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/ ,* 
-------