurrent installation of a soil vapor extraction system to con-
trol the migration of contaminant vapors, except when degrada-
tion is significant enough to insure no impacts; and
• capture or containment of groundwater to prevent off-site migra-
tion of dissolved residual hydrocarbon components when sensi-
tive receptors (e.g., domestic or public water supply wells) are
nearby.
In addition to the health and safety concerns cited above, one should be
aware of the potential for air sparging to result in modified aquifer condi-
3.11
-------
Process Identification and Description
tions, such as aquifer plugging because of iron precipitation or biomass
growth stimulated by increased oxygen supply rate.
3.1.3 Bioventing
Under natural conditions aerobic biodegradation rates are typically lim-
ited by oxygen supply rates in the subsurface. But the rate of oxygen resup-
ply to the subsurface is increased during the course of vapor extraction as
air is drawn from the atmosphere into the subsurface. Consequently, it has
been hypothesized that the enhanced oxygen delivery will affect an increase
in aerobic biodegradation of hydrocarbons (Texas Research Institute 1980
and 1984; Wilson and Ward 1986; Bennedsen, Scott, and Hartley 1987;
Conner 1988; and Ostendorf and Kampbell 1989). Field applications of soil
vapor extraction conducted with the purpose of enhancing biodegradation
have been reported by Ely and Heffner (1988), Staps (1989), and Hinchee et
al. (1989). To fully mineralize a hydrocarbon such as benzene (convert to
CO2, H2O, and biomass), a mass ratio of 3:1 g-oxygen/g-benzene is re-
quired. Alternate means of O2 delivery have been considered, including
addition as dissolved O2 in water. However, assuming oxygen saturations
of 8 mg/L O2 for air-saturated water, 40 mg/L O, for pure O2-saturated wa-
ter, and 250 mg/L O2 for a 500 mg/L H,O, application, 2,000, 400, or 70
pore volumes of water, respectively, would be needed to fully mineralize
about 1,000 mg/kg benzene in soil having a bulk density of 1,600 kg/m3
(100 lb/ft3) at a porosity of 30%. Relative to the use of water infiltration,
inducing air flow is a significantly more efficient means to deliver O2 to the
unsaturated zone.
A by-product of the aerobic degradation of hydrocarbons is COr Ely
and Heffner (1988) observed that CO2 was generated during soil vapor ex-
traction; this was interpreted to be an indication of hydrocarbon degradation
in the unsaturated zone. Furthermore, Kerfoot et al. (1988) and Woo and
Coleman (1988) observed depressed O,, and elevated CO2 concentrations in
soil-gas measurements taken at fuel hydrocarbon contaminated sites. It
should be noted, however, that CO2 may be produced by inorganic carbon-
ate reactions; thus, one must be careful when inferring biodegradation rates
from CO2 measurements (Hinchee and Ong 1992). Smith, Dupont, and
Hinchee (1991) reported that 50% of the approximately 40,000 kg (88,185
Ib) of jet fuel removed from the subsurface at Hill Air Force Base, Utah,
was removed by biological degradation, while the remainder was removed
3.12
-------
Chapter 3
by volatilization. Before bioventing, subsurface soil-gas oxygen concentra-
tions in areas of high jet fuel contamination were less than 1%. Following a
period of bioventing, much of the unsaturated zone formerly devoid of oxy-
gen contained between 15 to 20% oxygen. In the initial period, the extrac-
tion rate averaged approximately 700 L/sec (1,500 ftVmin), corresponding
to about 15 to 20 void volume exchanges per day. Biodegradation in this
period represented approximately 15% of total removal. This was followed
by a period of lower airflow. Extraction averaged approximately 47 L/sec
(100 ftVmin) (representing =1 void volume exchange per day). Moisture
and nutrient additions also were initiated during this period. During the low
airflow rate period, volatilization declined greatly and biodegradation re-
moval increased. The result was that biodegradation represented over 95%
of the removal. Moisture addition did increase the rate of biodegradation,
but it was not clear that nutrient addition had any beneficial effect (Dupont
et al. 1991). To date, it has not been shown in a field study that nutrient
addition has a beneficial effect on bioventing. Comparison of removal rates
attributed to volatilization and biodegradation during a bioventing study at
Tyndall Air Force Base, Florida, indicates that the mechanisms are of com-
parable effect (see figure 3.6 on page 3.14).
The physical setup for bioventing appears similar to standard vapor ex-
traction configurations, but the objectives of the processes differ. In soil
vapor extraction the goal is to maximize the rate of volatilization, which
typically translates to maximization of vapor extraction rates. The goal of
bioventing, however, is to optimize aerobic biodegradation. Operationally,
this often requires maintaining the oxygen content in the soil at some set
value (2 to 4% v/v); corresponding vapor-flow rates are then typically lower
than in soil vapor extraction.
Various bioventing configurations have been proposed by Hinchee et al.
(1989); two of these are illustrated in figure 3.7 (on page 3.15). Figure 3.7a
presents a standard soil vapor extraction configuration wherein the ex-
tracted vapors are treated aboveground. The system shown in figure 3.7b is
configured to eliminate the need for aboveground vapor treatment; the vola-
tilized hydrocarbon vapors are drawn through a "clean" soil zone in the
hope that they will be degraded in the subsurface before being extracted
from it. In another configuration that does not use any vapor extraction
wells (not shown), air is injected into the residual hydrocarbon zone at a
controlled rate so that contaminant vapors are degraded as they are driven
3.13
-------
Process Identification and Description
Figure 3.6
Comparison of Removal Rates Attributed to Volatilization and
Biodegradation During Soil Venting at the Tyndall AFB Site
B % Removal Attributed to Biodegradalion - V1
• Total Removal - Hexane Eq (Vol+Deg) - VI
60
90 120
Venting Time (days)
150
180
210
cu
S
Source-Miller 1990
away from that zone. This design is not discussed at length in this document
as it is neither an extraction nor a vapor-extraction based technology.
Bioventing clearly represents a positive advance in vapor extraction-
based applications at sites having aerobically-degradable contaminants in
the unsaturated zone. This technology extends the potential application of
soil vapor extraction to sites contaminated with semivolatile fuel hydrocar-
bons, while requiring minimal modification of soil vapor extraction system
design.
3.1.4 Thermal Enhancements of Soil Vapor Extraction
It is generally accepted that soil vapor extraction becomes less practi-
cable for compounds with vapor pressures less than 0.1 mm Hg to 1.0 mm
Hg at ambient temperatures. It is expected, however, that the range of ap-
3.14
-------
Chapter 3
Figure 3.7
Simplistic Bioventing Process Schematics
a)
Vapor
Treatment Unit
Oxygen Supplied From Atmosphere
Vapor
Treatment Unit
b)
Blower/Vacuum
'Pump
Oxygen Supplied From Atmosphere
| Zone of Vapor Treatment [
Based on Hinchee et al. 1989
3.15
-------
Process Identification and Description
plicability can be extended by heating the subsurface, as contaminant vapor
pressures increase with temperature (assuming that removal rates are
roughly proportional to the partial pressure of a compound).
In addition, it is expected that microbiological activity can be increased
by gently warming the subsurface to temperatures in the range of 30° to
40°C (86° to 104°F), as biodegradation rates are expected to double for
every 10°C (SOT) temperature increase when other factors are not limiting.
Temperatures above this range are likely to be inhibitive.
There are a number of alternative in situ heating methods currently being
studied, including steam injection, hot-air injection, electrical-resistance
heating, radio-frequency heating, thermal-conduction heating, warm water
injection, and solar heating. While heated air injection is probably the most
practicable if obvious safety concerns are addressed (see Subsection 3.1.2),
it is not a very efficient means for delivering heat to the subsurface, given
the relatively low-heat capacity of air. For this reason, the alternate energy
delivery systems listed above are being investigated.
Few reports concerning in situ heating methods have been published and
they are considered to be developmental. But, surveys of thermally-en-
hanced vapor extraction processes can be made by reviewing journal and
conference articles and the patent literature. For example, Bridges et al.
(U.S. 4,376,598) and Johnson et al. (U.S. 5,076,727) disclose processes for
radio-frequency heating, Nelson and Rau (U.S. 5,011,329) describe a hot-
gas injection system, and Johnson et al. (U.S. 5,114,497) describe a thermal
conduction scheme for the treatment of contaminated surficial soils. See
Section 3.7 for a more complete description of in situ heating methods.
3.2 Fundamentals and Basic Phenomena
The performance of vapor extraction-based processes can be related to
three primary factors: equilibrium partitioning into the vapor space in a
porous medium, vapor-flow characteristics in the porous medium, and
mass-transfer considerations and limitations. In the subsections immedi-
ately following, the subsurface distribution of contaminants is described
qualitatively and a mathematical framework for discussing each of the fac-
tors identified above is established. It will be seen that the ability to quan-
3.16
-------
Chapter 3
tify these phenomena plays an important role in the design of pilot-scale
tests, evaluation of the data, and design of remedial systems.
3.2.1 Qualitative Description of the Subsurface Distribution
of an Immiscible Liquid
Typically, sites that are considered candidates for soil vapor extraction
and other vapor extraction-based technologies are locations where petro-
leum products (e.g., gasoline and other fuels) or chlorinated solvents have
spilled or leaked into the subsurface. These materials are often called "im-
miscible liquids," because of their limited solubility in water at environ-
mental pressures and temperatures. In typically released quantities of the
spilled liquids, they do not dissolve completely in the soil moisture or
groundwater and, consequently, a separate nonaqueous (not dissolved in
water) liquid phase is introduced into the subsurface. Immiscible liquids
may be chemically homogeneous, that is, they may consist predominantly
of one molecular constituent, as in a spill of a solvent comprised almost
entirely of trichloroethylene (TCE). The immiscible liquid in spilled re-
fined petroleum products, however, is typically chemically heterogeneous,
being a mixture of molecular constituents. For example, gasoline is a mix-
ture of hundreds of hydrocarbon constituents.
Immiscible liquids are called "light nonaqueous phase liquids"
(LNAPLs) if their density is less than that of water, or DNAPLs if their
density is greater than that of water. For example, gasoline (density =0.8 g/
cm3 (50 lb/ft3)) is an LNAPL and TCE (density =1.5 g/cm3(95 lb/ft3)) is a
DNAPL. When a release of either occurs, the liquid is driven downward by
gravity. In the course of its migration, lateral spreading of the liquid occurs
because of variations in soil permeability and layering, and liquid is re-
tained in the soil pores because of capillary forces. The amount of immis-
cible liquid immobilized in this way per unit of contaminated sediment is
referred to as the "residual saturation" in the soil matrix. For example,
Hoag and Marley (1986) determined residual saturations of gasoline to be
26,000 and 44,000 mg gasoline/kg soil for a medium and fine sand, respec-
tively. In small releases, all the liquid eventually becomes trapped and is
immobile; however, in more extensive spills, the capacity of the soil to
immobilize the liquid is exceeded and immiscible liquid drains to the water
table. At this point, LNAPLs spread laterally across the capillary fringe,
while DNAPLs continue to migrate below the water table until they become
immobilized or encounter a confining soil layer. Although the water table
3.17
-------
Process Identification and Description
provides an initial physical barrier for LNAPLs, a significant volume of
LNAPLs is often found trapped in a "smear zone" beneath the water table
because of natural water table fluctuations and remedial activities (e.g.,
pump-and-treat operations). This resulting distribution of immiscible liquid
in and near groundwater controls the rate at which constituents solubilize
and enter underlying groundwater, as well as the rate at which constituents
can be removed by vapor extraction. Essaid, Herkelrath, and Hess (1991)
provide a detailed description of the distribution of air, oil, and water adja-
cent to the water table at a U.S. Geological Survey research site near
Bemidji, Minnesota.
Layered lithologic units complicate the conceptualization of immiscible
contaminant migration outlined above. For example, a lens of clay or fine
silty sand can provide a physical barrier between the contaminant source
and the water table, thereby causing the immiscible liquid to "pond" on top
of the less permeable unit. As will be seen, heterogeneities in lithology also
affect the distribution of air flow in the unsaturated zone for a given vapor
extraction installation. Therefore, identification and characterization of
subsurface heterogeneities is essential for the designing of vapor extraction
systems.
Although physical properties of the liquid (e.g., density and surface ten-
sion) and the subsurface media control the initial migration and distribution
of immiscible liquids, the partitioning of constituents of these liquids into
aqueous, gaseous, and sorbed phases is also significant. The partitioning of
constituents into the aqueous phase affects water quality problems resulting
from a release, while partitioning into the vapor phase is a controlling factor
for vapor extraction-based technologies (one of the goals is to remove the
vapors efficiently). Although trapped immiscible liquids are said to be
immobile, constituents of these liquids continue to migrate in the environ-
ment as a result of partitioning. Constituents that dissolve in the soil mois-
ture are predominantly transported by leaching and groundwater movement,
while those partitioned into the vapor phase migrate as a result of vapor-
phase diffusion or induced vapor flow (as in the application of vapor extrac-
tion-based technologies). Historically, aromatic hydrocarbons (e.g.,
benzene, toluene, and xylenes) have been of primary concern in petroleum
product spills, since they are the most soluble constituents (see Subsection
3.2.3) and benzene is a known human carcinogen (US EPA IRIS Database
1993). Fortunately, these compounds also partition significantly into the
vapor phase and are often effectively removed by vapor extraction-based
3.18
-------
Chapter 3
technologies (see figure 3.8 on page 3.20; see also figure 3.9 on page 3.21).
Total volatile hydrocarbons consist of the sum of volatilized constituents of
the petroleum product that have migrated away from the oil body.
The microbial degradation of contaminants provides another mechanism
for the attenuation of compounds present in the unsaturated zone. In aero-
bic degradation, bacteria use oxygen to break down contaminants into car-
bon dioxide, water, and biomass (cells). Microbial degradation rates,
however, are often limited by the diffusive recharge of oxygen from the
atmosphere into the contaminated zone under natural conditions. Thus,
another beneficial consequence of vapor extraction is that of greatly en-
hancing the rate at which oxygen is recharged, often resulting in increased
microbial degradation rates. This is the basis for bioventing. See figures
3.10a and 3.10b (page 3.22); at this site, the microbial degradation rate in
the most highly-contaminated zone surrounding the former location of the
leaking tank is limited by the rate of oxygen diffusion.
3.2.2 Quantitative Description of Multiphase Partitioning
The extent and rate of the partitioning among vapor, aqueous (soil mois-
ture), sorbed, and immiscible (residual liquid or solid) phases in the subsur-
face described in Subsection 3.2.1, above, play a significant role in
contaminant migration and, consequently, influences the performance of
any vapor extraction-based process. Thus, it is important to quantify this
phenomena. In this section the working approximation used by
Corapcioglu and Baehr (1987), Johnson, Kemblowski, and Colthart (1988),
and Johnson et al. (1990) is followed in order to establish a framework for
mathematically describing the chemical partitioning within soil matrices.
The goal is to develop a model that can be used to predict equilibrium vapor
concentrations of compounds in the subsurface. The use of this information
in estimating the performance of vapor extraction-based processes will be
demonstrated in subsequent sections.
In general, the total volumetric concentration T (g-j/cm3-soil) of compo-
nent j can be distributed in the subsurface between vapor, aqueous (soil
moisture), sorbed, and immiscible phases as described by the mass balance:
r, = eacj + ewCj
3.19
-------
Process Identification and Description
3.20
-------
Chapter 3
» Figure 3.9
Cross Section Illustration of Concentration of Total Volatile Hydrocarbons
Detected in Soil Gas at USGS Study Site Near Bemidji.MN
s
2
I
s
s
436
434 -
50 100
Distance trom Source, in Meters
200
-10-
Total Volatile Hydrocarbons, in grams per cubic meter
Source Hull, Landon, and Pfannkuch 1991
where 9 , 9w, and 9 denote the volumetric contents (cmVcm'-soil) of the
vapor, aqueous, and immiscible phases, respectively, and pb is the soil bulk
density (g-soil/cm3-soil). The volumetric concentrations of component j in
the vapor, aqueous, and immiscible phases are denoted G, C, and I (g-j/
cm3-vapor, -water, or -immiscible phase), respectively. The symbol S rep-
resents the amount of component j that is sorbed to the solid matrix per unit
mass of soil (g-j/g-soil). For equation [3.1J to be useful, supplemental
relations describing the equilibrium between phases must be supplied. The
idealized relationships that are standardly applied in this analysis are exam-
ined next.
Raoult's Law (Atkins 1978) is assumed to quantify the equilibrium be-
tween immiscible and gaseous phases; the gaseous phase concentration G
(g-j/cm3-vapor) of constituent j is related to the mole fraction % (dimension-
less) of constituent j in the immiscible phase and the saturated vapor con-
centration G"" (g-j/cm'-vapor):
GJ -
[3.2]
3.21
-------
Process Identification and Description
Figure 3.10
Areal Distribution of (a) Oxygen and (b) Carbon Dioxide as a
Percentage of the Soil Gas at a Depth of 6 ft BGS at the Galloway
Township, NJ, Gasoline-Spill Site, December 1989
•/
\
V
\
10 20
EXPLANATION
^—10— Normalized P«rc*nlig« o» oxygen by vWufli*. Contour Interval B percent
* Neittd v»por probei # Mul[l-laval oround-wat*r s*mi
VW4 Vtpof proM ni
\
20 30 40 SO FEET
EXPLANATION
- 10 - NormalliM
* N»lad v«pd> probe*
5 Therm Mar nea
t Vapor nxlr«ctl<
I Vapor prob* m
Source: Baehr and Hurl 1991
3.22
-------
Chapter 3
Here G"' is defined:
RT
[3.3]
where Mw (g-j/mole) is the molecular weight of the jth constituent, R is the
gas constant (82 cm3-atm/mole-K), T is the absolute temperature (K) and
Pv is the vapor pressure (atm) over the pure constituent j at temperature T.
For most compounds, Pv can be found in standard reference books (e.g.,
Verschueren 1983) for typical ambient temperatures (i.e., 20° to 25°C (68°
to 77°F)). To extrapolate to other temperatures, the Clausius-Clapeyron
equation (Atkins 1978) for liquid-vapor equilibrium is often used:
P] = exp
AH
-
RT
+ B
[3.4]
where AHv is the molar heat of vaporization (cal/mole), R is the gas con-
stant (1.99 cal/mole-K) and B is a unitless constant. Values for AHv can be
found in standard thermodynamic tables (e.g., Weast 1970), then B can be
determined by solving equation [3.4] for Et at a known temperature and
reference vapor pressure.
The equilibrium between aqueous and immiscible phases is assumed to
be described by an expression analogous to equation [3.2]:
Cj = Xtf [3.5]
where C^ (g-j/cm3-H2O) is the concentration of the jlh constituent in the
aqueous phase and Cs<" (g-j/cm3-H2O) is the solubility of the pure constitu-
ent j in water.
Henry's Law defines the equilibrium between aqueous and gaseous
phases as follows:
GJ = HjC] [3.6]
where H is the Henry's Law partition coefficient. To maintain consistency
with equations [3.2], [3.5], and [3.6], H is defined as follows:
3.23
-------
Process Identification and Description
(-ISCU
Hi = i [3.7]
Equation [3.7] also applies to regions of the unsaturated zone where the
immiscible phase is not present.
The mass of constituent j adsorbed to solid surfaces per unit mass of
subsurface soil, S , is often described as a function of the aqueous phase
concentration and a sorption coefficient ks . The simplest expression of this
relationship is a linear isotherm:
Sj = ks.Cj [3.8]
The sorption coefficient is generally taken to be a function of the fraction
of organic carbon f^ in the solid matrix (except for very low f ) and is de-
termined experimentally or estimated from one of several published correla-
tions (Lyman, Reehl, and Rosenblatt 1982). One expression that is
commonly used for gasoline-range compounds is the Karickhoff ( 1 98 1 )
equation:
*,,, = 0.63kowjfoc [3.9]
where kow denotes the octanol- water coefficient for compound j.
See table 3.1 (on page 3.25) for values of Gsal , S , H , and kow for se-
lected hydrocarbons. The reader can consult sources cited in the table for
values for other compounds.
Equilibrium expressions [3.1] through [3.9], subject to the constraint:
= 1 n= # of components [310]
provide a working approximation for equilibrium partitioning in porous
media. In general, solution of this set of equations is cumbersome and re-
quires iteration; however, useful approximations result when the limits of
"high" and "low" total volumetric concentrations are evaluated. Figure
3.11 (on page 3.26) presents the results of model calculations for gasoline in
3.24
-------
Chapter 3
Table 3.1
Thermodynamic Data for Selected Hydrocarbons (T = 2CTC)
benzene
n-hexane
toluene
o-xylene
of
[g/cm3- vapor x 10']
0.303
0617
0 133
0037
Csat
[g/cm3-H2O x 10']
1.78
00095
0.515
0.175
HJ
[cm'-HjO/cm'-air]
0.17
64.9
0.26
021
KOWJ
[cm3-H2O/g-soil]
135
8710
490
589
Gsat calculated from Equation [3 3) and values from Weast 1970
C83' from Verschueren 1983
Another source of basic data MacKay and Shiu 1981
a sandy soil. In this figure, the "full" model refers to solutions of equations
[3.1] through [3.10]. In the limit of high concentrations (>500 mg-gasoline/
kg-soil), the vapor concentrations become independent of the total gasoline
concentration and depend only on composition (the inflection point in fig-
ure 3.11 (on page 3.26) is an artifact of the calculation procedure and is not
expected to be a "real" phenomenon). Thus, the vapor-phase concentration
is adequately predicted by equations [3.2] and [3.3], rewritten here as:
G> =
As illustrated by figure 3.11 (on page 3.26), the maximum vapor concen-
tration achievable in the porous medium is that value predicted by equation
[3.1 1]. At most sites, equation [3.11] is likely to provide a reasonable esti-
mate of vapor concentrations and partitioning corresponding to conditions
before remediation. To estimate this upper bound, one needs to know only
two chemical-specific properties ( Pv. and Mw ), as well as an approximation
of the spilled product (immiscible phase) composition (x ). For refined
hydrocarbon mixtures, such as gasoline, hundreds of compounds can be
identified; thus, it is often necessary to approximately define the composi-
tion by grouping compounds into constituent classes, as illustrated for gaso-
3.25
-------
Process Identification and Description
Figure 3.11
Comparison of \fapor Concentration Prediction Models
100
c3 io
Equation [3-12]
Equation [3-11]
Full Model
T = 20'C
100 1000 10000
Residual Soil Concentration (mg-gasohne/kg-soil)
1000
100 ;
> 10 -
H
m
Equation [3-12]
Equation [3-11]
Full Model
T = 20"C
100 1000 10000
Residual Soil Concentration (mg-gasolme/kg-soil)
* denotes the sum of benzene, toluene, ethylbenzene, and xylenes vapor concentrations
Reprinted by permission of CRC Press from 'Estimates of Hydrocarbon Vapor Emissions Resulting from Service Station
Remediations and Buried Gasoline-Contaminated Soils* by PC Johnson, M B Hertz, and D L. Byers, Petroleum
Contaminated Soils, Vol 3, editors, P.T Kostecki, and E.J. Calabrese, Lewis Publishers, Copyright 1990 by CRC Press.
3.26
-------
Chapter 3
line in table 3.2. Another example of grouping compounds can be found in
table 2 of Johnson et al. (1990). Note that a single hydrocarbon can define
a class (e.g., benzene) if the hydrocarbon partitioning properties in the mix-
ture are representative of the range of interest. The approximate composi-
tion given in table 3.2 was based on the packed-column gas chromatograph
analysis of a regular, leaded gasoline reported by Bruell and Hoag (1984).
A constituent class is assigned properties by identifying a composite chro-
matographic peak with a known standard. Note that for this gasoline, the
aromatic constituents, benzene, toluene, and three xylene isomers, made up
20.1 % of the product, but account for about 80.6% of the total volatile hy-
drocarbons partitioned in the aqueous phase, because nonaromatic hydro-
carbons have very low solubilities. Note also that the gaseous phase
partition (the portion available for venting) is significant for all classes of
Table 3.2
Gasoline Hydrocarbons Grouped in Constituent Classes
1
2.
3.
4.
5.
6.
7.
8.
9.
10
benzene
toluene
Cg aromatics
C9-C,, aromatics
C5 alkenes
C5-C6 alkenes
C6 napthenes
C7-C|] alkanes
C6-Cn alkenes
C7-Cn napthenes
Mw,J
g/mole
78
92
106
132
70
83
84
113
103
98
XJ
unitless
0044
0064
0093
0 173
0073
0252
0.035
0241
0.017
0.008
Gsat
g/cm3x 101
0303
0 134
0044
0.014
169
0.70
0.55
025
013
0.12
.-.sat
g/cm'x 101
1 78
0.515
0156
0040
0.203
0021
0.055
0002
0030
0.030
G . _ saturated vapor concentration
Csal - Saturated dissolved concentration - solubility
Source' Corapcioglu and Baehr 1987
3.27
-------
Process Identification and Description
hydrocarbons. Another method of identifying classes of constituents, based
on the boiling points of known standards and their retention time for a se-
lected chromatographic method, is discussed by Johnson et al. (1990).
In the limit of low hydrocarbon concentrations, the equilibrium partition-
ing model reduces to:
ET.
G, = -F '—I T [3.12]
1 [HA + ew + PA,,]
which predicts that the equilibrium vapor concentration is proportional to
the residual concentration of j in the subsurface and is independent of com-
position as exhibited by the results presented in figure 3.11 (on page 3.26).
This low concentration regime is sometimes referred to as the "Henry's
Law Limit", and its effect may be significant near the completion of
remediation by soil venting or where there are very highly water soluble
compounds, such as oxygenated hydrocarbons. Equation [3.12] indicates
that venting is inherently less efficient as the process progresses because the
vapor concentration of compound j decreases in proportion to decreases in
the concentration of j in the subsurface. This has serious practical implica-
tions at sites where the goal is to achieve extremely low levels (ppb range)
of residual contamination.
3.2.3 Basis for Mathematical Models of Induced Air Flow in
the Unsaturated Zone
The success of a vapor extraction installation is fundamentally dependent
upon the establishment of an air-flow field that intersects the contaminant in
the unsaturated zone. The air-flow field developed for a given configuration
of extraction and injection wells depends on many factors, including the
pump(s) employed, screened intervals of the wells, depth to groundwater,
and the spatial distribution of air permeability in the unsaturated zone. A
mathematical model of induced airflow in the unsaturated zone provides a
valuable tool for simultaneously considering the effect of these factors. The
following discussion outlines the derivation of the general model presented
in detail by Baehr and Hull (1991) and then presents solutions for a number
of simplified scenarios.
3.28
-------
Chapter 3
Some readers will note similarities between the mathematical framework
given below and that commonly used when developing mathematical mod-
els of groundwater flow phenomena. The main physical difference between
the two being that vapors are compressible under most practical situations,
while groundwater is not. If compressibility is neglected, vapor and
groundwater flow equations are identical, and all of the analytical and nu-
merical tools developed for groundwater applications are directly appli-
cable, upon substitution of the appropriate fluid properties (Massman 1989).
Even when compressibility is retained in the development (as is done here),
the groundwater flow-vapor flow similarity is retained in many cases. For
example, it well be seen in Section 3.2.6 that equations governing the pres-
sure (head) field are similar in form as long as the vapor pressure field
equation is expressed in terms of the square of the pressure. To date, this
similarity has not been fully exploited in the development of predictive
tools for vapor flow.
The equation governing airflow in an unsaturated, porous medium origi-
nates with a mathematical expression of mass conservation:
? • (paqa) = 0 [3.13]
where pa (g/cm3) is the density of the vapor phase, 6a (cm3-vapor/cm3-soil)
is the vapor-filled porosity, and t(s) is the time. The specific discharge
(darcy velocity) vector qu (cm/s) is assumed to be related to the fluid poten-
tial O (cmVs2) through the following form of Darcy 's Law:
q, = -- k • VO [3.14]
where ii (g/cm-s) and k (cm2) denote the vapor phase viscosity and air per-
meability tensor, respectively. For compressible fluids (i.e., gases), is
given by (Hubbert 1940):
r dP
J^r
P. [3.15]
3.29
-------
Process Identification and Description
where z (cm) represents the elevation, P (g/cm-s2) is the vapor phase pres-
sure, and Po (g/cm-s2) is a reference vapor phase pressure. To complete the
mathematical formulation, a constitutive expression relating vapor-phase
density pa and pressure P is needed. Here, the Ideal Gas Law is adopted:
[3-16]
R T
where Mw (g/mole) is the average molecular weight of the vapor phase, R is
the universal gas constant (82. 1 cm3-atm/mole-K), and T (K) is the absolute
temperature.
By first substituting equation [3.16] into equation [3.15] and then assum-
ing that Mw is constant and the first term on the right side of equation [3.15]
is negligible in comparison to the second term (a good assumption for flows
induced by pressure gradients), a simplified expression for the fluid poten-
tial can be derived. When this expression is inserted into equation [3.14],
Darcy's Law reduces to:
-------
Chapter 3
proximation of flow to a trench or flow in a laboratory soil column and the
latter is often used as an approximation of flow to a vertical well. If a ho-
mogeneous media with constant air permeability and steady-state flow is
assumed, then equation [3.18] reduces to:
0 = - (linear flow) [3. 19a]
dx
0 = -—[r—1 (radial flow) [3.19b]
r dr \ dr
In each case boundary conditions must be specified at two locations;
typically these take the following form:
P = PW x = 0; P = Palm x = Latm (linear flow) [3.20a]
P = PW r = Rw; P = Patm r = Ralm (radial flow) [3.20b]
where Paim denotes atmospheric pressure and Pw is the pressure applied at
the vapor extraction location. It is important to note that pressures given
here are absolute pressures and not the gauge pressures (P = (P - Patm)) typi-
cally measured in the field. The solutions for the pressure distribution and
flow rates at the extraction locations for each scenario are:
P2 =
(P2 - P2}
\ aim w)
x + P2 (linear flow) [3.2la]
= Pi + (PL - ^) (radial flow) [3.21b]
The corresponding relationships between volumetric extraction rates Q
(cmVs) and applied vacuums Pw (g/cm-s2) are:
3.31
-------
Process Identification and Description
Q =
= HW\-
L P
atm w
(linear flow) [3.22a]
Figure 3.12
One-dimensional Vapor Flow Scenarios
a) Linear Flow
Vapor Flow
b) Radial Flow
Ratm
Vapor Flow
3.32
-------
Chapter 3
Q =
= H
(radial flow) [3.22b]
Here the convention that flow rates are positive if a vacuum is applied at
the designated extraction location (i.e., x = 0 and r = Rw) is used. The thick-
ness of the permeable zone in both cases is designated H, and W represents
the width of the extraction location in the linear flow case (HW is then the
cross-sectional extraction flow area).
Figure 3.13
Predicted Vapor Flowlines to a Vapor Extraction Trench
percentage of total flow to trench
entering ground surface between this
location and trench well head. Note
that only half-plane is shown.
7.5% 15% 22.5% 30%
37.5%
Isotrppic and uniform permeability
distribution (i.e. equivalent
horizontal and vertical
permeabilities)
10 20
Distance [ft]
Source: Johnson and Ettinger 1992
3.33
-------
Process Identification and Description
If applying a one-dimensional solution given by either equation [3.21]
or [3.22], one must fully understand the nature of the geometric approxima-
tion; otherwise, false conclusions can result from the analysis. For ex-
ample, in the case of a single extraction well, the source of air to the well is
the atmosphere and therefore all flow paths intersect the land surface. Fig-
ure 3.13 illustrates the predicted two-dimensional flow paths induced by an
extraction trench in homogeneous porous media. The solution given by
equation [3.21b] is to be interpreted as an approximate pressure distribution
with r representing distance along the upwardly curving flow path that in-
tersects the atmosphere at a distance Raim where the pressure is atmospheric
Patm. Obviously, a flow path originating near the top of the well screen in-
tersects the land surface at a distance shorter than one originating near the
bottom of the well screen. Therefore, the "radius of influence", Ralm, pre-
dicted in applying the one-dimensional solution here represents some sort of
average over all flow paths. Clearly, higher dimensional solutions are re-
quired to rigorously analyze flow induced by wells. Use of a higher dimen-
sional model that includes the depth component, z, will result in a more
realistic definition of the zone of significant vapor flow.
It is worthwhile noting at this point that equations derived from analyses
of simplistic scenarios are misused quite frequently in practice. For ex-
ample, equations [3.21b] and [3.22b] contain the parameter, Ratm, which is a
quantity that arises as a consequence of the one-dimensional analysis. But
it is necessary to specify that atmospheric pressure is maintained some fi-
nite distance from the well, and this distance is referred to as the "radius of
influence". A useful feature of equations [3.21b] and [3.22b] is that well
flowrate predictions are not especially sensitive to changes in this distance
over a reasonable range of values (15 to 60 m (50 to 200 ft)). However,
some practitioners have inappropriately associated Ratm with a measure of
the remedial effectiveness of a vertical extraction well. Consequently,
equations [3.21b] and [3.22b] are often improperly used to empirically de-
termine some measure of remedial effectiveness from pilot-test data.
Despite their limitations, the simplistic one-dimensional solutions are
useful in identifying parameters that affect air-flow behavior as well as in
estimating ranges of possible behavior that might be observed in the field.
For example, the one-dimensional radial flow solutions give a useful first-
order approximation of behavior that is typically seen in the field with verti-
cal well systems. The characteristic shape of the spatial pressure
distribution shown in figure 3.14 (on page 3.35) indicates that measured
subsurface vacuums are likely to decrease rapidly within short distances of
3.34
-------
Chapter 3
Figure 3.14
Predicted and Measured Steady-State Radial Pressure Distributions
q,
3?
100
-100
Well Screen Depth 45-50 ft BGS
Estimated Radius of Influence = 60ft
Prediction - Injection Test
t
Prediction - Vacuum Test
20 40
Distance from \acuum/Injection Well (ft)
60
Reprinted by permission of CRC Press from "Soil Venting at a California Site- Field Data Reconciled with Theory
Hydrocarbon Contaminated Soils and Groundwater Analysis, Fate, Environmental and Public Health Effects" by PC
Johnson, C C Stanley, D L Byers, D.A Benson, and M A Acton, Remediation, Vbl I, editors, PT Kostecki and E J
Calabrese, Lewis Publishers Copyright 1991 by CRC Press
the extraction well. From this, one can deduce that subsurface soil gas pres-
sure monitoring probes should be installed relatively close to extraction/
injection wells, if a significant reading is expected. In addition, as the volu-
metric flow rate is predicted to increase roughly in proportion to the ap-
plied vacuum (for low to moderate vacuums) and the soil permeability, then
estimates of the subsurface permeability at a site can be used with equation
[3.22b] to approximate flow rates that might be achieved in a field-pilot test
(see figure 3.15 on page 3.37). In this figure, Q* (standard ftVmin) denotes
volumetric flow rates normalized to atmospheric pressure (i.e.,
\atm)
3.35
-------
Process Identification and Description
3.2.5 Two-Dimensional Vapor Flow Scenarios
Two-dimensional analyses result from the retention of an added spatial
dimension or time. For a homogeneous porous media and transient radial
flow, the approximate solution (for P2 = PPatm) to equation [3.18] is
(Johnson et al. 1990):
Patm~ PW = 4*H(k/n) )e^dX [3'23]
4t/>omf
where all parameters are defined above. If ( ' e^ )<0.1, then equation
[3.23] can be linearized to the form:
P - P =
aim w
Q
- 0.5772-]
[3.24]
Equations [3.23] and [3.24] can be used to estimate the time to establish
steady flow or to estimate the integrated subsurface air permeability from
transient field subsurface pressure measurements. Equation 3.24 actually
predicts that steady-state is never truly reached, so it is necessary to look at
the time it takes to achieve some percentage of the long-term value. For
relatively permeable soil types (e.g., sands and gravels) it is predicted that
steady-state flow conditions are achieved in a matter of minutes. Steady
conditions will take longer to establish in clayey soils; however, typical
transient responses reach steady conditions in a matter of minutes or hours.
This valuable information allows one to focus on predicting steady flow
fields when using more sophisticated flow models. It should be noted that
the time to reach steady conditions also increases with distance from the
extraction well.
Other two-dimensional solutions are found by adding the depth dimen-
sion, z, in the polar coordinate system. This allows development of a more
representative steady-flow model. Assuming that the unsaturated zone is
homogeneous and anisotropic with respect to air-phase permeability (i.e.,
vertical and horizontal components of air permeability are defined and are
constant over the domain), the air-flow equation [3.18] for the axisymmetric
(single well) two-dimensional polar coordinate system reduces to:
3.36
-------
Chapter 3
Figure 3.15
Predicted Steady-State Flow Rates (Per Unit Well Screen Depth) for a
Range of Soil Permeabilities and Applied Vacuums (Pw)
100
10 -
01 -
.001 -
.0001
Pw = 0.40 atm = 20 3 ft H2O
Pw = 0.60 aim = 13.6f(H2O
Pw = 0 80 atm = 6.8 ft H2O
P» = 0 90 atm = 3.4 ft H20
'w = 0.95atm=1.7ftH2O
Medium
Sands
Coarse
Sands
.01
.1 1 10
Soil Permeabilty (darcy)
100
- 110
1100
- 0.11
E
I
?
a
- o.on
0.0011
1000
[ft H2O] denote vacuums expressed as equivalent water column heights
Reprinted by permission of the National Ground Water Association from "A Practical Approach to the Design, Operation,
and Monitoring of In Situ Soil-Venting Systems" by PC Johnson, C C Stanley M.W Kemblowski, DI Ryers, and J.D
Colthart, Ground Water Monitoring Review, Spring, 1990 Copyright 1990 by the Ground Water Publishing Company
- 0
[3.25]
where, r and z are polar coordinates aligned along the major axes of the air
permeability tensor with radial and vertical components kr and kz (cm2).
A computer code recently developed by Joss and Baehr (1993a), which is
in the public domain, implements solutions to equation [3.25] subject to
various characterizations of the land surface boundary condition. This code
3.37
-------
Process Identification and Description
is an adaptation of the groundwater flow code: MODFLOW. The solutions
are derived according to the method presented by Baehr and Hull (1991).
Code output includes air pressure and flow vectors in the simulated domain
surrounding a single injection or extraction well. Analytical solutions for
two-dimensional flow to an extraction trench are also presented by Johnson
and Ettinger (1994), who derived their solutions using conformal mapping
procedures, and Shan, Falta, and Javandel (1992). Numerical solutions of
two-dimensional flow problems are presented in a series of articles by Wil-
son and others (Gannon et al. 1989; Gomez-Lahoz, Rodriguez-Maroto, and
Wilson 1991; Mutch and Wilson 1990). Flow lines depicted in these ar-
ticles are useful in developing a practical intuition concerning air-flow dy-
namics.
3.2.6 General Air-Flow Modeling Considerations
The most general steady-state modeling and design problem is, of
course, fully three dimensional. A three-dimensional model allows simulta-
neous simulation of multiple wells, heterogeneity in air-phase permeability
due to unsaturated zone stratigraphy, and variable moisture content, as well
as a variety of boundary conditions. The steady-state air-flow equation for
a three-dimensional, cartesian coordinate system is as follows:
= 0 [3.26]
where x, y, and z are cartesian coordinates aligned along the major axes of
the permeability tensor with components kix, kyy, and kzz. A computer code
developed by Joss and Baehr (1993b), also in the public domain, solves
equation [3.26], subject to a variety of boundary conditions that can be
encountered at field sites. The code was developed by adapting the U.S.
Geological Survey groundwater flow simulator, MODFLOW, when the
similarity between the groundwater flow equation and equation [3.26] was
recognized. Other computer codes, such as CSUGAS (Sabadell, Eisenbeis,
and Sumada 1988), developed at Colorado State University, are also avail-
able for similar purposes.
3.38
-------
Chapter 3
Analogous to the use of a groundwater flow model in quantifying aquifer
flow, an air-flow model can be used in a calibrative mode or a predictive
mode. When it is used in a calibrative mode, the objective is to estimate the
spatial distribution of air permeability in the unsaturated zone. Air perme-
ability depends on location because of two factors: (1) the nature of the
lithologic unit (e.g., coarse or medium sand) and (2) the occupation of void
space by the aqueous and immiscible phases. An air-flow model can be
calibrated against data collected during a pneumatic pump test (typically
involving a single well) to obtain air permeability estimates. When used in
a predictive mode, the spatial distribution of air permeability is assumed,
and a model can be used to predict air pressure and flow over the domain
under many proposed well configurations in order to aid in determining an
optimal system design. Pressure vs. discharge relationships at extraction
wells can be simulated to assist in the selection of pumps. The effect of
heterogeneity in air permeability, caused by layers of different lithologic
units on the distribution of flow velocity in the domain, can be simulated to
estimate cleanup times. A model used in the predictive mode is able to
calculate a zone of influence based on flow velocity, rather than on an arbi-
trary criterion based on measured air pressure.
3.2.7 Transport Considerations and the Estimation of Re-
moval Rates and Residuals
As mentioned previously, the two main factors that govern the perfor-
mance of vapor extraction-based processes are: (1) chemical partitioning to
the vapor phase (see Subsection 3.2.2) and (2) vapor flow behavior (see
Subsections 3.2.3 through 3.2.6). In combination, they determine subsur-
face hydrocarbon transport and consequently, the removal rates obtained
with any vapor extraction-based system. Subsections 3.2.7 through 3.2.11
examine different ways to combine the information presented in Subsec-
tions 3.2.2 through 3.2.6 in order to estimate the performance of vapor ex-
traction-based processes. In Subsections 3.2.4 through 3.2.6, the general
approach is outlined, and varying levels of simplification are discussed.
Upon review of the literature, one finds that many different mathematical
formulations have been used to describe the vapor-phase transport of
chemicals in porous media. Most models account for convective and diffu-
sive transport in similar ways. The differences, however, primarily result
3.39
-------
Process Identification and Description
from assumptions used to describe local chemical equilibria and local mass
transfer rates between various chemical phases.
The most mathematically convenient approach is to assume that there are
no local mass transfer resistances and that partitioning between various
chemical phases can be described in terms of strictly linear equations. In
this case, the equations governing chemical transport are identical to those
commonly used in groundwater transport modeling, and any of a number of
existing groundwater contaminant transport tools can be used to predict the
performance of vapor extraction systems.
Unfortunately, chemical partitioning in soils does not always behave as
predicted by mathematically-convenient linear models, especially in the
concentration range for which vapor extraction technologies are most com-
monly applied. In the approach described below, a more complete math-
ematical approximation of chemical partitioning in soil matrices is
presented. The assumption of local equilibrium, however, is retained. This
assumption is justified since the intent is to develop models that will predict
optimal vapor extraction system performance. It is implicit that any real
system will perform less effectively.
The degree to which local mass transfer rates limit vapor extraction per-
formance is unknown at this time. The mathematical framework can be
easily expanded to account for local resistances through the introduction of
a number of mass transfer coefficients. In the absence of data, however,
most users will be forced to (arbitrarily) select values for these parameters.
Johnson et al. (1989) presented an analysis of vapor-phase equilibrium on
the pore scale, and concluded that vapor-phase mass transfer resistances
would be negligible for typical subsurface vapor velocities induced by soil
vapor extraction. Intraparticle mass transfer resistances have been ob-
served, however, during soil-water partitioning experiments. It is possible
that these could play a significant role in limiting vapor extraction perfor-
mance in the instance of low-contaminant concentrations. In Subsection
3.2.11 macro-scale mass transfer limitations resulting from heterogeneous
geology and contaminant distributions, are discussed together with the ef-
fect that these have on vapor extraction performance.
Whether or not the user is operating under a system controlled by signifi-
cant mass transfer resistances impacts the optimal operating strategy that is
used. In the absence of mass transfer resistances (macro- or micro-scale),
removal rates always increase with increased flow rates. Thus, the system
3.40
-------
Chapter 3
operates at the maximum flow rate achievable by the system. In this case,
systems can be designed based on economic constraints, performance goals
(remediation time, cleanup level etc.), and a design parameter that repre-
sents the minimum volume of vapor required to achieve the desired degree
of remediation. The latter is determined form chemical equilibria models,
or laboratory soil column experiments. This approach is outlined in Sub-
section 3.2.9.
When mass transfer resistances govern system performance, the operat-
ing strategy is quite different. The user will find that the removal rate does
not always linearly increase with flow rate. In fact, as the flow rate is in-
creased the user will eventually observe minimal changes in removal rate
with increasing flow rate. This is generally reflected by decreases in offgas
concentrations with increases in extraction flow rate. Thus, the operating
strategy will be to find the minimum flow rate that achieves the maximum
removal rate.
Over the course of most vapor extraction projects, users will find that
both operating strategies need to be used. During the initial phase of opera-
tion, most systems perform as if there are no mass transfer limits. Later,
mass transfer resistances become more important.
Again, the reader is cautioned that local equilibrium is assumed in the
following mathematical development, and it is not implied that mass trans-
fer resistances do not play important roles at some sites. Rather, the intent
is to develop screening-level models that will predict optimal vapor extrac-
tion system performance. It should be understood that any real system will
perform less effectively. Such models are often used to identify sites, or
conditions, for which vapor extraction is not likely to be successful.
In the most general sense, prediction of the response in the subsurface to
the installation and operation of a vapor extraction-based process requires
prediction of the induced air-flow field, local equilibrium partitioning of
hydrocarbons, and the rate of hydrocarbon transport through the subsurface.
The first two are predicted by equations presented in Subsections 3.2.2
through 3.2.6, and the latter is typically described by the equation:
[3.27]
3.41
-------
Process Identification and Description
for each chemical component. Here T (g-j/cm3-soil) is the total volumetric
content of component j, q (cm/s) is the specific discharge (Darcy velocity)
vector, G (g-k/cm3-vapor) is the vapor-phase concentration of k, D (crnVs)
is the effective dispersion/diffusion coefficient, and A/(T) (g-j/cm3-soil/s)
represents losses due to degradation. Underlying equation [3.27] is the
assumption that convection occurs most significantly in the vapor phase and
that vapor-phase dispersion is the most significant dispersive mechanism,
Solution of a three-dimensional transient version of equation [3.27] would
provide the most complete tool for predicting vapor extraction-based sys-
tem performance; however, a model that combines a multiconstituent equi-
librium model with three-dimensional air flow is not currently available in
the public domain. Even if such a model were available, it might be too
costly because of the data and time it would require for application. As
with numerical air-flow models, it may be that detailed numerical transport
codes are best used for visualization purposes and developing an under-
standing of what may occur for a given scenario, rather than general appli-
cation at all sites.
3.2.8 Nondimensional Solutions - Constant Flow/Constant
Vapor Concentration
The simplest of all transport models is a nondimensional model that
combines vapor concentration G (g-j/cm3) and flow-rate estimates Q (cm3/
s) to produce first-order removal rate estimates. The removal rate of com-
pound j, R (g-j/s), and time to achieve cleanup T (s) are given by:
R — QG (nondimensional estimate) [3.28]
T = VSO!l I Rj (nondimensional estimate) [3.29]
Here, Q and G are calculated as explained in Subsections 3.2.2 through
3.2.6, and V denotes the volume of soil (cm3) containing the average total
volumetric contaminant concentration (g/cm3). While this model ne-
glects a number of significant phenomena and is not likely to ever produce
accurate estimates for sustained vapor extraction system performance, it
does provide a quick means of estimating maximum system performance;
no system could ever perform better than this reference performance. The
3.42
-------
Chapter 3
removal rate given by equation [3.28] might also be regarded as an estimate
of what might be observed when a system is first turned on, before compo-
sitional changes and mass-transfer limitations act to reduce the removal rate
with time.
3.2.9 One-Dimensional Models - Equilibrium-Based Well-
Mixed Systems
The next level of sophistication in modeling removal due to volatiliza-
tion is to incorporate compositional changes with time. In this approach,
the equilibrium partitioning model described in Subsection 3.2.2 is used and
the subsurface is treated as a well-mixed system. Changes in the amount of
residual contaminant over time are then described by:
j- = -QGJ - A/(7y ) [3.30]
where all variables are as defined above, with A/(T ) representing losses due
to biodegradation. This approach has been used by Baehr, Hoag, and
Marley (1989) to describe the variation in effluent concentration for col-
umn-venting experiments with gasoline-contaminated sand. Figure 3.16
(on page 3.44) is a plot of the total hydrocarbon removal rate as a function
of time for one of the experiments. In this experiment, airflow was held
constant and all air passing through the column was in contact with gaso-
line-contaminated pore space. The decline in removal rate was due entirely
to weathering of the residual gasoline; as more volatile hydrocarbons were
removed, total hydrocarbon concentration in the pore space declined be-
cause of compositional changes. Humps in the model prediction curve
presented in figure 3.16 (on page 3.41) are due to removal of constituent
classes from the residual product. This approach was used also by Johnson,
Kemblowski, and Colthart (1988) and Johnson et al. (1990, 1992) to de-
velop a screening-level design criteria for vapor extraction systems. They
used the model to predict a minimum vapor volume requirement a (cm3-
vapor/g-initial residual) to achieve a desired level of cleanup. Predicted
values of a, for gasoline, for example, range from 25,000 to 125,000 cm3-
vapor/g-initial residual (400 to 2,000 ft3-vapor/lb-initial residual), depend-
ing on the grade of gasoline and degree of preweathering in the subsurface.
The minimum cleanup time estimate T is given by:
3.43
-------
Process Identification and Description
Q
[3.31]
Alternately, one could use equation [3.31] to establish a minimum vapor-
flow requirement Q to achieve cleanup in a specified time. The reader is
cautioned that equation [3.31] is a lower-bound estimate of the cleanup time
and is intended only for screening-level evaluation.
If it is assumed that chemical equilibria between phases in the soil matrix
are described by linear relationships and that the system is well-mixed, then
equation [3.30] can be solved. It can be observed that this solution will
predict soil and extracted vapor concentrations that decrease exponentially
with time, as presented by Roy and Griffin (1991).
Figure 3.16
Comparison of Total Hydrocarbon Fluxes from Steady-Flow Venting
Experiments to Predictions Obtained from the Mathematical Model
2.5 ,.
2.0 -
a
I 15
« 1.0
I
20
40 60
Time, in Minutes
80
100
Flow rate = 1.43 L/mm
Source. Baehr, Hoag, and Marley 1989
3.44
-------
Chapter 3
3.2.10 Two-Dimensional Model - Cleanup Along Streamlines
The concept introduced in Subsection 3.2.9 of a minimum vapor volume
requirement a (cm3-vapor/g-initial residual) to achieve a desired level of
cleanup can be built upon to develop a transient model for remediation
along a streamline, or flow path. In this model, clean air enters at the edge
of the contaminated zone and a "cleaning front" propagates toward the ex-
traction point.
Let a (cm3-vapor/g-initial residual) denote the volume of vapor required
to achieve the desired remediation per unit mass of residual hydrocarbon.
This parameter may be determined from a lab experiment or by modeling
(as described above). Then, if (g/cm3-soil) denotes the volumetric
hydrocarbon concentration, q (cm/s) the specific discharge, t (s) time, and !;
the location of the cleaning front, then:
[3.32]
dt a
Consider one-dimensional radial flow to a vertical well in which -q = Q/
(2itrH), and Q (cmVs), r (cm), and H (cm) denote the total volumetric flow
rate to the well, distance from the well, and thickness of the permeable
zone, respectively. Equation [3.32] becomes:
dt 2nrH a
The solution is:
[3.33]
nHa
[3.34]
where Ro is the radius of the immiscible contamination away from the ex-
traction point at t = 0. Equation [3.34] implies that the time required to
clean the flow path T for this example is given by:
nR2H a
[3.35]
3.45
-------
Process Identification and Description
which the reader can verify is equivalent to equation [3.31].
A two-dimensional analysis using proof similar to that of the above one-
dimensional analysis was used by Johnson and Ettinger (1994) to predict
the performance of a vapor extraction trench.
3.2.11 Mass-Transfer Limitations, Nonideal Scenarios, and
Transient Effects
The performance of vapor extraction systems under ideal conditions
where the induced vapor flow passes through the contaminant source and
there is equilibrium partitioning was estimated in the preceding sections.
Scenarios where this may not be the case are considered in this section.
Figure 3.17 (on page 3.47) illustrates three common scenarios where the
removal rates inherently must be less than that predicted by the ideal mod-
els presented above. Geological heterogeneities also frequently result in
non-ideal conditions.
In the scenario depicted in figure 3.17a, a fraction <|> of the air drawn
through the soil bypasses the contaminant zone as a result of nonhomoge-
neous contaminant distribution or improper extraction well placement.
Clean air reaching the extraction well then dilutes the contaminant-laden
vapors in the extraction well. An estimate of the removal rate R (g-j/cm2-s)
is then:
# = (1-0)*™ [3.36]
where Rmax (g-j/s) represents the removal rate estimated for the ideal case.
The corresponding time for cleanup T (s) estimate is increased relative to
the minimum cleanup time estimate T (s) by:
T = Tmm [3.37]
In the scenarios of figures 3.17b and 3.17c, the induced vapor flow path
parallels the residual contaminant but does not pass through it. This might
occur with flow past pooled liquid contaminant (figure 3.17b) or with con-
taminants trapped in low-permeability zones surrounded by more permeable
layers. In these instances, the removal rate is limited by vapor-phase diffu-
3.46
-------
Chapter 3
Figure 3.17
Limiting Model Scenarios for Removal Rate Estimates
a)
Vapor Flow
Vapor Flow
Vapor Flow
SideView
Top View
b)
vapor concentration
profile
Vapor Flow
Vapor Concentration = 0
gy»5ijgisit ^^^StSm **!
1I3B«C, •• ?; ^^.sTV^JPSrt^i
Impermeable Layer _. .. _ '
Liquid Contaminant
C)
"Wet" Zone with Residual Contamination
Reprinted by permission of the National Ground Water Association from 'A Practical Approach to the Design, Operation,
—•..—*-_:„ -i,- **. o.:,,..-..--o u.. ^ lohnson, C.C.Stanley, M.W.Kemblowski, D.I.Rysrs, andJ.D.
I. Copyright 1990 by the Ground Water Publishing Company
nQpi H nou ujr ^n>
-------
Process Identification and Description
sion, which is usually described in terms of an effective diffusion coeffi-
cient Deff (cm2/s). The diffusive flux F. (g-j/cm2-s) across an incremental
distance Ax (cm) can be written:
Ax
[3.38]
where AG} (g-j/cm3-vapor) denotes the change in vapor concentration across
that distance.
For the scenario depicted in figure 3.17b (on page 3.47) and one-dimen-
sional radial-flow, Johnson et al. (1990) solved equation [3.27] to derive an
efficiency factor r| relative to the ideal case (R = QG) for a single compo-
nent contaminant pool:
(P -P }
\ atm well /
?2-*,
[3.39]
where:
Deff = effective diffusion coefficient (cm2/s),
H = thickness of zone permeable to airflow (cm),
(4. = vapor viscosity (g/cm-s) ==0.00018 g/cm-s,
k = permeability to vapor flow (cm2),
Ratm = radius of pressure influence of extraction well (cm),
R,,,,,,, = well radius (cm),
= atmospheric pressure (1.016 x 106 g/cm-s2),
= absolute pressure at venting well (g/cm-s2), and
R, < r < R2 defines the region in which contaminant is present.
As an example, consider a 5.1 cm (2 in.) radius vapor extraction well
installed in a sandy soil (k = 1 x 10~8 cm2 (1 x 10" ft2)) with a blower con-
nected to maintain a 0.90 atm (0.91 x 106 g/cm-s2 (1.5 lb/in.2 = 41 in. H2O
gauge-vacuum)) pressure at the well. The objective is to remediate a zone
well
well
3.48
-------
Chapter 3
extending to R2 = 900 cm (30 ft) from the well. Assuming that H = 300 cm
(10 ft), Ratm = 1,200 cm (40 ft), and Deff = 0.0014 cm2/s (1.5 x 10'6 ft2/s) then
equation [3.39] predicts the removal rate will be only 9% of the ideal case
removal rate.
Figure 3.17c (on page 3.47) is representative of the situation wherein a
clay lens containing residual contamination is surrounded by more perme-
able sandy soils. In the model of Johnson et al. (1990), a drying front de-
velops in the less-permeable zone and grows away from the permeable/
impermeable soil interface. Thus, the distance over which vapors diffuse to
reach the flowing-vapor stream increases with time; consequently, as pre-
dicted by equation [3.38], the rate of removal R (g/s) decreases with time:
[3.40]
where all variables are as defined above. The corresponding drying zone
thickness 5 (cm) is given by:
Figure 3.18 (on page 3.50) presents results of a sample calculation for
benzene (G = 0.00032 g/cm3 @20°C (0.02 lb/ft3)) at a residual level of T =
0.017 g/cm3 (1.1 lb/ft3) and R2 = 900 cm (30 ft). These results indicate U
would take approximately one year to clean a 150 cm (5 ft) layer.
Equation [3.41] predicts a removal rate vs. time dependence that is very
similar to what is observed at many vapor extraction sites, although the
decline with time is typically attributed to changes in the residual composi-
tion. Diffusion limitations can also produce the same apparent behavior;
thus, care must be taken to not over interpret field data. For example, vapor
composition analyses with time are necessary to distinguish between the
two phenomena; mass-transfer limited scenarios are often characterized by
large decreases in removal rate without accompanying significant changes
in composition.
3.49
-------
Process Identification and Description
Figure 3.18
Estimated Maximum Removal Rates and Residual Hydrocarbon
Reduction for a Venting Operation Limited by Diffusion.
1000
100.
Benzene (20'C)
RI =5.1 cm
R2 =900 cm
200
• ioo
100
200 300
Time (d)
400
500
Reprinted by permission of the National Ground Water Association from "A Practical Approach to the Design, Operation,
and Monitoring of In Situ Soil-Venting Systems" by PC Johnson, C C Stanley, M W Kemblowski, D L Byers, and J D
Colthart, Ground V&ter Monitoring Review. Spring, 1990 Copyright 1990 by the Ground Water Publishing Company
It is also possible to observe very rapid declines in removal rate with
time during the first few days of operation of a vapor extraction system that
is not attributed to either phenomenon discussed above. Usually, this is an
indication that the airflow has swept those contaminant vapors that had
originally diffused away from the vapor source back to the extraction point.
Once the first few "pore volumes" of vapors are removed from the subsur-
face, decreases in removal rate with time tend to be less drastic, unless ac-
companied by significant changes in water-table height, vapor flow-rate,
etc.
3.2.12 Microbiological Processes (Bioventing)
As explained in Section 3.1, inducing vapor flow through the subsurface
not only increases removal rates through volatilization, but also enhances
aerobic microbial degradation as a result of accelerating the resupply of
oxygen to the subsurface. Bacteria utilize oxygen to break down degrad-
3.50
-------
Chapter 3
able contaminants into carbon dioxide, water, and biomass (cells); for ex-
ample, consider the degradation of n-hexane:
C6HH + — O2 -» 6CO2 +1H2O + biomass [3.42]
Based on equation [3.42], 3.5 g-O2 are required per gram of n-hexane
degraded. In recent years, numerous laboratory-scale investigations have
been conducted in order to study the aerobic degradation of hydrocarbons
and the effects of changes in moisture content, nutrient addition, contami-
nant concentration and composition, and other process variables on degra-
dation rate. Results indicate that many contaminants of concern are
biodegradable and indigenous microorganism populations can carry out the
task, if subsurface conditions are amenable. In this monograph,
bioremediation is discussed only as it relates to soil vapor extraction. The
reader is referred to the monograph in this series, Innovative Site
Remediation Technology: BIOREMEDIATION, for an in-depth presentation
of bioremediation.
Quantifying degradation rates in the field can be difficult, and the data
are often subject to a range of interpretations. As equation [3.42] indicates,
assessment can be based on contaminant disappearance (soil sampling),
oxygen gas depletion (soil-gas sampling), or carbon dioxide gas formation
(soil-gas sampling). Since there is inherently a large degree of uncertainty
associated with the first option, reported bioventing rates are typically based
on results of soil-gas analyses combined with theory. The monitoring and
interpretation of field data are discussed in detail in Section 3.6. Final veri-
fication is generally based on soil core analysis after prolonged operation
(months).
Microbiological activity has been reported at temperatures varying from
-12° to 100°C (54° to 212°F) (Brock, 1970); however, the optimal range for
biodegradation of most contaminants is much narrower. Individual micro-
organisms may tolerate a temperature range of up to about 40°C (104°F).
Generally, biodegradation rates double for every 10°C (5^°F) temperature
increase, up to some inhibitive temperature. The van't Hoff-Arrhenius
equation expresses this relationship quantitatively as:
Y = Ae'E'/RT [3.43]
3.51
-------
Process Identification and Description
where:
Y = temperature-corrected biodegradation rate,
A = baseline reaction rate,
Ea = activation energy,
R = gas constant, and
T = absolute temperature.
Miller (1990) found Ea equal to 8 to 13 kcal/mole for in situ biodegrada-
tion of jet fuel. In the 17° to 27°C (63° to 8TF) range, the van't Hoff-
Arrhenius relationship accurately predicted biodegradation rates.
Figure 3.19 (on page 3.53) illustrates the effect of temperature on JP-4
jet fuel biodegradation in soils collected from Eielson Air Force Base near
Fairbanks, Alaska. See also figure 3.20 (on page 3.55). The colder-region
organisms biodegrade jet fuel at higher rates at 20°C (68°F) than do organ-
isms adapted to temperate climates.
3.2.13 Air Injection Below the Groundwater Table
The capacity of an air-sparging system for volatilizing hydrocarbons in
the saturated zone or resupplying an aquifer with oxygen is dependent upon
the way in which the air is distributed when injected into the saturated zone.
Although it is convenient to visualize uniform bubbles simply rising upward
and outward from air injection points, the actual air distribution patterns are
likely to be quite complex. A working description of the likely flow and
distribution of injected air and its effect on system performance follows.
In order for air to be injected beneath the groundwater table, air must
first displace water from the injection well, the air-sparging well packing
(or diffuser), and the formation. At a minimum, therefore, an initial pres-
sure Pmin (g/cm-s2) is required equal to:
"min = Pv/aterS"water + "diffuser "*" "soil [3-44J
where:
Pwater = density of water (1 g/cm3),
g = acceleration due to gravity (980 cm/s2),
3.52
-------
Chapter 3
Figure 3.19
Biodegradation of JP-4 Jet Fuel in Soils Collected Near Fairbanks, Alaska
Days
Source: Wyzaand Hinchee 1990
H
water
p
diffuser
= depth below water table to top of injection well screen
(or diffuser) (cm),
= air-entry pressure for injection well packing or diffuser
(g/cm-s2), and
= air entry pressure for the formation (g/cm-s2).
3.53
-------
Process Identification and Description
Equation [3.44] can also be written in terms of pressure head, where the
pressure head H., expressed as a height of water column is related to the
pressure P by:
H.t = '— [3.45]
rwatero
Equation [3.44] becomes:
#min = H water + Hdiffuser + Hso,l t3-46]
and Hmm, Hdiffuser, and Hwil denote the minimum pressure head for sparging,
air-entry pressure head for the diffuser (or packing), and air-entry pressure
head for the formation, respectively. For reference, 1 atm = 1 x 106 g/cm-s2
(33.9 ft H2O).
The air-entry pressure required varies with the type of soil, with values
ranging from approximately 1 cm (0.4 in.) for coarse sands to >1 m (1.1 yd)
for very fine-grained soils, such as silts and clays. In practice, one often
encounters layered lithologies.
The relationship between the three factors dictating the minimum
sparging pressure, Hnun, has implications also for the distribution of air in
the vicinity of the sparging well. When Hwa)er is much greater than Hdiffuser
and H^, it is likely that air will enter the formation primarily near the top
of the sparging well screen. Thus, if standard well constructions (i.e., slot-
ted pipe surrounded by a sandy packing material) are used in coarse soils, it
does not make much sense to extend the well screen to more than about 1
meter. To achieve more uniform distribution of air entry across the air
injection well, therefore, it is necessary to increase the value of Hdiffuser,
which can be accomplished either by increasing the resistance to airflow
across the well screen (using a diffuser) or by using a much finer well pack-
ing material.
As air enters the formation at the injection well, it displaces water and
begins to travel outward and upward toward the vadose zone as a result of
the applied pressure, buoyancy forces, and macroscale characteristics of the
formation. Normally, uniform displacement fronts are envisioned when one
fluid displaces another in a porous medium; however, in this case a much
3.54
-------
Chapter 3
different air-flow path results because the air is much less viscous than
water and the formation is more permeable to airflow than water under
typical sparging conditions. The air is said to have greater "mobility" than
the displaced water and therefore, the flow is "unstable" because even very
small heterogeneities in the formation cause preferential air-flow channel-
ing through "fingers" (see figure 3.21 on page 3.56). Channeling becomes
more pronounced as the degree of heterogeneity increases. Air injection
may also cause soil fracturing and weakening of the soil stability.
Figure 3.20
Biodegradation of JP-4 Jet Fuel at 2CTC in Various Soils Collected Near
Fairbanks, Alaska; Fallen, Nevada; and Panama City, Florida
Days
Source. Wyza and Hinchee 1990
3.55
-------
Process Identification and Description
Large-scale variations in the formation, such as those caused by distinct
strata or lenses, where the less-permeable (and higher air entry pressure)
strata are not continuous and the air passes around the units as it finds a
pathway through the saturated zone, will also affect the air distribution, as
illustrated in figure 3.22 (on page 3.57). Now consider the effect on the
path when air is injected into a stratum of greater permeability than the
continuous stratum lying above it. An air "bubble" will spread horizontally
beneath the less permeable layer until the pressure in the bubble increases to
the air-entry pressure of the confining unit or until it intercepts a permeable
vertical pathway (such as a monitoring well). The practitioner must be
aware that this vapor-flow distribution may actually result in increased
transport of hydrocarbon away from the source.
Figure 3.21
Observed Air Channel Pattern in Uniform Mixture of
0.75 and 0.3 mm Glass Beads
Overburden With
4mm Beads
Mixture ofTVo
Size Beads
Source: Jietal. 1993
3.56
-------
Chapter 3
Figure 3.22
Observed Air Channel Pattern in Stratified Medium
Source: Ji et al. 1993
The foregoing discussion may help provide an intuitive picture of air
distribution and flow resulting from air injection beneath the water table
and roughly identify inherent advantages and limitations of this technology.
Given current limitations, the exact numbers, size, or spacing of air chan-
nels that are likely to be formed during air sparging cannot be predicted.
Nonetheless, the formation of distinct air channels (as opposed to bubbles)
significantly affects the capacity of the system for maximizing the rate of
volatilization and enhancing microbial degradation of residuals trapped
beneath the groundwater table. In the flow visualization work of Ji et al.
(1993), small bubbles were observed only when glass bead diameters were
>2 mm (0.08 in.). Given the transport discussion in Subsections 3.2.8
through 3.2.11, effective remediation of residuals trapped within the air
channels can be expected, but removal of contaminants from water satu-
3.57
-------
Process Identification and Description
rated zones will be limited severely by diffusion limitations, unless signifi-
cant mixing of the water-saturated zones is induced by the airflow (e.g.,
pulsed sparging). Similarly, the transport of oxygen into the water-satu-
rated formation will be limited by diffusive processes. The removal of
hydrocarbons from low permeability unsaturated formations by diffusive
processes was discussed in Subsection 3.2.11 (see equation [3.40], for ex-
ample). That scenario is similar to that of the volatilization of hydrocarbons
from a water-saturated zone into an air channel. Removal rates from water-
saturated zones, however, are expected to be much lower, as diffusion coef-
ficients in water are four orders of magnitude less than diffusion
coefficients in air.
3.3 Characterization Activities
Competent characterization of the site and assessment of the hydrocar-
bons present are needed to provide sufficient information to assess the fea-
sibility of vapor extraction-based technologies and to design effective
remedial systems. As will be shown in Section 3.4, there are a number of
different methods for assessing feasibility and designing systems, and each
practitioner must select the approach that best meets the requirements and
constraints of a given site. Data requirements vary from site to site and
consequently, a fixed set of actions will not be prescribed here. Instead, a
range of characterization steps will be presented in this section from which
the reader may select after identifying his or her particular data require-
ments based on the guidance given in Section 3.4. In addition, this section
will illustrate the effective presentation of information so that it can be used
in the feasibility screening and design process described in Section 3.4.
A sequence of characterization activities is outlined in table 3.3 (on page
3.59), in order of priority based on needs for feasibility screening and de-
sign and, to some degree, on the difficulty of obtaining the information.
Each of the characterization activities is discussed and a few common meth-
ods for performing each are briefly described below; however, the reader
should note that there is a wide range of tools that can be used to meet many
of the data collection objectives. For a general introduction to site assess-
ment activities, see the API Publication 1628: A Guide to the Assessment
and Remediation of Underground Petroleum Releases.
3.58
-------
Chapter 3
3.3.1 Contaminant Assessment
The goal of the contaminant assessment is to: (1) develop a "picture" of
the subsurface, such as that shown in figure 3.23 (on page 3.60), where the
contaminant distribution is superimposed upon the geological cross-section
Table 3.3
Site Characterization Activities
Activity #
Description
Data Reduction and
Presentation
Requirements
Preliminary Characterization Activities
1 Hydrocarbon Assessment (See §3 3 1) See figure 3.23
• vertical/horizontal delineation See figure 3.24
• hydrocarbon characterization (type, boiling point
distribution, regulated component identification)
2 Geologic/Hydrogeologic Assessment (See §3 3 2) See figure 3 23
• identification of soil strata
• permeability assessment (core (ests, sieve analysis, etc )
• static wrier table determination (and seasonal fluctuations)
• subsurface conduits, piping, tanks, obstructions, etc
Laboratory Characterization Activities (See §3.3.3)
3 Laboratory Soil Column Feasibility Studies (optional) See figure 3 26
Field Pilot-Scale Activities (See §3.3.4)
4 Airflow-vs-Applied Pressure/Vacuum Test See figure 3 28
• vacuum test for vapor extraction wells
• pressure test for air injection wells
5 Effluent Vapor Characterization-vs-Time See figure 3 30
• total hydrocarbon concentrations
• regulated compound speciation
• hydrocarbon characterization (i.e. boiling point distribution)
• O2/CO2 speciation
6 Subsurface Pressure Distribution See figure 3.32
• as function of depth and distance
• steady-state and transient measurements
7 Subsurface Vapor Concentration Distribution See figure 3.33
• as function of depth and distance
• hydrocarbon concentrations and composition
• O2/CO2 speciation
8 Oroundwater Elevation Changes Resulting from See figure 3.35
Air Extraction/Injection
9 Groundwater Monitoring See figure 3.36
• hydrocarbon levels
• dissolved oxygen
10 Tracer Gas Tests See figure 3 37
3.59
-------
Process Identification and Description
Figure 3.23
Recommended Presentation of Total Hydrocarbon
Distribution and Subsurface Geology
North
South
o—i
10-
20-
30-
40_
50-
60—'
Static Ground
\Vater1able
L
i
I\ Tank
Sandy \ Backfill
Clay \ (former tank
\ location) -
V*.*-
J
f
- 0.02
- 0.0
- 0.0
>
^ r, **
•o.z
t
- 0.0
- 1.7
1
-24 t
- 73
1
L 9.5 '
L \
Fine to
Coarse Sand
'
k
Silty Clay
&
Clayey Silt
r
Medium Sand
1
-0.5 /
m~ 1 ,*7 2.
r*r
- 512
- 5.4
- 8577
-341
- 653
- 3267
- 1237
- 23831
- jjiy
- 1.7
- 0.8
A £__
— -*f.(^- — f — — — — —
- 0.3
- 8.2
- 214
r-31
|- 967
- 971
L 28679
- 23167 -i
- 0.31
- — 1.-2
- 0.44
- 0.17
- 8.8
- - 0.63
- 1.5
- 0.86
- 23
- 1.6
- 3.2 T
HB-17 HB-10 HB-5 HB-3
SCALE (ft)
I 1
0
10
20
Soil concentrations given in mg/kg-soil.
Reprinted by permisssion of CRC Press from 'Soil \6nting at a California Site Field Data Reconciled with Theory.
Hydrocarbon Contaminated Soils and Groundwater: Analysis, Fate, Environmental and Public Health Effects' by PC.
Johnson, C.C. Stanley, D.L. Byers, O.A. Benson, and M A. Acton, Remediation, Vol. I, editors, RT Kostecki and E.J.
Calabrese, Lewis Publishers Copyright 1991 by CRC Press
of the area to be treated and (2), sufficiently characterize the contaminant so
that an assessment of its treatability can be made. Possible contaminant
characterization methods are described and possible approaches for deter-
mining the spatial extent of contaminants are discussed in this subsection
and characterizing the subsurface is discussed in Subsection 3.3.2.
3.60
-------
Chapter 3
Under the current state-of-the-practice, contaminant composition is as-
sessed based on results of laboratory analyses of soil and groundwater
samples obtained from soil borings and groundwater monitoring wells.
Laboratories are most often requested to analyze samples for total petro-
leum hydrocarbons (TPH) and specific compounds of regulatory interest
(e.g., benzene). Recall, however, that the performance of vapor extraction-
based technologies is linked to the partitioning of contaminants in the sub-
surface and the partitioning of any single compound may be dependent on
all others present in the matrix (see Section 3.2). Thus, while typical labo-
ratory analyses have some value from a regulatory perspective, they provide
an incomplete picture for the purposes of feasibility screening and design.
Because a complete analysis of many contaminant mixtures is not practi-
cable (e.g., there are >100 components in a typical gasoline), simpler cost-
effective methods for providing the needed information are sought. One of
these is the characterization of contaminant distributions by boiling point
fractionation, represented in figure 3.24 (on page 3.62). This can be done in
a number of ways; the simple method presented below utilizes laboratory
results that are generated in standard analyses, but are not reported.
In the course of many routine contaminant analyses a liquid or vapor
contaminant sample is injected into a gas chromatograph equipped with a
flame ionization detector (GC-FID). The chromatographic column then
separates contaminant components as they are swept by a carrier gas toward
the detector, which provides an electrical signal reflecting intensity vs. time
to a recorder/integrator. Concentrations are assigned to individual compo-
nents with knowledge of their elution/retention time and intensity vs. con-
centration calibrations. When requested to report the concentrations of
individual components, laboratories focus on specific peaks and ignore the
rest of the chromatograph. If, however, the elution times of a series of
"marker" compounds (e.g., a series of normal alkanes: methane, ethane,
propane, n-butane, etc.), are known then the sum of all contaminants that
elute between two known compounds can be quantified and expressed as a
fraction of the total contaminants detected. For characterization purposes
then, these fractions or groups of compounds, can be treated as "pseudo-
compounds," having chemical properties that are some average of the two
bounding marker compounds. For example, if boiling point ranges are
assigned to these fractions, a "boiling point curve" such as that shown in
figure 3.24 (on page 3.62) can be generated. This information provides a
valuable picture of the distribution of contaminants in a mixture and can be
3.61
-------
Process Identification and Description
used to carry out some of the partitioning calculations outlined in Subsec-
tion 3.2.2. This approach is illustrated by Johnson et al. (1990), who com-
pare predictions for a gasoline mixture of 72 hydrocarbons with that of an
equivalent nine-component approximate mixture.
It should be noted that the characterization method described above is
especially useful in assessing the behavior of complex mixtures in processes
that are governed by component vapor pressures (e.g., soil vapor extraction
of typical residual levels of hydrocarbons); this is because compounds are
roughly separated by vapor pressure in a GC-FID analysis. The same is not
true for processes governed by solubility or partitioning from an aqueous
phase, unless the contaminant mixture is composed of components having
chemical structures similar to the marker compounds. For example, in
Figure 3.24
Boiling Point Distribution Curves for Samples of
Fresh" and "Weathered" Gasolines
o.o
-40
Tb CO
Reprinted by permission of the National Ground Water Association from 'A Practical Approach to the Design, Operation,
and Monitoring of In Situ Soil-Venting Systems" by PC. Johnson, C.C Stanley, M.W. Kemblowski, D I. Ryers, and J 0.
Coltnart, Ground Water Monitoring Review, Spring, 1990. Copyright 1990 by the Ground Water Publishing Company
3.62
-------
Chapter 3
gasoline there is a mixture of aromatic and aliphatic hydrocarbons, of which
the aromatic compounds are by far the most soluble. If the method given
above is followed and fractions are characterized by a series of n-alkanes,
the average solubility value assigned to fractions containing the aromatic
compounds will be a significant underestimation of their true equilibrium
partitioning into an aqueous phase. In the absence of any knowledge of the
contaminant mixture, one empirical method of working around this problem
is to perform a second GC-FID analysis on a "small" aqueous sample that
has been allowed to equilibrate with a "large" hydrocarbon sample previ-
ously characterized by the above method. As in the residual analysis de-
scribed above, the chromatographic output from the aqueous phase analysis
is again divided by hydrocarbon fractions and the concentration in each
fraction is quantified. For partitioning calculations then, the "pseudo-com-
pound" solubilities are set equal to the measured concentrations from this
experiment.
In determining the spatial distribution of contaminants, first, the areal
extent of contamination is assessed (horizontal delineation), and then a
vertical profile of the area is determined. Detailed contaminant composi-
tion analyses (such as that described above) are time-consuming and costly,
but are fortunately needed for only a limited number of samples when in-
vestigating single-source release sites. There are a number of field investi-
gation methods for contaminant delineation that vary in terms of cost, ease,
reliability, ability to provide information on vertical definition, and use of
existing wells. The object is to increase the speed of sample collection or
decrease the complexity and cost of analyses and to this end, the preferred
methods often involve some cost-saving gross measurement of contaminant
concentrations, such as TPH or the detection of some easily identifiable
"marker" compound, such as a highly-volatile or highly-soluble component.
See table 3.4 (on page 3.64) for qualitative assessments of the performance
of a number of these methods, which are briefly described below.
Soil-Gas Surveys. In a soil-gas survey, a hollow tube with a sampling
interval at the bottom (slots or holes) is driven into the soil. A vapor
sample is then drawn up through the tube and collected for laboratory
analysis or direct field analysis. Complex soil-gas surveys involve multi-
level sampling of many areas; however, for site screening, "shallow" soil-
gas sampling (<1.5 m (5 ft) BGS) is most often practiced. A positive result
(i.e., detectable contaminant vapors) of a soil-gas analysis is a reliable indi-
cator of the presence of contaminants somewhere in the subsurface; unfor-
3.63
-------
Process Identification and Description
Table 3.4
Methods for Hydrocarbon Delineation
Melhod
Soil Gas Survey
Product Thickness
Groundwater Sampling
Cone Penetrometer
Soil Sampling
Ease
High
High
High
Moderate
Low
Cost
Low
Low
Moderate
Moderate
High
Reliability
Vadose1 GW2
Low
Moderate
Low
High
Moderate
Low
Moderate
High
High
Moderate
Vertical Definition
Vadose1 GW2
Possible
No
No
Yes
Yes
No
No
Possible
Yes
Yes
1 - unsaturated zone characterization
2 - saturated zone characterization
tunately, a negative result cannot be interpreted to indicate the absence of
contaminants, except in the very localized sampling area. Soil-gas sam-
pling can be performed relatively rapidly and at a low cost; however, proper
analysis requires knowledge of the subsurface geology and contaminant
composition. Deeper soil-gas surveying has been reported successful for
dissolved contaminant plume delineation, although this conclusion is not
widely accepted (soil-gas surveying may prove to be useful for shallow,
soluble groundwater plumes or free-phase liquid plumes). It is most likely
to be effective for the detection of volatile compounds in relatively perme-
able and homogeneous soils or in other words, in those scenarios favorable
for the application of vapor extraction-based technologies. If sampling and
analysis are performed properly, soil-gas data provide a direct assessment
of the maximum extractable contaminant vapor concentrations that could be
measured at the start of a vapor extraction operation. However, it should be
noted that no quantitative correlation exists between measured soil-gas
concentrations and hydrocarbon levels in soils for high concentrations of
residual contaminants (a conclusion supported by the equilibrium partition-
ing analysis presented in Section 3.2). Soil-gas surveys are also limited by
materials that the probes can be pushed through, and may not be practicable
in areas of urban fill (concrete, rubble, and brick).
3.64
-------
Chapter 3
Product Thickness Measurements. The appearance of free-phase liquid
contaminants in monitoring wells indicates soils highly saturated with con-
taminants near or below the water table; but it does not give any indication
of the extent of contaminants in the vadose zone. Where there are
LNAPLs, visible free-phase contaminants indicate that the contaminants are
distributed at or near the water table (or capillary fringe). If the contami-
nant has been present for a while, there is also a high probability of residu-
ally-saturated soils above and below the water table because "smearing" of
LNAPL during the rise and fall of the water table over the hydrogeological
cycle. The presence of DNAPLs also indicates contaminant-saturated soils
at or below the water table; however, the densities of these materials do not
constrain them to the water table interface.
Ground-water Sampling. Unless residual contamination is known to have
affected groundwater (i.e., free-phase mobile LNAPL or DNAPL is
present), groundwater sampling results are of limited value in delineating
unsaturated zone contamination. Groundwater levels respond to soil con-
tamination if: (1) groundwater is in contact with the impacted soil, (2) there
is significant recharge that leaches through impacted soils, or (3) there is
significant contaminant vapor migration. As in soil-gas sampling, there-
fore, positive results indicate the presence of contaminants somewhere in
the subsurface, but the absence of measured dissolved contaminant plumes
does not necessarily indicate the absence of vadose zone contamination.
Groundwater sampling results are valuable for use in mapping contami-
nants beneath the water table if there is some presampling knowledge of the
vertical contaminant distribution. Where LNAPLs are known to be present,
the focus is on sampling groundwater in the vicinity of the water table (or
the zone over which the groundwater level has fluctuated) because that is
the zone in which LNAPLs are most likely to be present. Distribution of
DNAPLs is strongly dependent on the release size and the subsurface geol-
ogy. The ability to predict their probable location is limited, except where
distinct confining strata are present.
Groundwater samples can be obtained from traditional monitoring wells
or through use of newer sampling methods, such as the cone penetrometer,
or of the direct-push technologies (Geoprobe™, Hydropunch™, etc.).
Cone Penetrometer. The cone penetrometer is a tool adapted from geo-
logic practice that can be used to simultaneously determine subsurface li-
3.65
-------
Process Identification and Description
thology and contaminant extent based on soil-gas and groundwater analy-
ses. It typically consists of a narrow (<5-cm (2 in.) diameter) cylindrical
rod equipped with pressure transducers and a sampling port. As the rod is
driven into the soil, the normal pressure exerted by the soil on the rod tip
and tangential resistance exerted on the rod sleeve are recorded and the ratio
of the two are related to soil structure. The data is then used to deduce soil
properties. More advanced units can perform pore pressure dissipation tests
in order to quantify point permeability. The sampling port allows access to
the collection of pore water, groundwater, or soil-gas samples. For certain
geological conditions (one must be able to drive the cone through the sub-
surface), the cone penetrometer is an attractive tool for performing rapid
preliminary assessments and can be used to identify locations for permanent
monitoring wells.
Soil Sampling. Soil sampling refers to the physical collection of soil
from below grade by hand augering or other drilling or direct push tech-
niques. Soil samples are collected at selected depth intervals, preserved,
and either sent to laboratories for analysis, or analyzed in the field by a
screening technique (usually headspace vapor analysis). Soil sampling is
the traditional method of contaminant delineation; however, in interpreting
the results, the user must be familiar with biases introduced by sample han-
dling practices and soil heterogeneities (which are also present to some
degree for all sampling methods).
3.3.2 Geologic/Hydrogeologic Assessment
In a comprehensive geologic/hydrogeologic assessment, one should lo-
cate distinct geologic strata, subsurface conduits and obstructions (piping,
tanks, etc.), and the water table and determine groundwater gradient,
groundwater velocity, and interval of observed water table fluctuations. In
addition, knowledge or estimates of the physical properties of each of the
strata is needed (e.g., permeability, moisture content, organic carbon frac-
tion). The most important, with respect to vapor-extraction based technolo-
gies, is an estimate of the permeability to airflow. This may be obtained
through permeability tests on soil cores or estimated through correlations
based on grain-size analysis. It should be noted that results from either
approach may not always be indicative of actual field conditions, as the lab
tests are often performed on disturbed soil samples, and generalized correla-
tions also have some degree of uncertainty. Even when performed on "un-
3.66
-------
Chapter 3
disturbed" samples, these tests provide only a localized measure of vertical
permeability.
Most geologic cross-sections are currently developed through analysis of
drilling logs and site records (of tanks, piping, etc.); although other geo-
physical techniques, such as the cone penetrometer (see Subsection 3.2.1),
are beginning to play a more important role in this aspect of site assessment.
If borings are to be drilled, it is worthwhile considering installing monitor-
ing wells, piezometers, or vadose zone monitoring installations in them at
the same time, as these will be required if field-pilot testing is conducted.
At this point, it is worthwhile stressing the need to develop detailed geo-
logic cross-sections when considering vapor extraction-based technologies.
The level of detail necessary varies with technology; for example, air
sparging performance is more affected by subtle permeability changes than
vapor extraction or bioventing. At a minimum, distinct geologic units need
to be identified so that the system designer can form a clear conceptual
picture of the induced airflow. Vapor flow rate and flow path are two of the
three most significant factors influencing vapor extraction-based system
performance. They are controlled primarily by the site hydrogeology and
well construction. Knowledge of subsurface geology is essential for the
evaluation of processes using air injection, as there is serious concern that a
misapplication of the technology can result in detrimental environmental
and health effects. In air sparging, airflow away from the injection well is
strongly influenced by relatively minor lithological changes and the identi-
fication of possible preferential channels and barriers to flow is critical.
3.3.3 Laboratory Soil Column Tests
In some instances, it becomes necessary to empirically measure, rather
than predict, the optimal system performance at a given site. Historically,
such cases have comprised a small percentage of the total number of
vacuum vapor applications. Most are associated with sites where remedial
cost estimates are very large, and it is very important that realistic perfor-
mance expectations be determined. Bench-scale experiments might also be
a cost-effective means of identifying how changing process conditions
might affect field-scale performance. For example, the effect of variations
in flow rate on possible micro-scale (intraparticle) mass transfer limitations
could be studied. Historically, bench-scale treatability tests have been of
the kind depicted in Figure 3.25 (on page 3.68). Here, a core of the material
3.67
-------
Process Identification and Description
Figure 3.25
Basic Laboratory Soil Column Treatability Test Schematic
Valve i
II II
11 II
Cold Trap
Carbon Filter
Pressure
Gauge
Inlet
Sweep Gas Carbon Trap
(Air, or and/or Dryer
Nitrogen)
Continuous Hydrocarbon Analyzer
Temperature
Gauge
Flowmeter - -
Humidifier
3.68
-------
Chapter 3
to be treated is placed in a soil column, or series of columns. Each column
is configured so that air flows in one end of the column and out the other,
either by applying a pressure or a vacuum to one end. Flow rates, pres-
sures, and contaminant concentrations in the vapor streams may be mea-
sured during the test. Soil samples are analyzed before and after the test.
However, it must be recognized that laboratory tests are being conducted
on disturbed soil samples that are inherently unrepresentative of field condi-
tions for a number of reasons. Data from such tests have historically been
overemphasized and misinterpreted. Even if intact cores are used, air is
constrained to flow vertically in the laboratory test, which is not likely the
primary air-flow direction in the field. Thus, it is arguable that air perme-
ability, mass-transfer limitations, and "real-time" loss rates cannot be realis-
tically measured in the laboratory-scale test depicted in figure 3.25 (on page
3.66). In addition, it must be remembered that the sample to be tested is
only representative of a very localized area of the site under consideration.
There are many reasons that bench-scale laboratory tests are rarely per-
formed. Two obvious ones are cost and time. However, there also signifi-
cant state-of-the-art limitations in the ability to properly extrapolate
bench-scale results to field designs. Two of the most important factors in
any in situ remedial technology application — macro-scale mass transfer
limitations and geological heterogeneity are not represented in typical soil
column type experiments. A case in point is the work by Hinchee and
Arthur (1991). They found that in soil column experiments that the applica-
tion of nutrients increased the rate of biodegradation in Hill AFB soil
samples. Dupont, Doucette, and Hinchee (1991), however, found that nutri-
ent addition in the field had no effect on bioventing rates.
The true value of laboratory column tests is that they enable one to deter-
mine achievable cleanup levels, assess the accuracy of chemical partitioning
models that might be used for predictive purposes, and to approximate po-
tential biodegradation rates. The goal, then, is to conduct the tests in a man-
ner that allows simulation of the most efficient hydrocarbon removal in the
field. This is accomplished with low vapor flow rates. The evaluation of
pore-scale equilibrium presented by Johnson et al. (1990) suggests that pore
velocities less than 0.1 cm/sec are appropriate and this is roughly equivalent
to a flow rate per unit cross-sectional area of 0.03 cm3/cm2-s (0.06 ftVft2-
min). In any case, practitioners should ensure that velocities bench-scale
tests are representative of expected full-scale field conditions.
3,69
-------
Process Identification and Description
Suggested Soil Column Feasibility Test Protocol. A suggested protocol
for conducting column tests is given below. This approach differs some-
what from that given in the US EPA's Guide for Conducting Treatability
Studies under CERCLA: Soil Vapor Extraction (1991b) in that here the use
of modeling is less integral to the experimental protocol and adjustments
during the test are based solely on experimental data.
1. Calibrate all analytical instruments and process monitors (flow
meters, pressure gauges, etc.);
2. Collect representative core samples in the field and ship to test-
ing facility. Assure sample is preserved;
3. Construct five soil columns using 5.1 cm (2 in.) diameter x 31
cm (12.2 in.) glass columns (approximately 1 kg (2.2 Ib) soil).
Sample each and analyze for:
• total contamination concentration (using a solvent extraction/
GC method),
• boiling point distribution (as described in Subsection 3.3.1),
• compounds of regulatory interest,
• soil moisture content,
• organic carbon fraction,
• bulk density and total porosity, and
• enumeration of hydrocarbon-degrading bacteria (optional);
4. Based on results of the total hydrocarbon analyses, compute an
estimate of the total contaminant mass (mg) in the column MHC(t
= 0) = Mmil C^., (t = 0), where MMI, is the mass of soil in the
column (kg), and Csoi| (t = 0) is the total contaminant concentra-
tion (mg/kg) in the soil at the start of the test;
5. Place soil columns in the apparatus shown in figure 3.25 (on
page 3.66). Keep inlet and outlet valves closed and allow to
equilibrate for at least one hour or until the temperature of the
soil core stabilizes at ambient temperature. Then sample the soil
gas and analyze for:
• total contaminant concentration,
• boiling point distribution (as described in Subsection 3.3.1),
3.70
-------
Chapter 3
• compounds of regulatory interest, and
• O2/CO2 (optional);
6. Open inlet and outlet valves and start airflow through the col-
umn at a constant flow rate of 0.04 L/min (0.0014 ftVmin) (for
sampling purposes it is more convenient to apply a pressure
upstream). If it is desired to distinguish between aerobic biodeg-
radation and volatilization, then N2 sweep gas may be used in-
stead of air with some columns. In most cases, it is desirable to
pass the inlet air/vapor stream through a bubbler to humidify the
air and prevent moisture loss during the test. If aerobic biodeg-
radation experiments are to be conducted, CO2-free air should be
utilized at the sweep gas;
7. Monitor total contaminant concentration continuously (or as
often as possible) in the effluent vapor stream Cvapor (mg/L). It is
important that an appropriate analyzer be used, such as a flame
ionization detector (FID) for most hydrocarbons. Based on this
data and measured flow rates Q (L/min), compute continuously
the total hydrocarbon mass lost by volatilization M)osi(t) (mg) as
time t according to the expression:
QCmpordt = 2,-(GC,,opor(f, - AO + QCvapor(t, ,,^ „
1=0 '=1 I • ' J
For biodegradation tests, one can use the CO2 effluent data or O2
consumption data to estimate mass losses;
8. When data indicates that 50%, 75%, 90%, and 95% of the total
initial hydrocarbon mass has been lost by volatilization (M|ost/
MHC(t = 0) = 0.5, 0.75, 0.90, and 0.95), then sacrifice a column
and perform the analyses specified under item 3 of the protocol.
The fifth column should be used as a duplicate of one of these
(most often the end of the test);
9. Effluent vapor samples should be collected and analyzed as
described under item 5 whenever the effluent vapor concentra-
tion (as measured by the on-line total hydrocarbon analyzer)
decreases by 50% from the concentration of the previous analy-
sis; and
3.71
-------
Process Identification and Description
10. In addition to the chemical analysis and flow rate data collec-
tion, the pressure drop across the column and temperature in the
column should be monitored.
Data from the test should be reduced and displayed as shown in figure
3.26 (on page 3.73). If biodegradation is to be studied, plots showing efflu-
ent CO2 and O2 vapor concentrations as a function of time should also be
constructed. Note that all measurements are being presented as a function
of the volume of sweep gas passed through the column, normalized to the
initial mass of contaminant. This normalization approach may not be ap-
propriate for biodegradation studies. If the flow rate is held constant during
the test, this is equivalent to Qt/MHC(t = 0). This is the proper way of reduc-
ing the data for a volatilization-dominated process as suggested by theory
presented in Section 3.2. It is also common to see data presented as a func-
tion of the volume of sweep gas passed through the column normalized to
the volume of vapor in one vapor-filled "pore volume" in the column; how-
ever, this normalized dependence on pore volumes is appropriate only when
the partitioning of contaminants into the vapor phase is linearly proportional
to the concentration of contaminants in the soil matrix or when the samples
have been obtained from a uniformly-contaminated site. Use of these data
in the system design process will be demonstrated in Section 3.4.
3.3.4 Field Pilot-Scale Activities
Field pilot-scale activities, items 4 through 10 of table 3.3 (on page
3.59), are focused upon the in situ measurement of soil permeability to
vapor flow, zone of vapor extraction, extracted gas concentration and com-
position, aerobic biodegradation rates, and flow balancing requirements (for
injection/extraction systems). Regardless of the level of complexity, all
activities require a minimum test system consisting of the following:
• test vapor extraction well;
• test air injection well (if air sparging is being evaluated);
• vacuum pump and/or blower to induce air flow;
• vapor treatment system (if required);
• calibrated flow meter(s); and
• calibrated pressure/vacuum gauge(s).
3.72
-------
Chapter 3
Figure 3.26
Hypothetical Laboratory Soil Column Test Results
40
35.
,-, 30
° ,<:
X 25.
•1 20.
a
<£ 10 j
Q_
•^— Inlet Pressure
Flowrate— ^•
^—Outlet Pressure
•0.08
•0.06
' 0.04
•0.02
0 50 100 150 200
V/MHC (t=0) [L-vapor/g-hydrocarbon]
Total Hydrocarbon Vapors
Total Residual Hydrocarbons
0
0 50 100 150 200
V/MHC (t=0) [L-vapor/g-hydrocarbon]
D BP Range 1
D BP Range 2
D BP Range 3
• BP Range 4
BP Range 5
0 8.3 35 93 129 227
V/MHC (t=0) [L-vapor/g-hydrocarbon]
D BP Range 1
D BP Range 2
Q BP Range 3
• BP Range 4
BP Range 5
0 8.3 35 93 129 227
V/MHC (t=0) [L-vapor/g-hydrocarbon]
Soil Column ID: 5.1 cm
Soil Column Length' 31 cm
Mass of Soil: 1.1 kg
BP Range 1:<28°C
BP Range 2: 28 -80'C
BP Range 3: 80- lll'C
BP Range 4: 111-144'C
BP Range 5: >144"C
3.73
-------
Process Identification and Description
Depending on the information desired, additional characterization activi-
ties may also require the following:
• sampling ports in the process lines;
• in situ monitoring installations (for both vadose and saturated
zones);
• sampling devices (sampling pumps, syringes, etc.);
• analytical instruments (hydrocarbon analyzer, gas chromato-
graph, etc.);
• tracer gas delivery system; and
• tracer gas monitoring system.
The goal and requirements of each of the activities listed in table 3.3 (on
page 3.59) are discussed in more detail below. It is useful at this point,
however, to discuss the proper installation of the "minimum test system,"
since this is a central feature of all the activities.
As a first consideration, pilot vapor extraction and air injection test wells
should be placed within the area to be treated by a full-scale system. This
typically means that extraction wells are placed within the contaminated
soil zone and screened intervals are selected so as to induce air flow
through or past (in the case of highly heterogeneous media) the zone con-
taining contaminants. For some bioventing scenarios discussed in Subsec-
tion 3.1.3, test wells may actually be placed outside the zone containing
contaminants. The installation of injection wells should be based on similar
considerations. Air injection wells used for air sparging are typically in-
stalled where residual contamination is suspected and care must be taken to
screen the wells in a narrow interval (0.3 to 1.0 m (1 to 3.3 ft)) below the
suspected depth of impacted soils. At sites where a number of distinct
zones are to be treated and a full-scale system is likely to include wells
screened in each zone, more than one test well is appropriate. In practice,
existing groundwater monitoring wells are often used for pilot-scale testing;
however, the reader is cautioned that this is appropriate only in cases where
the capillary fringe area is the zone of interest. Otherwise, pilot-tests con-
ducted with these wells may not be representative of actual full-scale opera-
tion.
Care should be taken in connecting test wells, flow meters, pressure
gauges, and blowers or vacuum pumps at the manifold. Since most blow-
3.74
-------
Chapter 3
ers/vacuum pumps are driven by fixed-speed motors, extraction/injection
flow rates are often controlled by installing gate/block/globe valves and an
air inlet/outlet pipe on the manifold as shown in figure 3.27. Although a
single in-line valve is sufficient to control the injection/extraction flow
rates, the air inlet/outlet pipe is typically included to allow the same level of
control, while also preventing blower/pump overheating. It is very impor-
tant to insure that flow meters and pressure/vacuum gauges are placed be-
tween the wellhead and first encountered valve or piping junction,
otherwise the flow rate and applied pressure/vacuum at the wellhead cannot
be measured accurately. Unfortunately, this is not always done in practice,
Figure 3.27
Simplified Field Pilot Test Schematic for Vapor
Extraction-Based Technologies
Air Outlet
Pipe.
Vapor-
Liquid
Pressure Separator
Air Inlet
Pipe
Blower
i Valves
Vacuum Vapor Treatment
Pump or Umt
Blower
3.75
-------
Process Identification and Description
and these measuring devices are often found incorrectly located between the
blower/vacuum pump and an air inlet/outlet valve.
3.3.4.1 Vapor Flow vs. Applied Vacuum/Pressure Test
In order to select an appropriate vapor extraction blower/vacuum pump,
it is necessary to measure extraction flow rates as a function of applied
vacuum for each test well. For a pilot-test system connected as shown in
figure 3.27, this is accomplished through the following steps:
1. Open the air inlet valve;
2. Close the valve leading to the wellhead;
3. Turn on the blower/vacuum pump so that air is drawn in only
through the air inlet line;
4. Open fully the valve leading to the wellhead;
5. Once the flow rate has stabilized, record the wellhead vacuum
and flow rate;
6. In a series of increments, slowly close the air inlet valve until
fully closed; and
7. For each increment, allow the flow rate to stabilize and record
the wellhead vacuum and flow rate.
For systems using air injection, an air injection test is conducted by a
procedure similar to the extraction test described above. The recommended
sequence is:
1. Open air outlet valve;
2. Close the valve leading to the wellhead;
3. Turn on the blower so that air is being forced out only through
the air outlet line;
4. Open fully the valve leading to the wellhead;
5. Once the flow rate (if any) has stabilized, record the wellhead
pressure and flow rate;
6. In a series of increments, slowly close the air inlet valve until
fully closed; and
3.76
-------
Chapter 3
7. For each increment, allow the flow rate to stabilize and record
the wellhead pressure and flow rate. In air injection for air-
sparging systems, no flow will be measured until the minimum
pressure required to initiate flow is exceeded. Record the pres-
sure at which flow is first initiated.
If these are the only data desired, the injection/extraction test can usually
be conducted within a few hours, since flow rates typically stabilize (for all
practical purposes) within a few minutes. Data from these tests should be
presented as shown in figure 3.28. These are recommended methods only
and there are other acceptable methods of displaying the data. Flow rates
should be reported in "standard" flow rate units (the equivalent volumetric
flow at 1 atm pressure); if Q is the flow rate measured at an absolute pres-
sure P, then the standard flow rate Q* = Q(P/1 atm).
3.3,4.2 Extracted Vapor Characterization vs. Time
Whether evaluating soil vapor extraction, air sparging, bioventing, or any
other variation of vapor extraction technology, there is a need to determine
the extraction-vapor concentrations and compositions that are likely to be
observed. This information, along with knowledge of possible extraction
Figure 3.28
Presentation of (a) Extraction and (b) Injection Test Data
30
a)
x
U
E
§ 10-
0
b)
40
0 30-
'E 10-
0 100 200 300 400
V&cuum [in. H20]
406 in. H2O = 1 psi
) 400 800 1200
A Pressure [in. H2O]
3.77
-------
Process Identification and Description
well flow rates and regulatory requirements, is used to determine what pro-
cess modifications (vapor treatment units or lower flow rates) are necessary
to comply with emissions requirements.
The opportunity to obtain this information is presented during the extrac-
tion/injection test described above in Subsection 3.3.4.1. Total vapor con-
centrations can be measured as a function of time with an on-line total
hydrocarbon analyzer (flame-ionization detectors are recommended for
most hydrocarbons), and vapor samples can be collected and subjected to
gas chromatographic analyses. Although the collection and analyses of
these samples is not a complicated process, the following measures should
be incorporated into any sampling plan:
1. Samples should be collected between the extraction wellhead
and any air inlet line; and
2. The test should to be conducted for a long enough period to
assure that vapor concentrations are representative of extended
system operation; the test should be conducted long enough to
extract several (probably >5) pore volumes of soil gas.
The first measure assures that representative samples of the extracted
vapors are obtained. Care should be taken to assure that sampling ports are
not placed within a few feet of any air inlet junction, as significant back-
mixing may occur near the junction. One must also recognize that the va-
por samples are being withdrawn from a system under vacuum and this
warrants special care. The recommended sampling procedure is to draw the
sample through on-line analyzers or into sampling bags without having it
pass through a pump. This is easier to do with an on-line analyzer, as a
sampling pump can usually be installed downstream of the detector. Ob-
taining bag samples requires a little more creativity. It is usually accom-
plished as shown in figure 3.29 (on page 3.79). The sampling bag is
connected to a port within a chamber that can be sealed and evacuated. The
exterior port is then connected to the process sampling location and by
evacuating the sealed chamber, a sample is drawn into the sampling bag
without having passed through a sampling pump.
The second measure listed above is important because samples obtained
at the start of a vapor extraction pilot test are not representative of sustained
system operation. When flow is initiated in a pilot test, one typically ob-
serves relatively high extracted vapor concentrations that decrease rapidly
3.78
-------
Chapter 3
Figure 3.29
Schematic of Sampling Apparatus Used When Sampling
Vapors Under Vacuum Conditions
Sampling
Bag
Evacuation Pump
over a period of a few hours to a few days to some more stable level (at
least, the rate of decline in concentration is much slower than observed in
the initial start-up period). This is because the initiation of subsurface va-
por flow draws vapors from the contaminant source as well as from other
areas to which contaminant vapors have migrated from the source over
time. Until these vapors are recovered by the extraction well, the measured
extracted vapor concentration is elevated above levels that will be observed
during more sustained operation of the system. Consequently, it is useful to
estimate how long a given test must be conducted. Here the approach used
by Johnson and Stabenau (1991) is followed, which approximates this tran-
sient period Tslart up (s) as the time required to sweep one "pore volume" of
vapors through the flow zone:
start-up^-'/
[3.48]
where EA denotes the air-filled void fraction in the subsurface (0.30 is a
good approximate value), Qwe)| (cmVs) represents the volumetric flow rate to
the extraction well, and the flow zone has been approximated by a cylinder
3.79
-------
Process Identification and Description
of radius RF (cm) and height Hp (cm). In the absence of any other informa-
tion, RF can be estimated to be roughly equal to the depth to top-of-screen
for the well (HF). For an extraction well screened from 3to4m(10tol3
ft) BGS pulling 0.01 m3/s (=20 standard ftVmin), equation [3.48] predicts
the transient period to last approximately 45 minutes. Data collected during
this test should be reduced and displayed as shown in figure 3.30.
3.3.4.3 Subsurface Vapor-Phase Pressure Distribution
When subsurface monitoring installations are available, the subsurface
vapor-phase pressure distribution resulting from injection/extraction tests
can also be monitored. This information is used to assess the zone of con-
Figure 3.30
Presentation of Extracted Vapor Analyses from Pilot Test
100
80-
60-
40-
20-
150
0 0.03 0.08 0.16 0.24 0.4 1.2 2.1
Time Since Start-Up [d]
D BP Range 1:<28"C
D BP Range 2: 28-80'C
D BP Range 3: 80-1 ll'C
• BP Range 4: 111-144'C
• BPRange5:>144'C
_§
.1 100-
50-
Q = 12 SCFM
0.0 0.5 1.0 1.5 2.0 2.5
Time Since Start-Up [d]
Adapted from Johnson, PC., Stanley C.C , Byers, D.L, Benson, D A , Acton, M.A, Soil Vbntmg at a California Site: Held
Data Reconciled with Theory, 274-275, in Hydrocarbon Contaminated Soils and Groundwater, Vbl. 1, Kosteckl, PI,
Calabrese, E.J., Eds., Lewis Publishers (subsidiary of CRC Press), Boca Raton, Florida, 1991. With Permission.
3,80
-------
Chapter 3
tainment and can be used also with permeability distribution data and flow
modeling to gain a better understanding of the subsurface vapor-flow pat-
terns. These pressure monitoring installations may simply consist of soil-
gas probes driven into the subsurface, or they may be dedicated installations
that have been installed in soil borings during the hydrocarbon/geologic
assessment phase of site characterization. Care should be taken in the inter-
pretation of data collected from driven soil-gas points as leakage or short-
circuiting is a risk. This is particularly true of shallow (less than 1 to 2 m (3
to 6.5 ft)) driven points. See figure 3.31 which presents an example instal-
lation. Based on the theoretical analysis presented in Subsection 3.2.4, it is
recommended that monitoring points be placed relatively close (1, 3, 7, 15
Figure 3.31
\tadose Zone Monitoring Installation
1/8" OD Teflon Tubing
0
Borehole
Depth BGS
Ground Surface (ft)
-0
Box Containing Vapor Sampling
Ports &Thermocouples
r'PVCPipe
Coarse Packing
GS
Cement/Bentonite
10
--20
--30
--40
Reprinted by permission of CRC Press from "Soil tenting at a California Site: Field Data Reconciled with Theory;
Hydrocarbon Contaminated Soils and Groundwater: Analysis, Fate, Environmental and Public Health Effects' by PC.
Johnson, C C. Stanley, D.L. Byers, D.A Benson, and M.A Acton, Remediation, Vbl. I, editors, RT Kbstecki and E.J.
Calabrese, Lewis Publishers. Copyright 1991 by CRC Press.
3.81
-------
Process Identification and Description
m (3.2, 10, 23, 49.2 ft)) to the test injection/extraction well, since the vapor-
phase subsurface pressure dissipates rapidly with distance from the injec-
tion/ extraction point.
The extraction/injection tests described in Subsection 3.3.4.1 present the
opportunity for measuring transient pressure changes and steady-state pres-
sure distributions for each change in extraction/injection vapor-flow rate.
Transient data can be used as an alternate means of obtaining soil perme-
ability data, as explained in Subsection 3.2.5. The steady-state data are
used to define the zone of vapor containment and may be used with perme-
ability distribution data and flow modeling to gain a better understanding of
the subsurface vapor-flow patterns. It may not be practicable to collect
transient data in very permeable soils (medium to coarse sands), as the flow
field is established within a very short period.
Transient data are often presented, as shown in figure 3.32a, in accor-
dance with the theory presented in Subsection 3.2.5. The presentation of
steady-state data varies, depending on the density of sampling points. For
sparse data, the presentation is usually similar to that of figure 3.32b, as
Figure 3.32
Presentation of Subsurface Pressure Monitoring Results from Pilot Test
a) Transient Results b) Steady-State Results
b) 100
ff «J -
0-2-
af
c
'? -4-
(§
s
CJ
Q -6-
3
x
|-8-
"°\"
0
D
D 0
D
D
a
n
a HB-7D (r=3.4 m) ° a o a
* HB-6D (r=16m) a
o HB-14D(r=9.8m) Demn = 40ft
°r,
£'
o.
-100
m
ioo°
Depth = 40 ft
100
Reprinted by permission of CRC Press from "Soil Vfentmg at a California Site Field Data Reconciled with Theory
Hydrocarbon Contaminated Soils and Groundwater Analysis, Fate, Environmental and Public Health Effects* by PC.
Johnson, C.C Stanley, D L. Byers, D A Benson, and M.A Acton, Remediation, Vbl I, editors, RT Kostecki and E J.
Calabrese, Lewis Publishers Copyright 1991 by CRC Press
3.82
-------
Chapter 3
suggested by the theory presented in Subsection 3.2.5. More extensive data
permit presentation of interpolated contour plots superimposed on the geo-
logic cross-sections.
3.3.4.4 Subsurface Vapor Concentration Distribution and In
Situ Respirometry
Subsurface monitoring installations used for vapor-phase pressure moni-
toring can also be used to collect soil-gas samples before and during a pilot
test. Procedures discussed in Subsection 3.3.4.2 for collecting vapor
samples under vacuum conditions should be followed.
The relative impact of bypassing (air flows to extraction well without
passing through the zone of contamination) and mass-transfer limitations on
vapor extraction-based system performance can be assessed by comparing
soil-gas concentrations with levels measured in extraction wells. Typically,
measured soil-gas concentrations in the contaminant-impacted zone are
greater than concentrations measured in extraction well soil gas (as this
represents an average concentration of all vapors reaching the well), and the
difference between the two gives a qualitative indication of the degree to
which flow is being induced through the contaminant zone. In addition, the
measurements can be used as supporting evidence in defining the zone of
vapor flow; once the transient period (discussed in Subsection 3.3.4.2) is
over, a reduction in soil-gas concentrations is typically observed in areas in
which flow has been induced.
These measurements play a significant role in the pilot testing of air-
sparging (or any other air injection) systems, as they can quickly indicate if
there is uncontrolled pressure-driven migration of vapors away from the
source area. When this migration occurs, monitoring locations are installed
between the zone of treatment and any sensitive receptors (basements, etc.).
In situ soil-gas measurements are used also to assess aerobic contaminant
degradation rates. In an "in situ respirometry test," the extraction/injection
flows are stopped and the disappearance of O2 and appearance of CO2 are
monitored with time. It is important that this test also be conducted in a
"background" area in order to assess the "natural" subsurface respiration
rate. The background area ideally is similar with regards to geological and
microbial conditions and differs only in that no contaminant is present. See
figure 3.33 (on page 3.84) for data resulting from a test. Interpretation of
these data are discussed in Subsection 3.5.3.3.
3.83
-------
Process Identification and Description
Figure 3.33
Oxygen Utilization and Carbon Dioxide Production in Various
Phases of a Bioventing Project at Tyndall AFB, Florida
50 75
Time (Hours)
50 75
Time (Hours)
Source- Miller 1990
3.3.4.5 Groundwater Elevation Changes
During a vapor extraction pilot test, it is important to monitor groundwa-
ter elevation changes where contaminants are distributed throughout soils
located in close proximity to the water table. Subsurface vapor-phase pres-
sure changes caused by inducing flow affect the groundwater elevation. An
applied vacuum of "X" in. H2O will cause a steady-state rise in the water
table of "X" in. (similarly, positive pressures resulting from injection can
cause water table depression). Monitoring these elevation changes is im-
portant, as they can limit the range of vacuums that may be applied without
pulling water up the pipe or saturating the contaminant zone or they may
indicate that a groundwater drawdown system is necessary.
Groundwater elevation changes during air-sparging tests are taken as
indications of airflow in the saturated zone by some practitioners. Tempo-
rary groundwater mounding has been observed in close proximity to air-
sparging wells.
Monitoring of elevation changes is important and care must be taken in
selecting a method that does not involve changing the subsurface pressure
distribution, as when a monitoring well cap is opened. There are a number
of practical methods for taking these measurements involving the use of dip
tubes, pressure transducers, and interface probes. See figure 3.34for illus-
3.84
-------
Chapter 3
tration of two possible approaches. See figure 3.35 (on page 3.86) for data
resulting from a pilot test.
3.3.4.6 Groundwater Monitoring
Groundwater monitoring can also be conducted during field-pilot tests
and samples obtained from monitoring wells or piezometers can be ana-
Figure 3.34
Groundwater Elevation Monitoring Approaches: a) Direct
Measurement b) Indirect Measurement
a)
Air-Tight Monitoring Well
Cap/Water Sensor Assembly
See Detail
\
Pressure Gauge
Connection
Wire to Sensor
Double TeHon
Inner Septa Seal
'..,.'. t.;...
1.1 1 ....
Monitoring
Well Cap
= Hmcasured-AP[in.H20]
3.85
-------
Process Identification and Description
lyzed for dissolved contaminants and dissolved oxygen. With soil vapor
extraction standing alone, the information resulting from a short-term test
may not be very revealing. In air sparging, however, it is possible that sig-
nificant changes in some indicator parameters might be observed during a
short-term test.
It is important to recognize that groundwater samples obtained during air
sparging may, or may not, be representative of aquifer conditions. Because
of formation stratification and other subsurface heterogeneities, air channels
that propagate outward from the injection point may intersect a monitoring
well and short-circuit to the vadose zone through the well casing or bore-
hole annulus. If this occurs, analysis of groundwater monitoring samples
will generally reflect higher dissolved oxygen and lower contaminant levels
than those that exist in the aquifer. Use of multiple, nested piezometers will
help eliminate this potential confusion. See figure 3.36 (on page 3.87) for
sample data from an extended air-sparging pilot test.
Figure 3.35
Measured Water Table Upwelling
0.5 .
o
a?
0.4 .
0.3.
0.2.
0.1 .
n - Vacuum Increase
• - Water Table Upwelling
[ft H2O] denote vacuums expressed as equivalent water column heights
Reprinted by permisssion of CRC Press from "Soil Venting at a California Site' Field Data Reconciled with Theory.
Hydrocarbon Contaminated Soils and Groundwater: Analysis, Fate, Environmental and Public Health Effects" by PC.
Johnson, C.C. Stanley, O.L. Byers, D.A. Benson, and M.A.Acton, Remediation, Vol. I, editors, PT Kbstecki and E.J.
Calabrese, Lewis Publishers. Copyright 1991 by CRC Press.
3.86
-------
Chapter 3
Figure 3.36
Groundwater Monitoring Results During Application of Air Sparging
Qfi 4
E 2
1
•s
-4
-B- %C
^^
^^
hange BTEX ^^
^"^
•
^ 1
4
2
-4
0 5 1 0 1 5 20 25 30 35 40 45 50 55 60 65 70
Distance to SPVE-2 (feet)
? 10
O
(L>
CJ
3 4
8 2
r gg I
ft - ^
i
^ % i
^ % ••«
% p ^
5O [Before Sparging]
3O (After Sparging]
ll
12 16 20 30 52
Distance (feet)
Source- Feltenetal 1992
Test Date 3/1 8/92
3.3.4.7 Tracer-Gas Tests
The use of tracer-gas tests is probably the most straightforward and de-
finitive way to assure containment of injected vapor streams during air
sparging or during other processes that utilize air injection. They can also
be used to better define the flow field.
3.87
-------
Process Identification and Description
In a tracer-gas test, an inert, easily-detectable tracer gas is fed into the
injection vapor stream or into a subsurface location. Its rate of recovery in
the extraction system and in soil-gas samples is then monitored to assess
ability to control the vapor-flow field. Injection and extraction flow rates
can be tuned to assure containment of the vapor-flow field as recovery effi-
ciency is monitored and analysis of soil-gas samples permits a better picture
of vapor-flow paths to be developed. Transient analyses can also be used to
gain a better understanding of travel times or vapor velocities.
Two tracer gases frequently used in environmental applications are SF6
and He. See figure 3.37 for sample data resulting from an extraction/injec-
tion test.
3.4 System Design
The goal in designing vapor extraction-based processes is to specify
system components and operating conditions that will meet remedial goals
(i.e., cleanup levels and specified duration), while operating within con-
straints (i.e., costs, emissions limits, etc.). Unfortunately, because of the
Figure 3.37
Presentation of Tracer Test Results
S 80-
40-
20-
oi-
operating regime
0 2 4 6 8 10
Total Extraction Flow/Total Injection Flow
3.88
-------
Chapter 3
compounding of uncertainties and inherent limitations arising from natural
heterogeneities, site characterization data, and predictive capabilities, even
well-reasoned goals cannot realistically be expected to be met with a high
degree of confidence. Consequently, the design of vapor extraction-based
processes continues even after the initial system is installed and started.
System performance must be monitored, the results interpreted, and system
modifications made accordingly. Of all the information contained in this
section, the most important message for the reader is that all systems should
be of robust design and flexible so as to handle a wide range of system ex-
pansions and operating conditions. The additional cost incurred in
building-in adequate flexibility is usually small compared to total project
costs and this incremental investment will often provide long-term savings.
As with the presentation of characterization activities (Section 3.3), the
intent here is not to prescribe a standard system design process. Instead, a
range of possible design approaches that encompass most current practices,
as well as suggested improvements, is presented. For each approach pre-
sented, data requirements are specified and the advantages and disadvan-
tages of that approach are discussed. The design approaches, summarized
in table 3.5 (on page 3.90), are discussed in the order of complexity and
expertise required.
The system design process should not be initiated until a feasibility
evaluation has been made. In making a feasibility evaluation, however,
aside from employing the very general tables, graphs, or guidelines avail-
able, one of the best methods (but at many sites not cost-effective) is to
design a system and then assess its implementability. An example of this
approach, which is a variation of that initially presented by Johnson et al.
(1990, 1991), is presented in Chapter 4.0. Experience teaches that some of
the most simplistic design practices are actually more valuable for estimat-
ing feasibility than for designing systems.
Along with the operating conditions, the following components are typi-
cally specified in a soil vapor extraction-based system design:
• number of vapor extraction wells;
• number of air injection wells;
• well location(s);
• well construction(s) (depth, screened interval, materials, etc.);
3.89
-------
Process Identification and Description
Table 3.5
Soil Vapor Extraction-Based Processes Design Approaches
Approach
• empirical approach
• minimization of capital
expenses
• radius of influence-
based approaches
• screening level model-
based approaches
• detailed numerical
modeling and
optimization-b«sed
approaches
Required
Information1
• 1,2
• 1, 2, inventory
of existing
equipment
• 1,2,4,52, 6
• 1,2,32,4,5,6,
economic data
• -1,2,3,4,5,6,
7, 8, 9,
economic data
Advantages
• quick, easy, low skill
level required
• quick, easy, minimizes
new capital
expenditure.
maximizes use of
existing equipment
• insures containment of
hydrocarbon vapors
• little effort required,
design based on
desired performance,
cost of analyses not
prohibitive
• design can be
optimized & based on
desired performance
Disadvantages
• unknown system
performance, technology
may not even be
applicable
• unknown system
performance, technology
may not even be
applicable
• unknown system
performance, does not
insure remediation in
reasonable time frame
• requires higher level of
expertise & ability to
interpret data
• requires highest level of
expertise & ability to
interpret data, cost may
be prohibitive
1 - refers to activities defined in table 3 3
2 - optional, not always used in this approach
• extraction blower(s) or vacuum pump(s);
• injection blower(s);
• vapor treatment unit(s);
• equipment manifolding & piping; and
• instrumentation and controls (flow meters, sampling ports, vapor
concentration monitoring, control valves, pressure or vacuum
relief valves, etc.).
The method for prescribing each component under the various design
approaches is discussed in the subsections immediately following. With the
exception of the number of wells selected, well locations, and the operating
conditions, strategies for determining the design parameters are relatively
3.90
-------
Chapter 3
similar for all approaches. To avoid repetition, the methods for prescribing
the common system components are discussed first.
3.4.1 Extraction/Injection Well Construction
For the most part, vapor extraction-based processes have predominantly
employed vertical wells as a means of directing the flow of vapors. In light
of recent advances in horizontal drilling techniques, however, the use of
horizontal wells is being explored, as these offer some unique advantages
(vapor extraction beneath buildings, greater effective airflow per well in a
narrow depth interval). In addition, vapor extraction trenches have been
used for shallow groundwater sites. Despite the obvious physical differ-
ences in each of these scenarios, the following factors must be addressed
when specifying well construction. These are:
• size of conduit (pipe);
• conduit material;
• length of interval perforated or "screened" to vapor flow;
• the packing and screened-interval specifics;
• method of well installation in the subsurface; and
• minimization of vapor flow "short circuiting."
Well diameters ranging from 1 to 6 in. are commonly employed in ex-
traction/injection well construction. Smaller diameter wells present an
advantage in that they may be driven to depths of about 9 m (30 ft) if the
site geology is amenable; however, the drawback is that there may be sig-
nificant line pressure drops when high vapor flow rates are required. See
Peters and Timmerhaus (1980) for estimation of pressure drops due to flow
in pipes. Larger diameter wells are almost always installed by first drilling
a borehole and then placing the well casing within. A permeable (less resis-
tant to vapor flow than the surrounding formation) packing material is
placed in the annulus around the screened interval, and the remaining annu-
lar region is then grouted (using cement-bentonite grout) to prohibit short-
circuiting of the airflow. An example of this kind of construction is shown
in figure 3.38 (on page 3.92). For most applications PVC piping is ad-
equate; however, other materials should be used in high-pressure, high-
temperature scenarios, or extremely adverse environmental conditions. As
to systems using air injection wells, it is important to recognize that air
3.91
-------
Process Identification and Description
Figure 3.38
Schematic of Standard Extraction/Injection Wells
Pressure/ Vacuum Gauge
and Sampling Port
Suspected
. Hydrocarbon
- Containing
Zone
Saturated
Zone
passing through an injection blower may undergo a significant temperature
rise (especially for high-injection pressures). In this case, the practitioner
should consider the use of injection well materials capable of withstanding
high-temperature conditions.
Screened interval locations should be chosen to maximize airflow
through the desired zone. In the absence of accessible predictive tools, the
practitioner must develop a good intuitive feel for subsurface vapor-flow
paths and the influence of geologic conditions and man-made obstructions
and conduits. In vapor extraction systems, the screened interval is usually
set across the hydrocarbon impacted zone, unless it extends to ground sur-
3.92
-------
Chapter 3
face, or over a very wide interval (e.g., 15 to 30 m (50 to 100 ft)). In this
case, nested wells are often used to better control vapor flow through the
entire zone of contamination. In air-sparging injection wells, the top of the
screened interval must be set below the water table and the zone of residual
contamination; the screened interval itself is typically short (0.5 to 1.0 m
(1.6 to 3.3 ft)), and is often placed not much below the contaminant-im-
pacted zone, as there are concerns about the ability to control the air-flow
pathways, even though in cases of two contaminant zones this may lead to
limited zones of influence.
Screen size should be selected to maximize the area open for vapor flow,
while maintaining an open conduit. Packing materials should be at least as
permeable to vapor flow as the formation itself.
3.4.2 Vapor Treatment
Vapor treatment units are selected to allow the operator to meet regula-
tory requirements, which often take the form of (1) a prescribed percent
reduction in contaminant concentrations across the treatment unit (e.g., 95%
reduction in total contaminant concentration), (2) a required limit on emis-
sion rate of specific or total hydrocarbons (e.g., 0.45 kg/day (1 Ib/day) total
hydrocarbon emission restriction), or (3) some combination of the two. In
some areas it is permissible to discharge untreated vapors, if dispersion
modeling results indicate adequate concentration reduction within a given
distance of the source stack.
A variety of units meeting most needs are currently offered by vendors
for lease or purchase. Most fall under one of the following categories:
• thermal oxidation - vapors are destroyed by combusting them in
a chamber that is designed to maintain a temperature of >930°C
(1,700°F) and a gas residence time of about one second. Supple-
mentary fuel is required when influent contaminant concentra-
tions drop below explosive levels. Many geometries are
available, ranging from units that look like rectangular boxes to
others that resemble flare stacks. Thermal oxidizer units are the
most robust in terms of being able to handle a wide range of
vapor concentrations and are often more reliable (from an equip-
ment standpoint) than other kinds of units; however, operating
3.93
-------
Process Identification and Description
expenses for supplemental fuel can become prohibitive at lower
vapor concentrations;
internal combustion engines - in one of the more novel ap-
proaches to vapor treatment, extraction wells are connected to
the intake manifold of an internal combustion engine, which
then serves as the extraction pump as well as the vapor treatment
device. The vapors fuel the engine; supplemental fuel is added
if the vapor concentration is too low. Concerns with noise,
maintenance, and supplemental fuel costs limited the use of
early versions of these units; however, recent advances in the
design have led to the development of versatile and easily trans-
portable units;
catalytic oxidation - vapors are passed through a catalyst bed
maintained at an elevated temperature (often 260° to 370°C
(500° to 700T)), where they are combusted. Careful control of
the temperature is essential to prevent melt-down of the catalyst.
These units are efficient processors of lower vapor concentration
streams (<1% v/v); higher concentration streams must be diluted
with air. (See the monograph in this series, Innovative Site
Remediation Technology: THERMAL DESTRUCTION, for a
more extensive discussion of catalytic oxidation.);
adsorption units - contaminant vapors are adsorbed out of the
flowing-gas stream onto a support material, such as vapor-phase
activated carbon or zeolite beds. Typically, these units do not
have large capacities and must be run in a recycle mode when
contaminants are to be desorbed (e.g., by steam stripping) when
the capacity has been exceeded. These units are not used as
extensively as those described above, but they do have the ad-
vantage of being able to recover contaminants, rather than
merely destroying them. In addition to solid adsorption materi-
als, relatively nonvolatile liquid hydrocarbons may also be used
to adsorb volatile compounds from vapor streams; and
biological processes - biologically-based units are currently in
the development and evaluation phase. Vapors are transferred to
a bacteria-containing media (e.g., liquid in a fluidized biological
reactor or soil/peat in a "biofilter") where contaminant-degrad-
ing bacteria convert them to carbon dioxide and water. (See
3.94
-------
Chapter 3
monograph in this series, Innovative Site Remediation Technol-
ogy: BIOREMEDIATION, for a more extensive discussion of
biological processes.)
When selecting a vapor treatment unit it is necessary to specify the range
of influent vapor concentrations, range of total vapor flow rates to be pro-
cessed, and treatment requirements. At this point, the user should select the
most cost-effective option, or combinations of technologies. Some indica-
tion of the applicability of a number of options has been given in the pre-
ceding text. Purchase prices for many off-the-shelf units (appropriate for
flows <235 L/sec (500 standard ftVmin)) currently range from $40,000 to
$150,000. Some care should be taken in selecting a vapor treatment strat-
egy, as this often limits the range of potential operating conditions. In
many cases, the incremental cost for larger capacity units (e.g., cost differ-
ential between a 47 and 95 L/sec (100 and 200 standard ftVmin unit)) is not
significant, and it is worthwhile investing the money to allow greater flex-
ibility.
3.4.3 Extraction Pump/Blower and Injection Blower
In order to select extraction blowers, vacuum pumps, and injection blow-
ers, the flow rate necessary at a given design vacuum or pressure is speci-
fied. This information is collected during the characterization activities
outlined in Section 3.3. Based on these data, pump and blower manufactur-
ers can specify which of these units is needed based on measured perfor-
mance or "pump" curves for their equipment. Given a range of
manufacturers and types of blowers and pumps, one generally chooses a
unit based on cost, maintenance requirements, and experience with similar
units. Since extraction and injection blowers and pumps are the heart of
most above-ground vapor extraction equipment, it is recommended that
most systems be designed with redundant blowers and pumps to allow for
periodic failure and maintenance. Practitioners should also assess the need
for explosion-proof equipment, as extracted vapor concentrations often
exceed lower explosive limits.
Although lower cost blowers have historically been the predominant
choice for vapor extraction, there is now a trend to install more versatile
vacuum pumps, since the incremental cost for upgrading is usually not sig-
nificant in relation to total remediation costs. Costs for units with capacities
up to about 95 L/sec (200 standard ftVmin) range from $1,000 to $10,000.
3.95
-------
Process Identification and Description
3.4.4 Instrumentation
Process monitoring is the focus of the discussion in Section 3.6; how-
ever, it is worthwhile to summarize here some of the instrumentation needs
for any system design.
At a minimum, flow rates, applied vacuums or pressures, and extracted
vapor concentrations should be monitored for each well, as well as for the
total system. Except for total system flow and concentration measurements,
sampling ports and measuring devices must be placed between extraction
wellheads and the first downstream flow restriction or manifold junction; in
injection wells, they are placed between injection wellheads and the first
upstream flow restriction or manifold junction. If the system requires N
extraction wells and M injection wells, at a minimum (N+M+1) flow meters
and pressure/vacuum gauges and (N+l) vapor sampling ports are required.
If tracer tests are to be conducted, (N+M+1) vapor sampling ports should be
employed as well.
Total hydrocarbon measuring devices (e.g., continuous flame ionization
detectors or explosimeters) are often placed in-line to assess changes in the
total concentrations of contaminants in the extracted vapors. These are
sometimes integral components of the vapor treatment unit.
In addition, there may be instrumentation associated with vapor treat-
ment units, such as thermocouples, to measure catalyst bed temperatures.
Many newer units are outfitted with a dial-in modem that provides notifica-
tion of system shut-down.
3.4.5 Manifolds
The connecting of system components at the manifold has a significant
effect on the operation and monitoring of vapor extraction-based processes;
yet traditionally, little attention has been paid to this aspect of system de-
sign. As a first rule, the designer must allow for flexibility and future ex-
pansion. This means that the system design must easily accommodate the
installation of additional wells, blowers, or vapor treatment units with mini-
mal cost and disruption of an operating system. Valves must be installed to
allow independent control of flow to and from each well, blower, and vapor
treatment unit. Furthermore, the piping should be installed in a manner that
facilitates replacement of equipment, especially blowers, pumps, and instru-
mentation.
3,96
-------
Chapter 3
Care should be taken to locate instrumentation where it can be readily
monitored and replaced. Guidelines given above for the placement of flow
meters, pressure gauges, and sampling ports should be followed.
3.4.6 Surface Seals and Passive Inlet Wells
The effect and usefulness of surface seals and passive air inlet wells in
controlling vapor-flow paths are often debated among practitioners. Sur-
face seals are touted as a means of expanding the radial influence of vapor
extraction wells. Passive inlet wells supposedly allow better direction of
the flow of air through a zone of contamination.
When employing vapor extraction wells and trenches at shallow depths
(top of screened interval <3 m (10 ft) BGS) and in homogeneous soils, the
user typically finds that the zone of influence of the well or trench does not
extend much beyond a distance equivalent to the depth to top of the screen.
The result is that a greater areal density of wells or trenches is required to
assure vapor containment. In order to combat this problem, it has been
suggested that sealing the ground surface might extend the vapor flow path
and provide better control. Conceptually, this idea is acceptable and expec-
tations are consistent with modeling results; however, before surface seals
are installed at all sites, a determination should be made whether a "true"
surface seal can be achieved in the field. It is unlikely, that a pneumatically
effective surface seal can be achieved in the field by the methods typically
proposed to prevent water infiltration (plastic liners and asphalt). A com-
mon assumption is that pavement and roads provide good pneumatic sur-
face seals. Unfortunately, this may not be the case, since most roads and
paved areas have cracks and joints that may act as air entry points, and they
overlay an extremely permeable bedding material layer. In some cases,
however, this bedding material is saturated (especially when adjacent to
over-irrigated planters), and this limits the surface leakage of air. There
have been reports1 of surface covers employing composite soil:geomem-
brane sections with bedding layers compacted wet of optimum. These types
of seals are specifically designed to prevent air infiltration and seem to
achieve the theoretical effect of influencing air flow pathways. Methods
such as these, however, are more substantial and involve more engineering
1. Hartley, James D. Personal Communication re McClellan AFB Operable Unit D
Cover and Site S Soil Vapor Extraction System, February 28, 1995
3.97
-------
Process Identification and Description
than the most commonly proposed methods (single-layer plastic sheeting
and asphalt).
Although ground surface covers cannot be automatically assumed to be
effective surface air seals, they may still provide benefits by limiting water
infiltration into the subsurface and reducing surface emissions of contami-
nants. In the latter case, contaminant vapors may collect in the permeable
subgrade of an asphalt pavement but be limited in terms of surface flux by
the small openings in the cracks or joints of the overlying pavement. Sys-
tem designs can incorporate surface seals with extraction ports to limit con-
taminant emissions to the atmosphere. Such surface seals, or vapor
shrouds, are often integral components of the thermally-enhanced vapor
extraction-based processes, discussed in Section 3.5.
The use of passive air inlet wells has been proposed where it is desirable
to better direct the air-flow field. In concept, air is drawn into the subsur-
face through the passive air inlet well as a result of air being withdrawn
from an extraction well. As with surface seals, however, there is little evi-
dence showing any benefit in using passive air inlet wells. In the authors'
view, based on field experience and theoretical considerations, such wells
will significantly affect the flow field only if they are placed very close to
the extraction well. This may be best understood by recalling that a rapid
decline in vacuum with distance is typically observed within merely a few
feet from a vapor extraction well. By placing a passive inlet well at a dis-
tance where the induced vacuum is normally small, the driving force for
flow from that region is only incrementally changed.
3.4.7 Design Approaches
Approaches used in designing vapor extraction-based processes, listed in
table 3.5 (on page 3.88), are discussed in the subsections immediately fol-
lowing. They are then compared in Subsection 3.4.8.
3.4.7.1 Empirical Approach
In the empirical approach, the most simplistic of the design approaches,
the number of wells, well locations, and blower and vapor treatment re-
quirements are specified based on previous experience, general guidelines,
and intuition. The only essential information is a baseline geologic/
hydrogeologic assessment summary indicating the general location of con-
taminant-impacted soils.
3.98
-------
Chapter 3
This design process can be easily and quickly performed. It requires a
minimum level of skill in design and little understanding of vapor extrac-
tion-based processes. Unfortunately, performance of the resulting system
may be problematic and may even cause detrimental effects (e.g., drawing
vapors to sensitive receptors), just as can poorly-designed air-sparging sys-
tems.
In this design approach, the maximum system performance is limited by
characteristics of individual system components. In addition, the practitio-
ner risks purchasing and installing equipment that may not be appropriate.
3.4.7.2 Matching System Design to Existing Equipment
In this approach, the use of existing equipment is maximized in order to
minimize capital expenses. In the extreme case, a vendor will merely con-
nect a portable skid-mounted vapor extraction system to existing groundwa-
ter monitoring wells and then, the only installation cost is that of connecting
the equipment.
Design data requirements are the same as those for the approach dis-
cussed in Subsection 3.4.7.1 and this approach suffers from the same draw-
backs. But, of course, it does have the advantage of minimizing installation
costs. Although these savings are likely to be offset by the increased oper-
ating costs of a very inefficient system, it is an approach often used in order
to avoid the expense of customized systems. Drawbacks can be minimized
by purchasing very robust equipment (e.g., vacuum pumps instead of blow-
ers and large capacity vapor treatment units).
3.4.7.3 Radius-of-lnfluence Approaches
Radius-of-influence approaches are currently the most frequently used.
The number of wells and their spacing are based solely on the subsurface
pressure distribution measured during a pilot-scale test.
The basic approach is outlined in figure 3.39 (on page 3.100). First,
steady-state subsurface pressure distribution data are plotted as a function of
distance from the pilot extraction well, usually on log-log or semi-log coor-
dinates, and a best-fit line is then drawn. The "radius of influence" of the
well is then graphically determined; it is that distance from the well where
the pressure distribution has been extrapolated to reach some specified
value. In practice, these specified values typically are 0.1 in. H2O, 1.0 in.
H2O, or 10% of the applied vacuum at the extraction wellhead.
3.99
-------
Process Identification and Description
Figure 3.39
Flowchart tor Radlus-of-lnfluence Design Approach
so
1. Conduct pilot test and measure:
• steady-state pressure distribution
• flowrate -vs- pressure
O, 40-
33
e
=• 30 •;
20-
10
0
.1 1 10 AlOO
Distance from \tell [ft]
2. Plot steady-state pressure distribution:
• draw best-fit line through data
• graphically determine Rj
3. Select well locations so that radii of
influence overlap zone to be remediated
Suspected Zone of
Hydrocarbons
Once the radius of influence (R:) has been determined, circles with this
radius are drawn on a site map locating wells such that areas circumscribed
by the circles overlap the zone to be remediated. After well locations have
been specified, the designer selects extraction/injection pumps/blowers and
a vapor treatment unit, often by one of the two approaches described in
Subsections 3.4.7.1 and 3.4.7.2.
3.100
-------
Chapter 3
For air sparging, this approach has been extended to include the analysis
of pilot-test unsaturated zone pressure distribution data, water table eleva-
tion data, and groundwater dissolved oxygen concentrations in an effort to
establish an equivalent air-sparging radius of influence.
In the design of air-sparging vapor extraction systems, air-sparging well
locations are selected first so that their radii of influence encompass the
zone to be treated. Extraction well locations are then selected to assure that
the extraction radii of influence extend between and around the air-sparging
injection well radii of influence.
This approach is widely practiced because it is simple and graphical and
closely imitates common groundwater recovery design practices. Unfortu-
nately, this approach, at best, merely assures containment of contaminant
vapors (the purpose of groundwater recovery system design) and does not
provide for estimation of performance or long-term costs (Johnson and
Ettinger 1994, illustrate that this design approach can result in extremely
inefficient system designs in some cases).
3.4.7.4 Screening-Level Model Approaches
The major limitation of the simplistic design approaches discussed above
lies in their failure to consider probable system performance factors
(cleanup levels, duration of remediation, etc.) or constraints (costs, regula-
tory requirements, etc.). In order to incorporate these factors, the practitio-
ner must apply predictive models that facilitate estimation of system
performance as a function of a wide range of parameters, such as geologic
conditions, contaminant type, and number of wells. Section 3.2 focused on
the development of mathematical descriptions of relevant phenomena and
showed that there were numerous modeling approaches ranging in complex-
ity from analytical solutions to numerical algorithms requiring computer-
based solutions. This section addresses the use of more accessible
screening level models in the system design process. As an example, the
approach presented by Johnson et al. (1990, 1991), which forms the basis
for the Hyperventilate (Johnson and Stabenau 1991) software guidance
system (distributed by the US EPA) is adopted. The data requirements
include the following:
• extraction flow rate as a function of applied vacuum for concep-
tual well design;
3.101
-------
Process Identification and Description
• estimate of average residual soil concentration and volume of
impacted soil;
• remedial goals - cleanup levels and remediation time;
• minimum volume of vapor to achieve required cleanup;
• steady-state pressure distribution from the well;
• effluent vapor concentration data from pilot test;
• geological cross-section map with contaminant distribution; and
• constraints - costs, regulatory requirements, etc.
The results of Subsection 3.2.9 can be applied in relating (1) the number
of wells (Nwells) required to achieve the goals as a function of the flow rate
to a single well (Q(AP) (cnvVs)), (2) target remediation time (tclesilup(s)), (3)
average soil contaminant concentration ( (mg/cm3-soil)), (4) volume
of contaminant-containing soil (Vwi/ (cm3-soil)), and (5) the parameter a (1-
vapor/g-initial contaminant), which represents the minimum volume of
vapor required to achieve cleanup per unit mass of contaminant as follows:
* _ Vsoil < Csoil
The value of a can be determined by considering equilibrium modeling
results (Subsection 3.2.9), possible mass-transfer resistances (Subsection
3.2.1 1), and laboratory soil column tests (Subsection 3.3.3). In the
Hyperventilate approach, a is determined by first predicting (or measuring)
the ideal equilibrium-based model result (Subsection 3.2.9) and then multi-
plying it by an efficiency factor determined from screening-level mass-
transfer models (Subsection 3.2.1 1) or by estimating subsurface vapor
dilution obtained from comparisons of soil gas and extracted vapor concen-
trations in the pilot test. Although the above discussion is oriented toward
contaminant removal through volatilization, if biodegradation rates are
known, they can also be incorporated in the modeling predictions of a.
The number of wells, Nwdls, predicted by equation [3.49] should be re-
garded as a minimum estimate of wells required, as any real system per-
forms less efficiently than an ideal system. Equation [3.49] provides a
screening-level tool relating the number of wells, cleanup objectives, and
operating conditions (flow rate). After either extraction pump characteris-
3.102
-------
Chapter 3
tics (required total flow and AP at each well) or the number of wells is
specified, the remaining parameter is calculated from equation [3.49]. Once
the number of wells and applied pressure are specified, steady-state pres-
sure distribution data and the approach described in Subsection 3.4.7.3 are
used in an attempt to assure that the system will contain the vapor. If it
does not, additional wells will be needed. Figure 3.40 is a sample
worksheet for this design approach.
Although this approach is discussed mainly in relation to vapor extrac-
tion systems, it is not difficult to extend it to the sparging of residual con-
Figure 3.40
Sample Worksheet for a Screening-Level Model Design Approach
Description of Data or Calculation
Value/
Result
Units
Site Data
• attach geological cross section showing contaminant
sampling results, subsurface conduits, and groundwater
elevation
• attach contaminant analyses results
• average contaminant concentration in treatment zone
• estimated volume of soil in treatment zone VM1j
• approximate soil density p^,,
• design flow rate per well Q@AP (attach pilot test data)
• radius-of-influence (attach steady-state pressure dist. figure)
Remedial Goals
• cleanup target
• desired duration of remediation Tc
• cost goal
:leanup
Calculations
• minimum volume of vapor required (ideal case) a
• efficiency factor (describe basis for choice) e
' Nw=ll! = (Vsollpsoll
-------
Process Identification and Description
taminants in the saturated zone. Equation [3.49] would be interpreted as
relating the minimum number of air injection wells required to the flow rate
per air injection well and desired remediation duration.
At this point, an estimated total project cost can be calculated by specify-
ing the costs for blowers, wells, piping, vapor treatment, and operation and
maintenance, and another equation can be generated predicting total project
cost as a function of duration of remediation. These operations are illus-
trated through an example in Chapter 4.0.
3.4.7.5 Detailed Numerical Modeling and Optimization-Based
Approaches
The approach presented in Subsection 3.4.7.4 can be extended through
use of more complex mathematical models. Such models will usually re-
quire more data than are readily available; however, there are sites for
which the projected cost of remediation warrants the expense of acquiring
the needed data. To be used most effectively, complex modeling must be
performed iteratively. The predictive model must be continuously refined,
calibrated, and updated based on performance data collected from system
operation. To date, accepted and validated models are not accessible to the
general practitioner, although they may soon be (e.g., Joss and Baehr 1994).
3.4.8 Comparison of Various Design Approaches
To compare the several design approaches, figures 3.41 (on page 3.105)
and 3.42 (on page 3.106) set forth summaries of results of many of the site
characterization activities described in Section 3.3 for an example site. Al-
though the example site has a number of distinct strata, for the sake of sim-
plicity, this comparison focuses on the fine to medium sandy layer located
between the clay layer and water table at a depth interval of about 13 to 15
m (43 to 49 ft) BGS. The determinations resulting from the several designs
are tabulated in table 3.6 (on page 3.107).
For the design approaches outlined in Subsections 3.4.7.1 and 3.4.7.2,
only the geological cross-section and contaminant distribution data pre-
sented in figure 3.41 are needed. Experience has shown that it is likely that
a practitioner following one of these methods (and in the absence of pilot-
test data) would install a system consisting of a single vapor extraction well
screened roughly between 11 and 16 m (35 to 55 ft) BGS. This well might
actually be an existing groundwater monitoring well. It would be connected
3.104
-------
Chapter 3
Figure 3.41
Total Hydrocarbon Distribution and Subsurface
Geology for Sample Design Problem
North
South
fl-
ic—
©
§ 20 —
VI
|
I 30-
3
I 40-
s
50-
60 —
Static Ground
\Vater Table
L
k
i
- 0.3
>
" tr.2
j
t
- 0.02
- 0.0
- 0.0
>
1
- 0.0
- 1.7
>
-24 t
- 73
>
L 9.5
L \ Tank
k Sandy \ Backfill
Clay \ (former tank
r V location) -
y _ _ _
Fine to
Coarse Sand
r
k
Silty Clay
&
Clayey Silt
1
I
Medium Sand
r
/
/
-0.5 /
"».U 4-
™r
- 512
- 5.4
- 8577
"Ijll
- 653
- 3267
- 1237
- 23831
- jJiy
- 1.7
- 0.8
A £_ .
— -*f.^ — — ~~ — ^^
- 0.3
- 8.2
- 214
_2| _
i- 967
h 971
_ 28679
- 23167
- 0.31
1 T
— —1:2
- 0.44
- 0.17
- 8.8
- - 0.63
- 1.5
- 0.86
- 23
\ 1.6
p 3.2 T
HB-17 HB-10 HB-5 HB-3
SCALE (ft)
10
20
Soil concentrations given in mg/kg-soil.
Reprinted by permisssion of CRC Press from 'Soil Vbnting at a California Site: Field Data Reconciled with Theory
Hydrocarbon Contaminated Soils and Groundwater: Analysis, Fate, Environmental and Public Health Effects" by RC.
Johnson, C.C. Stanley, D.L Byers, O.A Benson, and M.A Acton, Remediation, \bl. I, editors, RT. Kostecki and E.J.
Calabrese, Lewis Publishers. Copyright 1991 by CRC Press.
to a 47 L/sec (100 standard ft3/min) capacity blower and a vapor treatment
unit. It is also very likely that in these geologic conditions, the user will not
have anticipated the need for a groundwater drawdown system and that he
or she will soon find the extraction flow decreasing to zero as the water
level rises above the screened interval. Water would not be pulled all the
3.105
-------
Process Identification and Description
way up the well to ground surface, as the maximum upwelling that can be
created by a perfect vacuum is 10.33 m (33.9 ft).
In the radius-of-influence design approach, described in Subsection
3.4.7.3, the information in figure 3.41 is used along with the pilot-test data
in figure 3.42. Based on the steady-state pressure data in figure 3.42, com-
mon practice would yield a radius of influence of approximately 18 m (60
ft). In comparison, figure 3.41 indicates that most hydrocarbons are distrib-
uted within a radial distance of 9 m (30 ft), and it appears that a single well
Figure 3.42
Summary of Design Data for Sample Design Problem
30
20 -
10 -
*Water Level Constant @ 49 ft
0 100 200 300 400
\Scuum [in. H2O]
100.
0-
-100
Depth = 40 ft
1 10 100
Distance from Wfell [ft]
150
I 100-
50-
Q = 12 SCFM
0.0 0.5 1.0 1.5 2.0 2.5
Time Since Start-Up [d]
Soil Column Test Results
Total Hydrocarbon Vapors
Total Residual Hydrocarbons
V/MHC (t=0) [L-vapor/g-hydrocarbon]
3.106
-------
Chapter 3
Table 3.6
Results of Example Design Problem
Design Approach
Parameter
Design Specifics
• number of wells
• total flow rate [SCFM] @ required
vacuum [in. H2O]
• vapor treatment unit capacity [SCFM]
A, B
1
available
blower
100
C
1
available
blower
100
D
45
450 @
28 in
450
E
NA
NA
NA
System Expectations
• hydrocarbon level reduction [%]
• duration of cleanup [d]
360
NA
NA
A,B- Intuition/Experience-Based Empirical Approach and Minimization ol Capital Expenses (Subsection 3471 and
3472)
C - Radius ol Influence-Based Approach (Subsection 3473)
D - Screening Level Model-Based Approach (Subsection 3474)
E - Detailed Numerical Modeling and Optimization-Based Approaches (Subsection 3475)
NA - this level ol design not yet justifiable for this site
? - unknown for this design procedure
should be sufficient for remediation. Therefore, the vapor extraction system
designed by this process is identical to that resulting from the methods of
Subsections 3.4.7.1 and 3.4.7.2. It is likely, however, that in this case the
practitioner would recognize the necessity of incorporating a groundwater
drawdown system, as significant upwelling would have been discovered in
the pilot test.
Under the approach described in Subsection 3.4.7.4, full use is made of
the complete pilot-test data. In order that equation [3.49] may be used, the
following parameter values are assigned:
soil
= 71 (30 ft)2 x 6 ft = 17,000 ft3 = 4.8 x 106 m3
= 20,000 mg/kg-soil = 0.020 g/g-soil x 1.7 g/cm3-soil =
0.034 g/cm3-soil
a = (200 L/g-hydrocarbon (3,200 ft3/lb-hydrocarbon) (ideal
case))/(0.50 (efficiency)) = 400 L/g-hydrocarbon (6,400
ft3/lb-hydrocarbon)
3.107
-------
Process Identification and Description
Q(AP) = 10 standard ft3/min @ 100 in. H2O = 4.7 L/sec
= 360 d = 3.1x10? sec
The volume of soil, V^,, is obtained by approximating the hydrocarbon-
containing zone as a cylinder of radius 9 m (30 ft) and depth 1.8 m (6 ft), as
estimated from the data in figure 3.41 (on page 3.105). The average soil
concentration in this zone is assumed to be 20,000 mg-hydrocarbon/
kg-soil, consistent with data given in figure 3.41. In order to assign a value
to the minimum volume of vapor required to achieve the desired degree of
remediation a, the soil column test data and pilot-test vapor concentration
data in figure 3.42 (on page 3.106) are used. Soil column test data indicate
that about 200 L-vapor/g-initial residual (3,200 ftVlb-initial residual) hydro-
carbon is required to achieve a reduction of 95% in the total hydrocarbon
concentration under ideal conditions (it is more likely that modeling results
would be sufficient in this case, but soil column data are used for purposes
of illustration), while effluent vapor concentration monitoring results from
the pilot test suggest that only half the extracted flow is actually passing
through hydrocarbon-containing soils. Thus, a is modified to account for
this efficiency (about 50%) and a value a = 400 L-vapor/g-initial residual
(6,400 ftVlb-initial residual) hydrocarbon is assigned. Flow-rate data com-
bined with knowledge of groundwater drawdown system limitations indi-
cate that the maximum extraction flow rate is likely to be about 4.7 L/sec
(10 standard ftVmin) per well in this situation, where airflow is confined
between the clay layer and groundwater. In this example, the goal of a
remediation duration of one year (360 d = 3. 1 x 107 sec) is assumed. As
indicated in table 3.6 (on page 3.107), the prediction resulting from equa-
tion [3.49] is that a minimum of 45 wells will be required to remediate this
site. Consequently, the system must be sized to handle and treat 45 x 4.7 U
sec (10 standard ftVmin) = 212 L/sec (450 standard fWmin) of vapor at 100
in. H2O gauge vacuum.
For this example, more sophisticated predictive models were not used,
mainly because it is not likely that they will be used in the near future for
sites of this scale. The reader should see, however, that there might be a
certain benefit in using more sophisticated models for this site, as the large
number of wells predicted results mainly from flow restrictions imposed by
groundwater upwelling. With more complex predictive tools, the user
could explore complex system designs using air injection wells and state-of-
3.108
-------
Chapter 3
the-art groundwater extraction systems to partially overcome this constraint.
It should be noted that even though 45 wells would not be a typical design,
it is unlikely that other technologies currently available would be effective
in remediating this site in the desired time frame.
In comparing the system designs resulting from the several approaches,
as table 3.6 (on page 3.107) indicates, one sees little difference among the
systems designed under the methods of Subsections 3.4.7.1, 3.4.7.2, and
3.4.7.3. In each case, a simple, single extraction well system is prescribed.
But the predictions resulting from the more sophisticated (and reliable)
screening level model approach yield a prohibitively high density of extrac-
tion wells, a result that should lead the user to question the efficacy of vapor
extraction at this site. Based on the screening-level model predictions, sys-
tems installed by the first three methods are unlikely to achieve satisfactory
remediation in a reasonable time frame, although there is nothing in the
design procedure to indicate this result. Thus, this exercise illustrates the
inherent drawbacks of simplistic design practices that are commonly used.
3.4.9 Other Design Considerations
There is an increasing trend toward using integrated and phased ap-
proaches wherein multiple technologies are used together or in sequence to
achieve the desired remediation goal. In many cases, additional advantages
result from combining technologies. In the preceding example, a groundwa-
ter recovery system could be used to contain the soluble groundwater plume
as well as to enhance the vapor extraction system performance by minimiz-
ing groundwater upwelling. The practitioner is urged to be alert to possible
synergistic effects as system design progresses.
3.5 Costs
It is common to see remediation costs of ex situ technologies (incinera-
tion, thermal desorption, soil washing, etc.) quoted on a per volume of soil
treated basis, and it would be useful if costs of vapor extraction-based tech-
nologies were similarly quoted. Unfortunately, given the dependence on
site geology, type of contaminant, and other factors, especially depth to
contamination, costs of vapor extraction-based technologies cannot be
3.109
-------
Process Identification and Description
meaningfully expressed on a per-volume basis. Estimates can be made for
major capital equipment costs, and these have been included in the text.
The major portion of total remedial costs in most cases is that for operating
labor, maintenance, and monitoring. These costs can best be minimized by
specifying systems under a design method that incorporates remedial goals
and by employing equipment requiring little maintenance.
3.6 Performance Monitoring
As observed in Section 3.4, the design of vapor extraction-based systems
is, for all practical purposes, a continuous process that begins with the ini-
tial pre-construction design and continues after the system is installed and
operating. Monitoring data are relied upon in assessing system perfor-
mance, calibrating models, and guiding necessary operational changes and
equipment modifications. System monitoring and data presentation require-
ments have been presented to some extent in the discussion of field-pilot
tests (Subsection 3.3.4) and system design (Section 3.4), and the reader is
referred to those discussions. Here, data and data presentation options for
full-scale continuously operating systems are discussed in more detail.
Once again, there is a wide range of monitoring options and it is up to the
practitioner to select monitoring requirements based on the particular need
for information. Requirements are presented here in relative order of im-
portance in assessing system performance. While there is flexibility in
choosing monitoring strategies, as table 3.7 (on page 3.111) indicates, there
is a minimum level of information that must be gathered in order to make
basic performance evaluation decisions.
3.6.1 Primary Process Variables (Vapor Flow Rates, Pressure,
Extracted Gas Characterization)
The most straightforward means of assessing vapor extraction-based
process performance is to monitor the flow and composition of the ex-
tracted gases. This is the minimum monitoring required and is done to track
mass-removal rates, compositional changes, and mass- and vapor-flow
rates. Interpretation of the data can lead to identification of permeability
changes and mass-transfer limitations.
3.110
-------
Chapter 3
Table 3.7
Process Monitoring Options and Data Interpretation
Data Interpretation/Analysis Requirement Data Collection Requirement
concentration vs. time 1
composition vs. time
flowrate vs. time
applied pressure/vacuum vs. time
mass removal rate [mass/timel vs. time
cumulative removed by volatilization [mass]
identify mass transfer limitations
aerobic biodegradation contribution to removal rate [mass/time] vs. time 1,2,6*
aerobic biodegradation contribution to cumulative removed [mass]
total remediation costs [$] vs. time 1, 2°, 3
cost per mass of hydrocarbon removed [$/kg-removed] vs. time
effect of environmental factors [qualitative] 1, 2^, 4
in situ assessment of treatment with time [qualitative areal impact] 1,2*1,4*. 5,6'', 8°, 9s
define zone of vapor containment [qualitative areal impact] 1, 5*, 7, 11 *
closure monitoring report 1, 2b. 3*, 4*, 5, 7, 8, 9, 10, 11*
areal impact of air sparging 1, 2, 4*, 5*. 6*, 7, 8*, 9,10,11*
effect of water table elevation changes 1,2,4,5,6,7,9,10
injection/extraction flow rate optimization 1, 2, 3,4, 5, 6, 7, 8, 9, 10, 11
flow field definition
b - applicable for bioventing applications, * - optional, or as required. s - relevant to air sparging
Data Collection Requirement Key
1) process monitoring data- extraction/injection flow rate(s) and vacuum(s)/pressure(s), extraction vapor concentration
and composition
respiratory gas (02, C02) monitoring of extracted vapor stream
cost monitoring- capital, operation and maintenance, and utilities costs
environmental monitoring- temperature, barometric pressure, precipitation, soil moisture content
„ in situ soil gas monitoring: vapor concentration and composition
6) in situ soil gas monitoring respiratory gases (CO2 and O2)
7) subsurface pressure distribution monitoring
~ soil samples
groundwater monitoring
10
groundwater elevation monitoring
tracer gas monitoring
Flow rate Q (volume/time) may be measured by a number of means and
should be corrected to some volume per unit of time Q* at a standard pres-
sure and temperature and is expressed:
Q* = Q (P/l atm)(293 °K/T) [3.50]
where P (atm) and T (°K) are the absolute pressure and absolute tempera-
ture measured at the flow-rate measuring device, respectively. Examples of
3.111
-------
Process Identification and Description
acceptable expressions of flow rate units are standard mVhr or standard ft3/
min (this implies flow rates corrected to 1 atmosphere and 20°C (68°F)).
The use of standard units is especially important, as most gas analyses are
expressed on similar bases and these two values are multiplied to assess
mass-removal rates.
A variety of methods are available for measuring gas-flow rates. Pilot
tubes or orifice plates combined with an inclined manometer or a differen-
tial pressure gauge are acceptable for measuring flow velocities of at least
400 m/min (1,300 ft/min). For lower flow rates, a rotometer will typically
provide a more accurate measurement. As mentioned previously (Subsec-
tion 3.3.4 and Section 3.4), flow meters should be installed to measure flow
rates from all extraction/injection wells, as well as combined total extrac-
tion/injection flow. In order to be able to express the measured flow rate on
a standard (1 atm, 20°C) basis, the pressure at the point of flow measure-
ment must be known. All extraction well flow meters and pressure gauges
must be placed between the wellhead and first downstream junction or
valve (or upstream, in the case of air-injection wells). There are also other
guidelines for flow-meter placement that are specific to different types of
flow meters (e.g., placement at least 10 pipe diameters away from constric-
tions).
Extraction gas pressure should be monitored at the extraction wellhead.
Typically, this is done by a permanently installed pressure gauge or a
"quick release" connection that facilitates measurement. The pressure mea-
surements required for flow-rate measurement are also useful in interpreting
system operation and performance. Pressure changes over time (at constant
flow rate) indicate soil-gas permeability changes and usually are the result
of soil-moisture changes (due to upwelling, infiltration, or drying). Figure
3.43 (on page 3.113) presents pressure and flow rate data for a vapor extrac-
tion system on the same graph. It can be seen that a permeability reduction
is occurring with time, as the flow rate is decreasing with time while the
applied vacuum is held constant. In this case, the reduction was attributed
to groundwater elevation changes (Johnson et al. 1991). Similar injection
pressure vs. flow-rate plots should be made for air sparging, bioventing, and
other vapor extraction-based technologies employing air injection wells.
Exhaust gas is monitored in order to determine contaminant removal
rates and assess mass-transfer limitations. Gas composition measurements
typically include, at a minimum, some measure of the contaminant concen-
3.112
-------
Chapter 3
tration and composition and may include respiratory gas measurements
(Subsection 3.5.2). A variety of techniques are available for measuring
contaminant concentration in the extracted gas; the choice in a given situa-
tion may be dictated by regulations or permitting procedures. Typically, a
flame ionization detector (FID), a photoionization detector (PID), or an
electrochemical detector (e.g., a "hot wire") is used. If the contaminant is a
chlorinated organic, an electron capture detector (BCD) or Hall-type detec-
Figure 3.43
Presentation of "Minimum" Data Collection
Needs for Vapor Extraction Systems
20
20 40 60 80 100 120
Time [d]
20 40 60 80 100 120
Time [d]
0 20 40 60 80 100 120
Time [d]
100
•5 80J
I
-------
Process Identification and Description
tor may be more appropriate. The detector may be coupled with a gas chro-
matograph to separate peaks.
For most nonhalogenated organics, the FID is currently the recom-
mended detector. Individual peaks of a chromatographic column are quan-
tified and used to assess composition. In the absence of chromatographic
separation, the total detector response is used as an indication of total con-
taminant concentration. For some organics, such as benzene, PID detectors
are often used because of their high sensitivity; however, this sensitivity is
compound-specific and highly variable. Thus, the PID is a poor indicator of
total contaminant concentration and should not be used for this purpose
unless it is known that the PID is equally responsive to all compounds in the
hydrocarbon vapor stream. A PID usually works best when a single com-
pound is present and its response is known. A hot wire detector is used to
monitor explosive environments and is adequate for monitoring total con-
taminant response at higher concentrations (above 100 ppmv).
Currently, concentrations are most often reported by laboratories as
ppmv, parts per million by volume (sometimes called L/L). This is a mea-
sure of the partial pressure of the gas and should not be confused with parts
per million by mass or mass per volume (i.e., mg/kg or mg/1). Concentra-
tions may be expressed also as mass per unit volume of vapor, such as )Jg/
m3 or mg/1. The basic relationship between partial pressure and mass per
unit volume is:
IQ~6MW
[3.51]
where MW (|Ug/mole) denotes the molecular weight of the contaminant used
to calibrate the detector (may not be the actual contaminant being moni-
tored) and R represents the gas constant (8.2 x 10"5 m3-atm/mole-K), and T
= 293 °K (20°C). Equation [3.51] is essentially the Ideal Gas Law where
C (ppm ) 10"6 represents the partial pressure of the gas being monitored.
It is important to recognize that expression of gas concentrations in volume/
volume units is meaningless unless one also specifies the calibration com-
pound. Thus, a total contaminant concentration of 100 ppmv measured on a
portable FID calibrated to methane must be expressed as 100 ppmv-methane
to have meaning (e.g., a gasoline vapor stream reported to have a total con-
taminant concentration of 100 ppmv-methane is not equivalent to a reported
total concentration of 100 ppmv-hexane).
3.114
-------
Chapter 3
For vapor extraction-based technologies, vapor concentrations should be
reported and recorded in mass/volume units, as this facilitates the calcula-
tion of removal rates Rv (mass/time) and any confusion is eliminated, where
Rv is the product of the flow rate (volume/time) and vapor concentration
Cva or (mass/volume):
[3.52]
Here, use of the flow rate expressed in standard units, Q*, is indicated as
most gas samples are analyzed from sample containers maintained at 1 atm
pressure. The cumulative contaminant recovered by volatilization, Tv
(mass), is then computed by integrating the recovery curve over time:
t
Tv = \Rvdt [3.53]
(=0
Figure 3.43 (on page 3.1 13) presents sample total hydrocarbon vapor
concentrations, calculated removal rates, and cumulative amount recovered
by volatilization for a vapor-extraction application.
As a general rule for vapor extraction-based technologies, the contami-
nant concentration in extracted gas usually declines over time, unless sig-
nificant process modifications are made after start-up (such as the
installation of an air-sparging system). In many soil vacuum extraction
systems, three performance phases are evident. These are illustrated con-
ceptually in figure 3.44 (on page 3.1 16). During the initial "flushing
phase", the first volume of soil gas is extracted from the contaminated soil.
Contaminant vapor concentrations are relatively high during this phase,
reflecting the long-term soil-gas equilibria condition. During the second
"evaporation phase", contaminants are rapidly removed from the more per-
meable soils, through which extracted gas readily flows. As these more
permeable channels become cleaner, the final, or "diffusion phase," begins.
During the diffusion phase, contaminants are removed more slowly from
the less-permeable portions of the formation and the removal rate is limited
by diffusion from the less-permeable soils into the more-permeable flow
channels (Subsection 3.2.1 1). In some cases, all three phases are not evi-
dent and the shape of the curve depends greatly on the contaminant mixture
and site conditions.
3.115
-------
Process Identification and Description
Figure 3.44
Generalized Concentration or Mass of Contaminant Observed
In Gas Extracted from a Vacuum Extraction System
Time
If a vapor extraction system is shut down during the diffusion phase,
contaminant vapors continue to diffuse away from the contaminant source
and the average soil-gas concentration in the /one of induced airflow slowly
increases. Consequently, higher concentrations are observed when the
system is restarted. This phenomenon is referred to as the "restart spike."
Figure 3.45 (on page 3.117) illustrates this phenomenon at a tetrachloroeth-
ylene (PCE)-contaminated site. In this study, both extraction and air
injection were used.
While total hydrocarbon analysis of extracted vapors is used to deter-
mine the removal rate and cumulative amount removed, this data cannot be
used to determine whether declines in vapor concentration are the result of
reductions in residual contaminant levels, composition changes (i.e.,
"lighter" compounds being removed first), or mass-transfer limitations
(Subsection 3.2.11). In addition, these data are usually insufficient to sat-
isfy regulations, which may require specification of selected compounds
(e.g., benzene). For this reason, it is recommended that a compositional
analysis be performed on selected extraction gas samples. Composition
analyses are discussed in Subsection 3.3.1.
3.116
-------
Chapter 3
In a highly diffusion-limited situation, the relative concentrations of
high-vapor pressure and lower-vapor pressure compounds will not change
greatly over time, even if the total contaminant concentration drops substan-
tially. Conversely, if a site is not highly diffusion-limited, the relative ratio
of highly volatile to less volatile contaminants will decrease more rapidly.
As an example, figure 3.43 (on page 3.113) presents composition data from
the first 120 days of vapor extraction at a service station site (Johnson et al.
1991). The general trend exhibited by the vapor concentration data is con-
sistent with expectations; vapor concentrations are declining substantially
during this time frame. In addition, the vapors of the relatively "heavier"
hydrocarbon components are gradually becoming richer. The composi-
tional change over the 120 day period, however, cannot account for the 100-
fold decrease in vapor concentration and this indicates remediation is being
controlled by mass-transfer limitations.
Figure 3.45
Tetrachloroethylene (PCE) Concentration in
Extracted Gas from a Site Near Milan, Italy
80 100
Time (Days)
Data from Castalia 1990
3.117
-------
Process Identification and Description
3.6.2 Respiratory Gas Monitoring
For bioventing applications, or where an aerobic biodegradation contri-
bution to removal is to be estimated, the data discussed in Subsection 3.5.1
should be supplemented by respiratory gas (O2 and CO2) monitoring data.
Monitoring the extracted vapor stream is discussed in this subsection and
using in situ soil-gas respiratory gas monitoring is discussed in Subsection
3.5.5.3. Here, it is assumed that complete aerobic degradation of the con-
taminant occurs; for example, the degradation of benzene is expressed as:
C6H6 + — O2 -» 6CO2 + 3H2O [3.54]
Equation [3.54] indicates that by monitoring oxygen utilization or CO2
production one should be able to estimate aerobic biodegradation rates. An
important consideration commonly overlooked in practice, however, is that
there may be other in situ processes (natural respiration, abiotic reactions,
etc.) that use or produce CO2 and O2. Currently, it is felt that monitoring O2
use provides a better estimate of biodegradation than does measuring CO2
production, because of the presence of carbonates, especially in sandy soils.
Working on a diesel-contaminated site in the Netherlands, van Eyk and
Vreeken (1989) found that much of the CO2 produced by biodegradation
went into carbonate formation and was not evolved as gaseous CO2.
Hinchee and Ong (1992), working at a number of bioventing sites, found
that, in low-pH soils (pH <6.0), biodegradation estimates based on CO2
were comparable to estimates based on O2. At higher pH levels, however,
CO2-based estimates were found to substantially underestimate biodegrada-
tion.
Typically, at the onset of remediation, O2 concentrations are low (<5% v/
v), and CO2 (>15% v/v) concentrations are high. Over time, increases in O2
and decreases in CO2 are observed in the extracted vapor streams as well as
in soil-gas monitoring probes. When the O2 and CO2 concentrations ap-
proach background levels, biodegradation has slowed and either the site is
clean of biodegradable compounds, or some other limitation to biodegrada-
tion has intervened.
Figure 3.46 (on page 3.119) presents data collected at a JP-4 contami-
nated site. The O2 and CO2 concentrations in the extraction gas over a 68-
day period are illustrated. The venting system was shut off on day 48,
3.118
-------
Chapter 3
Figure 3.46
Oxygen and Carbon Dioxide Concentrations in the Venting of
Gas Versus Time at the Hill AFB, Utah, Soil Venting Site
December 18,1988 to April 1,1989
25
20 -
10-
5 -
Background O2 Concentration
—T~
10
20 30 40 50
Cumulative Venting Time (Days)
60
70
Source Smith, Dupont, and Hinchee 1991
corresponding to the recorded drop in O2 concentration. The "background"
CO2 and O2 levels are extraction gas measurements from a similar vapor
extraction process conducted at a nearby uncontaminated site.
In using either O2 or CO2 measurements to estimate biodegradation, it is
necessary to estimate background O2 consumption and/or CO2 production.
For reference, the atmosphere contains approximately 20.9% O2 and 0.35%
CO2. Natural, uncontaminated soil gas, however, contains somewhat less
O2 and more CO2 than the atmosphere because of natural soil respiration
(figure 3.46). This was shown by both Dupont, Doucette, and Hinchee
(1991) and Miller, Hinchee, and Vogel (1991). Vent wells and soil vapor
extraction systems were installed in nearby uncontaminated soils. These
soil vapor extraction systems were operated at rates similar to the soil vapor
extraction systems in contaminated soils in order to quantify the back-
3.119
-------
Process Identification and Description
ground respiration rate. This approach, although ideal, is not feasible at
many sites.
Given O2 and/or CO2 levels (% by volume) in the extracted vapors for
the remediation and background sites, loss rates due to aerobic biodegrada-
tion, Rb (mass/time), are estimated through the following process:
1. Calculate molar uptake of oxygen, Mb (moles O2/time), correct-
ing oxygen depletion for natural respiration from a "back-
ground" site:
.
d)\ 100% )RT{m* -vapor)
where Cb (% O2), Cr(% O2), P (atm) , R (8.2 x ICr5 atm-m-Vmole-
K), and T (°K) denote the concentration of O2 measured in ex-
tracted vapors at the background site (under similar operating
conditions), concentration of O2 measured in extracted vapors at
the remediation site, pressure at the flow rate monitoring point,
gas constant, and absolute temperature at the monitoring point,
respectively; and
2. Divide Mb ((moles-O2/d) by the stoichiometric ratio S (moles O2/
moles contaminant), and multiply by the contaminant (C) mo-
lecular weight, Mw (g/mole), to obtain the contaminant degrada-
tion rate Rb (g/d):
* C ' [3.56]
^2 i moles-C
"h(~~d~)~'"(d ksjca^^o,
For example, assuming the degradation of hexane (C6HI4):
C6//14 + y02 -* 6CO, + 1H20 [3.57]
then S = 19/2 moles of O2 are consumed for each mole of hydro-
carbon degraded.
An alternative approach based on carbon isotope analysis of extraction
gas samples can be used to distinguish, at least qualitatively, between back-
ground CO2 and CO2 resulting from hydrocarbon degradation. Here, advan-
3,120
-------
Chapter 3
tage is taken of the knowledge that petroleum hydrocarbons tend to have a
higher ratio of 12C/13C relative to other carbon sources in the subsurface.
Therefore, by measuring the 12CO2/13CO2 ratio in extracted gas from con-
taminated and uncontaminated sites, an inference as to the source of the
CO2 can be made. Aggarwal, Means, and Hinchee (1991) describe a
method for determining the source of CO2-hydrocarbon biodegradation vs.
background respiration based on the stable carbon isotope composition of
the soil gas. This can be a relatively inexpensive (=$25 to $150/sample)
qualitative approach to determining whether aerobic biodegradation of hy-
drocarbons is occurring.
3.6.3 Cost
Costs should be monitored during the remediation process in order to be
able to evaluate the cost-effectiveness of operating the vapor extraction-
based system. Following are cost components of vapor extraction-based
systems:
• Capital Cost: The cost of system design, construction, and in-
stallation; it usually includes costs of permitting and pre-opera-
tional studies;
• Utilities: Includes electrical usage by blower/vacuum pump and
auxiliary equipment operation, as well as the cost of any supple-
mental fuel for offgas treatment;
• Operation and Maintenance: The cost of basic system operation,
maintenance, and repair; in most cases, labor is the most signifi-
cant component of this cost;
• Monitoring: The most variable cost associated with a vapor
extraction-based technology system's operations, and at many
sites the largest component; it consists of analytical charges and
labor, with labor frequently being the largest element; and
• Miscellaneous: All other costs, including equipment or space
rental, ongoing permit costs, and the loss of property use associ-
ated with the vapor extraction-based system.
To evaluate cost-effectiveness of the vapor extraction-based technology
system, costs and mass of contaminant removed/degraded should be plotted
vs. time. For example, the unit cost for contaminant removal in dollars per
pound recovered can be calculated and displayed as a function of time. The
3.121
-------
Process Identification and Description
resulting plot usually resembles an exponential increase, reflecting that as
the remediation nears completion cost per unit mass of contaminant re-
moved increases. In some cases, it may be possible to negotiate an end
point based upon this economic analysis.
3.6.4 Environmental Factors
In addition to the process variables discussed above, certain environmen-
tal variables may influence vapor extraction-based technology performance.
Even if they do not directly affect performance, they may influence inter-
pretation of monitoring data. These include barometric pressure, tempera-
ture, relative humidity, and rainfall.
A variety of conventional techniques are available to measure barometric
pressure. In addition, in most areas, weather reports that include accurate,
reliable barometric pressure measurements can be obtained. It is more com-
mon, however, simply to assume standard barometric pressure, adjusted for
altitude. Barometric pressure fluctuations can significantly affect in situ
vapor-phase pressure measurements, as these are always relative to the
current atmospheric pressure. This effect is most significant when attempt-
ing to measure induced pressure changes that are of the same magnitude as
the barometric pressure fluctuations (as when measuring pressure distribu-
tions away from injection/extraction wells during short-term pilot tests).
Such effects can be significant in regions with large diurnal temperature
fluctuations (e.g., daytime temperature of 38°C (100°F) to nighttime tem-
perature of 16°C (61°F))..
Air temperature fluctuations over a few hours or days have little or no
impact on subsurface temperature. In situ soil temperatures and extracted
gas temperature may be simply monitored with thermocouples and ther-
mistors. Some authors have reported using extracted gas and in situ tem-
perature measurements as qualitative indicators of microbiological activity.
For the most part, however, temperature measurements currently do not
play an important role in most monitoring plans, except in very harsh cli-
mates and where there are technology modifications discussed in Section
3.6.
Soil moisture content can have a significant impact on vapor extraction-
based system performance. A high-soil moisture content can cause reduc-
tions in soil-gas permeability and diffusivity, which result in poor gas flow
3.122
-------
Chapter 3
and poor mass-transfer characteristics. A low-soil moisture content (below
the wilting point) can slow microbial degradation. Dupont, Doucette, and
Hinchee (1991) found soil moisture to be a very important variable in bio-
venting. Another consideration is that under very dry conditions, a soil's
adsorptive capacity for hydrocarbons increases, thereby adversely affecting
the hydrocarbon/soil equilibria.
Measurements of relative humidity in the extraction gas and ambient air
(or injection air), coupled with rainfall measurements, can be used to de-
velop a rough in situ water-mass balance. The difficulty lies in estimating
infiltration from rainfall. This exercise, however, is useful in determining
the rate of drying induced by a vapor extraction-based process. Soil mois-
ture may also be measured directly. Smith, Dupont, and Hinchee (1991)
utilized neutron probe access tubes to measure in situ soil moisture. Figure
3.47 illustrates the results of their measurements before and after the surface
application of water at a site at Hill AFB, Utah. Soil moisture may also be
Figure 3.47
Cumulative Hydrocarbon Removal (Volatilized and
Biodegraded) Hill AFB, Utah,Soil Venting Site
(December 18,1988 to November 14,1990)
JFMAMJJASOND
1989
Date
JFMAMJ JASON
1990
Source: Smith, Dupont, and Hinchee 1991
3.123
-------
Process Identification and Description
measured gravimetrically in collected soil samples. The authors are not
aware of any optimal soil moisture levels; the practitioner should be aware
that system performance can be enhanced or impaired by changes in soil
moisture levels, and there are few practical options for controlling the soil
moisture (capping to reduce infiltration, irrigating to increase moisture con-
tent, etc.). In general, in arid environments and high-permeability soils,
low-soil moisture effects are more likely to be seen. In more moist environ-
ments and low-permeability soils, high-soil moisture effects are more likely
to be seen.
3.6.5 Subsurface Monitoring
When the extracted hydrocarbon vapors and environmental factors only
are monitored, ability to assess system performance is limited; in particular,
these data generally cannot be used to determine the extent of remediation
or the factors and phenomena that are limiting system performance. Ex-
tracted gas vapors represent a volume average of all vapors flowing to a
given extraction well. Therefore, measurements are biased toward those
zones that supply greater flow, as is illustrated in figure 3.48 (on page
3.125). At best, monitoring aboveground process streams allows estimation
of the removal rates by volatilization, cumulative amount volatilized, and,
in some cases, aerobic biodegradation rates.
To gain a better picture of the extent of remediation and processes that
affect system performance, some level of in situ monitoring is required. A
number of options for doing this are discussed in the Subsections immedi-
ately following.
3.6.5.1 In Situ Soil-Gas Monitoring
Soil-gas monitoring is a quick and effective means for assessing the ef-
fect of remediation in an area around a soil-gas monitoring location. Soil-
gas samples can be obtained over time and analyzed for total hydrocarbon
concentration and composition. Over time, a reduction in concentration and
shift in composition will be seen in those areas that are being remediated.
Table 3.8 (on page 3.126) presents soil-gas data from a vapor extraction
site. These data illustrate how some areas at a vapor extraction site were
more effectively remediated over time. For example, the area surrounding
location HB-7D did not show as significant a reduction in total hydrocarbon
concentrations as other areas.
3.124
-------
Chapter 3
Figure 3.48
Schematic of Hypothetical Subsurface Flow and Relation
Between Extracted Gas Concentration and In Situ Soil
Gas Concentrations in Stratified Flow System
HC = 100 ppmv; O2 = 18%; CO2 = 1
• Extracted
Vapors
Multilevel
Soil Gas
Monitoring
Point
E V T T Y
| HC = 5000 ppny O2 = 1%; CO2 = 14%]
[Diffusive Flux|
• [Advective Flow]
[HC = 80 ppmv; O2 = 19%; CO2 = 1
t
[Diffusive Flux |
Low
Permeability
High
Permeability
Low
Permeability
HC = 100 ppmv, O2 = 1%, CO2 =15%
Soil-gas samples can be extracted from small diameter tubes driven into
the soil, or from tubes permanently installed in a borehole as shown in fig-
ure 3.49 (on page 3.127), which also shows the optional installation of ther-
mocouples). Although the construction of in situ soil-gas monitoring points
is relatively simple, a number of factors deserve careful attention. In gen-
eral, numerous narrowly-screened points are more useful than one single,
broadly-screened point. Small-diameter tubing (0.6 cm (1/4 in.) or even 0.2
cm (1/16 in.)) is usually preferred to larger diameter tubing, as this reduces
the volume of gas in the sampling tube and allows easier, more effective
purging and sampling. Often the tubing is strapped to the outside of a sup-
3.125
-------
Process Identification and Description
Table 3.8
Vadose Zone Monitoring Well Vapor Concentrations and Compositions
Measured During a Vapor Extraction Application
Well ID
Distance from
Venting Well Total
HB-25 Depth Hydrocarbons
Date (ft) (ft) (mg/1)
% Vapor Composition
Boiling Point Fraction
2 3 4 .5
HB-6S
HB-6S
HB-6D
HB-6D
HB-6D
HB-6D
HB-7S
HB-7S
HB-7S
HB-7S
HB-7M
HB-7M
HB-7M
HB-7M
HB-7D
HB-7D
HB-7D
HB-14S
HB-14S
HB-14M
HB-14M
HB-14D
HB-14D
HB-14D
HB-14D
1/24/89
12/9/89
1/24/89
1/25/89
2/28/89
12/9/89
1/24/89'
1/25/89
2/28/89
12/9/89
1/24/89
1/25/89
2/28/89
12/9/89
1/24/89
2/28/89
12/9/89
1/24/89
12/9/89
1/24/89
12/9/89
1/24/89
1/25/89
2/28/89
12/9/89
8
8
8
8
8
8
15
15
15
15
15
15
15
15
15
15
15
30
30
30
30
30
30
30
30
10
10
40
40
40
40
10
10
10
10
25
25
25
25
40
40
40
10
10
25
25
40
40
40
40
13
0.15
308
168
91.8
372
32.7
289
11.5
3.2
298
474
202
166
239
241
106
63
02
18.2
0.42
251
199
no
44
15.4
133
32.0
37.8
27.7
13.1
45.6
519
27.0
14,2
362
34.8
173
0.5
43.2
286
269
58.7
27 1
47.8
17.9
40.0
43.4
6.4
143
34.6
18.7
49.1
50.0
52.9
39.1
33.3
34.6
25.2
38.5
508
26.8
32.7
5.6
48.3
42.4
41.5
270
333
29.7
34.2
45.9
450
443
407
30.8
25.3
17.6
11.8
18.0
38.1
15.9
10.7
32.2
107
12.3
27.0
38.6
46.7
8.2
25.5
27.9
12.7
124
21.4
27.6
13.2
10.8
40.4
362
19.2
26.0
1.2
0.4
1.4
9.2
5.2
2.8
15.7
51.3
0.7
11.4
11 4
44.7
0.3
3.5
3.4
1.6
14.3
1 1
13.4
0.8
0.8
8.1
77
0.0
16.7
0.0
0.0
0.0
0.6
0.0
0.0
0.0
20.0
0.0
0.0
0.0
24
00
00
02
0.0
12.9
00
6.8
0.0
0.0
0.8
1.1
Source' Johnson etal. 1991
Boiling Point Fractions. 1- methane - isopentane
2: isopentane - benzene
3' benzene - toluene
4: toluene - xylenes
5: >xylenes
<28'C)
28 - 80'C)
80- 111'C)
111 -144'C)
>144'C)
porting rod, such as a one-inch PVC pipe. Hinchee (unpublished data) has
installed as many as 12 discretely-screened sampling points in a single 30 m
(100 ft) boring. Collection of soil-gas samples under vacuum conditions is
discussed in Subsection 3.3.4.2.
3.126
-------
Chapter 3
Figure 3.49
Typical Vapor Extraction Monitoring Point Construction Detail
Finish Concrete
to Drain Away
from Box
r
Water Tight C
Iron Well Bo
Q
T
V?
Gravel (for Box Drainage) \
Bentonite y
Gravel /
Bentonite ]
Backfill
Bentonite
Gravel
Bentonite
Backfill
Bentonite
Gravel
>-
ast
X
uick Couples
/ Metal Tags
' / } 1 Ground;
^^r / •< Finish at Grade Also
3 3 ^/ 1 Acceptable
1,
1
"*^
i
— ^~_
) ;
i
\ i
1
i \
2ft
r
k
ift
f
i
2ft
r
\
1 t
>
) :
>
\ i
2ft
r
k
1ft
r
»
2ft
r
7
^ ~~^ 1/4" Polyethylene Tubing
\ or Other Material
) :
:
2ft
r
h
1.5ft
r
Bore Hole
hermocouple
with Leads
Dimensions will vary for specific installations
3.127
-------
Process Identification and Description
3.6.5.2 Subsurface Vapor Phase Pressure Distribution
The installations used for soil-gas monitoring can also be used for sub-
surface vapor phase pressure monitoring. As in the case of field pilot-scale
tests, these data can be used to assess the zone of containment of a vapor
extraction-based system. In addition, the data can be used to estimate the
subsurface pressure gradient and that estimate can be combined with perme-
ability information to calculate vapor-flow rates through different soil zones
(note that the absence or presence of a measured pressure does not indicate
flow or remediation). Over time, changes in subsurface pressure distribu-
tions can indicate changes in other parameters, such as soil permeability.
The practitioner should be equipped with a wide range of gauges, as pres-
sures (vacuums) measured in extraction wells and subsurface monitoring
points can vary from a few tenths of an inch to a few feet of water; injection
pressures may be as high as several psi. The theoretical limit for vacuum
measurements is =10 m (33 ft), but in actual practice, it is rare that a
vacuum in the formation of more than 1.5 to 3 m (5 to 10 ft) of water is
achieved.
3.6,5.3 Subsurface Respiratory Gas Monitoring
In bioventing or other vapor extraction-based systems the monitoring of
subsurface respiratory gases (O2 and CO2) can be an important part of a
vapor extraction-based process evaluation if the operator wishes to assess
the contribution of microbiological degradation. To ensure maximum bio-
degradation, minimum oxygen levels of 2% to 5% v/v should be maintained
throughout the formation (Miller 1990). Smith, Dupont, and Hinchee
(1991) found that, although exhaust gas O2 rose to 20% within 2 months of
the initiation of venting, 6 months were required for O2 to rise above 5% in
some parts of the site.
Transient analyses of soil-gas samples can be used to estimate in situ
biodegradation rates. These in situ respiration tests are conducted by shut-
ting down the extraction system and measuring O2 depletion and CO2 pro-
duction. Figure 3.50 (on page 3.129) presents plots of measurements.
Consistent with figure 3.50, oxygen utilization appears to be constant (oxy-
gen concentrations decrease linearly over time), based on the limited data
published to date. Following are the steps in converting soil-gas oxygen
depletion data to degradation rates:
3.128
-------
Chapter 3
Figure 3.50
Oxygen Utilization and Carbon Dioxide Production in Various
Phases of a Bioventing Project at Tyndall AFB, Florida
201
50 75
Time (Hours)
Source- Miller 1990
1. Correct oxygen depletion for natural respiration from a "back-
ground" site (in practice, this background rate is very low, near
zero, at most sites);
2. If data are linear and zero-order kinetics are assumed, calculate
the slope of the O2 concentration vs. time plot. The magnitude
of the slope is defined to be k' (%/h);
3. Convert k' (%/h) to ((moles-O2/m3-vapor)/d) by using the Ideal
Gas Law:
moles-Oi \_k'(%lh) (h\ P I moles-O2}
(m3 - vapor)-d) 100 UJtf7\m3 -vaporJ [3'58^
where P (atm) denotes the total pressure in the subsurface (1 atm
after extraction is terminated), R (m3-atm/mole-K) denotes the
universal gas constant (8.2 x 105 m3-atm/mole-K), and T (°K) is
the absolute temperature;
4. Divide k' ((moles-O2/m3-vapor)/d) by the stoichiometric ratio S
(moles O2/moles contaminant) to obtain the contaminant degra-
dation rate k" ((moles-contaminant/m3-vapor)/d):
3.129
-------
Process Identification and Description
moles - O2
,„, moles - contaminant},, {(m3 -vapor) —u j ,~ •»,
V (m3 - vapor) -d ) ^ molesO2 \ " J
moles contaminant)
For example, assuming the degradation of hexane (C6H14):
Q#,4 + y 02 -» 6C02 + 7//20 [3.60]
then S = 19/2 moles of O2 are consumed for each mole of hydro-
carbon degraded; and
Convert to K ((g-contaminant/kg-soil)/d), the mass degradation
rate of contaminant, normalized per unit mass of contaminant-
containing soil, by multiplying k" by the molecular weight of
contaminant MW (mg-contaminant/mole-contaminant), the va-
por-filled porosity eA (m3-vapor/m3-soil), and dividing by the
soil bulk density pb (kg-soil/m'-soil):
I ing — contaminant } ,1 moles — contaminant \f mg — contaminant \
A - = - 1 - - A/Vv - I
^ (kg — soil) — d j \ (m~ —vapor) — d ) \moles — contaminant)
[3.61]
-soil m -soil
This should be considered an estimate of biodegradation requir-
ing numerous assumptions on the limited number of sites at
which in situ respiration tests have been used to estimate biodeg-
radation rates and at which soil hydrocarbon concentration re-
duction data was available, this estimate appears to be within a
factor of 2 or 3 (Miller 1990; Downey 1993; Wilson 1994; and
Hinchee 1994).
3.6.5.4 Soil Borings and Site Sampling
In many cases, collection and analysis of soil samples is required by
regulatory agencies in order to demonstrate proof of remediation. Evalua-
tion of process performance typically is based on limited initial and final
soil samplings and, at times, on some intermediate sampling. The difficulty
3.130
-------
Chapter 3
in using soil sampling to evaluate the process lies in the extreme variability
inherent in analyses of soils from a site contaminated by hydrophobic or-
ganics. Working at a hydrocarbon-contaminated site at Tyndall AFB,
Florida, Miller (1990) established two treatment plots, each approximately
2.4 by 4.8 m (8 by 16 ft) in area and 1.5 m (5 ft) deep. Despite every effort
to select uniformly contaminated plots, the 21 soil samples collected
yielded coefficients of variation of approximately 100%. Prior to treatment
of a JP-4 contaminated site at Hill AFB in Utah, Oak Ridge National Labo-
ratory personnel analyzed 259 soil samples from the soils beneath a spill of
94,630 L (25,000 gal). The highest concentration exceeded 20,000 mg/kg
and many samples had concentrations above 1,000 mg/kg. Of all the
samples, however, only 77 (30%) showed contamination above detection
limits. The point is, although soil sampling is an important part of verifying
remediation broadly, soil sampling data must be carefully interpreted when
used to quantitatively evaluate process performance.
3.6.5.5 Groundwater Sampling
Like soil sampling, groundwater sampling is often required to demon-
strate system effectiveness. Many closure plans include some period of
posttreatment groundwater monitoring (typically, one year) to demonstrate
long-term effectiveness. Since the goal in many cases is to protect ground-
water quality, there is reason to include groundwater sampling in monitor-
ing plans. In processes aimed directly at treating the saturated zone (e.g.,
air sparging), these data are relevant in assessing system performance and
assuring containment. Typically, the groundwater samples will be analyzed
for selected contaminants and dissolved oxygen. In air sparging, some have
used these data to assess the area being affected by sparging wells. It is
assumed that elevated dissolved oxygen levels and decreased contaminant
concentrations are indications that the sampling point lies within the zone of
sparge-air flow. Great care should be taken in interpreting groundwater
monitoring data, as monitoring wells are ideal short-circuiting pathways for
both injection and extraction gas flow.
3.6.5.6 Water Table Fluctuations
It is well known that vapor extraction-based processes can induce signifi-
cant water table elevation changes because of changes in the subsurface
pressure. Figure 3.51 (on page 3.132) illustrates this phenomenon for a
3.131
-------
Process Identification and Description
vapor extraction system; in which it can be seen that water table elevation
changes occurred in response to applied vacuum changes at an extraction
well.
Water table fluctuations are important for at least two reasons. First,
from a practical standpoint, it is possible for the water table to rise into an
extraction vent and block or reduce airflow. In low-permeability soils,
where higher vacuums may be required, this can be a serious limitation.
Second, in rising, the water table may cover or mask some of the contami-
nated soil. This is particularly true for LNAPLs, such as a fuel hydrocar-
bons, which tend to spread on top of the water table. In many cases,
changes observed in extracted contaminant vapor concentrations are the
direct result of water table elevation changes; water table rises tend to corre-
late with decreases in contaminant vapor concentrations. Water table eleva-
tion changes may result from seasonal changes in precipitation,
groundwater recovery process changes, or changes in other external pump-
ing conditions (e.g., periodic use of a neighboring well). Obviously, mea-
Figure 3.51
Vapor Flowrate and Groundwater Upwelling Dependence on
Applied Vacuum (a) and,Transient Watertable and Subsurface
Vacuum Response (b) at a Vapor Extraction Application
(a)
0.5
(b)
0.4-
£,0.3-
I
£ 0.2
0.1 -
0.0
a - Vacuum Increase
• - Water Table Upwelling
.1
D CCP
1 10
Time (min)
4-
3-
2-
- Flowrate
- Upwelling
100 10 20 30 40 50 60
Vacuum (in. H2O)
50
40
-30
-20
10
0
* Ft H,O denote vacuums expressed as equivalent water column heights
" Relative to level 8 20 in H2O vacuum
Source: Johnson etal 1991
3.132
-------
Chapter 3
surement of water table elevations during the operation of vapor extraction-
based processes is very important. See Subsection 3.3.4.5 for further dis-
cussion.
3.6.5.7 Tracer Gas Tests
Inert tracer gases (e.g., He and SF6) can be used to gain a better under-
standing of the flow dynamics of vapor extraction-based processes. For
example, a small quantity of tracer gas can be injected into the subsurface
some distance from an extraction well, and its appearance at the extraction
well can be monitored in order to assess the induced flow strength through
the injection location.
Tracer gas tests are currently the best means to quantitatively assess flow
dynamics, although Kerfoot (1992) has reported the development of an in
situ tool that purportedly measures soil-gas velocities directly. The most
valuable use of tracer gases at this time, however, is in the assessment of
vapor extraction-based processes that involve air injection. As mentioned
in Subsection 3.1.2, there is significant concern that improperly operated
systems will accelerate the uncontrolled migration of contaminant vapors
away from the treatment area. Tracer gas tests provide a simple and quick
experimental means to define appropriate safe operating conditions for
wells. They are conducted by mixing a tracer gas with the injection air
stream and then monitoring it in the extraction system and near any sensi-
tive vapor receptors (buildings, conduits, etc.). Injection and extraction
rates are then modified until all of the tracer gas is recovered.
3.7 Technology Variations: Combined In
Situ Soil Heating and Vapor Extraction
This monograph principally addresses soil vapor extraction and, to a
lesser extent, the complementary technologies of air sparging and biovent-
ing. In the authors' experience, these are currently the soil vapor extrac-
tion-based technologies most likely to be applied. There are, however,
other vapor extraction-based technologies that deserve mention, technolo-
gies that combine in situ soil heating and vapor extraction.
3.133
-------
Process Identification and Description
Conventional vapor extraction equipment is combined with a means by
which energy is supplied to the subsurface in order to elevate the subsurface
temperature. This is done to broaden the range of applicability of vapor
extraction-based technologies by enhancing either biological activity
(through gentle heating to 30 to 40°C (86 to 104°F); see Subsection 3.2.12)
or the rate of volatilization (through aggressive heating to >100°C (212°F)).
As a general rule, biological rates are assumed to double for every 10°C
(SOT) increase in temperature; nonetheless, of the two mechanisms, the
potential for increasing removal by volatilization is much greater. This can
readily be seen through an assessment of increases in pure component vapor
pressures with equation [3.4] (see Subsection 3.2.2):
[3.4]
RT
where AHv is the molar heat of vaporization (cal/mole), R is the gas con-
stant (1.99 cal/mole-°K), and B is a unitless constant. Values for AHv can
be found in standard thermodynamic tables (e.g., see Weast 1970); then, B
can be determined by solving equation [3.4] for B at a known temperature
and reference vapor pressure. Figure 3.52 (on page 3.135) shows the de-
pendence of pure component vapor pressure on temperature for a range of
hydrocarbons. If it is assumed that the soil can be heated to a given tem-
perature, most of the screening models derived in Section 3.2 can still be
used to estimate system performance. At a screening level, one can expect
removal rates to increase in proportion to increases in vapor pressure.
In diffusion-limited formations, beneficial effects in addition to the vapor
pressure increase result from increases in temperature. For an absolute
temperature change from T, to T2, vapor-phase diffusion coefficients in-
crease roughly by the factor (T/T,)3'2, thereby helping to enhance the rate of
vapor transport from low-permeability zones to regions of high- vapor flow.
In addition, for temperature increases above 100°C (212°F), the production
of steam from low-permeability, moist soils and its pressure-driven flow to
regions of high-vapor flow can help drive contaminant vapors out of the
low-permeability zones at a rate much higher than the natural diffusion rate.
To determine whether acceptable performance can be achieved at a given
temperature, a heat input requirement is first calculated, realistic heat input
rates are assessed, and the time required to achieve a target temperature is
estimated. Sufficient energy must be supplied to heat the soil, raise the
3.134
-------
Chapter 3
Figure 3.52
Vapor Pressure for Various Organic Compounds
as a Function of Temperature
10,000
1,000-
100-
10-
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Temperature (°C)
Data from: Perry, Chilton, and Kirpatrick 1973; and Boublik, Fried, and Hala 1984.
temperature of the soil moisture and contaminants, cause the soil moisture
and contaminants to volatilize, and then raise the gases to the desired tem-
perature. In order to raise the temperature of dry uncontaminated soil from
T, to T2, the minimum energy input must be AHs (cal/g-soil):
3,135
-------
Process Identification and Description
CP,SOII(T2-T{) [3.62]
where C soil (cal/g-°K) denotes the mean heat capacity of the soil over this
temperature range. Table 3.9 presents thermal data for various soils. If
water is present in the soil, the temperature is elevated above the boiling
point of soil moisture, and the process occurs at constant pressure, then the
following additional energy input AHw (cal/g-soil) can be approximated as:
A//w = 6W(Cp^(Tb - r,) + Affv + C^_v(T2 - Tb)) [3.63]
where 9w (g-H2O/g-soil), Cpw., (cal/g-liquid-water/°K), AHy (cal/g-H2O), and
C w v (cal/g-water-vapor/°K) represent the soil moisture content, mean heat
capacity of liquid water, heat of vaporization for water, and the mean heat
capacity of water vapor, respectively. An equation similar to equation
Table 3.9
Thermal Properties of Selected Materials
Porosity
Soils
Sand
Clay
Peat
Miscellaneous Materii
Water, liquid
Water, ice
Gasoline
Benzene
Kerosene
Wood
Air (dry at STP)
04
04
0.4
04
0.4
0.4
0.8
0.8
0.8
ils
Volumetric
Wellies;,
00
02
04
00
02
04
0.0
04
0.8
Thermal
Conductivity
(lO-'Cal/cm
sec-'C)
07
42
52
06
2.8
38
0 14
0.7
1 2
Volumetric Heat
Capacity Cp
(cal/cm'-'C)
03
05
07
03
05
07
035
0.75
1 15
1 0
045
0.76
0.56
0.67
0 64-0.93
0 00028
Source
a
a
a
a
a
a
a
a
a
b
a
b
b
b
b
b
Sources (a) Hillel 1982, (b) Perry, Chilton, and Kirkpatrick 1973
3.136
-------
Chapter 3
[3.63] can be formulated also for contaminants present (if they are solids at
ambient temperature, heats of melting must be incorporated). An equation
for heat required for air in the pore space can also be written, but this value
will be negligible in comparison with the energy already required, as heat
capacities of gases are much smaller than those of solids and liquids.
Given a total energy requirement, the minimum heating period duration
can be estimated by simply dividing the total energy requirement AHT (cal/
g-soil) by the expected heating rate QH (cal/d). This value and an estimate
of time for remediation at that target temperature are two quantities used to
estimate the efficacy and cost of any in situ soil heating plan. In many of
the heating methods described here, the rate of heating is likely to be the
limiting step in remediation.
A number of thermal enhancement processes being used or under devel-
opment are discussed below, and exhibit significant variations in heating
rates, temperature limits, energy efficiency, complexity, and cost.
3.7.1 Steam Stripping
In situ steam stripping typically involves:
• delivering steam to the contaminated zone via injection wells;
• heating the contaminated zone to vaporize the contaminants or
increase their mobility; and
• creating a pressure gradient to control movement of the contami-
nants and of the steam condensate front to a recovery point.
The injected steam travels some distance from the injection point and
then condenses. The energy lost because of cooling and condensation is
transferred to the formation. Contaminants in the soil may vaporize (due to
increased vapor pressure), become dissolved in the condensate front (due to
increased solubility), or be displaced (due to a reduction in viscosity and
capillary forces). These processes must be controlled to minimize possible
detrimental effects, such as contaminant "smearing" or enhanced vapor
transport away from the source area.
In situ steam stripping may be successful where the organic contaminant
consists of compounds with a low solubility in water and a boiling point
below about 250°C (482°F) (Nunno et al. 1989; Murdoch et al. 1989). The
process relies upon the same principles as conventional steam stripping in
two-phase distillation. Contaminants are vaporized then condensed at the
3.137
-------
Process Identification and Description
steam front, and are often displaced as much as volatilized. They may be
carried to the surface as a mixture of wet steam and organic vapor. A
vacuum system collects the vapor mixture for removal of the organic con-
taminants. The water vapor is condensed, and the insoluble organic phase
is collected for recycling or disposal. The offgas and water may be treated
further to remove any remaining traces of organics.
In situ steam stripping requires steam generation and injection, vapor
containment, and offgas collection and treatment system. The objectives of
design of injection and extraction wells and process flow are to permeate
the steam uniformly through the target zone, cause the condensate front to
be driven to the extraction wells, and assure containment of vapor migra-
tion. Often, impermeable surface covers are used to help control the flow
field.
3.7.1.1 Steam Stripping Technology
Three steam stripping methods have been used in field tests:
Method A - Steam injected into supply wells is drawn through the
contaminated zone by vacuum;
Method B - Steam and hot air are injected from drill bits as the
bits are rotated in the contaminated zone; and
Method C - Steam is injected below a bank of dense contaminant
to lift it into a stream of hot water. The hot water is
injected so as to move laterally over the bank of con-
tamination in order to intercept the rising contaminant
and carry it to a collection well.
Each method employs different injection and control techniques with the
object of achieving acceptable contaminant removal along with efficient use
of steam, given the combination of contaminants and geology at the site.
In Method A, steam injection wells are installed to inject steam_at or
below the level of the contamination. The wells are placed throughout the
formation so as to fully saturate the contaminated area. Well spacing de-
pends on the permeability of the formation. In field tests, well spacing of
approximately 1.8 m (6 ft) has been used at sites having a variety of soils,
including sandy silts, clay, and sand (Udell and Stewart 1989; Nunno et al.
1989). Steam pressure typically is about 40 kPa (6 psig) for well depths of
approximately 6 m (20 ft). Pressures as high as 79 kPa (11.5 psig) have
3.138
-------
Chapter 3
been used in laboratory tests. The laboratory studies indicate better con-
taminant removal at higher steam pressure, if the pressures can be reached
without fracturing the formation (Lord et al. 1990a & b). The contaminant
can be withdrawn using vacuum extraction wells in the formation (Udell
and Stewart 1989) or by applying a vacuum at the soil surface under a flex-
ible membrane liner covering (Lord et al. 1990a & b) or a vacuum bell
(Nunno et al. 1989).
Following are reported removal efficiencies under Method A:
• benzene, toluene, ethylbenzene, and xylene: 20% in clay and
99.5% in sand (Nunno et al., 1989);
• naphthalene: 60% in clay and 99.9% in sand (Nunno et al.
1989);
• polycyclic aromatic hydrocarbons: 35% in bog and 97% in sand
(Nunno et al. 1989);
• phenol: 20 to 80% in clay or sand (Nunno et al. 1989); and
• acetone, xylene, ethylbenzene, and 1,2-dichlorobenzene: 98%
(Udell and Stewart 1989).
Laboratory experiments have indicated that steam stripping efficiency may
depend on the polarity of the compound as well as its vapor pressure (Lord
etal. 1990a&b).
Method B employs a process tower, which supports a pair of cutting
blades at the end of a hollow shaft. The cutting blades are rotated in oppo-
site directions as they are lowered vertically into the soil. The blades can
reach depths of 9.1 m (30 ft). The cutting heads break up the soil and
thereby assure the uniform flow of gases. Steam, at 200°C (392°F), and
compressed air, at 135°C (275°F), are piped through the shafts to nozzles
located on the cutter blades. A steel shroud (3.0 m (10 ft) by 1.8 m (6 ft) by
2.1 m (7 ft)) covers a 2.2 m (7.3-ft) by 1.2 m (4 ft) area of soil undergoing
treatment. Larger areas are treated by multiple application of overlapping
treatment blocks (La Mori 1989).
A blower keeps the shrouded area under vacuum in order to enhance the
flow of gases from the soil and to prevent leakage to the outside environ-
ment. The offgases are passed to the treatment train where the water and
organics are removed by condensation in coolers and treated in carbon ad-
sorption beds. The air is filtered and recycled to the soil by a compressor.
3.139
-------
Process Identification and Description
Water is removed from the liquid stream by a four-stage separator followed
by batch distillation and is then recycled to a cooling tower. The condensed
organics are collected for recycle or disposal (La Mori 1989).
Removal of volatile organic compounds (EPA Method 8240) is reported
as greater than 96%, with average pretreatment soil concentrations in the
range of 1,114 to 3,954 mg/kg. Removal of semivolatile organic com-
pounds (EPA Method 8270) is reported as ranging from 11% to 93%, with
average pretreatment soil concentrations ranging from 1,014 to 12,116 mg/
kg (La Mori 1989).
Method C is used to effect the contained recovery of oily wastes. Injec-
tion and production wells are drilled in a pattern designed to sweep oily
waste accumulations with steam and hot water. Low-quality steam is in-
jected below the deepest penetration of organic liquids. The steam con-
denses, causing an upward flow of hot water. The upward flow dislodges
and sweeps organic liquids up into more permeable regions. Hot water is
injected above the natural impermeable barriers heating and mobilizing the
main accumulation of oily wastes. After organic liquids are mobilized
above the impermeable barriers, hot water injection into and water with-
drawal from the production wells are controlled so as to sweep accumulated
oily wastes through the more permeable regions.
Oily wastes are contained vertically by controlling temperatures during
hot water displacement. Downward penetration of oily wastes is reversed
by the thermal expansion of the heated organic liquids. Flotation of the
heated organic liquid phase is limited by injecting cooler water above the
oily waste accumulations.
One-dimensional experiments showed a relationship between residual
saturation of the oily contaminant and the temperature of the displacing hot
water. As the sweep-water temperature increased from ambient to approxi-
mately 69°C (157°F), for tests without surfactant addition, the reduction in
oily waste residual increased from less than 20% to more than 60% for soil
with up to a nominal 3% by weight initial oily waste saturation. Under
similar conditions, tests with surfactants increased the organic reduction to
approximately 90%. Three-dimensional simulations were used to validate
the operation of the cold-water cap and the displacement efficiencies shown
in the one-dimensional tests (Johnson and Guffey 1990).
3.140
-------
Chapter 3
3.7.1.2 Advantages and Disadvantages of Steam Stripping
Because the petroleum industry uses steam injection for oil recovery,
there is extensive experience with the technique. Steam injection can be
applied to semivolatile as well as to volatile compounds. The organics can
often be collected as a separate phase for reprocessing and reuse.
Depending on the extent of contamination, the amount of contaminant
present, and formation characteristics, treatment times can be significantly
reduced compared to those of ambient temperature treatment systems.
Treatment times ranging from hours to days have been reported (La Mori
1989; Udell and Stewart 1989).
In impermeable formations, the steam flows may be too small to allow
practicable treatment (Ghassemi 1988). Success of in situ steam stripping
operations can also be limited where the formation or the contamination is
heterogeneous. Interbedding of permeable and impermeable layers in the
subsurface can lead to steam flow around the impermeable layers, rather
than through them. Transport out of the impermeable regions is limited by
the diffusion rate, which can be slow. Cyclic steam injection, however, can
improve steam use efficiency in these cases (Udell and Stewart 1989;
Briggs 1989). There is also concern about practical control of the steam
condensate front in the absence of any lower confining unit; this front may
contain, or displace, significant amounts of contaminant. If not properly
controlled, the contaminant may be driven deeper, or become "smeared"
throughout the subsurface.
Steam injection requires additional capital equipment and energy. This
increases the cost and complexity over that of ambient temperature
remediation systems. The high temperatures will also adversely affect bio-
logical degradation processes.
An inherent limitation of steam injection systems lies in their inability to
heat formations to temperatures significantly greater than 100°C (212°F), as
steam temperatures are limited by injection and subsurface pressures.
Steam stripping may not be efficient for removal of higher boiling point
compounds, such as some aliphatic and aromatic fractions of jet fuels and
gasoline, chlorobenzene, trichloroethylene, dichloroethane, and
tetrachloroethane.
3.141
-------
Process Identification and Description
3.7.2 Radio Frequency Heating
Radio frequency (RF) heating has the potential of increasing subsurface
temperatures well above the 100°C (212°F) boiling point of water, allowing
more rapid removal of higher boiling point compounds than does steam
injection. Energy is delivered to the subsurface via radio-frequency waves,
which excite molecular motion and induce heating (much in the same way a
microwave oven heats food).
In addition to standard soil vapor extraction equipment, the system uti-
lizes electrodes or antennae connected to a radio-frequency generator.
These transmit radio-frequency waves into the formation where some of the
energy is absorbed for heating. The exciter array electrodes can be inserted
into holes drilled into the formation or positioned on the soil surface. A
modified radio transmitter serves as the power source. The broadcast fre-
quency is in the industrial, scientific, and medical band. Operating fre-
quency is chosen based on the dielectric properties of the soil and the areal
extent of the contamination.
RF heating occurs through ohmic and dielectric mechanisms. Ohmic
heating results from a voltage drop pushing electrons up into the conduction
band and moving them through the soil mass, producing resistance heating.
For the most efficient and uniform heating with RF power, ohmic heat input
should be kept to a minimum by limiting the induced voltage drop in the
soil mass.
Dielectric heating results from distortion of the atomic or molecular
structure in response to an applied electric field. Typically, the dipole mo-
ments of the molecules in a polar substance are randomly oriented. The
application of an external electric field will cause the dipole moments to
begin to align.
The dielectric constant is a critical parameter in RF heating system de-
sign. It is dependent on the moisture content of the soil. As formation heat-
ing proceeds, water is driven off, resulting in a drop in the loss tangent.
Maintaining efficient coupling to the RF field with the formation when the
moisture is removed is a major challenge in RF heating system design. Sys-
tems typically maintain coupling by changing the broadcast frequency, the
electrical properties of the network used to match the exciter array to the
soil mass, or both. Johnson, Otermat, and Chou (1991) describe a means
for overcoming some of these problems by injecting a warm, moist vapor
3.142
-------
Chapter 3
stream across the target depth to maintain moisture (and hence, energy ab-
sorbance) in the target zone.
3.7.2.1 Status of RF Heating Technology
To the authors' knowledge, RF heating/vapor extraction has had limited
field application. A number of laboratory-scale treatability tests have been
conducted. The authors understand that the US Air Force Armstrong Labo-
ratory Environmental Division, Tyndall AFB, Florida, plans a full-scale
demonstration test in the future. (There have been field demonstrations of
an RF process for enhanced recovery of oil from oil shale and tar sand de-
posits.)
A limited field-scale demonstration test was conducted at the Volk Air
National Guard Base. The contaminated site had served as a fire-fighting
training area for more than 25 years. Waste oils, fuels, and other hydrocar-
bons were burned in a flare pit to simulate aircraft fires. An estimated
189,000 L (50,000 gal) of hydrocarbon materials had migrated into the soil
around the pit. During the test at this site, RF heating raised the soil tem-
perature to 100°C (212°F) in 2 days and to 150°C (302°F) in 8 days. RF
heating was applied for 12.5 days.
3.7.2.2 Advantages and Disadvantages of RF Heating
RF heating has potential as a soil heating method, as the user can poten-
tially achieve more rapid heating rates and more uniform heating than with
any competing technology. In addition, no fluid injection is required for
heat delivery, and the entire system can be operated under vacuum condi-
tions to insure containment.
Capital equipment costs (mainly the RF generator) currently appear to be
limiting the development and application of this technology. For some
large sites, it is possible that this technology may be cost effective; how-
ever, unless pilot-scale tests can be conducted with costs similar to those of
competing technologies, it should not be considered seriously in remedial
strategy planning. In addition, it must be noted that operation of this pro-
cess requires a higher level of sophistication than most remediation pro-
cesses.
As with other heating processes, the high temperatures will inhibit bio-
logical activity in the soil. High temperatures may also have an effect on
3.143
-------
Process Identification and Description
the soil structure and induce cracking, which may, however, present some
advantages in low permeability formations.
3.7.3 Joule Resistance Heating
Another potentially efficient soil heating method uses the soil as the
conduction path for electrical current. Energy lost due to electrical resis-
tance heats the formation. A major challenge in using this technology lies
in the inherent drop in electrical conduction with decreased soil moisture
content. Because the current density is highest near the electrodes, soil in
this area dries faster. As a result, the electrical resistance increases, and the
heating rate decreases and becomes highly uneven (if voltage is held con-
stant).
Permafrost has been melted prior to construction using in-ground resis-
tance heaters ("conductive heating"; see Subsection 3.6.4). The efficiency
of the method, however, declines as the volume melted increases, because
of the increased path length for heat conduction. Maksimenko (1984) pro-
posed a two-stage process using resistance heating to start the melting pro-
cess and then applying the resistance heaters as electrodes in order to
continue heating by electrical conduction through the melted permafrost.
One process specifically designed to remediate contaminated sites uses
an array of conductors formed by inserting metal pipes into the contami-
nated soil (Heath 1990). An electrical current is passed between electrodes
to heat the soil enough to remove most of the soil moisture and any volatile
contaminants. Water vapor and volatile organics are collected by conven-
tional vacuum extraction techniques. Heating and moisture removal also
preconditions the soil by making it permeable to gas flow. The soil can be
vented through the pipe-electrodes themselves. When the bulk of the soil
moisture is removed, the voltage is increased to stimulate in-place oxidation
of any nonvolatile organics. The process is reported to cleave the nonvola-
tile organics into smaller, lighter components, which then volatilize. The
oxidation products are collected in the vacuum extraction system.
In situ vitrification (Koegler 1989; Buelt and Westsik 1987) also em-
ploys electrical conduction in soil but operates at temperatures high enough
to melt the soil. To the authors' knowledge, this process has not been used
with soil vacuum extraction, although a vapor shroud is employed at ground
surface to collect any vapors driven upward by the heating.
3.144
-------
Chapter 3
3.7.3.1 Status of the Joule Resistance Heating Technology
A pilot test of in situ electrical heating of soil has been conducted in the
field for permafrost melting. In-ground electric heaters melted permafrost
to a radius of 2.3 m (7.5 ft) in 450 hours of heating with a specific power
input of 27.4 kWh/m3 (0.8 kWh/ft3). Bench-scale studies of hazardous ma-
terial destruction by electrical resistance heating are reported to remove
95% of 2-chlorophenol and 25 to 99% of a semivolatile/nonvolatile organic
mixture. Sand, silty loam, and bentonite clay were used in the tests. Re-
portedly, the organics were partially volatilized and partially decomposed,
producing peroxide in the presence of moisture and O~ by corona discharge
in air.
To the authors' knowledge, joule resistance heating has not been used at
field scale in conjunction with vacuum extraction. It is the authors' under-
standing that such field tests are being contemplated by Battelle's Pacific
Northwest Laboratories, Richland, Washington.
3.7.3.2 Advantages and Disadvantages of Joule Resistance
Heating
Joule resistance heating is an emerging and, as yet, unproven technology.
As such, little information is available on its feasibility. Equipment costs
appear high, at least of the same magnitude as for RF heating. The elec-
trodes are thin rods or tubes that can be installed with little or no distur-
bance of the soil; thus, they are much easier to install than steam or hot air
injection wells. Because the power source is electrical current, the equip-
ment required for the supply of energy is typically less complicated than
that used for steam generators, air heaters, or RF transmitters.
The key challenge in designing in situ joule resistance soil heating sys-
tems is to maintain uniform conduction through the soil mass during heat-
ing. Heating drives off soil moisture, thereby reducing conductivity, and
causing uneven heating and inefficient use of electrical energy. This prob-
lem can be somewhat mitigated through electrode design and placement.
3.7.4 Conductive Heating
In conductive heating, a heat source is placed on the soil surface or in-
serted into the formation; the temperature of the heater is raised, and then
the soil is heated by conduction. Although a number of heater designs are
3.145
-------
Process Identification and Description
possible, the most practicable is likely to be an off-the-shelf electrical resis-
tance heater. Down-hole heaters are known to have been applied for the
purpose of enhancing oil recovery, but their use for remediation is not
widely reported. Johnson et al. (1992) describe a conductive heating soil
vapor extraction process for the remediation of surficially-contaminated
soils, which achieved a >99% reduction in chlorinated hydrocarbon concen-
trations in surficial (0 to 0.6 m (0 to 2 ft) BGS) soils (Johnson, unpublished
data).
Conductive heating presents an advantage in that it is probably the least
sophisticated heating technology discussed here. In addition, capital costs
may be low in comparison with other methods described above, as electrical
resistance heaters can be purchased off-the-shelf. A disadvantage is that
heat conduction through soils is inherently very slow and large temperature
gradients must be maintained to insure acceptable heating rates.
3.7.5 Modifying Soil Surface
Soil temperature can be affected by regulating the incoming and outgo-
ing radiation, or by changing the thermal properties of the soil. Only rela-
tively small temperature increases can be effected, but no auxiliary energy
input is needed. For this method to be effective, surface modifications must
increase the mean annual surface temperature. The problem lies in maxi-
mizing the flow of incoming solar radiation, minimizing the reflection and
radiation of energy from the soil surface, and retarding or preventing heat
losses to the atmosphere through conduction, convection, and evaporation.
The goals are to increase the heat absorption of the soil during warm or
sunny periods and to reduce heat loss during cold or dark conditions.
Simple approaches can be taken such as removing vegetation and/or tilling
the surface, mulching with organic material or plastic sheeting, or irrigating
in order to increase the heat capacity and thermal conductivity of the sur-
face soil.
The simplest approach is to remove vegetation. A heavy mat of vegeta-
tion will absorb the solar radiation before it reaches the ground and thus
reduce soil heating. In periods when the soil is emitting heat, however, the
vegetation helps to insulate the soil and reduce heat loss. The net effect of
removing vegetation is typically an increased average flow of heat into the
soil; however, the competing effects result in reduced overall efficiency.
3.146
-------
Chapter 3
Greater efficiency can be obtained by using clear polyethylene sheets.
The ground is covered with sheeting to increase radiation collection during
the day and reduce convective and conductive heat loss at night. The rela-
tively short-wavelength solar radiation readily passes through the clear
plastic, but the longer wavelength radiation emitted by the soil does not.
The most significant problems reported in the use of polyethylene are deg-
radation of the material from exposure to cold and wind damage.
The soil can be irrigated to increase moisture content, thereby increasing
the soil's thermal conductivity. The effect of the increase in thermal con-
ductivity is somewhat offset by an increase in specific heat due to the water.
Saturating soil with water, however, typically has the net effect of increas-
ing the heat transfer rate. Irrigation has the least effect in peat and the
greatest effect in sandy soils.
3.7.5.1 Status of Soil Surface Modification Technology
The technology of soil surface modification has undergone pilot-scale
field testing in cold climates. Various surface treatments have been ex-
plored for applications in improving crop growth and melting permafrost in
cold regions in tests unrelated to contaminated soil remediation. A litera-
ture search produced no reports indicating that surface modification to pro-
duce soil heating has been demonstrated for remediation of contaminated
sites.
Clear polyethylene has been applied to increase soil temperature for
improved seed germination (Dinkel 1966) at a site near Palmer, Alaska.
This work demonstrated an increase of 16.7°C (62°F) at 2.5 cm (1 in.) depth
in plots covered with clear polyethylene.
In a study of permafrost thawing (Nicholson 1978), soil temperatures
were modified by removing vegetation to increase energy input in the sum-
mer and by adding snow fences to increase snow cover and improve insula-
tion in the winter. Temperatures in the treated plot were significantly
higher than in the control plot. Over the 5-year test period, the temperature
at a depth of 10 m (33 ft) in the test plot rose from about -2.2°C (36°F) to
about -0.4°C (32°F). Of particular note, the large drop in winter tempera-
ture was avoided. The winter temperature in the control plot at 0.25 m
(0.82 ft) dropped to about -17°C (2°F), while the winter temperature in the
test plot at 0.25 m (0.82 ft) only dropped to about -2°C (62°F).
3.147
-------
Process Identification and Description
Small plots were also used to test the effects of a variety of surface treat-
ments. Soil temperature increases reported for the surface treatments were
about:
• 2°C (62°F) with vegetation stripped;
• 4.5°C (40°F) clear polyethylene applied over natural vegetation;
and
• 5.5°C (42°F) clear polyethylene applied over stripped soil.
Problems with polyethylene sheets — becoming brittle in winter cold
and being easily destroyed by the wind — were noted. Dusting with carbon
black and using a black dye reportedly delayed degradation.
Field tests of various surface treatments were also conducted at six test
plots near Fairbanks, Alaska, from 1980 through 1983 (Esch 1984). Perma-
frost at the test site exceeded 30 m (98 ft) in depth, with an average tem-
perature of -1°C (34°F) at a depth of 10 m (33 ft). The tests showed that
simply stripping away the surface vegetation caused a drastic change in the
energy balance of a previously undisturbed surface. Removing vegetation
resulted in an increase in thaw depth of 2.0 m (6.6 ft) in 4 years. Light-
colored gravel alone did not increase the thaw depth, but did improve con-
trol of vegetation regrowth. Gravel darkened with asphalt provided a thaw
depth of 2.5 m (8.2 ft) in 4 years. Clear polyethylene film, applied just a
few centimeters above the soil surface, increased the 4-year thaw depths by
15% when applied to a stripped soil plot, and by 17% when applied over an
asphalt-coated gravel pad. Covering the asphalt-coated gravel pad with a
clear polyethylene film to create a greenhouse effect resulted in an increase
to 3.0 m (9.8 ft) thaw depth in 4 years.
3.7.5.2 Advantages and Disadvantages of Soil Surface Modifi-
cation
Soil surface modification is a very simple technology that has the poten-
tial of supplying heat to depths of up to several meters at very low cost. It
provides in situ bulk heating without requiring major external energy input
or expensive capital equipment.
The low-energy density that is available limits the temperatures that can
be achieved. Soil heating occurs through conduction from the surface
downward. Thermal diffusivity of the soil is sufficiently low that tempera-
ture changes occur in a period of days or weeks. The method is most appli-
3.148
-------
Chapter 3
cable in improving biotreatment rates, particularly in cold climates. It is
unlikely that application of this method would produce temperatures result-
ing in significant increases in the vaporization rate of water or stripping of
contaminants.
3.7.6 Hot Air or Hot Gas Injection
Hot air injection can raise soil temperature; however, because of the very
low heat capacity of gases, it has limited application. For example, air in-
jection at a temperature of 100°C (212°F) into a sandy soil with a heat ca-
pacity of 0.5 cal/cm3-°C (590 Btu/fiVF) at a typical soil venting air-flow
rate of 100 m3/hr (50 ftVmin) with a radius of influence of 8 m (26 ft), a 4-
m-thick (13 ft-thick) soil column at 10°C (50°F), would warm at the average
rate of approximately 0.15°C/day (32.3°F/day). As the soil warmed, the rate
of warming would decrease. Generally, at the air-flow rates used in soil
venting, air must be warmed to several hundred degrees centigrade in order
to add sufficient heat to warm soils at a rate sufficiently high to be usable.
Hot gas injection appears particularly appealing when thermal treatment
of offgas air is required. Hot offgas resulting from incineration represents
wasted heat that potentially may be recovered by injection into the soil for
use in removing contaminants. In a case reported by Oak Ridge National
Laboratory (1990), the incinerator was distant from the gas injection wells,
resulting in long piping runs and high-heat losses. If biodegradation is an
important part of the treatment regime, the reduced oxygen content of the
incinerator offgas could be a concern. Oxygen effects can be avoided by
using an air-to-air heat exchanger. The heat exchanger transfers heat from
the offgas to the, venting air while keeping the gases apart. The additional
equipment required increases the system's complexity and cost and inevita-
bly results in additional heat losses. Another significant problem may result
when hot gas is used in attempting to warm soils for bioremediation; if gas
temperatures rise more than 20 to 30°C (68 to 86°F) above ambient, micro-
organisms near the injection point may be inhibited or killed.
High temperatures help offset the low-heat capacity of the air, but create
other problems. To carry significant amounts of heat in air, temperatures
must be above 300°C (572°F). With these temperatures, substantial insula-
tion may be required to control heat losses in the piping to the inlet well.
Also, the high temperatures needed will damage the materials used in typi-
cal wells. In order to withstand elevated air temperatures and temperature
3.149
-------
Process Identification and Description
cycling, the wells require more expensive materials and more complex de-
signs. Although not reported in the literature, the effects of high-air inlet
temperatures and temperature cycling on conventional injection wells could
pose problems. Steel pipe could be used in place of the more typical plastic
piping. Even if steel injection wells are used, however, there would be
another drawback. The bentonite, which seals most wells, would not main-
tain integrity at temperatures above 100°C (212°F). Cycling, that is, heating
followed by cooling in a well, will result in substantial expansion and con-
traction of the well piping. A conventional concrete or cement seal could
lose its integrity under these conditions. For substantial, long-term use of
high-temperature air injection, new and much more costly injection well
designs would likely be required.
Another heat source that has been used to generate hot air is solar heat-
ing. Under this concept, air is drawn through a flat-plate solar collector by
a blower, which then discharges to an air injection well. The increase in air
temperature imposed by the collector is limited. The system is reportedly
used to increase the temperature to improve biodegradation rates (Billings
1991).
Hot air injection has been used in conjunction with steam heating to
assure that the stripped organics remain in the gas stream. Two of the sys-
tems reviewed applied this technique — the Toxic Treatments - NovaTerra
(USA), Inc., Detoxifier (La Mori 1989) and the ENSR groundwater cleanup
system (Smith, Aiken, and Tursman 1990). Air injection in these systems,
however, follows steam injection and its purpose is to maintain organics in
the vapor state; hot air is not injected for bulk soil heating.
3.7.7 Fiber Optic Heating
The sun is a low-intensity source of energy for soil heating. Under ideal
conditions the power density available from the sun is about 1 kWh/m2 of
collector. Parabolic collectors coupled to optical fibers are being tested for
using solar energy to heat soil in situ (Houthoofd, McCready, and Roulier
1991). Collectors have been used in the past for direct heating of air and
water. Significant heat losses during transmission may make this process
undesirable for in situ heating of soil.
Compound parabolic concentrators, which have the advantage of collect-
ing scattered sky light as well as direct sunlight, are modified for coupling
to a cable of optical fibers. The optical fiber has the potential of transfer-
3.150
-------
Chapter 3
ring the solar radiation with high efficiency over long distances. Since the
solar energy is conducted as light rather than as a hot fluid, thermal losses
along the transmission line are greatly reduced. The collector surfaces
would also remain relatively cool and radiate little energy because conver-
sion to heat occurs at the radiator or at the end of the optic cable, rather than
at the collector.
3.7.8 Warm Water Injection
Water has a much higher heat capacity than air. In soils sufficiently
permeable to make water injection practical, warm water injection may
represent a reasonable approach to increasing temperatures in order to in-
crease biodegradation rates. At a site at Eielson Air Force Base, Alaska,
encompassing an area of 230 m2 (2,500 ft2), water at a temperature of ap-
proximately 30°C (86°F) was injected at a flow rate of approximately 6 L/
min (1.6 gal/min). Through the winter, soil temperature was maintained
near 10°C (50°F) in the warmed area, and dropped below 0°C (32°F) in the
unheated areas (Leeson et al. 1993).
The primary drawback of warm water injection lies in obtaining and
delivering an adequate water supply. If permeability is low, the infiltration
rates may be limited. In addition, water can be added only at temperatures
below =100°C (212°F), limiting both heat input and the upper temperatures
that can be obtained. Accordingly, warm water injection is probably not a
practical approach to heating in order to improve volatilization rates for
high-boiling point compounds.
3.8 Summary of Good Practices
Although soil vapor extraction is relatively well-developed, there are
opportunities for significantly improving system performance and reducing
costs. This is true because the state of the practice is largely empirically
based and many common practices and beliefs have no apparent technical
basis.
This section presents opportunities for improving soil vapor extraction
system performance, but they are applicable to all vapor extraction-based
processes. Similarly, the flow diagram in figure 3.53 (on page 3.153), the
3.151
-------
Process Identification and Description
current state of the practice of key elements which are addressed below, is
generally applicable to all such processes, provided minor modifications are
made.
3.8.1 Site Characterization
Site characterization is performed in order to delineate the extent of con-
tamination and identify potential migration pathways. In a basic assess-
ment, soil cores are collected, groundwater monitoring wells are installed,
and samples are taken. The soil and groundwater samples are analyzed in
order to identify specific contaminants of concern and measure their con-
centrations. Sometimes, physical and chemical properties of the soils are
also measured.
Too often site assessments are conducted without first determining data
needs for the decisions to be made. Total project costs could often be re-
duced with a little preplanning. At many sites (e.g., sites of underground
storage tank releases), the investigator has an idea of the contaminant
source, contaminant type, and subsurface stratigraphy. At these sites, the
following measures may lead to cost savings:
• At a minimum, sufficient data should be collected to enable
drawing a subsurface cross-section map that identifies soil struc-
ture, contaminant distribution (levels and location), and the loca-
tion of other significant features (groundwater table, tanks,
subsurface conduits, etc.);
• If preliminary data indicate that vapor extraction may be appli-
cable, soil borings drilled during the site assessment phase
should be used for the installation of pilot-test vapor extraction
wells and vadose zone monitoring point installations. Soil-gas
samples can then be collected shortly after the basic assessment;
• Contaminants, especially complex mixtures, are more easily
characterized by the distribution of compounds (e.g., boiling-
point distribution) rather than by specific compound. Typically,
as a result of regulatory requirements, soil samples are analyzed
for a limited number of compounds of concern (e.g., benzene,
toluene, xylenes) and some measure of total contaminant levels
(e.g., total petroleum hydrocarbons). These data are often of
3.152
-------
Chapter 3
Figure 3.53
In Situ Soil Venting System Design Process
Process
| Leak or Spill Discovered |
Emergency Response & Abatement
Output
Background Review
Site Characteristics'.
Subsurface Characteristics
• soil stratigraphy
• chacteristics of distinct soil layers
(permeability estimates, soil types)
• depth to groundwater
• groundwater gradient
• seasonal water table fluctuations
• aquifer permeability (estimate)
• subsurface & above-ground temperature
Contaminant Delineation
• extent of free-phase hydrocarbon
• distribution of contaminant in vadose zone
• distribution of contaminant in saturated zone
• extent of soluble contaminant plume
• compsoition of contaminant
• soil vapor concentrations (optional)
Define Clean-up Objectives
Screen Treatment Alternatives
Air Permeability Test U
Groundwater Pump Test
I System Design]
T ,
System Operation & Monitoring
• removal rate estimates
1 vapor flowrate estimates
1 final residual levels & composition
1 air permeability of distinct soil layers
1 radius of influence of vapor wells
1 initial vapor concentrations
1 drawdown determination
1 radius of influence
• pumping rate determinition
1 number of vapor extraction wells
1 vapor well construction
• vapor well spacing
• instrumentation
• vapor treatment system
»flowrate (vacuuum) specifications
• groundwater pumping system specifics
1 venting recovery rates
1 changes in vadose zone contamination
(target levels based on
exposure assessment)
System Shut-Off "Clean" Site
3.153
-------
Process Identification and Description
limited value in evaluating the applicability of vapor extraction-
based processes and that of many other remedial processes;
• Field analytical methods should be used to supplement labora-
tory methods when possible. More data can be collected for the
same, or less, cost; and
• Soil property analyses should be minimized. For each distinct
soil layer, it is useful to have a measure of the moisture contenl,
organic carbon fraction, and permeability. Laboratory physical
property test results (i.e., permeability) should not be overem-
phasized because the tests are conducted on disturbed samples
and the results are of limited value.
3.8.2 Defining Remedial Objectives
It is not uncommon for vapor extraction-based systems to be designed
and installed without target cleanup goals being set. Consequently, closure
requirements are often negotiated after prolonged operation of the system.
In many cases, the additional operating time (and costs) yields no additional
benefits, because inherent limitations of system or site condition (e.g., try-
ing to remediate residuals trapped beneath the water table by soil venting
alone). Therefore, to assure the least cost:
• Remedial objectives (target cleanup levels and duration of
remediation) should be defined and constraints (costs, physical
boundaries, community imposed, etc.) should be identified be-
fore remedial technologies are selected.
• Remedial objectives should be set with a view to protecting
human health and the environment and be derived on a site-
specific basis in consideration of reasonable potential beneficial
uses.
3.8.3 Screening Treatment Alternatives
In the screening of treatment alternatives, all relevant factors, and not
just the technical aspects should be considered. Vapor extraction-based
technologies should be selected, of course, only if they are considered likely
to meet the remedial objectives, given site-specific constraints. It is impor-
tant to identify limitations and to assess the degree of uncertainty obtained
in applying vapor extraction-based technologies. In some cases, a cost-
3.154
-------
Chapter 3
benefit analysis is useful. As demonstrated in Section 4.3, this requires
tools to predict long-term system performance and costs. The availability
and development of such tools are discussed in Subsection 3.8.7.
3.8.4 Field Tests
Currently, a typical vapor extraction field test is of short duration (a few
hours). Competent practitioners measure flow rates, applied vacuums, and
vapor concentrations at each extraction well. In addition, the subsurface
pressure distribution away from the extraction well is also measured at a
few locations. Often, to avoid the expense of installing test wells and va-
dose monitoring installations (which could have been installed during the
site assessment as discussed above), existing groundwater monitoring wells
are used. Consequently, pilot-test observations (flow rates, vapor concen-
trations, etc.) are often not representative of full-scale system performance.
To generate more relevant data for full-scale system designs, it is important
to insure that guidelines similar to those given in Section 3.3 are followed
when conducting such field tests.
3.8.5 System Design
Designers need to recognize the inherent uncertainties associated with in
situ remedial systems. Therefore, designs should be flexible enough to al-
low future system expansions and a range of operating conditions. This
matter is discussed further in Section 3.8.7.
3.8.6 Operation and Monitoring
In practice, much of the monitoring of soil vapor extraction systems is
driven by regulatory requirements. Typically, vapor extraction flow rates,
system vacuums, and total contaminant vapor concentrations are measured
on a periodic basis. In addition, permit conditions often require that emis-
sions of specific compounds (e.g., benzene) be quantified and that vapor
treatment unit performance be documented. But, monitoring plans are sel-
dom designed with a view to optimizing system performance, the signifi-
cant cost of remediation notwithstanding.
Monitoring plans should provide for measurements that yield sufficient
information about system performance to enable process adjustments so
that remediation can be optimized. As explained in Section 3.5, a wide
3.155
-------
Process Identification and Description
range of parameters can be observed at soil venting sites, yet the data typi-
cally collected in present practice can not be used to distinguish among
several possible causes. For example, based on total hydrocarbon vapor
concentrations, system performance data from diffusion-limited sites often
resemble similar data from ideal sites; only if composition (e.g., boiling
point distribution) data are collected can the cause affecting system perfor-
mance be discerned.
3.8.7 Engineering Analysis
Engineering analyses are required to assess the feasibility of vapor ex-
traction-based processes, interpret field-test results, design systems, and
optimize system performance. To date, engineering analyses have usually
been limited to determining the "radius of influence" RT from pilot test data,
drawing circles of radius R, on a site plan map to determine well locations,
and computing total hydrocarbon removal rates from vapor concentration
and vapor flow rate data. The first two practices have no clear technical
basis, and the third yields an incomplete picture of system performance.
Yet, practitioners generally are not inclined to perform additional analyses.
This may be because of the perception that: (1) engineering analysis, or
"modeling," is a complex process that yields little return for the investment,
(2) vapor extraction is a "forgiving" process in that any system, however
poorly designed and operated, will eventually remediate a site if given
enough time (and money), (3) vapor extraction is not competitive enough to
bear the cost, or (4) site owners are not sufficiently informed of the need to
provide an incentive to develop and utilize more sophisticated analyses
procedures.
To understand the role of engineering analysis, consider the extreme case
where no analysis is performed and systems are installed based solely on
past experience. If the contaminant and site conditions are identical to an-
other site that has been remediated, then reasonable confidence can be
placed in expectations of system performance. If site conditions are similar,
but not identical, predictions of system performance are less certain. With a
new site that is dissimilar to any previous one, there is a high degree of
uncertainty as to whether vapor extraction can successfully remediate that
site. Rarely are identical sites encountered in practice; thus, there will al-
ways be a large degree of uncertainty when decisions are based on experi-
ence alone. Uncertainty about system performance means uncertainty about
costs. Vapor extraction-based technologies can be costly, therefore, there is
3.156
-------
Chapter 3
a clear economic incentive to reduce the uncertainty in expectations/predic-
tions of system performance.
Engineering analysis is the tool to help reduce uncertainty. It supple-
ments field experience enabling the extrapolation of observations at one site
in order to predict behavior at a dissimilar site, gain a better understanding
of the parameters that most influence system behavior, and predict changes
in system performance resulting from modifications of the system or
changes in operating conditions. There is a wide spectrum of possible engi-
neering analysis approaches ranging from the current approach generally
applied (little or none) to the complex modeling of vapor flow and contami-
nant transport in the subsurface, such as that reported by Benson (1992).
Because of costs associated with engineering analysis and the typical qual-
ity of site characterization data, it is unlikely that complex modeling will be
routinely conducted for the majority of vapor extraction sites; however,
there is an intermediate level of analysis that is justifiable by the potential
for cost savings. Some examples are the site-specific flow modeling by
Baehr and Hult (1991), screening/feasibility model development
(Massmann 1989; Johnson, Kemblowski, and Colthart 1988; Johnson et al.
1990, 1991; Baehr, Hoag, and Marley 1989), and the use of models for
"visualization" purposes (Wilson, Clarke, and Clarke 1988; Gannon et al.
1989; Mutch and Wilson 1990; Gomez-Lahoz, Rodriguez-Maroto, and
Wilson 1991).
Another factor that may be limiting practitioners' use of engineering
analysis is the lack of readily accessible computing tools. If programmed in
a user-friendly manner, their use can minimize the time and costs of analy-
ses. For soil venting applications, accessible predictive tools are presently
primarily limited to air-flow models, such as CSUGAS (Sabadell,
Eisenbeis, and Sumada 1988) and AIRFLOW (Joss and Baehr 1993a), and
the screening-level flow, partitioning, and transport models such as
Hyperventilate (Johnson and Stabenau 1991) (other models are currently
available, but this Monograph does not provide provide an exhaustive list-
ing or evaluation of each). For bioventing and air sparging, such tools are
essentially nonexistent. The use of engineering analysis, or modeling, will
likely increase in the future as additional computing tools become available.
3.157
-------
-------
Chapter 4
POTENTIAL APPLICATIONS
Given enough time, vapor extraction-based technologies will "work" at
any sites that yield air from the subsurface, because every contaminant
incorporated in soil has a nonzero vapor concentration in the pore vapor
space. Hence, as long as vapors are withdrawn from the subsurface, there is
removal of contaminant. Additional measures, such as the enhancement to
biodegradation resulting from increased oxygen supply rates, may act to
accelerate the treatment process to some degree. Yet, there are bound to be
constraints limiting the range of conditions under which vapor extraction-
based technologies will be practicable. An attempt to define this range of
conditions is made in this chapter.
4.1 Review of Reported Applications
The first method for determining likely conditions is based on practical
experience. Reports are gathered from pilot- and full-scale studies, and
then the authors attempt here to correlate a relationship between the degree
of "success" and geologic conditions, contaminant properties, and system
variables (equipment, flow rates, etc.). This is similar to the approach taken
by Hutzler, Murphy, and Gierke (1989) in defining the state of the practice
for vapor extraction. The writers compiled a number of reports of applica-
tions and summarized information obtained from each. In reviewing this
compilation and reports from other applications, it appears that vapor ex-
traction is most often applied at sites where hydrocarbon fuels (gasoline,
aviation fuels, etc.) and solvents have contaminated soils. Examples are
service stations, electronics manufacturing sites, military bases, and dry
cleaners. From the limited geological information documented, it appears
that vapor extraction systems are often installed at sites with very perme-
able (sandy) soils. Beyond these observations, very little can be concluded
from the documented applications. Most case histories fail to report final
4.1
-------
Potential Applications
soil and groundwater contaminant levels, costs, and original target cleanup
levels; thus, it is difficult to judge whether or not the application was "suc-
cessful." It should be noted that identifying unfavorable conditions for
vapor extraction technologies based on these case histories is difficult, as
they are biased strongly toward those sites where something significant
occurred (rapid cleanup, high removal rates, etc.). To date, the engineered
use of bioventing and air sparging has not been as extensive as that of vapor
extraction; the few reported bioventing studies have been at aviation fuel
(Hinchee et al. 1989; Miller et al. 1990) and "heavier" fuel releases
(Downey and Guest 1991). Air sparging has been applied to remediate
chlorinated solvents (Middleton and Miller 1990; Bohler et al. 1990; Brown
and Fraxedas 1991; Brown, Herman, and Henry 1991; Kaback et al. 1991),
gasoline (Ardito and Billings 1990; Marley 1991; Kresge and Dacey 1991),
and aviation gas (Griffin, Armstrong, and Douglas 1991).
4.2 Review of Recommendations
In the second method for determining conditions under which vapor
extraction-based technologies will likely be applicable, the authors refer to
the literature. Some writers define the range of applicability in terms of
chemical properties; compounds with vapor pressures >0.5 mm Hg and
dimensionless Henry's Law constants exceeding 0.01 are often considered
candidates for vapor extraction (Hutzler, Murphy, and Gierke 1989).
Brown, Herman, and Henry (1991) suggest that compounds with dimen-
sionless Henry's Law constants >103 are "strippable" and amenable to air
sparging. In order to incorporate soil permeability as an additional crite-
rion, others have resorted to the creative use of charts, tables, and
nomographs (US EPA 199la and b). Figure 4.1 (on page 4.3) is an example
of one of the more informative representations of conventional thinking; the
use of vapor extraction is recommended at "permeable" (sandy) sites con-
taminated by "volatile" compounds (gasoline, solvents, etc.), and its use is
discouraged for less volatile compounds and lower permeability soils
(clays). This guidance is clearly consistent with the case histories discussed
in Section 4.1, above, as well as with basic vapor flow and contaminant
partitioning considerations. Based on the limited data to date, it appears
that this guidance is also roughly applicable to air sparging applications.
For reference, soils with permeabilities to air flow exceeding 10'8 cm2 (10~3
cm/s hydraulic conductivity) are commonly regarded as "permeable."
4.2
-------
Chapter 4
Figure 4.1
Soil Vapor Extraction Applicability Nomograph
Vapor
Pressure
Soil Air
Permeability
Time
Since
Release
Naphthalene
Aldicarb
Low
(Clay)
Weeks
Months
Years
Weeks
Months
Years
Weeks
Months
Years
Match Point
Vapor pressure in mm Hg
Source: US EPA 1991a
4.3 Quantifying Applicability
In summary, based on case studies and published guidance, vapor extrac-
tion-based technologies are likely to be most successful at sites where vola-
tile compounds have impacted permeable soils. It is important to note,
however, that this guidance is only intended to roughly define the range of
applicability, and not to limit it. Obviously, other equally significant fac-
tors such as contaminant distribution, remedial objectives, economics,
equipment availability, community pressures, and physical boundaries are
neglected in this initial evaluation of applicability. Having recognized this,
other investigators are focusing on the development of methods incorporat-
4.3
-------
Potential Applications
ing all relevant parameters in order to determine applicability on a site-by-
site basis. These efforts typically involve a quantitative prediction of long-
term system performance based on site characteristics, contaminant
properties, and pilot-scale data. For example, Johnson et al. (1990) and
Johnson and Stabenau (1991) compare remedial objectives (cleanup levels,
time to achieve cleanup) with screening level predictions of system perfor-
mance in order to determine vapor extraction applicability. This approach
was illustrated in Subsection 3.4.7. Accepting the uncertainties associated
with screening-level predictions, the major deficiency of the approach out-
lined by Johnson et al. (1991) is the failure to explicitly incorporate eco-
nomic constraints. A more general method for quantitatively defining the
range of potential applications for vapor extraction-based technologies is
described below and then illustrated through a simplistic example.
Figure 4.2 (on page 4.5) outlines one process for determining the appli-
cability of vapor extraction-based technologies. A vapor extraction-based
technology is potentially applicable, of course, if remedial goals can be met
given constraints imposed at the site. The remedial goals include target
cleanup levels and an acceptable time frame for remediation and are based
on considerations of risk-reduction and preservation of reasonable potential
beneficial uses. Site-specific data, including geological conditions contami-
nant characteristics are also required and pilot test data are considered, if
available. Typical constraints include total project cost, permitting require-
ments, physical boundaries, community-imposed limitations, and equip-
ment availability.
Given the necessary inputs, the objective is to determine whether there
are technically feasible process configurations and operating conditions that
will satisfy both the remedial goals and imposed constraints. Therefore, a
tool is needed to accurately predict the performance and costs of a given
system. Unfortunately, under the current state-of-the-practice, such predic-
tive tools are not accessible to most practitioners (see Section 3.8.7 for dis-
cussion). Therefore, a screening level analysis is presented here to illustrate
the benefits of the general approach outlined in figure 4.2 (on page 4.5). In
this example, the authors evaluate the applicability of vapor extraction for
the remediation of an operating service station and, for this purpose, col-
lected the following data and information:
Site characterization data:
m Based on inventory records 22,700 L (6,000 gal) of regular
gasoline have leaked from an underground storage tank;
4.4
-------
Chapter 4
Figure 4.2
Applicability Decision Process
Define Objectives:
• risk reduction
• preservation of reasonable
potential beneficial uses
Assess Site-Specific
Conditions:
• soil stratigraphy
• soil properties (permeability,
etc)
• depth to groundwater
• groundwater gradient
• seasonal water table
fluctuations
• aquifer characteristics (yield,
etc)
• distribution of contaminant
• extent of soluble contaminant
plume
• composition of contaminant
• contaminant properties
• pilot test data
V J
-
[Determine Remedial Goals: |
• cleanup targets
• time frame for cleanup 1
1
Is this Technology Appropriate at
My Site?
•What range of process conditions
(equipment & operational
procedures) are necessary to meet
the remedial goals9
•Are any of those conditions
consistent with the constraints9
1
r >i
• possible system designs
• estimated costs
• remediation potential
V J
*-
( ~\
Identify Constraints:
• costs (capital & operation)
• physical boundaries
• permitting
• political actions
• equipment availability
V J
m Free-product skimmer pumps have recovered 3,785 L (1,000
gal) of gasoline floating on the water table;
• Soils between the release point and groundwater are composed
of fine and silty sands; and
• It is anticipated that a groundwater pumping system will have to
be installed along with any vapor extraction system in order to
counteract the upwelling induced by the vapor extraction wells.
Remedial goals and economic constraints:
• Achieve a 90% reduction in total gasoline levels and reduce
benzene levels in soils so that groundwater cleanup standards
(based on preservation of reasonable potential beneficial uses)
are met;
• Effect remediation within a 20 year period; and
4.5
-------
Potential Applications
• Projected cost of vapor extraction must be less than $500,000,
which is the total cost estimated for an alternate excavation and
disposal option.
Design considerations:
m Based on screening-level analyses, it is estimated that in order to
achieve the desired reduction in gasoline and benzene levels, a
minimum of 100 L-atr/g-gasoline must flow through the zone of
contamination. Given that a significant fraction of the residual
is located near the water table, an efficiency factor of 20% is
assumed. Thus, the screening-level analysis is based on a vapor
requirement of 100/0.2 = 500 L-air/g-gasoline (see Johnson et al.
1990, Johnson and Stabenau 1991 for more details on the
screening analysis);
• A pilot test was conducted using an existing groundwater moni-
toring well and it is estimated that the maximum practicable
vapor-flow rate from vapor extraction wells is 9.5 L/sec (20
standard ft3/min); and
• Local air emissions requirements mandate a 90% reduction in
total hydrocarbon concentration of extracted vapors prior to
discharge.
Economic analysis (all costs given in current $):
• There will be a minimum installation cost of $20,000;
• Cost for vacuum pump/blower is $5,000 per 47 L/sec (100 stan-
dard ftVmin) capacity;
• Installation cost for vapor extraction wells is $5,000 per well;
• Cost for vapor treatment is $100,000 per 230 L/sec (500 stan-
dard ftVmin) capacity;
• Operation & maintenance costs are estimated to be $15,000 per
year; and
• All costs quoted above include piping and miscellaneous ex-
penses.
Based on the above information, the next step in the applicability analy-
sis is the calculation of the number of wells, Nwells, required based on the
500 L-air/g-gasoline vapor requirement. If:
4.6
-------
Chapter 4
MR = mass of residual gasoline in soil = 1.5 x 107 g = 5,000 gal
Qwell = flow rate to a single well = 9.5 L/sec = 20 standard ftVmin
T = time period for remediation; 0 < T < 20 years
a = vapor volume requirement = 500 L-air/g-gasoline
then the required number of wells is related to these other parameters by:
N =^- [41]
Dwells ~. L^-1J
xfwell ~
where the value Nwe|ls is rounded up to the next largest integer (i.e., Nwe]ls =
2.3 is rounded up to Nwells = 3). The estimated total cost of remediation by
vapor extraction is then:
Cost = $20,000 (minimum installation cost)
+ $5,000 Nwc|ls (extraction well installation)
+ $5,000 Nb|owers (blower/vacuum pump cost)
+ $ 100,000 NVT (vapor treatment unit cost)
+ $15,000 T (operation & maintenance costs) [4.2]
where the number of blowers, Nblowers, and the number of vapor treatment
units NVT are based on the total system flow rate and their individual capaci-
ties (i.e., 47 L/sec (100 standard ftVmin) for blowers, 236 L/sec (500 stan-
dard ftVmin) for vapor treatment units).
Figure 4.3 (on page 4.8) displays results of this screening analysis,
equipment required, and associated costs with different remediation periods.
For short remediation times (<2 years), the total cost is dominated by capi-
tal equipment costs associated with the large number of wells, blowers, and
vapor treatment units required. For longer remediation times, capital equip-
ment costs are relatively constant, and the increase in total cost with in-
creased remediation time is due to increased total operation and
maintenance (O&M) costs. It should be noted that the shape of the total
cost vs. remediation time curve shown in figure 4.3 is similar to curves that
would be generated from more detailed analyses. There is typically an
optimal design remediation time corresponding to the minimum total cost.
In figure 4.3, the minimum total remediation cost is =$200,000, which cor-
responds to systems designed for 2 to 4 year remediation durations. In this
example, therefore, vapor extraction is judged to be applicable because the
estimated total cost associated with it is less than the cost of the excavation
and disposal option.
4.7
-------
Potential Applications
Figure 4.3
Results of Applicability Screening Analysis
Equipment
1000000
800000-
600000
Number of Wfells
Number of Blowers
Number of \&por Treatment
Units
5 10 15
Remediation Time [years]
Cost
5 10 15
Remediation Time [years]
20
However simplistic, this example illustrates the importance of incorpo-
rating all relevant factors into the applicability screening analysis. Had the
desired time frame for remediation been less than a year or had there been
options having total estimated costs significantly less than $200,000, vapor
extraction would have been judged not to be applicable, despite the fact that
the contaminant is "volatile" and subsurface soils are "permeable." The
results of this analysis also hint at some of the benefits of air sparging and
bioventing. Air sparging is expected to lower remediation costs where
residuals trapped within or beneath the groundwater capillary fringe zone
contribute to unacceptable groundwater concentrations; capital equipment
and operational cost savings are realized because groundwater pumping
systems are not required to lower the water table and expose soils for vapor
extraction and long-term operating expenses are reduced because the time
frame for remediation is shortened. Engineered bioventing systems offer
capital equipment cost reductions by eliminating vapor treatment equip-
ment, the need for an explosion-proof blower, and reducing the blower size.
In a bioventing operation, lower air-flow rates frequently reduce operation,
monitoring, and maintenance costs on a daily basis, but they can increase
the time required for cleanup. The net impact on cost depends on site-spe-
cific factors.
4.8
-------
Chapter 5
PROCESS EVALUATION
Throughout this monograph the authors stress that the practicable degree
of remediation through vapor extraction-based processes is limited by site
characteristics and contaminant properties. The actual degree of
remediation achieved, of course, also depends on specific skill and knowl-
edge of the practitioner, which is reflected in the system design and opera-
tion. Unfortunately, even very inefficient remedial processes may appear to
be highly successful in the eyes of less-informed practitioners.
At this point, it would be useful to present a tabulation of case study
results clearly showing the demonstrated capabilities of vapor extraction-
based processes. Review of available data, however, shows that most case
studies are merely expositions of anecdotal data and pseudo-science that
add very little to basic knowledge of vapor extraction-based processes.
Buscheck and Peargin (1991) surveyed 143 vapor extraction projects in an
attempt to correlate performance with relevant parameters and concluded
that the majority of sites were insufficiently monitored to provide useful
information. For this reason, the authors choose not to present such a tabu-
lation, as they feel the results can be very misleading. Instead, they incor-
porate results of studies throughout the text to illustrate key concepts.
Following are the two key questions that generally are not answered in
case studies:
• What triggered the decision to turn the system off?; and
• Were the cleanup goals achieved (what was the condition of the
site after treatment)?
In practice, it is not always apparent from process monitoring when a
vapor extraction-based system has achieved its cleanup goals. Some ven-
dors claim vacuum extraction cleanups at some sites in fewer than 90 days.
Dupont, Doucette, and Hinchee (1991) report completion of a combined
vacuum extraction/bioventing remediation of a jet fuel-contaminated site in
5.1
-------
Process Evaluation
about 2 years. Based on the authors' experience, most systems installed at
service stations operate for periods greater than one year and less than five
years; this is supported by the survey reported by Buscheck and Peargin
(1991).
Ultimately, soil sampling is often required to verify cleanup and in many
cases some residual hydrocarbons are detected. To avoid the cost of prema-
ture sampling, a pragmatic approach is to operate the system until both the
contaminant and the respiratory gas concentrations approach background
levels, no significant restart spike is observed, and oxygen utilization does
not exceed background. These conditions indicate that the vacuum extrac-
tion system has accomplished all it can, and that either the site is clean or
the residual contamination is not responding to vacuum extraction treat-
ment.
It is also possible to define an end point based on economic analysis. As
mentioned in Section 3.5, the cost-effectiveness of an extraction system can
be tracked over time, and the decision to cease operation can be pegged to a
given cost per unit contaminant removed.
5.2
-------
Chapter 6
LIMITATIONS
By examining the data from the available laboratory-, pilot-, and full-
scale studies and reviewing the conditions that affect the range of applica-
bility and methods used to decide applicability (see Section 4.1 through
4.3), one can begin to understand some of the limitations of vapor extrac-
tion-based technologies. Table 6.1 (on page 6.2) summarizes a number of
scenarios in which the chances of conducting "successful" vapor extraction-
based remediation are diminished. In most of these scenarios there is a
condition that limits the rate of remediation (e.g., soil permeability or soil
heterogeneity), or affects design (e.g., shallow water tables or strict air
emission requirements). Thus, in these scenarios, vapor extraction-based
processes may not be cost competitive with other remedial options.
In addition, there is a limitation common to all in situ technologies.
Given the information available from typical site characterization efforts,
systems are selected, designed, and operated under conditions of consider-
able uncertainty. Thus, there is no guarantee that any in situ process will
perform as expected, even under seemingly ideal conditions. To some de-
gree, the effect can be minimized by incorporating some factor for uncer-
tainty into the analysis and by designing flexible systems that can be
expanded and operated over a range of conditions.
6.1
-------
Limitations
Table 6.1
Limitations of Vapor Extraction-Based Processes
Scenario
Processes Affected
Limitations
desired remediation time is short
(<6 months)
subsurface soils are highly
heterogeneous with large air
permeability contrasts, or soils are
fractured
vapor extraction, air
sparging, bioventmg
vapor extraction,
bioventing
most economically-feasible vapor
extraction-based system designs
achieve remediation in 0.5 - 3
year time frame
remediation rate is likely to be
limited by mass transfer rates
(diffusion-controlled), air-flow path
is difficult to control
subsurface soils are highly
heterogeneous with large air
permeability contrasts, or soils are
fractured
air sparging
mjecled air-flow path is difficult to
control
residual contamination is located in
the groundwater capillary fringe
vapor extraction, may be difficult to extract vapors and
bioventmg deliver O2 without causing
groundwater upwelhng to saturate
contaminated zone, groundwater
drawdown system may be required
contaminant does not partition well
to vapor phase (i e , vapor pressure
<1 mm Hg, or dimensionless
Henry's Law constant <0 001)
vapor extraction, air
sparging
volatilization removal rates will be
low, unless extracted/injected
vapor flow rates are very high,
will be difficult to reach soil
cleanup levels <1 mg/kg
soil is not very permeable to air-flow
(i.e., permeability <10 9 cm2)
soil ib not very permeable to air-flow
(i e., permeability <10'n cm2)
compound is resistant to
biodegradation
vapor extraction, air
sparging
bioventing
bioventing
volatilization removal rates will be
low, unless vapor concentrations
are very high
oxygen delivery (and biodegradation)
rate will be low
low biodegradation rates
very little above-ground surface area
available
above-ground vapor
extraction, above-ground
bioventmg
above ground processes require
dedicated area for duration of
remediation (often >3 months)
shallow (<5 ft BOS) contamination
vapor extraction,
bioventmg, air sparging
"radius of influence" of vapor
extraction wells is small, economics
may favor other options (i.e.,
excavation & disposal)
shallow groundwater (<5 ft BOS)
vapor extraction,
bioventmg, air sparging
"radius of influence" of vapor
extraction wells is small, economics
may favor other options (i.e.,
excavation & disposal)
very strict air emissions
requirements
vapor extraction, air
sparging, bioventing
vapor treatment costs can make vapor
extraction uneconomical relative to
other options
BGS - Below Ground Surface
6.2
-------
Chapter 7
TECHNOLOGY PROGNOSIS
Currently, there are few effective in situ processes for treating soils con-
taminated with volatile components, and soil vapor extraction continues to
be the preferred option for treating soils contaminated by fuel tank releases.
Similarly, there are few alternatives to air sparging for residuals trapped
below the water table and to bioventing for treating less-volatile biodegrad-
able compounds. Of the vapor extraction-based technologies, soil vapor
extraction is the most fully developed, having been practiced longest and at
the largest number of sites. Results from laboratory and modeling studies
have helped develop a basic understanding of the processes affecting soil
vapor extraction system performance. There are, nonetheless, identifiable
needs for further development and demonstration, discussed in Section 7.1
(see Section 3.8 for recommendations for improving the state of the practice
of soil vapor extraction). Air sparging and bioventing, on the other hand,
are promising developmental technologies and a niche has been identified
for each. Needs for further development and demonstration are discussed in
Sections 7.2 and 7.3. Other vapor extraction-based processes, such as in
situ soil heating, should be considered experimental at this time.
7,1 Soil Vapor Extraction
Although this section addresses soil vapor extraction, it applies to all
vapor extraction-based processes.
Under the current state of the practice, systems are designed based on the
"radius of influence" calculated from pilot-test data (this is an extension of
groundwater "capture zone" design practices). A serious defect in this ap-
proach is that systems are designed without considering remedial objec-
tives, especially time and cost. Although a poorly designed system may
7.1
-------
Technology Prognosis
eventually remediate a site, the cost will likely be substantially increased.
While appealing in its simplicity, the radius of influence approach has no
technical basis, and there is no obvious relationship between the radius of
influence (as it is commonly defined by subsurface pressure distributions)
and the zone of soil remediated by a vapor extraction well in a given period,
as illustrated by Johnson and Ettinger (1994). Thus, there is a need for
design methods to incorporate long-term system performance predictions,
remedial goals, economic, and other considerations.
To cost effectively operate a vapor extraction-based system, one must
have a good understanding of induced vapor-flow patterns in the subsur-
face. Vapor-flow patterns are deduced from subsurface pressure measure-
ments, knowledge of the site stratigraphy, hydrogeology, and vapor-flow
models. It would be useful to have practical devices that directly measure
vapor-flow paths and flow velocities in situ, such as that described by
Kerfoot (1992).
When vapor treatment is required, it is often a significant part of total
vapor extraction system costs. Presently, most vapor treatment units (e.g.,
thermal oxidizers, catalyst beds, and internal combustion engines) combust
vapors; however, lower cost, noncombustion alternatives, including biodeg-
radation- and absorption-based processes, are being evaluated.
Another factor that may be limiting practitioners' use of engineering
analysis (see Subsection 3.8.7) is the lack of readily accessible computing
tools. If programmed in a user-friendly manner, their use can minimize the
time and costs of analyses. For soil venting applications, accessible predic-
tive tools are limited primarily to air-flow models, such as CSUGAS
(Warner et al. 1991) and AIRFLOW (Joss and Baehr 1993a), and the
screening-level flow, partitioning, and transport models, such as
Hyperventilate (Johnson and Stabenau 1991). For bioventing and air
sparging, such tools are essentially nonexistent. The use of engineering
analysis, or modeling, will likely increase in the future as additional com-
puting tools become available.
7.2
-------
Chapter 7
7.2 Air Sparging
As observed in Subsection 3.1.2, considerable research and demonstra-
tion must be conducted before a consistent and reliable design approach
will be realized. To date, limited information exists on the assessment,
monitoring, performance analysis, modeling, and engineering design neces-
sary to implement air sparging with confidence and predictability. With
proper assessment, design, and operation, however, air sparging may be-
come a powerful remedial technology. Its potential for circumventing the
need for prolonged groundwater pumping systems certainly makes it attrac-
tive.
The effectiveness of air sparging is dependent upon how well the in-
jected air travels through the saturated zone. Although there is a large body
cf literature focused on multiphase flow (e.g., air and water) there is cur-
rently little practical understanding of how these processes influence the
range of behavior likely to occur in porous media at air-sparging sites. As a
result, many air-sparging systems are being installed on a trial-and-error
basis. To accelerate the development and understanding of the capabilities
of this process, the following are needed:
• a better understanding of the distribution of air resulting from
subsurface injection and how it is affected by soil structure and
process variables;
• measuring devices to help better delineate the effect of an air-
sparging well at a given site;
• guidelines for applicability and field testing based on fundamen-
tal considerations; and
• well-documented studies illustrating the performance and long-
term impact of air sparging at a number of sites.
It is also important to note that, as explained in Subsection 3.1.2, air is
being forced into the subsurface and the possibility of vapor migration to
nearby buildings or utility corridors is a concern. Thus, it would be useful
to have a monitoring method that insures proper control of vapor-flow
paths.
7.3
-------
Technology Prognosis
7.3 Bioventing
Like air sparging, bioventing is also a developing process, and many
questions remain unanswered. Following are some that need to be ad-
dressed:
• What situations are amenable to biodegradation?
• What range of biodegradation rates can be achieved in the field?
• Are O2/CO2 vapor concentrations reliable indicators of biodegra-
dation? and
• What is the most cost-effective practice of the process?
7.4
-------
Appendix A
LIST OF REFERENCES
Ardito, C.P. and J.F. Billings. 1990. Alternative remediation strategies: the
subsurface volatilization and ventilation system. In Proc. Petroleum Hydrocar-
bons and Organic Chemicals in Ground Water: Prevention, Detection, and
Restoration. Houston, Texas. Oct. 31 - Nov. 2.
Aggarwal, P.K., J.L. Means, and R.E. Hinchee. 1991. Formulation of nutrient
solutions for in situ bioremediation. In In Situ Bioreclamation, eds. R.E.
Hinchee and R.F. Olfenbuttel, 51-66. Stoneham, Mass: Butterworth-
Heinemann.
Atkins, P.W. 1978. Physical Chemistry. Oxford University Press.
Baehr, A.L. and G.E. Hoag. 1988. A modeling and experimental of induced
venting of gasoline-contaminated soils. Soils Contaminated with Petroleum:
Environmental and Public Health Effects, eds. E.J. Calabrese, P.T. Kostecki,
113-124. John Wiley & Sons.
Baehr, A.L., G.E. Hoag, and M.C. Marley. 1989. Removing volatile contami-
nants from the unsaturated zone by inducing advective air-phase transport. J. of
Contaminant Hydrology 4:1.
Baehr, A.L. and M.F. Hull. 1991. Evaluation of unsaturated zone air permeabil-
ity through pneumatic tests. Water Resources Research 27(10).
Bennedsen, M.B., J.P. Scott, and J.D. Hartley. 1987. Use of vapor extraction
systems for in-situ removal of volatile organic compounds for soil. In Proc.
National Conference of Hazardous Wastes and Hazardous Materials. Hazard-
ous Materials Control Research Institute. Washington, D.C.
Benson, D.A. 1992. A two dimensional, finite difference model of the phase
distribution and vapor transport of multiple compounds. In Proc. of the Third
Annual West Conference on Hydrocarbon Contaminated Soils and Groundwa-
ter: Analysis, Fate, Environmental and Public Health Effects, and Remediation.
Long Beach, Calif. March 9-12.
Billings, O.K. 1991. Personal communication.
A.I
-------
List of References
Bohler, U., J. Brauns, H. Hotzl, and M. Nahold. 1990. Air injection and soil air
extraction as a combined method for cleaning contaminated sites - observations
from test sites in sediments and solid rocks. In Contaminated Soil '90, eds. F.
Arendt, M. Hinsenveld, and W.J. van den Brink. Boston: Kluwer Academic
Publishers.
Boublik, T., V. Fried, and E. Hala. 1984. The Vapour Pressures of Pure Sub-
stances. New York: Elsevier.
Briggs, P.J. 1989. A Simulator for the recovery of heavy oil from naturally
fractured reservoirs using cyclic steam injection. In Proc. of the 6th Middle
East Oil Show, Paper 17954. Manama, Bahrain. March 11-14.
Brock, T.D. 1970. Biology of Microorganisms. Englewood Cliffs, N.J.:
Prentice-Hall, Inc.
Brown, R.A., and R. Fraxedas. 1991. Air sparging - extending volatilization lo
contaminated aquifers. In Proc. of the Symposium on Soil Venting. R.S. Ken-
Environmental Research Laboratory. Houston. April 29 - May 1.
Brown, R.A., E. Henry, and C. Herman. 1992. Treatment of a solvent contami-
nated site with air sparging/soil vapor extraction. In Proc. American Institute of
Chemical Engineers, Spring National Meeting. New Orleans.
Brown, R.A., C. Herman, and E. Henry. 1991. The use of aeration in environ-
mental clean-ups. In Proc. Haitech International Pittsburgh Waste Conference.
Pittsburgh. May 14-16.
Bruell, C.J. and G.E. Hoag. 1984. Capillary and packed-column gas chromatog-
raphy of gasoline hydrocarbons and EDB. In Proc. Petroleum Hydrocarbons
and Organic Chemicals in Groundwater-Prevention, Detection,and Restora-
tion. National Water Well Association. Houston. Nov.5-7.
Buelt, J.L., and J.H. Westsik. 1987. In situ vitrification — preliminary results
from the first large-scale radioactive test. PNL-SA-15277. Richland, Wash.:
Pacific Northwest Laboratory.
Buscheck, T.E. and T.R. Peargin. 1991. Summary of a nation-wide vapor ex-
traction system performance study. In Proc. Petroleum Hydrocarbons and
Organic Chemicals in Ground Water Conference: Prevention, Detection, and
Restoration. National Water Well Association; American Petroleum Institute.
Houston.
Castalia, 1990.
Conner, J.R. 1988. Case study of soil venting. Pollution Engineering 7:7478.
Corapcioglu, M.Y. and A.L. Baehr. 1987. A compositional multiphase model
for groundwater contamination by petroleum products, 1. theoretical consider-
A.2
-------
Appendix A
ations. Water Resources Research 23(1).
Dinkel, D.H. 1966. Polyethylene mulches for sweet corn in northern latitudes.
In Proc. Am. Hort. Soc. 89:497.
Downey, D.C. 1993. Personal communication form Doug Downey, Senior
Engineer, Engineering Science, Denver, Colo.
Downey, D.C. and P.R. Guest. 1991. Physical and biological treatment of deep
diesel-contaminated soils. A field scale investigation of enhanced petroleum
hydrocarbon biodegradation in the vadose zone at Tyndall AFB, Florida. In
Proc. Petroleum Hydrocarbons and Organic Chemicals in Ground Water:
Prevention, Detection, and Restoration. Houston. Nov. 20 - 22.
Dupont, R.R., W. Doucette, and R.E. Hinchee. 1991. Assessment of in situ
bioremediation potential and the application of bioventing at a fuel-contami-
nated site. In In Situ and On-Site Bioreclamation, eds. R.E. Hinchee and R.F.
Olfenbuttel, 262-82. Stoneham, Mass.: Butterworth-Heinemann.
Ely, D.L., and D.A. Heffner. 1988. Process for in-situ biodegradation of hydro-
carbon contaminated soil. U.S. Patent 4,765,902.
Esch, D.C. 1984. Surface modifications for thawing of permafrost. In Proc.
Third International Specialty Conference — Cold Regions Engineering, North-
ern Resource Development. Edmonton, Alberta, Canada. April 4-6.
Essaid, H.I., W.N. Herkelrath, and K.M. Hess. 1991. Air, oil, and water distri-
butions at a crude-oil spill site. Bemidji, Minnesota. U.S. Geological Survey
Toxic Substances Hydrology Program. In Proc. Technical Meeting. U.S. Geo-
logical Survey Water-Resources Investigations Report 91-4034. Monterey,
Calif. March 11-15.
Farrell, J. and M. Reinhard. 1994. Desorption of halogenated organics from
model solids, sediments and soil under saturated conditions. Environmental
Science and Technology. 28(1): 53-72.
Felten, D.W., M.G. Leahy, L.J. Bealer, and B.A. Kline. 1992. Site remediation
using air sparging and soil vapor extraction. Presented at NGWA Petroleum
Hydrocarbons and Organic Chemicals in Groundwater: Prevention, Detection,
and Restoration. Houston. Nov.
Gannon, K., D.J. Wilson, A.N. Clarke, R.D. Mutch, Jr., and J.H. Clarke. 1989.
Soil clean-up by in-situ aeration, II. Effects of impermeable caps, soil perme-
ability, and evaporative cooling. Sep. Sci. Technol. 24:831.
Ghassemi, M. 1988. Innovative in situ treatment technologies for cleanup of
contaminated sites. J. Hazard. Mater. 71:189.
A.3
-------
List of References
Gomez-Lahoz, C., J.M. Rodriguez-Maroto, and D.J. Wilson. 1991. Soil clean
up by in situ aeration. VI. Effects of variable permeabilities. Sep. Sci. Technol.
26:133.
Griffin, C.J., J.M. Armstrong, and R.H. Douglas. 1991. Engineering design
aspects of an in situ soil vapor remediation system (sparging). In In Situ
Bioremediation. Stoneham, Mass.: Butterworth-Heinemann.
Heath, W.O. 1990. In situ heating to destroy and remove organics from soils, fn
Proc. Department of Energy/Air Force Joint Technology Review Meeting, Soil
and Groundwater Remediation. Atlanta. Feb. 6-8.
Hillel, D. 1982. Introduction to soil physics. New York: Academic Press.
Hinchee, R.E. and M. Arthur. 1991. Bench scale studies of the soil aeration
process for bioremediation of petroleum hydrocarbons. Applied Biochemistry
and Biotechnology 28/29:901.
Hinchee, R.E., D.C. Downey, and T. Beard. 1989. Enhancing biodegradation of
petroleum hydrocarbon fuels through soil venting. In Proc. Petroleum Hydro-
carbons and Organic Chemicals in Ground Water: Prevention, Detection, and
Restoration. National Ground Water Association, american Petroleum Institute,
Houston. Nov. 15 - 17.
Hinchee, R.E., D.C. Downey, J.K. Slaughter, and M. Westry. 1989. Enhanced
Biorestoration of Jet Fuels: a Full-Scale Test at Elgin Air Force Base. Air
Force Engineering and Services Report ESL/TR/88-78. Fla.
Hinchee, R.E., and S.K. Ong. 1992. A rapid in situ respiration test for measur-
ing aerobic biodegradation rates of hydrocarbons in soil. / of the Air & Waste
Mgmt.Assoc. 42(10): 1305-1312.
Hinchee, R.L. 1994. Unpublished data from numerous sites. Battelle, Colum-
bus, Ohio.
Hoag, G.E. and B. Cliff. 1988. Use of soil venting for the treatment/reuse of
petroleum-contaminated soils. Soils Effects, eds., E.J. Calabrese, P.T. Kostecki,
301-316. John-Wiley & Sons.
Hoag, G.E. and M.C. Marley. 1986. Gasoline residual saturation in unsaturated
uniform aquifer materials. J. Environ. Eng. Div. Am. Soc. Civ. Eng. 112(3).
Houthoofd, J.M., J.H. McCready, and M.H. Roulier. 1991. Solar heating with
optical fibers. In Solar heating technologies for in situ treatment: a review.
Remedial action, treatment, and disposal of hazardous waste, 199-200. In Proc.
Seventeenth Annual RREL Hazardous Waste Research Symposium. Risk Re-
duction Engineering Laboratory, Office of Research and Development. EPA/
600/9-91/002. Washington, DC. April 9-11.
A.4
-------
Appendix A
Hubbert, M.K. 1940. The theory of groundwater motion. J. of Geology. 48(8).
Hull, M.F., M.K. Landon, and H.O. Pfannkuch. 1991. Field validation of con-
ceptual models of mobilization and transport of volatile petroleum derivatives
in the unsaturated zone near Bemidji, Minnesota. In Proc. Technical Meeting.
U.S. Geological Survey Water Resources Investigations Report 91-4034. U.S.
Geological Survey Toxic Substances Hydrology Program. Monterey, Calif.
March 11-15.
Hutzler, N.J., B.E. Murphy, and J.S. Gierke. 1989. State of technology review
soil vapor extraction systems. EPA/600/2-89/024. U.S. Department of Com-
merce NTIS document PB89-195184.
Ji, W., A. Dahmani, D. Ahlfeld, J.D. Lin, and E. Hill. 1993. Laboratory study
of air sparging: air flow visualization. Ground Water Monitoring and
Remediation, Fall 1993.
Johnson, Jr., L.A., and F.D. Guffey. 1990. Contained recovery of oily wastes
(CROW). Draft final report, US EPA Agreement CR-815333. Laramie, Wyo-
ming.
Johnson, P.C. and R.A. Ettinger. 1992. Some considerations for the design of in
situ vapor extraction extraction systems: radius of influence -vs- radius of
remediation. Ground Water Monitoring and Remediation, Summer 1994.
Johnson, P.C., M.W. Kemblowski, and J.D. Colthart. 1988. Practical screening
models for soil venting applications. In Proc. Petroleum Hydrocarbons and
Organic Chemicals in Ground Water Conference. National Water Well Asso-
ciation, American Petroleum Institute. Houston.
Johnson, P.C., M.W. Kemblowski, and J.D. Colthart. 1990. Quantitative analy-
sis for the cleanup of hydrocarbon-contaminated soils by in situ soil venting.
Ground Water 3(28): 413.
Johnson, P.C., C.C. Stanley, M.W. Kemblowski, D.L. Byers, and J.D. Colthart.
1990. A practical approach to the design, operation, and monitoring of in situ
soil-venting systems. Ground Water Monit. Rev. Spring.
Johnson, P.C. and A.J. Stabenau. 1991. Hyperventilate - A Software Guidance
System Created for Vapor Extraction Applications. Shell Oil Company (see
also USEPA 510-CB-92-001 for software orders).
Johnson, P.C., A.L. Otermat, and C.C. Chou. 1991. In situ decontamination of
spills and landfills by focussed microwave/radio-reequency heating and a
closed-loop vapor flushing and vacuum recovery system. U.S. Patent
5,076,727. Dec. 31.
Johnson, P.C., C.C. Stanley, D.L. Byers, D.A. Benson, and M.A. Acton. 1991.
Soil venting at a California site: field data reconciled with theory. Hydrocarbon
A.5
-------
List of References
contaminated soils and groundwater: analysis, fate, environmental and public
health effects. Remediation, Vol I., eds P.T. Kostecki and E.J. Calabrese. Lewis
Publishers.
Johnson, P.C., J.D. Colthart, A.L. Otermat, D.A. Weingaerfner, C.C. Chou,
D.L. Byers, S.M. Steerns, A.R. Marsden, Jr., and G.M. Deeley. 1992. Soil
Decontamination. U.S. Patent 5,114,497. May 19.
Joss, C.J. and A.L. Baehr. (1993a). AIRFLOW- An adaptation of the ground
water flow code mod/low to simulate 3-d airflow in the unsaturated zone. To be
published as a U.S. Geological Survey Open File Report.
Joss, C.J. and A.L. Baehr. (1993b). AIRTEST- A computer code to simulate 2-d
radially symmetric flow in the unsaturated zone. To be published as a U.S.
Geological Survey Open File Report.
Kaback, D.S., B.B. Looney, C.A. Eddy, and T.C. Hazen. 1991. Innovative
ground water and soil remediation: in situ air stripping using horizontal wells.
In Proc. Fifth National Outdoor Action Conference on Aquifer Restoration,
Ground Water Monitoring, and Geophysical Methods. Las Vegas. May 13-16.
Karicknoff, S.W. 1981. Semi-empirical evaluation of sorption of hydrophobic
pollutants on natural sediments and soils. Chemosphere. 10(8): 833.
Kerfoot, H.B., C.L. Mayer, P.B. Durgin, and J. D'Lugosz. 1988. J. Groundwa-
ter Monitoring Review. 8:67.
Kerfoot, W.B. 1992. Direct measuring soil gas velocity probes for subsurface
vapor control on sparging sites. In Proc. Sixth National Outdoor Action Confer-
ence on Aquifer Restoration, Ground Water Monitoring, and Geophysical
Methods. Las Vegas. May 11-13.
Koegler, S.S. 1989. Project summary 116-B-6-1 crib in situ vitrification dem-
onstration project. PNL-SA-16553. Richland, Wash.: Pacific Northwest Labo-
ratory.
Kresge, M.W., and M.F. Dacey. 1991. An evaluation of in situ groundwater
aeration. Paper presented at HazMat International '91. Atlantic City.
La Mori, P.N. 1989. In situ hot air/steam extraction of volatile organic com-
pounds. In Solid/liquid separation: waste management and productivity en-
hancement, ed. H.S. Muralidhara, 386-395. Columbus, Ohio: Battelle Press.
Leeson, A., R.E. Hinchee, J. Kittel, G. Sayles, C.M. Vogel, and R. N. Miller.
1993. Optimizing bioventing in shallow vadose zones and cold climates. J. of
Hydrogeological Sciences 38(4): 283-95.
Lord, A.E., D.E. Hullings, R.M. Koerner, and J.E. Brugger. 1990a. Vacuum-
assisted steam stripping to remove pollutants from contaminated soil: a labora-
tory study. In Proc. Sixteenth Annual Hazardous Waste Symposium on
A.6
-------
Appendix A
4
t
Remedial Action, Treatment of Hazardous Waste, 377. EPA/600/9-90/037. US
EPA, Cincinnati.
Lord, A.E., R.M. Koerner, D.E. Hullings, and I.E. Brugger. 1990b. Laboratory
studies of vacuum-assisted steam stripping of organic contaminants from soil.
In Proc. Fifteenth Annual Research Symposium for Land Disposal, Remedial
Action, Incineration, and Treatment of Hazardous Waste, 124. EPA/600/9-90/
006. US EPA, Cincinnati.
Lyman, W.J., W.F. Reehl, D.H. Rosenblatt. 1982. Handbook of chemical prop-
erty estimation methods. New York: McGraw-Hill.
MacKay, D. and W.Y. Shiu. 1981. A critical review of Henry's Law constants
for chemicals of environmental interest. J. Physical Chemistry Reference Data
10(4): 1175.
Maksimenko, E.S. 1984. Combined method for electric thawing of permafrost.
Soil Mechanics Found. Eng. 21(6): 261.
Marley, M.C. 1985. Quantitative and qualitative analysis of gasoline fractions
stripped by air from the unsaturated zone. M.S. Thesis, Department of Civil
Engineering, University of Connecticut.
Marley, M.C. 1991. Air sparging in conjunction with vapor extraction for
source removal at VOC spill sites. In Proc. Fifth National Outdoor Action
Conference on Aquifer Restoration, Ground Water Monitoring, and Geophysi-
cal Methods. Las Vegas. May 13-16.
Marley, M.C. and G.E. Hoag. 1984. Induced soil venting for the recovery /
restoration of gasoline hydrocarbons from the vadose zone. In Proc. Petroleum
Hydrocarbons and Organic Chemicals in Ground Water Conference., National
Water Well Association; American Petroleum Institute. Houston.
Massmann, J.W. 1989. Applying groundwater flow models in vapor extraction
system design. J. Environ. Eng. 115(1). 129.
Middleton, A.C., and D.H. Hiller. 1990. In situ aeration of groundwater a tech-
nology overview. Paper presented at A Conference on Prevention and Treat-
ment of Soil and Groundwater Contamination in the Petroleum Refining and
Distribution Industry. Montreal, Quebec, Canada. Oct. 16 - 17.
Miller, R.N. 1990. Afield scale investigation of enhanced petroleum hydrocar-
bon biodegradation in the vadose zone combining soil venting as an oxygen
source with moisture and nutrient addition. Ph.D. Diss. Utah State University,
Logan, Utah.
Miller, R.N., R.E. Hinchee, C.M. Vogel, R.R. DuPont, and D.C. Downey.
1990. A field scale investigation of enhanced petroleum hydrocarbon biodegra-
dation in the vadose zone at Tyndall AFB, Florida. In Proc. Petroleum Hydro-
A.7
-------
List of References
carbons and Organic Chemicals in Ground Water: Prevention, Detection, and
Restoration. Houston. Oct. 31 - Nov. 2.
Miller, R.N., R.E. Hinchee, and C. Vogel. 1991. A field-scale investigation of
petroleum hydrocarbon biodegradation in the vadose zone enhanced by soil
venting at Tyndall AFB, Florida. In In Situ and On-Site Bioreclamation, Vol. 1,
eds. R.E. Hinchee and R.F. Olfenbuttel, 283-301. Stoneham, Mass:
Butterworth-Heinemann Publishers.
Murdoch, L., B. Patterson, G. Losonsky, W. Harrar, and K. Dotson. 1989. A
review of innovative technologies of delivery or recovery for the remediation of
hazardous waste sites. Contract No. 68-03-3379, Work Assignment 1-11. US
EPA, Risk Reduction Engineering Laboratory. Cincinnatti.
Mutch Jr, R.D. and DJ. Wilson. 1990. Soil clean-up by in-situ aeration. IV.
anisotropic permeabilities. Sep. Sci. Technol. 25:1.
Nicholson, R.H. 1978. Permafrost modification by changing the natural energy
budget. In Proc. Third International Conference on Permafrost. Edmonton,
Alberta, Canada. July 10-13.
Nunno, T., J. Hyman, P. Spawn, J. Healy, C. Spears, M. Brown, and C. Jonker.
1989. In situ steam stripping of soils. International Fact Sheet, Assessment of
International Technologies for Superfund Application — Technology Identifi-
cation and Selection. EPA/600/2-89-017. US EPA.
Oak Ridge National Laboratory. 1990. Enhancement of removal by heating.
Oak Ridge National Laboratory Draft Report, Section G. Submitted to
Armstrong Laboratory, Environics Division, US Air Force, Tyndall AFB, Fl,
Osejo, R.E. and DJ. Wilson. 1991. Soil clean up by in situ aeration. IX. diffu-
sion constants of volatile organics and removal of underlying liquid. Sep. Sci.
Technol. 26(12): 1433.
Ostendorf, D.W. and D.H. Kampbell. 1989. Vertical profiles and near surface
traps for field measurement of volatile pollution in the subsurface environment.
In Proc. New Field Techniques for Quantifying the Physical and Chemical
Properties of Heterogenous Aquifers. National Water Well Association. Dallas.
Perry, R.H., G.H. Chilton, and S.D. Kirkpatrick. 1973. Chemical engineers'
handbook. 5th ed. New York: McGraw-Hill.
Peters, M.S. and K.D. Timmerhaus, 1980. Plant design and economics for
chemical engineers, p. 511. McGraw-Hill.
Rodriguez-Maroto, J.M., C. Gomez-Lahoz, and D.J. Wilson. 1991. Soil clean
up by in situ aeration. VIE. effects of system geometry on vapor extraction
efficiency. Sep. Sci. Technol. 26:1051.
Roy and Griffen. 1991. /. of Hazardous Materials. 26:301-17.
A.8
-------
Appendix A
Sabadell, G.P., J.J. Eisenbeis, and O.K. Sumada. 1988. The three-dimensional
model CSUGAS: a management tool for the design and operation of soil vent-
ing systems. HMCRI Superfund '88 Conference. Washington, D.C. Nov.
Shan, C.S., R.W. Falta, and I. Javandel. 1992. Analytical solutions for steady
gas flow to a soil vapor extraction well. Water Res. Res. 28(4): 1105.
Smith, G.J., R.R. Dupont, and R.E. Hinchee. 1991. Hill AFB JP-4 site (building
914) remediation. A Report to Hill AFB, Utah.
Smith, G.J., J.W. Aiken, and J.F. Tursman. 1990. A systems approach for rapid
cleanup of chlorinated solvents in groundwater. In Proc. First USA/USSR Joint
Conference on Environmental Hydrology and Hydrogeology. Leningrad,
USSR. June 18-21.
Sparks, D.L. 1989. Kinetics of soil chemical processes. San Diego: Academic
Press.
Staps, J.J.M. 1989. International evaluation ofin-situ biorestoration of con-
taminated soil and groundwater. Dutch National Institute of Public Health and
Environmental Protection (RIVM). RIVM Report No.738708006. Bilthoven.
Texas Research Institute. 1980. Laboratory scale gasoline spill and venting
experiment. American Petroleum Institute Interim Report No. 77435JST.
Washington, D.C.
Texas Research Institute. 1984. Forced venting to remove gasoline vapor from
a large-scale model aquifer. American Petroleum Institute Final Report 82101-
F:TAV. Washington, D.C.
Thornton, J.S., and W.L. Wootan, Jr. 1982. Venting for the removal of hydro-
carbon vapors from gasoline contaminated soil. J. Environmental Science and
Health A17(l): 31.
Trewis, C.C. and J.M. Macinnis. 1992. Vapor extraction of organics from sub-
surface soils, is it effective? Environ. Sci. Technol. 26(10): 1885-7.
Udell, K.S., and L.D. Stewart, Jr. 1989. Field study of in situ steam injection
and vacuum extraction for recovery of volatile organic solvents. UCB-
SEEHRL Report 89-2, University of California. Berkeley, Calif.
US EPA. 1988. Assessment of international technologies for Superfund appli-
cations, 37-40. EPA/540/2-88/011. OSWER. Washington, D.C. Sept.
US EPA. 199la. Soil vapor extraction technology reference handbook. EPA/
540/2-91/003. CERI. Cincinnati. Feb.
US EPA. 1991b. Guide for conducting treatability studies under CERCLA: soil
vapor extraction interim guidance. EPA/540/2-91/019A. Office of Emergency
and Remedial Response. Washington, D.C. Sep.
A.9
-------
List of References
US EPA. 1993. Innovative treatment technologies: annual status report. 5th ed.
Number 5. EPA/542-R-93-003. NCEPI. Cincinnati. Sep.
van Eyk, J., and C. Vreeken. 1989. Venting-mediated removal of diesel oil
from subsurface soil strata as a result of stimulated evaporation and enhanced
biodegradation, 475-485. In Hazardous Waste and Contaminated Sites.
Envirotech Vienna, Vol. 2, Session 3. ISBN 389432-009-5. Westarp Wiss.:
Essen
Vershueren, K. 1983. Handbook of environmental data on organic chemicals.
New York: Van Nostrand Reinhold Co., Inc.
Weast, R.C. 1970. Handbook of chemistry and physics. 51st ed. Cleveland: The
Chemical Rubber Company.
Wehrle, K. 1990. In situ cleaning of CMC contaminated sites: model-scale
experiments using the air injection (in situ stripping) method in granular soils.
In Contaminated Soil '90, eds. F. Arendt, M. Hinsenveld, and W.J. van den
Brink, 1061-2. Boston: Kluwer Academic Publishers.
Wilson, D.J. 1990. Soil clean-up by in-situ aeration, v. vapor stripping from
fractured bedrock. Sep. Sci. Technol. 23:991.
Wilson, D.J., A.N. Clarke, and J.H. Clarke. 1988. Soil clean-up by in-situ aera-
tion, i. mathematical modelling. Sep. Sci. Technol. 23:991.
Wilson, D.J., A.N. Clarke, and R.D. Mutch Jr. 1989. Soil clean-up by in-situ
aeration. III. passive vent wells, recontamination, and removal of underlying
non-aqueous phase liquid. Sep. Sci. Technol. 24:939.
Wilson, J.T. 1994. Personal communication from John T. Wilson, Senior Mi-
crobiologist, US EPA, Ada, Okla.
Wilson, J.T., and C.H. Ward. 1986. Opportunities for bioremediation of aqui-
fers contaminated with petroleum hydrocarbons. /. of Industrial Microbiology.
27:109.
Woo, S. and E. Coleman. 1988. Effect of low oxygen condition in service sta-
tion vadose wells on hydrocarbon vapor measurements. In Proc. HazMaCon
88. Association of Bay Area Governments. Anaheim, Calif.
Wyza, R., and R.E. Hinchee. 1990. Unpublished data, Battelle, Columbus,
Ohio.
A.10
-------
Appendix B
SUGGESTED READING LIST
Ahlfeld, D., A. Dahmani, M. Farrell, and W. Ji. 1992. Detailed field measure-
ments of an air sparging pilot test. Submitted to Groundwater.
Hinchee. R.E., D.C. Downey, R.R. DuPont, P.K. Aggarwal, and R.N. Miller.
1991. Enhancing biodegradation of petroleum hydrocarbons through soil vent-
ing. J. of Hazardous Mater. 27:315.
Hunt J.R., N. Sitar, and K.S. Udell. 1988. Nonaqueous phase liquid transport
and cleanup I. Analysis of mechanisms. Water Resources Research 24(8):
1247.
Hunt J.R., N. Sitar, and K.S. Udell. 1988. Nonaqueous phase liquid transport
and cleanup I. Experimental studies. Water Resources Research 24(8): 1259-
69.
Marley, M.C., M.T. Walsh, and P.E. Nangeroni. 1990. A case study on the
application of air sparging as a complimentary technology to vapor extraction
at a gasoline spill site in Rhode Island. In Proc. Hydrocarbon Contaminated
Soils and Groundwater Conference, University of Massachusetts, Amherst,
Mass.
McAuliffe, C. 1966. Solubility in water of paraffin, cycloparaffin, olefm, acety-
lene, cycloolefin, and aromatic hydrocarbons. J. Physical Chemistry 70(4):
1267.
Mendoza, C.A. and E.G. Frind. 1990. Advective-dispersive transport of dense
organic vapors in the unsaturated zone: 1. model development. Water Re-
sources Research. 26(3): 379.
Muskat, M. and H.G.Botset. 1931. Flow of gas through porous materials, Vol 1
In Physics.
D.S. GOVERNMENT PRINTING OFFICE: 1995-620-508/82045
B.I
------- |