PB87-182291
BACKGROUND DOCUMENT ON BOTTOM LINER PERFORMANCE
IN DOUBLE-LINED LANDFILLS AND SURFACE IMPOUND-
MENTS
U.S. Environmental Protection Agency
Washington, DC
Apr 87
LIBRARY
EJBD
ARCHIVE
EPA
530-
sw-
87-
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
PB87-182291
EPA/530-SW-87-013
BACKGROUND DOCUMENT
ON
BOTTOM LINER PERFORMANCE
IN
DOUBLE-LINED LANDFILLS
AND SURFACE IMPOUNDMENTS
Prepared for
.o
c£ U.S. Environmental Protection Agency
.CO ^ Office of Solid Waste
-"-J.-n ^ Washington, D.C. 20460
~-. ;'•.•> .2: st
'n o
Work Assignment No. 7
r.;5 6 .~~ o (Amendment 4)
I* s >,
-g O
GeoServices Inc. Consulting Engineers
1200 South Federal Highway, Suite 204
Boynton Beach, Florida 33435
April 1987
O
CD
"
O
vr
^
REPRODUCED BY
U.S. DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL
INFORMATION SERVICE
SPRINGFIELD. VA 22161
-------
DISCLAIMER
This Report was furnished to the U.S. Environmental Protection
Agency by GeoServlces Inc. Consulting Engineers, 1200 S. Federal
Highway, Boynton Beach, Florida 33435, under subcontract to NUS
Corporation and 1n fulfillment of Contract No. 68-01-7301, Work
Assignment No. 7, Amendment 4. The primary authors of this document
are Drs. R. Bonaparte, J.F. Beech, and J.P. Glroud. The opinions,
findings, and conclusions expressed are those of the authors and not
necessarily those of the Environmental Protection Agency or
cooperating agencies. Mention of company or product names 1s not to
be considered an endorsement by the Environmental Protection Agency.
-------
TABLE OF CONTENTS
Chapter 1
INTRODUCTION
1.1 LEGISLATIVE HISTORY AND CURRENT REQUIREMENTS
1.1.1 Legislative and Regulatory History
1.1.2 Legislative Requirements
1.1.2.1 Current Requirements for Top Liner
1.1.2.2 Current Requirements for Bottom Liner
1.1.3 Proposed Double Liner Rule of March 28, 1986
1.1.4 Proposed Liner/Leak Detection Rule (Pending)
1.2' PURPOSE AND SCOPE OF THE BACKGROUND DOCUMENT
1.2.1 Purpose of the Background Document
1.2.2 Sources of Available Data
1.2.3 Scope of the Background Document
-------
Chapter 2
DEFINITIONS AND CONCEPTS
RELATED TO DOUBLE LINER SYSTEMS
2.1 INTRODUCTION
2.1.1 Purpose of this Chapter
2.1.2 Organization of this Chapter
2.2 WASTE MANAGEMENT UNITS
2.2.1 Introduction
2.2.1.1 Definition
2.2.1.2 Purpose of this Section
2.2.2 Description
2.2.2.1 Types of Waste Management Units
2.2.2.2 Geometry of Waste Management Units
2.2.2.2.1 Surface Impoundments
2.2.2.2.2 Landfills
2.2.2.2.3 Waste Piles
2.2.3 Ground Pollution Mechanism
2.2.3.1 Surface Impoundments
2.2.3.2 Landfills
2.2.3.3 Waste Piles
2.3 LINING SYSTEMS USED IN WASTE MANAGEMENT UNITS
2.3.1 Introduction
2.3.1.1 Importance of Lining Systems
2.3.1.2 Scope of this Section
2.3.1.3 Definition of Lining Systems
2.3.2 Materials Used in Lining Systems
2.3.2.1 Introduction
2.3.2.2 Liner Materials
2.3.2.2.1 Introduction
2.3.2.2.2 Compacted Soils
2.3.2.2.3 Geomembranes
2.3.2.3 Drainage Materials
2
-------
2.3.2.3.1 Introduction
2.3.2.3.2 High-Permeability Soils
2.3.2.3.3 Synthetic Drainage Materials
2.3.2.4 Transition Materials
2.3.2.4.1 Filters
2.3.2.4.2 Protective Layers
2.3.2.5 Reinforcement Materials
2.3.3 Double Liners
2.3.3.1 Introduction
2.3.3.1.1 Definitions
2.3.3.1.2 Terminology Related to Double Liners
2.3.3.2 Use of Double Liners in Waste Management Units
2.3.3.2.1 Current Regulations
2.3.3.2.2 Examples of Uses of Double Liners in
Waste Management Units
2.3.3.2.3 Influence of Liner on Leak Detection
2.4 LEAKAGE DEFINITION AND DETECTION
2.4.1 Definitions
2.4.1.1 Leak and Leakage
2.4.1.2 Leak Size and Leakage Rate
2.4.1.3 Leakage Collected and Leakage Out of the Unit
2.4.2 Leak Detection System
2.4.2.1 Definition
2.4.2.2 Purpose of Leak Detection
2.4.2.3 Performance Characteristics of Leak Detection
Systems
2.5 EPA LIQUIDS MANAGEMENT STRATEGY
2.5.1 Introduction
2.5.2 EPA Liquids Management Strategy
2.5.3 Performance Criteria for Evaluation of Bottom Liners
2.5.3.1 Leak Detection Sensitivity
2.5.3.2 Leachate Collection Efficiency
2.5.3.3 Leakage Out of the Unit
2.5.3.4 Breakthrough Time
-------
Chapter 3
PERFORMANCE OF COMPACTED SOIL LINERS
3.1 INTRODUCTION
3.2 FACTORS AFFECTING COMPACTED SOIL LINER PERFORMANCE
3.2.1 Nature of Compacted Soils
3.2.2 Hydraulic Conductivity
3.2.2.1 Compaction Effort and Water Content
3.2.2.2 Secondary Structures
3.2.2.2.1 Definition of Secondary Structures
3.2.2.2.2 Natural Secondary Structures
3.2.2.2.3 Construction Related Secondary
Structures
3.2.2.2.4 Environmentally Related Secondary
Structures
3.2.2.3 Interactions Between Compacted Soil and Leachate
3.2.3 Capillary Stresses
3.2.4 Settlement
3.2.5 Conclusions
3.3 CASE HISTORIES OF CLAY LINING SYSTEM PERFORMANCE
3.3.1 Overview of Case Histories
3.3.2 Summary of Case Histories
3.4 ANALYSIS OF PERFORMANCE - STEADY STATE SATURATED FLOW (ID)
3.4.1 Introduction
3.4.2 Overview of Analysis
3.4.3 Leak Detection Systems Sensitivity
3.4.4 Leachate Collection Efficiency
3.4.5 Leakage Out of the Unit
3.4.5.1 Hydraulic Head on Bottom Liner
3.4.5.2 Liner Thickness
3.4.5.3 Liner Hydraulic Conductivity
3.4.6 Breakthrough Time
3.4.6.1 Hydraulic Head on Bottom Liner
3.4.6.2 Liner Thickness
3.4.6.3 Liner Hydraulic Conductivity
-------
3.4.7 Summary
3.5 ANALYSIS OF PERFORMANCE - PARTIALLY SATURATED FLOW (ID)
3.5.1 Introduction
3.5.2 Overview of Analysis
3.5.2.1 Description of Model
3.5.2.2 Summary of Analysis Performed
3.5.3 Leak Detection Sensitivity
3.5.4 Leakage Out of Unit
3.5.4.1 Effect of Soil Suction Stress
3.5.4.2 Effect of Hydraulic Head
3.5.4.3 Effect of Liner Thickness
3.5.4.4 Effect of Liner Hydraulic Conductivity
3.5.5 Breakthrough Time
3.5.5.1 Effect of Soil Suction Stress
3.5.5.2 Effect of Hydraulic Head
3.5.5.3 Effect of Liner Thickness
3.5.5.4 Effect of Liner Hydraulic Conductivity
3.5.6 Summary
3.6 ANALYSIS OF PERFORMANCE - PARTIALLY SATURATED-2D
3.6.1 Introduction
3.6.2 Overview of Analysis
3.6.2.1 Description of UNSAT2D Program
3.6.2.2 Summary of Study
3.6.3 Initial Leak Detection Time
3.6.4 Leachate Collection Efficiency
3.6.5 Leakage Out of Unit
3.6.6 Breakthrough Time
3.6.7 Summary
3.7 COMPARISON OF RESULTS
3.8 SUMMARY
-------
Chapter 4
PERFORMANCE OF COMPOSITE LINERS
4.1 INTRODUCTION
4.2 FACTORS AFFECTING THE PERFORMANCE OF COMPOSITE LINERS
4.2.1 FML Related Issues
4.2.1.1 Types of Geomembranes
4.2.1.2 FML Performance
4.2.1.2.1 FML Permeation
4.2.1.2.2 Defects
4.2.1.2.3 Damage During Manufacture, Fabrication
or Installation
4.2.1.2.4 Operational Damage
4.2.1.2.5 Conclusions
4.2.2 Composite Liner Performance
4.2.2.1 Effect of Compacted Soil Hydraulic Conductivity
4.2.2.2 E-ffect of Contact Between Soil and FML
4.3 LEAKAGE MECHANISMS THROUGH COMPOSITE LINERS
4.3.1 Introduction
4.3.2 Leakage Due to Permeation Through FML
4.3.3 Frequency and Size of FML Defects
4.3.4 Analytical and Model Studies
4.3.5 Conclusions on Leakage Through Composite Liners
4.4 PERFORMANCE OF COMPOSITE LINERS - 1-D STEADY-STATE ANALYSIS
4.4.1 Introduction
4.4.2 Leakage Through an Intact FML
4.4.2.1 Procedure
4.4.2.2 Leak Detection Sensitivity
4.4.2.3 Leachate Collection Efficiency
4.4.2.4 Leakage Out of Unit
4.4.2.5 Summary
4.4.3 Leakage Through Holes in FML Component of Composite Liner
4.4.3.1 Procedure
4.4.3.2 Leak Detection Sensitivity
4.4.3.3 Leachate Collection Efficiency
4.4.3.4 Leakage Out of Unit
4.4.3.5 Summary
-------
4.4.4 Leakage Through a Typical Composite Liner
4.5 PERFORMANCE OF COMPOSITE LINERS - 2-D TRANSIENT ANALYSIS
4.5.1 Introduction
4.5.2 Overview of Analysis and Results
4.5.3 Leak Detection Sensitivity
4.5.4 Leachate Collection Efficiency
4.5.5 Leakage Out of the Unit
4.6 COMPARISON OF RESULTS OBTAINED FROM ANALYTICAL AND NUMERICAL
MODELS
4.7 SUMMARY AND CONCLUSIONS
-------
Chapter 5
COMPARISON OF COMPACTED SOIL
AND COMPOSITE BOTTOM LINER PERFORMANCE
5.1 INTRODUCTION
5.1.1 Purpose
5.1.2 Organization of this Chapter
5.1.3 Comments on Data
5.1.4 Presentation of Data
5.2 LEAK DETECTION SENSITIVITY
5.2.1 Definition and Importance
5.2.2 Comparison of Compacted Soil and Composite Bottom Liners
5.3 LEACHATE COLLECTION EFFICIENCY
5.3.1 Definition and Importance
5.3.2 Comparison of Compacted Soil and Composite Bottom Liners
5.4 LEAKAGE OUT OF WASTE MANAGEMENT UNIT
5.4.1 Definition and Importance
5.4.2 Comparison of Compacted Soil and Composite Bottom Liners
5.5 SUMMARY OF COMPARISONS
8
-------
Chapter 6
CURRENT PRACTICE
IN LINING SYSTEM DESIGN
6.1 INTRODUCTION
6.2 RESULTS OF EPA SURVEY
6.2.1 Information on the Survey
6.2.1.1 Reason for the Survey
6.2.1.2 Questions 1n the Survey
6.2.1.3 Results of the Survey
6.3 CONCLUSIONS
6.3.1 Owners or Operators Opted for Composite Bottom Liner
6.3.2 Assessment of Adverse Impact
-------
Chapter 7
SUMMARY
7.1 SUMMARY OF COMPARATIVE PERFORMANCE
10
-------
Appendix A
CASE HISTORIES OF COMPACTED SOIL
LINING SYSTEM PERFORMANCE
A.I OVERVIEW OF CASE HISTORIES
A. 2 LANDFILLS
A.2.1 Sanitary Landfill Sarnia, Ontario
A.2.2 Four Compacted Clay Lined Landfills in Wisconsin
A.2.3 Hazardous Waste landfill in Illinois
A.2.4 Field Scale Test Liner
A. 3 SURFACE IMPOUNDMENTS
A.3.1 Three Surface Impoundments in Texas
A.3.1.1 Two Pounds in Central Texas
A.3.1.2 Evaporation Pond in North Texas
A.3.1.3 Brine Ponds in Southern Texas
A.3.1.4 Conclusions
A.3.2 Cooling Pond in Mexico
A.3.3 Two Prototype Compacted Clay Lining Systems
A. 4 SUMMARY
11
-------
Appendix B
ANALYSIS OF LEAKAGE
THROUGH COMPOSITE BOTTOM LINERS
B.I INTRODUCTION
B.I.I Scope
B.I.2 Organization
B.2 LEAKAGE DUE TO PERMEATION THROUGH FML
B.2.1 Permeameter Tests
B.2.2 The Concept of Coefficient of Migration
B.2.3 Water Vapor Transmission Tests
B.2.4 Relationships Between Various Expressions of Flow Rate
B.2.5 Leakage Rate Evaluation
B.2.6 Migration of Chemicals
B.3 FREQUENCY AND SIZE OF FML DEFECTS
B.3.1 Purpose
B.3.2 Data from Construction Quality Assurance
B.3.3 Data from Forensic Analyses
B.3.4 Conclusions on Frequency of Defects
B.3.5 Estimation of Size of Defects
B.3.6 Standard Hold Size and Frequency
B.4 ANALYTICAL STUDIES
B.4.1 Introduction
B.4.1.1 Purpose of the Section
B.4.1.2 Leakage Mechanisms
B.4.1.3 Organization of this Section
B.4.2 Analyses Assuming Perfect Contact
B.4.3 Analyses Assuming Flow Between FML and Soil
B.4.4 Free Flow through Holes in the FML
B.5 LABORATORY MODELS
B.5.1 Introduction
B.5.2 Review of Tests by Brown et al.
B.5.3 Review of Tests by Fjkuoka
B.6 CONCLUSIONS ON LEAKAGE THROUGH COMPOSITE LINERS
B.6.1 Conclusions from Analytical Studies
B.6.2 Conclusion from Model Tests
12
-------
B.6.3 Conclusions for Leakage Rate Evaluation
13
-------
Appendix C
SUMMARY OF TWO DIMENSIONAL PARTIALLY
SATURATED FLOW ANALYSIS RESULTS
[DATA FROM RADIAN, 1987]
C.I DESCRIPTION OF RESULTS
14
-------
CHAPTER 1
INTRODUCTION
-------
1.1 LEGISLATIVE HISTORY AND CURRENT REQUIREMENTS
1.1.1 Legislative and Regulatory History
Under Section 3004 of the Resource Conservation and Recovery Act
(RCRA), owners and operators of treatment, storage, and disposal
facilities (TSDFs) are required to comply with standards "necessary to
protect human health and the environment". Regulations that
established the major components of these standards were issued on May
19, 1980 (45 FR 33221); these were the first national standards that
defined acceptable management practices for hazardous waste. They
established EPA's first phase of requirements under Section 3004 of
RCRA for owners and operators of facilities that treat, store, or
dispose of hazardous wastes. These standards included Part 265
requirements applicable during the interim status period and Part 264
requirements applicable to permits.
On July 26, 1982 (47 FR 32274), EPA promulgated technical and
permitting standards under Part 264 for landfill, waste pile, surface
impoundment, and land treatment units. These regulations consisted of
a set of design and operating standards separately tailored for each
type of unit. The design and operating standards required units
(other than land treatment units) to have a liner and leachate
collection system to prevent migration of wastes to the subsurface
soil or to ground water or surface water during the active life of the
unit.
The Hazardous and Solid Waste Amendments (HSWA) to RCRA became law
November 8, 1984. Under Section 3004(o) and 3015 of these amendments,
certain surface impoundments and landfills must have "two or more
liners and a leachate collection system above (in the case of a
landfill) and between such liners." This is the minimum technological
requirement for new units, replacement units, and lateral expansions
of existing units which are not subject to statutory variances.
1.1.2 Legislative Requirements
The double liner system requirements developed under HSWA are
1-1
-------
intended to satisfy the policy objective stated in RCRA 3001(a)(4) to
"assure that hazardous waste management practices are conducted in a
manner which protects human health and the environment". An interim
minimum double liner requirement was included in HSWA (RCRA
3004(o)(5)(B)) to provide time (30 months) for EPA to develop
regulations and guidance documents on double liner systems. The
interim double liner system requires:
"a top liner designed , operated, and constructed of materials
to prevent the migration of any constituent Into such liner
during the period such facility remains in operation
(including any post-closure monitoring period), and a lower
liner designed, operated and constructed to prevent the
migration of any constituent through such liner during such
period. For the purpose of the preceding sentence, a lower
liner shall be deemed to satisfy such requirement if it is
constructed of at least a 3-foot thick layer of recompacted
clay or other natural material with a permeability of no more
than 1 x 10~7 centimeter per second."
The current regulations on double liner systems codified in the
Code of Federal RegulationsfCFR) promulgated July 15, 1985 (50 FR
28747, 28748) reflects the interim minimum technology standards stated
in RCRA 3004 (o)(5)(B). In language that almost identically tracks the
statute, the regulations (Parts 264 and 265) require new and
replacement units and lateral expansions of units at surface
impoundments and landfills to have a double liner system with a
leachate collection and removal system that protects human health and
the environment.
1.1.2.1 Current Requirements for Top Liner
The top liner of a double liner system required by the current
regulations must be "designed, operated, and constructed of materials
to prevent the migration of any constituent Into such liner during the
period such facility remains in operation (Including any post-closure
monitoring period).
1-2
-------
1.1.2.2 Current Requirements for Bottom Uner
The bottom liner of a double liner system required by the current
regulations must be "designed, operated, and constructed to prevent
the migration of any constituent through such Uner during such
period" (the active life and post-closure care period).
The regulations state that the bottom liner requirement can be
satisfied by a bottom liner that is at least a 3-foot thick layer of
recompacted clay or other natural material with a permeability
(hydraulic conductivity) of no more than 1 x 10 centimeter per
second (cm/s).
1.1.3 Proposed Double Liner Rule of March 28. 1986
On March 28, 1986, EPA proposed double liner and leachate
collection system requirements for landfills and surface impoundments
(51 FR 10706-10723). These proposed regulations are intended to
codify the minimum technology double liner requirements mandated by
the HSWA for landfills and surface impoundments. The March 28, 1986,
proposed rule requires new units, lateral expansions, and replacements
of existing units at landfills and surface impoundments to have two or
more liners and a leachate collection and removal system above (for
landfills) and between the liners. The liner system proposed in the
March 28, 1986 Federal Register consists of:
• A top liner, designed, constructed and operated to prevent
migration of liquids Into 1t; and
• One of two possible bottom liners;
—A bottom liner designed, constructed and operated so that
liquids do not migrate through it. The minimum standard is a
1-m (3.0-ft.) layer of compacted soil with a maximum hydraulic
conductivity of 1 x 10~* m/s (1 x 10~7 cm/s);
1-3
-------
OR
--A composite bottom liner made of two components. The upper
component would be designed, operated, and constructed to
prevent migration of hazardous constituents into it and a lower
component designed, operated, and constructed to minimize
migration of hazardous constituents through it if the upper
component were breached before the post-closure care period
ends. At a minimum, this lower bottom liner component must be
a compacted soil with a maximum hydraulic conductivity of 1 x
10 "' m/s (1 x 10 "7 cm/s).
While the rule proposed for codification on March 28, 1986, does
not provide minimum specifications for the geomembrane (FML)
components of the top and bottom liners, EPA has developed guidance on
minimum specifications for these materials. According to the Draft
Minimum Technology Guidance on Double Liner Systems of May 24, 1985
(EPA 530-SW-85-012):
• The geomembrane top liner should be at least 0.75 mm (30 mil)
thick, if it is protected in a timely manner after placement;
if it is not protected in a timely manner, the top FML should
be at least 1.12 mm (45 mil) thick.
• The upper geomembrane component of a bottom liner should be at
least 0.75 mm (30 mil) thick.
Examples of double liner systems 1n hazardous waste management units
are given in Chapter 2.
1.1.4 Proposed Liner/Leak Detection Rule (Pending)
EPA is currently developing proposed regulations for leak
detection systems intended to satisfy the HSWA statutory requirements
for leak detection systems at landfill, surface impoundment, waste
pile and land treatment units. EPA is planning to require minimum
system performance criteria for leak detection systems as well as
minimum specifications for components of the leak detection system in
1-4
-------
the proposal on leak detection. In addition to the leak detection
regulations, EPA is also planning to extend the minimum technology
double liner requirements to waste piles. It also will establish
construction quality assurance (CQA) requirements for owners or
operators of hazardous waste management units.
1.2 PURPOSE AND SCOPE OF THE BACKGROUND DOCUMENT
1.2.1 Purpose of the Background Document
In its development of the Proposed Double Liner Rule and the
pending Proposed Liner/Leak Detection Rule, and through its ongoing
research and development efforts, EPA has gathered data that indicates
that compacted low-permeability soil bottom liners provide a lower
level of performance capability in double liner systems than composite
bottom liners comprised of a geomembrane upper component and a
compacted low-permeability soil lower component. To announce this
data to the public, EPA plans to Issue a Notice of Data Availability
on Bottom Liners in April 1987. In addition to announcing the data,
the Notice will provide a summary of the data and a discussion of
their significance. The purpose of this technical background document
is to fully document all relevant data relating to the comparative
performance of compacted soil and composite bottom liners.
The purpose of this document 1s to: (1) present data relating to
the comparative performance of compacted low-permeability soil and
composite bottom liners; (2) compare the capabilities of compacted
low-permeability soil and composite bottom liners to satisfy the
statutory goal of RCRA to prevent the migration of hazardous
constituents from the hazardous waste management unit and detect
leakage through the top liner at the earliest practicable time; and,
(3) quantify performance differences between double liner systems with
compacted low-permeability soil and composite bottom liners. To the
extent possible, comparisons will be presented in relation to the
following three criteria that EPA believes relate to protection of
human health and the environment for land disposal units:
1-5
-------
• leak detection capability;
• leachate collection system efficiency; and
• leakage volume Into and out of the bottom liner.
1.2.2 Sources of Available Data
Several sources of data have been reviewed and used during
preparation of this technical background document. These Include:
• case study information on observed leakage rates through
compacted soil liners;
• case study information on the integrity (number of holes) of
Installed geomembrane liners;
• case study Information on leakage rates through large-scale
model composite liners;
• analytical and numerical studies of flow into, through, and out
of compacted low-permeability soil and composite bottom liners;
• analytical and numerical studies of differences in the
performance of "leak detection systems due to different
(compacted low-permeability soil or composite) bottom liners;
and
• responses to an EPA questionnaire on current bottom liner
construction practices at hazardous waste management
facilities.
1.2.3 Scope of the Background Document
This technical background document is divided into seven chapters,
each of which is summarized below. Note that 1n the remainder of this
document, compacted low-permeability soil liners are referred to
simply as compacted soil liners.
1-6
-------
Chapter 1 presents an introduction to the technical background
document. This Introduction includes a description of the Legislative
History leading to the Notice of Data Availability on Bottom Liners.
This introductory chapter also describes the purpose and scope of the
background document.
Chapter 2 provides a brief description of key concepts which are
used in this document. They include such introductory materials as
definitions of lining system, double liner, single liner, and
composite liner; leachate collection and removal systems; functions of
lining system components; and, important characteristics of lining
system components. Chapter 2 also discusses EPA's "liquids management
strategy" and describes the lining system performance criteria
(detection capability, leachate collection system efficiency, and
leakage volume out of the unit) useful in evaluating to what degree
bottom liners meet EPA's goal of preventing migration through the
lining system and out of the unit. This chapter is included as
background to ensure that all readers are familiar with the basic
concepts that are fundamental to understanding the significance of the
data presented herein, as well as to ensure that all readers have a
source of basic information to assist them in interpretation of the
data.
Chapter 3 presents a discussion of the performance of compacted
soil liners. This chapter reviews case studies of leakage through
compacted soil liners. This data can be used to estimate achievable
soil hydraulic conductivities in the field and to estimate the
breakthrough times and flows out of the unit associated with compacted
soil liners. Chapter 3 also presents the results of analytical and
numerical studies of the performance of compacted soil bottom liners.
The analytical investigation is based on the application of Darcy's
law. The numerical investigations make use of the SOILINER and
UNSAT2D computer models. The UNSAT2D model is also used to
investigate overall LDCRS/bottom liner system performance.
Chapter 4 is concerned with the performance of composite bottom
liners. Case study data is presented and then used to draw
conclusions regarding the number and types of holes in properly
1-7
-------
constructed FML (geomembrane) liners. Small-scale laboratory test
results of seepage through composite liners are also reviewed. The
analytical and numerical methods presented in Chapter 3 are also used
in Chapter 4 to investigate the performance of composite bottom
liners. The UNSAT2D model is also used here to- investigate overall
LDCRS/bottom liner system performance.
Chapter 5 summarizes the comparative performance of compacted soil
and composite bottom liners. Comparisons are made by excerpting data
from Chapters 3 and 4, respectively, and then comparing it in terms of
LDCRS detection sensitivity and collection efficiency, as well as the
the cumulative leakage through the bottom liner and time for leakage
to break through the bottom liner.
Chapter 6 summarizes current practices in double liner system
design at hazardous waste management units. This summary is based on
the results of an EPA survey conducted 1n January and February 1987.
Chapter 7 provides a concise summary of the findings presented in
background document.
1-8
-------
CHAPTER 2
DEFINITIONS AND CONCEPTS
RELATED TO DOUBLE LINER SYSTEMS
-------
2.1 INTRODUCTION
2.1.1 Purpose of this_C_hap_ter
The purpose of this chapter is to ensure that all readers are
familiar with the important basic concepts of hazardous waste
management units, waste containment, lining systems, leachate
collection and removal systems, and leak detection systems as well as
the materials used to construct lining systems, leachate collection
and removal systems, and leak detection systems.
The concepts presented in this chapter provide many of the
foundations for EPA's "liquids management strategy" and "systems
approach" to waste containment. This strategy is discussed herein,
and the lining system performance criteria which are influenced by the
bottom liner and which are relevant to the strategy are defined.
2.1.2 Organization of this Chapter
This chapter is comprised of four sections devoted respectively to
hazardous waste management units, lining systems, leakage, and EPA's
liquids management strategy. A brief outline of each section is as
follows:
• Section 2.2 gives a general description of the various
hazardous waste management units such as landfills, surface
impoundments, and waste piles, and discusses ground pollution
mechanisms which may be associated with leakage from these
units.
• Section 2.3 presents the various types of lining systems used
in hazardous waste management units and the materials used to
construct these lining systems. Also, Section 2.3 defines
basic lining system elements such as double liners, composite
liners, leachate collection and removal systems, and leak
detection systems.
2-1
-------
• Section 2.4 discusses the concept of leakage (what is leakage?
what is a leak?), the purpose of leak detection systems (why is
it important to detect leakage? complementarity of leak
detection and leachate collection), and performance
characteristics of leak detection systems.
• Section 2.5 discusses EPA's goal for bottom liners and defines
lining system performance criteria that are relevant to
achieving this goal.
2.2 WASTE MANAGEMENT UNITS
2.2.1 Introduction
2.2.1.1
"Waste management unit" 1s a generic term which 1s used in this
report to describe land disposal units used to treat, store or dispose
of hazardous waste. These units include: landfills, surface
impoundments, and waste piles.
2.2.1.2 Purpose of th1s_Sectlon
It is not possible to discuss leakage without a knowledge of:
• the containment facilities from which leakage 1s taking place;
and
• the lining systems through which leakage is taking place.
The purpose of Section 2.2 1s to briefly describe surface
impoundment, landfill and waste pile units, and to discuss pollution
mechanisms that may be associated with leakage from these units. The
next section (2.3) will be devoted to lining systems used in those
units.
2-2
-------
2.2.2 Description
2.2.2.1 Types of Waste_Management_Units
Three types of waste management units are considered: landfills,
surface impoundments, and waste piles. These three types of units are
illustrated in Figure 2-1 and their usage is as follows:
• landfills are used for permanent disposal of solid waste
(hazardous waste in "hazardous waste landfills" or municipal
waste in "sanitary landfills");
• surface impoundments are used to store liquids (with, possibly,
particles in suspension, which settle progressively) or sludges
(which consolidate progressively); and
• waste piles are used for temporary storage of solid waste.
2.2.2.2 Geometry_of_Waste_Management_UnUs
2.2.2.2.1 Surface Impoundments
The overall shape of surface impoundments is roughly that of an
inverted truncated pyramid with "side slopes" and a "bottom". The
side slopes can be as steep as permitted by geotechnical
considerations and they typically range between 2H/1V and 4H/1V, while
the bottom is nearly horizontal with just the slope (e.g., 2%)
required for the drainage layer if there is a double liner.
2.2.2.2.2 Landfills
The lower part of a landfill has roughly the shape of an inverted
truncated pyramid, like a surface impoundment. This is the part of a
landfill which is lined prior to waste placement. The side slopes of
the bottom part of a landfill can be as steep as permitted by
geotechnical considerations and they typically range between 2H/1V and
4H/1V, while the bottom is nearly horizontal with just the slope
(e.g., 2%) required for the drainage layer(s) that is (are)
incorporated into the lining system.
2-3
-------
The upper part of a landfill includes a cap which is placed on top
of the waste to close the landfill after completion of waste placement
operations. The cap is a lining system used to prevent (or, at least,
minimize) penetration of rain water into the landfill.
Large landfills may be divided into cells which are operated
sequentially.
2.2.2.2.3 Waste Piles
A waste pile can have any shape compatible with waste stability.
The lining system placed under the waste pile is nearly horizontal,
with just the slope (e.g., 27.) required for the drainage layer(s) that
is (are) incorporated into the lining system.
2.2.3 Ground Pollution Mechanism
2.2.3.1 Surface Impoundments
A surface impoundment can cause pollution of soil and ground water
if the hazardous liquid contained in the impoundment leaks through the
lining system and into the ground.
In rare occasions, waves of liquids stored in surface impoundments
have overtopped the crests of the impoundments thereby causing ground-
water pollution.
2.2.3.2 Landfills
The mechanism by which a landfill can cause soil and ground-water
pollution includes two steps:
• first, leachate is generated in the landfill; and
• then, pollution occurs if some leachate migrates through the
lining system into the ground.
2-4
-------
Leachate can be produced by two mechanisms, Intrusion of water
into the waste and generation of leachate within the waste:
• Intrusion of Water in the Waste. The main cause of leachate
production is infiltration of rain water into the waste. The
rain water seeping through the waste becomes progressively
polluted and the resulting polluted liquid is called
"leachate". In exceptional cases, leachate can be produced by
intrusion of ground water into the waste (if the ground water
table rises), or, even more exceptionally, by intrusion of
flood water into the waste.
• Generation of Leachate within the Waste. Leachate can
originate in the waste if liquid is entrapped in the waste
during waste placement. Drums containing liquids are not
allowed in hazardous waste landfills, and the only possibility
for entrapping liquids 1s through moisture in the waste or in
the earth used for the dally covers (i.e., the layers of
compacted earth, placed every day on the waste). Part of the
moisture Included in the waste or the daily covers can be
expelled by consolidation (I.e., decrease in volume of the
waste and the daily covers due to compression caused by the own
weight of the waste and the dally covers).
To prevent pollution of soil and ground water by landfills, all
efforts should be made to prevent production of leachate:
• A low-permeability cap must be placed on the landfill
immediately after completion of waste placement operations to
prevent intrusion of rain water.
• Selection of landfill location and appropriate design should
prevent intrusion of ground water and flood water.
• Waste and daily cover material should not contain excess
liquids.
Since leachate production cannot be totally prevented, especially
during landfill operation (i.e., during waste placement) when rain can
2-5
-------
fall freely on the landfill, a lining system is necessary at the
bottom and on the side slopes of the landfill.
2.2.3.3 Waste_mes
The two-step mechanism by which waste piles can cause soil and
ground-water pollution is similar to the mechanisms related to
landfills which were described 1n Section 2.2.3.2. Waste piles are
temporary storage units and the waste is normally removed after some
time.
2.3 LINING SYSTEMS USED IN WASTE MANAGEMENT UNITS
2.3.1 Introduction
2.3.1.1
From the above discussion it 1s clear that the lining system
placed on the bottom and the side slopes of a waste management unit
has a critical role: the ground 1s polluted as soon as liquid leaks
through the lining system. Therefore it Is essential to have a good
knowledge of lining systems prior to discussing leakage.
2.3.1.2 Scope of this Section
The purpose of this section 1s to provide basic Information on the
types of lining systems used 1n hazardous waste management units, and
on the materials used to construct these lining systems. This section
should familiarize the reader with the vocabulary used to describe
lining systems.
This section will address the following: definition of lining
systems, materials used 1n lining systems, double liners, and
composite liners. (Experience shows that it is not practical to
discuss double liners and composite liners without a knowledge of
materials used to construct lining systems.)
2-6
-------
2.3.1.3 Definition o
The terms "liner" and "lining system" are not synonymous.
A liner is a low-permeability barrier used to impede liquid or gas
flow. Note that "low permeability" is used, and not "impermeable".
If there was such a thing as an impermeable barrier, it would be
possible to prevent leakage, and many of the discussions and
considerations presented in this background document would be
pointless. Although it may be possible that a glass is impermeable to
water, in modern technology there is no material that is impermeable
at the scale of a waste management unit where the area to be lined can
be as large as tens of hectares (dozens of acres).
Since no liner is impermeable, pollution control can only result
from a combination of liners and drainage layers, performing
complementary functions:
• Liners (which are low-permeability barriers) impede the flow of
undesirable (polluted) liquids toward the ground.
• Drainage layers (which have a high permeability) convey the
undesirable flow away from the ground.
Such combination of liners and drainage layers is called a "lining
system".
2.3.2 Materials Used in Lining Systems
2.3.2.1 Introduction
Materials used in lining systems include:
• low-permeability materials to construct the liners;
• high-permeability materials to construct the drainage*layers;
2-7
-------
• transition materials (or Interface materials) acting as filters
or protective layers (I.e., providing filtration or protection)
between various layers of a lining system; and
• reinforcement materials which increase the strength of a lining
system (if required).
These materials are briefly discussed below.
2.3.2.2 L!D5rJta1:er1als
2.3.2.2.1 Introduction
Low-permeability materials used 1n civil engineering to construct
liners include: compacted low-permeability soils, geomembranes,
concrete, and asphaltic concrete. Concrete and asphaltic concrete are
not used in hazardous waste units for the following reasons:
• Concrete liners tend to undergo much cracking and therefore
tend to leak significantly.
• Asphaltic concrete cannot be used because asphalt has a poor
resistance to attack by many chemicals typically found in waste
management units.
Therefore, only low-permeability soils and geomembranes are
discussed in this document.
2.3.2.2.2 Compacted Soils
Compacted low-permeability soils used to construct liners include:
clay, silty clay, clayey sands, and silty sands. If such soils are
not available at the site, it is possible to make a low-permeability
soil by mixing bentonite with sand. Bentonite 1s composed of
extremely small particles of sodium montmor1llon1te. When it is dry,
it becomes a powder which can be put 1n bags, and 1s purchased and
transported lik? cement.
2-8
-------
2.3.2.2.3 Geomembranes
- Definition
Geomembranes are low-permeability membranes used in civil
engineering as fluid barriers. By definition, a membrane is a
material that is thin and flexible.
- Examples
All geomembranes presently used in hazardous waste management
units are synthetic geomembranes. (Asphaltic geomembranes, which are
used for lining water storage facilities, are not used in hazardous
waste units because they do not have adequate resistance to chemical
attack.) Typical examples of geomembranes used in hazardous waste
units include: high density polyethylene (HOPE) geomembranes; linear
low density polyethylene (LLDPE) geomembranes; polyvinyl chloride
(PVC) geomembranes; and chlorosulfonated polyethylene (CSPE)
geomembranes.
- Terminology
The term geomembrane is often used by the engineering community in
place of the term "flexible membrane liner" (FML). EPA is using the
term "flexible membrane Uner" or FML to be consistent with the
terminology used in the past in documents discussing waste management
units. Therefore, for consistency with previous EPA documentation
"flexible membrane liner" or "FML" will be used in the remainder of
this document to describe synthetic membranes used as low-permeability
liners.
2.3.2.3 Drainage Materials
2.3.2.3.1 Introduction
High-permeability materials used to construct drainage layers
include: high-permeability soils, synthetic drainage materials, and
2-9
-------
pipes. High-permeability soils and synthetic drainage materials are
discussed below.
2.3.2.3.2 High-Permeability Soils
High-permeability soils Include a wide variety of sands and
gravels ranging from fine to coarse In size and well-graded to uniform
in gradation. Selection of a high-permeability soil for specific
conditions must consider the following:
• the drainage layer should be able to collect and rapidly remove
liquids entering the leak detection, collection and removal
system as a result of leakage through the top liner;
• the high-permeability soils should not damage FMLs when the
FMLs are directly 1n contact with the soils; and
• the drainage layer should be physically compatible with
transition materials to prevent any potential migration of the
transition materials into the drainage layer which could lead
to clogging.
2.3.2.3.3 Synthetic Drainage Materials
Synthetic drainage materials are made of planar structures which
are thick enough to convey fluids in their plane. Synthetic drainage
materials are usually made from polymers. Typical polymers include
polypropylene, polyester, polyethylene. These polymers are highly
inert to biological and chemical degradation.
Four types of synthetic drainage materials are currently
available. These are thick needlepunched nonwoven geotextiles,
geonets, geomats and corrugated or waffled plates. With the exception
of needlepunched nonwoven geotextiles, these materials can be combined
with geotextile filters to form drainage geocomposites.
2-10
-------
2.3.2.4 Transition Materials
Transition materials include filters or protective layers.
2.3.2.4.1 Filters
Filters are located between the drainage layer and the soil to be
protected. They usually consist of a granular layer or a combination
of granular layers, or a geotextlle. Their function 1s to allow free
flow into the drainage layer and at the same time prevent the
migration of particles of the protected soil into the drainage layer.
2.3.2.4.2 Protective Layers
Protective (cushion) layers are located between the drainage layer
and the FML. Their function is to protect the FML from damage by the
drainage material. Cushion layers usually consist of a sand layer or
a thick needlepunched nonwoven geotextile.
2.3.2.5
Reinforcement materials are typically placed in a soil layer.
Typical functions include reinforcing the lining system on steep
slopes to prevent sliding along the slope, reinforcing slopes to
prevent slope failure, or bridging over cavities, depressions or soft
spots. The materials most frequently used in reinforcement
applications at waste management units are geogrids and geotextiles.
2.3.3 Double Liners
2.3.3.1 10 1™^ 2*120
2.3.3.1.1 Definitions
- Double Liner
A "double liner lining system" simply called a "double liner
system" or a "double liner" is a lining system which includes two
2-11
-------
liners with a leachate collection and removal system between the two
liners.
Clearly, two liners In contact (I.e., without a leachate
collection and removal system between the two liners) do not
constitute a double Uner (they constitute a single Uner, as
discussed below).
- Single Liner
A lining system which includes only one liner is called a "single
liner".
- Composite Liner
A composite liner 1s a Uner comprised of two or more low-
permeability components of different materials in contact with each
other. For example, a FML and a clay layer placed 1n contact with
each other constitute a composite liner (a FML composite liner).
Composite liners do not constitute a double liner because there is no
leachate collection and removal system between the two low-
permeability components.
The purpose of a FML-compacted soil composite liner is to combine
advantages of FMLs and compacted soils. FMLs have a much lower
permeability than compacted soils, but they may have holes through
which large leakage can occur if the FML 1s placed on a pervious
medium and then subjected to a hydraulic head on Us top surface. The
leakage rate through a FML hole Is reduced 1f there 1s compacted low-
permeability soil under the FML.
2.3.3.1.2 Terminology Related to Double Liners
- Terminology Related to the Liners
In this document, the upper liner of a double liner is called "top
liner" and the lower liner is called "bottom liner". We recognize
that this terminology may be confusing since the term "bottom liner"
2-12
-------
may be mistaken for "bottom lining system", i.e., the lining system
located at the bottom of a waste management unit.
"Top liner" is synonymous with "upper liner" or "primary liner".
"Bottom liner" is synonymous with "lower liner" or "secondary
liner".
- Terminology Related to the Leachate Collection and Removal Systems
In all waste management units lined with a double liner there is a
pervious layer between the two liners. This layer is called the
"leachate collection and removal system (LCRS) between the liners".
If this system is also used as a leak detection system (LDS), its name
becomes "leak detection, collection, and removal system" (LDCRS).
While 1n surface Impoundments there 1s only one pervious layer
(i.e., the LDCRS mentioned above), there are two pervious layers in
landfills: the LDCRS and the layer located above the top liner and
called the "leachate collection and removal system (LCRS) above the
top liner".
2.3.3.2 Use of Double Liners in Waste Management Units
2.3.3.2.1 Current Regulations
Current EPA regulations (40 CFR Parts 264 and 265) require a
double liner system 1n all new hazardous waste landfill and surface
impoundment units. Furthermore, as discussed in Chapter 1, the two
liners comprising the double liner system should meet the following
requirements:
• "A top liner designed, operated, and constructed of materials
to prevent the migration of any constituent into such liner
during the period such facility remains in operation (including
any post-closure monitoring period)".
2-13
-------
• "A bottom liner designed, operated and constructed to prevent
the migration of any constituent through such liner during such
period. For the purpose of the preceding sentence, a lower
liner shall be deemed to satisfy such requirement if it is
constructed of at least a 3-foot thick layer of recompacted
clay or other natural material with a permeability of no more
than 1 x 10~7 centimeter per second".
According to the Draft Minimum Technology Guidance on Double Liner
Systems of May 24, 1985 [USEPA, 1985]:
• The top liner FML should be at least 0.75 mm (30 mil) thick, if
it is protected in a timely manner after placement; if it is
not protected 1n a timely manner the top liner FML should be at
least 1.15 mm (45 mil) thick.
• The upper FML component of a bottom composite Uner should be
at least 0.75 mm (30 mil) thick.
2.3.3.2.2 Examples of Uses of Double Liners in Waste Management Units
- Types of Double Liners Used in Waste Management Units
From the above discussion, it appears that four types of double
liners are currently permitted by existing EPA regulations (Figure
2-2). Such double liners can be used for landfills, surface
impoundments, and waste piles. The double Uner using two composite
liners (Figure 2-2(b)), called "double composite liner", is
increasingly used In order to minimize the amount of leakage through
the top liner while maximizing the collection efficiency of the LDCRS.
- Caution on the Use of Top Composite Liners in Surface Impoundments
The use of a top composite Uner in a surface Impoundment requires
special caution.- If the FML (which 1s the upper component of the top
composite liner) Is not covered with a heavy material (such as a layer
of earth, or concrete slabs), and 1f there 1s leakage through the FML,
liquids tend to accumulate between the low-permeability soil (which is
2-14
-------
the lower component of the top composite liner) and the FML since the
submerged portion of the FML (whose specific gravity is close to 1) is
easily uplifted. Then, if the impoundment is rapidly emptied, the FML
is subjected to severe tensile stresses because the pressure of the
entrapped liquids is no longer balanced by the pressure of the
impounded liquid. Therefore, a top composite liner should always be
loaded, which is automatically the case in a landfill or in a waste
pile, and which must be taken into account in the design of a liquid
impoundment.
2.3.3.2.3 Influence of Liner on Leak Detection
The LDCRS between the top and bottom liner is also used as a leak
detection system to form a leak detection, collection, and removal
system (LDCRS). The leakage that is collected has migrated through
the top liner and flows, in the LDCRS, over the top surface of the
bottom liner. It appears that the two liners have the following
influence:
• The top liner governs the amount of leakage entering the LDCRS.
Many hazardous waste management units include a top composite
liner in order to minimize leakage through the top liner.
• The bottom liner has a major influence on the performance of
the LDCRS. As will be shown in Chapter 5, a compacted soil
liner allows greater leakage into and through the bottom liner
than does a composite. For this reason, a composite (Figure 2-
3 (a)) is preferable to compacted soil (Figure 2-3 (b)).
As will be shown in Chapter 6, owners and operators of hazardous
waste management units rarely use compacted soil bottom liners
(Figure 2-3 (a)) because of the performance deficiencies associated
with them in comparison to composite bottom liners (Figure 2.3 (b)).
2-15
-------
2.4 LEAKAGE DEFINITION AND DETECTION
2.4.1 Definitions
2.4.1.1 L§?k_30
-------
(Note: 1 hectare = 100 m x 100 m = 10 000 m2.)
The following conversions apply:
1 gall on/acre/day = 1.08 x 10~11 m/s
9.35 liters/hectare/day
- 0.935 liters/1000 m2/day
1 liter/hectare/day » 1.16 x 10~l* m/s
0.11 gallons/acre/day
= 0.1 liters/1000 mVday
1 liter/lOOOmVday = 1.16 x 10"11 m/s
1.1 gallon/acre/day
- 10 liters/hectare/day
1 m/s - 8.64 x 1010 Uters/lOOOmVday
- 8.64 x 1011 liters/hectare/day
- 9.24 x 1010 gallons/acre/day
From a practical standpoint, the approximate conversion can be
used:
1 liter/lOOOmVday = 1 gallon/acre/day
1 Ltd = 1 gpad
2.4.1.3 Leakage Collected and Leakage Out of the Unit
The possible fates of liquids entering a double liner system are
shown in Figure 2-4.
The leakage discussed 1n the previous sections is the leakage that
the LDCRS system is intended to collect and detect. This is the
leakage through the top liner (C in Figure 2-4).
The leakage out of the unit, which is the leakage through the
bottom liner (J in Figure 2-4), is only a fraction of the leakage
through the top liner. Other fractions include:
2-17
-------
• leakage entrapped in the LDCRS by absorption, capillarity,
ponding, etc. (F in Figure 2-4);
• leakage collected at the LDCRS sump (G in Figure 2-4); and
• leakage absorbed in the bottom liner (I in Figure 2-4).
If the LDCRS is properly designed, the liquid head on the bottom
liner is very small, and leakage through the bottom liner (which is
governed by head on the bottom liner) is very small. This is
consistent with the EPA's goal of protecting human health and
environment through system impermeability and not liner
impermeability. No liner is perfectly impermeable but proper design
can almost achieve system impermeability.
2.4.2 Leak Detection System
2.4.2.1 Definition
In the context of this background document, leak detection refers
to leakage through the top liner and, therefore, a leak detection
system is a system which Is placed between the two liners of a double
liner system to monitor the leakage through the top liner.
2.4.2.2 Pu rpose_of_Leak_Detec tton
As Indicated 1n the definition given in Section 2.4.2.1, the
purpose of a leak detection system 1s to monitor leakage through the
top liner. Monitoring leakage through the top liner is an important
component of EPA's systems approach to the containment of hazardous
constituents using double liner systems.
2-18
-------
2.4.2.3 Performance Characteristics of_Leak_Detect1on Systems
The important performance characteristics of leak detection
systems include:
• the leak detection sensitivity, which is the smallest leakage
rate that can be detected by the considered leak detection
system;
• the detection time (i.e., the time necessary to detect a leak),
which is a function of the leakage rate;
• the leachate collection efficiency, which is the ratio between
the leakage that is collected at the sump of the LDCRS and the
leakage that actually passes through the top liner into the
LDCRS.
2.5 EPA LIQUIDS MANAGEMENT STRATEGY
2.5.1 Introduction
The previous sections of this chapter have described waste
management units, lining systems, and leakage through lining systems.
This section of Chapter 2 discusses EPA's "liquids management
strategy" for land disposal units. From an understanding of this
strategy, key lining systems performance criteria are identified which
will be used in subsequent sections to compare the performance of
compacted soil and composite bottom liners and establish the best
demonstrated available technology (BOAT) for double liner systems.
2.5.2 EPA Liquids Management Strategy
The fundamental goal of EPA's hazardous waste management
regulations is the protection of human health and the environment. To
fully understand the relationship of this document to the hazardous
waste land disposal regulatory program promulgated on July 26, 1982,
the "liquids management strategy" must be considered.
2-19
-------
Since the onset of the hazardous waste land disposal program,
EPA's strategy for protecting human health and the environment has
been to set a "no migration" lining system goal for land disposal
units. Congress perpetuated this performance goal in Section
3004(o)(5)(6) of the 1984 RCRA amendments by providing an interim
design that uses a top liner "designed, operated and constructed to
prevent migration of any constituent Into 1t "and a bottom liner"
designed, operated and constructed to prevent the migration of any
constituent through such liner." EPA's Proposed Double Liner Rule of
March 28, 1986, maintains EPA's goal of preventing migration of
constituents out of the hazardous waste management unit.
While the EPA's position has been and continues to be to prevent
hazardous constituent migration out of the unit, it recognizes that
the "no migration" goal is not always achievable. However, through
the EPA's "liquids management strategy" and through the use of BOAT
for double liner systems, 1t 1s believed that waste management units
with double liner systems can come very close to the "no migration"
goal (see discussion in Section 2.4.1.3 on leakage out of the unit).
EPA's liquids management strategy has two main objectives: (i)
minimize leachate generation in the waste management unit (which was
discussed in Section 2.2.3.2); and (11) maximize leachate removal from
the waste management unit at the earliest practical time. It is
through these two operational objectives that EPA will achieve the
Congressional goal of preventing migration of hazardous constituents
out of the unit.
This background document applies to the second part of the
"liquids management strategy", namely, maximizing leachate removal
from the waste management unit. The double Uner system is the
mechanism by which leachate collection and removal can be maximized.
The top and bottom liner together with the LCRS above the top liner
(in the case of landfills) and the LDCRS between the liners function
in an integrated, Interdependent manner to prevent leachate migration
out of the unit by maximizing its collection and removal. Each of the
system elements reinforces and supports the other: the liners serve as
2-20
-------
a barrier to leachate migration and facilitate its collection and
removal; the leachate collection and removal system (LCRS) above the
top liner in landfills enables collection and removal of leachate and
minimizes the buildup of the liquid pressure on the top liner; the
leachate collection and removal system between the liners serves to
minimize the buildup of head on the bottom liner; and the leak
detection system provides the owner or operator and EPA with
notification of leakage through the top liner, which enables the
review of existing conditions and may lead to the taking of certain
response activities.
In this integrated system, the bottom liner serves several
functions. These include:
• maximizing the detection capability of the leak detection
system to enable leak detection at the earliest practicable
time (RCRA 3004(o)(4)(A)); detection sensitivity, defined
subsequently, is a key detection capability performance
criterion;
• maximizing leachate collection and removal in the LDCRS; this
is achieved by having a LDRCR/bottom liner system with as high
a collection efficiency as possible; the key performance
criterion here is leachate collection system efficiency; and
• minimizing the migration of leakage into and through the bottom
liner; this is achieved by minimizing hydraulic head on the
bottom liner (which is accounted for by having an LDCRS with a
sufficiently permeable drainage media) and, for a given
hydraulic head, by choosing a bottom liner which prevents to
the extent technically feasible migration of hazardous
constituents out of the unit.
2.5.3 Performance Criteria for Evaluation of Bottom Liners
In Section 2.5.2 several performance criteria for evaluation of
the comparative performance of bottom liners were identified based on
EPA's liquids management strategy. These criteria are defined below.
2-21
-------
For clarity an Illustration of a compacted soil bottom liner with some
typical dimensions is given in Figure 2-3 (a). An illustration of a
composite bottom liner with some typical dimensions is given in Figure
2-3(b).
2.5.3.1 Leak_Detect1on_Sens1t1v1 ty
The detection sensitivity is the smallest leakage rate through the
top liner (E In Figure 2-4) that can be detected in the LDCRS sump
within a reasonable amount of time. The hydraulic conductivity of the
bottom liner (or top component of a composite bottom liner) is the
variable which most influences detection sensitivity. The smaller the
hydraulic conductivity, the better the leak detection sensitivity.
2.5.3.2 Leachate Collect1on_EffU1ency
The leachate collection efficiency 1s the ratio of the leakage
collected at the LDCRS sump (G 1n Figure 2-4) divided by the.leakage
entering the LDCRS (E 1n Figure 2-4). There are two measures of
collection efficiency: (1) cumulative collection efficiency measured
from the time of unit start-up to any other point 1n time; and (ii)
steady-state collection efficiency at any point in time after the
LDCRS has "wetted up". The two factors that most Influence collection
efficiency are the hydraulic conductivity of the bottom liner and the
capillary tension in the LDCRS drainage media (the latter factor
affects the cumulative collection efficiency only).
2.5.3.3 Leakageju t_of _ the_U n 11
Leakage out of the unit refers to leakage that passes through the
bottom liner into the ground (J in Figure 2-4). The factor which most
influences leakage out of the unit is the hydraulic conductivity of
the bottom Uner.
2.5.3.4
While breakthro'ugh times are not considered to be critical
performance criteria within the context of EPA's "liquids management
2-22
-------
strategy", they do provide useful information on compacted soil liner
behavior and are included herein for completeness. Breakthrough time
refers to the time from when leakage first enters the LDCRS (E in
Figure 2-4) until the time it first passes through the bottom liner
and enters the ground (J in Figure 2-4). The two factors which most
influence breakthrough time are the hydraulic conductivity and
thickness of the compacted soil bottom liner.
2-23
-------
, uia&le.
(a) Landtll
imrva, SMS
0 I
(t>) Suriaca. Ihiboundment"
V
g Sust
(c) Waste. Rle.
Figure 2-1. Waste management units.
Waste.
lininq. SuSletn
-------
Top L/ne< ; LDCRS '
So+"torv\
Ca)
Linear Cor^pcicted Soil L- FM L
, LDC'g-S
. 4
(
Liner- - , . . 4
( | > 0.«U (3
Liner soi > 0.
i I t
Top Line*- Compaei-ed Sot I
Z)
Linav
FML
Cowpacf(fd Soi, j >0.qm(3ft) L< 10-V/s
'
Figure 2-2. Four types of double liner systems used in hazardous
waste management units.
2-25
-------
A. COMPACTED SOIL BOTTOM LINER
LOCKS
f.'lter
H
C
Com packed
////// J/J JJJ// M /////////
Sub grade
6. COMPOSITE BOTTOM
V
LDCRS
H
X
Ce?mpacfed
so
T= 1.0 4o 2.5
(40 fp lOOrvi.l]
(iypica 1 )
/IT///// I 111111111111 /11
Subg«-ade soil
Figure 2-3.
Illustrations and typical properties of bottom
liners: (a) compacted soil; and (b) composite.
2-26
-------
A = leachate collected in the LCRS sump
Leachate Collection and
Removal System (LCRS)
B = leachate stored* in LCRS
C = leachate from the LCRS into top liner
Top Liner
D = leachate stored* in top liner
E = leakage from the top liner into the
LDCRS
Leak Detection
Collection and
Removal System
(LDCRS)
G = leakage collected in the LDCRS sump
F = leakage stored* in LDCRS
H = leakage from the LDCRS into the
bottom liner
Bottom Liner
I - leakage stored* in the bottom liner
Ground
J - leakage from the bottom liner
into the ground
* Stored liquids due to capillarity, absorption, etc.
Figure 2-4,
Fate of liquids entering a double liner system at a
landfill unit.
2-27
-------
CHAPTER 3
PERFORMANCE OF COMPACTED SOIL LINERS
-------
3.1 INTRODUCTION
This chapter addresses the performance of compacted low-
permeability soil liners. The performance of composite liners will be
addressed in Chapter 4.
The performance of compacted soil liners can be affected by a
variety of parameters. The purpose of this section 1s to review these
parameters and study how they affect soil liner performance. In
Section 3.2, the factors affecting compacted soil liner performance
are described. In Section 3.3, a literature review of case histories
documenting compacted soil liner performance is summarized. A summary
of each case history is presented in Appendix A. In Section 3.4, a
one-dimensional saturated flow analysis is made. In Section 3.5, a
one-dimensional partially saturated flow analysis is presented. In
Section 3.6, a two-dimensional partially saturated flow analysis is
presented. In Section 3.7, a comparison of these three analyses Is
made. In Section 3.8, the information presented In the chapter is
summarized and conclusions are drawn.
Compacted soil lining system performance can be evaluated in terms
of four criteria:
• leak detection capability (defined by leak detection
sensitivity);
• leachate collection system efficiency;
• leakage into the bottom liner and out of the unit; and
• breakthrough time.
The analyses presented in Sections 3.4, 3.5, and 3.6 evaluate the
effect of compacted soil bottom liner properties on lining system
performance, using the above four criteria.
3-1
-------
3.2 FACTORS AFFECTING COMPACTED SOIL LINER PERFORMANCE
3.2.1 Nature of Compacted Soils
In order to discuss the factors affecting compacted soil liners it
is first necessary to understand the nature of compacted soil.
Compaction is the densification of soil through the application of
mechanical energy. The strength, ductility, permeability, and
structure of a compacted soil will be affected by the method of
compaction, level of compactlve effort, and water content at which the
soil is compacted. The dry density and water content of a soil are
used to evaluate the degree to which the soil has been compacted. The
typical variation of dry density as a function of water content is
shown in Figure 3-1. It can also be observed that, for a given level
of compactive effort, the dry density of the soil will first increase
with increasing water content, and then decrease. The water content
corresponding to the apex of the curve is called the optimum water
content, while the dry density corresponding to the apex 1s called the
maximum dry density. Not shown in Figure 1 is the fact that, at a
given water content, the dry density of the soil increases with
increasing compactive effort.
Compaction curves, such as the one shown in Figure 3-1, are
usually obtained by performing Standard Proctor or Modified Proctor
laboratory tests [ASTM D698 and ASTM D1557, respectively]. These
tests are used to determine the maximum dry density attainable, for a
given level of compactive effort, 1n a standard laboratory mold. The
standard and modified test are carried out 1n essentially the same
manner, except that 1n the modified test the soil is compacted into
the mold in thinner lifts (I.e., 5 lifts Instead of 3) and higher
compaction energy is used. The test results are used to develop
specifications for the compaction water content and compactive effort
in the field. They are also used as part of the Construction Quality
Assurance (CQA) program to determine 1f an adequate degree of field
compaction has been achieved. This Is the same procedure used to
verify the construction procedure of earthen fills. The only
difference between a traditional earthen fill and a compacted soil
Uner is that different water contents and compactive efforts may be
3-2
-------
specified so as to minimize permeability (liners) rather than maximize
strength (earthen fills). Compacted soil liners will usually be
compacted at higher water contents (at water contents wet of optimum)
than earthen fills.
3.2.2 Hydraulic Conductivity
Permeability 1s a critical parameter which describes the rate and
volume of flow through a compacted soil liner. The degree of
permeability of a soil 1s often expressed in terms of its hydraulic
conductivity (or coefficient of permeability), k (m/s):
k = V/(Ait) (Equation 3-1)
which is the volume, V (m1), of fluid passing per unit area, A (m2),
per unit hydraulic gradient, 1 (m/m), per unit period of time, t (s).
The hydraulic conductivities of natural soils vary over many orders of
magnitude. The hydraulic conductivity of a soil compacted to a given
water content and dry density is highly dependent upon its post-
compaction degree of saturation (Sr). The higher the Sr, the higher
the hydraulic conductivity. Typical saturated (Sr « 100%) hydraulic
conductivities for a range of soil types are summarized in Table 3-1.
The degree of saturation of a compacted soil liner is high (e.g.,
above 90% or more). Therefore, the saturated hydraulic conductivity
is usually conservatively used in the design of soil liners.
The saturated hydraulic conductivity of a soil Is affected by a
variety of factors such as: compaction water content and compactive
effort; natural, construction-related, or environmentally-related
secondary structures; and Interactions between the soil and permeating
liquid. The effect of each of these factors will be reviewed below,
separately, even though they can be interrelated.
3.2.2.1 Compact 1[on_Ef f ort_and_Water_Content
The hydraulic conductivity of a compacted soil will depend on the
compaction effort and the water content at which it is compacted. At
a given water content, the hydraulic conductivity of a soil will
3-3
-------
decrease with Increasing compactlve effort. At a given dry density,
the hydraulic conductivity of a soil will be one to two orders of
magnitude higher 1f compacted dry of optimum, than 1f compacted wet of
optimum, as can be seen In Figure 3-1. It is important that dry
density and water content be controlled when trying to achieve a low-
permeability soil, because just a small decrease 1n compaction water
content (from wet of optimum to dry of optimum) may result 1n an order
of magnitude decrease in hydraulic conductivity.
3.2.2.2 Secondary Structures
3.2.2.2.1 Definition of Secondary Structures
A secondary structure [Terzaghi and Peck, 1967] 1s a random or
repeating feature which disrupts the continuity of a soil mass and
thereby affects the performance of the soil mass (e.g., permeability,
strength, compressibility, etc.). The discussion here 1s limited to
the effect of secondary structures on permeability. Examples of
secondary structures are root holes, unremolded soil clods,
desiccation cracks, and construction related secondary structures such
as zones of variable density or water content which result in
preferential fluid pathways through the compacted soil mass.
Secondary structures can result from natural, construction-related, or
environmental factors; each type of secondary structure 1s discussed
below.
3.2.2.2.2 Natural Secondary Structures
Examples of natural secondary structures are root holes, fissures
resulting from large ground deformations, or sllckenslded surfaces.
In general, these structures are formed over time and are commonly
found 1n natural soil formations. They are not expected to exist in
recently compacted soil liners. However, it should be realized that
they can exist 1n the natural soil subgrade upon which the compacted
soil Uner Is placed.
3-4
-------
3.2.2.2.3 Construction Related Secondary Structures
The hydraulic conductivity of a compacted soil mass can be
affected by the presence of secondary structures generated during
construction. Typically, soil Is compacted in 150 mm (6 in) thick
lifts and nonhomogenelty can occur within a given lift and along the
interfaces between lifts. These zones of nonhomogeneity, or secondary
structure, may be reflected in either zones of higher hydraulic
conductivity within the compacted soil mass, or as a soil mass with a
higher hydraulic conductivity in the horizontal direction than in the
vertical direction.
The hydraulic conductivity within a given soil lift can vary by
several orders of magnitude because of such factors as inadequate
compactive effort, variation in soil water content, and variations
in the soil structure. If a good bond is not developed between
adjacent soil lifts, the Interface between lifts can act as a
preferred horizontal flow path, with a hydraulic conductivity higher
than that in the vertical direction. If the zones of high hydraulic
conductivity of a compacted soil mass become connected, seepage can
occur at rates well in excess of the rates predicted from laboratory
hydraulic conductivity tests.
Secondary structures 1n soil liners can also result from soil
clods which are present in the soil prior to Its placement. These
clods are not always remolded during the compaction process. As a
result, the compacted soil can have a structure composed of soil clods
surrounded by reworked soil. The structure of the clods will be the
same as the natural soil deposit from which the soil was taken. The
structure of the natural soil clods which have not been remolded by
the compaction process will present a discontinuity to uniform
hydraulic flow through the liner. The presence of clods (and stones
or cobbles with dimensions that are a large fraction of the lift
thickness) are one of the primary causes of nonhomogeneous compacted
soil masses. It should be realized that clods are broken down in the
laboratory and stones and cobbles are screened from the sample and
therefore the effects of both are not accounted for In permeability
tests on laboratory compacted samples.
3-5
-------
3.2.2.2.4 Environmentally Related Secondary Structure
As reported in the case studies in Appendix A, a primary cause of
secondary structures In compacted soil liners is excessive drying
after placement. A compacted soil can be placed at the correct
moisture content and compacted to the specified dry density. If,
however, it 1s not properly maintained between the time of compaction
and placement of the next lining system component, drying can occur by
moisture evaporation from the soil surface. Drying of the son will
result in desiccation cracks opening up in the soil. This is
especially true 1n soils with very low hydraulic conductivities
because the clay minerals which impart this low hydraulic conductivity
usually have the property of undergoing shrinkage upon drying. It has
been reported that cracks up to 150 to 200 mm deep (6 to 8 in) can
occur within one day of placement 1f the surface 1s not properly
protected [Ghasseml et al., 1983]. The presence of cracks or fissures
will act as channels for the passage of liquids, and as a result, the
effective hydraulic conductivity of the soil mass may be several
orders of magnitude higher than the hydraulic conductivity of intact
soil samples of typical laboratory size.
3.2.2.3 interact
Hydraulic conductivity tests performed in the laboratory normally
use distilled or tap water as the permeating liquid. A number of
studies have evaluated the hydraulic conductivities of soils using
permeants other than water. These permeants are Intended to simulate
the leachate generated 1n waste disposal units [Bowders et al., 1986;
Brown et al., 1983; Fernandez and Qulgley, 1985; Gordon and Forrest,
1981; Pierce and Peel, 1985]. The studies have indicated that some
permeants generate higher hydraulic conductivities in some soils than
those obtained using distilled water. These differences in hydraulic
conductivities may be due to any one of several factors Including
chemically induced soil structural changes and solution or
precipitation of sol Ids. The chemical composition of the permeating
fluid should be considered when designing a waste disposal unit.
3-6
-------
3.2.3 Capillary Stresses
In partially saturated fine grained soils the existing water is
held at the soil particle contacts. A tension force exists on the
surface of this held water, resulting 1n the soil being hygroscopic.
The degree to which the soil 1s able to take on water is often
expressed in terms of the negative pore water stresses which exists in
the soil, which are called capillary stresses (or suction stresses).
The existence of capillary stresses will alter the hydraulic
conductivity of the soil. Capillary stress 1s Inversely proportional
to the degree of saturation, Sr. The larger Sr, the smaller the
capillary stress. The smaller the capillary stress, the higher the
hydraulic conductivity. An example of the variation of hydraulic
conductivity as a function of capillary stress (suction) Is shown in
Figure 3-2.
3.2.4 Settlement
All subgrades will undergo settlement when loaded. The magnitude
of settlement will depend upon the subgrade soil properties and the
magnitude of the applied load. The performance of both compacted soil
liners and leachate collection and removal systems can be affected by
settlements in the supporting subgrade.
3.2.5 Conclusions
A number of factors can Influence the performance of compacted
soil liners, as described In this section. These factors affect
performance primarily through their effect on hydraulic conductivity.
For the purpose of this study, which 1s to compare compacted soil and
composite bottom liners, the most Important factors affecting the
hydraulic conductivity of a compacted soil liner constructed using a
specific soil are:
• compactlve effort
• compaction water content; and
3-7
-------
• secondary structures.
These factors must be considered when determining the expected
hydraulic conductivity of a compacted soil liner. For the purposes of
this background document, it is important to understand that various
factors can affect the hydraulic conductivity of a compacted soil
liner, so that the variation can be accounted for in subsequent
analyses. The degree to which the hydraulic conductivity may vary in
the field can be estimated from a review of case histories of
compacted soil liner performance.
3.3 CASE HISTORIES OF CLAY LINING SYSTEM PERFORMANCE
3.3.1 Overview of Case Histories
Appendix A summarizes the documented performance of compacted soil
lining systems for both landfills and surface impoundments. The
landfill facilities reviewed were primarily for the containment of
sanitary waste, however, one case history described the performance of
a clay lining system at a hazardous waste landfill. Another case
history reports the results from a large scale field test. The
surface impoundment case histories describe facilities which were
constructed to hold fresh water, salt water (brine), or contaminated
liquid. The landfill case histories are summarized in Section A.2 of
Appendix A and the surface impoundments are summarized 1n Section A.3.
3.3.2 Summary of Case Histories
In all cases when field and laboratory data were available, it was
found that the field measured hydraulic conductivity was as much as an
order of magnitude or more higher than the laboratory measured
hydraulic conductivity. The range of values obtained and the probable
cause of the difference between the laboratory and field measured
values are summarized 1n Table 3-2. Only the more conclusive case
histories have been summarized 1n Table 3-2. It can be observed that
several factors can have an effect on the performance of a compacted
soil liner. Most of these factors relate to secondary structures
(nonuniformities) in the compacted soil liner caused by:
3-8
-------
• desiccation cracks;
• soil clods;
• spatial variation in compactive effort and compaction water
content; and
• quality assurance of construction operation.
It should be realized that the above variables, which affect field
performance, are seldom reproduced in standard laboratory tests.
Therefore, achieving the desired hydraulic conductivity in the
laboratory does not guarantee the same value will be obtained in the
field. However, it is possible to achieve a hydraulic conductivity of
10"' m/s (10~7 cm/s) in the field and this 1s the design objective of
the EPA. Achieving this goal requires the proper soils, compaction
-procedures, and construction conditions. Recognizing the number of
factors which can affect the hydraulic conductivity, and that the
design goal' is not always achieved, the subsequent analyses will
consider both a standard hydraulic conductivity of 10"' m/s (10~7
cm/s) and a lower bound hydraulic conductivity of 10"' m/s (10~*
cm/s). This lower value of conductivity might be representative of a
compacted soil liner with some degree of secondary structure resulting
from nonuniform compaction conditions, soil clods, drying or other
factors. These conductivities will be used in the investigation of
leak detection system sensitivity, leachate collection efficiency,
leakage out of the unit, and breakthrough time for compacted soil
liners.
3.4 ANALYSIS OF PERFORMANCE - STEADY STATE SATURATED FLOW (ID)
3.4.1 Introduction
In this section, analyses of compacted soil bottom liner
performance are presented. These analyses are based on steady-state
saturated flow in one dimension. The analysis of lining system
performance can utilize analytical or numerical models having various
degrees of complexity, as Indicated 1n Figure 3-3. The analyses
3-9
-------
presented 1n this section use the simplest of the three models
presented 1n Figure 3-3, that of one-dimensional steady-state
saturated flow. This simple model readily permits the evaluation of a
wide variety of scenarios. The model will be used to evaluate the
effect of several parameters on the leak detection systems
sensitivity, leachate collection efficiency, leakage out of the unit,
and breakthrough time. In Section 3.7, the results obtained here will
be compared with those from the one-dimensional (Section 3.5) and two-
dimensional (Section 3.6) partially saturated flow analyses.
In the following calculations it will be assumed that the
compacted soil Uner 1s in a saturated state and that its
permeability does not change with time. It 1s also assumed that the
compacted soil 1s homogeneous (I.e., It does not possess secondary
structures such as root holes or desiccation cracks).
3.4.2 Overview of Analysis
. Flow through a homogeneous saturated soil can be modeled using
Oarcy's equation:
v - k1 (Equation 3-2)
where v .- apparent fluid velocity (m/s); and k « hydraulic
conductivity (also called coefficient of permeability (m/s)).
The apparent velocity Is the quantity of water that flows 1n a
unit period of time across a unit area, perpendicular to the direction
of flow. It should be noted that water actually flows through the
voids at a higher rate than that given by the apparent velocity. The
actual velocity or seepage velocity, vs, 1s equal to:
vs a v/n (Equation 3-3)
where v » apparent velocity and n is the porosity of the soil. The
apparent velocity should be used when calculating the quantity of flow
through a section, while the seepage velocity should be used to
calculate the time It takes a unit of liquid to flow a given distance.
3-10
-------
Therefore, the seepage velocity is used to calculate breakthrough
time.
If the surface area, A, of the compacted soil liner Is known, then
the volume of flow per unit time can be established from the following
relationship:
Q - vA - k1A (Equation 3-4)
The application of Equations 3-2 and 3-4 can be applied to a
compacted soil Uner, such as that shown 1n Figure 3-4. For the case
shown in Figure 3-4, Darcy's equation can be rewritten as:
(h+H)
v = k (Equation 3-5)
H
where h - hydraulic head acting on the bottom Uner, and H - thickness
of the compacted soil liner. For a given cross-sectional area, A,
the volume of flow per unit time, 0 (m'/s), 1s:
(h+H)
Q - k A (Equation 3-6)
H
where all of the units have been defined above.
From Equations 3-4 and 3-5 it can be observed that for a given area,
A, the performance of the compacted soil liner will be dependent upon
three variables:
• h - hydraulic head;
• H - compacted soil Uner thickness;
• kc - hydraulic conductivity of the compacted soil liner;
The remainder of Section 3.4 investigates the effect of hydraulic
head, compacted soil liner thickness, and compacted soil hydraulic
3-11
-------
conductivity on leak detection system sensitivity, leachate
collection efficiency, leakage out of the unit, and breakthrough
time.
3.4.3 Leak Detection Systems Sensitivity
In general, the leak detection system sensitivity 1s dependent
upon the properties of both the LDCRS and the bottom Uner. In the
event of a concentrated leak through the top liner, a two-dimensional
analysis (and Ideally, a three-dimensional analysis) Is required to
evaluate detection sensitivity. However, 1f uniform leakage through
the top liner 1s considered, 1t 1s possible to establish a lower bound
for detection sensitivity using a one-dimensional analysis.
The minimum top liner leakage rate that can be detected must be
greater than the rate at which liquid will flow, due to gravity, Into
the bottom liner, with the hydraulic head, h, just equal to zero.
Under this condition, the minimum leakage rate will be Independent of
the hydraulic head on the liner (1t Is zero) or the thickness of the
liner (the hydraulic gradient 1s one). Therefore, the minimum
detectable leakage rates for hydraulic conductivities of 10"* and 10~*
m/s (1Q~* and 10~7 cm/s) are:
Hydraulic Leakage Rate
Conductivity liters/1000m*/day
m/s (cm/s) or (gpad)
10"' (10-«) 860
ID"' (ID'') 86
It should be noted that these rates are the theoretical minimum
detectable leakage rates for uniform top liner leakage throughout the
waste management unit. The actual minimum detectable leakage rate is
site-specific and will depend on many factors (e.g., type of leak,
location of leak, effective hydraulic conductivity of bottom liner,
and design of the LDCRS).
3-12
-------
3.4.4 Leachate Collection Efficiency
In general, the steady-state leachate collection efficiency is
dependent upon the properties of both the LDCRS and the bottom liner.
However, 1t 1s possible to evaluate the collection efficiency using
one-dimensional saturated flow by making two simplifying assumptions.
The first assumption is that leakage through the top liner is uniform.
The second assumption is that a head of 0.03 m (0.1 ft) acts on the
bottom liner, Irrespective of the rate of leakage through the top
liner. This Is believed to be a conservative assumption because the
hydraulic head on the bottom liner should normally be small.
The calculated steady-state collection efficiencies (7.) for a
range of top Uner leakage rates and bottom liner hydraulic
conductivities of 10"' m/s (10~7 cm/s) and 10~* m/s (10~* cm/s) are:
Leakage Rate Collection Efficiency %
Through Top FML
liters/1000m2/day kc=10~' m/s 10"' m/s
or gpad (10~r cm/s) (10~* cm/s)
10 00
100 11% 0
1000 91% 11%
The above results Indicate that the leakage rate through the top
liner must be approximately 1000 Ltd (gpad) for a collection
efficiency greater than 90%, 1f the bottom liner hydraulic
conductivity 1s 10~' m/s (10~7 cm/s). The leakage rate must be even
larger to get a comparable efficiency when the hydraulic conductivity
is 10"' m/s (10~' cm/s).
3.4.5 Leakage Out of the Unit
3.4.5.1 Hydraulic Head on Bottom Liner
The variation of leakage out of the unit as a function of the
hydraulic head acting on the Uner is studied for the case of a 1 m (3
3-13
-------
ft) thick compacted soil liner with a hydraulic conductivity of 10~f
m/s (10~7 cm/s). For these calculations, a range of hydraulic heads
from 0.03 to 3.0 m (0.1 to 10 ft) was conservatively assumed. For
this range of hydraulic heads the calculated leakage out of the unit
Is:
Leakage Out
Hydraulic of Unit
Head Uters/lOOOmVday
m (ft) or (gpad)
0.03
0.06
0.3
3.0
(0.1)
(0.2)
(1.0)
(10)
89
92
112
344
From these results 1t can be seen that the steady-state leakage out of
the unit Is not greatly Influenced by the hydraulic heads acting on
the bottom liner as long as the hydraulic head 1s about 0.3 m (1 ft)
or less. Further for these hydraulic heads, 1t Is observed that In
terms of orders of magnitude, the steady-state seepage through a
bottom liner 1s approximately equal to the detection sensitivity
associated with the bottom Uner. The result for a hydraulic head of
3.0 m (10 ft) shows about 3 times more steady-state leakage out of the
unit than for the smaller hydraulic heads. This case represents the
"upper bound" of hydraulic head on a bottom Uner for a surface
Impoundment that has undergone catastrophic failure of the top liner
and LDCRS.
3.4.5.2 Liner Thickness
The effect of Uner thickness on leakage out of the unit was
Investigated for a compacted soil bottom Uner with a hydraulic
conductivity of 10"' m/s (10~7 cm/s) subjected to 0.03 m (0.1 ft) of
head. The leakage out of the unit was calculated for Uner thicknesses
varying from 1 to 3 in (3 to 10 ft). The leakages out of the unit for
this range of thicknesses are:
3-14
-------
Leakage Out
Liner of the Unit
Thickness 11ters/1000m2/day
m (ft) or (gpad)
1 (3) 89
2 (6) 87
3 (10) 87
It can be seen that increasing the bottom liner thickness win not
significantly reduce the leakage out of the unit.
3.4.5.3 Liner Hydraulic Conductivity
The influence of compacted soil liner hydraulic conductivity on
steady-state leakage out of the unit was investigated for the case of
a 1 m (3 ft) thick soil Uner subjected to a 0.03 m (0.1 ft) hydraulic
head, and hydraulic conductivities of 10~' and 10~§ m/s (10~7 and
10"' cm/s). The calculated leakages for these hydraulic
conductivities are 89 Ltd (gpad) and 890 Ltd (gpad), respectively. It
can be observed (as expected) that the leakage out of the unit is
directly proportional to the hydraulic conductivity.
3.4.6 Breakthrough Time
3.4.6.1 Hydraulic Head on Bottom Liner
The effect of hydraulic head on breakthrough time for a 1 m (3
ft.) thick compacted soil liner with a hydraulic conductivity of 10~*
m/s (10~7 cm/s) is analyzed. Since breakthrough time Is a transient
phenomenon, It will be Interesting to compare the results of this
calculation with subsequent calculations which assume partially
saturated flow. For these calculations, the compacted soil bottom
liner is assumed to have a porosity of 0.5. Hydraulic heads ranging
from 0.01 m (0.03 ft) to 0.3 m (1.0 ft) are considered. The
breakthrough times for the assumed range of hydraulic heads are:
3-15
-------
Hydraulic Breakthrough
Head Time
m (ft) (years)
0.01 (0.03) 16
0.06 (0.2) 15
0.3 (1) 12
The breakthrough time 1s not greatly Influenced by the hydraulic head
acting on the bottom lining system, as long as the head remains within
the assumed range.
A hydraulic head of 0.3 m (1 ft.) 1s conservatively used in the
subsequent discussions on the effect of liner thickness and liner
hydraulic conductivity.
3.4.6.2 Liner TMckness
The Influence of compacted soil liner thickness on the theoretical
breakthrough time 1s illustrated in Figure 3-5, which shows the
variation of breakthrough time as a function of thickness, for three
different soil hydraulic conductivities. These breakthrough times
were calculated considering a soil porosity, n, equal to 0.5. The
discussion here will be limited to only one hydraulic conductivity, k=
10"' m/s (10~7 cm/s). For the curve Indicated in.Figure 3-5, it can
be observed that, for a compacted soil liner less than 1 m (3 ft.)
thick, the breakthrough time 1s very sensitive to liner thickness.
The calculated breakthrough time for a range of thicknesses (and k =
10"' m/s (10~7 cm/s)) are summarized as follows:
Liner Breakthrough
Thickness Time
m (ft) (years)
0.5 (1.5) 5
1 (3) . 13
2 (6) 28
3-16
-------
If the desired design goal 1s considered to be no breakthrough
during at least the active life and post-closure care period (about
30 years) of the unit, then, from the above results (based on one-
dimensional steady state saturated flow) it 1s likely that a 1 m (3
ft.) compacted soil liner with an average hydraulic conductivity of
10"' m/s (10~7 cm/s) will not meet the design goal based on saturated
flow. For such a design goal, a compacted soil liner will be required
with a thickness greater than 2 m (6 ft.), and with kc = 1 x 10 "' m/s
(1 x 10~7 cm/s) or better in all areas to be lined (I.e., minimal
secondary structures).
3.4.6.3 Liner Hydraulic Conductivity
The influence of compacted soil liner hydraulic conductivity on
breakthrough time is Illustrated in Figure 3-5, which shows the
variation of breakthrough time as a function of thickness, for three
different soil hydraulic conductivities. It can be observed that the
three curves have the same shape and that, for a given liner
thickness, changing the hydraulic conductivity by an order of
magnitude will change the breakthrough time by about an order of
magnitude.
It was shown in Section 3.3 that the hydraulic conductivity of a
compacted soil liner can vary by up to several orders of magnitude and
that the field value depends on many factors. Thus, the expected
value of hydraulic conductivity must be carefully evaluated if a
reasonable evaluation of breakthrough time (as well as other
performance criteria) 1s to be made.
3.4.7 Summary
The effect of hydraulic head, Uner thickness, and hydraulic
conductivity on compacted soil bottom Uner performance was evaluated
using a model based on one-dimensional steady-state saturated flow.
The results presented 1n this section, along with additional results,
are summarized 1n Table 3-3. It can be observed from the results that
the hydraulic head does not significantly affect detection
sensitivity, leakage out of the unit, or breakthrough time.
3-17
-------
Increasing the liner thickness will Increase the breakthrough time,
but will not influence the detection sensitivity or leakage out of the
unit. The hydraulic conductivity has the most effect on bottom liner
performance. An order of magnitude decrease 1n the hydraulic
conductivity will decrease the breakthrough time and increase the
leakage out of the unit proportionally. Possibly the most significant
observation 1s that with compacted soil bottom liners leakage out of
the unit will be large (1f there is leakage through the top liner),
and collection efficiencies will be low, even in units meeting current
EPA design requirements (kc - 1 x 10"' m/s (1 x 10"7 cm/s) and H - 1 m
(3 ft)).
3.5 ANALYSIS OF PERFORMANCE - PARTIALLY SATURATED FLOW (ID)
3.5.1 Introduction
In Section 3.4 the performance of a saturated compacted soil liner
was analyzed. However, undermost field conditions, compacted soil
liners will be In a partially saturated state. As was discussed in
Section 3.2.3, capillary stresses exist 1n a partially saturated soil
which will influence the liner performance. In Section 3.5 the effect
of partial saturation on one-dimensional flow through a compacted soil
liner will be Investigated.
3.5.2 Overview of Analysis
3.5.2.1 Ascription of Model
The analysis of one-dimensional flow through a partially saturated
soil mass is performed using the SOILINER computer model [GCA, 1986].
The program uses the finite difference technique to solve nonlinear
equations describing unsaturated one-dimensional flow. The SOILINER
program 1s capable of simulating multllayered systems, variable
initial moisture content, and changing boundary conditions. The
program user can specify an Initial suction stress distribution 1n the
compacted soil liner and natural soil, corresponding to the desired
partially saturated state. The computer program will then
incrementally alter the suction stress distribution to correspond to
3-18
-------
the change 1n soil moisture content that will occur as fluid flows
through the soil Hner. The relationship between moisture content and
soil suction stress 1s based on characteristic moisture curves,
Included 1n the computer program. Curves for 12 different soil types
are available. The computer program also Incrementally changes the
hydraulic conductivity to correspond to the moisture content
distribution which exists at a given point in time. A complete
description of the mathematical model and soil characteristic curves
can be found in the user's manual [GCA, 1986] and will not be repeated
here.
3.5.2.2 Summary_of Analyses Performed
The primary purpose of the analyses performed was to study the
effect of soil suction in conjunction with the three parameters
studied in Section 3.4 (hydraulic head, hydraulic conductivity, and
compacted soil Hner thickness) on compacted soil bottom Hner
performance (FMLs cannot be adequately modeled using SOILINER). The
simple lining system shown in Figure 3-6 was analyzed, 1n order to
simplify interpretation of the results, and to allow a comparison with
the data presented 1n Section 3.4. The comparison of the results are
presented in Section 3.7. The parameters, which were varied, included
initial soil suction stress, \|>; hydraulic head acting on the bottom
liner, h; soil liner thickness, H; and hydraulic conductivity of the
compacted soil liner, kc. The effects of varying the above parameters
on detection sensitivity, leakage out of the unit, and breakthrough
time are presented below. It 1s noted that the effect of the natural
soil hydraulic conductivity, kn, and depth of the water table, h,,
were also Investigated. It was found that variations in these
parameters had only a minor effect on leakage out of the unit and
breakthrough time.
3.5.3 Leak Detection Sensitivity
In general, leak detection sensitivity is dependent upon the
properties of both the LDCRS and the bottom liner. For concentrated
leakage, a two-dimensional analysis 1s required to evaluate detection
sensitivity. However, 1f uniform leakage through the top Hner 1s
3-19
-------
considered, 1t 1s possible to use a one-dimensional analysis to
establish a lower bound for the detection sensitivity.
The minimum leakage rate which can be detected must be greater
than the rate at which liquid can flow, due to gravity, Into the
bottom liner. In the case of partially saturated flow, leakage Into
the bottom Hner will occur due to capillary suction which acts 1n
addition to gravity. The capillary stresses will be 1n effect until
saturated conditions are reached, after which point only gravity
forces will drive the flow. As flow occurs through the bottom Hner,
the degree of saturation of the bottom Hner will Increase with time
and, In turn, the magnitude of the capillary stresses will decrease.
Therefore, the Influence of the capillary stresses will decrease with
time. This effect can be observed 1n the following results, which
show the variation of leakage Into the bottom liner as a function of
time until steady-state conditions are reached (Initial suction stress
V - -274 kN/m1 (-40 ps1)):
Leakage Into
Bottom Liner
Time (liters/1000m2/day
(years) or gpad)
0.5 200
1.0 . 156
2.5 120
4.5 105
5.9 (steady state) 97
It should be noted that the above numbers are for a 0.08 m (0.25
ft) head acting on the bottom Hner and are considered to overestimate
the minimum detectable uniform leak by approximately 10%. From the
above, 1t can be observed that the detection sensitivity will change
with time until steady-state conditions are achieved within the bottom
liner.
3-20
-------
3.5.4 Leakage Out of Unit
3.5.4.1 Ef f ect_of _So 1 1 _Suc t 1 on_Stress
The effect of soil suction stress, i|», on leakage out of the unit
was analyzed by varying the Initial suction stress 1n the soil and
holding other variables constant, for the soil liner shown in Figure
3-6. The variables held constant and their respective values are:
hydraulic head on liner, h - 0.08 m (0.25 ft)
soil liner thickness, H - 1 m (3 ft)
soil liner hydraulic conductivity, kc = 1x10"' m/s (lx!0~7 cm/s)
natural soil hydraulic conductivity, kn - IxlO"4 m/s (lxlO~2 cm/s)
depth to water table, h! - 1 m (3 ft)
The soil suction stresses ranged from -1 kN/m2 (-0.14 ps1) to -550
kN/m2 (80 ps1). A value of -1 kN/m2 (-0.14 ps1) Is representative of
an almost completely saturated soil liner, while a value of -549 kN/m2
(-80 psi) Is representative of the suction stresses 1n a soil liner
constructed with a plastic clay with a water content close to the
plastic limit. Obviously, this later value represents an extreme
condition. In all cases, the steady-state leakage out of the unit was
97 Ltd (gpad). The only role of the soil suction stresses are to
change the rate at which this steady-state condition 1s achieved. It
is useful to note that the calculated values for steady-state leakage
out of the unit are very consistent with those calculated using
Darcy's law 1n Section 3.4.5.2.
3.5.4.2
The effect of hydraulic head, h, on the steady-state rate of
leakage out of the unit was analyzed by varying the hydraulic head
acting on the bottom liner over a range from 0.03 m (0.1 ft) to 0.3 m
(1.0 ft). Other variables were held constant as follows:
3-21
-------
initial suction stress - -274 kN/m2 (-40 ps1)
soil .liner thickness, H - 1 m (3 ft)
soil liner hydraulic conductivity, kc - 1x10"' m/s (IxlO"7 cm/s)
natural soil hydraulic conductivity, kn - IxlO"4 m/s (IxlO'2 cm/s)
depth to water table, h, - 1 ra (3 ft)
The calculated steady-state rate of leakage out of unit for the
following hydraulic heads are:
Steady-State Rate
of Leakage
Hydraulic Out of Unit
Head liters/1000m*/day
m (ft) or gpad
0.03 (0.1) 91
0.08 (0.3) 97
0.15 (0.5) 105
0.30 (1.0) 118
It can be observed that for the considered range of hydraulic heads,
the head acting on the bottom liner has little effect on the steady-
state rate of leakage out of the bottom liner.
3.5.4.3 §!!ect_of_L1ner_Th1ckness
The effect of compacted soil liner thickness on the steady-state
rate of leakage out of the unit was analyzed by varying the liner
thickness and holding other variables constant for the soil liner
shown in Figure 3-6. Other variables were held constant as follows:
initial suction stress, ty - -274 kN/m1 (-40 ps1)
hydraulic head, h - 0.08 m (0.25 ft)
soil liner hydraulic conductivity, kc « 1x10"* m/s (IxlO"7 cm/s)
natural soil hydraulic conductivity, kn « IxlO"4 m/s (IxlO'2 cm/s)
depth to water table, h, - 1 m (3 ft)
3-22
-------
The calculated steady-state rate of leakage out of the unit for
the following liner thicknesses are:
Steady-State Rate
of Leakage
Soil Liner Out of Unit
Thickness IHers/lOOOmVday
m (ft) or gpad
0.3 (1) 147
1.0 (3) 97
2.0 (6) 90
3.0 (10) 87
It can be observed that the steady-state rate of leakage out of the
unit 1s not greatly Influenced by the compacted soil Uner thickness
as long as the liner 1s at least 1.0 m (3 ft) thick.
3.5.4.4 Effeet_of_L 1ner_Hydrau11c_Conduct1ylty
The effect of compacted soil Uner hydraulic conductivity on the
steady-state rate of leakage out of the unit was analyzed by varying
the liner hydraulic conductivity and holding other variables constant
for the soil liner shown 1n Figure 3-6. The variables held constant
and their respective values are:
initial suction stress, i|r - -274 kN/m2 (-40 ps1)
hydraulic head, h • 0.08 m (0.25 ft)
soil liner thickness, H • 1 m (3 ft)
natural soil hydraulic conductivity, kn = IxlO"4 m/s (lxlO~2 cm/s)
depth to water table, h, - 1 m (3 ft)
Liner hydraulic conductivities of 10~l°, 10~', and 10"' m/s (10~9,
10"7, and 10~* cm/s) were investigated. The calculated steady-state
rates of leakage out of the unit are:
3-23
-------
Steady-State Rate
of Leakage
Liner Hydraulic Out of Unit
Conductivity liters/1000m2/day
m/s (cm/s) or gpad
10~l° 10"' 10
10"' 10"7 97
10~* 10"' 970
It can be observed that the steady-state rate of leakage out of the
unit 1s highly dependent upon the liner hydraulic conductivity. The
calculated leakage rates are very consistent with those obtained In
Section 3.4.5.3 using Darcy's law.
3.5.5 Breakthrough Time
3.5.5.1 §f!?ct_ofJo11_Suct1on_Stress
The effect of the Initial soil suction stress, i|», on breakthrough
time was analyzed by varying the Initial suction stress in the soil
and holding other variables constant, for the soil liner shown In
Figure 3-6. The variables held constant and their respective values
are:
hydraulic head on liner, h « 0.08 m (0.25.ft)
soil liner thickness, H - 1 m (3 ft)
soil Uner hydraulic conductivity, kc - 1x10"' m/s (IxlO"7 cm/s)
natural soil hydraulic conductivity, kn - lxlO~4 m/s (IxlO"2 cm/s)
depth to water table, h, - 1 m (3 ft)
Soil suction stresses ranging from -1 kN/m2 (-0.14 psi) to -550
kN/m2 (-80 psi) were Investigated. The soil suction stress 1n the
compacted soil liner is taken to be constant with depth, and in the
natural soil it is assumed to have'a minimum (maximum suction) at the
liner Interface and vary linearly with depth to zero at the ground
water table, as shown in Figure 3-6.
3-24
-------
The calculated breakthrough time for the stated range of soil
suction stresses are:
Initial Suction Breakthrough
Stress Time
kN/m2 (ps1) (years)
- 1
- 10
- 69
-137
-274
-549
(- 0.1)
(- 1-4)
(-10.0)
(-20.0)
(-40.0)
(-80.0)
13.2
12.2
11.3
11.1
10.9
10.8
It can be observed from the above results that breakthrough time
decreases with Increasing suction stress. However, the reduction 1s
not large and the breakthrough time 1s almost constant for suction
stresses less than approximately -100 kN/m1 (-14 ps1).
3.5.5.2 §f!?£t_of_Hydraunc_Head
The effect of hydraulic head, h, on breakthrough time was analyzed
by varying the hydraulic head on the bottom liner from 0.03 m (0.1 ft)
to 0.3 m (1 ft). Other variables were held constant as follows:
initial suction stress - -274 kN/ma (-40 ps1)
soil liner thickness, H - 1 m (3 ft)
soil liner hydraulic conductivity, kc - lxlO~f m/s (lxlO~7 cm/s)
natural soil hydraulic conductivity, kn - IxlO"*4 m/s (lxlO~* cm/s)
depth to water table, h, - 1 m (3 ft)
The calculated breakthrough times for the following hydraulic
heads are:
3-25
-------
Hydraulic Breakthrough
Head Time
m (ft) (years)
0.03 (0.1) 12.0
0.08 (0.3) 10.9
0.15 (0.5) 9.9
0.30 (1.0) 8.5
It can be observed that If the hydraulic head acting on the bottom
liner-is decreased from 0.3 m (1 ft) to 0.03 m (0.1 ft), then the
breakthrough time Is decreased by approximately 30 percent.
3.5.5.3 §ffect_of_L1ner_Th1ckness
The effect of liner thickness on breakthrough time was analyzed by
varying the Uner thickness and holding other variables constant, for
the soil liner shown In Figure 3-6. The variables held constant and
their respective values are:
initial suction stress, i|> - -274 kN/m* (-40 psi)
hydraulic head, h - 0.08 m (0.25 ft)
soil liner hydraulic conductivity, kc - 1x10"' m/s (IxlO"7 cm/s)
natural soil hydraulic conductivity, kn « IxlO"4 m/s (IxlO"2 cm/s)
depth to water table, h, - 1 m (3 ft)
The calculated breakthrough times for the following Uner
thicknesses are:
Soil Liner Breakthrough
Thickness Time
m (ft) (years)
0.3 (1) 2
1.0 (3) 11
2.0 (6) 25
3.0 (10) ' 42
3-26
-------
It can be observed that breakthrough time 1s greatly Influenced by the
1iner thickness.
3.5.5.4 §ff§£t_of Liner Hydraulic Conductivity
The effect of Hner hydraulic conductivity on breakthrough time
was analyzed by varying the liner hydraulic conductivity and holding
other variables constant, for the soil Hner shown 1n Figure 3-6. The
variables held constant and their respective values are:
initial suction stress, « - -274 kN/m2 (-40 psi)
hydraulic head, h * 0.08 m (0.25 ft)
soil Hner thickness, H - 1 m (3 ft)
natural soil hydraulic conductivity, kn = IxlO"4 m/s (lxlO~2 cm/s)
depth to water table, t^ - 1 m (3 ft)
Liner hydraulic conductivities of 10"10, 10~», and 10'* m/s (10"',
10"7, and 10~* cm/s) were studied and the calculated breakthrough
times are shown below:
Liner Hydraulic Breakthrough
Conductivity Time
m/s (cm/s) (years)
10"10 10"' 109
10"' 10"7 11
10"' 10"§ 1
It can be observed that the breakthrough time Is highly dependent upon
the liner hydraulic conductivity.
3.5.6 Summary
The effect of hydraulic head, liner thickness, and hydraulic
conductivity on compacted soil bottom liner performance was evaluated
using a one-dimensional partially saturated flow model. The study
showed that partial saturation had only a small effect on the
performance of compacted soil liners. This effect is largest when the
3-27
-------
degree of saturation 1s smallest and reduces as conditions approach
steady state. The results presented 1n this section are summarized in
Table 3-4. It can be observed from the results that the hydraulic
head does not greatly affect leak detection sensitivity, leakage out
of the unit, or breakthrough time. Increasing the liner thickness
will Increase the breakthrough time; liner thickness has little
influence, however, on detection sensitivity and leakage out of the
unit. Hydraulic conductivity 1s the variable that most Influences
bottom liner performance. An order of magnitude decrease 1n the
hydraulic conductivity will decrease the detection sensitivity and
breakthrough time proportionally and will Increase leakage out of the
unit by an order of magnitude. In summary, the results obtained using
SOILINER and kc - 1 x 10~7 cm/s Indicate detection sensitivities in
the range of 100 Ltd (gpad), collection efficiencies that are zero
below the detection sensitivity and remain low until very large
leakage rates are encountered, and the potential for large amounts of
leakage out of the unit (approximately 100 Ltd (gpad)) prior to leak
detection and any possibility of response actions.
3.6 ANALYSIS OF PERFORMANCE - PARTIALLY SATURATED-2D
3.6.1 Introduction
The UNSAT2D computer program [Radian, 1987] was used to implement
a two-dimensional study of compacted soil bottom Uner performance.
This program employs a numerical methodology founded 1n the finite
element method and Is described 1n Section 3.6.2. Also presented 1n
Section 3.6.2 1s an overview of the results obtained. In Section
3.6.3, the performance of LDCRS with compacted soil bottom 11ners are
evaluated 1n terms of leak detection sensitivity. In Section 3.6.4
the performance of LDCRS with compacted soil bottom 11ners are
evaluated 1n terms of leachate collection efficiency. In Section
3.6.5, leakage out of units with compacted soil bottom liners is
investigated. In Section 3.6.6, the breakthrough times which can be
expected with compacted soil bottom liners are presented.
3-28
-------
3.6.2 Overview of Analysis
3.6.2.1 Descr1pt^on_of_UNSAT2D_Program
UNSAT2D Is a two-dimensional finite element computer program
prepared by S.S. Papadopulos & Associates, Inc. to simulate soil
moisture movement within waste disposal units Including landfills,
surface impoundments, and waste piles. Input parameters to the
program include water movement across model boundaries and/or
hydraulic head on model boundaries, unit geometry and materials,
material properties and initial moisture conditions in the unit and
surrounding soils. The program simulates the transient-state
distribution of hydraulic head and soil moisture within the unit for
each defined time step.
The program simulates a two-dimensional section through a waste
management unit. In formulation of the program it has been assumed
that adjacent parallel sections are Identical in their physical and
hydrological characteristics. As a result, the program can only model
linear tears in the FML (holes cannot be modeled).
The program models FMLs and geotextiles as one-dimensional
(linear) elements which have zero moisture storage. Leakage across
the element Is proportional to the head difference across the element.
Soil and waste are modeled by two-dimensional,. triangular elements
which have moisture storage capacity.
3.6.2.2 Summary of Study
The UNSAT2D program is capable of modeling a complete waste
management unit including lining system, waste, and cover systems. In
this investigation only the performance of the LDCRS and compacted
soil bottom liner are analyzed because the intent of this study is the
influence of the bottom liner on lining system performance. The
lining system modeled in UNSAT2D 1s shown 1n Figure 3-7.
3-29
-------
In contrast to the analyses 1n Section 3.4 and 3.5 where hydraulic
head on the bottom liner was a constant, UNSAT2D holds the leakage
rate through the top liner constant and the hydraulic head on the
bottom liner 1s allowed to vary. The leakage rate through the top
liner was controlled by varying the hydraulic head on the top liner
and the properties of the top Uner material. Three types of top
liner leaks were considered 1n the UNSAT2D numerical simulations
(Figure 3-7):
• uniform leakage through the entire top Uner (uniform leak);
• leakage through a portion of the top Uner on facility side
slope (sldewall leak); and
• leakage through a portion of the top liner on the bottom of the
unit (bottom leak).
A limitation of the program 1s that the smallest top liner
sidewall and bottom leak which could be analyzed Is 3 m (10 ft) wide.
In reality, FML top liner field defects are more likely to be small
tears or punctures, typically only a few millimeters (fraction of an
Inch) in diameter. Thus, the UNSAT2D top liner leak probably better
represents leakage through a composite top liner than through a top
liner consisting of an FML alone.
In addition to varying top Uner leakage rate and type of leak,
the bottom liner hydraulic conductivity and thickness were also varied
in the UNSAT2D simulations. The above four parameters were varied and
seven different combinations were analyzed. The seven combinations
are summarized 1n Table 3-5.
For the seven cases presented 1n Table 3-5 the following
evaluations are presented In subsequent sections:
• initial leak detection time;
• leachate collection efficiency;
• leakage out of the unit; and
• breakthrough time.
3-30
-------
Additional Information on each of the UNSAT2D numerical
simulations can be found in Appendix C.
3.6.3 Initial Leak Detection Time
Leak detection sensitivity was previously defined as the minimum
rate of top Uner leakage that can be detected in the LDCRS sump.
Detection sensitivity 1s calculated based on saturated, steady-state
conditions. Associated with detection sensitivity is "leak detection
time", which is the time between when leakage enters the LDCRS (when
it has just passed through the top liner) to the time it appears in
the LDCRS sump. Leak detection time can also be calculated assuming
saturated steady-state conditions. When calculated on this basis, it
is referred to as the steady-state leak detection time. Another
parameter related to LDCRS performance 1s the initial leak detection
time. This parameter corresponds to the time required to detect
leakage after a leak first occurs. It 1s different than steady-state
leak detection time because It accounts for the delay in leak
detection due to the requirement to "wet up" the LDCRS drainage media
prior to the Initiation of drain flow (this is due to capillary
suction in the LDCRS) and because it accounts for the loss of liquid
into the bottom liner due to bottom liner permeability. Both of these
factors are ignored in steady-state leak detection time calculations.
The initial leak detection time represents the behavior of a
relatively dry LDCRS during the early active life of a waste
management unit.
For a given top liner leakage rate and unit geometry, the factors
that most Influence the initial leak detection time are the capillary
suction in the LDCRS and the hydraulic conductivity of the bottom
liner. In this section, the effect of bottom liner hydraulic
conductivity on the initial leak detection time is investigated using
the UNSAT2D computer model.
The variation in Initial leak detection time with leakage rate is
summarized in Table 3-6 for different lining systems. Three different
leakage rates are presented for the case of a 1 m (3 ft) thick
compacted soil liner with a hydraulic conductivity of 10"' m/s (10~7
3-31
-------
cm/3). It can be observed that at a leakage rate of approximately
1000 Ltd (gpad) the Initial leak detection time 1s a few months.
However, 1f the rate of leakage Is an order of magnitude less, 100 Ltd
(gpad), the leak 1s not detected within a 10-year period. The long
time required to Initiate drain flow for these cases are due to two
causes: (1) the kc - 1 x 10~T cm/s compacted soil Uner permits
leakage out of the unit at a rate In the range of 100 Ltd (gpad); and
(11) the numerical simulations with UNSAT2D used an LDCRS drainage
media with k - 1 x 10~* cm/s, which corresponds to a fine sand with
capillary suction of about 1.0 m (3 ft). Therefore, a large volume of
leakage through the top Uner will be held by capillary suction in the
LDCRS which will significantly delay the initiation of drain flow.
The results presented in Table 3-6 show leakage rates of
approximately 1000 Ltd (gpad) will go undetected for longer than 10
years if the hydraulic conductivity of a 1 m (3 ft) thick compacted
soil liner is only 10"' m/s (10~» cm/s). This indicates ;that the
initial leak detection time Is very dependent upon the hydraulic
conductivity of the compacted soil bottom liner.
One procedure for Improving the overall performance of a compacted
soil liner is to increase Us thickness. The leakage rate for a 2 m
(6 ft) thick compacted soil bottom liner (k - 10~7 cm/s) is also
summarized 1n Table 3-6. It can be observed that at a leakage rate
of approximately 1000 Ltd (gpad) Increasing the thickness has little
effect on the Initial leak detection time.
3.6.4 Leachate Collection Efficiency
Leachate collection efficiency was defined 1n Section 2.5.3.2 as
the ratio of the liquid collected in the LDCRS sump divided by the
leakage through the top liner.
This section reviews the effect of compacted soil bottom liner
hydraulic conductivity and thickness on the collection efficiency of
the LDCRS assuming.uniform leakage through the top liner. Figure 3-8
shows collection efficiencies as a function of time for three
scenarios with roughly equivalent top liner leakage rates (Q):
3-32
-------
• 1 m (3 ft) thick compacted soil liner, kc - 10"» m/s (10~7
cm/s), Q - 801 Ltd (gpad);
• 1 m (3 ft) thick compacted soil liner, k - 10"' m/s (10~*
cm/s), Q - 928 Ltd (gpad); and
• 2 m (6 ft) thick compacted soil Hner, k - 10~f m/s (10~7
cm/s), Q - 800 Ltd (gpad).
It can be observed that Increasing the thickness of the compacted
soil bottom Hner results 1n only a small Improvement 1n the
collection efficiency of the system. More Important 1s the effect of
hydraulic conductivity of the compacted soil liner on leachate
collection efficiency. It can be observed that an order of magnitude
increase in hydraulic conductivity results in zero leakage being
collected from the LDCRS at the end of ten years for a rate of uniform
top liner leakage in the range of 1000 Ltd (gpad).
3.6.5 Leakage Out of Unit
Leakage out of the unit 1s plotted as a function of time in Figure
3-9 for the following three scenarios with roughly equivalent top
liner leakage rates (Q):
• 1 m (3 ft) thick compacted soil Hner, kc - 10"' m/s (10~7
cm/s), Q • 801 Ltd (gpad);
• 1 m (3 ft) thick compacted soil liner, kc « 10"' m/s (10"'
cm/s), Q - 928 Ltd (gpad); and
• 2 m (6 ft) thick compacted soil liner, kc - 10"' m/s (10~7
cm/s), and Q • 800 Ltd (gpad).
The results plotted 1n Figure 3-9 are for roughly equivalent top liner
leakage rates in the range of approximately 1000 Ltd (gpad). At
comparable leakage rates into the LDCRS, a significant Increase
(approximately 400%) in leakage out of the unit occurs when the
3-33
-------
hydraulic conductivity 1s decreased by an order of magnitude. In
addition, Increasing the thickness of the compacted soil bottom liner
from 1 to 2 m (3 to 6 ft) has only minimal effect on the leakage out
of the unit.
The plot of leakage from the unit as a function of time are
approximately linear and the rate of leakage from the unit 1s the
slope of the line. The rate of leakage from the unit 1s plotted as a
bar chart In Figure 3-10. It can be observed from Figure 3-10 that
almost all of the leakage through the top liner (approximately 1000
Ltd (gpad) for the cases considered here) can flow through the Uner
If the hydraulic conductivity Is 10"§ m/s (10~* cm/s).
3.6.6 Breakthrough Time
The effect of hydraulic conductivity and thickness on breakthrough
time 1s presented In Table 3-7. For 1 and 2 m (3 and 6 ft) thick
compacted soil liners with a hydraulic conductivity of 10~f m/s (10~7
cm/s), breakthrough will occur several years after leakage 1s detected
1n the drain. However, 1f the 1 m (3 ft) thick compacted soil liner
has a hydraulic conductivity of only 10~* m/s (10~* cm/s), then
breakthrough will occur very early In the Hfe of the facility,
without the leak ever being detected. Also, 1f the leakage rate 1s
less than approximately 100 Ltd (gpad) then breakthrough will occur
through a 1 m (3 ft) thick compacted soil Uner with a hydraulic
conductivity of 10~* m/s (10~7 cm/s) before leakage 1s detected.
3.6.7 Summary
The results of two-dimensional numerical simulations of partially
saturated flow In a LDCRS/compacted soil Uner system were presented.
These simulations were carried out by Radian Corporation using the
finite element computer model UNSAT2D. The results presented above
and summarized In Table 3-8 centered on the effect of soil Uner
thickness and hydraulic conductivity on the initial leak detection
time, leachate collection efficiency, leakage from unit, and
breakthrough time.
3-34
-------
3.7 COMPARISON OF RESULTS
In this section, the results presented in sections 3.4, 3.5, and
3.6 are compared. The purpose of the comparison is to draw
conclusions regarding both the methods used for analyzing bottom liner
performance and on the performance of the bottom liners. The results
of the one-dimensional saturated flow analyses are summarized in Table
3-3 and the results of the one-dimensional partially saturated flow
analyses are presented 1n Table 3-4. The results of the two-
dimensional partially saturated flow analyses are presented in Table
3-8. Comparison of the results will be made in terms of leak
detection sensitivity, steady-state collection efficiency, steady-
state leakage out of the unit, and breakthrough time.
Bar charts summarizing leak detection sensitivity are shown in
Figures 3-11 and 3-12 for hydraulic conductivities of 10~' m/s (10~7
cm/s) and 10~* m/s (10~* cm/s), respectively. The result in Figures
3-11 and 3-12 show good correlation 1n the prediction of leak
detection sensitivity for all three methods of analysis. Therefore,
one dimensional saturated flow can be used to evaluate detection
sensitivity. The results also show that there is a direct
relationship between the detection sensitivity and the hydraulic
conductivity of the soil.
A comparison of calculated leachate collection efficiencies using
the one-dimensional steady state saturated flow model and the two-
dimensional partially saturated transient flow model, are shown in
Figures 3-13 and 3-14, for a 1 m (3 ft) thick compacted soil liner
with a hydraulic conductivity of 10"' m/s (10~7 cm/s). The results
plotted in Figures 3-13 and 3-14 are for leakage rates through the top
liner of approximately 100 Ltd (gpad) or 1000 Ltd (gpad). Good
correlation between the two methods was obtained for the considered
leakage rates.
Steady-state rates of leakage out of the unit are shown in Figures
3-15, 3-16, 3-17, for the following Uner conditions:
3-35
-------
• 1-m (3-ft) thick compacted soil (kc = 10~7 cm/s);
• 1-m (3-ft) thick compacted soil (kc « 10~* cm/s); and
• 2-m (6-ft) thick compacted soil (kc - 10~7 cm/s).
It can be observed from these figures that the steady-state rates
of leakage out of the unit predicted by all three methods are
approximately the same, and therefore the rate of leakage out of the
unit can be estimated using one-dimensional saturated flow analyses.
The results also show that the hydraulic conductivity of the soil has
a direct Impact on the rate of leakage from the unit.
Breakthrough times for three different compacted soil bottom
liners are compared In Figures 3-18, 3-19, and 3-20:
• 1-m (3-ft) thick compacted soil (kc - 10"7 cm/s);
• • 1-m (3-ft) thick compacted soil (kc - 10~* cm/s); and
• 2-m (6-ft) thick compacted soil (kc = 10~7 cm/s).
Comparison of the results presented in the three figures show that
there is fairly good correlation between the breakthrough times
predicted by the three methods of analysis. The one-dimensional
steady-state saturated analysis appears to provide reasonable, but
slightly unconservatlve, results. The results also show that
breakthrough time can be Increased by Increasing liner thickness.
Breakthrough time 1s Inversely proportional to the hydraulic
conductivity of the compacted son used to construct the bottom liner.
3.8 SUMMARY
In this section the performance of compacted soil bottom liners
was evaluated. First, review of the factors affecting the performance
of compacted soil Hner was made. Of all the factors which can affect
the performance 1t was shown that the hydraulic conductivity of the
soil is influenced by a wide range of events. Next, a literature
3-36
-------
review of the performance of compacted soil liners was presented and
it was found that although a hydraulic conductivity of 10"' m/s (10"'
cm/s) can be achieved, a variety of factors can prevent the desired
hydraulic conductivity from being obtained. Therefore, hydraulic
conductivities of 10"' m/s (10"' cm/s) and 10~* m/s (10"' cm/s) were
subsequently analyzed.
Analysis of compacted soil Uner performance was made using three
different approaches:
• saturated, steady-state, one-dimensional;
• partially saturated, transient, one-dimensional; and
• partially saturated, transient, two-dimensional.
The analyses revealed that the three methods of analysis give
comparable results and that simple one-dimensional, steady-state
saturated flow analyses are suitable for evaluating many aspects of
bottom Uner performance. Compacted soil bottom liner thickness was
found to have only a minor effect on all aspects of bottom liner
performance except breakthrough time. However, the hydraulic
conductivity of a compacted soil liner was found to have a very
significant effect on bottom liner performance. This 1s important
because, as shown 1n Section 3.3, the in situ hydraulic conductivity
of compacted soil liners can vary over several orders of magnitude.
If the compacted soil hydraulic conductivity 1s less than the
specified value:
• the leak detection sensitivity Is diminished;
• the leachate collection efficiency is reduced;
• the rate of leakage from the unit is increased; and
• the breakthrough time 1s decreased.
3-37
-------
Table 3-1. Typical soil hydraulic conductivities. Holtz and Kovacs
[1981]
*M»I"C ••*«•» *• *•*•#
T.nn.lHH
{
{
o* CO*H»C'**W ol
COf "iC't'ii OF rtrtwf Aiiii'r
lO* 10' 10 10 ' 10 ' 10 ' 10'' 10 ' 10 • IQ • 10 • 10 '
+
+
I I
f*tft. m*«Hl
fitod iff. tititilittf div rfnwfii. HC.
1 1-
l -*«» »«
anon l«4 ix
Kmicttf . •wMtwrtf elm: fcKWnd OC d«r<
(*.». -»« oobml. I
••1mn»rv<>MM Mfill • f
WO*
•Ho >tqw*d to (Mi nxoi.l
Cvwm fcnd mi •" itOiM ttK;
lilt If »|f**M
FaNtnt |«t>*d Pnmttmrttr;
Htn^ol^
If 1 .H»rv«tl*»vivl«l Only
*IO*V •*l«HDf*WMl frf
10' 10' 1.0 IO * 10 ' 10 •* 10 4 10 * 10 * tO ' 10 ' 10 '
*Ov* M **
3-38
-------
Table 3-2. Comparison of field and laboratory hydraulic
conductivities.
Field Hydraulic
IxlO"1 m/s
Laboratory Hydraulic
Conductivity
<«/*}
8.4x10'"
to 2.5x10'' a'C
3x10'' b'd
IxlO"'
to 2x10"' b'd
Conduct ivi ty
(m/s)
3.3x10""
to 2.7x10'"
2x10''
to 2 x 10"' °
1x10''
to 4x10'' D
Probable Cause
for Difference
joints and fissures
in natural material
cracks caused by
drying
cracking caused by
drying
Reference
Griffin et al.
Daniel [1984]
Daniel [1984]
[1985]
- natural soi1
- compacted soil liner
- measured in field test
- baciccalculated
< 10'
m/s
limitations of field
placement and com-
paction procedure
(even with construc-
tion quality
assurance)
Auvinet and Espinosa
[1981]
9x10''
4x10''
m/s
m/s
1x10'"
2x10'"
m/s
m/s
variation in liner
due to soil clods.
cracks, or variation
in compactive
effort
Day and Daniel [1985]
3-39
-------
Table 3-3. Summary of one-dimensional saturated flow analyses.
• I 1x10"' m/s (1x10"' cm/s) I 1x10"* m/s (1x10"' cm/s) I
• | 25 am (1 In.) | 0.3 m (1 ft) | 25 inn (1 in.) | 0.3 m (1 ft) |
H -
Leak
Detection Sensitivity
Ltd (gpad)
1 m
(3 ft)
86
2 m
(6 ft)
86
1 m
(3 ft)
86
2 m
(6 ft)
86
1 m
(3 ft)
860
2 m
(6 ft)
860
1 m
(3 ft)
860
2 m
(6 ft)
860
Breakthrough Time
(years)
Leakage Out of the
Unit
Ltd (gpad)
15
89
31
88
12
112
28
99
1.5
890
3.1
880
1.2
1120
2.3
990
3-40
-------
Table 3-4. Summary of one-dimensional partially saturated flow
analyses .
Leak
Detection Sensitivity
Ltd (gpad)
Breakthrough Time
(years)
Steady State Rate
Leakage out of
the Unit
Ltd (gpad)
kc - 1x10"' m/«
n • 76 im
H - 1 m
(3 ft)
-100
11
97
(1x10*' on/s)
i (3 in.)
2 m
(6 ft)
-100
25
90
kc - 1x10"' m/
h • 76 nn
H • 1 m
(3 ft)
-1000
7
890
s (1x10"* OT/S)
i (3 in.)
2 m
(6 ft)
-1000
2.5
870
a y - -274 kN/m' (-40 ps1)
3-41
-------
Table 3-5. Compacted soil Uner scenarios analyzed by Radian using
UNSAT2D.
BOTTOM LINER
1
1
1
1
1
1
1
m
*c
m
"c
m
KC
m
"C
m
*c
m
"c
m
(3
st
(3
(3
(3
(6
3
(3
2
(3
ft)
.10"
ft)
10"
ft)
10"
ft)
10"*
ft)
10"
ft)
10"
ft)
thick
' m/s
thick
' m/s
thick
• m/s
thick
m/s
thick
* m/s
thick
1 m/s
thick
DESCRIPTION
compacted soi
(10"' cm/3)
compacted soi
(10~7 cm/s)
compacted soi
(10~7 cm/s)
compacted soi
(10~» cm/s)
compacted soi
(10"' cm/s)
compacted soi
(10~7 cm/s)
compacted soi
1
1
1
1
1
1
1
liner
liner
1 iner
liner
1 iner
liner
liner
LEAK TYPE
MODELED
uniform
uniform
uniform
uniform
uniform
sidewall
bottom
TOP LINER
LEAKAGE RATE
Ltd (gpad)
101
801
1419
928
.
800
60
49
kc = 10"' m/s (10~7 cm/s)
3-42
-------
Table 3-6. Effect of compacted soil bottom liner hydraulic
conductivity on initial leak detection time for various
top liner leakage rates. [Data from Radian, 1987]
Top Liner Initial Leak
Bottom Liner Leakage Rate Detection Time
Description (Ltd or gpad) (Years)
compacted soil
kc = 10"' m/s (10~7
1 m (3 ft) thick
cm/s)
1419
801
101
0.14
0.26
> 10
compacted soil
kc - 10~" m/s (10"8 cm/s) 928 > 10
1 m (3 ft) thick
compacted soil
kc = 10~f m/s (10~7 cm/s) 800 0.19
2 m (6 ft) thick
3-43
-------
Table 3-7. Effect of bottom liner hydraulic conductivity on
breakthrough time for top liner leakage rates of about
1000 Ltd (gpad). [Data from Radian, 1987]
Initial Leak
Bottom Liner Breakthrough Detection Time
Description Time (Years) (Years)
compacted soil
kc = 10"' m/s (10~7 cm/s)
1 m (3 ft) thick
4.4
4.5
9.6
0.14
0.26
> 10
compacted soil
kc - 10"' m/s (10"' cm/s) 1.0 > 10
1 m (3 ft) thick
compacted soil
kc = 10"» m/s (10~7 cm/s) - 11 0.19
2 m (6 ft) thick
3-44
-------
Table 3-8. Summary of two-dimensional partially saturated flow
analyses. [Data from Radian, 1987]
H
Steady State
Leakage Out of
the Unit
Ltd (gpad)
Steady State
Collection
Efficiency
Ltd (gpad)
Initial Leak
Detection Time
(years)
Breakthrough
Time (years)
kc » 10"' m/s (10"' cm/s)
Leakage Rate -
100 Ltd
(gpad)
1 m
(3 ft)
94 .
- 0
>10
9.6
2 m
(6 ft)
-
- 0
-
-
Leakage Rate -
1000 Ltd
{gpad)
1 m
(3 ft)
174
- 80
0.3
4.5
2 m
(6 ft)
142
- 80
0.2
11
kc - 10"' m/s (10"' cm/s)
Leakage Rate -
100 Ltd
(gpad)
1 m
(3 ft)
100
0
>10
-
2 m
(6 ft)
100
0
>10
-
Leakage Rate -
1000 Ltd
(gpad)
1 m
(3 ft)
909
- 0
>10
1.0
2 m
(6 ft)
-
- 0
-
-
3-45
-------
<
i
g
>:
1
i
!
•
(0)
IZ 13
14
15 16 17 18
19
122
•*
'a
Sll8
=
M
*
1
* £
O
(6)
— Shows chongem nxxslure
ond density from perme - .
otion
12 13 14 15 16 17 18 19
Woler Content, in X
Figure 3-1. Examples of compaction curves and effect of compaction on
hydraulic conductivity. [Lambe, 1958]
3-46
-------
io-
io-io
X
o
>-
I—
>
D
O
2
O
O
ID'"
10"
PROBE I
— USED IN
COMPUTER
• »' \°
•: V*
* °\, t
. 'V°
"Xn a
• ^S, °% »0
70
r.O 60
SUC1ION. AIM
BO
100
Figure 3-2. Effect of capillary stresses (suction) on hydraulic
conductivity. [Hamilton et al., 1981]
3-47
-------
Model
Solution
Technique
Purpose
Major Assumption
or Limitations
saturated
flow
(1-0)
partially
saturated
flow
(1-D)
partially
saturated
flow
(2-0)
hand
calculations
finite
difference
method
finite
element
method
straightforward
analysis of a
range of cases
account for
partial
saturation
account for
partial
saturation
and 2-D
only valid for
saturated flow
one-dimensional
accuracy of
analysis is
limited by finite
element mesh
Figure 3-3. Summary of models used to analyze lining systems,
3-48
-------
LDCRS
• k
H
Conn cached soil
Subq«"ade soi
Figure 3-4. Illustration of a LDCRS and compacted soil Hner.
3-49
-------
RELATIONSHIP
AMb TH
H TiM£,
COMDUCJIVITY of Bc
OF BOTTOM LlMER
_;iy v
ID
F
o?
c^ ° '10
O a)
/
—
CD
ID STEADY
STATE FLOW
COMPACTED SOIL.
Figure 3-5. Variation of theoretical breakthrough time as a function
of hydraulic conductivity and soil liner, thickness (ID
steady-state analysis).
3-50
-------
Son- POKE
/ /•'
SOIL.
50CTIOW
STRESS
/,//•/ r / J S
\
\
\
\
X V
H
SOIL.J
A/ATUf?.AL SOIL ,
Figure 3-6. Idealization of compacted soil Hner for analysis using
one-dimensional partially saturated flow computer model
SOILINER.
3-51
-------
f.-. COMPACTED CLAY
OflAWING NOT TO SCALE
Figure 3-7. Lining system modeled using the UNSAT2D program, to
study bottom liner performance. [Radian, 1987]
3-52
-------
ACHATt COLLECT/ON EFFICIENCY AS A Fu-NCTlOfv/ OF TiH£
COMPACTED SOIL BOTTOM
SO -
u
vj
U^-
UL «
LJ o"
O
o
CJ
40
20
ZD TRAMSIENT
FLOW
1o
DURATION op LEAKAGE
Figure 3-8. Collection efficiency as a function of time for a rate of
uniform top Uner leakage 1n the range of 1000 gpad (Ltd);
2D transient analysis using (UNSAT2D). [Data from Radian,
1987]
3-53
-------
LEAKft&E OUT OF UNIT AS A FUNCTION OPT
£OMPACTEb
-------
LEAKAGE OUT Of
COMPACTED SOIL BOTTOM LINERS
ct 0«.
APPROXIMATE
Tf/ftOU&H TOP
C*Lt
i
M-
ID TRANS I£A/T
FLOW
HI
Figure 3-10. Comparison of leakage out of unit using 2D transient
analysis (UNSAT2D). [Data from Radian, 1987]
3-55
-------
COMP>UISOA/ OF LEAK DtTECTIOM
SOIL BOTTOA1 L
*. LtcL)
06
VI
ID
l~ ^ Ul
n 1-
< <
Ul <£
I- ^
vft \-
t ^
o <
-4 0-
v/1
>.
2
Uj
lu
Ul
»-
a:
of
i-
«t
QL
H =
Figure 3-11. Comparison of leak detection sensitivity of compacted
son bottom liners obtained using different analysis
methods.
3-56
-------
Of LEAK DETECTION!
SOIL BOTTOM L
LtcL)
SATURATE A/VALV6.15
A^AL/S
10 TRA/^>i
PARTIALLV
/J
111
UJ
/Q.
UJ
Ul
H =
Figure 3-12. Comparison of leak detection sensitivity of compacted
soil bottom liners obtained using different analysis
methods.
3-57
-------
& LEACHATE COLLECTION EFFICIENCY
CO»\PACTBt> SOIL. BOTTOM
11%
Ul
a
IU
fc
v)
I
ID
0, -
H =
~»
-j
e*
2
ID-'/*[S (ifl'^/s')
10O L-bA. (^ (idU^Lj
(3f*)
Figure 3-13. Comparison of collection efficiencies using different
analysis methods and for a top liner leakage rate of 100
Ltd (gpad).
3-58
-------
COMPARISON OP L^ACM^TE COLLECT/ON
SO/L .BOTTOM L I A/£ fcS
81%
•£
V)
IU ^
£ ^
vn *
jQ
^
< ^
\- K
m */?
vi >
^ ~l
h 5
Q o/
^
t-4 OL.
k^- JO'1 r
Q r 600
H ^ i™
Figure 3-14. Comparison of collection efficiencies using different
analysis methods and a top liner leakage rate of - 1000
Ltd (gpad).
3-59
-------
COKPAC.I90AJ OF STEA&V- STATE LEAKA6E OUT OF
SOIL BOTTOM i//V£/?5
14-1*4
Ui
h
UJ
fc
v/l
UJ
h
h
«c
UJ
«c
h
2
UJ
»
2
<
"
A
ul
>»
-J
<
IU
h
h
«t
\n
I
OL
, OP
A RftN&E OF HXDRAULIC HCADS OH
THE BOTTOM LINER OR A RAN6E OF
TOP LIHER RRTES OF
Figure 3-15. Comparison of steady-state rate of leakage out of unit
for a 1-m (3-ft) thick compacted soil bottom liner with
kc - 1C"' m/s (10"' cm/s).
3-60
-------
OF STEAby-STATE LEAkA,&E OUT DF UV / T
SO/L BOTTOM L/A/££<>
odL oc.
142_
V/l
LU
I-
J)
i
X
15
/o
e
V-
s
T
UJ V>
I I
<*
VI
T
*t
-r
a
Figure 3-16. Comparison of steady-state rate of leakage out of unit
for a 2-m (6-ft) thick compacted soil bottom Uner
with kc => 10"' m/s (10~7 cm/s).
3-61
-------
OF STEADY-STATE LEAKA&E OUT OF UM IT
O>MPft<:TE.£> SO/L BOTTOM
Bio-mo
\n
Hi
RANfcE OF RESULTS
909
8<*0
V7 UJ
5 l-
«c <*:
>
2
ex.
V)
IU
H = IAA (3-tt)
Figure 3-17. Comparison of steady-state rate of leakage out of unit
for a 1-m (3-ft) thick compacted soil bottom liner with
kc = 10~' m/s (10~f cm/s).
3-62
-------
OF
COMP/S6TE1> SO/L BOTTOM LlA/££S
12-J5"
v/1
UJ
U
**
iS *
y> >•
Gl'
H *
Figure 3-18. Comparison of breakthrough times for a 1-m (3-ft) thick
compacted soil bottom liners with kc - 10~* m/s (10~7
cm/s).
3-63
-------
OF
CONtPA£T£b SOIL BOTTOM
T/KES
-Z8 - 31
J§$$$
t
ll
»
•2 -J
-f. ~4
o/ ^-
« -M.
H K
& *JL
r* t±.
(^t]
Figure 3-19. Comparison of breakthrough times for a Z-m (6-ft) thick
compacted soil bottom liners with kc - 10"' m/s (10~7
cm/s).
3-64
-------
COMPARISON OF 0CEAKTROU&H T/MES
CON\PACTE£ SOIL BOTTOM
1-1.5
uj
A
-c
UJ
te
VI
V)
>-
-j
I-
-c
v/)
RAA/6E OF
1.0
: t
i •"
i/> ^
2 5
t- p
<3
-------
CHAPTER 4
PERFORMANCE OF COMPOSITE LINERS
-------
4.1 INTRODUCTION
This chapter addresses the performance of composite bottom liners
composed of FML top components and compacted soil bottom components.
The performance of compacted soil liners was addressed in Chapter 3.
In this chapter, the factors affecting the performance of composite
liners are addressed and analyzed. To start, Section 4.2 presents a
discussion of the factors which influence the performance of composite
bottom liners. In Section 4.3, mechanisms for leakage through
composite liners are reviewed. In Section 4.4, the results of an
analytical (one-dimensional, steady-state) study of composite bottom
liner performance are presented. In Section 4.5, the results of a
numerical (two-dimensional, transient) study of composite bottom liner
performance are presented. In Section 4.6, the results of the
analytical and numerical studies are compared. In Section 4.7,
conclusions are drawn.
In Sections 4.4, 4.5, and 4.6 the performance of composite liners
is evaluated in terms of:
• leak detection sensitivity;
• leachate collection efficiency; and
• leakage out of the unit.
4.2 FACTORS AFFECTING THE PERFORMANCE OF COMPOSITE LINERS
4.2.1 FML Related Issues
4.2.1.1 Types of Geomembranes
Before addressing the factors affecting FML performance it is
worthwhile to briefly introduce the types of FMLs available.
Geomembranes include polymeric and asphaltic materials. When
reference is to polymeric materials only, the term FML can be used (as
discussed in Chapter 2). Polymers are chemical compounds of high
molecular weight. Only synthetic polymers are used to make FMLs. The
4-1
-------
most common types of polymers presently used as base products in the
manufacture of FMLs can be classified as shown in Table 4-1. A
description of the types of polymeric FMLs has been provided by Giroud
and Frobel [1984] and Giroud [1984f].
4.2.1.2 FML Performance
4.2.1.2.1 FML Permeation
A common misconception regarding FMLs is that they are
impermeable, that Is, no fluid will pass through an intact FML.
However, it is important to realize that all materials used as liners
are at least slightly permeable to liquids or gases and a certain
amount of permeation through liners should be expected. Additional
leakage results from defects such as cracks, holes and faulty seams
[Giroud, 1984c].
4.2.1.2.2 Defects
As mentioned above, a certain amount of leakage can occur through
FML defects. The most common type of FML defects are holes that
result from improper design, defective manufacturing or defective
installation. The size of holes may vary from pinholes up to seam
defects or tears several centimeters (several inches) long. In
general, FML holes control the amount of leakage through the FML. A
number of references are available which discuss the various types of
defects which have been observed in FML lined waste management units
[Bass, et al., 1985; Giroud, 1984a; Giroud, 1984b; Mitchell, 1984].
f
4.2.1.2.3 Damage During Manufacture, Fabrication or Installation
FML performance can be affected by events during manufacture,
fabrication, shipping, handling, storage and Installation.
There are essentially five types of defects that can develop in
FML sheeting material during the manufacturing process. All are
caused by some Irregularity or eccentricity of the sheet extrusion
process. These Include:
4-2
-------
• pinholes caused by moisture in the system during extrusion;
• holes caused by moisture or other irregularity during
extrusion;
• craters created by foreign matter in the extrudate;
• small bumps caused by excess concentrations of carbon black;
• insufficient thickness caused by the extrusion setting or feed
process; and
• scratches or gouges, caused by impact or contact with external
objects.
FMLs are susceptible to damage 1f improperly shipped, stored or
handled. Improper handling can puncture or tear the rolls. Dragging
rolls or panels of material can cause abrasion-related damage that is
usually not repairable. In this case, the damaged FML must be
rejected or wasted.
Storage of the FML on site requires a location that provides
protection from damage or contamination due to wind, dust, dirt, rain,
or ultraviolet exposure.
Proper installation is required to ensure that the FML functions
as designed. Installation problems can be caused by improper seaming,
vehicular traffic, or debris.
The purpose of the seam is to provide liner continuity between
individual FML panels. For this reason the seam must exhibit, at a
minimum:
• continuity along the entire length of the seam; and
• seam strength and ductility consistent with that of FML panels.
4-3
-------
Leakage through a FML liner 1s more likely to occur at defects in
the seam than anywhere else in the FML. Welds, whether Involving
fusion, extrusion or chemical bonding must be carried out with the
utmost care.
Weather can affect seam quality. It 1s often difficult to meet
seam performance criteria under adverse weather conditions. Usually
seaming 1s not permitted during periods of either very hot or very
cold temperatures. Seaming should not be undertaken when moisture is
present on the FML. The quality of seams is also affected by the
presence of dust, dirt or other Impurities. The FML should be as
clean as possible prior to seaming.
Construction equipment should never be allowed directly on a FML.
Usually, a protective cover about 0.3 m (1 ft.) thick is specified to
protect the FML. During placement of the protective cover, equipment
should only operate on the already placed protective cover and not on
the FML.
Debris can damage a FML liner. Cigarette butts can burn a hole
through a FML. Sharp tools and knives should never be used directly
on a FML but rather on a protective surface (e.g., a piece of wood or
scrap FML).
4.2.1.2.4 Operational Damage
In addition to the damage which can occur during manufacture,
fabrication or installation, FML damage can occur during the operation
of a waste management unit. In general, operational damage will
result from the improper use of heavy equipment on the lining system.
A properly designed lining system will have a protective cover to
distribute vehicle loads and minimize the effect of operational
errors. On the bottom of landfills, a 0.6 m (2 ft.) thickness of soil
1s used as a protective cover against equipment damage. However, it
is sometimes difficult to place a protective cover on side slopes and
they may have only minimal protection.
In general, equipment does not operate within a surface
4-4
-------
impoundment. As a result surface impoundments are not always
constructed with a protective cover. Therefore, the FMLs 1n surface
impoundments may be more susceptible to accidental damage than
landfills.
4.2.1.2.5 Conclusions
Leakage through a FML can occur because of fluid permeation
through "intact" portions of the FML and through defects in the FML.
Defects are by far the largest cause of leakage through FMLs. FML
defects can occur during FML manufacture, FML installation or unit
operation. The number and size of defects for composite bottom liners
will be quantified (based on case histories and judgment)
subsequently, and in Appendix B, and will serve as the basis for
leakage calculations.
4.2.2 Composite Liner Performance
4.2.2.1 Effeet_of_Compacted_So 1^_Hydrau VI c_Conduc11y 1 ty
In Chapter 3 it was demonstrated that if the desired hydraulic
conductivity of a compacted soil is not achieved, a significant
increase in flow through the compacted soil would result. The
hydraulic conductivity of a compacted soil portion of composite bottom
liner is also important to overall liner performance. However, the
hydraulic conductivity of the compacted soil component of a composite
is not as Important as the hydraulic conductivity of a compacted soil
liner. This 1s because 1n a composite, the upper FML limits access of
leachate to any pathways through the compacted soil. In the event
there is a hole in the FML, flow through the composite liner will be
influenced by the hydraulic conductivity of the lower compacted soil
component. The larger the hydraulic conductivity, the larger the flow
through the hole.
4.2.2.2 Effect of Contact_Between_So^l_and_FML
The effect that a hole in the FML component of a composite liner
will have on flow through the liner also depends on the quality of the
4-5
-------
contact between FML and soil. Perfect contact over the entire bottom
liner is not achievable. There will always be wrinkles in the FML or
unevenness 1n the subgrade soil surface. As a result, flow through a
leak 1n the FML will be larger than that predicted by theoretical
solutions assuming perfect contact. The effect of imperfect contact
and the quantification of the increase in leakage is addressed
subsequently and in Appendix B.
4.3 LEAKAGE MECHANISMS THROUGH COMPOSITE LINERS
4.3.1 Introduction
In Section 4.2 it was pointed out that leakage can occur through
either an intact FML or through holes 1n the FML. In Appendix B to
this report, a detailed discussion and analysis of leakage into and
through composite liners is presented. The purpose of Section 4.3 is
to briefly present the conclusions of Appendix B. The four main
topics of Appendix B are:
• leakage due to permeation through a FML;
• frequency and size of FML defects;
• analytical and model studies related to leakage through
composite liners due to a holes In the FML; and
• conclusions on leakage through composite bottom liners.
4.3.2 Leakage Due to Permeation Through FML
In Section B.2 of Appendix B an extensive discussion of permeation
through "intact" FMLs 1s presented. FML permeation may be attributed
to vapor diffusion, flow through microscopic holes and possibly other
mass transfer mechanisms. Results from permeameter tests (Figure B-l)
and vapor transmission tests are interpreted in terms of the
coefficient of migration of the FML, ug, which has been defined by
Giroud et al. [1987]:
4-6
-------
and
v = Q/A = ug/T
kg = Ug/h
(Equation 4-1)
(Equation 4-2)
where: Q = flow rate due to permeation through the FML (m'/s); ug »
coefficient of migration of the FML (mz/s); T - FML thickness (m); kg
= "equivalent hydraulic conductivity" of FML to be used with Darcy's
Equation (m/s); and h » hydraulic head acting on the FML (m).
Values of the coefficient of migration derived from the results of
permeameter tests and water vapor transmission tests are given in
Appendix B and summarized below:
h = 1 mm | h = 30 mm
CSPE
HOPE
1.3 x 10"1§
3 x 10~20
3.3 x 10~IC
1.5 x 10"17
I Values of u (mVs)
From these values of ji and from a knowledge of the hydraulic heads
acting in the permeameter and water vapor transmission tests, the
following "equivalent hydraulic conductivities" can be calculated:
| h-lmm I h - 30 mm
CSPE
HOPE
1
.3
3
x
x
10"
10~
1 6
2 7
ll.
1
5
x
x
10"
10~
1 4
1 *
Values of kg (m/s) I
4-7
-------
From the above results 1t appears that a FML "equivalent hydraulic
conductivity" of kg « 1 x 10"14 m/s (1 x 10~la cm/s) would provide a
conservative measure of permeation through an FML. This value of kg
will therefore be used 1n subsequent calculations of permeation
through an FML. In addition, a value of kg - 1 x 10"11 m/s (1 x 10"ll
cm/s) will be used in subsequent calculations to assess the
sensitivity of bottom liner performance to the selected value of kg.
This latter value of kg 1s very conservative and might be considered
to represent a "worst case scenario" for permeation through an intact
FML.
4.3.3 Frequency and Size of FML Defects
Sections B.3.2 and B.3.3 of Appendix B present data from six case
histories on the observed frequency of seam defects in FMLs installed
with and without construction quality assurance monitoring. In
Sections B.3.4 and B.3.6 the data from these case histories are
analyzed and the following conclusions are drawn on "standard" defect
frequencies and sizes which are used In subsequent calculations of
leakage through holes in FML composite liners.
• the "standard" defect (hole) area selected is 1 cm2 (10~4 m2 or
0.16 in2);
• the "standard" frequency of defect (hole) 1s one per 4000 m*
(one per acre).
The standard hole size and frequency have been selected with the
assumption that Intensive quality assurance monitoring will be
performed. The standards given above are believed to be conservative
for project where there is intensive quality assurance. These
standards do not, however, take into account cases where design flaws
or poor construction practices would lead to many seam defects or a
large tear 1n the FML.
4.3.4. Analytical and Model Studies
Sections B.4 and B.5 of Appendix B present the results of
4-8
-------
analytical calculations and model scale tests of leakage through holes
in FMLs (alone and as part of a composite liner). The analytical
calculations (B.4) and the model scale test results (B.5) are
described in detail. From the analytical calculations, conclusions
can be drawn on the quantities of leakage that would flow through a
hole in the FML component of a composite bottom liner. These
conclusions are presented in the next section.
4-3.5 Conclusions on Leakage Through Composite Liners
Section 8.6 of Appendix B provides detailed conclusions regarding
leakage through composite liners. From these conclusions, Figure 4-1
and Table 4-2 have been developed.
• Figure 4-1 gives the leakage rate and the radius of the wetted
area for the case of leakage through an FML hole in the bottom
composite Uner.
• Table 4-2 gives the leakage rate through composite bottom
liners due to FML permeation and FML holes.
Table 4-2 was established using the information on permeation
through FMLs presented in Section 4.3.2 and on leakage through FML
holes summarized in Figure 4-1. To the best of our knowledge, Table
4-1 summarizes the best demonstrated available technology (BOAT) on
leakage rates through composite bottom liners.
Table 4-2 will be conservatively interpreted (a poor contact
between the FML and compacted soil components of the composite bottom
liner will be considered) in subsequent calculations when assessing
composite bottom liners performance. For this condition leakage
through a "standard" FML defect (hole) will equal 1 Ltd (gpad) If kc -
1 x 1Q~' m/s (1 x 10'* cm/s), and 0.1 Ltd (gpad) if kc - 1 x 10"» m/s
(1 x 10~7 cm/s).
4-9
-------
4.4 PERFORMANCE OF COMPOSITE LINERS - 1-D STEADY-STATE ANALYSIS
4.4.1 Introduction
In this section, the results of analytical calculations are
presented and are used to evaluate the performance of composite bottom
liners. As discussed 1n Section 4.3, leakage through an FML can be
due to either of two causes:
• Leakage due to permeation through a FML without any holes, as
discussed 1n Section 4.3 and Appendix B. In this case, the
rate of liquid or vapor movement through the FML is not
significantly affected by the hydraulic conductivity of the
compacted soil under the FML since even low-permeability soils
are much more permeable than FMLs and act as permeable media
when placed under FMLs.
• Leakage through holes in a FML placed on a layer of compacted
low-permeability soil to form a composite liner.
The calculations presented in Section 4.4 are based on the
procedures developed in Appendix B and summarized in Section 4.3. The
calculations are presented first (Section 4.4.2) for the case of
liquid migration through an Intact FML (no holes). Second (Section
4.4.3), calculations are presented based on leakage through a hole in
the FML of a composite liner. Third (Section 4.4.4), the two types of
leakage are added together to evaluate the overall performance of
composite liners. In all three cases, the performance of composite
liners 1s evaluated In terms of leak detection sensitivity, leachate
collection efficiency, and leakage out of the unit.
4.4.2 Leakage Through an Intact FML
4.4.2.1
In Section 4.3 "equivalent hydraulic conductivities", kg, of 10~14
to 10~13 m/s (10~12 to 10~11 cm/s) were conservatively selected for
FMLs based on the results of permeameter tests and water vapor
4-10
-------
transmission tests. These "equivalent hydraulic conductivities" are
valid only for small hydraulic heads on the bottom liner (up to 0.03 m
to 0.1 m (0.1 to 0.3 ft)). This range of hydraulic heads is larger
than the range expected on the bottom liner and are therefore also
conservative. For this limited range of hydraulic heads, Darcy's
Equation can be used to approximate the performance of an Intact FML:
v - kg h/T (Equation 4-3)
Q » kg h A/T (Equation 4-4)
where: Q = leakage rate (m3/s); kg = "equivalent hydraulic
conductivity" of the FML (m/s); h = hydraulic head (m); A = surface
area of the FML (m2); and T = thickness of the FML (m).
In order to ensure the validity of Darcy's Equation for permeation
of an FML; a hydraulic head of not more than 30 mm (0.1 ft.) was
considered to act on the FML. FML thicknesses, T, of 1 mm (40 mil) or
2 mm (80 mil) were considered, along with "equivalent hydraulic
conductivities" of 10~14 and 10~13 m/s (10~12 and 10~11 cm/s). These
values will be used along with Equations 4-1 and 4-2 to determine:
• leak detection sensitivity;
• leachate collection efficiency; and
• leakage out of the unit;
for an intact FML.
4.4.2.2 Leak_Detect1on_Sensit1y1ty
In general, leak detection sensitivity is dependent upon the
properties of both the LDCRS and the bottom liner. In the event of
concentrated leakage through the top liner a two-dimensional analysis
is required to evaluate leak detection sensitivity. However, if
uniform leakage through the top liner is considered, 1t is possible to
establish a lower bound for leak detection sensitivity using a one-
dimensional steady-state saturate flow analysis.
4-11
-------
The minimum leakage rate that can be detected must be greater than
the rate at which liquid can flow by gravity into the bottom liner,
without having liquid build up 1n the LDCRS. Under this condition,
the minimum leakage rate will be independent of the hydraulic head on
the liner (it 1s zero) or the thickness of the Uner (the hydraulic
gradient is one). Therefore, the minimum detectable leakage rates for
hydraulic conductivities of 10~14 and 10"1* m/s (10"ia and 10~" cm/s)
are:
Hydraulic Leakage Rate
Conductivity Ltd
m/s (cm/s) or (gpad)
10~" (HT11) 0.01
10'" (10~la) 0.001
It should be noted that these rates are the theoretical minimum
detectable leakage rates for uniform leakage. For concentrated
leakage, the theoretical minimum detectable leakage rates will be
smaller because the area of bottom liner wetted by the leak is
smaller. However, these leakage rates are so small that no collection
system is sensitive enough to detect this level of leakage. In
practical terms, therefore, an Intact FML provides "absolute"
detection sensitivity. The above rates of leak detection sensitivity
for kg - 1 x 10~14 m/s (1 x 10"12 cm/s) Is very consistent with the
value given for FML permeation 1n Table 4-2. Also, the detection
sensitivity will not be Influenced by the hydraulic conductivity of
the underlying compacted soil layer, because even a low-permeability
soil is much more permeable than an Intact FML.
4.4.2.3 Lea cha te_Co He c 11 on_EffJ[c 1 en cy
In general, the steady-state leachate collection efficiency is
dependent upon the properties of both the LDCRS and the bottom Uner.
To evaluate leachate collection efficiency using one-dimensional
steady-state saturated flow, three simplifying assumptions are
4-12
-------
required. The first assumption Is that leakage through the top liner
is uniform. The second assumption is that a head of 30 mm (0.1 ft)
acts on the bottom liner, regardless of the flow through the top
liner. The third assumption is that any liquid that does not flow
into the bottom liner 1s collected.
Using the above assumptions, the calculated steady-state leachate
collection efficiencies (%) for a range of leakage rates and hydraulic
conductivities of 10~l* m/s (10"ia cm/s) and 1CT1* m/s (10~11 cm/s)
are (T = 1 mm (40 mil)):
Steady-State
Collection Efficiency (%)
Top Liner
Leakage Rate kg=10"14 m/s kg-10~l* m/s
Ltd (gpad) (l(Tia cm/s) (10"11 cm/s)
0.01 . 0 0
0.1 80 0
1.0 98 80
10 99.8 98
The above results are conservative, because at low rates of top
liner leakage the hydraulic head on the bottom liner will be less than
that assumed for the steady-state calculations (30 mm (0.1 ft)).
The above results indicate that even with this conservative
assumption, high steady-state collection efficiencies can be achieved
in LDCRS systems underlain by intact FMLs (even when the top liner
leakage rate 1s small, e.g., 1.0 Ltd (gpad)). Also, the steady-state
leachate collection efficiency of an intact FML will not be influenced
by the hydraulic conductivity of the underlying compacted son layer.
4-13
-------
4.4.2.4 Leakage_Out_of_Un1t
The rate of liquid permeation through an intact FML can be
approximated using Equation 4-4 and a hydraulic head of 30 mm (0.1
ft):
Rate of Permeation
FML Thickness Ltd (FML)
kg»10~14 m/s kg=10~13 m/s
mm (mil) (10~ia cm/s) (10'11 cm/s)
1 (40) 0.02 0.2
2 (80) 0.01 0.1
These results give higher permeation rates than Table 4-2 because
the results in this section are based on the conservative assumption
that h - 30 mm (0.1 ft) and use kg values larger than those reported
in Section 4.3.2 and used to establish Table 4.2. Even with these
very conservative assumptions, however, the rate of liquid permeation
through an intact FML Is very small. Also, the rate of liquid
permeation through an intact FML is not influenced by the hydraulic
conductivity of the underlying soil layer.
4.4.2.5 Summary
The calculated leak detection sensitivity, steady-state collection
efficiency and leakage out of the unit for an intact FML overlying a
compacted soil layer are summarized In Table 4-3. This table has been
established using conservative assumptions regarding FML performance.
From this summary it can be observed that a composite liner with an
Intact FML provides a very high level of performance in terms of the
lining system performance criteria Important to EPA's liquid
management strategy. In Section 4.4.3 the effect of a hole in the FML
will be addressed.
4-14
-------
The effect of FML defects on composite liner performance Is
evaluated next, In Section 4.4.3.
4.4.3 Leakage Through Holes 1n FML Component of Composite Liner
4.4.3.1
The mechanism by which leakage occurs through a defect or hole in
a FML component of a composite liner was described 1n detail in
Appendix B and summarized 1n Section 4-3. The conclusions drawn in
Appendix B and Section 4-3 will be used to evaluate:
• leak detection sensitivity;
• leachate collection efficiency; and
• leakage out of the unit
of a composite liner with a hole in the FML component.
4.4.3.2 LeakJDe t e c t 1 on_Sen s 1 t 1 v 1 ty
In general, the presence of a hole will not influence the leak
detection sensitivity of a FML, because the area of the hole is very
small in comparison to that of the surrounding Intact FML. Therefore,
leak detection sensitivity can be taken to be that of an intact FML.
It was shown in Section 4.4.2.2 that detection sensitivity is 0.01 Ltd
(gpad) for kg - 10~" m/s (10"11 cm/s) and 0.001 Ltd (gpad) for kg =
10"14 m/s (10 " cm/s)).
4,4.3.3 Le acha te_Co 1 lee 1 1 on_Ef f ±c 1 ency
The steady-state leachate collection efficiency will be dependent
upon the number of holes in the FML, as shown 1n Figure 4-2. A FML
installed with good construction quality assurance would be expected
to have not more than 3 to 5 defects per hectare (1 to 2 per acre),
and possibly fewer. It can be observed from Figure 4-2 that the
steady-state collection efficiency of the lining system would be
affected by a few FML holes only at very low rates of leakage through
the top liner. Even in the extreme event of 25 defects per hectare
(10 defects per acre) a relatively high steady-state collection
4-15
-------
efficiency (95%) would be achieved at a leakage rate equal to 20 Ltd
(gpad).
4.4.3.4
The leakage that can occur through a hole in the FML component of
a composite Uner can be obtained from the results summarized in Table
4-2. The leakage that can occur through a single "standard" FML
defect under a hydraulic head of 30 mm (0.1 ft) and conservatively
assuming poor contact between FML and compacted soil layers is:
Soil Hydraulic Rate of Leakage
Conductivity Out of Unit
m/s (cm/s) Ltd (gpad)
10"f (10~7) 0.4 (0.1)
10'* (10~4) 4 (1)
A comparison of the amount of leakage due to one "standard" FML
defect per acre and due to permeation through an intact FML is shown
in Figure 4-1.
4.4.3.5
The leak detection sensitivity, steady-state leachate collection
efficiency, steady-state leakage out of the unit that can be expected
through a "standard" FML defect are summarized In Table 4-4. It can
be observed from the results 1n Table 4-4 that a 1 cm2 (10~4 m2 or
0.16 in2.) standard defect will not substantially Impact the leak
detection sensitivity, leachate collection efficiency and leakage out
of the unit. The evaluation made 1n this section, along with the
evaluation of composite liners with Intact FML components made in
Section 4.4.2, will be drawn together in Section 4.4.4 and conclusions
will be made for the case of leakage through a composite due to both
FML permeation and FML defects.
4-16
-------
4.4.4 Leakage Through a Typical Composite Liner
The performance of a composite liner will depend upon the number
of FML defects and the amount of liquid that migrates through the
intact portions of the FML. The performance of an intact FML was
evaluated in Section 4.4.2 and the influence of a FML defect was
evaluated in Section 4.4.3. These results are combined in this
section to evaluate the overall performance of a composite bottom
liner. In this evaluation, 5 defects per hectare (2 defects per acre)
are considered. This number of defects is considered to be an upper
bound of the number expected in a properly installed FML with a good
construction quality assurance program. For this composite liner the
calculated performance under 30 mm (0.1 ft) of head is summarized in
Table 4-5. From this table, and Table 4-3, it can be observed that
the presence of a few "standard" FML defects does not greatly alter
the overall performance of the lining system.
Several observations can be made regarding leak detection
sensitivity, steady-state collection efficiency and leakage out of the
unit.
• The theoretical leak detection sensitivity of an LDCRS with a
properly designed and constructed composite bottom liner is
much less than one Ltd (gpad). A few "standard" FML defects
have a negligible effect on the leak detection sensitivity of a
lining system with a composite bottom liner.
• The theoretical steady-state leachate collection efficiency for
composite bottom liners with an Intact FML component is high
and remains high even when the FML has several "standard"
defects.
• The theoretical steady-state leakage out of a unit with a
composite bottom liner having an intact FML is much less than 1
Ltd (gpad). Just as important, the leakage out of the unit
remains less than 1 Lt'd (gpad) even when the FML has several
"standard" defects.
4-17
-------
With respect to breakthrough time, the time to breakthrough a
composite bottom liner with an Intact FML 1s very long. The time to
breakthrough a composite liner with a FML component having macroscopic
defects will not be much different than that for a compacted soil
liner alone. However, the quantity of leakage associated with
breakthrough of the composite will be at least several orders of
magnitude less than the quantity of leakage associated with
breakthrough a compacted soil Uner. Since the quantity of leakage
associated with breakthrough of a composite 1s very small, the use of
breakthrough time as a lining system performance criterion becomes
unnecessary.
4.5 PERFORMANCE OF COMPOSITE LINERS - 2-D TRANSIENT ANALYSIS
4.5.1 Introduction
The evaluation of composite bottom Uner performance in the
previous section was limited to one dimensional, steady-state
saturated conditions. Two-dimensional transient analyses of
partially saturated flow has been carried out by Radian Corporation,
Austin, TX, using the UNSAT2D finite element computer program [Radian,
1987]. The Radian evaluation 1s an extension of their work on
compacted soil bottom liners, presented previously in Chapter 3.
Details of the UNSAT2D computer program can be found in Chapter 3.
Detailed results from the Radian numerical simulations are presented
In Appendix C.
An overview of the analysis and results obtained are presented in
Section 4.5.2. Leak detection sensitivity of composite bottom liners
1s discussed In Section 4.5.3. Leachate collection efficiency of
composite liners 1s discussed In Section 4.5.4. Leakage Into and
through composite bottom liners is reviewed In Section 4.5.5.
4.5.2 Overview of Analysis and Results
The basic lining system analyzed using UNSAT2D was reviewed in
Section 3.6.2.2 and is not repeated. The only addition to the
simulations carried out for compacted soil liners is a thin, very low-
4-18
-------
permeability layer with zero liquid storage capacity. This thin layer
is placed over the compacted soil bottom liner to simulate the FML
component of a composite bottom liner. In the UNSAT2D simulations, the
migration of liquid across this very thin layer is described by the
FML leakance, L, which has units of s'1. The velocity of liquid
migration across the membrane is equal to the leakance of the FML
multiplied by the hydraulic head differential across the FML. In a
double-liner system, the hydraulic head on top of the bottom liner
is almost always very small. However, in the UNSAT2D simulations the
capillary suctions acting on the bottom of the FML component of the
bottom liner is significant. In these simulations, this capillary
suction is equivalent to a hydraulic head of 3.4 m (11.1 ft) acting on
the FML. To counteract the effect of the large hydraulic gradient set
up in the numerical simulations by the action of capillary suction
pulling water through the FML, a very low leakance value was selected
by Radian. In UNSAT2D simulations, a leakance of 7 x 10~14 s"1 was
selected to be used with a compacted soil capillary suction of 3.4 m
(11.1 ft) of negative head. These values correspond almost exactly to
a 1-mm (40-mil) thick FML with kg - 1 x 10~" m/s (1 x 10~12 cm/s)
subjected to a hydraulic head of 30 mm (0.1 ft). Radian also carried
out simulations with a FML leakance of 3 x 10~11 s"1. This leakance
is 430 times larger than the one for an "intact" FML and can be
considered to approximately represent an FML bottom liner that has
undergone "significant" deterioration. Finally, Radian carried out
several numerical simulations with an intermediate leakance value, L -
3 x 10~l* s~l, that has undergone "some" deterioration.
The analysis of composite bottom liners Is more limited in scope
than the analysis of compacted soil liners presented in Chapter 3.
The parameters which were varied were leakage rate, leak location, and
rate of permeation of the bottom liner. Four different combinations
of the above parameters were analyzed, and the parameters and results
are summarized in Table 4-6.
4.5.3 Leak Detection Sensitivity
Leak detection sensitivities estimated from numerical simulations
using UNSAT2D are presented in Table 4-5. The detection sensitivity
4-19
-------
for leakance, L - 7 x 10~14 s~l, and a hydraulic gradient of one is
estimated to be about 0.001 Ltd (gpad). This value of detection
sensitivity is very good. In fact the bottom Uner leakage rates
associated with this detection sensitivity are so small that no
collection system is sensitive enough to detect this level of leakage.
Table 4-5 also shows the leak detection sensitivity of a FML with L =
3 x 10"11 s~l. This sensitivity can be Interpreted as being one
associated with an FML that has undergone extensive deterioration over
much of Its area. The detection sensitivity for this case is on the
order of 0.4 Ltd (gpad). From this* it can be observed that the leak
detection sensitivity of a composite bottom liner with an FML
component that has undergone some degree of deterioration is still
good.
4.5.4 Leachate Collection Efficiency
The steady-state leachate collection efficiency results obtained
from the UNSAT2D numerical simulations are shown in Table 4-5. It can
be seen that for an Intact FML with L - 7 x 10'1* s~l the steady-state
leachate collection efficiency 1s very high for top liner leakage
rates of 60 Ltd (gpad). From Table 4-5, it can be deduced that even
for a very small top liner leakage rate of 1 Ltd (gpad), the steady-
state leachate collection efficiency is about 98%. It can also be
seen that as the FML leakance Increases, the steady-state collection
efficiency decreases. For Instance, for a FML leakance of L - 3 x 10"
11 s~l, the steady-state collection efficiency will be about 50% for
top liner leakage rates of about 15 Ltd (gpad). Figure 4-4 shows the
generalized relationship between steady-state collection efficiency,
rate of uniform top liner leakage and FML leakance derived from the
results of the UNSAT2D numerical simulations performed by Radian.
4.5.5 Leakage Out of the Unit
The computer program UNSAT2D was used to obtain results for
leakage Into the bottom liner for several different FML likenesses and
rates of uniform top liner leakage. These results are shown in Table
4-6. For the most part, leakage Into the bottom liner 1s close to but
slightly larger than leakage out of the unit. The difference between
4-20
-------
leakage Into and leakage out of the bottom liner is due to the fact
that some of the liquid that migrates into the bottom liner remains
held in the pore space of the bottom liner by capillary suction.
Table 4-6 shows that the steady-state leakage into the bottom
liner is very small for a composite bottom liner having an intact FML
with L = 7 x 10"14 s'1. Even for an FML with a "relatively high"
leakance of 3 x 10~12 s~l, leakage Into the bottom liner is only on
the order of 1 Ltd (gpad). Figure 4-5 shows the relationship between
cumulative leakage into the bottom liner FML leakance and time. For
comparative purposes the cumulative leakage into a compacted soil
liner with kc = 1 x 10~7 cm/s is also shown in Figure 4-5.
4.6 COMPARISON OF RESULTS OBTAINED FROM ANALYTICAL AND
NUMERICAL MODELS
A comparison of the results obtained using the one-dimensional
steady-state analytical model and the two-dimensional transient
numerical model (UNSAT2D) is presented in this section. The analytical
results were presented in Section 4.4 and the numerical results can be
found in Section 4.5.
The numerical results indicate a leak detection sensitivity on the
order of 0.001 Ltd (gpad) for a composite liner with an intact FML
(leakance equal to 7 x 10~14 s~l). The analytical calculations
resulted in a leak detection sensitivity of about 0.001 Ltd (gpad)
for an FML with an "equivalent hydraulic conductivity" kc - 1 x 10~12
cm/s. The numerical and analytical results are therefore consistent.
Leachate collection efficiencies for an LDCRS underlain by an
intact FML can be estimated from Tables 4-4 and 4-6 for top liner
leakage rates of 1 Ltd (gpad) and 10 Ltd (gpad). For the comparison
presented below, the FML thickness is assumed to be 1.0 mm (40 mils)
and the "equivalent hydraulic conductivity", kc, 1s assumed to be 1 x
10~12 cm/s (L - 7 x 10~14 s"1):
4-21
-------
Collection Collection
Model Efficiency (%) Efficiency (%)
1 Ltd (gpad) 10 Ltd (gpad)
analytical 98% 99.8%
numerical 98% 99.8%
The analytical calculations using an FML with 2 "standard" defects
can be compared to the numerical simulations using a leakance of L -
3 x 10~12 s"1. Comparison of the steady-state leachate collection
efficiencies (assuming kc - 1 x 10~7 cm/s) are shown below:
Collection Collection
Model Efficiency (%) Efficiency (%)
1 Ltd (gpad) 10 Ltd (gpad)
analytical 80% 98%
numerical 10% 91%
The above comparisons are very consistent for the case of intact
FMLs. In fact, the results for the "defective" FMLs also compare
favorably, considering the differences in assumptions between the
analytical and numerical models. The results show that the numerical
simulations using a FML with a leakance of L - 3 x 10~1Z s"1 result in
collection efficiencies somewhat less than those from the analytical
calculations using a FML with 2 "standard" defects.
The steady-state leakage Into (or out of) a composite bottom liner
with an intact FML can be estimated from Tables 4-3 and 4-6 and the
following assumptions: kg - 1 x ID'11 cm/s, L - 7 x 10~l* s~l, kc - 1
x 10~12 cm/s, T - 1 mm (40 mils), and H - 1 m (3 ft). For these
conditions:
Leakage out of Unit
Model Ltd (gpad)
analytical 0.02
numerical 0.02
4-22
-------
Below, the analytical calculations using a FML with 2 "standard"
defects are again compared to the numerical simulations using a
leakance, L = 3 x 10~1Z s"1. The following comparisons are for
steady-state leakage out of the unit:
Leakage out of Unit
Model Ltd (gpad)
analytical 0.2
numerical 0.9
The comparisons again show good consistency between the analytical
and numerical approaches.
4.7 SUMMARY AND CONCLUSIONS
In Chapter 4 the performance of composite liners was evaluated.
It was shown that leakage can result from liquid migration through an
intact FML and from flow through FML holes. Analytical procedures for
calculating liquid migration through an intact FML and leakage due to
holes were summarized from a detailed presentation of these subjects
in Appendix B. It was found that leakage through holes is much larger
than leakage due to liquid migration through an intact FML. It was
also shown that a properly Installed FML with good construction
quality assurance monitoring should have not more than 3 to 5 defects
(holes) per hectare (1 to 2 per acre), and possibly less. It was
shown that leakage through a composite bottom liner with this number
of FML defects is not large.
The analytical procedure presented and the observations regarding
defects were used to evaluate the performance of composite bottom
liners. The evaluation showed that a LDCRS underlain by composite
liner containing two standard defects per acre and subjected to a
conservative 30 mm (0.1 ft) hydraulic head can be used to detect
leaks much smaller than 1 Ltd (gpad). In addition, a LOCRS underlain
by a composite liner can collect almost all of the liquid leaking
through the top liner and can limit the rate of leakage out of the
4-23
-------
unit to less (and possibly much less) than 1 Ltd (gpad). This level
of performance is considered to be very good.
The results of numerical simulations of composite bottom liner
performance were also presented. These simulations were carried out
using the 2-D finite element computer program UNSAT2D. The
simulations are valuable because they can account for the operation of
the entire unit, from start-up through the post-closure care period.
They are also useful because they account for the effects of partial
saturation of the soil components of the lining system. The results
obtained from UNSAT2D were used to evaluate leak detection time,
leachate collection efficiency and leakage out of the unit for lining
systems with composite bottom liners. The results obtained from this
evaluation were shown to be very consistent with the results obtained
from the simpler 1-D steady-state analytical calculations. Taken
together, the analytical and numerical analyses provide a consistent
evaluation of composite bottom liner performance.
4-24
-------
Table 4-1. Summary of FML Polymers used to manufacture FMLs.
CATEGORY
POLYMER
SYMBOL3
Thermoplastics
Polyvinyl Chloride
Oil-resistant PVC
Thermoplastic Nitrile-PVC
Ethylene Interpolymer Alloy
PVC
OR-PVC
TN-PVC
EIA
Crystal line
Thermoplastics
Low Density Polyethylene
High Density Polyethylene
High Density Polyethylene-
Alloy
Polypropylene
Elasticized Polyolefin
LOPE
HOPE
HDPE-A
Thermoplastic
Elastomers
Chlorinated Polyethylene CPE
Chlorinated Polyethylene- CPE-A
Alloy
Chlorosulfonated Polyethylene CSPE
("Hypalon")
Thermoplastic Ethylene-Propylene T-EPDM
Diene Monomer
Elastomers
Isoprene-Isobutylene Rubber
("Butyl Rubber")
Ethylene-Propylene Diene
Monomer
Polycholoroprene ("Neoprene")
Epichlorohydrin Rubber
IIR
EPDM
CR
CO
symbols used by the National Sanitation Foundation (NSF) Joint
Committee on Flexible Membrane Liners (FML))
Note: Polymers are usually compounded with various additives such
as fillers, fibers, carbon black, plasticizers, stabilizers,
antioxidants, fungicides, and other polymers. These additives perform
various functions without altering the very low permeability of the
base product.
4-25
-------
Table 4-2. Leakage rates through composite liners. Leakage due to
permeation is obtained from Table B-6 (rounding up the
figures) and leakage due to holes is obtained from Figure 4-
1, as a function of the quality of contact between the FML
component and the compacted soil component of the composite
liner. This table has been established with: hole area = 1
cm* (0.16 in2.); compacted soil thickness = 0.9 m (3 ft);
FML thickness - 1 mm (40 mils); and frequency of holes = 1
per 4000 m2 (1 per acre).
Quality
of
contact
Good
Poor
Leakage
mechanism
Permeation
Hole
TOTAL
Permeation
Hole
TOTAL
Low-Permeability
Compacted Soil
Hydraulic Conductivity,
KC
10~* m/s
(10~* cm/s)
0.001
0.2
0.2
0.001
1
1
10"' m/s
(10~7 cm/s)
0.001
0.02
0.02
0.001
0.1
0.1
Values of leakage rate
in Ltd or gpad
4-26
-------
Table 4-3. Performance of an Intact FML (Note: Analysis assumes 1-D
steady-state saturated flow).
Detection Sensitivity
Ltd (gpad)
Steady-State
Collection a
Efficiency, %
Steady-State
Leakage Out of
Unit
Ltd (gpad)
kg-
"c -.
(10
1 mmb
(40 mil)
0.001
98
0.02
10" '" ml
I0"'m/s
1 cm/s)
2 mn
(80 mil)
0.001
99
0.01
! (10""
kc - 11
(10"'
1 rnn
(AQ mil)
0.001
98
0.02
:m/ s )
)'• m/s
cm/s)
2 mn
(80 mil)
0.001
99
0.01
kg-
kr - 11
(10"'
1 mm
(40 mil)
0.01
80
0.2
10'" ml
}"' m/s
:m/s)
2 tan
(80 mil)
0.01
90
0.1
5 (10'"
kc ' 1
{10"'
1 nm
(40 mil)
0.01
80
0.2
cm/ s )
D"' m/s
cm/ s )
2 nin
(80 mil)
0.01
90
0.1
a - leakage rate through top liner assumed to be 1 Ltd (gpad) for collection efficiency calculation;
b - numbers refer to FML thickness.
4-27
-------
Table 4-4. Performance of a composite liner with a single "standard"
defect 1n the upper FML component.3 (Note: This table
assumes zero liquid migration through intact portions of
FML. Analysis assumes 1-D steady-state flow.)
kc - 10"' m/s (10~* cm/s) I kc - 10~' m/s (10~7 cm/s)
Leak
Detection5' c
Sensitivity
Ltd (gpad)
Steady-State
Collection0* d
Efficiency, %
Steady-State
Leakage Out of
Unit Ltd (gpad)
0.01
95
4
(D
0.01
99.5
0.4
(0.1)
b
c
d
hole area - 1 cm* (10"* or 0.16 1nJ) and hydraulic head on
composite liner - 30 mm (0.1 ft);
- kr
10
•I 3
m/s (10"" cm/s);
- one hole per acre;
- leakage rate through top liner assumed to be 20 Ltd (gpad)
for collection efficiency calculation.
4-28
-------
Table 4-5. Performance of composite bottom liners with 2 "standard"
FML defects per acre. (Note: This table includes liquid
migration due to FML permeation (Table 4-3). Analysis
assumes 1-D steady-state flow.)
Leak
Detection Sensitivity
Ltd (gpad)
Steady-state
Collection °
Efficiency. X
Steady-state
Leakage Out of
Unit
Ltd (gpad)
kg-
kc ' 1
1 mmc
(40 mil)
0.001
90
2.0
10"' ' ml
D"1 cm/s
2 mn
(80 mil)
0.001
90
2.0
S (10'"
kc - 10
1 mm
(40 mil)
0.001
99
0.2
cm/s)
'' cm/s
2 frm
(BO mil)
0.001
99
0.2
kg.
kc - 1
1 mn
(40 mil)
0.01
89
2.2
10~" ml
0"* cm/s
2 mn
(00 mil)
0.01
89
2.1
s (10'"
kc . 10
1 mn
(40 mil )
0.01
98
0.4
cm/s)
"' cm/s
2 mn
(80 mil )
0.01
98
0.3
a - hole area « 1 cm* (10"* m1 or 0.16 in') and hydraulic head on composite liner - 30 nro (0.1 ft);
b - leakage rate through top liner « 20 Ltd (gpad);
c - numbers refer to FML thickness.
4-29
-------
Table 4-6. Performance of composite bottom liners based on results
of UNSAT2D numerical simulations carried out by Radian.
(Note: L - 7_ x 10~14 s"1 corresponds to an intact FML; L
= 3 x 10 " s ' corresponds to a highly deteriorated FML).
Leak
Detection Sensitivity
Ltd (gpad)
L - 7 x 10~'4 s"1
0 - 700
Ltd (gpad)
.001a
Q -60
Ltd (gpad)
. 001
L - 3 x 10"" $"'
Q - 1240
Ltd (gpad)
.04
90
Ltd (gpad)
.04
L • 3 x 10'" s"1
0-780
Ltd (gpad)
0.4
90
Ltd (gpad)
0.4
Steady - state
Collection Efficiency
(X)
Steady - state
Leakage Into
Bottom Liner
Ltd (gpad)
100
0.02
99.9
0.02
99.9
0.9
99
0.9
99
7.6
92
7.6
Notes: a - leak detection sensitivity results are estimated rather than calculated values.
4-30
-------
CN»IT F0«.
MCAb I, • JO «"*t
-------
- STAT6
WITH MULTIPLE
o
z
u.
o
{-
O
_J
O
50
RA16 OF
LEAKAGE THeo06H THE TOP LINER
Figure 4-2. Steady-state leachate collection efficiency of a
composite liner with multiple imperfections.
4-32
-------
LEAKAGE OUT OF THE UNIT
(LEAKAGE IMTO THE Boll CM L/AJEfO
Cgpad °" Ltd)
1.0
i_
o £
,0
L's
-4-
0 v)
-??
O
1 ^^
o
STA1E
U= O.O3
H • 1^
§f -s 5" ^'T
ci^0 as= 0 ^
-~s "^ £ ^J ^
b 3 o J. "^^3 £
r ^ ^ F* """
ii " v " i H
°- o U a^ .• w *> H
^-»/
-------
- STATE COLLECTION EFFICIENCY
OBTA/NEb FROM UMSATZC? MMEKICAL SlMl/LAl
U
Z
u»
OJ
•* 50
O
uu
0
.1
10
FLOW
L = FML LEAKANCE
100
100O
RATE. OF
(gpad
TOP
Ltd)
Figure 4-4.
Comparison of steady-state leachate collection
efficiencies of LDCRS with composite bottom liners of
various leakances. [Data from Radian, 1987]
4-34
-------
CUMULATIVE LEAKAGE INTO THE BOTtOM LINER
OBTAINED FRoM USAT2D NUMERICAL SIMULATIONS
\(J
rt.
UJ
>£- A f— 1^1
Ox-x 1P*1
ZD TRANSIENT
FLOW
^o+e: O refers 4-0
CO 0
lij
<
UJ
lu
>
§
s:
D
o
\o'
101
10'
10*
10
.-iz
Q- ^OLtd (gpod>
5<10 ns'
Q= '1O Ltd (gpad )
Ltd (cjpud)
10'
1O1
10"
10'
Figure 4-5.
DURATION OF LEAKA6E
Cumulative leakage into the bottom liner for different
FML leakances and for a rate of top liner leakage, 0,
in the range of 100 Ltd (gpad). [Data from Radian,
1987]
4-35
-------
CHAPTER 5
COMPARISON OF COMPACTED SOIL
AND COMPOSITE BOTTOM LINER PERFORMANCE
-------
5.1 INTRODUCTION
5.1.1 Purpose
The purpose of this chapter 1s to summarize and compare the
performance of compacted soil and composite bottom liners. This
comparison is presented in terms of the system performance criteria
which are believed to be most critical to meeting EPA's goals of
preventing migration of hazardous constituents from landfill and
surface impoundment units to the extent technically feasible. These
criteria are:
• leak detection sensitivity of the LDCRS;
• leachate collection efficiency of the LDCRS; and
• leakage Into the bottom Hner and out of the unit.
5.1.2 Organization of this Chapter
The remainder of this chapter is comprised of four sections
organized as follows:
• Section 5.2 presents data on leak detection sensitivity.
Specifically, this section compares the sensitivity of
equivalent LDCRS having either compacted soil or composite
bottom Hner systems.
• Section 5.3 presents data on leachate collection efficiency.
Specifically, this section compares the collection efficiency
of equivalent LDCRS having either compacted soil or composite
bottom liner systems. Comparisons are made 1n terms of both
steady-state and cumulative collection efficiencies.
• Section 5.4 presents data on total leakage Into the bottom
liner and out of the unit. Specifically, this section compares
the steady-state and cumulative leakage out of equivalent waste
management units having either compacted soil or composite
bottom liner systems.
5-1
-------
• Section 5.5 presents a summary of comparisons and draws
conclusions.
5.1.3 Comments on Data
The data presented in this chapter were developed from studies,
literature reviews, field reviews and calculations which were
presented in Chapter 3 for compacted soil bottom liners and Chapter 4
for composite bottom liners. The reader is encouraged to study
Chapters 3 and 4 so as to better understand the comparisons presented
in this chapter.
The numerical data presented in this chapter were developed from
calculations assuming either one-dimensional, saturated steady-state
flow or two-dimensional transient flow. In Chapter 3 it was shown
that one-dimensional saturated steady-state analysis results are in
good agreement with results obtained from more detailed 'calculations
assuming one-dimensional transient flow through partially saturated
soils (using the SOILINER computer model) and two-dimensional
transient flow through partially saturated soils (using data from
Radian obtained with the UNSAT2D computer model). In Chapter 4 it was
shown that the one-dimensional saturated steady-state analysis gives
results which are very consistent with those obtained from the two-
dimensional partially saturated transient analysis (UNSAT2D).
The one-dimensional steady-state analyses presented in this report
are appropriate for Investigating the leakage through a section of
bottom liner due to an overlying hydraulic head. This type of
analysis Is not appropriate to evaluate overall facility performance
(which includes modeling: (1) the entire lining system; (11) time
from the start of a unit's active life through the post-closure care
period; and (111) the time-dependent unit boundary conditions). For
this, a two-dimensional (and Ideally three-dimensional) model which
can account for time dependence 1s needed. The UNSAT2D model fits
into this category. It can be used to compare the performance of
equivalent facilities subject to equal top liner leakage rates and
having either compacted soil or composite bottom liners.
5-2
-------
5.1.4 Presentation of Data
For the sake of clarity of presentation, the data presented in
this chapter is primarily in the form of simple bar charts comparing
leak detection sensitivity, leachate collection efficiency and total
leakage out of the unit for compacted soil and composite bottom
liners. These bars charts were derived from more detailed data tables
and graphs presented in Chapters 3 and 4.
Table 5-1 presents a summary of bottom liner dimensions and
properties used to generate the bar charts for one-dimensional,
saturated steady-state leakage analysis presented in this chapter.
5.2 LEAK DETECTION SENSITIVITY
5.2.1 Definition and Importance
The leak detection sensitivity 1s the smallest leakage rate
through the top liner (E in Figure 2-4) that can be detected in the
LDCRS sump within a reasonable period of time. The RCRA amendments of
November 1984 create a statutory requirement for leak detection
systems at hazardous waste management units. A small detection
sensitivity standard, based on BOAT for leachate collection and
removal systems, is an important feature of leak detection capability.
It ensures that the owner or operator will have the ability to monitor
his unit for even very small rates of leakage through the top liner.
Detection of small rates of leakage 1s Important to ensure they are
collected. Detecting small rates of leakage 1s also crucial for
compliance with the statutory requirement to detect leakage at the
earliest practicable time.
5.2.2 Comparison of Compacted Soil and Composite Bottom Liners
Figure 5-1 compares the "steady-state" leak detection sensitivity
of compacted soil and composite bottom liners. It was shown In
Chapters 3 and 4 that steady-state, saturated analyses and transient,
partially saturated analyses provide comparable detection
sensitivities. It can be seen that the detection sensitivity of an
LDCRS underlain by a 3 ft (1 m) thick layer of compacted soil with a
5-3
-------
hydraulic conductivity of 1 x 10~7 cm/s 1s about 86 gallons per acre
per day (gpad) or liters per 1000 m2 per day (Ltd). In other words,
based on the one-dimensional steady-state analysis, a uniform top
liner leakage rate smaller than 86 Ltd (gpad) would never be
collected.
In reality, a concentrated top liner leak smaller than 86 Ltd
(gpad) may be detected because the wetted area associated with a
concentrated leak will be Just a portion of the bottom liner surface
area. However, establishing a detection sensitivity criterion based
on an assumed uniform leak Is entirely acceptable.
From Figure 5-1 it can be seen that an LDCRS with a composite
bottom liner can theoretically detect leakage rates 10,000 to 100,000
times smaller than a compacted soil bottom Uner. This difference in
detection capability 1s dramatic. The magnitude of the difference may
be better Illustrated on a logarithmic scale, as shown in Figure 5-2.
In fact, the detection sensitivity associated with' a composite bottom
liner exceeds the practical capabilities of typical liquids removal
systems (sumps and pumps) to collect and remove the leakage. It is
further noted that a few small "standard defects" (see Table 5-1) have
a negligible Influence on the leak detection sensitivity of a
composite bottom liner.
Figure 5-1 also shows that hydraulic conductivity has a
significant Influence on the detection sensitivity of LDCRS with
compacted soil bottom liners. This figure clearly demonstrates the
importance of achieving specified soil hydraulic conductivities in the
field.
5.3 LEACHATE COLLECTION EFFICIENCY
5.3.1 Definition and Importance
The leachate collection efficiency Is the ratio of the leachate
collected in the LDCRS sump (G in Figure 2-4) divided by the leakage
entering the LDCRS through the top Uner (E In Figure 2-4). There are
two measures of leachate collection efficiency: (1) cumulative
leachate collection efficiency which 1s based on the total leakage
entering the LDCRS from the beginning of a unit's active life to any
5-4
-------
other point in time; and (ii) the steady-state leachate collection
efficiency at any point in time. The cumulative leachate collection
efficiency is smaller than the steady-state leachate collection
efficiency because the cumulative efficiency considers the leakage
held in the LDCRS drainage media by capillary tension as uncollected
leakage.
Leachate collection efficiency 1s a very Important concept in
EPA's liquids management strategy for protecting human health and the
environment. This strategy for management of land disposal units
(except land treatment units) has two parts: (1) minimize leachate
generation in the waste management unit; and (ii) maximize collection
and removal of leachate from the unit at the earliest practicable
time. Clearly, a high leachate collection efficiency is a
prerequisite for meeting the second part of EPA's liquids management
strategy. Without a high collection efficiency, leachate collection
and removal cannot be maximized.
5.3.2 Comparison of Compacted Soil and Composite Bottom Liners
The steady-state leachate collection efficiencies of LDCRS with
compacted soil and composite bottom liners are compared in Figure 5-3.
This comparison has been made for an FML with an "equivalent hydraulic
conductivity", kg = 1 x 10'11 cm/s. (This value represents a very
conservative upper bound of "equivalent hydraulic conductivities" for
FML liners. As was shown In Table 4-2, the "equivalent hydraulic
conductivities" of most FMLs will be at least an order of magnitude
smaller than the value used to generate Figure 5-3.) It can be seen
that the collection efficiency for an LDCRS with a compacted soil
bottom liner 1s zero until the rate of uniform leakage exceeds the
leak detection sensitivity. In contrast, the collection efficiency of
a LDCRS with an Intact composite bottom liner is very high, even for
top liner leakage rates as low as 1 Ltd (gpad).
Figure 5-3 also shows the effect of a small hole ("standard
defect") in the FML component of a composite bottom liner. It was
shown in Chapter 4 and Appendix B that one small defect per acre is
probably a reasonable, conservative assumption for a well designed and
constructed facility with a good construction quality assurance
program. This small defect is assumed to be square with a side
5-5
-------
dimension of 0.4 in (10 mm). Figure 5-3 shows that the effect of one
"standard defect" on collection efficiency is small, even at top liner
rates of uniform leakage of 1 Ltd (gpad). Figure 5-3 also shows the
effect of a composite bottom liner with one standard defect in the FML
and with a compacted soil component hydraulic conductivity of 1 x 10~*
cm/5. This case represents a composite liner having an FML component
with a defect and a compacted soil component that does not meet
specification. Under this set of "problem" conditions, the leak
detection sensitivity of the LDCRS reduces to 1 gpad (which 1s still
almost 1000 times greater than the leak detection sensitivity of the
compacted soil alone). It is also clear from Figure 5-3 that the
leachate collection efficiency for compacted soil liners increase
rapidly once the rate of leakage exceeds the liner's leak detection
sensitivity. At a top liner leakage rate of 100 Ltd (gpad), the
collection efficiency for a compacted soil liner with kc = 1 x 10~7
cm/s in all areas is only 11%. At a top liner leakage rate of 1000
Ltd (gpad), however, the collection efficiency has increased to 91%.
EPA believes that most leaks win be 1n the'range of 100 Ltd (gpad) or
less.
Figure 5-4 compares the leachate collection efficiencies of
composite liners with varying numbers of FML "standard defects". It
can be seen that even with 25 "standard defects" per acre (which is an
unrealistically high number for a unit with good CQA), the leachate
collection efficiency of the LDCRS 1s very good and 1s far better than
those for the compacted soil bottom liners shown 1n Figure 5-3. This
result is significant and shows that even a very poorly Installed
composite liner provides a system with a much higher collection
efficiency than provided by a compacted soil bottom liner with kc - 1
x 10~7 cm/s 1n all areas.
Figure 5-5, 5-6, and 5-7 present bar charts comparing the leakage
collection efficiencies (based on the ID steady-state analysis) of
LDCRS with various types of bottom liners. These charts are Included
for completeness and clarity of presentation.
•
Figure 5-8 presents the steady-state leachate collection
efficiencies obtained from waste management unit simulations using the
two-dimensional numerical model UNSAT2D. Figure 5-8 corresponds to a
top liner uniform leakage rate of about 1000 Ltd (gpad). Figure 5-8 is
5-6
-------
similar to Figure 5-7 (ID steady-state analysis) as they both compare
the performance of LDCRS at top liner rates of uniform leakage of 1000
Ltd (gpad). The following comparisons are made:
Leachate collection efficiency
ID steady-state 2D transient
Composite (intact) > 99.9 > 99.9
Compacted soil (kc - 1 x 10"T cm/s) 91 78
Compacted soil (kc - 1 x 10~* cm/s) 11 10
Inspection of these results show that the ID steady-state analysis
and the 2D transient analysis give very consistent results for LDCRS
leachate collection efficiencies. In both cases, the composite bottom
liner provides a more efficient system than the compacted soil bottom
liner with kc - 1 x 10~7 cm/s, which In turn provides a significantly
more efficient system than the compacted soil bottom Uner with kc • 1
x 10~* cm/s.
The steady-state leachate collection efficiencies obtained from
UNSAT2D and from the ID analyses can be compared for top liner rates
of uniform leakage in the 50 Ltd (gpad) range. When this 1s done, the
following results are obtained:
Leachate collection efficiency
ID steady-state 2D transient
Composite (Intact) > 99 > 99
Compacted soil (kc = 1 x 10~7 cm/s) 0 0
Compacted soil (kc - 1 x 10"* cm/s) 0 0
Inspection of these results shows that at this rate of uniform top
liner leakage, the compacted soil liners provide zero leachate
collection efficiency, while the Intact composite provides a very high
collection efficiency. This difference Is dramatic.
Figures 5-9 and 5-10 present cumulative leachate collection
efficiencies from waste management unit simulations using UNSAT2D and
5-7
-------
a top liner rate of uniform leakage of about 1000 Ltd (gpad). It can
be seen that the cumulative efficiency increases with time after unit
start-up. The efficiency is low at the time of unit start-up since
much of the Initial leakage is held within the LDCRS by capillary
tension. Depending on the leakage rate and the hydraulic conductivity
of the LDCRS drainage media, the duration of this transient "wetting
up" period can be significant. This period will be relatively short,
however, for a large top liner leakage rate of 1000 Ltd (gpad). Once
the "wetting-up" of the LDCRS 1s complete, the cumulative collection
efficiency begins to approach the steady-state efficiency; the
difference between the two 1s largely accounted for by the leakage
stored in the LDCRS by capillary tension (as noted in Figures 5-9 and
5-10).
Figures 5-11 and 5-12 present collection efficiencies from waste
management unit simulations using UNSAT2D and a top liner rate of
leakage in the range of 50 Ltd (gpad). The teak type used In these
numerical simulation 1s a sldewall' leak. Figure 5-11 presents the
steady-state collection efficiency 10 years after facility start-up.
It can be seen that the compacted soil collection efficiency is zero,
which is logical since the rate of leakage is below the detection
sensitivity of the bottom liner. The steady-state collection
efficiency associated with the intact composite bottom Uner is in
excess of 99%. The cumulative collection efficiencies after 10 years
are shown In Figure 5-12. The collection efficiency for a LDCRS
underlain by an Intact FML 1s 95%. This efficiency 1s "relatively low"
because, even after 10 years of unit operation, the LDCRS continues to
entrap new leakage through capillary tension. Almost 5% of the liquid
that has passed through the top liner 1s held in the LDCRS sand by
capillary tension. In the UNSAT2D numerical simulations, the drainage
media in the LDCRS 1s assumed to have a saturated hydraulic
conductivity of 1 x 10"' cm/s. This corresponds to an initial
capillary tension of about 0.5 m (1.5 ft) of negative hydraulic head.
Thus, there exists sufficient capillary tension to essentially
saturate the entire LDCRS drainage media. For a 0.3 m (1 ft)
thickness of sand with a porosity of 30%, the void space in the sand
that is available to entrap leakage Is approximately 100,000
gallons/acre. If this capillary tension 1s destroyed (either through
the use of a more permeable drainage medium such as gravel or a
synthetic drainage material), the leakage storage capacity of the
5-8
-------
LDCRS would be dramatically reduced and the associated collection
efficiencies would be increased.
Figures 5-11 and 5-12 are interesting because they show the effect
of a major imperfection in the FML component of a bottom composite
liner. These results from the UNSAT2D numerical simulations indicate
a significant effect of a major bottom liner FML defect. The
imperfection modeled in UNSAT2D consists of an approximately 3 m (10
ft) long bottom liner sidewall leak directly under a top liner
sidewall leak. This represents a very major breach of the FML
component of a composite bottom liner. The leak is simulated in the
numerical model by increasing the hydraulic conductivity of the
leaking portion of the FML.
5.4 LEAKAGE OUT OF WASTE MANAGEMENT UNIT
5.4.1 Definition and Importance
Leakage out of the unit refers to leakage that passes through the
bottom liner into the ground (J in Figure 2-4). A related performance
variable is leakage into the bottom liner (H in Figure 2-4).
Leakage out of the unit is a very important concept in EPA's
liquids management strategy. As described 1n Section 5.3.1, one part
of this strategy is to maximize leachate collection and removal from
the unit. This can only be achieved If leakage out of the unit Is
minimized to the extent technically feasible. Further, EPA's goal for
lining systems since promulgation of technical and permitting
standards under Part 264 on July 26, 1982 (47 FR 32274) has been to
prevent migration of wastes to the subsurface soil or ground water.
While EPA recognizes that absolutely achieving this goal is not
technically achievable at present, they believe that BOAT can come
very close to preventing migration. The degree to which a lining
system satisfies this goal is reflected by the leakage out of the
unit.
5.4.2 Comparison of Compacted Soil and Composite Bottom Liners
Steady-state leakage out of the unit (or leakage into the bottom
liner) for compacted soil and composite bottom liners with a 0.03 m
5-9
-------
(0.1 ft) hydraulic head on the bottom liner are compared in Figure 5-
13. This figure is based on one-dimensional, saturated steady-state
analysis. A bottom liner hydraulic head of 0.03 m (0.1 ft) is
believed to represent a worst case scenario. (Also, 0.03 m (0.1 ft)
corresponds approximately to the capillary rise in a sand LDCRS
drainage media with a hydraulic conductivity equal to about 1 cm/s.)
From Figure 5-13 it can be seen that an intact composite bottom
liner with an "equivalent hydraulic conductivity" of the FML component
of kg - 1 x 10~12 cm/s permits almost 5,000 times less leakage out of
a unit than a 1 x 10~7 cm/s compacted soil liner alone. Figure 5-14
indicates that increasing the thickness of the compacted soil has a
negligible effect on the steady-state leakage out of the unit. Figure
5-15 is interesting because it shows total leakage out of the unit for
composite liners having FMLs with defects. It can be seen that the
leakage through a composite liner with kg - 1 x 10~12 cm/s and kc - 1
x 10~7 cm/s and with one "standard defect" per acre (which is believed
to be conservative for a properly designed and constructed unit) is
still on the order of one thousand times smaller than the leakage
through a compacted soil liner with kc - 1 x 10~7 cm/s in all areas.
Even if the number of defects were Increased to 10 (which might be
considered to represent a "problem" site) leakage through the
composite is on the order of 100 times smaller than leakage through
the compacted soil.
Figure 5-16 compares cumulative leakage out of the unit for times
up to 10 years, for units with compacted soil and composite bottom
liners. Figure 5-17 presents the same Information as Figure 5-16, but
on a log-log plot and for a duration of leakage up to 27 years (10,000
days). The presentation of data 1n the format given in Figure 5-17
allows for the evaluation of cumulative leakage out of the unit, for
any duration of leakage and for any type of bottom Uner.
Figures 5-18 to 5-22 present comparative results on cumulative
leakage into the bottom Uner which were obtained using UNSAT2D
numerical simulations. Figures 5-18 to 5-22 correspond to a top liner
rate of uniform leakage on the order of 1000 Ltd (gpad). In these
simulations, the leakage into the bottom liner associated with an
Intact composite liner is on the order of 5000 times smaller than that
associated with a compacted soil liner with kc - 1 x 10~7 cm/s.
5-10
-------
Figures 5-21 and 5-22 present UNSAT2D simulations for a sidewall
leak and a leakage rate of about 50 Ltd (gpad). These results from
the UNSAT2D numerical simulations indicate a significant effect of a
major bottom liner FML defect (I.e., a defect 3 m (10 ft) long). This
represents a very major breach of the FML component of a composite
bottom liner.
5.5 SUMMARY OF COMPARISONS
A summary is presented in Table 5-2 of the comparisons made using
results from the one-dimensional steady-state analyses. These results
are based on uniform leakage through the top liner, on a hydraulic
head on the bottom Uner equal to 0.3 m (0.01 ft), and on kc =« 1 x
10~7 cm/s and kg - 1 x 10'11 cm/s. From this summary, the following
observations are drawn regarding the performance criteria critical to
establishing BOAT for LDCRS and bottom liner systems and fundamental
to EPA's liquids management strategy.
• Leak detection sensitivity - The theoretical leak detection
sensitivity of an LDCRS with a properly designed and
constructed composite bottom Hner is much less than one Ltd
(gpad). A few "standard" FML defects have a negligible effect
on the detection sensitivity of a lining system with a
composite bottom liner. By comparison, the leak detection
sensitivity of a compacted soil bottom liner 1s on the order of
100 Ltd (gpad) with kc - 1 x 10~' cm/s 1n all areas of the
liner.
• Leachate collection efficiency - The theoretical steady-state
leachate collection efficiency for composite bottom liners with
intact FML 1s 1n excess of 99%, even for relatively low top
liner leakage rates such as 20 Ltd (gpad). Just as important,
the leachate collection efficiency remains high even when the
FML component of a composite bottom Hner has several
"standard" defects (more defects than would be expected in a
properly designed and constructed lining system). In contrast,
the theoretical steady-state collection efficiency of a
compacted soil bottom liner is zero for all rates of uniform
5-11
-------
top liner leakage up to approximately the leak detection
sensitivity (on the order of 100 Ltd (gpad) for a compacted
soil bottom liner with kc = 1 x 10~7 cm/s, and on the order of
1000 Ltd (gpad) for a compacted low-permeabiliHy soil bottom
liner with kc - 1 x 10~* cm/s).
• Leakage into the bottom liner and out of the unit - The
theoretical steady-state leakage out of a unit with a composite
bottom liner with an intact FML is much less than 1 Ltd (gpad).
Just as important, the leakage out of the unit remains less
than 1 Ltd (gpad) even when the FML component of a composite
bottom liner has several "standard" defects (more defects than
would be expected in a properly designed and constructed lining
system). In contrast, for a uniform hydraulic head on the
bottom liner of 0.03 m (0.1 ft), the leakage out of a unit with
a compacted soil bottom liner and kc - 1 x 10"' cm/s in all
areas of the liner is on the order of 100 Ltd (gpad).
5-12
-------
TABLE 5-1. Bottom liner dimensions and properties for one-dimensional,
saturated steady-state leakage analysis.
Compacted Soil
Thickness - H - 1 m (3 ft)
Hydraulic conductivity
standard - kc = 1 x 10~7 cm/s
alternate - kc = 1 x 10~* cm/s
Hydraulic head on liner - h - 0.03 m (0.1 ft)
Composite (Intact)
Compacted soil thickness - H = 1 m (3 ft)
Hydraulic conductivity
standard - kc = 1 x 10 7 cm/s
alternate - kc = 1 x 10~* cm/s
FML thickness
standard - T - 1.0 mm (40 mils)
alternate T - 2.0 mm (80 mils)
Hydraulic conductivity
standard - kg = 1 x 10~12 cm/s
alternate - kg = 1 x 10"11 cm/s
Hydraulic head on liner h - 0.03 m (1 ft)
Composite (with defect)
Same as intact composite except for defect.
FML defect size - 1 cm x 1 cm (0.4 in x 0.4 in)
Number of defects
standard - 1 defect per acre
alternate - multiple defects per acre
5-13
-------
TABLE 5-2. Summary of comparative results based on one-dimensional
steady-state analysis1.
Detection Sensitivity
(9l«d)
Steaay-stdte collection
efficiency (0 ' 20 gpad)
1
Contacted toll
(•eet design spec.)
kc • 1 < 10'* o/s
86
0
2
Convicted Sail
(Ooesn't meet >ptc.)
kc . 1 i 10*' w/s
860
0
3
CcoMslte
(no ocfects)
kg . 1 i 10'" ai/s
0.001
99.91
4
Comostte
(o«« Detect)
k, . 1 « 10'" w/s
0.001
99.4X
5
Composite
(ten aefects)
k, . 1 I 10"" cm/S
0.001
951
5le»oy-state collection
e"i:>»»cy (0 • 100 gpad)
Steady-state collection
efficiency (0 - 1000 gpad)
Steady-state
leakage out of the unit
(goad) (0 • 20 gpad)
Steady-state
leakage out of unit
(gpad) 0 * 1000 gold)
Cumulative leakage out of
unit after 10 years
(gal/acre) (0 • 20 gpad)
Cumulative leakage out of
unit after 10 years
(gal/acre) (0 • 1000 gpad)
Ill
911
20
89
7 i 10-
J • 10'
0
11
20
890
7 i 10'
J « 10'
99.981
> 99.991
0.02
0.02
7 i 10'
7 i 10'
99.91
> 99.991
0.12
0.12
< i 10*
4 I 10'
991
> 99.991
1.0
1.0
4 » 10'
4 I 10'
Note: ' In ill cases. H . 1 n (3 ft); ti . 0.0] • (O.I ft): I • 1 im (40 mils);
k, . 1 i 10"" c»/l: ana kc . 1 > 10"' CB/S unless otherwise noted.
5-14
-------
LEAK DETECTION SENSITIVITY
•" Ltd)
860
__ *
O f-
o ^
a ^
£"T
o „
ID STEADY-
STATE FLOW
t
H
rf> ~
S.^
E *
66
O.OO1
Figure 5-1.
Comparison of leak detection sensitivity (minimum leakage
rate through the top liner needed to detect leakage) of
LDCRS with compacted soil and composite bottom liners (ID
steady-state analysis).
5-15
-------
LEAK DETECTION SENSITIVITY
°" Ltd)
>-
I-
IU j
V) '
T)
Z o
VJ
VU
t
10'
^-x
± 10*
10"
10
f?o»\f)c 'Tor
Co>npac)ed soils
ID SIEADY -
STATE FLOW/
Ranqc -for
Composites (^r\^ac^)
I0
's
10
10
"'
10
10
'1
10" 10" 10"'*
HYORAUHC
Figure 5-2. Comparison of leak detection sensitivity of LDCRS with
compacted soil and composite bottom liners (ID steady-
state analysis).
5-16
-------
5T£AT>Y- STATE COLLECTION
COMPACTED SOIL AN[> COMPOSITE
100 r-
Oi
o
u_
u.
o
H
-------
COMPOSITE WITH MULTIPLE IMPERFECTION
100 r-
U_
vu^
w£
_
O
t-
O
U)
_i
_J
O
0
10
ZO
3.0
50
RME OF
LEAKAGE TH£006H THE TOP LlMCK
0" ltd")
Figure 5-4. Comparison of steady-state leachate collection
efficiencies of LDCRS with compacted soil and composite
bottom liners. Composite bottom liners include FMLs with
defects (ID steady-state analysis).
5-18
-------
ST£At>y-5TAT£
COLLECTIONS
TOP Lll
2.0 GPAD (LTD)
(%)
o v/i
2
•O
-------
ST£Al?y-STATe CULLECIIOM fcFFIdl
£ OF UNIFORM TOP L/A/E/?
EQUALS tOO GPAD (LTD)
C V*)
M C Y
6E
/O
o
5
O
;V- ^ i
O "~
I""
O O
°
,
I-
1
1
5
^
0 o
O "~"
a
-a
V?
1
VI
O'o
*^~ ,
'Cn >•
Q ^_
J"«
o "
o *
a-
fl.D
IP STEADV-
5TATE FLOU)
Figure 5-6.
Comparison of steady-state leachate collection
efficiencies of LDCRS with compacted soil and composite
bottom liners (ID steady-state analysis).
5-20
-------
KATE OF
COLLECTION! EF Fldl £ M C
TOP L//VE£
1000 GPDCLTb)
100
' *2
v^N »™»
•*- >
,j V.I V) i
ts O
£ ^/
^ v» 0— '
O ~ |V
:t- 7 ^ °
vA — ^ T-^
O
I"" * "
0 0 •o
^J -^ -^ r
100 qq q qc| c|
1
"^
— «
d
~? ^
c? 5
x) o
^ ><
V* s "
o
1~"
I
-5
13
?:
V/l
SJ-
Vj'-Q
rt T-
M **
X T—
5 '
0 <
^ _<—
^ II
o -£
^1.0
71 ^I_
O 5
vi O
r*-
w o
O "
o
Q—
O y
IP
5TATE FLOU)
Figure 5-7. Comparison of steady-state leachate collection
efficiencies of LDCRS with compacted soil and composite
bottom liners (ID steady-state analysis).
5-21
-------
5TEA\DNf-STATE
FOR AM ** \ODO
LEAKAGE
COLLECT
G?PAl>
ANI>
ON
TOP LINER
L£A £
-4-
U
-V-
c
'V)
ZD TRANSi
FLOV-J
o
0 "
t
V*
"
0 0
10
Figure 5-8.
Comparison of steady-state leachate collection
efficiencies of LDCRS with compacted soil and composite
bottom liners (20 transient analysis). [Data from
Radian, 1987]
5-22
-------
W
CUMULATIVE COLLECTION EFfltl ZUCy AT
FOR AAJ -1000 GPAb(LTD) TOP
LEAKAGE RATE -AA/& UNIFORM LEAK
VJ
cs
•- O
IT
O '
78
2D
FLOW
o £
\n o
"0 '
_«J ^
tj „
O ,^_
S -
o .,
Ncrlc;
O.SV. of
held m Uic LbCttS
H
- to
Q ^
o ^~
<3 "
10
Figure 5-9.
Comparison of cumulative leachate collection efficiencies
of LDCRS with compacted soil and composite bottom liners
(2D transient analysis). [Data from Radian, 1987]
5-23
-------
CUMULATIVE COLLECTION
FOR AM
lOQO
RATE
TOP L/NEK
1 FLOW
u.
u.
UJ
•2.
o
a)
o
100
50
ZD TKANK.1ENT
FLOW
: Almost 0.57. of
top liner leafcraae '5
Ueld in
Capill««-»|
Com a c.-fed Soil
Z 3 4 5 h * B
DORM-ION OF LEAKAGE
10
Figure 5-10. Comparison of cumulative leachate collection efficiencies
of LDCRS with compacted soil and composite bottom liners
(2D transient analysis). [Data from Radian, 1987]
5-24
-------
STEADY- STATE COLLECTIOM
FO£ A^ -N- 50 6 P ATX LTD) TOP
LEAKAGE. RATE ANb SibEWALL
AT 10
o
l
H
-^L
V
u
2D TRANSIENT
FLOW
-------
CUMULATIVE COLLECTION B
FOR Ati -^ So GfAD
/?/\TE
AT 10
TOP
0
a
U)
2D TRANSIEMT
FLOW
fe*!
U
£
-------
1--
lL
O
UJ
_J
LEAKAGE OUT OF UNIT
(LEAKAGE INTO BOTTOM L
(gpad or Ltd)
V)
— £
o ^
v> ^
o
o-
£
U'o
137
o ^_
o u
O^i
ID STEADY
STATE FLOW
Vl= 0.03m (0. 1 f))
T' I.On.^ (-10 WM|
H 1 m (3-H)
s .«
«-°
0 Z.
O.OZ
TYPE OF LINER
Figure 5-13. Comparison of leakage out of the unit (or leakage into
the bottom liner) for units with compacted soil and
composite bottom liners (ID steady-state analysis).
5-27
-------
LEAKAGE OUT OF UNIT
(LEAKA66 INTO BOTTOM!
SOIL L/MER.S
~ Ltd)
w >
890
VJ S^
CD
o2
r—
* £
880
1D STEADY
STATE PLOW
H- 0.03m (0
V)
0 —
7 ^
,_ N
88
Figure 5-14. Comparison of leakage out of the unit (or leakage into
the bottom liner) with various compacted soil bottom
liners (ID steady-state analysis).
5-28
-------
LEAKAGE OUT OF UNIT
(Njo BOTTOM
(Tgpad of Ltd)
v/)
-n
S_
ID STEADY
STATE FLOW
Vl = 0 03rv> (0. 1 -f-l)
T' 1.0«^ (-10 VTM|)
K = 1 m ( 5 {-t )
o
o
V
TS
"0
c
p
4-
o
ex""
Q—
o
VJ
1.2
0.3
o.ia
Figure 5-15. Comparison of leakage out of the unit with compacted soil
and composite bottom liners. Composite bottom liners
include FMLs with defects (ID steady-state analysis).
5-29
-------
CUMULATIVE LEAKAGE OUT OF
CLEAKAGE WTO BOTTOM L/
FOR A 0.03m (0.1ff) HE/M> oN A BOTTOM LlMER.
loooor
8000 -
i- §
O ^- o
°I>
UJ i_ v
/\ * *•
Gooo -
^" I
-\b -^
\- vO *
4000
u)
O
zooo
ID STEAbV -
STATE FLOVv)
Compos i
(1 defect)
DURATION OF LEAKAGE
Figure 5-16. Comparison of cumulative leakage out of the unit (or
leakage into the bottom liner) for units with compacted
soil and composite 'bottom liners (ID steady-state
analysis).
5-30
-------
FOR A
^ s~\
°v
u-^-s
t *
E^:
o
o
o
^
«*
IU L_ i.
^o J!
2*-
UJP
-i t-
z
UJ ""
^\b
H vfl
^d
U
o
n
VI
o
10
10"
E LEAKA&e OUT OF UN
OLEAVLA&E INTO BOTTOM L/NE^O
0.03^(0.1^ HEAO OM A BOTTOM
1t> STEADY-
STATE FLOV\J
10'
DURATION! Op LEAKAGE
Figure 5-17. Comparison of cumulative leakage out of the unit (or
leakage into the bottom liner) for units with compacted
soil and composite bottom liners (ID steady-state
analysis).
5-31
-------
CUMULATIVE LE/U/\6E /A/TO
BOTTOM LIN^K. A/=ie/^ 3 MONTHS*
FOR AN*- 1000 GPADO.TZ>) TOP
RATE AND UNIFORM LEA
/acre. " //fos//000*0
31,000
— »n
0 «
U) *
u
IV
0
0 ^
O o
0 -*
2D TRANSIENT
FLOVO
o
V) 0
O U4 (/J
V) \^
r4- *
IJ'n w^
O O ^. o
ti ^ «/» r* .
0 Jl 0 V
^* ^ (l
jl° UJ
5,000
1
Figure 5-18. Comparison of leakage into the bottom liner after 3
months for units with compacted soil and composite bottom
liners (2D transient analysis). [Data from Radian, 1987]
5-32
-------
AM
CUMULATIVE LEAKAGE INTO THE
" L/A;F* AFTE/e 10 ^EA;?S
7000 .GP4D art.) TOP UAJER
RA7Z AMD UtilFOKM LEAK
(ga 110ns /acre »»•
"
o
a
£ "
O v
2D TRANSIENT
FLOW)
5
-4- *--
O K
-------
CUMULATIVE LEAJCA6E 1A/7D THE BoTTOKI
FOR AM MOOO 6PADCLTD) TOP
LEAKA6E RATE AND UNIFORM
llpns/atvc. ^ /»f<2vS
IU
I
o
CO
UJ
u
3.0
MJ v> 3ft
\n " *"V
10
ZD TRAMSIENT
FLOvJ
CompQcfed Soi
kc= "6
3 4- 5 6 7- Q
DURATION OF L.E.AKA6E
Figure 5-20. Comparison of leakage Into the bottom liner out of the
unit with compacted soil and composite bottom liners
(2D transient analysis). [Data from Radian, 1987]
5-34
-------
CUWULATWE
BOTTOM L/WE* AFT£/e
AM ^-50 GPAD CLTD)
/?AT£ AA/D
MONTHS
L/NE&
LE/U
3ODO
— v/l
'o £
Compacf ec/
fce- / « »o-
ZD TRANSIENT
FLOVO ,
I&OO
H-_
~0'tf
-J-*
O V "
OC-J
00
"t^ *""
w n
JA
Figure 5-21. Comparison of cumulative leakage into the bottom liner
for units with compacted soil and composite bottom
liners. Composite bottom liners includes an FML with a
major defect (2D transient analysis). [Data from Radian,
1987]
5-35
-------
CUMULATIVE LEAKAGE /A/TO
THE BOTTOM LINER AnEg 10
50 GPA&CLTD) TOP UNIR
LEA
-.- . *f —
(go lions/acre T lifer-S/IOOO m1.
o
v
ZD
.FLOVO
1C.O.OOO
o
V» 0
o r
S."
o
o
a
0
47.000
Figure 5-22. Comparison of cumulative leakage into the bottom liner
for units with compacted soil and composite bottom
liners. Composite bottom liners includes an FML with a
major defect (20 transient analysis). [Data from Radian,
1987]
5-36
-------
CHAPTER 6
CURRENT PRACTICE
IN LINING SYSTEM DESIGN
-------
6.1 INTRODUCTION
This chapter presents a summary of current practice in lining
system design of hazardous waste landfill and surface impoundment
units regulated under RCRA (40 CFR Parts 264 and 265). This summary
is from a survey conducted by EPA on the number of hazardous waste
landfill and/or surface Impoundment units for which permit
applications were received since November 8, 1984, having either
compacted soil bottom liners or composite bottom liners in their
double liner system design. The survey was conducted in January and
February 1987.
6.2 RESULTS OF EPA SURVEY
This section discusses the results of the survey of hazardous
waste permit applications by EPA region. It includes the questions,
the data and the assumptions.
6.2.1 Information on the Survey
6.2.1.1 Reason for the Survey
In developing the final rule for double liner and leachate
collection systems for landfills and surface impoundments, the Land
Disposal Branch of the Office of Solid Waste sought to determine
whether eliminating the option of a compacted soil bottom liner in the
double liner system design would have an adverse Impact on the
regulated community.
6.2.1.2 Quest1ons_ln_the_Survey
Each of the EPA's ten regions were asked a set of questions in a
memorandum from the Director of the Office of Solid Waste dated
January 20, 1987. The questions were:
6-1
-------
1. How many new landfill and surface Impoundment units
are Included 1n permit applications submitted to your
region?
2. How many new landfill and surface impoundment units
have double liners where the lower liner 1s a
compacted soil (clay) Hner?
3. How many new landfill and surface Impoundment units
have double liners where the lower Hner 1s a
composite (a FML on top of a compacted soil) Hner?
4. How many new landfill and surface Impoundment units
do not have detailed plans that provide enough
information to make this determination?
6.2.1.3 Results of the Survey . .....
JA.M'l^//
The results of the survey are shown 1n Taple 6-1. In the 10 EPA
Regions, 183 applications for hazardous wastej^or surface Impoundment
units had been received. Of those applications:
* 7 units used compacted soil bottom liners 1n their designs;
• 152 units used composite bottom liners in their designs;
• there was not enough Information on 24 units to make a
determination on the type of bottom liner.
Therefore, of the 159 units for which permit applications have
been filed since November 8, 1984 and for which sufficient Information
was provided:
• 95.6 percent of the units used composite bottom liners in the
double Hner system design;
• 4.4 percent of the units used compacted soil bottom liners 1n
the double liner system design.
6-2
-------
6.3 CONCLUSIONS
6.3.1 Owners or Operators Opted for Composite Bottom Liner
The survey conducted by EPA on current double Uner system design
practices, as Indicated by permit applications for hazardous waste
surface impoundment and landfill units, shows owners or operators at
these units have overwhelmingly opted to utilize a composite bottom
liner. The state of practice in double liner system design for
hazardous waste management facilities today is to use composite bottom
liners in double liner systems.
6.3.2 Assessment of Adverse Impact
The survey indicates there will be minimal impact on the regulated
community if composite bottom liners are required in double liner
system design. The total number of units for which permit applications
have been submitted since November 8, 1984 which do not have composite
bottom liners is estimated to be 7, or less than 4 percent (although
this number may be slightly higher, depending on the determination of
design for those units in the undetermined category).
6-3
-------
Table 6-1. Summary of February 1987 EPA survey of current practice for
design of bottom liners at hazardous waste management units
for which permit applications have been submitted since
November 8, 1984.
No. of Compacted Soil
Region Units Bottom Liner
Region I 8 0
(Boston)
Region II
(New York)
Region III
(Philadelphia)
Region IV
(Atlanta)
Region V
(Chicago)
Region VI
(Dallas)
Region VII
(Kansas City)
Region VIII
(Denver)
Region IX
(San Francisco)
Region X
(Seattle)
5
4
20
85
15
11
22
7
6
0
0
0
3
3
0
1
0
0
Composite
Bottom Liner
0
4
4
20
68
12
11
21
7
5
Undetermined
8
1
0
0
14
0
0
0
0
1
TOTALS
183
152
24
6-4
-------
CHAPTER 7
SUMMARY
-------
7.1 SUMMARY OF COMPARATIVE PERFORMANCE
Based on comparisons of performance of compacted soil and
composite bottom liners, the following observations were drawn in
Chapter 5.
Leak detection sensitivity - The theoretical leak
sensitivity of an LDCRS with a properly designed and
constructed composite bottom liner is much less than one Ltd
(gpad). A few "standard" FML defects have a negligible effect
on the detection sensitivity of a lining system with a
composite bottom liner. By comparison, the leak detection
sensitivity of a compacted soil bottom liner is on the order
of 100 Ltd (gpad) with kc - 1 x 10~7 cm/s in all areas of the
liner.
Leachate collection efficiency - The theoretical steady-state
leachate collection efficiency for a composite bottom liner
with an intact FML is in excess of 99'/., even for relatively
small leakage rates such as 20 Ltd (gpad). Just as important,
the collection efficiency remains high even when the FML
component of a composite bottom liner has several "standard"
defects (more defects than would be expected in a properly
designed and constructed lining system). In contrast, the
theoretical steady-state collection efficiency of a compacted
soil bottom liner is zero for all rates of uniform top liner
leakage up to approximately the leak detection sensitivity
(on the order of 100 Ltd (gpad) for a compacted soil bottom
liner with kc - 1 x 10"7 cm/st and on the order of 1000 Ltd
(gpad) for a compacted soil bottom liner with kc - 1 x 10 *
cm/ s ) .
Leakage into and out of the bottom liner - The theoretical
steady-state leakage out of a unit with a composite bottom
liner with an intact FML Is much less than 1 Ltd (gpad).
Just as important, the leakage out of the unit remains less
than 1 Ltd (gpad) even when the FML component of a composite
bottom liner has several "standard" defects (more defects
than would be expected in a properly designed and constructed
lining system). In contrast, for a uniform hydraulic head on
7-1
-------
the bottom liner of 0.03 m (0.1 ft), the leakage out of a
unit with a compacted soil bottom Uner with kc » 1 x 10~7
cm/s 1s on the order of 100 Ltd (gpad).
Based on the above summary, It can be concluded that properly
designed and constructed composite liners Incorporating a FML upper
component and a compacted soil lower component represent current best
demonstrated available technology (BOAT). It 1s believed that a
composite bottom liner, used 1n conjunction with a properly designed
and constructed double liner system, can come very close to meeting
the goal of Congress and EPA of preventing migration of hazardous
constituents through the lining system and Into the ground. In
contrast, compacted soil bottom liners will significantly limit the
migration of leakage through them, but they do not provide a level of
performance comparable to, or even close to, composite bottom liners,
and therefore do not represent the best demonstrated available
technology (BOAT).
7-2
-------
REFERENCES
-------
I. CITED REFERENCES
Auvinet, G. and Espinosa, J., "ImpermeabiUzation of a 300-Hectare
Cooling Pond", Permeability and Groundwater Containment Transport.
ASTM STP 746, T.F. Zimmie and C.O. Rlggs, Eds., American Society
for Testing and Materials, Philadelphia, PA, 1985, pp. 151-167.
Bass, J.M., Lyman, W.J77~""3md~.Tratnyek, J.P., "Assessment of Synthetic
Membrane Successes and Failures at Waste Storage and Disposal
Sites". Report to EPA, NTIS No. PB85-245 637/AS, Arthur D. Little
Inc., 1985.
Bowders, J.J., Daniel, D.E., Broderjckj G.P., and Liljestrand, H.M.,
"Methods for Testing the Compatibility of Clay Liners with
Landfill Leachate", Hazardous and Industrial Solid Waste Testing:
Fourth Symposium. ASTM STP 886, O.K. Petros, Jr., W.J. Lacy, and
R.A. Conway, Eds., American Society for Testing and Materials,
Philadelphia, PA, 1986, pp. 233-250.
Brown, K.W., Green, J.W., and Thomas, J.C., "The Influence of Selected
Organic Liquids on the Permeability of Clay Liners", Proceedings
of the Ninth Annual Research Symposium. Ft. Mitchell, KY, May
1983, pp. 114-125.
Brown, K.W., Thomas, J.C., Lyhon, R.L., Jayawickrama P., and Bahrt,
S.C., "Quantification of Leak Rates Through Holes in Landfill
Liners". USEPA Report CR 810940, Cincinnati, OH, [DATE MISSING ON
THE REPORT].
Daniel, D.E., "Predicting Hydraulic Conductivity of Clay Liners",
Journal of Geotechnical Engineering. American Society of Civil
Engineers, Vol. 110, No. 2, Feb 1984, pp. 285-300.
Daniel, D.E., Trautwein, S.J., and McMurtry, D.C., "A Case History of
Leakage from a Surface Impoundment", Seepage and Leakage from Dams
and Impoundments. American Society of Civil Engineers, 1985, pp.
220-235.
Day. S.R. and Daniel, D.E., "Hydraulic Conductivity of Two Prototype
Clay Liners", Journal of Geotechnical Engineering, Vol. 11, No. 8,
Aug 1985, pp. 957-970.
Faure, Y.H., "Nappes Etanches; Debit et Forme de TEcoulement en Cas
de Fuite". Thesis, University of Grenoble, France, Dec 1979,
263 p.
-------
Faure, Y.H., "Design of Drain Beneath Geomembranes: Discharge
Estimation and Flow Patterns 1n the Case of Leak", Proceedings of
the International Conference on Geomembranes. Vol. 2, Denver, CO,
Jun 1984, pp. 463-468.
Fernandez, F. and Qulgley, R., "Hydraulic Conductivity of Natural
Clays Permeated with Simple Liquid Hydrocarbons", Canadian
Geotechnical Journal. Vol. 22, 1985, pp. 205-214.
Fukuoka, M., "Outline of Large Scale Model Test on Waterproof
Membrane". Unpublished Report, May 1985, 24 p.
Fukuoka, M., "Large Scale Permeability Tests for Geomembrane-Subgrade
System", Proceedings of the Third International Conference on
Geotextlles. Vol. 3, Vienna. Apr 1986. pp. 917-922.
GCA Corporation,. "SolHner Model - Documentation and User's Guide
(Version 1)". EPA/530-SW-86-006, U.S. Environmental Protection
Agency, Cincinnati, OH, Apr 1986.
Ghassemi, M., Haro, M., Metzger, J., Powers, M., Quinlivan, S.,
Scinto, L., and White, H., "Assessment of Technology for
Constructing and Installing Cover and Bottom Liner Systems for
Hazardous Waste Facilities; Volume II. Technical Analysis". Final
Report, EPA Contract No. 68-02-3174, Work Assignment No. 109,
Washington, D.C., Apr 1983, 102 p.
Giroud, J.P. and Frobel, R.K., "Geomembrane Products", Geotechnical
Fabrics Report. Fall 1983, pp. 38-42.
Giroud, J.P., "Geotextlle and Geomembranes - Definitions, Properties
and Design". Industrial Fabrics Association International, St.
Paul, MN, 1984a, 325p.
Giroud, J.P. and Bonaparte, R., "Waterproofing and Drainage:
Geomembranes and Synthetic Drainage Layers", R.I.L.E.M. Symposium
No. II. Liege, Belgium, Jun 1984b.
Giroud, J.P., "Impermeability: The Myth and a Rational Approach",
Proceedings of the International Conference on Geomembranes. Vol.
1, Denver, CO, Jun 1984c, pp. 157-162.
Giroud, J.P. and Stone, J.L., "Design of Geomembrane Liner for the
Proton Decay Experiment", Proceedings of the International
Conference on Geomembranes. Vol. 2, Denver, CO, Jun I984d, pp.
469-474.
-------
Glroud, J.P., "Geomembrane Liner: Accidents and Preventive Measures",
Proceedings of the International R.I.L.E.M. Symposium on "Plastic
and Rubber Waterproofing in Civil Engineering". Session 4, Liege,
Belgium, Jun 1984e.
Giroud, J.P., "Assessment of Synthetic Membrane Performance at Waste
Disposal Facilities". Contract 68-03-1772, Woodward-Clyde
Consultants, Chicago, IL, Nov 1984f, 59p.
Giroud, J.P. and Fluet, J.E., Jr., "Quality Assurance of Geosynthetic
Lining Systems", Journal of Geotextiles and Geomembranes. Vol. 3,
No. 4, 1986, pp. 249-287.
Giroud, J.P., Bonaparte, R., and Beech, J.F., "Leakage Through
Geomembrane~rfners". to be published, 1987.
Goodal, D.C. and Quigley, R.M., "Pollutant Migration from Two Sanitary
Landfills Near Sornla, Ontario", Canadian Geotechnlcal Journal.
Vol. 14, 1977, p. 223.
Gordon, B.B. and Forrest, M., "Permeability of Soils Using
Contaminated Permeant", Permeability and Groundwater Contaminant
Transport. ASTM, STP 746, T.F. Zimmie and C.O. Riggs, Eds.,
American Society for Testing and Materials, Philadelphia, PA,
1981, pp. 101-120.
Gordon, M.E., Huebner, P.M., and Kmet, P., "An Evaluation of the
Performance of Four Clay-Lined Landfills in Wisconsin",
Proceedings of the Seventh Annual Madison Waste Conference.
Madison, WI, 1984, pp. 399-460.
Griffin, R.A., Cartwright, K., Dumontelle, P.B., Follmer, L.R., Stohr,
C.J., Johnson, T.M., Killey, M.M., Hughes, R.E., Herzog, B.L., and
Morse, W.J., "Investigation of Clay Soil Behavior and Migration of
Industrial Chemicals at Wllsonvllle. Illinois", Proceedings of the
Ninth Annual Research Symposium. Ft. Mitchell, KY, May 1983, pp.
70-79.
Griffin, R.A., Herzog, B.L., Johnson, T.M., Morse, W.J., Hughes, R.E.,
Chou, S.F.J., and Follmer, L.R., "Mechanisms of Contaminant
Migration Through a Clay Barrier — Case Study, Wilsonville,
Illinois", Proceedings of the Eleventh Annual Research Symposium.
Cincinnati, OH, Apr 1985, pp. 27-38.
Hamilton, J.M., Daniel, D.E., and Olsen, R.E., "Measurement of
Hydraulic Conductivity of Partially Saturated Soils", Permeability
and Groundwater Contaminant Transport, ASTM STP 746, T.F. Zimmie
and C.O. Riggs, Eds., American Society for Testing and Materials,
Philadelphia, PA, 1981, pp. 182-196.
-------
Haxo, H.E., Mledema, J.A., and Nelson, N.A., "Permeability of
Polymeric Membrane Lining Materials", Proceedings of the
International Conference on Geomembranes, Denver, CO, Jun 20-24,
1984, pp. 151-156.
Holtz, R.D. and Kovacs, W.D., "An Introduction to Geotechnical
Engineering". Prentice-Hall, Inc., Englewood Cliffs, New Jersey,
1981, 733 p.
Kastman, K.H., "Hazardous Waste Landfill Geomembrane: Design,
Installation and Monitoring", Proceedings of the International
Conference on Geomembranes. Vol. 1, Jun 1984, pp. 215-220.
Lambe, T.W., "Compacted Clay: Engineering Behavior" Journal of Sol 1
Mechanics and Foundations Division. Vol. 125, Part I, May 1958,
p. 718.
Mitchell, D.H., "Technology for Uranium Mill Ponds Using
Geomembranes". NUREG/CR-3890, PNL-5164, Prepared for the U.S.
Nuclear Regulatory Commission, Washington, D.C., Dec 1984.
Peirce, J.J. and Peel, T.A., "Effects of Inorganic Leachates on Clay
Soil Permeability", Proceedings of the Eleventh Annual Research
Symposium. Cincinnati, OH, Apr 29 - May 1, 1985, pp. 182-189.
Radian Corporation, "Technical Data Summary; Hydraulic Performance of
Minimum Technology Double Liner Systems". Final Report, USEPA
Contract No. 68-01-7310, Task 7.4, Austin, TX, Mar 1987, 64 p.
Rogers, C.E., "Engineering Design for Plastics". E. Baer, Relnhold
Publ. Corp., New York, 1964, pp. 609-688.
Rogowski, A.S., "Hydraulic Conductivity of Compacted Clay Soils",
Proceedings of the Twelfth Annual Research Symposium. Cincinnati,
OH, Aug 1986, pp. 29-39.
Sherard, J.L., "The Upstream Zone in Concrete-Face Rockfill Dams",
Proceedings of a Symposium on Concrete-Face RockfUl Dams -
Design. Construction. and Performance. Sponsored by the
Geotechnical Engineering Division of the American Society of Civil
Engineers, J.B. Cooke and J.L. Sherard, Eds., Detroit, MI, Oct
1985, pp. 618-641.
Stone, J.L., "Leakage Monitoring of the Geomembrane Liner for the
Proton Decay Experiment", Proceedings of the International
Conference on Geomembranes. Vol. 2, Jun 1984, pp. 475-480.
-------
Telles, R.W., Unger, S.L., and Lubowltz, H.R., "Technical
Considerations for De Mlnimls Pollutant Transport by Polymeric
Liners". Contract No. 68-03-3218, U.S. Environmental Protection
Agency, Cincinnati, OH, Sep 1986, 80 p.
Terzaghi, K. and Peck, R.S., "Soil Mechanics In Engineering Practice".
John Wiley & Sons, New York, 1967, 729 p.
USEPA, "Minimum Technology Guidance on Double Liner Systems for
Landfills and Surface Impoundments — Design, Construction, and
Operation". Draft Second Version, EPA 530-SW-85-012, U.S.
Environmental Protection Agency, Cincinnati, OH, May 1985, 71 p.
II. ADDITIONAL REFERENCES - NOT CITED
Boutwell, G.P. and Donald, V.R., "Compacted Clay Liners for Industrial
Waste Disposal", Presented at the ASCE National Meeting. Las
Vegas, NV, Apr 1982.
Boynton, S.S., "An Investigation of Selected Factors Affecting the
Hydraulic Conductivity of Compacted Clay". M.S. Thesis, University
of Texas, Geotechnical Engineering Thesis GT83-4, Geotechnical
Engineering Center, Austin, TX, 1983, 79 p.
Boynton, S.S. and Daniel, D.E., "Questions Concerning Hydraulic
Conductivity of Compacted Clay", Journal of Geotechnical
Engineering. Vol. Ill, No. 4, 1985.
Day, S.R., "A Field Permeability Test for Compacted Clay Liners". M.S.
Thesis, University of Texas, Austin, TX, 1984, 105 p.
Herzog, B.L. and Morse, W.J., "A Comparison of Laboratory and Field
Determined Values of Hydraulic Conductivity at a Waste Disposal
Site", Proceedings of the Seventh Annual Madison Waste Conference.
University of Wisconsin-Extension, Madison, WI, 1984, pp. 30-52.
Harrop-Wil1iams, K., "Clay Liner Permeability: Evaluation and
Variation", Journal of Geotechnical Engineering. ASCE, Vol. Ill,
No. 10, Oct 1985, pp. 1211-1225.
Rogowski, A.S., "Effectiveness of a Compacted Clay Liner in Preventing
Ground Water Contamination", Proceedings of the Fifth National
Symposium and Exposition on Aquifer Restoration and Ground Water
Monitoring. Columbus, OH, May 1985, pp. 412-429.
-------
Rogowskl, A.S., WeinMch, B.E., and Simmons, D.E., "Permeability
Assessment In a Compacted Clay Liner", Proceedings of the Eighth
Annual Madison Waste Conference. Department of Engineering and
Applied Sciences, University of Wisconsin-Extension, Madison, WI,
Sep 1985, pp. 315-336.
Rogowskl, A.S. and Richie, E.B., "Relationship of Laboratory and Field
Determined Hydraulic Conductivity 1n Compacted Clay Soils",
Proceedings of the Mid-Atlantic Industrial Waste Conference.
University Park, PA, 1984, pp. 520-533.
-------
APPENDIX A
CASE HISTORIES OF COMPACTED SOIL
LINING SYSTEM PERFORMANCE
-------
A.I OVERVIEW OF CASE HISTORIES
This Appendix summarizes the documented performance of compacted
soil lining systems for both landfills and surface impoundments. The
landfill facilities reviewed were primarily for the containment of
sanitary waste, however, one case history described the performance of
a clay lining system at a hazardous waste landfill. Another case
history reports the results from a large scale field test. The
surface impoundment case histories describe facilities which were
constructed to hold fresh water, salt water (brine), or contaminated
liquid. The landfill case histories are summarized in Section A.2 and
the surface impoundments are summarized 1n Section A.3.
A. 2 LANDFILLS
A.2.1 Sanitary Landfill Sarnla. Ontario
Goodall and Quigley [1977] studied cation migration from two
sanitary landfills near Sarnia, Ontario, Canada. These two sites are
referred to as th-e Confederation Road site and the Blackwell Road
site. Only the performance of the lining system at Confederation Road
Site will be summarized here.
At the location of the Confederation Road landfill, approximately
41 m (135 ft) of water-laid glacial till overlie Devonian limestone
and shale. The glacial till consist primarily of silty clay with 40-
50% clay sized particles. The till deposit 1s slightly
overconsolidated except for a desiccated crust above 7 m (23 ft).
Fissuring has developed as a result of desiccation and the fissure
spacing decreases with depth.
The landfill was excavated in the natural till to a depth of 5.6 m
(18 ft) which corresponds to the upper boundary of the undesiccated
soil. The landfill started operation 1n 1967 and was closed in 1971.
Selected cation (calcium, magnesium, sodium, and potassium)
concentrations were measured beneath the Confederation Road site and
A-l
-------
compared with background concentrations. From this comparison, It was
estimated that the cations had migrated approximately 300 mm (1 ft)
Into the natural clay below the landfill.
Typical values of sllty clay till hydraulic conductivity In the
vicinity of the Sarnla landfills measured using a variety of test
methods are shown In Table A-l. An exploration of the site revealed
a downward hydraulic gradient of 0.16 to 0.25 beneath the landfill.
Using this range of hydraulic gradient, hydraulic conductivities of 2
to 3 x 10"10 m/s (2 to 3 x 10~' cm/s) which are based on the data 1n
Table A-l, a porosity of 0.3, and a period of 6 years, Goodall and
Qulgley [1977] calculated an expected cation migration distance of 30
to 50 mm (1 to 2 In.), which 1s less than the estimated cation
migration distance. Goodall and Qulgley [1977] applied the concept of
molecular diffusion to the Confederation Road site to explain the
observed distance of 300 mm (1 ft) in 6 years.
If the calculations of Goodall and Quigley [1977] are repeated
using hydraulic conductivities of 1 x 10~f m/s (1 x 10"? cm/s) and 1 x
10"' m/s (1 x 10~" cm/s) cation migration distances due to seepage of
approximately 120 to 1200 mm (5 to 50 in.) are calculated, which bound
the observed cation migration distance. These hydraulic
conductivities are larger than those found in Table A-l. However,
they may be attributable to fine desiccation cracks existing in the
soil. Such cracks exist 1n a vertical direction and may not be
accounted for in a borehole hydraulic conductivity test.
A.2.2 Four Compacted Clay Lined Landfills 1n Wisconsin
Gordon et al. [1984] documented the performance of four clay-Hned
landfills in Wisconsin, three of which are described here. The three
facilities are for the containment of sanitary waste and all three
lining systems were composed of a granular leachate collection system
over a compacted clay Hner. The compacted clay liners are on the
order of 1.2 to 1.5 m (4 to 5 ft) thick. One of the landfills had
been in operation for 8 years, the other two were in operation for 4
or 5 years. The landfill which was In operation for 8 years indicated
leakage through the bottom of the clay liner after 8 years. The
A-2
-------
liners that had been 1n operation about 5 years showed no sign of
leakage.
•
At the 8 year old site, the 1.2 to 1.5 m (4 to 5 ft) thick
compacted clay liner has prevented the migration of pollutants for not
more than 8 years. Back calculations based on saturated flow and a
breakthrough time of 8 years give a clay liner hydraulic conductivity
of approximately 5 x 10~* m/s (5 x 10~7 cm/s).
A.2.3 Hazardous Waste Landfill 1n Illinois
Griffin et al. [1983] and Griffin et al. [1985] presented the
results of a study of organic contaminants migration at a hazardous
waste landfill. Hazardous waste was burled in 26 trenches. A
compacted clay liner was used 1n at least one of the trenches, but for
the most part the operation relied upon the natural clay tills at the
site to contain the waste. Routine monitoring of wells revealed that
organics had migrated as far as 15 m (50 ft) 1n a three year period,
which is 100 times to a 1000 times faster than anticipated. These
anticipated times were based on hydraulic conductivities measured in
the laboratory. Subsequent field hydraulic conductivity tests
indicated 1n situ permeabilities which were one to two orders of
magnitude larger than those measured 1n the laboratory. The
laboratory hydraulic conductivity of the soil layer with the highest
degree of contamination ranged from 3.3 x 10~11 to 2.7 x 10~10 m/s
(3.3 x 10"' to 2.7 x 10"' cm/s), while the field hydraulic
conductivities ranged from 8.4 x 10~10 to 2.5 x 10"' m/s (8.4 x 10"'
to 2.5 x 10"' cm/s). The results show the types of variations between
field and laboratory measured hydraulic conductivities and provide an
indicator of the effect of scale on this parameter.
A.2.4 Field Scale Test Liner
Rogowski [1986] reported on results of a field scale research
facility constructed to evaluate the hydraulic conductivity of
compacted soil liners. The facility enables construction of a
compacted soil liner 9.1 m (30 ft) by 22.9 m (75 ft) in area and 0.3-
m (1-ft) thick. A set of collector drains are situated beneath the
A-3
-------
liner and a set of 250 Infiltration, cylinders are located on top of
the Uner. Compacted clay liners can be constructed 1n the facility
using standard field equipment and procedures. Variables measured
during testing Include Infiltration, exflltratlon, and compacted soil
density.
The soil used by Rogowskl to construct the clay Uner was defined
as a cherty silt loam composed primarily of 11 lite and kaolinite, with
small amounts of montmorillonlte. The laboratory hydraulic
conductivity of this soil measured In a falling head permeameter is
about 1 to 2 x 10~l° m/s (1 to 2 x 10"g cm/s). Shortly after the
start of the field test (which Involved the ponding of water on top of
the 11ner)t exflltratlon out of the bottom of the Uner was observed.
The rate of exflltratlon Increased steadily during the first 7 to 8
months after ponding and then began to decline. Based on Infiltration
rates after 9 months, and measured hydraulic gradients, hydraulic
conductivity values were calculated. The results of this calculation
are provided 1n Figure A-l. The conductivities of 10"7 m/s (10"*
cm/s) are believed to have been Influenced by the proximity of the
soil to the edge of the liner box. It can be seen that the remaining
hydraulic conductivities range over several orders of magnitude, from
about 10"* to 10"10 m/s (10~* to 10"' cm/s). It 1s observed that this
range of hydraulic conductivities has a lower limit corresponding to
the hydraulic conductivity measured In the laboratory and an upper
limit several orders of magnitude larger than the hydraulic
conductivity measured In the lab. As of this date, no clear
explanation has been obtained for the observed distribution of
calculated hydraulic conductivities.
A.3 SURFACE IMPOUNDMENTS
A.3.1 Three Surface Impoundments 1n Texas
Daniel [1984] documented four case histories where laboratory
hydraulic conductivity tests underpredlcted the hydraulic
conductivities back calculated from measurements of seepage through
compacted soil liners. The surface Impoundments were used to retain
fresh water, salt water, or contaminated liquid. Three of the case
A-4
-------
histories are summarized here 1n this section, the fourth is
summarized in Section A.3.2.
A. 3.1.1 Iwo_Ponds_1n_Central_Texas
Two ponds, each covering 0.8 ha (2 acres) and 1.5 m (5 ft) deep,
were constructed at a manufacturing plant to hold fresh water. Each
pond was lined with a 0.3 m (1 ft) clay liner compacted in two lifts
with sheepsfoot rollers. A geotechnlcal consultant recommended, on
the basis of laboratory results (standard Proctor compaction tests and
triaxial permeability tests), that the clay liner be compacted wet of
optimum. Several days after installation of the clay liner the
geotechnical consultant was asked to Inspect the liner/ Moisture
content measurements of the clay liner at the time of inspection
revealed that it had either been compacted dry of optimum or it had
dried since the time of placement. Attempts were made to fill the
pond and the rate of leakage through the liner was several hundred
times larger than anticipated. The hydraulic conductivities of the
clay liners were back calculated from the observed rates of leakage
and were found to be on the order of 2 x 10~* to 5 x 10"1 m/s (2 x
10"' to 5 x 10~* cm/s). To reduce the leakage rate it was necessary
to drain the pond, then remove and recompact the clay liner wet of
optimum (field moisture content measurements were not taken). Water
was pumped into the ponds within 30 minutes of recompaction to prevent
desiccation. This procedure reduced the leakage rate enough so that
the pond could be filled to the design depth and operated
satisfactorily. The back calculated hydraulic conductivity of the
recompacted liner was estimated to be 5 x 10"' m/s (5 x 10"* cm/s).
A.3.1.2 Evaporation Pond 1n North_Texas
A 10 ha (25 acre) evaporation pond was constructed at a power
plant by mixing bentonite into the upper 200 mm (8 in) of natural soil
and recompacting this mixture. The pond went Into operation in 1970
and by 1978 it was apparent that leakage from the pond had
contaminated nearby wells. The average rate of leakage over the eight
year period was back calculated to be approximately 40 000
liters/lOOOmVday (40,000 gallon/acre/day). The hydraulic conductivity
A-5
-------
was back-calculated to be approximately 3 x 10~4 m/s (3 x 10~* cm/s).
Samples of the soil were recompacted 1n the laboratory and the
hydraulic conductivity was measured using water from the pond as the
permeant. The laboratory measured hydraulic conductivities ranged
from 2 x 10"f to 2 x 10"7 m/s (2 x 10~7 to 2 x 10"' cm/s).
The pond was taken out of service and a new pond was constructed
with a compacted soil Uner consisting of three 150 mm (6 1n) thick
lifts of a mixture of bentonite and local soil. The performance of
the new lining system was not documented, however, collection wells
were Installed to recover the contaminated ground water.
A.3.1.3 §Mne Ponds 1n Southern Texas
Two ponds (referred to as eastern and western ponds) were
constructed at a chemical plant for the purpose of retaining a 25%
brine solution. The ponds were constructed by excavating to a depth
of 1.5 m (7 ft) below the ground surface, and then lining the
excavations with 0.6 m (2 ft) of compacted clay. The construction
procedure was poorly documented. The ponds were not put Into service
until two years after construction.
Contamination was detected in a nearby monitoring well within one
month of putting the eastern pond Into service. The brine was
transferred to the western pond. The compacted clay liner in the
eastern pond was removed and recompacted, however, this effort had no
effect on the performance of the pond. During reconstruction of the
eastern pond the western pond was found to be leaking. Monitoring of
the western pond over a six month period provided data from which a
hydraulic conductivity of 1 x 10~7 to 2 x 10~7 m/s (1 x 10~' to 2 x
10~* cm/s) was back calculated. This range of hydraulic conductivity
is approximately two orders of magnitude larger than those measured on
laboratory tests performed on undisturbed samples taken from the ponds
just prior to filling with brine. The laboratory tests were permeated
with the brine for two weeks and the measured hydraulic conductivity
was on the order of 1 to 4 x 10~* m/s (1 to 4 x 10~7 cm/s).
A-6
-------
The leakage from the eastern pond was eventually minimized by
Installing a FML, while the use of the western pond was limited.
•
A. 3.1.4 Conclusions
Several Important conclusions can be drawn from the case histories
summarized in Daniel [1984]. Drying of compacted soil liners can
result in an increase in permeability and every effort should be taken
to cover the compacted soil surface as soon as possible after
construction. Construction quality assurance should be implemented to
verify that the desired state of compaction is achieved. Drying and
improper placement can increase a soil layer's hydraulic conductivity
by up to one to two orders of magnitude.
A.3.2 Cooling Pond in Mexico
Auvinet and Esplnosa [1981] summarized the results of field and
laboratory hydraulic conductivity tests. The field tests included
permeation of a 2 m by 2 m (6 ft by 6 ft) by 0.6 m (2 ft) thick clay
liner, and a larger 50 m by 50 m (150 ft by 150 ft) test pond
surrounded by 6 m (20 ft) dikes. These test sections were
constructed as part of a study for a 300 hectare (740 acre) compacted
clay lined cooling pond.
Great care was required in the construction of the lining system
to achieve the desired hydraulic conductivity. The construction
procedure used was to first thoroughly mix the soil while adding water
(to a water content slightly greater than the optimum Proctor water
content) and allowing it to cure for approximately one week. This was
done to achieve a uniform water content throughout the soil, which is
essential for obtaining a low-permeability soil. The compaction
procedure was as follows:
11 • The surface on which the lining was to rest was watered and
recompacted with several passes of a crawler-type tractor.
After that, the surface was smoothed and sealed with passes of
a heavy (12 tonne) farm tractor.
A-7
-------
• The homogenized and cured material for the first layer was
brought 1n and spread by motorscrapers.
• A uniform 20-cm thickness was given to the loose layer with the
blade of the crawler-type tractor.
• Water was added by sprinkling to raise the water content of the
clay to about 5 percent above the Proctor optimum value.
• New passes of the crawler tractor helped to homogenize the
material.
• The clay was remolded by passes of the heavy farm tractor
(eight passes as an average).
• When the surface was sufficiently smooth to be traveled, the
same process was repeated for the next layer."
Also, special attention was given to dry spots and areas which did
not visually appear to be at the desired state of compaction.
The back-calculated hydraulic conductivity of the field test was
slightly higher than 1 x KT'm/s (1 x 10"' cm/s), while hydraulic
conductivities obtained from laboratory triaxial permeability tests on
undisturbed field samples were 10~10 m/s (10"' cm/s) or lower.
The hydraulic conductivities achieved in the field tests were
considered acceptable and the construction procedures used in the
tests were adapted to construct the main cooling ponds.
It can be concluded that even with good control on the placement
and compaction of a soil liner, the field permeability may be less
than that achieved in the laboratory.
A-8
-------
A.3,3 Two Prototype Compacted Clay Lining Systems
Day and Daniel [1985] present the results of two prototype clay
liners which were designed, constructed, and tested to measure field
hydraulic conductivity. A different clay type was used for each liner
and their properties are summarized in Table A-2. Clay 1 classified
as a low plasticity clay (CL) while Clay 2 classified as a high
plasticity clay (CH). Each clay liner tested was approximately 6-m
(20-ft) on a side and 150-mm (6-1n.) thick. Underlying the compacted
clay liner was a geotextile lateral drain. The geotextlle drain was
underlain by a FML. Any leakage through the compacted clay liner was
transmitted by the geotextlle drain to a monitoring station, where the
outflow was measured. From the measured rate of outflow, the field
hydraulic conductivities were calculated to be approximately 9 x 10~'
m/s (9 x io~* cm/s) and 4 x io~* m/s (4 x 10~* cm/s) for Clays 1 and
2, respectively. However, 1t was noted by the authors that "Based on
the average compaction water content and dry unit weight of each liner
and the results of the permeability tests on the laboratory-compacted
samples, the hydraulic conductivities that would be expected for the
liners would be approximately lx 10~10 m/s (1 x 10 "' cm/s) for Clay 1
and 2 x i(T11 m/s (2 x 10"1) cm/s) for Clay 2."
From the above results 1t can be observed that the hydraulic
conductivities measured 1n the field were approximately 1000 times
larger than that measured in the laboratory. Day and Daniel [1985]
suggested several reasons for this large difference. First, clods of
soil up to 100 mm (4 in.) were present in the field while the soil
used in the laboratory tests contained clods smaller than 30 mm (1
in.). Second, it is possible that secondary structures such as
cracks, fissures, joints, etc., were present in the field liner but
not in the lab samples. Third, zones of poor compaction may have
resulted in a clay liner of variable hydraulic conductivity.
Field hydraulic conductivity tests with the ring infiItrometer
were also performed. These tests measured a hydraulic conductivity of
the clay liner which was slightly larger but within an order of
magnitude of the hydraulic conductivity back-calculated from ponding
test.
A-9
-------
It can be concluded that laboratory testing cannot account for all
of the potential secondary structures which may exist 1n a compacted
clay liner, and large scale In situ tests seem to provide a better
Indication of the actual field hydraulic conductivity.
A.4 SUMMARY
In all cases where field and laboratory data were available, It
was found that the field measured hydraulic conductivity was up to an
order of magnitude or more higher than the laboratory measured
hydraulic conductivity. The range of values obtained and the probable
cause of the difference between the laboratory and field measured
values are summarized 1n Table A-3. Only the more conclusive case
histories have been summarized 1n this table. It can be observed that
many factors can have an affect on the performance of a compacted soil
liner. Most of these factors relate to secondary structures
(nonuniformltJes) In the compacted soil liner caused by:
• desiccation cracks;
• soil clods;
• spatial variation 1n compactlve effort and compaction water
content; and
• quality assurance of construction operation.
It should be realized that the above variables, which effect field
performance, are seldom reproduced 1n standard laboratory tests.
Therefore, achieving the desired hydraulic conductivity in the
laboratory does not guarantee the same value will be obtained in the
field. However, It 1s possible to achieve a hydraulic conductivity of
10"' m/s (10~7 cm/s) 1n the field and this is the design objective of
the EPA. Achieving this goal requires the proper soils, compaction
procedures, and construction conditions. Recognizing the number of
factors which can affect the hydraulic conductivity, and chat the
design goal is not always achieved, the analyses presented 1n Chapter
3 of the background document should consider both a standard hydraulic
A-10
-------
conductivity of 10 * m/s (10~7 cm/s) and a lower bound hydraulic
conductivity of 10~* m/s (10~* cm/s). This lower value of
conductivity might be representative of a compacted soil Uner with
some degree of secondary structure resulting from nonunlform
compaction conditions, soil clods, drying or other factors.
Therefore, these hydraulic conductivities will be used in the
investigation of detection sensitivity, collection efficiency, leakage
from the unit, and breakthrough time for compacted soil liners
presented in Chapter 3.
A-ll
-------
Table A-l. Typical values of silty clay till
conductivities. [Goodall and Quigley, 1977]
hydraul1c
Depth
(m)
1.5
1.5
1.5
2.7
5.5
5.5
5.5
6.7
6.7
7.6
8.2
9.1
II. 6
M.3
17.1
18.3
21.6
27.4
Coefficient of permeability
Method k (cm/s at 7'C)
Lab— direct, constant head
Lab — direct, constant head
Lab — Harvard miniature (remoulded)
Field — uncased auger hole
Lab — consolidation
Lab — consolidation
Lab — consolidation
Lab — consolidation
Lab — consolidation
Field — Ogunbadcjo piezometer
Field — borehole 4, Confederation
Lab — direct, falling head
Field— borehole 1, Dlackwell
Field — borehole 8, Confederation
Lab — direct, falling head
Field — Ogunbadcjo piezometer
Lab — direct, falling head
Field — Ogunbadcjo piezometer
6.3x 10-'
1.7x ID'*
l.Ox I0-7
1.8x 10-T
2.2x 10-
2.4x 10-
2.4x 10-
2.3x 10-
1.5x 10-
1.2x 10-
1.3x 10-
1.7x 10-
1.6x 10-
1.6x 10-
2.6x 10-
5.2x 10-
2.9x 10-
3.5x 10-
A-12
-------
Table A-2. Properties of clays used in construction of prototype
compacted clay liner. [Day and Daniel, 1985]
Property
(1)
Liquid limit, as a percentage
Plastic limit, as a percentage
Plasticity index, as a percentage
Percent passing No. 200 sieve
Soil classification
Clay 1
(2)
30
19
11
80
CL
Clay 2
(3)
72
27
45
50
CH
A-13
-------
Table A-3. Comparison of field
conductivities.
and laboratory hydraulic
Field Hydraulic
Conductivity
(m/s)
Laboratory Hydraulic
Conductivity
(m/s)
Probable Cause
for Difference
Reference
a.
-------
Figure A-l.
Distribution of hydraulic conductivity in 0.3 m (1.0 ft)
thick compacted clay liner after 9 months of ponding
(liner dimensions, 9.1 x 22.9 m (30 x 75 ft)).
[Rogowski, 1986]
A-15
-------
APPENDIX B
ANALYSIS OF LEAKAGE
THROUGH COMPOSITE BOTTOM LINERS
-------
B.I INTRODUCTION
B.I.I Scope
Appendix B discusses leakage through composite liners with a view
to evaluate leakage through bottom liners of hazardous waste
management units.
Leakage through composite liners should first pass through the FML
which is the upper component of the composite liner. Leakage through
a FML is due to permeation through the FML and flow through holes in
the FML. Evaluation of leakage through a composite liner due to holes
in the FML is a complex matter and, accordingly, several sections of
this appendix are devoted to this topic.
B.I.2 Organization
Appendix B is organized as follows:
• Section B.2: leakage due to permeation through the FML.
• Section B.3: discussion on frequency and size of FML defects.
• Section B.4: analytical studies related to leakage through
composite liners due to holes in the FML.
• Section B.5: model tests on composite liners with a hole in
the FML.
• Section B.6: conclusions on leakage through composite liners
specific to bottom liners.
B.2 LEAKAGE DUE TO PERMEATION THROUGH FML
B.2.1 Permeameter Tests
Tests conducted at the University of Grenoble (France) by Giroud
from 1973 to 1978 and, then, by Gourc and Faure using a permeameter
B-l
-------
similar to those used to measure clay permeability (Figure B-l) have
shown that water flows through an FML.
Results of these tests have been published by Glroud [1984a,
1984c]. In these publications, Darcy's equation has been used to
Interpret the test results and calculate equivalent hydraulic
conductivities which vary significantly with the hydraulic head (and,
consequently, the hydraulic gradient).
B.2.2 The Concept of Coefficient of Migration
It 1s preferable to Interpret the permeameter tests discussed
above using the following equation proposed by Glroud et al. [1987]:
v - Q/A - Ug/T (Equation B-l)
where: Q - flow rate due to permeation through the FML; A « surface
area of the considered FML; Q/A - flow rate per unit area; ug =
coefficient of migration of the FML; and T - FML thickness.
Recommended SI units are: 0 (m*/s), A (m2), Q/A (m/s), yg (mVs), and
T (m).
Values of the coefficient of migration for various FMLs are given
in Table B-l. Although there are not enough test results to draw a
firm conclusion, 1t appears that the coefficient of migration
increases as the hydraulic head Increases up to some maximum value,
umax* For heads larger than approximately 10 meters (30 ft), u =
umax. The value of u^x depends on the polymer used to make the FML.
•The value of u Is obviously zero for a hydraulic head equal to zero.
Therefore, the typical shape of the curve of the coefficient of
migration versus hydraulic head 1s given as shown in Figure B-2.
It is difficult to conduct water permeability tests on FMLs with a
head of water smaller than 5 m (16 ft) because the flow rates are too
small to be accurately measured. The hydraulic heads that are
relevant to hazardous waste management units are usually smaller than
5 m (16 ft). Therefore 1t is useful to complement results from the
permeameter tests cited above by results from water vapor transmission
B-2
-------
tests which are typically conducted with a pressure on the order of 1
to 10 kPa (0.15 to 1.5 psi), I.e., a hydraulic head on the order of
0.1 m to 1 m (4 in. to 40 in.).
B.2.3 Water Vapor Transmission Tests
Water vapor transmission tests are typically performed on thin
membrane materials because the mechanism for fluid transport through
membranes is believed to be one of molecular diffusion through a
nonporous membrane [Haxo et al., 1984]. With this mechanism,
transport through the membrane involves three steps: (i) dissolution
of the fluid into the membrane; (ii) diffusion of the fluid through
the membrane; and (1i1) evaporation or dissolution of the fluid on the
downstream side of the membrane. According to Haxo et al. [1984], the
major driving force for the movement of a given fluid through a
membrane 1s its concentration gradient across the membrane. In the
case of water, the Important concentration gradient is suggested to be
the water vapor pressure, and moisture Is thought to move through the
membrane by water vapor diffusion. It is important to note that water
vapor diffusion decreases when the thickness of the membrane
increases, but is not dependent on the hydraulic head acting on the
membrane.
Haxo et al. [1984] have described a water vapor transmission test
(ASTM E96, Procedure BW) and have used it to measure water vapor
transmission rates for the range of FML materials given in Table
B-2. Values of water vapor transmission rates obtained from other
sources are given 1n Table B-3.
Knowing the water vapor transmission rate of a given FML obtained
in a given test, the quantity of vapor permeating through this FML can
be calculated using Fick's equation:
M/(At) - (WVT) (T0/T) (Ap/Ap,) (Equation B-2)
where: M - mass of vapor migrating through the FML; A - FML surface
area; t - time (i.e., duration of the permeation); WVT - water vapor
transmission rate; T0 = FML thickness used in the water vapor
B-3
-------
transmission test; T - considered FML tmcicness; ap - vapor pressure
difference between the two sides of the considered FML; and Ap0 •
vapor pressure difference between the two sides of the FML used in the
water vapor transmission test. Recommended SI units are: M (kg), A
(ma), t (s), WVT (kg/m'.s), T, and T (m), and Ap and Ap. (N/m2).
(Note: 1 g/m'.day - 1.16 x 10"1 kg/m'.s).
Vapor pressure 1s given by:
p . ps H (Equation B-3)
where: ps « vapor pressure at saturated point; and H - relative
humidity.
Therefore, Equation B-2 can be written as follows:
M/(At) - (WVT) (T./T) (AH/AH.) (Equation B-4)
where: AH - relative humidity difference between the two sides of the
considered FML; AH. - relative humidity difference between the two
sides of the FML used In the water vapor transmission test; and other
notation as for Equation B-2.
It should be pointed out that the use of Equations B-2 and B-4
should be restricted to pressures that are not too different from the
pressures typically used to conduct the water vapor transmission test
(e.g., pressures on the order of 1,000 to 10,000 Pa (0.15 to 1.5 psi),
i.e., hydraulic heads on the order of 0.1 to 1 m (4 to 40 in.) of
water).
According to Pick's equation (Equation B-2), there is no
permeation through an FML 1f the relative humidity 1s the same on both
sides of the FML. This is the case, 1n particular, if there is water
on both sides, even 1f there Is a pressure difference. This is in
disagreement with results obtained using a permeameter, which were
presented 1n Section B.2.1. More research 1s therefore needed on
this subject.
B-4
-------
B.2.4 Relationships between Various Expressions of Flow Rate
In order to use water vapor transmission test results to
complement permeameter test results, 1t 1s necessary to establish
relationships between the various coefficients used to express flow
rate.
An equivalent hydraulic conductivity for FMLs can be obtained by
expressing flow rate through a FML using Darcy's equation:
v « Q/A - kg 1 (Equation B-5)
where: v - apparent velocity of the flow; 0 - flow rate; A = area
perpendicular to the flow; kg - equivalent hydraulic conductivity of
the FML; and i - hydraulic gradient.
By comparing Equation B-l with Equation B-5, it appears that:
yg - kg h (Equation B-6)
where: ug - coefficient of migration of the FML; kg - equivalent
hydraulic conductivity of the FML; and h - hydraulic head.
Recommended SI units are: ug (mVs), kg (m/s), and h (m).
By comparing Equation B-5 (Darcy's equation) with Equation B-2
(Pick's equation), it appears that:
WVT = p kg/g T - p kg h/T (Equation B-7)
By combining Equations B-6 and B-7, it comes:
WYT - p Ug/T (Equation B-8)
where: kg - FML equivalent hydraulic conductivity; g - acceleration
of gravity; T « FML thickness; WVT - FML water vapor transmission
rate; p = pressure; p = liquid density; h = hydraulic head; and pg =
B-5
-------
coefficient of migration. The recommended SI units are: kg (m/s), g
(m/s1), T (m), WVT (kg/m».s), p (Pa), p (kg/m'), h (m), and ug (mVs).
A useful conversion factor for WVT 1s:
lg/ma/day - 1.16 x NT' kg/m'/s
Using Equation B-8, the measured water vapor transmission (WVT)
values given in Tables B-2 and B-3 have been converted into values of
the coefficient of migration. It is interesting to see in Table B-2
that series of tests on a given product (e.g., series of four tests on
PVC) with various thicknesses generally give consistent values of the
coefficient of migration.
There are not enough values 1n Tables B-l, B-2 and B-3 to
establish a complete table of values of coefficient of migration, jjg,
for FMLs. It 1s therefore necessary to draw curves such as those in
Figure B-3 to make Interpolations and extrapolations for small values
of the hydraulic head. Also, Tables B-l, B-2 and B-3 contain
discrepancies and apparently erratic results due to the difficulty of
the tests and the sometimes great differences between FMLs of the same
type. Therefore some averaging was necessary. Values of the
coefficient of migration from Tables B-l, B-2 and B-3 are summarized
in Table B-4. Figure B-3 was established using values of the
coefficient of migration given 1n Table B-4.
The large discrepancy between water vapor transmission rates
measured on PVC at 0.14 m head (Table B-2) and 0.6 m head (Table B-3)
probably result from the fact that the PVC tested at 0.14 m head was a
FML made of plastlclzed PVC and the PVC tested at 0.6 m was pure PVC.
Plasticized PVC 1s swelled by the plastlclzers and tends to be more
permeable than pure PVC (such as the stiff PVC used to make bottles,
which has a very low permeability).
B.2.5 Leakage Rate Evaluation
From Figure B-3, It Is possible to establish Table B-5 which gives
our best estimate of coefficient of migration values from the analyzed
data. From Table B-5, 1t 1s possible, using Equation B-l, to
B-6
-------
establish Table B-6 which gives leakage rates due to permeation
through FMLs, assuming an FML thickness of 1 mm (40 mils).
B.2.6 Migration of Chemicals
Many types of FMLs swell when placed in contact with chemicals.
As a result, the distance between polymeric chains increases and
permeability increases. Therefore, an FML can have a low permeability
to water and a high permeability to some chemicals. Data regarding
permeation of FMLs by chemicals can be found in [Haxo et al., 1984]
and [Telles et al., 1986].
B.3 FREQUENCY AND SIZE OF FML DEFECTS
B.3.1 Purpose
The purpose of this section is to evaluate the size and frequency
of defects which can occur in a FML. This information is necessary
for making analytic calculations to evaluate leakage through top
liners (FML alone as well as composite liners). Although this
section is devoted to all types of defects, it focuses primarily on
seam defects because forensic analyses have shown that leakage
through FML liners is often due to defective seams, and the most
complete documentation of FML defects is for seam defects.
This section is organized as follows: first, data from
construction quality assurance and forensic analyses are reviewed,
then conclusions are drawn from these data.
B.3.2 Data from Construction Quality Assurance
- Small Liquid Reservoir
This project, constructed in 1981, is described in detail by
Giroud and Stone [1984], and Stone [1984]. Information regarding seam
defects can be summarized as follows.
B-7
-------
The double Uner system Includes two 2.5 mm (100 mil) thick HOPE
FMLs which were welded using an automated extrusion welder.
Ultrasonic testing, carried out as part of the quality control and
quality assurance program, showed that approximately 0.5% of the seam
length was defective. The detected defects were repaired and the
reservoir was filled with water. Leakage occurred and an inspection
showed that leakage was taking place through approximately 0.015% of
the seam length. The ratio 0.5/0.015 shows that, in this project,
intensive quality assurance divided the length of defects by
approximately 30.
This project is particularly interesting because it provides an
evaluation of the benefits from construction quality assurance.
- Large Landfill with Single Liner
Kastman [1984] Indicates that in a carefully monitored landfill
liner installation done in 1983, approximately one defect every 15 m
(50 ft) of seam was detected and repaired, as part of the quality
assurance process. The liner was a 1 mm (40 mil) thick HOPE FML and
seaming was achieved with a fillet extrusion weld done using a hand
welder.
- Large Landfill with Double Liner
Giroud and Fluet [1986] report the result of an analysis conducted
on the basis of data collected during the quality assurance process of
liner Installation in a large landfill, lined 1n 1985 with an HOPE
FML. The surface area of the liner is approximately 35 000 m2
(350,000 fta) and seam length is approximately 5000 m (16,000 ft).
During the quality assurance process, an average of approximately one
seam defect every 9 m (30 ft) of seam length was discovered and
repaired.
- Large Landfill with Single Liner
This case history presents the results of an analysis conducted on
the basis of data available in GeoServlces files. The data were
B-8
-------
collected during the Installation of the lining system 1n a large
landfill, in 1987, as part of the quality control provided by the FML
installer and quality assurance provided by an independent firm. The
surface area of the liner is approximately 53 000 m* (570,000 ft1) and
seam length is approximately 8000 m (26,000 ft). The liner was a 1.5
mm (60 mil) thick HOPE FML. Half of the seam length was welded using
a hand welder which made fillet extrusion welds; the other half was
welded using an automated flat welder. An average of approximately
one seam defect every 11.5 m (38 ft) of seam length was discovered by
the FML installer and the independent quality assurance firm. All
these defects were repaired. Seam inspection was performed first by
the installer, and then by the Independent firm after the installer
had completed his Inspection. The Installer detected approximately
one seam defect every 17 m (56 ft) of seam length. The independent
firm detected approximately one seam defect every 35 m (115 ft) of
seam length.
This project 1s Interesting because 1t provides an evaluation of
the benefits from construction quality assurance. The independent
firm discovered additional seam defects, after the installer had
completed his quality control inspection. The defects discovered by
the independent firm totaled one third of the total seam defects. The
benefits of quality assurance are probably greater than that: 1t is
probable that, without the continuous presence at the site of the
independent quality assurance firm, the FML installer would have found
fewer defects than he did as part of his quality control effort.
B.3.3 Data from Forensic Analyses
- Smal1 Indoor Tank
The following case history is reported by Giroud and Fluet [1986].
A power generating station required a small acid holding tank,
which was constructed of concrete and lined in 1985 with a high
density polyethylene (HOPE) FML which required approximately 100 m
(300 ft) of field seaming. The seams were fillet welds done with a
hand welder. The design and installation included no third party
B-9
-------
quality assurance, but careful quality control of seaming was provided
by the installer, using visual Inspection and vacuum box.
Upon completion of the liner installation, the tank was filled
with water to check for leaks. The Uner did leak, so the tank was
emptied, repairs were made and the tank was filled again. This cycle
was repeated several times, with leaks found on every filling. Leaks
were found at 15 different locations, I.e. an average of one leak per
7 m (23 ft) of seam.
- Large Surface Impoundment
The following case history 1s reported by Giroud and Fluet [1986].
A large reservoir, lined with a single reinforced chlorosulfonated
polyethylene (CSPE-R) FML, had been constructed to contain phosphoric
acid. The reservoir was approximately 3 m (10 ft) deep and its
surface area was approximately 20 000 m1 (200,000 ft1).
One year after the first filling, the reservoir suddenly emptied.
The analysis of the failure Indicated that phosphoric acid, leaking
through several defective seams, attacked the ground, creating
cavities. The largest cavity was one meter (three feet) in diameter
and half a meter (20 Inches) deep. Under the pressure of the
impounded liquid, the FML spanning this largest cavity burst,
releasing all of the impounded phosphoric add Into the ground.
Quality assurance during installation had consisted of only two
one-day visits by an engineer who specialized in roofing membranes.
Therefore, it 1s not surprising that defective seams were not detected
prior to filling.
During the forensic analysis, visual observation showed that
approximately 0.1% of the seam length was defective. It is probable
that a higher percentage would have been obtained if a vacuum box had
been used instead of the visual Inspection.
B-10
-------
B.3.4 Conclusions on Frequency of Defects
- Consistency of the Observations
Sections B.3.2 and B.3.3 present data related to frequency of seam
defects. Some of these data are expressed as an average seam length
exhibiting one defect (e.g., one defect per 7 m (23 ft) of seam),
while other data are expressed as percentage of defective seam length
(e.g., 0.5% of the total seam length was defective).
If an average length of seam defect (prior to quality assurance)
of 25 mm (1 in.) 1s considered, a percentage of defective seam length
of 0.5% 1s equivalent to one defect every 5 m (17 ft). Therefore, the
above observations appear to be consistent.
- Conclusion Regarding Frequency of Seam Defects
It is not possible to draw general conclusions from only six
cases. However, since the observations made in these six cases were
consistent it is possible to draw the following tentative conclusions:
• An average of one defect per 10 m (30 ft) of seam can be
expected without quality assurance.
• An average of one defect per 300 m (1,000 ft) of seam can be
expected with reasonably good installation, adequate quality
assurance, and repair of noted defects. (Quality assurance
followed by adequate repair drastically decreases the number of
seam defects but may not totally eliminate them.)
The average of one seam defect per 10 m (30 ft) without or before
quality assurance will probably decrease in the future as a result of
the increasing use of new, automated methods of seaming which are now
available. However, the number of seam defects after quality
assurance may not decrease significantly because, in the present state
of practice for construction quality assurance, great emphasis is put
on finding seam defects and repairing them. Nonetheless, the better
seaming methods that are now available are highly beneficial for at
B-ll
-------
least the following reasons: (1) less seam repair 1s required during
installation; (11) frequency of destructive seam testing may be
decreased; (111) quality assurance effort may shift toward other areas
where improvement is sorely needed such as connections of FMLs with
appurtenances and placement of drainage materials (which 1s essential
for the functioning of leak detection systems); and (1v) stronger
seams that are less likely to fall when subjected to stresses.
As a result of the above discussion, a frequency of one defect per
300 m (1,000 ft) of seam can be used as a working assumption. If FML
panels 6 to 10 m (20 to 30 ft) wide are used, one defect per 300 m
(1,000 ft) of seam is equivalent to 3 to 5 seam defects per hectare (1
to 2 seam defects per acre) of Installed FML.
As soon as possible. these tentative conclusions must be
supplanted and modified as required by conclusions established on a
broader base of well documented case .histories. In the meantime (and"
in the absence of better data), a frequency of one or two defects per
4000 m2 (acre) will be used in calculations for estimating leakage
rate in order to size leak detection drainage layers. This frequency
is assumed to include all types of defects, not only seam defects.
B.3.5 Estimation of Size of Defects
The seam defect documentation reported above addressed primarily
the frequency of seam defects. Extensive documentation of defect size
does not exist. On the basis of interviews with quality assurance
personnel it appears that the maximum size of defects which may still
exist after Intensive quality assurance is equivalent to hole
diameters on the order of 1 to 3 mm (0.04 to 0.12 in.) for seam
defects and maybe up to 5 mm (0.2 in.) for special areas such as
connections of FML with appurtenances.
There are also defects that cannot be observed by the quality
assurance personnel, such as: (1) puncture of the FML during
Installation of the protective earth cover; and (11) puncture of the
FML as a result of stresses due to the weight of waste or traffic
related to the operation of the hazardous waste management unit.
B-12
-------
Therefore, for design purposes It may be appropriate and conservative
to consider a hole larger than the expected size of defects at the end
of FML installation (which were estimated above as 5 mm (0.2 in.)
maximum in diameter).
B.3.6 Standard Hole Size and Frequency
For the consistency of calculations and discussions supporting the
Notice of Data Availability, it is recommended that a standard hole
size and frequency be selected. The same standard hole size and
frequency will also be useful as guidance for designers of leak
detection systems.
As a result of the above discussions, a standard hole area of 1
cm: (10"* m* or 0.16 in2.) has been selected, and, on the basis of the
discussion presented in Section B.3.4, a frequency of one standard
hole per 4000 m* (acre) is considered. The standard hol-e area and
frequency are used in this background document for calculations done
to evaluate leakage rates, and they are recommended, as well, for
design calculations.
It should be kept in mind that the standard hole size and
frequency have been selected with the assumption that intensive
quality assurance monitoring will be performed. Also, the standard
hole size and frequency do not take into account cases where design
flaws or poor construction practices would lead to many seam defects
or a large tear in the FML.
B.4 ANALYTICAL STUDIES
B.4.1 Introduction
B.4. l.l Purpose_of_the_SectU>n
This section discusses leakage through composite liners due to a
hole in the FML. The purpose of this discussion is to draw practical
conclusions regarding the evaluation of leakage rate through composite
top liners.
B-13
-------
B . 4 . i . 2
A composite Uner is comprised of a FML (which is the upper
component of the composite Hner) and a low-permeability compacted
soil layer (which 1s the lower component of the composite Hner). If
there 1s leakage through a composite Hner, the leachate first
migrates through the FML, then may travel laterally in the space, if
any, between the FML and the low-permeability compacted soil, and,
finally, migrates through the low permeability soil.
There are two mechanisms by which leakage can migrate through a
FML:
• permeation through the FML (I.e., flow through a FML that has
no defects); and
• flow through holes in the FML.
Leakage rate due to permeation through the FML should not be
significantly affected by the presence of the low-permeability
compacted soil layer under the FML because even a soil with a very low
permeability Is still very permeable as compared to a FML without
holes and pinholes. The case of permeation through a FML without
holes was discussed in Section B.2.
The leachate that has passed through the FML can flow laterally to
a certain extent between the FML and the low-permeability compacted
soil, before it migrates through the low permeability soil. This is
possible if there 1s a space between the FML and the low permeability
soil.
B.4.1.3 Organization of this Section
Two types of analytical studies can be found 1n the literature:
• analytical studies assuming that thare 1s perfect contact
between the FML and the low-permeability compacted soil, and,
B-14
-------
consequently, that the leachate does not flow laterally between
the FML and the low-permeability compacted soil; and
• analytical studies assuming that leachate flows laterally
between the FML and the low-permeability compacted soil before
it migrates through the low permeability soil.
B.4.2 Analyses Assuming Perfect Contact
- Assumptions
Faure [1979] has made an extensive study of the leakage rate
through a composite liner due to a hole 1n the FML, assuming perfect
contact between the FML and the underlying low permeability soil.
First, Faure considered two simple two-dimensional cases:
• flow net established by considering that the entire soil layer
is saturated (Figure B-4 a); and
• radial flow (Figure B-4 b) which leads to a convenient close
form solution for the leakage rate (the radial flow was thought
to be a reasonable assumption for thick soil layers, but in
fact is not, as shown by Faure (see Figure B-7)).
These two types of flow lead to absurd results (such as flow rate
increasing when soil thickness Increases). However, those cases are
useful because Faure showed that they provide upper boundaries for the
actual flow rate through the composite Uner when the FML and the
underlying soil are in perfect contact. Also the leakage rate in the
case of the radial flow is expressed by a close form solution for the
three-dimensional case (circular hole), which provides a convenient
upper boundary for the three-dimensional case. This is very useful
because the three-dimensional case is very difficult to analyze and
this upper boundary is one of the few theoretical data available for
the three-dimensional case.
A lower boundary of the leakage rate is obtained by assuming that
the flow is vertical (Figure B-4 c).
B-15
-------
The actual flow 1f the FML and the low-permeability compacted soil
are 1n perfect contact 1s shown 1n Figure B-4 d. This has been
demonstrated in the two-dimensional case by:
• Faure [1979] who used numerical methods; and
• Sherard [1985] who traced flow nets by trial and error.
Both Faure and Sherard have shown that, 1n a two-dimensional flow:
• there is horizontal flow in the soil along a portion of the
Interface (although there 1s no flow between the FML and the
soil because there Is no space between the FML and the soil
when perfect contact 1s assumed); and
• there 1s a phreatic surface beyond which the soil is not
saturated.
These qualitative characteristics of the flow are certainly also
applicable to the three-dimensional case (circular hole). Typical
flow nets for the two-dimensional case are given in Figure B-5 and a
chart giving the location of the phreatic surface In the two-
dimensional case Is presented 1n Figure B-6.
- Leakage Rates for the Two-dimensional Case
Leakage rates obtained with the various assumptions discussed
above are given 1n Figure B-7 adapted from Faure. This figure shows
that:
• absurd results are obtained with the upper boundaries, cases
(a) and (b), when the low-permeability compacted soil
thickness, H, is large; and
• case (c) Is a very low lower boundary when the low-permeability
compacted soil thickness, H, is large.
B-16
-------
A chart giving the actual leakage rate (I.e., the leakage rate
obtained in case d) when the FML and the underlying soil are 1n
perfect contact has been prepared by Faure [1979, 1984] for the two-
dimensional case (Figure B-8). The results given by Sherard [1985]
for a limited number of cases are consistent with Faure's. Faure's
chart (Figure B-8) 1s used with the following equation:
Q/B - C kc (H + h) (Equation B-9)
where: Q - leakage rate; B - length of the slot in the direction
perpendicular to the figure; Q/B • leakage rate per unit length; C =
dimensionless coefficient given by the chart; kc - hydraulic
conductivity of the low-permeability compacted soil underlying the
FML; H - thickness of the low-permeability compacted soil; and h =
hydraulic head on top of the FML.
The equation .for the two-dimensional radial flow (case (b) in
Figures B-4 and B-7) which gives an upper boundary for the actual
leakage rate Is obtained by integrating Darcy's equation for a
circular domain:
Q/B - u kc (h + H)/Log (2H/b) (Equation B-10)
where: Q « leakage rate; B - length of the slot in the direction
perpendicular to the figure; Q/B - leakage rate per unit length; kc =
hydraulic conductivity of the low-permeability compacted soil; h =
head on top of the FML; b - width of the slot; and H - thickness of
the low permeability soil underlying the FML. Recommended SI units
are : Q (mj/s); Q/B (m'/s/m, I.e., m«/s); kc (m/s); h (m); b (m); and
H (m).
The equation for the vertical flow (case (c) in Figures B-4 and
B-7), which gives a lower boundary for the flow rate, is obtained by
writing Darcy's equation for a rectangular domain:
Q/B - kc b (h + H)/H (Equation B-ll)
where the notation is the same as above.
B-17
-------
This lower boundary gives a good approximation of the actual
leakage rate 1f the ratio between the width of the FML hole and the
thickness of the low permeability soil 1s large, which 1s rare.
The upper boundary provided by the radial flow (Equation B-10) is
excessively high 1n many cases and Increases when H/h 1s large and
increases, as shown 1n Figure B-5. Since the leakage rate cannot
increase 1f the thickness of the soil layer increases, the upper
boundary Is Increasingly far from the actual leakage rate when H/h
increases and, therefore, cannot be used as an approximation for the
actual leakage rate.
Equation B-10 can be arbitrarily transformed by replacing h + H by
h, which gives:
Q/B « it kc h/Log (2H/b) (Equation B-12)
As it turns out, this equation can be used for large values of H/h
where 1t gives a lower boundary (case (ba) in Figure B-7) of the
actual leakage rate which is not too far from the actual value (case
(d) in Figure B-7).
These considerations regarding boundaries will be useful to
provide guidance for an approximate evaluation of the leakage rate in
the three-dimensional case (circular hole) where the actual value of
the leakage rate is not known.
- Leakage Rate for the Three-Dimensional Case
In the case of a three-dimensional flow (circular hole), the
actual flow is certainly limited by a bell-shaped phreatic surface
similar to the phreatic surface of the two-dimensional flow (case (d)
1n Figure B-4, and Figure B-5). However, no analytical or numerical
study is presently available to the best of our knowledge. An upper
boundary and a lower boundary are available and they are expressed by
close-form solutions.
B-18
-------
The equation related to the three-dimensional radial flow (similar
to the two-dimensional case (b) In Figure B-7), which gives an upper
boundary for the actual leakage rate, is obtained by Integrating
Darcy's equation for a spherical domain:
Q - IT kc (h + H) d/(l - 0.5d/H) (Equation B-13)
where: Q = leakage rate; kc = hydraulic conductivity of the low-
permeability compacted soil; h - hydraulic head on top of the FML; d =
diameter of the circular hole; and H = thickness of the low
permeability soil. Recommended SI units are: Q (m'/s), kc (m/s), h
(m), d (m), and H (m).
The equation related to the vertical flow (similar to the two-
dimensional case (c) in Figure B-4), which gives a lower boundary for
the actual leakage rate, is obtained by writing Darcy's equation for a
cylindrical domain:
Q - kc a (h + H)/H (Equation B-14)
where: a = surface area of the hole 1n the FML (a = rr d2/4 if the
hole is circular); and other notation as above.
As discussed for the two-dimensional case, Equation B-13 can be
rewritten as follows:
Q - n kc h d/(l - 0.5d/H) (Equation B-15)
It is possible that this equation gives a lower boundary of the
actual leakage rate when d/H is small (like Equation B-12 for the two-
dimensional case). It is Interesting to note that Equation B-15 tends
toward a very simple limit when d/H tends toward zero:
Q « TT kc h d (Equation B-16)
where: Q = leakage rate; kc - hydraulic conductivity of the low-
permeability compacted soil underlying the FML; h = hydraulic head on
top of the FML; and d = diameter of the circular hole in the FML.
B-19
-------
Due to the lack of any better solution, Equation B-16 will be used
as an approximation for the actual leakage rate.
Another approach for evaluating leakage rate In the three-
dimensional case Is to use the chart established by FaurTTm^Jthe two-
dimensional case (Figure B-8) and modify Equation B-9 by replacing the
length B of the slot by the perimeter nd of the circular hole (and not
half the perimeter, nor the diameter of the hole as one may be tempted
to do):
«^»
Q - TT C kc (H + h) d (Equation B-17)
where: Q - leakage rate; C - dlmenslonless coefficient given by
Faure's chart (Figure B-8); kc - hydraulic conductivity of the low-
permeablHty compacted soil; H - thickness of the low permeability
soil layer; h - hydraulic head on top of the FML; and d - hole
diameter.
B.4.3 Analyses Assuming Flow between FHL and Soil
- Introduction
Analytical studies have been conducted by Fukuoka [1986] and Brown
et al. [no date]:
• Fukuoka considers the case where there is a geotextlle (without
a hole) between the FML (with a hole) and the soil. The liquid
leaking through the FML hole first flows horizontally in the
geotextlle, then vertically through the soil layer.
• Brown et al. consider that there 1s a space between the FML and
the soil layer. The liquid leaking through the FML hole first
flows horizontally in the space, then vertically through the
soil layer.
B-20
-------
- Flow in Geotextile between FML and Soil
Fukuoka [1986] considers that there is a geotextile between the
FML and the low permeability soil, and that the leachate flows
horizontally and radially within the geotextile before it flows
vertically in the soil underlying the geotextile. Although
geotextiles are not used in composite liners, the analysis made by
Fukuoka is pertinent to composite liners because similar equations can
be used for flow in the narrow space between a FML and the soil.
The following differential equation has been established by
Fukuoka [1986]:
(1/r) (dh/dr) + d'h/dr* - h kc/(H9) (Equation B-18)
where: r = radius from center of hole; h - hydraulic head, at radius r
in the geotextile; kc - hydraulic conductivity of the low-permeability
compacted soil underlying the geotextile; H - thickness of the soil
layer; and 8 » hydraulic transmissivity of the geotextile.
The only assumption is that the flow in soil is vertical. No
assumption 1s made regarding the hydraulic head in the geotextile.
This head decreases from a maximum value at the FML hole, to zero at
the periphery of the wetted portion of the geotextile. Consequently,
flow through soil is faster at the center of the wetted area than at
the periphery. Solving the above equation would give the radius of
the wetted area and would allow to determine the leakage rate.
Fukuoka did not solve the equation, but the solution proposed by Brown
et al. for Equation B-24, which is similar, can be adapted to Equation
B-18 if the thickness of the geotextile (and, therefore, its
transmissivity) is assumed not to vary with the radius r (while, in
B-21
-------
fact it varies since the effective stress on the geotextile varies
with the radius r).
Equation B-18 was established by combining Darcy's vertical flow
in the soil with Darcy's radial flow 1n the geotextile, Qr, which 1s
governed by the classical differential-equation:
Qr - - 2 IT r kp s dh/dr (Equation B-19)
where: kp - hydraulic conductivity of the geotextile 1n the direction
of Us plane; s - thickness of the geotextile (U&r, spacing between
FML and soil); and other notation as above.
This equation can also be written:
Qr - - 2 ir r 9 dh/dr (Equation B-20)
where: 8 - hydraulic transm1ss1vity of the geotextile.
- Flow in Space between FML and Soil
This study was made by Brown et al. principally to extrapolate
results obtained with their small diameter model to real situations
where the flow may laterally extend over a large area.
The approach used by Brown et al. 1s similar to Fukuoka's. They
combine vertical Darcy's flow In the low-permeability compacted soil
with radial flow In the space between the FML and the underlying soil.
Brown et al. Integrated Newton's equation for viscous fluids in a
circular domain and demonstrated that the radial flow 1s governed by:
Qr - - [ir r s3 p g/(6 n)] (dh/dr) (Equation B-21)
where: r « radius from center of hole; s - spacing between FML and
low permeability soil; p - density of leachate; g - acceleration of
gravity; n • viscosity of leachate; and h - hydraulic head at radius.r
in the space between FML and soil.
B-22
-------
By comparing Equations B-20 and B-21, it appears that a space s
between the FML and the underlying soil is equivalent to a hydraulic
transmissivity 9 given by:
9 - p g s'/(12 n) (Equation B-22)
For example, using the density (p - 1000 kg/mj) and the viscosity
(n • 10~* kg/ms) of water, this equation shows that a spacing s « 1 urn
is equivalent to a hydraulic transmissivity of 8 x 10~* m'/s, and a
spacing s = 0.1 mm is equivalent to a hydraulic transmissivity of 8 x
10~8 mz/s. These transmissivity values are consistent with
transmissivities of synthetic drainage layers.
The differential equation obtained by Brown et al. is:
d(r dh/dr)/d r - [12 n kcr/(P g s')] (1 + h/H) (Equation B-23)
which can be written:
(1/r) (dh/dr) + dah/dra -
[12n kc/(p g s')] (1 + h/H) (Equation B-24)
Combining Equation B-22 and B-24, it appears that Equation B-24
[Brown et al.] Is Identical to Equation B-18 [Fukuoka, 1986] except
for the last term, h/H for Fukuoka and (1 + h/H) for Brown et al.
(This discrepancy must be elucidated.) Brown et al. solved this
differential function using Bessel functions to interpret results from
their laboratory model (see Section B.5.2). However, the charts they
proposed for field conditions were established with a. simplifying
assumption: the hydraulic gradient for the vertical flow in soil is
one. In other words, they assume that the hydraulic head on top of
the low-permeability soil is zero. This assumption is valid only if
the hydraulic head on top of the FML is much smaller than the
thickness of the low-permeability soil layer. This assumption is:
• always acceptable for bottom liners;
• never acceptable for surface impoundment top liners; and
B-23
-------
• probably acceptable 1n most cases of top liners for landfills.
With the simplifying assumption of a gradient of one 1n the soil,
the differential equation governing the flow becomes, as Indicated by
Brown et al.:
dh 6 n kc
_ . - (r - IT RVr) (Equation B-25)
dr p g s*
which gives the following relationship [Brown et al.]:
h + H - [3 n kc rfl/(4 P 9 s*)]
[2 (2R/d)2 Log (2R/d) - (2R/d)a + 1] (Equation B-26)
where: h • hydraulic head on top of the FML; H - thickness of the
low-permeability soil layer; n • viscosity of the leachate; kc =
hydraulic conductivity of the low-permeability compacted soil; d =
diameter of the hole 1n the FML; p • density of the leachate; g =
acceleration of gravity; s - spacing between the FML and the low-
permeability compacted soil; and R • radius of the wetted area.
Equation B-26 gives the radius of the wetted area if the spacing s
between the FML and the low-permeability compacted soil is known.
Guidance regarding selection of spacing values can be obtained through
backcalculation of Brown et al.'s test results (see Section B.5.2).
When the radius R of the wetted area 1s known, the leakage rate
can be determined by using the following equation which derives from
Darcy's equation with the assumption that the hydraulic gradient is
one in the low-permeability compacted soil:
Q - u R* kc (Equation B-27)
The above equations were used by Brown et al. to establish charts
giving the leakage rate and the radius of the wetted area (Figures B-9
through B-12). To summarize results presented 1n these charts and
extrapolate or interpolate them, we propose the following equations:
B-24
-------
Q - 0.7 a0'1 kc°-" h (Equation B-28)
R - 0.5 a"" kc-«"" h— (Equation B-29)
These empirical equations are only valid with the units Indicated:
0 - leakage rate (m'/s); a - surface area of FML hole (ma); kc «
hydraulic conductivity of low-permeabil1ty compacted soil (m/s); h =
hydraulic head on top of FML (m); and R - radius of wetted area
between FML and soil (m).
B.4.4 Free Flow through Holes 1n the FML
- Purpose
The case of free flow through holes 1n the FML provides an upper
boundary for the flow rate which could happen 1n the case of a large
space between a FML with a hole and the underlying low-permeability
compacted soil.
- Basic Equation for Leakage Rate
Assuming that there 1s a large empty space under a FML with a
hole, Bernoulli1's equation for free flow through orifices can be used
to evaluate the leakage rate through the hole:
Q - C a 7 2gh (Equation B-30)
where: Q - leakage rate;- h - hydraulic head on top of the FML; a =
hole surface area; and g - acceleration of gravity. C is a
dimensionless coefficient, valid for any Newtonian fluid, and is
related to the shape of the edges of the aperture; for sharp edges, C
» 0.6. Recommended SI units are: Q (m*/s), h (m), a (ma), and g
(m/s').
B-25
-------
- Radius of Wetted Area
The "wetted area" Is the area where leakage flows between the FML
and the underlying low-permeability compacted soil before 1t seeps
Into the soil.
By combining Equations B-27 and B-30, It appears that, If the
spacing between the FML and the soil 1s large enough to ensure free
flow, the radius of the wetted area 1s given by:
n Ra kc - 0.6 a >/ 2 g h (Equation B-31)
hence:
R - 0.44 a°" (2 g h) ••" kc"°" (Equation B-32)
and, In the case of a circular hole:
R - 0.39 d (2 g h)"" kc~°" (Equation B-33)
where: R - radius of the wetted area; a - hole area; d - hole.
diameter; g - acceleration of gravity; h - hydraulic head on top of
FML; and kc - hydraulic conductivity of the low-permeabmty compacted
soil underlying the FML. Recommended SI units are: R (m), a (m2), d
(m), g (m/s2), h (m), and kc (m/s).
- Calculations
Equation B-30 has been used to calculate leakage rates for two
typical holes:
• a 2 mm (0.08 in.) diameter hole which is typical of a small
hole due to defective seaming (as discussed 1n Section B.3.5);
and
• a 11.3 mm (0.445 in.) diameter hole which 1s the standard 1 cm2
hole recommended for design, as indicated 1n Section B.3.6).
B-26
-------
Results from these calculations are given 1n Table B-9. Hydraulic
heads considered 1n these calculations are as follows:
• 0.03 m (0.1 ft) which is an average head that can normally be
expected on the top liner of a landfill with a well designed
and constructed leachate collection and removal system.
• 0.3 m (1 ft) which is the maximum head considered in the design
of the leachate collection and removal system of a landfill.
• 3 m (10 ft) which is a typical head on the top liner of a
surface impoundment.
B.5 LABORATORY MODELS
B.5.1 Introduction
Tests to evaluate leakage through composite liners due to a hole
in the FML were conducted by Fukuoka [1985, 1986] and Brown et al. [no
date]. It is important to recognize that neither the Brown et al.
tests or Fukuoka tests were developed to model leakage through
composite bottom liners under field conditions. The Brown et al.
tests were preliminary and conceptual 1n nature. The Fukuoka tests
did not even directly relate to field conditions existing at landfills
and surface impoundments. However, both sets of tests (and in
particular the Brown-et al. tests) can be used to develop an
understanding of the mechanics of flow through composite liners and to
relate design equations to field condition.
In both cases, tests were conducted with a FML having a circular
hole, and various hole diameters were used in both testing programs.
Additional tests by Brown et al. included FML flaws that are not
circular such as slits or seam defects. The tests were intended to be
full-scale models of the reality since hole size, FML thickness, and
(approximately) soil layer thickness were similar to what they are in
B-27
-------
the field. However, the permeameters used had a limited diameter
(e.g., 0.6 m for Brown et al., and 1.5 m for Fukuoka) and the
extension of lateral flow between the FML and soil was limited by the
walls of the permeameter.
In the tests conducted by Brown et al., the FML was always covered
by 0.15 m (6 In.) of gravel to ensure contact between FML and soil,
and, in some tests, an additional load up to 160 kPa (3340 psf)
(equivalent to 10 m of soil) was applied to evaluate the effect of
overburden pressure. In many of the tests conducted by Fukuoka, the
FML was not covered, and the only load applied on the FML was the
water pressure.
Water heads in Brown et al. tests were up to 1 m, while in Fukuoka
tests, they were up to 40 m. Tests by Brown et al. were conducted for
landfill applications while Fukuoka was working on the design of a
large dam and reservoir.
Fukuoka used only a PVC FML, while Brown et al. considered a
variety of FMLs: HOPE, PVC, CSPE, and EPDM, with various thicknesses.
Tests by Fukuoka as well as tests by Brown- et al. showed that
there Is flow between the FML and the soil. Some of the tests
conducted by Fukuoka and by Brown et al. Included a geotextile between
the FML and the soil. With a geotextile, flow between the liners
would be expected and the liners do not constitute a true composite
liner.
B.5.2 Review of Tests bv Brown et al.
These tests are presented in a report by Brown et al. [no date],
B-28
-------
- Description of the Tests
Tests were conducted in a 0.6 m (24 in.) diameter permeameter.
Hole diameters ranged between 0.8 mm (1/32 in.) and 13 mm (1/2 in.),
and non-circular holes such as slits and seam defects were considered.
The FMLs were: HOPE (0.8 mm to 2.5 mm) (30 to 100 mils); PVC (0.5
to 0.8 mm) (20 to 30 mils); CSPE (0.9 to 1.15 mm) (36 to 45 mils); and
EPDM (0.8 mm) (30 mils).
In some tests, geotextiles were included between the FML and the
soil. The geotextiles were needlepunched nonwovens with a mass per
unit area of 250 to 350 g/ma and a thickness (under no load) on the
order of 2.5 to 4 mm.
The soils used were a silty sand (k - 2 x 10~* m/s), and a clayey
silt (k - 2 x 10"' m/s).
- Approach
The diameter of the permeameter used by Brown et al. was small
(0.6 m) and lateral flow could not extend beyond a radius of 0.3 m as
it would have in most cases without the limitation imposed by the
permeameter walls. Therefore, the calculations presented in Section
B.4.3 were used to backcalculate the value of the spacing between the
FML and soil from the test results. The value of the spacing thus
obtained can then be used in similar equations to determine the radius
of the wetted area and, therefore, the leakage rate in actual
situations where lateral expansion of the flow is not impeded by
permeameter walls. The backcalculated spacing values are as follows:
0.02 mm for clayey silt regardless of FML
0.08 mm for silty sand and flexible FML (PVC)
0.15 mm for silty sand and stiff FML (HOPE)
B-29
-------
Spacing between the FML and the soil, and, therefore, the leakage
rate, appears to Increase 1f the FML stiffness Increases (at least in
the case of the more permeable soil). It also appears that spacing
increases 1f the soil 1s coarse, which 1s Illustrated by:
0.02 mm - d,0 of clayey silt
0.08 mm - d,% of silty sand
The above spacing values are related to the case of a FML with
15 cm of gravel overburden. This 1s an unreal1st1cally low overburden
pressure 1n comparison to those typically encountered 1n the field.
Following Is a review of the Influence of various parameters on
test results.
- Effect of Overburden Pressures
When a compress1ve stress of 160 kPa (equivalent to 10 m of soil)
is applied on a 0.75 mm (30 roil) thick HOPE FML placed on a soil with
a hydraulic conductivity of 2 x 10~* m/s, the flow rate through a
FML hole is divided by 200 and the backcalculated theoretical spacing
between FML and soil Is divided by 10 (there are no results for the
soil with a hydraulic conductivity of 2 x 10~' m/s).
- Effect of Flaw Shape
Erratic results were obtained with slits and seam defects on the
soil with kc - 2 x 10~* m/s:
Some tests showed that a 50 mm slit or seam defect is often
equivalent to a 0.5 to 1 mm diameter circular hole (however
other tests showed that a 50 mm seam defect can be equivalent
to a 75 mm diameter hole).
B-30
-------
• Tests showed that a 150 mm slit or seam Is often equivalent to
a 75 mm diameter circular hole (which 1s very different from
the 0.5 to 1 mm diameter circular hole Indicated above as
equivalent to a 50 mm seam defect).
It was difficult to compare slits, seams and circular holes with
the 2 x 10~s m/s soil because for that soil there 1s more lateral flow
and permeameter walls disturbed the flow.
- Conclusions from Brown et al.'s Tests
-- x
In order to extrapolate to field conditions, Brown et al. make the
following recommendations regarding the values of the spacing between
FML and soil to be used in the equations presented in Section B.4.3 to
evaluate leakage rate and radius of wetted area 1n actual field
conditions where lateral extension of flow is not impeded by wall
permeameter:
soil hydraulic FML-soil
conductivity, kc spacing, s
(m/s) (mm)
10~' 0.15
10"7 0.08
10"' 0.04
10"' 0.02
These values are the upper boundary of (or even larger than) the
backcalculated spacing values previously given 1n the discussion of
the approach. Also, these spacing values are for the case when there
is little or no overburden (e.g., 15 cm of gravel), and they are
expected to be smaller than 1n the case when there is a large
overburden. Therefore, for these two reasons, leakage rates
calculated by Brown et al. are likely to be conservative. Results of
the Brown et al. study indicate that there is a significant benefit of
a composite liner design incorporating a FML upper component and a
compacted soil lower component.
B-31
-------
B.5.3 Review of Tests by Fukuoka
These tests are described in [Fukuoka, 1985; and Fukuoka, 1986].
They were conducted for the design of the lining system for a dam and
a reservoir with a maximum water head of 40 m (130 ft). Although
these conditions are not representative of hazardous waste management
units, the study conducted by Fukuoka, when combined with the findings
of Brown et al., provide a good understanding of the mechanisms
governing leakage through composite liners.
- Description of the Test
All tests discussed below were conducted with the following
equipment, conditions, and materials: permeameter diameter 1s 1.5 m (5
ft); water pressure 1s 200 or 400 kPa (4,000 or 8,000 psf); soil
permeability 1s on the order of 10"? to 10~' m/s (10~* to 10~' cm/s);
soil thickness is 0.45 m (1.5 ft) when no soil cover is placed on the
FML and 0.225 m when a 0.225 m (0.75 ft) thick soil cover 1s placed on
the FML; the FML is a 1 mm (40 mil) thick PVC; the geotextile 1s a
needle-punched nonwoven geotextile (mass per unit area 450 g/m2 (13
oz/sq. yd), 4 mm (160 mil) thick, permeability 0.001 m/s (0.1 cm/s)
under no pressure and 0.0005 m/s (0.05 cm/s) under a 400 kPa (8,000
psf) pressure).
- Tests with FML Alone on Soil (no geotextile, no cover)
In this case, tests show that the diameter of the FML hole needs
to be larger than 2 mm (0.08 in.) approximately in order to ensure
that free flow through the hole (assuming there is nothing under the
FML) is larger than flow rate through soil alone. This indicates that
the soil layer has less Influence in reducing leakage rate in the case
of very small holes than in the case of large holes.
Tests showed that the leakage rate becomes equal to the leakage
rate with no FML at all when the diameter of the FML hole Is larger
than approximately 20 mm (3/4 in.) (Figure B-13). This indicates that
B-32
-------
leakage flows laterally between the FML and the soil and reaches the
walls of the permeameter (diameter 1.5 m (5 ft)) when the diameter of
the hole 1s 20 mm (3/4 1n.) or more. This also Indicates that the
pressure in the liquid located between the FML and soil Is the same as
the pressure on top of the FML.
Pressure measurements 1n the soil (Figure B-14a) showed that the
full water pressure Is applied on top of the soil, which confirms that
there is a space between FML and soil where water flows freely. In
other words the FML was slightly uplifted by water. (Note that
pressure on top of the FML, plus the weight of the FML (specific
gravity 1.2) exceeds the pressure under the FML by 2 Pa (0.04 psf).
This is an extremely small pressure (i.e., of the order of the
pressure exerted by a couple of sheets of paper 1n dry conditions) and
it is easily overcome by the stiffness of the FML, even a FML as
flexible as PVC - a PVC FML wrinkle can easily carry a couple of
sheets of paper.)
- Tests with FML on Geotextlle on Soil
The geotextile had no hole (only the FML had a hole). The
geotextile and the FML were not glued together (I.e., the FML was
simply laid on the geotextile). (This detail is important in the
discussion presented hereafter.)
When FML hole was smaller than 30-50 mm (1-2 in.) approximately,
flow rate was approximately 20 times smaller than flow rate through
soil alone. In other words, when FML hole diameter was smaller than
30-50 mm (1-2 1n.), using a geotextile under the FML decreased the
flow rate by approximately one order of magnitude or more.
Pressure measurements in the soil in the case of a 20 mm (3/4
in.) diameter FML hole (Figure B-14 b) showed that the water pressure
on the soil surface (i.e., under the geotextile) was roughly uniform
and 15 times smaller than the uniform pressure in the case without
geotextile between FML and soil. This Indicates that the head and,
consequently, flow rate was 15 times smaller with geotextile than
B-33
-------
without geotextile, which is consistent with the observations
mentioned above.
•
Pressure measurement 1n the soil In the case of a 50 mm (2 in.)
diameter FML hole (Figure B-15) showed that water pressure on the soil
surface was less uniform than 1n the case of a 20 mm (3/4 in.)
diameter FML hole. Pressures were larger 1n the vicinity of the hole
which indicated that there was water flowing 1n the geotextile within
a radius smaller than the radius of the test permeameter.
It may be concluded that FML, geotextile and soil stay in close
contact when the FML hole 1s smaller than 50 mm (2 in.)* This appears
clearly because:
• if water were accumulating between FML and geotextile, the
water pressure on the soil would be uniformly high, almost
equal to the water pressure on the FML (I.e., 200 or 400 kPa)
(4,000 or 8,000 psf) since geotextile permittivity (I.e.,
permeability/thickness) 1s much larger than soil permittivity
and, therefore, head loss through geotextile would be small;
and
• if water were accumulating between geotextile and soil, both
geotextile and FML would be uplifted and the water pressure on
the soil would be equal to the water pressure on the FML (i.e.,
200 or 400 kPa (4,000 or 8,000 psf)).
FML, geotextile, and soil stay 1n close contact because the
pressure on top of the FML (200 or 400 kPa) (4,000 or 8,000 psf) is
much higher than the pressure below the geotextile. The same would
happen with the FML alone (I.e., water pressure on top of the FML
would be higher than water pressure under the FML) if the FML were in
close contact with the soil. But, if the FML were not in close
contact, because of small soil surface Irregularities, and there were
preferential channels for the flow of water between the FML and soil,
water pressure between the FML and soil might become equal to water
pressure on top of the FML. If the soil surface were perfectly
smooth, and if the FML had no wrinkle, there would be no preferential
B-34
-------
path for the water: the FML and the soil would stay in close contact
(the same way two pieces of polished steel stick to each other because
there Is no air or water pressure between them).
- Tests with Earth Cover on the FML, but no Geotextile
In this case, the tests (conducted with FML hole diameter of 10
and 20 mm (3/8 and 3/4 in.)) show a flow rate reduction of the order
of 40% (i.e., a factor of 1.66) as compared to the case where there is
no earth cover on the FML (Figure B-13). The thickness of the earth
cover was 0.225 m (0.75 ft), and the thickness of the soil under the
FML was 0.225 m (0.75 ft) (i.e., a total soil thickness of 0.45 m (1.5
ft) as in the tests discussed above).
More tests would be necessary to draw conclusions, such as tests
with a permeable cover material and comparable tests with identical
low-permeability compacted soil layer thickness under the FML.
However, the tests by Fukuoka show that an earth cover, even on a
flexible FML such as PVC, does not have a marked effect on leakage
rate probably because it is not sufficient to force the FML into soil
irregularities.
B.6 CONCLUSIONS ON LEAKAGE THROUGH COMPOSITE LINERS
B.6.1 Conclusions from Analytical Studies
It appears that the theoretical analyses involved in the
apparently simple problem of leakage through a hole in a FML placed on
a low permeability soil to form a composite liner are extremely
complex.
If perfect contact between the FML and soil is considered, the
two-dimensional problem has been solved but the three-dimensional
problem still requires research. There is no satisfactory approximate
B-35
-------
solution and the analytical lower and upper boundaries are too far
from the actual solution to give valuable Information.
Differential equations have been proposed and some approximate
numerical solutions are available for the case of Imperfect contact
between the FML and soil. To use these equations, 1t 1s necessary to
know the spacing between the FML and the underlying Iow-permeab1l 1ty
soil. Spacing values backcalculated from model tests are only
preliminary and are probably smaller than actual spacing values In the
field. Field conditions listed below will affect actual site-
specific results, while the quality of FML - compacted soil contact
1s probably better In the laboratory than In the field, the laboratory
tests to date have been carried out at unrealIstlcally low overburden
pressures.
• subgrade surface preparation 1s not as good as 1n the model
tests; and
• FMLs have wrinkles and some of these wrinkles are probably not
flattened by overburden pressures.
As a result, actual leakage rates In the field will likely vary
from those calculated using equations Incorporating FML-soil spacings
backcalculated from model tests. Also, 1t 1s likely that there will
be some spatial variation throughout the Uner.
B.6.2 Conclusion from Model Tests
Tests show that, 1n all cases where a FML 1s placed in direct
contact with a low permeability soil, some liquid that has passed
through a hole in the FML flows laterally In the space between the FML
and the underlying soil. Tests show that, as a result of lateral
flow, leakage rates observed are higher than leakage rates which would
be obtained if there was a perfect contact between the FML and the
underlying soil. The degree of contact between the FML and soil in
the model tests can be considered good (smooth soil surface, no cracks
B-36
-------
in clay) but not perfect since flow takes place between the FML and
the soil.
From a construction standpoint, it is recommended to make every
effort to ensure a good contact between FML and low permeability soil
which includes: (1) having a low permeability soil with a smooth
surface and no cracks; and (1i) minimizing or eliminating wrinkles in
the FML. Ideally, the FML should be sprayed on the low permeability
soil instead of being made in a plant and transported to the site: in
this case, the contact may not only be "good" but "perfect".
From a design standpoint, 1t 1s necessary to take Into account the
flow of leachate between the FML and the soil for leakage evaluation
as well as for any other appropriate design consideration such as
damage caused to the soil layers by liquid flowing in the space
between the FML and the underlying soil layer.
Although the tests provided a good understanding of the mechanisms
involved, the diameter of the permeameter, the design parameters and
test conditions used by Brown et al. and Fukuoka limit the usefulness
of the test results when developing design recommendations. Although
extrapolation of test data to field conditions was done by Brown et
al. using a sound theoretical analysis, test conditions were too far
from actual conditions to ensure that extrapolated values are
adequate.
In spite of their limitations, the tests show that composite
liners are significantly more effective than low-permeability
compacted soil alone or FML alone.
B.6.3 Conclusions for Leakage Rate Evaluation
- Review of Methods for Leakage Rate Evaluation
A series of methods have been discussed to evaluate leakage rate
through a composite liner due to a hole in the FML. These methods can
be ranked as follows:
B-37
-------
• An absolute minimum of the leakage rate 1s given by the
vertical flow equation assuming perfect contact between the FML
and the underlying soil (Equation B-14).
• An approximate value (possibly an underestimate) of the leakage
rate 1n case of perfect contact between the FML and the
underlying soil 1s given by Equation B-16. Since this equation
has not been tested, It Is appropriate to have the absolute
minimum mentioned above to make sure that no absurd result Is
considered.
• Leakage rate obtained using charts prepared by Brown et al. on
the basis of their tests (Figures B-9 through B-12) or the
empirical equations we have proposed to summarize these charts
(Equations B-28 and B-29) may be smaller than actual leakage
rate because In the field FMLs have at least some wrinkles and
subgrade preparation Is not as good as In the model tests,
thereby allowing more flow between the FML and the soil In the
field than 1n the models. However, a counteracting Influence
is that the overburden pressure In the model tests was well
below overburden pressures representative of field conditions.
• Finally, leakage through a hole tn a FML alone (I.e., with
nothing underneath It) 1s certainly much larger than leakage
through a composite liner with the same FML hole, even in field
conditions with a far from perfect contact between the FML and
the underlying soil. This case, therefore, provides an
absolute maximum of the leakage rate.
A summary of pertinent equations is presented in Table B-8.
B-38
-------
- Leakage Rate and Radius Graphs
Because of the uncertainties in the analyses as well as the wide
variety of contact conditions, 1t is appropriate in each given case to
plot leakage rates obtained with all the methods described above in
order to make interpolations. It is also appropriate to use a semi-
logarithmic scale for the plot since leakage rates vary within a range
of several orders of magnitude, as is usually the case in hydraulic
problems. The graph in Figure B-16 has been established with a 1 cm2
hole, which is the recommended standard hole for design as indicated
in Section B.3.6. This graph has b.een established for a hydraulic
head of 30 mm (0.1 ft) on top of the FML. Numerical values used to
establish the graph in Figure B-16 are given in Table B-9.
Similarly, a graph can be established for the radii of wetted
areas (i.e., the area covered by leakage flowing between the FML and
the low-permeability compacted soil, before it flows into the
compacted soil) obtained with all the methods described above and
summarized in Table B-8. The radius graph related to a hydraulic head
of 30 mm (0.1 ft) on top of the FML is given in Figure B-17.
- Use of Leakage Rate and Radius Graph
The leakage rate graph permits the determination of the leakage
rate for any given field condition by interpolation between the best
case and the worst case (this worst case is unlikely at a unit with
CQA):
• In the best case: (1) the soil is well compacted, flat and
smooth, has not been deformed by rutting due to construction
equipment, and has no clods nor cracks; and (ii) the FML is
flexible and has no wrinkles.
• In the worst case: (i) the soil is poorly compacted, has an
irregular surface, and is cracked; and (ii) the FML is stiff
and exhibits a pattern of large, connected wrinkles.
B-39
-------
The conditions 1n the best case can be almost as good as the
conditions In the tests by Brown et al. and Fukuoka discussed 1n
Section B.5. Therefore, on the graphs, the best case for field
conditions Is represented by the vertical line corresponding to test
results.
In order to locate the worst field case we have used the radius
graph for a large head, I.e., 3 m (10 ft) (Figure B-21), and we have
assumed that the radius of the wetted area cannot exceed a value on
the order of 10-30 m (30-100 ft) for a value of the compacted soil
hydraulic conductivity of 10~* m/s (10~* cm/s). The location of the
worst case line thus obtained shows that the conditions 1n the worst
case are still much better than the case of free flow through holes 1n
the FML. Free flow 1s an extreme case which Is possible only 1f the
FML 1s very far from the low-permeability compacted soil over a very
large area (radius of 10 to 100 m), which 1s practically Impossible.
Between the best field case and the worst field case we have
selected a vertical line representing good field conditions and a
vertical line representing poor field conditions. As a result, it
appears in Figure B-16 that, for a head of 30 mm (0.1 ft), a leakage
rate of 0.8 liters/day (0.2 gallon/day) corresponds to good field
conditions and a hydraulic conductivity of kc - 10"' m/s (10~* cm/s)
for the low-permeability compacted soil underlying the FML. This
value of the hydraulic conductivity Is a conservative value to
consider in design since the required value of 10~* m/s (10~7 cm/s)
may not always be reached at the site. A less conservative
calculation could consider a compacted soil hydraulic conductivity of
kc - 10"* m/s (10~7 cm/s). In this case a leakage rate of 0.08
liters/day (0.02 gallon/day) 1s obtained. Poor field conditions would
give a leakage rate value of 4 liters/day (1 gallon/day) for kc - 10~'
m/s (10~* cm/s) and 0.4 liters/day (0.1 gallon/day) for kc - 10~* m/s
(10"f cm/s).
B-40
-------
- Leakage Rate due to Permeation and Holes
Leakage rates through composite liners due to a hole in the FML,
obtained from Figure B-16 are summarized in Table B-10, which also
gives leakage rate due to permeation obtained from Table B-6.
To the best of our knowledge, Table B-10 summarizes the best
demonstrated available technology on leakage rate through bottom
composite liners.
B-41
-------
Table B-l. Values of the migration coefficient, y, obtained from permeability
tests conducted at the University of Grenoble (France) with the
apparatus shown in Figure B-l.
FML Type
CSPE
Butyl
Butyl
EPDM
PVC
PVC
PVC
Asphaltic
Asphaltic
hydraulic head, h, in m
5
3.5x10-"
4.2xlO~11
10
3.8xlO~1J
7.7xlO~12
1.7xlO"IJ
l.lxlO"17
1.7xlO"12
1.6xlO'12
8.1xlO~ls
7.4xlO"11
1.6xlO~"
25
1.9xl(T12
6.7x10'"
3.2X10"11
50
S.OxlO"1*
3.9X10"11
2.9xlO'!1
2.3xlO"12
2.5xlO"12
2.1xlO"12
2.0xlO"12
6.5xlO~13
S.SxlO"11
75
7.4X10"11
4.5xlO~1J
100
5.5xlO"12
3.1xlO"lJ
3.0xlO"13
2.2xlO"12
l.lxlO"12
4.4xlO~1J
l.OxlO'12
Values of u in mz/s
B-42
-------
Table B-2. Water vapor transmission (WVT) rates of FMLs from [Haxo et.
al., 1984] and values of the coefficient of migration
derived from WVT values using Equation B-8 (See also Table
B-3). All these tests were conducted at 23°C with a
relative humidity difference of 50%, which is equivalent to
a pressure of 1.4 kPa, i.e., a head of 0.14 m of water.
Polymer
Thickness,
(mm)
Water Vapor
Transmission
WVT,
(g/m2.day)
Coefficient
of migration
y
(m2/s)
Butyl rubber
CPE
CSPE
ELPO
CO
EPDM
0.85
0.85
1.85
0.53
0.79
0.79
0.85
0.94
0.97
0.74
0.76
0.89
0.91
0.94
1.07
0.72
,160
,650
0.51
0.94
1.70
0.384
0.020
0.097
0.643
1.400
0.320
0.264
0.305
0.643
0.333
0.663
0.438
0.748
0.422
0.252
0.142
20.18
14.30
0.270
0.190
0.172
3.8 x 10"
2.0 x 10"
2.1 x 10'
3.9 x 10"
1.2 x 10'
2.9 x 10"
10'
•1 t
•I «
2.6 x
2.2 x
10"
7.2 x 10"
2.9 x 10"
5.8 x 10"
4.5 x 10'
7.9 x 10"
4.6 x 10"
3.1 x 10"
1.2 x 10'
•i s
2.7 x 10~13
2.7 x 10~13
1.6 x 10
2.1 x 10
3.4 x 10'
'I 8
' 1 «
• 1 »
B-43
-------
Table B-2, continued
Neoprene
Nitnle rubber
PB
PEEL
LOPE
HOPE
HDPE-A
PVC
PVC-E
PVC-OR
Saran Film
0
0
1
1
0
0
0
0
0
2
0
0
0
0
0
0
0
0
.51
.91
.27
.59
.76
.69
.20
.76
.80
.44
.86
.28
.51
.76
.79
.91
.83
.013
0
0
0
0
5
0
•
*
•
•
•
•
10
0.
0.
0.
0.
0
304
473
429
237
51
084
.50
0573
0172
0062
0472
4
2
1
1
2
4
,42
.97
.94
.85
.78
.17
563
1
5
6
4
4
6
2
5
1
1
4
1
1
1
1
2
4
8
.8
.0
.3
.4
.8
.7
.4
.0
.6
.8
.7
.4
.7
.7
.7
.9
.0
.5
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
10"1
10'1
10'1
10'1
10"'
10''
10" '
10'1
10''
10''
10~l
10"'
10"1
10~l
HT1
10~l
10'1
10"1
s
s
(
s
4
(
4
«
t
t
I
4
4
4
4
4
4
J
Abbreviations are defined in Table 4-1.
B-44
-------
Table B-3. Water vapor transmission (WVT) rates of FMLs [Rogers, 1964]
and values of the coefficient of migration derived from WVT
values using Equation 2.2-9. (See also Table B-2.)
Reference
FML Pressure
Type
P
(kPa)
Hypalon 6.4
Butyl 6.4
PVC 6.1
HOPE 0.92 6.4
0.94 5.8
0.95 6.1
0.96 5.8
Water Reference Coefficient
Vapor Thickness of
Transmission Migration
WVT T u
(g/m2.day) (rim) (mVs)
161 0.025 4.6xlO"14
26 0.025 7.5xlO~l*
32 0.025 9.2xlO~ls
28 0.025 S.lxlO"16
14 0.025 4.1x10""
6.7 0.025 1.9xlO~>s
4 0.025 l.lxlO"1*
Notes: (i) the test pressure, p, is derived from the test relative
humidity difference using Equation B-3; (ii) a 6 kPa pressure
is equivalent to a water head of 0.6 m (2 ft).
B-45
-------
Table B-4. Summary of values of the coefficient of migration, u, from
Tables B-l, B-2 and B-3.
Hydraulic
head
h
0.14 m
0.6 m
10 m
50 m
100 m
FML Type
CSPE
5xlO'lf
4.6xlO"14
3.8xlO~12
S.OxlO'12
5.5xlO~12
PVC
1.7xlO"14
9.2X10"1'
1.6xlO"12
Z.OxlO'12
l.OxlO'12
HOPE
1.7xlO~"
4.1xlO~li
-
.
-
Values of coefficient of migration, u (m*/s)
B-46
-------
Table B-5. Values of coefficient of migration resulting from
extrapolations and interpolations in Figure B-3.
FML
Type
CSPE
HOPE
Hydraulic head in m (ft)
0 m
(0 ft)
0
0
0.03 m
(0.1 ft)
3.5xlO~1'
1.5xlO~17
0.3 m
(1 ft)
l.BxlO"14
lxlO~l&
3 m
(10 ft)
GxlO"1 J
^M^
7xlO~14
Values of coefficient of migration, u, in mz/s
> 10 m
(> 30 ft)
6xlO"IZ
— •
lxlO~"
umax
Table B-6. Values of rate of leakage due to permeation through FML
derived from values of coefficient of migration given in
Table B-5, using Equation B-l and assuming an FML thickness
of 1 mm (40 mils).
FML
Type
CSPE
HOPE
Hydraulic head in m (ft)
0 m
(0 ft)
0
0
0.03 m
(0.1 ft)
0.035
0.0015
0.3 m
(1 ft)
1.5
0.1
3 m
(10 ft)
60
7
> 10 m
(> 30 ft)
600
100
Values of leakage rate in liters/lOOOmVday (Ltd) or
gallons/acre/day (gpad)
B-47
-------
Table B-7. Leakage rate due to holes
pervious medium such as
1n an FML placed on a very
a drainage layer. Note: the
11.3 mm diameter circular hole has a surface area of 1 crrr
Hydraulic head
Defect
diameter
0.03 m
(0.1 ft)
0.3 m
(1 ft)
3 m
(10 ft)
2 mm
(0.08 in.)
11.3 mm
(0.445 1n.)
125
(30)
1,260
(330)
400
(100)
4,000
(1,000)
1250
(300)
12,600
(3,300)
Values of leakage rate in liters/day
(gallons/day)
8-48
-------
Table B-8. Summary of equations giving leakage rate, Q, and radius of
wetted area, R, for composite liners when there is a hole in
the FML.
ABSOLUTE MINIMUM __^ (MIN) 1n Figures B-16 and 17
(Vertical flow) ^~"~-^
Q = kc a (h + H)/H (Equation B-14)
R = d/2
PERFECT CONTACT ^^ (P.C.) in Figures B-16 and 17
(Approximate value of Q given by radial flow)
0 - n kc h d (Equation B-16)
R » unknown
EXCELLENT CONTACT (TEST) in Figures B-16 and 17
(Empirical equations from model tests)
Q - 0.7 a0'1 kc°"8 h (Equation B-28)
R - 0.5 a0'06 kc~0>0' h"6 (Equation B-29)
LARGE SPACE BETWEEN (MAX) in Figures B-16" and 17
FML AND SOIL
(Q given by Bernouilli's equation)
0 - C a vTgh - 0.6 a / 2gh (Equation B-30)
R - 0.39 d (2 g h)002S kc~°" (Equation B-33)
where: kc = hydraulic conductivity of low-permeability compacted soil
underlying the FML; a = area of hole in FML; h =» hydraulic head on
FML; H = thickness of compacted soil layer; d = diameter of hole in
FML; and g = acceleration of gravity. Recommended SI units: kc
(m/s), a (m2), h, H, and (m); and g (m/s2). These units are mandatory
for the two empirical equations.
B-49
-------
Table B-9. Numerical values used to establish the graphs presented in Figures B-16
and B-17. This table has been established for a hydraulic head of
30 mm (0.1 ft) on top of the FML, a hole area of 1 cm* (0.16
In2.), and a low-permeability compacted soil thickness of 0.9 m (3 ft).
Hydraulic Conductivity of Compacted Soil
Underlying the FML
Leakage
Rate
9
(m'/s)
Radius
of
Wetted
Area
R
(m)
Case
Absolute minimum
Perfect contact
(approximate
theory)
Good contact
(model tests)
Free flow
(Bernoulli 1's
equation)
Absolute minimum
(hole radius)
Perfect contact
(unknown)
Good contact
(model tests)
Free flow
Equation
B-14
B-16
B-28
B-30
R - d/2
B-29
B-33
10"7 m/s
l.OxlO"11
l.lxlO"10
5.8x10"'
4.6x10"'
0.0056
-0.032(*)
0.14
12
10"' m/s
l.OxlO'12
1.1x10""
7.6xlO"18
4.6x10"'
0.0056
-0.032(*)
0.17
38
10~' m/s
l.OxlO"11
l.lxlO"12
l.OxlO"10
4.6xlO"6
0.0056
-0.032(*)
0.19
122
(*) Value obtained by interpolation in Figure B-17,
B-50
-------
Table B-10.
Leakage rates
permeation 1s
figures) and
Figure B-16,
between the
component of
through composite liners. Leakage due to
obtained from Table B-6 (rounding up the
leakage due to holes is obtained from
as a function of the quality of contact
FML component and the compacted soil
the composite liner. This table has been
1 cmz (0.16 in2.};
established with: hole area =
compacted soil thickness = 0.9 m (3 ft);
1 mm (40 mils); and frequency of holes
(1 per acre).
FML
= 1
thickness -
per 4000 m2
Quality
of
contact
Good
Poor
Leakage
mechanism
Permeation
Hole
TOTAL
Permeation
Hole
TOTAL
Low-Permeabil ity
Compacted Soil
Hydraulic Conductivity,
kc
10~" m/s
(10"* cm/s)
0.001
0.2
0.2
0.001
1
1
10"f m/s
(10"' cm/s)
0.001
0.02
0.02
0.001
0.1
0.1
Values of leakage rate
in Ltd or gpad
B-51
-------
Af& UNDER.
PRESSURE
Y/////////////A
WATER
FML
POROUS STONE
Figure B-l. Permeameter used to evaluate flow through Intact FMLs at
the University of Grenoble (France).
B-52
-------
KA)
Figure B-2.
Typical shape of the curve giving the coefficient of
migration, Ug, as a function of the hydraulic head, h.
B-53
-------
1E-1W
*« 1E-12,
.2 1E-13J
(D
C_
01
IE-14,
g
O)
o
CJ
1E-17
^ CSPE
a PYC
o HOPE
1E-3 1E-2
1E-1 IE 0
Hydraulic Head
El IE 2
IE 3
Figure B-3. Values of coefficient of migration, yq, for various FMLs
from Table 2.2-6.
B-54
-------
Figure B-4.
Flow nets for the four cases considered in two-
dimensional theoretical studies related to leakage
through composite liners due to hole in FML, assuming
perfect contact between FML and soil layer: (a) entire
soil layer saturated; (b) radial flow; (c) vertical flow;
(d) actual flow. As demonstrated by Faure [1979], the
actual flow is limited laterally by a phreatic surface.
Note that in cases (a), (b), and (d), there is flow in
the soil along the interface, although there is no flow
between the FML and the soil because there is no space
between the FML and the soil in the considered cases
since perfect contact is assumed.
B-55
-------
Figure B-5.
Typical flow nets for leakage through a composite liner
due to a FML hole (two-dimensional study assuming that
the FML and the underlying soil are in perfect contact)
(see case (d) in Figure B-4). The cases shown above are:
(a) b/H - 0.005 and h/H - 1; (b) b/H - 0.005 and h/H - 3;
(c) b/H - 0.05 and h/H - 1/3; and (d) b/H - 0.05 and h/H
- 1. Notation: b - width of infinitely long hole (slot)
in the FML; h - hydraulic head on top of the FML; and H «
thickness of the soil layer underlying the FML [Faure,
1979].
B-56
-------
Figure B-6.
Lateral extent of the phreatic surface limiting trie flow
in the soil layer due to a hole in the FMl. This chart
is related to the two-dimensional case (the hole is a
Slot of width b) and perfect contact is assumed between
the FML and the soil layer [Faure, 1979].
B-57
-------
Figure B-7.
Leakage rates through a composite liner due to a slot of
b m the FML (two-dimensional case), assuming
perfect contact between the FML and the soil.
Calculations were made with several assumptions regarding
(a) soil entirely saturated by the flow; (b,)
flow using Equation B-9; (b,) radial flow using
Equation B-ll; (c) vertical flow; (d) actual flow. Cases
•ough (d) are illustrated in Figure B-4.
B-58
-------
. 0
100
5.C
O.O 1
0.05
Figure B-8.
Chart giving dimensionless coefficient C to be used in
Equation B-9 which gives the leakage rate through a
composite liner due to a slot in the FML (two-dimensional
case). Coefficient C can also be used in Equation B-17
to make an approximate evaluation of the leakage through
a composite liner due to a circular hole in the FML
(three-dimensional case). Notation: h = hydraulic head
on top of the FML; b = width of the slot (to be replaced
by the diameter d of a circular hole when the chart is
used for the three-dimensional case); and H = thickness
of soil layer.
B-59
-------
50 60 7O 8O
90
IOO
Figure B-9.
Leakage through a composite liner due to a hole in the
FML [Brown et al.]. Chart giving the leakage rate, 0,
and radius, R, of the wetted area as a function of the
hydraulic head on the FML, for a compacted soil _hydraulic
conductivity kc - 3.4 x 10"' m/s (3.4 x 10 * cm/s).
Notation: d - diameter of the FML hole; and h =
hydraulic head on the FML. Note: although the chart in
[Brown et al.] is labeled "kc - 10~* cm/s", it seems to
us that it was established for 3.4 x 10"" cm/s.
B-60
-------
§
90
100
Figure B-10.
Leakage through a composite liner due to a hole in the
FML [Brown et al.]« Chart giving the leakage rate, 0,
and radius, R, of the wetted area as a function of the
hydraulic head on the FML, for a compacted soil hydraulic
10"7 m/s (3.4 x 10~s cm/s).
of the FML hole; and h =
Note: although the chart in
"kc = 10~* cm/s", it seems to
for 3.4 x 10"* cm/s.
conductivity kc
3.4 x
Notation: d = diameter
hydraulic head on the FML.
[Brown et al.] is labeled
us that it was established
B-61
-------
O
*M*
UJ
.10
.09
.08
.07
.06
•«
§
^ .04
.03
.02
.01
10 ZO SO 40 SO 60
tCAO (em)
70
eo
90
too
Figure B-ll
Leakage through a composite liner due to a hole in the
FML [Brown et al.]. Chart giving the leakage rate, 0,
and radius, R, of the wetted area as a function of the
hydraulic head on the FML for a compacted soil hydraulic
10~' m/s (3.4 x 10'' cm/s).
of the FML hole; and h =
Note: although the chart in
"kc » 10~* cm/s", it seems to
for 3.4 x 10"* cm/s.
conductivity kc * 3.4 x
Notation: d - diameter
hydraulic head on the FML.
[Brown et al.] is labeled
us that it was established
B-62
-------
.01
.009
.ooa -
.007 H
.006
.005
.004
.OOi
.002 -
.001
IO 20 3O 4O SO 60
HEAD (cm)
70
ao
90
96
90
85
BO
75
60
• 55
50
•45
• 4O
• 30
20
10
100
Figure B-12. Leakage through a composite liner due to a hole in the
FML [Brown et al.]. Chart giving the leakage rate, 0,
and radius, R, of the wetted area as a function of the
hydraulic head on the FML, for a compacted soil hydraulic
conductivity kc = 3.4 x 10"' m/s (3.4 x 10"' cm/s).
Notation: d = diameter of the FML hole; and h =
hydraulic head on the FML. Note: although the chart in
[Brown et al.] is labeled "kc - IO"7 cm/s", it seems to
us that it was established for 3.4 x 10"' cm/s.
B-63
-------
100
10
50
Diameter of Defect (nun)
Figure B-13. Leakage rates measured in tests conducted with a FML
having a circular hole [Fukuoka, 1985]: (A) no soil
cover on the FML, no geotextlle between the F-ML and the
soil; (B) there is a geotextile between the FML and the
soil, but there is no soil cover on the FML; and (C)
there is a soil cover on the FML and no geotextile
between the FML and the soil. Notation: 0 = leakage
rate measured in the tests; and QQ - leakage rate when
there is no FML (i.e., leakage rate governed by Darcy's
flow through the soil).
B-64
-------
Defect i s o-t lhe 4.-
y
Figure B-14. Water pressure in the soil under the FML in the case of a
20 mm (3/4 in.) diameter hole in a 1 mm (40 mil) thick
PVC FML; (a) the FML is placed directly on the soil; and
(b) there is a geotextile between the FML and the soil
[Fukuoka, 1985].
B-65
-------
rtgn
Olfl
Olfl
0.07
O 1Q
00.07
00.07
nm=;
»OOI OO.QI
OO.Ol
of
is ioo Kf*-
Figure B-15. Water pressure in the soil under the FML in the case of a
50 mm (2 in.) diameter hole in a 1 mm (40 mil) thick PVC
FML placed on a needlepunched nonwoven geotextile (mass
per unit area 450 g/m* (13 oz/sq. yd)) resting on the
soil [Fukuoka, 1985].
B-66
-------
Figure B-16. Graph giving the leakage rate in case of leakage through
a FML hole in a composite liner. The hydraulic head is
30 mm (0.1 ft) and the hole area is 1 cm2 (i.e., diameter
of 11.3 mm). Because of uncertainties in the analytical
analyses as well as the large influence of soil
conditions and contact between the FML and the soil, only
a range of values can be given. Field conditions can be
anywhere between the two extremes: (1) best, i.e., the
soil is well compacted, flat and smooth, has not been
deformed by rutting during construction, and has no clods
and cracks, and the FML is flexible and has no wrinkles;
and (2) the soil is poorly compacted, has an irregular.
surface and is cracked, and the FML is stiff and exhibits
a pattern of large, connected wrinkles. Abbreviations:
GOOD and POOR = good and poor field conditions; MIN,
P.C., TEST, and MAX are defined in Table B-8.
B-67
-------
Figure B-17.
Graph giving the radius of the wetted area in case 01
leakage through a FML hole in a composite liner. Th<
hydraulic head is 30 mm (0.1 ft) and the hole area is :
cm2 (i.e., diameter of 11.3 mm). Because of uncertaintie:
in the analytical analyses as well as the large influenci
of soil conditions and contact between the FML and thi
soil, only a range of values can be given. Fiel<
conditions can be anywhere between the two extremes: (l
best, i.e., the soil is well compacted, flat and smooth
has not been deformed by rutting during construction, an<
has no clods and cracks, and the FML is flexible and ha:
no wrinkles; and (2) the soil is poorly compacted, has ai
irregular surface and is cracked, and the FML is stif
and exhibits a pattern of large, connected wrinkles
Abbreviations: GOOD and POOR - good and poor fieli
conditions; MIN, P.C., TEST, and MAX are defined in Tabli
B-8.
B-68
-------
10 .
I _
3rw\
10'
- I6l
I
(PC)
-It.'
(TEST;
POM,
Figure B-18. Graph giving the radius of the wetted -area in case of
leakage through a FML hole in a composite liner. The
hydraulic head is 3 m (10 ft) and the hole area is 1 cm2
(i.e., diameter of 11.3 mm). Because of uncertainties in
the analytical analyses as well as the large influence of
soil conditions and contact between the FML and the soil,
only a range of values can be given. Field conditions
can be anywhere between the two extremes: (1) best,
i.e., the soil is well compacted, flat and smooth, has
not been deformed by rutting during construction, and has
no clods and cracks, and the FML is flexible and has no
wrinkles; and (2) the soil is poorly compacted, has an
irregular surface and is cracked, and the FML is stiff
and exhibits a pattern of large, connected wrinkles.
Abbreviations: GOOD and POOR = good and poor field
conditions; MIN, P.C., TEST, and MAX are defined in Table
B-8.
B-69
-------
APPENDIX C
Summary of
Two-Dimensional Partially
Saturated Flow Analysis Results
[Data f rom Radian, 1987].
-------
C.I DESCRIPTION OF RESULTS
The two-dimensional partially saturated flow analyses results
employed in the body of the report were developed by the Radian
Corporation [Radian, 1987]. These results are summarized in this
appendix by the following categories:
• compacted soil 1iners;
• composite soil liners with FML leakance, L = 7 x 10~14 s"1;
• composite soil liners with FML leakance, L = 3 x 10~:z s-I;
and
• composite soil liners with FML leakance, L = 3 x 10"11 s~l.
Two tables are presented for each lining system simulated: (i)
the first table describes the lining system and gives details of the
•various design variables, and (ii) the second table summarizes the
various leakages (or fluxes) calculated as a function of time.
It was noted by Radian that the calculated cumulative drainage
quantities provided in the second table include significant roundoff
error. Radian suggested that the best way to calculate the cumulative
drainage is as follows:
Cumulative Drainage = Cumulative Leakage - Cumulative Flux
into Bottom Liner - Liquid Stored in
LDCRS
Therefore, in the following tables, the reported cumulative
drainage at 40 years is reported using the above equation.
C-l
-------
Compacted Soil Liners
-------
TABLE A2-21. DESIGN VARIABLES FOR SIMULATION AH-1
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
252
120 feet
2%
10 feet at centerline
FML
800 gallons/acre-day
Uniform
Sand
10 ca/s
1 foot
1A.O cubic feet
Evenly spaced along lower slope
60 feet
Compacted soil
Clay
10~ C3/S
3 feet
127.9 cubic feet
Loam
10~ cn/s
Hydrostatic from water table
10 feet belcw top FML at centerline
-------
TAIiLE A2-22. EXPERIMEtfT AH-1—SIMULATION SUMMARY
Elapsed
Time
(weeks)
2
4
6
8
10
12
(drain
13.40
14
18
22
26
Leak
Rate
(gpad)
1145.53
1069.45
1016.05
975.10
939.71
904.42
flux begins)
874.28
861.48
795.40
794.68
794.59
(approximate time t
26.45
794.57
/if) Ypflru 794. 5?
Cumulative
Leakage
(gal/acre)
16.800
32.200
46.700
60.600
74.000
86.900
95.600
99.200
122.000
144,000
167.000
o steady-state)
169.000
n . fcon.nnn
Drain
Flux
(gpad)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
114.42
586.93
638.19
622.56
622.56
fi75-Sfi
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
0
385
10.500
27,900
45.400
47.400
aloifliLiiaa-
Flux
into
But torn
Liner
(gpad)
142.95
148.92
151.00
151.93
154.43
160.84
171.92
175.23
185.81
175.53
174.31
174.29
17V O/i
Cumulative
Flux into
Bottom
Liner
(gal/acre)
1,610
3.660
5,770
7,890
10,000
12,200
13.900
14.600
19.900
24.900
29.800
30.300
2 s^n.nnn
Flux
from
Unit
(gpad)
42.59
47.29
53.17
62.96
75.16
88.94
99.69
103.85
146.59
170.12
173.10
173.18
171. 94
Cumulative
Flux from
Unit
• (gal/acre)
548
1.1BO
1.880
2.690
3.660
4.810
5.740
6.170
9.680
14,200
19.100
19.600
? sin.nnn
//»•/•• .r
-------
TABLE A2-23. DESIGN VARIABLES FOR SIMULATION AH-2
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Nunber of drains
Drain locations
Drain spacing
Bottom Liner
Type
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
25%
120 feet
2%
10 feet at centerline
FML
101 gallons/acre-day
Uniform
Sand
10 cm/s
1 foot
11.7 cubic feet
Evenly spaced along lower slope
60 feet
Compacted soil
Clay
10 cm/s
3 feet
127.9 cubic feet
Loan
10 cm/s
Hydrostatic from water table
10 feet below top FKL at centerline
-------
TABLE A2-24. EXPERIMENT AH-2--SIMULATION SUMMARY
Elapsed
Time
(years)
1
2
3
4
5
6
Leak
Rate
(gpad)
111.17
104.72
101.26
99.00
97.45
96.45
(approximate time
6.44
40 Years
96.15
95.16
Cumulative
Leakage
(gal/acre)
43.600
82.800
120,000
157,000
193.000
228.000
to steady-state)
243,000
1,410,000
Drain
Flux
(gpad)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
0
,/ -/oo
n
V
Flux
into
Bottom
Liner
(gpad)
59.50
70.01
77.43
83.13
87.35
90.22
91.11
94.05
Cumulative
Flux into
Bottom
Liner
(gal/acre)
18.700
42,400
69,400
98.800
130,000
162,000
177.000
1,330,000
Flux
from
Unit
(gpad)
53.25
67.61
76.04
82.21
86.75
89.84
90.80
94.00
Cumulative
Flux from
Unit
(gal/acre)
13,700
36,100
62,400
91.400
122.000
155.000
169.000
1,320,000
/.,,.,/ ~
.CC ^J
-------
TABLE A4-5. DESIGN VARIABLES FOR SIMULATION AH-6
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
25%
120 feet
2%
10 feet at centerline
FML
62 gallons/acre-day
Sidewall
Sand
10 cm/s
1 foot
1A.O cubic feet
Evenly spaced along lower slope
60 feet
Compacted soil
Clay
10 co/s
3 feet
127.9 cubic feet
Loam
10~ cm/s
Hydrostatic from water table
10 feet below top FML at centerline
-------
TABLE A4-6. EXPERIMENT AH-6—SIMULATION SUMMARY
Elapced
Time
(yeara)
1
2
3
4
5
6
7
Look
Rate
(gpad)
61.10
61.37
61.38
61.38
61.38
61.38
61.38
(approximate time
7.38
40 Years
61,38
61.38
Cumulative
Leakage
(gal/acre)
23.000
45.400
67.800
90.200
113.000
135.000
157.000
to steady-state)
166.000
897.000
Drain
Flux
(f>pad)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0,00
0.00
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
0
0
' 0.
Flux
into
Bottom
Liner
(gpnd)
32.96
40.25
46.21
50,66
53.93
56.28
57.91
58.40
60.24
Cumulative
Flux into
Bottom
Liner
(gal/acre)
11,000
24.400
40.200
58.000
77,100
97.300
118.000
126.000
843,000
Flux
from
Unit
(gpad)
30.66
38.08
44.18
49.01
52.65
55.31
57.23
57.79
59.98
Cumulative
Flux from
Unit
(gal/acre)
8.870
21,500
36,500
53.600
72.200
92,000
113.000
121.000
834,000
• f r.cv
-------
TABLE A2-15. DESIGN VARIABLES FOR SIMULATION AG-2
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
74 feet
25%
76 feet
f\0f
J-x>
20 feet at centerline
FML
1,406 gallons/acre-day
Uniform
Sand
10 cm/s
1 foot
11.5 cubic feet
Evenly spaced along lower slope
38 feet
Compacted soil
Clav
10~ cn/s
3 feet
125.9 cubic feet
Loan
10 cm/s
Hydrostatic from water table
10 feet belov top FML at centerline
-------
TABLE A2-16. EXPERIMENT AG-2--SIMULATION SUMMARY
Elapsed Leak
Time
(weeks)
1
2
3
4
5
6
7
(drain
7.16
8
12
16
20
24
28
32
Rate
(gpad)
1,801
1.697
1,648
1,607
1.570
1.534
1,491
flux begins)
1,481
1.448
1.428
' 1.420
1.417
1.416
1.416
1.416
Cumulative
Leakage
(gal/acre)
13,300
25.400
37.100
48,400
59,500
70.300
80.900
82.600
91.100
131.000
171.000
211,000
251.000
290.000
330.000
Drain
Flux
(gpad)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
618.38
733.74
800.43
863.14
889.38
889.38
889.38
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
0
0
2,410
22.700
44.400
67.200
91.200
116.000
141,000
Flux
into
Bottom
Liner
(gpad)
119.21
189.72
193.41
196.16
200.32
206.23
223.52
229.99
227.73
190.92
184.06
180.49
178.28
177.02
176.36
Cumulative
Flux into
Bottom
Liner
(gal/acre)
594
1.770
3.130
4.490
5.880
7.310
8,810
9,080
10,500
16,100
21,400
26,400'
31,500
36,400
41,400
Flux
from
Unit
(gpad)
32.44
33.16
39.95
45.49
48.38
53.20
62.46
64.34
57.62
127.99
149.67
160.82
167.45
171.25
173.28
Cumulative
Flux from
Unit
(gal/acre)
254
483
732
1.030
1.360
1.710
2.110
2.190
2.560
5.370
9.310
13.700
18.300
23.000
27.900
(approximate time to steady-state)
34.17
1.416
40 Years 1.415
351.000
20.700.000
889.38
873. fll
154.000
,'f O 5 >•' u a °
- <• ; /
12 700 000
176.14
175.61
44.100
2.570.000
173.99
175.61
30.500
2.550.000
' 0" f
-------
TABLE A2-35. DESIGN VARIABLES FOR SIMULATION BA-5
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial ir.cisture condition
Water Table
Elevation
Surface Impoundment
2,0 acres
150 feet
30 feet
25%
120 feet
2%
10 feet at centerline
FKL
800 gallons/acre-day
Uniform
Sand
10 cn/s
1 foot
1&.7 cubic feet
Evenly spaced along lower slope
60 feet
Compacted soil
/
cm/s
6 feet
261.1 cubic feet
Loam
10 cn/s
Hydrostatic from water table
10 feet below top FML at cen-terline
-------
TABLE A2-36. EXPERIMEOT BA-5—SIMULATION SUMMARY
Klopued
Time
(months)
1
2
Leak
Rate
(gpad)
1039.30
949.11
Cumulative
Leakage
(gal/acre)
33,800
63.900
Drain
Flux
(gpad)
0.00
0.00
Cumulative
Drainage
(gal/acre)
0
0
Flux
into
Bottom
Liner
(gpad)
127.77
143.15
Cumulative
Flux into
Dot tooi
Liner
(gal/acre)
3.330
7.470
Flux
from
Unit
(gpad)
50.27
61.46
Cumulative
Flux from
Unit
(gal/acre)
1.290
3,000
(drain flux begins)
2.83 877.69 87.000
0.00
164.15
11,300
68.22
(approximate time to steady-state)
9.75 794.72 255,000 631.45 127.000
40 Yeara 794.62 11.600.000 631.45 -4.-190T000
143.60
142.76
43.400
2.090.000
139.80
142.76
4,640
3
4
5
6
7
8
9
861.46
796.19
795.24
794.98
794.84
794.77
794.72
91.400
116,000
140,000
165.000
189.000
213.000
237.000
126.25
631.91
631.45
631.45
631.45
631.45
631.45
556
12,700
31,900
51.800
71,800
91.900
112.000
168.04
167.64
153.54
147.97
145.34
144.33
143.84
12,200
17,500
22.400
26.900
31,400
35.800
40,200
69.18
74.82
98.40
120.92
132.32
136.93
138.95
4,990
7,170
9.790
13,200
17.100
21,200
25.400
28.600
2.070.000
«. v
-------
TABLE A2-37. DESIGN VARIABLES FOR SIMULATION BB-1A
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section -width
Sideslope length
Sideslope grade
Lower slope length
Lov>er slope grade
Impounded liquid depth
Top Liner
Type .
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
25%
120 feet
2%
10 feet at centerline
FML
928 gallons/acre-day
Uniform
Sand
10 cm/s
1 foot
7.9 cubic feet
Evenly spaced along lower slope
60 feet
Compacted soil
Clav
10 cm/s
3 feet
125.5 cubic feet
Loam
10 cn/s
Hydrostatic from water table
10 feet below top FKL at centerline
-------
TABLE A2-38. EXPERIMENT BB-1A--SIMULATION SUMMARY
Elapsed
Time
(months)
2
4
6
8
10
12
14
16
18
20
22
24
26
Leak
Rate
(gpad)
1.059
986
953
936
928
924
922
921
921
921
921
921
921
(approximate time
26.71
AH V ^> n r- f
921
QT1
Cumulative
Leakage
(gal/acre)
70.800
133.000
192.000
249.000
306.000
362.000
418,000
474.000
530.000
586.000
642.000
698.000
754.000
to steady-state)
774.000
i ^ «;nn nnn
Drain
Flux
(gpad)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
n nn
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
n~ .
Flux
into
Bottom
Liner
(gpad)
606
746
826
873
897
909
914
916
917
917
917
918
918
918
O1 H
Cumulative
Flux into
Bottom
Liner
(gal/acre)
29.100
70.800
119.000
171.000
225.000
280.000
335.000
391.000
447,000
502.000
558,000
614.000
670,000
690.000
i i inn nnn
Flux
from
Unit
(gpad)
555
731
819
869
896
908
914
916
917
917
917
917
918
918
O1 R
Cumulative
Flux from
Unit
(gal/acre)
18.000
58.100
106.000
157.000
211.000
266.000
321.000
377.000
433.000
489.000
544.000
600.000
656.000
676.000
i T inn nnn
/ /
-------
Composite Soil Liners
With
FML Leakance - 7 x 10~14 s"1
-------
TABLE
DESIGN VARIABLES FOR SIMULATION AM-*/
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
25Z
120 feet
21
10 feet at centerline
FML
60 gallons/acre-day
Sidewall
Sand
10~J cm/g
1 foot
37.0 cubic feet
Evenly spaced along lover slope
60 feet
Composite
7 X 10-'" 5-'
Clay
10 cm/6
3 feet
cubic feet
Loam
10 cm/s
Hydrostatic from water table
10 feet below top FML at centerline
-------
TABLE
EXPERIMENT AM--V—SIMULATION SUMMARY
Elapsed Leak
Time Rate
fm.-,M-*.) (gpad)
2
4
6
8
10
12
14
16
61.11
60.44
60.56
60.50
60,42
60.34
60.27
60.19
Cumulat ive
Leakage
(gal/acre)
4, 100
7,770
11,500
15,100
18,800
22,500
26,200
29,800
Drain Cumulative
Flux Drainage
(gpad) (gal/acre)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0
0
0
0
0
0
0
0
Flux Cumulative
into Flux into
Bottom Bottom
Liner Liner
(gpad) (gal/acre)
0.016
0.017
0.017
0.017
0.017
0.018
0.018
0.018
1
2
3
4
5
6
7
8
Flux Cumulative
from Flux from
Unit Unit
(gpad) (gal/acre)
-0.0394
0.0231
0.0236
0.0363
0.0418
0.0369
0.0410
0.0391
-11
-10
-8
-6
-4
-1
1
4
(drain flux begins)
16.52
17
19
21
23
60.17
60.15
60.02
59.89
59.81
(approximate
23.79
40 Years
59.79
59.71
30,800
31,700
35,300
39,000
42,600
0.00
22.23
44.47 2,
44.47 5,
44.47 8,
0
323
880
590
300
0.018
0.019
0.019
0.019
0.019
9
9
10
11
12
0.0304
0.0362
0.0289
0.0415
0.0396
4
5
7
10
12
time to steady-state)
44,000
873,000
44.47 9,
#57,
44.47 638-,-
360
0 VO
^ee-
0.019
0.019
13
277
0.025
0.019
13
285
o .5
-------
TABLE
DESIGN VARIABLES FOR SIMULATION C6-5
Description
7alue
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lover slope length
Lover slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
-450 feet
25Z
120 feet
21
10 feet at centerline
FML
78O gallons/acre-day
Sand
10 cm/s
1 foot
35.4 cubic feet
Evenly spaced along lover slope
60 feet
Composite
7 X \0~l
-------
TABLE
EXPERIMENT CB-J—SIMULATION SUMMARY
Elapsed
Time
(wee: >\i }
1
2
3
4
(drain
4 .33
5
8
11
14
17
20
23
26
Leak
Rate
(gpad)
970
943
918
891
Cumulative
Leakage
(gal/acre)
6,870
13,500
20,000
26,300
Drain
Flux
(gpad)
0
0
0
0
Cumulative
Drainage
(gal/acre)
0
0
0
0
Flux
into
Bottom
Liner
(gpad)
0.017
0.018
0.018
0.019
Cumulative
Flux into
Bottom
Liner
(gal/acre)
0. 12
0.24
0.36
0.49
Flux Cumulative
from Flux from
Unit Unit
(gpad) (gal/acre)
-2.29
-2.24
-2.20
-2.08
-15
-31
-46
-61
flux begins)
881
863
791
779
778
778
777
777
777
(approximate
28.93
111
40 Years 777
28,400
32,500
49,700
66,100
82,400
98,700
115,000
131,000
148,000
0
133
543
745
726
757
761
765
752
0
553
8,070
22,600
38,200
54,100
69,900
85,800
101,000
0.019
0.019
0.021
0.021
0.021
0.021
0.021
0.021
0.021
0.54
0.63
1.06
1.50
1.95
2.40
2.84
3.29
3.73
-2.12
-2.01
-1.85
-1.69
-1.49
-1.37
-1.22
-1.12
-1.00
-66
-76
-117
-154
-187
-217
-244
-269
-291
time to steady-state)
164,000
11,400,000
745
/.
747 -p
117,000
i^M^L
0.021
0.021
4.17
307
-0.90
0.02
-311
-126
/ ." =
-------
TABLE
DESIGN VARIABLES FOR SIMULATION
Description
Value
General Design Parameters
Facility Type
Flan Area
Half-section
Sideslope length
Sideslope grade
Lover slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
ISO feet
30 feet
25%
120 feet
2Z
10 feet at centerline
FML
49 gallons/acre-day
Sidewall
Sand
10~J cm/s
1 foot
35.4 cubic feet
Evenly spaced along lover slope
60 feet
Composite
1 X 10"'^ s '
Clay
10 cm/s
3 feet
125.2 cubic feet
Loam
10 cm/s
Hydrostatic from vater table
10 feet below top FML at centerline
Comment: Tear in upper liner directly above tear in bottom liner.
-------
TABLE
EXPERIMENT ost-v~SIMULATION SUMMARY
Elapsed
Time
( h • )
3
6
9
12
15
18
(drain
20
22
24
26
28
30
Leak Cumulative
Rate Leakage
(gpad) (gal/acre)
52.87
52. 15
51.67
51.28
50.89
50.43
4,900
9,690
14,400
19,100
23,800
28,400
Drain
Flux
(gpad)
0
0
0
0
0
0
.00
.00
.00'
.00
.00
.00
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
Flux
into
Bottom
Liner
(gpad)
11.
10.
11.
11.
11.
11.
42
98
15
35
56
86
Cumulative
Flux into
Bottom
Liner
(gal/acre)
1,
2,
3,
4,
5,
6,
760
770
780
810
860
920
Flux
from
Unit
(gpad)
5.79
8.60
9.74
10.47
10.99
11.43
Cumulative
Flux from
Unit
(gal/acre)
-6
685
1,530
2,450
3,440
4,460
flux begins)
50.04
49.62
49.31
49.08
48.91
48.79
(approximate time
30.94
40 Years
48.74
48.47 7
31,500
34,500
37,500
40;500
43,500
46,400
0
22
44
44
44
44
.00
.23
.47
.47
.47
.47
0
1,350
2,850
4,950
7,520
9,820
12.
12.
12.
12.
12.
12.
12
38
56
63
78
85
7,
8,
9,
9,
10,
11,
650
400
160
920
700
500
11.71
12.01
12.29
12.48
12.64
12.75
5,160
5,890
6,630
7,380
8,140
8,920
to steady-state)
47,800
10,000
44
44
.47
.47
11,100
441 500
c, i n'' r\f\ n
U-l U , UU U
12.
13.
88
03
11,
190,
800
000
12.78
13.03
9,280
187,000
r<
tr. L
i»< tft
:-3,-
-------
Composite Soil Liners
With
FML Leakance - 3 x 10"11 s"1
-------
TABLE A3-5. DESIGN VARIABLES FOR SIMULATION CZ-1
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
252
120 feet
2%
10 feet at centerline
FML
92 gallons/acre-day
Uniform
Coarse Sand
10 cm/s
1 foot
4.6 cubic feet
Evenly spaced along lower slope
60 feet
Composite _1
3 X 10 s
Clay
10 cm/fi
3 feet
125.A cubic feet
Loam
10 cm/s
Hydrostatic from water table
10 feet below top FML at centerline
-------
TABLE A3-6. EXPERIMENT CZ-1—SIMULATION SUMMARY
Elapoed
Time
(months)
2
4
6
8
10
Leak
Rate
(gpad)
97.46
95.36
94.33
93.33
92.52
Cumulative
Leakage
(gal/acre)
6.020
11.900
17.600
23.400
29.000
Drain
Flux
(gpad)
0.00
0.00
0.00
0.00
0.00
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
Flux
into
Bottom
Liner
(gpad)
0.82
0.85
0.86
0.87
0.87
Cumulative
Flux into
Bottom
Liner
(gal/acre)
47
98
150
202
255
Flux
from
Unit
(gpad)
0.03
0.17
0.31
0.42
0.45
Cumulative
Flux from
Unit
(gal/acre)
3
9
24
46
74
(drain flux begins)
11.12
12
13
14
15
92.14
91.97
91.88
91.83
91.81
(approximate time to
15.45
40 Years
91.81
91.77
32,200
34.600
37.400
40.200
43.000
Bteady-etate)
44.300
1.340,000
0.00
53.64
74.20
83.85
87.56
88.74
90.54
0
993
3.000
5.440
8.050
9.260
/j i!/j ,500
— l_r290_0QO—
0.87
0.88
0.88
0.88
0.88
0.88
0.87
285
308
335
362
388
400
12.800
0.54
0.60
0.62
0.65
0.68
0.68
0.80
93
108
126
146
166
175
11.400
»fu,
rs
>y «•*.'
-------
TABLE A3-7. DESIGN VARIABLES FOR SIMULATION CZ-2
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
25%
120 feet
2Z
10 feet at centerline
FML
1238 gallons/acre-day
Uniform
Coarse Sand
10
cm/ s
1 foot
4.6 cubic feet
Evenly spaced along lower slope
60 feet
Composite _.
3 X 10~1^ s
Clay
10 cm/s
3 feet
125.4 cubic feet
Loam
10 cm/s
Hydrostatic from water table
10 feet below top FML at centerline
-------
TABLE A3-B. EXPERIMEhfT CZ-2--SIHULATION SUMMARY
Elapsed
Time
(weeks)
1
2
3
4
(drain
4.94
5
7
9
11
13
15
17
Leak
Rate
(gpad)
1.316
1.311
1.305
1.299
flux begins)
1,287
1.286
1,264
1.250
1,242
1,240
1,237
1,237
(approximate time to
17.92
1,237
40 Years 1.237
^'/"'r/
Cumulative
Leakage
(gal/acre)
9.330
18.500
27,700
36.800
45,300
45,800
63.700
81.300
98,700
116.000
133.000
151.000
steady-state)
159.000
18.100,000
/«/- f ^
Drain
Flux
(gpad)
0.00
0.00
0.00
0.00
0.00
24.63
332.85
826.00
1.075.61
1.172.79
1.2U9.31
1.224.67
1.226.87
1.235.09
S C /i ~ r-
Cumu] «.tive
Drainage
(gal/acre)
0
0
0
0
0
9
2.740
11.200
24.800
40.700
57.400
74.500
82,400
18.000,000
Vc" *° V
Flux
into
Bottom
Liner
(gpad)
0.82
0.87
0.89
0.90
0.90
0.90
0.92
0.92
0.93
0.93
0.93
0.93
0.93
0.93
>
Cumulative
Flux into
Bottom
Liner
(gal/acre)
5
11
17
24
30
30
43
56
69
82
95
108
114
13.500
"0 9*/'»os/
Flux
from
Unit
(gpad)
-0.25
-0.07
-0.08
0.01
0.07
0.03
-0.02
-0.03
-0.02
0.19
0.17
0.29
0.11
0.32
A. f C
Cumulative
Flux from
Unit
(gal/acre)
0.7
0.9
1.1
1.4
1.6
1.6
1.9
2.6
3.4
4.8
6.6
8.9
10
4.650
-------
Composite Soil Liners
With
FML Leakance - 3 x 10~11 s"1
-------
TABLE A5-1. DESIGN VARIABLES FOR SIMULATION AM-2
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
15%
120 feet
2%
10 feet at centerline
FML
60 gallons/acre-day
Sidewall
Sand
10 cm/s
1 foot
37.6 cubic feet
Evenly spaced along lower slope
60 feet
Composite 1
3 X 10 l s"1
Clay
10~ cm/s
3 feet
125.2 cubic feet
Loam
10 cm/s
Hydrostatic from water table
10 feet below top FML at centerline
-------
TABLE A5-2. EXPERIMENT AM-2--SIMULATION SUMMARY
Elapsed
Time
Leak Cumulative
Rate Leakage
(months) (gpad) (gal/acre)
2
4
6
8
10
12
14
16
18
(drain
18.35
20
22
24
61.24
60.53
60.66
60.62
60.55
60.48
60.41
60.35
60.29
flux begins)
60.27
60.19
60.07
59.99
(approximate time to
25.21
40 Year
59.95
e 59.86
4.100
7.790
11,500
15.200
18.900
22,500
26.200
29.900
33,600
34.200
37.200
40.900
44.500
steady-state)
46.700
875,000
Drain
Flux
(gnad)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
44.47
44.47
44.47
44.47
44.47
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
0
0
0
0
1.520
4.120
6.830
8.470
j. <*ft „-,, -, .>
— 6£4rOOO-
Flux
into
Bottom
Liner
(gpad)
6.48
6.40
6.37
6.37
6.39
6.43
6.49
6.58
6.69
6.72
6.77
6.80
6.82
6.82
6.80
Cumulative
Flux into
Bottom
Liner
(gal/acre)
398
789
1,180
1,570
1.950
2,340
2.740
3.130
3.540
3,610
3.950
4,360
4.780
5.030
99.100
Flux
from
Unit
(gpad)
-0.03
1.03
2.14
3.08
3.76
4.35
4.80
5.17
5.47
5.54
5.74
5.97
6.15
6.25
6.80
Cumulative
Flux from
Unit
(gal/acre)
-2
26
125
285
494
742
1,020
1,320
1,650
1.710
1,990
2.350
2.720
2,940
96,900
<-'•) ,,>•-., s -
-------
TABLE A3-3. DESIGN VARIABLES FOR SIMULATION CB-3
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leak rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevation
Surface Impoundment
2.0 acres
150 feet
30 feet
25Z
120 feet
21
10 feet at centerline
FML
92 gallons/acre-day
Uniform
Sand
10 cm/s
1 foot
4.6 cubic feet
Evenly spaced along lower slope
60 feet
Composite _.
3 X 10 s
Clay
10 cm/s
3 feet
125.4 cubic feet
Loam
10 cm/s
Hydrostatic from water table
10 feet below top FML at centerline
-------
TABLE A3-4. EXPERIMENT CD-3--SIMULATION SUMMARY
Elapsed
Time
(months)
1
2
3
4
5
6
7
8
9
10
11
Leak Cumulative
Rate
(gpad)
98.60
97.63
96.57
95.67
95.15
94.68
94.17
93.70
93.28
92.91
92.57
Leakage
(gal/ncre)
3.040
6,020
8,980
11.900
14,800
17.700
20.600
23.400
26.300
29.100
31,900
Drain
Flux
(gpad)
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Cumulative
Drainage
(gal/acre)
0
0
0
0
0
0
0
0
0
0
0
Flux
into
Bottom
Liner
(gpad)
7.41
7.65
7.72
7.75
7.74
7.71
7.70
7.69
7.68
7.67
7.66
Cumulative
Flux into
Bottom
Liner
(gal/acre)
219
449
683
919
1,150
1.390
1,620
1.860
2,090
2.320
2.560
Flux
from
Unit
(gpad)
0.13
0.27
0.89
1.50
2.12
2.80
3.46
3.87
4.32
4.74
5.06
Cumulative
Flux from
Unit
(gal/acre)
14
19
36
71
127
202
295
405
530
667
816
(drain flux begina)
11.69
12
14
16
18
92.36
92.28
92.07
92.02
92.00
(approximate time to
18.46
40 Years
z
92.00
91.98
~'
33.800
34.700
40.300
45.900
51.500
steady-state)
52.800
1.340.000
>W /;. ^
0.00
25.91
71.11
81.71
83.46
83.72
84.10
,<«
0
176
3.590
8.350
13.400
14.600
/, / ifj 000
« /r-v 4/ o e .
7.66
7.65
7.62
7.58
7.56
7.55
7.51
e*rf * 3 f
2.720
2.790
3.250
3.720
4.180
4.280
110,000
" 3 J 5 /?,^ , /, Slj /
5.29
5.37
5.90
6.30
6.59
6.64
7.05
/
*f e
925
975
1.320
1.690
2.080
2.180
101.000
-------
TABLE A3-27. DESIGN VARIABLES FOR SIMULATION CB-4A
Description
Value
General Design Parameters
Facility Type
Plan Area
Half-section width
Sideslope length
Sideslope grade
Lower slope length
Lower slope grade
Impounded liquid depth
Top Liner
Type
Average leal rate
Leak location
Drainage Layer
Soil type
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Drain System
Number of drains
Drain locations
Drain spacing
Bottom Liner
Type
Bottom FML Leakance
Soil
Saturated hydraulic conductivity
Thickness
Initial moisture storage
Native Soil
Soil Type
Saturated hydraulic conductivity
Initial moisture condition
Water Table
Elevat ion
Surface Impoundment
2.0 acres
150 feet
30 feet
25%
120 feet
2Z
10 feet at centerline
FML
778 gallon/acre-day
Uniform
Sand
10"** cm/s
1 foot
35.4 cubic feet
Evenly spaced along lower slope
60 feet
3 xlO' s'1
Clay
10 cm/s
3 feet
125.2 cubic feet
Loam
10 cm/s
Hydrostatic from water table
10 feet below top FML at centerline
-------
TABLE -
EXPERIMENT CB- 4 A—SIMULATION SUMMARY
Elapsed Leak
Time
(weeks)
1
2
3
4
(drain
4.33
5
25
45
65
85
105
125
Rate
(gpad)
970
944
918
892
flux begins)
882
864
778
778
778
778
778
778
Cumulative
Leakage
(gal/acre)
6,870
13.500
20.000
26,400
28.400
32.500
142,000
251,000
360,000
469.000
578,000
687.000
Drain
Flux
(gpad)
0
0
0
0
0
129
745
767
756
756
723
762
Cumulative
Drainage
(gal/acre)
0
0
0
* 0
0
446
95.400
200.000
305.000
409.000
513.000
617,000
Flux
into
Bottom
Liner
(gpad)
7.19
7.38
7.56
7.76
7.84
7.97
8.16
7.88
7.75
7.68
7.65
7.63
Cumulative
Flux into
Bottom
Liner
(gal/acre)
50
101
153
207
225
262
1.430
2.550
3.650
4.730
5.800
6.870
Flux
from
Unit
(gpad)
-6.78
-6.21
-5.69
-5.13
-4.97
-4.60
1.78
4.71
6.18
6.91
7.27
7.45
Cumulative
Flux from
Unit
(gal/acre)
-36
-81
-123
-161
-173
-195
-297
177
952
1.880
2,870
3.900
(approximate time to steady-state)
133.53
778
40 Years 778
L./ ~> u , */ ~ y «:
o
733.000
11.400,000
.',c/ ,*
------- |