LEACHATE CLOGGING ASSESSMENT OF
GEOTEXTILE AND SOIL LANDFILL FILTERS
by
Robert M. Koerner and George R. Koerner
Geosynthetic Research Institute
Drexel University
Philadelphia, Pennsylvania 19104
Cooperative Agreement
CR-819371
Project Officer
Robert E. Landreth
Land Remediation and Pollution Control Division
National Risk Management Research Laboratory
Cincinnati, Ohio 45268
NATIONAL RISK MANAGEMENT RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
CONTACT
Robert Landreth is the EPA contact for this report. He is presently with the newly
organized National Risk Management Research Laboratory's new Land Remediation and Pollution
Control Division in Cincinnati, OH (formerly the Risk Reduction Engineering Laboratory). The
National Risk Management Research Laboratory is headquartered in Cincinnati, OH, and is now
responsible for research conducted by the Land Remediation and Pollution Control Division in
Cincinnati.
-------
DISCLAIMER
The information in this document has been funded wholly by the United States
Environmental Protection Agency under Cooperative Agreement No. CR-819371 to the
Geosynthetic Research Institute of Drexel University in Philadelphia, Pennsylvania. It has been
subjected to the Agency's peer and administrative review, and it has been approved for publication
as an EPA document. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
ii
-------
FOREWORD
The U. S. Environmental Protection Agency is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws, the
Agency strives to formulate and implement actions leading to a compatible balance between human
activities and the ability of natural systems to support and nature life. To meet this mandate,
EPA's research program is providing data and technical support for solving environmental
problems today and building a science knowledge base necessary to manage our ecological
resources wisely, understand how pollutants affect our health, and prevent or reduce
environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks from threats to
human health and the environment. The focus of the Laboratory's research program is on
methods for the prevention and control of pollution to air, land, water and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites and ground
water; and prevention and control of indoor air pollution. The goal of this research effort is to
catalyze development and implementation of innovative, cost-effective environmental technologies;
develop scientific and engineering information needed by EPA to support regulatory and policy
decisions; and provide technical support and information transfer to ensure effective
implementation of environmental regulations and strategies.
This publication has been produced as part of the Laboratory's strategic long-term research
plan. It is published and made available by EPA's Office of Research and Development to assist
the user community and to link researchers with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
ill
-------
ABSTRACT
The liquids management strategy for any municipal or hazardous waste landfill requires a
knowledgeable design strategy for the leachate collection system located at the base of the waste
mass. Such leachate collection systems generally consist of sumps, perforated pipes, drainage
materials (gravel soils or geonets) and filter materials (sand soils or geotextiles). The solid waste
mass lies above the filter, although sometimes a protective soil acts as a intermediate layer. As
leachate migrates through the waste mass it must be intercepted, collected and transported by the
leachate collection system. Since leachate is often high in suspended solids and microorganism
content, concerns over excessive clogging of the leachate collection system are often expressed.
More specifically, the filter is the target material due to its small opening spaces with respect to the
other materials that are involved. Thus, this project is completely oriented toward the filter
material of the leachate collection system. Both sand soil and geotextile filters are investigated,
although more emphasis is on geotextiles due to their greater current usage in this application.
A multifaceted approach leading to a design methodology was the focus of this study. The
project consisted of exhuming four sites-of-opportunity which essentially established ground
truth. Three of the sites had excessively clogged geotextiles and the fourth was marginally
adequate. Parallel to the field work was the laboratory evaluation of 12 commonly used filters; (10
geotextiles and 2 soils) using four different permeating liquids (water and 3 different leachates)
under three different accelerated leachate flow rates. This required 144 ASTM D1987 flow
columns to be constructed. Each were used for time periods of up to one year in order to establish
equilibrium permeability values. From this data, master curves of the twelve filters were
generated. Parallel to the field and laboratory efforts just noted was a computer modeling effort
(using the HELP model) to investigate flow rate sensitivity to different leachate collection materials
and to obtain the required, i.e., site specific, flow rates for the four exhumed sites.
These three activities (field exhuming, laboratory testing and computer analysis) were
brought together in a design model which can be used to calculate the factor-of-safety for a
leachate collection system filter in any site specific and material specific design. The formula is as
follows.
k,,
FS = allow
where
FS = factbr-of-safety against excessive filter clogging
kallow = allowable permeability for the specific filter material being considered
kreqd = required permeability for the site specific hydraulic and solid waste situation
DCF = drainage correction factor for the particular design geometry being considered
The DCF is a unique addition to the conventional type of factor-of-safety equations in the literature
and is necessary because some designs significantly limit the available flow area downstream of
iv
-------
the filter. Indeed, the incorporation of the DCF completely substantiated the field findings of the
exhumed sites. Example calculations of DCF for common design geometries are presented in the
report.
Some ancillary studies were also included, such as the "no filter" strategy and the use of
biocides within the various filter materials. Such design strategies are indeed possible but are felt
to be unnecessary in light of utilization of the design method. Clearly, landfill filters can be
designed so as not to result in excessive clogging of leachate collection systems. Design guidance
is presented in this report. For the case of mild leachates, recommendations are offered as to
typical geotextile filters which should result in conservative factors-of-safety against excessive
filter clogging. For more aggressive leachates, the procedures, test methods and computer
generated design flows detailed in this report must be utilized accordingly.
-------
vl
-------
CONTENTS
Disclaimer
Foreword
111
Abstract jv
List of Figures ix
List of Tables xv
Acknowledgements xvii
1. Introduction and Scope of Project 1
2. Sources of Leachate Collection Filter Clogging 7
3. Results of Previous Investigation 12
3.1 Phase I - Initial Flow Rate Evaluations 12
3.2 Phase E(a) - Improved Flow Rate Columns and Remediation Attempts 14
3.3 Phase H(b) - Biocide Treated Geosynthetics 17
3.4 Summary of the Previous Investigation 38
3.5 Recommendations of the Previous Investigation 20
4. Overview of This Project 21
5. Field Exhuming of Leachate Collection Systems 23
5.1 Field Exhuming Details 23
5.2 Field Exhumed Site #1 23
5.3 Field Exhumed Site #2 29
5.4 Field Exhumed Site #3 34
.5.5 Field Exhumed Site #4 39
5.6 Summary of Field Exhuming Study 43
6. Laboratory Investigations for Allowable Flow Rates 46
6.1 TestSet-Up 46
6.2 Test Procedure 51
6.3 Leachate, Filters and Flow Rates Used in Tests , 55
6.4 Test Results from Very High Flow Rates 57
6.5 Test Results from High and Intermediate Flow Rates 64
6.6 Comparison of Results Using Different Flow Rates 66
6.7 Summary of Laboratory Permeability Testing 66
7. Modeling for Required Flow Rates 71
7.1 Overview of the HELP Model 71
7.2 Use of the HELP Model for Exhumed Sites 72
7.3 Summary of HELP Model Utilization 75
vii
-------
Page
8. Design Method and Substantiation 77
8.1 Design Methodology and the Drainage Correction Factor 77
8.2 Substantiation of Design Methodology gl
9. Conclusions and Recommendations g2
10. References 87
Appendices
A. Results from Biocide Treated Geosynthetics 89
A-l Introduction go
A-2 Type of Biocide go,
A-3 Incorporation of the Biocide into Different Geosynthetics 90
A-4 Field Testing and Evaluation Procedures 91
A-5 Results of Test Series "A" 04
A-6 Results of Test Series "B" 04
A-7 Results of Test Series "C" 99
A-8 Conclusions of the Biocide Study 1Q2
B, Permeability Results for "High" and "Intermediate" Flow Rates 104
C. HELP Model Parametric Analysis
C-l Overview
C-2 Site Description
C-3 Landfill Description
C-4 Landfill with No Geotextile Filter/Separator
C-5 Landfill with a Geotextile Filter/Separator
C-6 General Comments
C-7 Conclusions of the HELP Model Analysis
D. The "No Filter" Design Scenario
D-l Overview '
D-2 Design Considerations [
D-3 "No Filter" Experiments and Results
D-4 Conclusions of the "No Filter" Study f
viii
-------
LIST OF FIGURES
Figure No.
1 Generalized Cross Section of Double Lined Landfill Liner and
Cover System 2
2 Common Cross Section of Leachate Collection System 3
3 Location of Removal Pipes Within a Drainage Layer 4
4 Mosaic of a Woven Geotextile Filter Under Very High Flow Rate
Conditions for 1, 3, 6 and 12 Month Time Periods at SOX Magnification 10
5 Mosaic of a Nonwoven Geotextile Filter Under Very High How Rate
Conditions for 1, 3, 6 and 12 Month Time Periods at 400X Magnification 11
6 Average Response of 96 Flow Rate Columns from Phase H(a) Activities 16
7 Flow Chart for This Project 22
8 Plan Views of Site #1 24
9 Cross Sectional Views of Site #1 25
10 Photographs of Site #1 27
11 Aerial and Plan of Site #2 30
12 Cross Section Views of Site #2 31
13 Photographs of Site #2 33
14(a) Plan View of Site #3 Showing the Dimensions of the Cell as well as the
Exhumed Area with Respect to the Different Waste Disposal Locations 35
14(b) Isometric View of Site #3 Showing the Geotextile Filter Separating the
Slurried Waste from the Pea Gravel and a Second Geotextile Filter Beneath
the Pea Gravel Which Also Socked the Perforated Pipe 35
15 Photographs of Site #3 36
16 Grain Size Distribution of Soils and Sludge at Site #3 38
17(a) Plan View of Site #4 Showing the Configuration and Dimensions of the
Cells. The Exhumed Wells were in the Area of the Oldest Cells and
Very Close to the Methane Condensing and Cleaning Plant 40
17(b) Isometric View of Site #4 Showing the Geotextile Filter Socking the
Perforated Pipe Which Was Used as Excavation Wells 40
18 Photographs of Site #4 42
ix
-------
Figure No.
Page
19 Various Leachate Collection Filter Configurations 44
20
--1987 Test Method ° ' ~"' ™t"M* "* AS™ 4g
21 Flow Columns and Support System with Associated Pipe Network 49
22 Reservoir Recirculation System
23 Plan View of The 48-Permeameter Set-Up 52
24 Sand Response Curves After 120 Days of Leachate Permeation 58
25 Woven Response Curves After 120 Days of Leachate Permeation 60
26 Nonwoven Needlepunched Curves After 120 Days of Leachate Permeation 61
27 Nonwoven Curves After 120 Days of Leachate Permeation 63
28 Response Curves of Experimental Geotextiles After 120 Davs of
Leachate Permeation ,.
o5
29(a) How Rate Curves for Sand Soils
29(b) How Rate Curves for Woven Geotextiles 69
29(c) Flow Rate Curves for Nonwoven Geotextiles 70
29(d) Master Curves for all Twelve Filters Evaluated (2 Sands and 10
Geotextiles) and Estimated Upper and Lower Bounds for
Extrapolation Back to'Typical" Site Specific Hux Values 70
30 HELP Model Simulation Process 71
31 HELP Model generated Values of Filter Permeability for Exhumed Sites 73
32 Various Leachate Collection Filter Configurations with the Associated
Range of Drainage Correction Factors 79
-------
APPENDIX A las
A-1 Effect of Geonet Biocide Content on System Permeability Under
Anaerobic Conditions (Test Series "A") 95
A-2 Effect of Geonet Biocide Content on System Permeability Under
Aerobic Conditions (Test Series "A") 96
A-3 Effect of Geotextile Biocide Content on System Permeability Under
Anaerobic Conditions (Test Series "B") 97
A-4 Effect of Geotextile Biocide Content on System Permeability Under
Aerobic Conditions (Test Series "B") 9g
A-5 Nonwoven Needle-Punched Effect of Geotextile Biocide Content on
System Permeability (Test Series "C") 100
A-6 Comparison of Nonwoven Needle-Punched Geotextiles with Varvine
Biocide Content (Test Series "C") 1()1
A-7 Comparison of Woven Monofilament Geotextile Opening Size with
Varying Biocide Content (Test Series "C") 103
APPENDIX B
B-1 Permeability Response Curves for Ottawa Sand Permeated with 4 Fluids at
a High Flow Rate of 3.8 I/week (1.0 gal/week) 105
B-2 Permeability Response Curves for Concrete Sand Permeated with 4 Fluids
at a High Flow Rate of 3.8 I/week (1.0 gal/week) 105
B-3 Permeability Response Curves for Geotextile Filter N 7 W Permeated with
4 Fluids at a High Flow Rate of 3.8 I/week (1.0 gal/week) 106
B-4 Permeability Response Curves for Geotextile Filter N 14 W Permeated
with 4 Fluids at a High Flow Rate of 3.8 I/week (1.0 gal/week) 106
B-5 Permeability Response Curves for Geotextile Filter N 32 W Permeated
with 4 Fluids at a High Flow Rate of 3.8 I/week (1.0 gal/week) 106
B-6 Permeability Curves for Geotextile Filter H 4 NPNW Permeated with
4 Fluids at a High Flow Rate of 3.8 I/week (1.0 gal/week) 10?
B-7 Permeability Curves for Geotextile Filter H 8 NPNW Permeated with
4 Fluids at a High How Rate of 3.8 I/week (1.0 gal/week) 107
B-8 Permeability Curves for Geotextile Filter H 16 NPNW Permeated with
4 Fluids at a High How Rate of 3.8 I/week (1.0 gal/week) 107
B-9 Permeability Curves for Geotextile Filter T 4 HBNW Permeated with
4 Huids at a High How Rate of 3.8 I/week (1.0 gal/week) 108
xi
-------
Figure No. _
Page
B-10 Permeability Curves for Geotextile Filter P 6 NPNW Permeated with
4 Fluids at a High Flow Rate of 3.8 I/week (1.0 gal/week) 108
B-11 Permeability Curves for Geotextile Filter A 10 W Permeated with 4
Fluids at a High Flow Rate of 3.8 I/week (1.0 gal/week) 109
B-12 Permeability Curves for Geotextile Filter N 22 NW/W Permeated with
4 Fluids at a High How Rate of 3.8 I/week (1.0 gal/week) 109
B-13 Permeability Response Curves for Ottawa Sand Permeated with 4 Fluids
at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week) no
B-14 Permeability Response Curves for Concrete Sand Permeated with 4 Fluids
at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week) 110
B-15 Permeability Response Curves for Geotextile Filter N 7 W Permeated with
4 Fluids at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week) 111
B-16 Permeability Response Curves for Geotextile Filter N 14 W Permeated with
4 Fluids at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week) 111
B-17 Penneability Response Curves for Geotextile Filter N 32 W Permeated with
4 Fluids at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week) 111
B-18 Permeability Curves for Geotextile Filter H 4 NPNW permeated with
4 Fluids at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week) 112
B-19 Permeability Curves for Geotextile Filter H 8 NPNW permeated with
4 Fluids at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week) 112
B-20 Penneability Curves for Geotextile Filter H 16 NPNW Permeated with
4 Fluids at a Intermediate How Rate of 0.95 I/week (0.25 gal/week) 112
B-21 Penneability Curves for Geotextile Filter T 4 NPNW Permeated with
4 Huids at a Intermediate How Rate of 0.95 I/week (0.25 gal/week) 113
B-22 Penneability Curves for Geotextile Filter P 6 NPNW Permeated with
4 Huids at a Intermediate How Rate of 0.95 I/week (0.25 gal/week) 113
B-23 Penneability curves for Geotextile Filter A 10 W Permeated with 4 Huids
at a Intermediate How Rate of 0.95 I/week (0.25 gal/week) 114
B-24 Penneability Curves for Geotextile Filter N 22 NWAV Permeated with
4 Huids at a Intermediate How Rate of 0.95 I/week (0.25 gal/week) 114
xii
-------
APPENDIX r
Leachate Collect™ System Used
1 18
i of Lateral Drainage Layer on Peak Daily Discharge
121
^^S^**MD^^<»'**™»
C-5 Effect of Evaporation Depth on Peak Daily Discharge (No Geotextile) 122
C-6 Effect of Runoff Curve Number on Peak Daily Discharge (No Geotextile) 122
C"? (NbSSS ***"* ^ PermeabUity on Peak D^ Discharge
«-mAaR,l,HM ,M~ r---.„..:,_, ^
*$£ My Discharge Assuming the Geotextile to be a Barrier Layer for
Different Waste Heights up to 100 ft y
127
C-10 Effect ofGeotextile Permeability on Peak Daily Discharge (K for
Lateral Drainage Layer =1 cm/sec) * 129
C-l 1 Effect of Geotextile Permeability on Peak Daily Discharge (K for
Lateral Drainage Layer = 0.1 cm/sec) 13()
C-12 Effect of Opnt^vtn* pom,.obility Peak Qn Daily Discharge (K for
131
i)ility on Daily Discharge (K for Lateral
cm/sec) 13"'
C-14 Variation in Peak Daily Discharge Assuming Geotextile to be a Vertical
Percolating Layer or a Barrier Layer for 5 ft Waste ! 34
C-15 Variation in Average Annual Totals Assuming Geotextile to be a Vertical
Percolating Layer or a Barrier Layer for 5 ft Waste 135
Vertical
136
C"1? K^l^r^r:!:""?" ™aimmi^^ "eoiexuie to oe a Barrier Layer for
136
xiii
-------
EgureNa
C-18 Average Annual Totals Assuming the Geotextile to be a Barrier Layer for
Different Waste Heights up to 100 ft 136
C-19 Effect of Lateral Drainage Layer and Geotextile Permeability on Peak
Daily Discharge (For No Waste Condition) 13g
C-20 Effect of Lateral Drainage Layer and Geotextile Permeability on Peak
Daily Discharge (For 5 ft Waste Condition) 138
C-21 Effect of Waste Height on Daily Discharge for Different Geotextile
Permeabilities (K for Lateral Drainage Layer = 0.1 cm/sec) 139
C-22 Effect of Waste Height on Daily Discharge for Different Geotextile
Permeabilities (K for Lateral Drainage Layer = 0.01 cm/sec) 140
C-23 Effect of Waste and Geotextile Permeability on Peak Daily Discharge
For 5 ft Waste (K for Lateral Drainage Layer = 1 cm/sec) 141
C-24 Effect of Geotextile Permeability on Daily Discharge at Seattle,
Washington (K for Lateral Drainage Layer = 1 cm/sec) ' 143
C-25 Effect of Geotextile Permeability on Daily Discharge at Phoenix,
Arizona (K for Lateral Drainage Layer = 1 cm/sec) ' 144
C-26 Effect of Geotextile Permeability on Daily Discharge for Different Cities
(For No Waste Condition)
C-27 Effect of Geotextile Permeability on Daily Discharge for Different Cities
(For 5 ft Waste Condition) 146
C-28 Effect of Geotextile Permeability on Daily Discharge For Different
Cities (For 10 ft Waste Condition) 147
C-29 Effect of Geotextile Permeability on Daily Discharge For Different
Cities (For 50 ft Waste Condition) 147
C-30 Effect of Geotextile Permeability on Daily Discharge For Different
Cities (For 100 ft Waste Condition) 147
APPENDED D
D-l Long Term Flow Permeameters Used for the Testing of Municipal
Solid Waste Placed Directly Above Granular Soils Without the Use
of Sand or Geotextile Filters Ts4
D-2 Particle Size Distribution Curves of Eight Soils Used in This Study I 155
D-3 Long Term How Results as per ASTMD1987 J 157
xiv
-------
LIST OF TABLES
labieNa
1 Range of Leachate Characteristics, after Chain and de Walle [4] 7
2 Indicators of Precipitate Clogging Parameters, after Driscoll [6] 9
3 Details of Municipal Solid Waste Landfill Leachates Evaluated and
Average Leachate Characteristics 12
4 Overview of the Leachate Collection Sites Exhumed 23
5 Results of Laboratory Tests on Materials of Site #1 28
6 Results of Laboratory Tests on Materials of Site #2 32
7 Permittivity Test Results for Site #3 (ASTM D-4491) 37
8 Permittivity Test Results of the Geotextiles of Site #4 41
9 Summary of the Leachate Characteristics of the Exhumed Field Sites 45
10 Overview of Exhumed Leachate Collection Systems 45
11 Description of Average Leachate Characteristics Evaluated in This Study 55
12 Description of Geotextile Characteristics Evaluated in This Study 55
13 Flow Rate Values Used for Accelerated Testing f-Safety Equation
(Equation 7) as Applied to Four Exhumed Field Sites 81
18 Recommended Geotextile Filters for Use with Relatively Mild
Landfill Leachates Which Have Low TSS and Low BOD5 Values 86
XV
-------
lableNa
APPENDIX A
A-l Conditions Within Flow Columns for Biocide Study 92
APPENDIX C
C-1 Description and Soil Properties of the Different Layers 119
C-2 Parameters Selected for the Sensitivity Analysis 120
C-3 Parameters Selected for the "No Filter" Parametric Study . 123
C-4 Parameters Selected for the Parametric Study with Filter 126
APPENDIX D
D-1 Granular Soil Characteristics for "No Filter" Permeability Study
xvi
-------
ACKNOWLEDGEMENTS
This work was funded by the U.S. Environmental Protection Agency under Cooperative
AgreementCR-819371 with Mr. Robert E. Landrethas the Project Officer. The support of the
Agency and the technical and administrative interaction with the Project Officer are sincerely
appreciated.
The cooperation of many municipal and hazardous waste landfill owner/operators is also
appreciated. At their request they wish to remain anonymous. Without such "ground truth"
investigations, however, the results of the study would not have been as definitive as we feel they
are.
The authors also thank the following EPA reviewers whose comments and suggestions
added greatly to the presentation of the results in the form of this final report.
Dr. Craig H. Benson - University of Wisconsin-Madison
Dr. Shobha K. Bhatia - Syracuse University
Dr. John J. Bowders, Jr. - University of Texas at Austin
Mr. Larry Lydick - National Seal Co.
Mr. Robert E. Mackey - Post, Buckley, Schuh & Jernigan, Inc.
Mr. Leo K. Overman - Colder Construction Services, Inc.
Mr. Larry W. Well - CH2M Hill, Inc.
xvii
-------
1. Introduction and Scope of Project
The concept of the containment system at a landfill is to envelope the solid waste and
isolate it from the surrounding environment. Insofar as total containment is concerned, landfills
consist of three major parts: A liner system beneath the waste, the liquids management system
within the waste, and cover system above the waste. Each system is quite complex having a
number of geosynthetic and natural soil components involved. While the cross section of Figure 1
emphasizes the use of geosynthetics in double lined facilities, it also clearly identifies the three
essential parts of a landfill containment system..
This report, and the study upon which it was developed, does not address the liner or
cover system, per se. Rather it addresses the liquids management system and, in particular, the
leachate collection system at the bottom of the solid waste mass, see Figure 2. Irrespective of the
method of ultimately handling the leachate, i.e., withdrawal on demand or leachate recycling, it
must flow within a drainage system consisting of granular soil or a geonet composite to a
perforated pipe network. Cross sections of several types of drainage and pipe systems are shown
in Figure 3. Once the leachate drains into the pipe, the flow rate increases rapidly and it continues
to flow gravitationallyto a low area, i.e., a sump, where it accumulates. From the sump, a
manhole or pipe riser allows for pumping and eventual removal of the leachate as per the site-
specific plan and permit.
Leachate collection systems must remain functioning for the service life and postclosure
care period of a landfill. Time frames of 30 years are generally required, although in reality, the
postclosure period could be longer depending upon the site-specific liquids management scheme,
e.g., those landfills using leachate recycling schemes.
A typical leachate collection system consists of the following components shown in Figure
2 (from the bottom of the solid waste downward):
• Filter layer—either sand or geotextile.
• Drainage material—either coarse sand, gravel, geonet or geocomposite.
• Perforated pipe—required for sand or gravel drainage materials.
• Sump—low area in the facility from which extends a vertical manhole or sidewall riser
in order to remove the leachate.
Note that a "soil protection layer" above the layer has not been specifically identified. When used,
this protection layer is usually native soil, and is often very low in its hydraulic conductivity or
permeability. If the permeability of the protection layer is less than that of the filter, it becomes a
de-facto liner and dominates the flow regime of the leachate collection system. The undesirable
result of "perched leachate" in the body of the waste would result, with possible cover seeps
and/or high hydraulic pressure exerted against the base and side slopes of the facility. Thus if a
soil protection layer is used above the filter, its permeability should be equal to that of the filter, or
higher, so as not to inhibit the flow of leachate. Note that "permeability", rather than hydraulic
conductivity will be used hi this report since it is the common term used in geosynthetics which is
the majority of this study.
-------
&H
-------
Dnfc
V V V V
Solid Waste
Solid Waste
Geotextile Filter
of concern
Drainage Gravel
Primary
eomembrane
Geotextile Cushion
Figure 2 - Common Cross Section of Leachate Collection System
-------
Excavated Trench Width
(a) Trench Type of Installation
(drawing not to scale)
Waste
Coarse
Drainage
Stone
Waste
(b) Embankment Type of Installation
(drawing not to scale)
Coarse
Drainage
Stone
(c) Embankment with V-Trench Type of Installation
(drawing not to scale)
Drainage
Layer
Geotextile
(Protection
Layer)
Geomembrane
Compacted
Clay Liner
Drainage
Layer
Geotextile
(Protection
Layer)
Geomembrane
Compacted
Clay Liner
Drainage
Layer
Geotextile
(Protection
Layer)
Geomembrane
Compacted
Clay Liner
Figure 3 - Location of Removal Pipes Within a Drainage Layer
4
-------
Considering the components of a leachate collection and removal system as just described,
it is felt that the filter layer will be the first component to excessively clog, if the phenomenon does
indeed occur. This is because the filter is the first layer of the drainage system that the leachate
encounters (thus the leachate has its highest sediment and microorganism content) and the voids of
the filter are the smallest of any component of the drainage system.
While such a filter could be either granular soil or geotextile, the current trend is toward the
use of geotextiles for a number of reasons:
• The savings in thickness provides additional landfill volume for placement of waste.
« A wide variety of geotextiles are readily available with various opening sizes via quality
controlled-manufacturing processes.
• Geotextiles are easily placed even in tight or poorly accessible locations.
• Geotextiles must be used to cover geosynthetic drainage systems, e.g., for geonets or
geocomposites, due to the large open spaces of these materials.
Regarding the design elements of any type of filter, including those used for leachate
collection systems, there are three primary considerations. The first and foremost filtration
criterion is to allow adequate flow of leachate into the underlying drainage layer so that a hydraulic
head does not build up in the solid waste mass. The second criterion is to adequately retain the
particles above the filter layer (solid waste or protection layer soil) so that the downstream drain
itself does not become excessively clogged with fine-grained particles. The third criterion is to
design the filter against long-term excessive clogging.
Hence, the proper design of a filter must establish a balance between flow capacity and
upstream particle retention as well as guard against the potential problem of long-term excessive
clogging. .
It must be recognized that some degree of flow reduction, i.e., clogging, of the filter is to
be expected. Such clogging can occur without adversely affecting the drainage system, at least
until the clogged filter begins to "starve" the downstream drain. At that point, leachate will begin
to build up into the solid waste as with the low permeability soil protection layers described earlier.
The implication of such buildup, called "perched leachate," is unknown but probably is not
desirable. In the extreme case, leachate may exit the cover soil of the facility in the form of
leachate seeps and exert hydraulic pressure on the base and side slopes which are the location of
sumps and pipe penetrations. Furthermore, if leachate recycling is the liquids management
scheme, the situation is aggravated due to the constant reintroduction of leachate back into the
waste mass. Leachate recycling will simply not work if the leachate collection filter becomes
excessively clogged. Therefore, "excessive clogging" is defined as the point in time when the
permeability of the filter renders the downstream leachate collection system ineffective for the site
specific application.
There exists many analytic attempts at achieving the above three criteria of adequate
permeability, proper retention of upstream soil, and long-term flow equilibrium. Christopher and
Fisher [1] summarize the existing criteria in each of the three above mentioned areas. Illustrative
-------
of many of these criteria is that recommended by the Federal Highway Administration as proposed
by Christopher and Holtz [2]. This set of criteria has recently been substantiated in the exhuming
of a series of different highway drainage systems (Koerner, et al. [3]).
It is felt, however, that the state-of-the-practice for filters in highway applications should
be challenged for the design of filters used in landfill leachate collection systems in regards to the
excessively clogging situation. This is primarily due to the fundamental differences between
leachate and water and the critical nature of the essentially inaccessible filter of a landfill leachate
collection system. Specifically, it is felt that the permeability and clogging criteria should be linked
together and that an overall global factor-of-safety against excessive clogging be introduced. For
the general landfill situation, soil retention is probably not as major of an issue as is permeability
and excessive clogging.
This report presents the findings of a multi-phased study in light of the previous
discussion. Field behavior, laboratory testing, computer modeling and design method
development are presented. In all cases focus is on the filter component of the leachate collection
system. Both geotextile and soil filters are addressed, but (as mentioned previously) the emphasis
is on geotextiles since they are used much more widely than sand filters at this time. Additionally,
the work has direct applicability to the drainage soil, sand or gravel, if no filter is used at all. This
"no filter" strategy is also investigated and is presented in an Appendix.
-------
2. Sources of Leachate Collection Filter Clogging
As stated in the introduction, the filter zone above the leachate collection system is an ideal
location for excessive clogging from several sources within the leachate. These sources include
paniculate clogging, biological clogging and precipitate clogging. The range of leachate
characteristics in Table 1 is such that particulates, microorganisms and precipitates are all common
to municipal solid waste (MSW) leachates. It is of interest to note that hazardous waste leachates
may be significantly less troublesome to leachate collection filters than MSW leachates, unless
such waste is co-disposed waste or has other extenuating circumstances.
Table 1 - Range of Leachate Characteristics, after Chain and de Walle [4]
Potential Clogging
Mechanism
Particulate
Property of Concern
Range of Values
(mg/1)
pH
Total solids (TS)
Total dissolved solids (TDS)
Total suspended solids (TSS)
j. vioi auaj^/vsiiu^u 3uu.ua ^ i oo,J
Chemical oxygen demand (COD)
Biochemical oxygen demand (BOD)
Total organic carbon (TOC)
3.7 - 8.5*
0 - 59,200
584 - 44,900
10-700
Biological
Precipitate
40 - 89,520
81-33,360
256 - 28,000
Specific conductance
Alkalinity (CaCO3)
Hardness (CaCO3)
Total phosphorus
Ammonia
Nitrate
Calcium
Chlorine
Sodium
Sulfate
Manganese
Magnesium
2,810 - 16,800
0 - 20,800
0 - 22,800
0-130
0- 1,106
0.2 - 10.29
60 - 7,200
4.7 - 2.467
0 - 7,700
1 - 1,558
0.09 - 125
17 - 15,600
*pH is in pH units, all others are in units of mg/1
Paniculate clogging is rather self explanatory. It is merely the settling out of suspended
particles from the leachate. Particulate clogging generally occurs at the upper surface of a filter.
This is sometimes referred to as formation of a surface cake. In contrast to this surface
phenomenon, depth filtration within the filter may also occur. This is the situation where
successively smaller particles in the leachate are removed from suspension within the thickness of
the filter. The mechanism of depth filtration incorporates embedment of suspended particles from
the leachate in the pores of the filter. In so doing the sediment that collects within the filter
acquires a gradation from large to small particles. It is possible that depth filtration is more
conductive to fluid flow than surface or cake filtration [5].
The clogging associated with biological growth is quite complex. It occurs when
microorganisms metabolize on and within the filter material. Biological growth depends on the
-------
presence of microorganisms, appropriate nutrients and environmental conditions which sustain
growth. Factors which influence biologicaLclogging include carbon-to-nitrogen ratio of the
leachate, rate of nutrient supply, concentration of micro and macro nutriets, moisture conditions of
the waste mass and temperature.
Anaerobic rather than aerobic conditions prevail in most locations in a landfill. Hence,
methanogenic bacteria are most prevalent. The quantity and quality of nutrients available to the
methanogenic bacteria is significant in this metabolism. Nutrients required by methanogenic
bacteria include; carbon, hydrogen, oxygen, nitrogen and phosphorous. In addition, they require
limited concentrations of trace metals such as sodium, potassium, calcium, and magnesium.
Biofilrn growth starts on the surface of the filter media and eventually incrusts it. Upon incrusting
the filter surface, it then moves into the pore spaces and further reduces flow through the filter.
Current practice at many landfills is to co-dispose sewage treatment plant sludge and
municipal solid waste. This practice ensures the presence of a population of methanogenic bacteria
as well as the needed nutrients needed for them to metabolize. In addition, leachate recirculation is
being practiced at a number of landfills and is currently permitted by Subtitle D regulations for
municipal solid waste landfills. Leachate recirculation is intended to make a "bioreactor" of the
landfill and to result in enhanced degradation of the waste mass by continuously recasting leachate
onto or into the waste. This practice further ensures the presence of a population of viable
microorganisms to degrade the waste mass and potentially cause excessive clogging of the filter of
the leachate collection system.
The final type of potential filter clogging is associated with chemical precipitation ft
occurs as a result of chemical processes which include the precipitation of calcium carbonate,
manganese carbonate and other insoluble forms such as sulfides, chlorides and silicates.'
Inorganic chemical precipitates can form when the pH exceeds 7. Hardness and total alkalinity of
the leachate are also important. Precipitation can be caused by the presence of oxygen, changes in
pH, changes in the partial pressure of carbon dioxide, or evaporation of residual liquid.
Biochemical precipitation can also exist. The biochemical mechanisms usually involve the
complexationof iron or manganese. The most frequently complexed metal is iron which results in
the formation of "ochre" deposits.
Both inorganic precipitate and biochemical precipitate clogging are iterative and sometimes
synergistic. When conditions are optimum, precipitate clogging can quickly decrease the
permeability of a filter. A modified list of precipitate clogging indicator parameters is given in
Table 2. 5
-------
Table 2 - Indicators of Precipitate Clogging Parameters, after Driscoll [6]
Corrosive Condition
Incrusting Condition
At pH less than 7
Dissolved oxygen in excess of 2 mg/1
Hydrogen sulfide (H2S) in excess of 1 mg/1
Total dissolved solids in excess of 1,000 mg/1,
indicating an ability to conduct electric current great
enough to cause electrolytic corrosion
Sulfate in excess of 300 mg/1
Carbon dioxide in excess of 50 mg/1
Chloride in excess of 500 mg/1
At pH greater than 7
Total iron (Fe) in excess of 2 mg/1
Total manganese (Mn) in excess of 1
mg/1 in conjunction with high pH and
the presence of oxygen
Total carbonate hardness in excess of
300 mg/1
Indeed, the combination of particulate, biological and precipitate potential for filter clogging
is of concern. For example, the scanning electron micrographs of Figures 4 (for woven
geotextiles) and 5 (for nonwoven geotextiles) illustrate the progressive reduction in void space in
the respective filter materials. However, it remains to be seen how these reductions effect (and to
what extent) the flow of leachate in landfill filters and if proper design can negate the phenomenon
to the point where acceptable performance of the filter can be achieved.
-------
(a) After 1 Month of Row
(b) After 3 Months of Flow
(c) After 6 Months of How
(d) After 12 Months of Flow
Figure 4 - Mosaic of Woven Geotextile Filter Under Very High Flow Rate Conditions for
1,3,6 and 12 Months Time Periods at 30X Magnification
10
-------
(a) After 1 Month of Flow
(b) After 3 Months of Flow
(c) After 6 Months of Flow
After 12 Months of Flow
Figure 5 - Mosaic of Nonwoven Geotextile Filter Under Very High Flow Rate Conditions for
1, 3,6 and 12 Months Time Periods at 400X Magnification
11
-------
3. Results of Previous Investigation
This project represents a continuation and expansion of a previous U.S. EPA sponsored
project [Assistance ID No. CR-814965] which was performed for the Agency by this same
research team between September 1, 1987 and August 30, 1990. So as to set the stage for the
work to follow the results from the earlier project will be briefly described. For complete details
see EPA/600/2-91/025 [7].
To investigate the behavior of several geotextile filters and a sand soil filter, six landfill
leachates were used under different experimental conditions. The characteristics of the leachates
are shown in Table 3. It was determined from particle size analyses that all of the sediment and
microorganisms contained in the six leachates fell into a relatively tight particle size distribution
wuhin the silt-size classification, i.e., they ranged from 0.074 mm to 0.002 mm.
Table 3 - Details of Municipal Solid Waste Landfill Leachates Evaluated and
Average Leachate Characteristics
Site
Designation
PA-1
NY-2
DE-3
NJ-4
MD-5
PA-6
pH
8.0
5.5
5.8
7.4
6.8
6.5
Average Leachate Characteristics
COD* (mg/1)
15,000
20,000
40,000
45,000
1,000
10,000
TS* (mg/1)
8,000
9,000
17,000
16,000
100
5,000
BODs* (mg/1)
2,000
5,000
24,000
25,000
150
2,500
,
BOD5 = biochemical oxygen demand at five days.
3-1 Phase I - Initial Flow ftate Evaluations
The first phase, which lasted for 12 months between September, 1987 and September,
1988, used flow boxes for aerobic evaluation of the filters and large containers for anaerobic
incubation of the various filters.
In the aerobic tests, 300 x 300 mm (12 x 12 in.) wooden flow boxes, 600 mm (24 in.)
high were used. The boxes were constructed using a base plate, a geonet drain, a geotextile filter,
and 150 mm (6 in.) of free draining sand. The remaining 450 mm (18 in.) of the boxes were'
empty so that falling head permeability tests could be conducted. Leachate passed through the
sand cover and geotextile filter and then flowed within the geonet, which was open at one end
only. The time for given quantities of leachate to pass through the system was measured. Each of
the six sites had at least four boxes, the only difference being the type of geotextile filter. Both
woven and nonwoven geotextiles were evaluated. They consisted of various polymer types and
manufacturing styles.
12
-------
The following findings were based on the flow rate behavior over the 12 month evaluation
period at each site.
(a) The flow rate measurements from the original values all decreased but varied
considerably.
(b) The relatively tightly woven geotextile filter, with a 4% open area, performed the
poorest. For each of the four different sites in which it was used, it clogged beyond
the detection limit. The time periods were from 4.5 to 12 months.
(c) Opening up the void space of the same type of woven geotextile to a 10% open area
helped considerably. Flow rates still decreased but were more equivalent to the needle
punched nonwoven geotextile types.
(d) The needle punched, nonwoven geotextiles performed equivalently. They were
similarly constructed but were of different polymer types. The results indicate that
polypropylene, polyester, and polyethylene fibers do not appear to give significantly
different values in their flow rate response behavior.
(e) A heat bonded, nonwoven geotextile was used at two sites. Its response was
somewhat poorer than that of the needle-punched nonwovens but better than that of the
4% open area woven geotextile.
(f) The Phase I study indicated that use of open woven geotextiles and each of the needle
punched, nonwoven geotextiles resulted in equilibrium flow conditions being
established between 6 and 12 months. The flow rate was reduced from as little as 20%
of the original values (at four sites) to as much as 80% (at two sites). These reductions
appeared to be related to the strength of the leachate insofar as their total solids (TS)
and microorganism content (BOD) were concerned. In the worst cases, flow rates
were usually greater than 0.68 ysec-m2 d.O gal/min-ft2). This is the equivalent to
58 x 107 1/ha-day (6.2 x 10? gal/acre-day), which far exceeds most design
requirements for leachate collection system filters.
(g) The cause of the flow reductions created somewhat of a dilemma. By cutting a cross
section of the boxes at the end of the 12 month period it was evident that the 150 mm (6
in.) of cover sand placed over the geotextile filter was a major source of the flow
reduction. The experiments showed that soil clogging can also be imporrtant. The soil
used was an open graded, rounded sand (Ottawa sand) having a permeability
coefficient of approximately 0.02 cm/sec (0.04 ft/min). Thus, it actually meets, and
even exceeds, most regulatory criteria for a drainage soil, let alone for a filter soil.
(h) Microscopic examination of the cross sectioned soil/geotextile systems showed heavy
paniculate clogging within the upper portion of the soil layer. Thereafter, the clogging
was either fibrous or consisted of very small clusters. Although not conclusively
proven, it was felt that the upper portion of the soil column filtered the suspended
solids out of the leachate and thereafter biological activity spread throughout the
13
-------
remaining portion of the soil column and into the underlying geotextile. This biological
activity took numerous forms including the deposition of precipitates in the soil and in
geotextile voids. Thus, different geotextiles (all other things being equal) responded
differently to a site specific leachate.
(i) The relative amounts of flow rate reduction between leachate sediment, biological
precipitates, and biological growth could not be distinguished in these particular tests.
The anaerobic tests were performed under completely submerged conditions in 210 litre
(55 gallon) drums. Twelve samples of each type of geotextile were suspended on stainless steel
racks and placed in leachate from the various sites. One sample of each type was removed for
testing each month. Four geotextile types were evaluated for each of the six landfill leachates.
After the samples were removed, they were brought to the laboratory and were tested for the
retained flow capability and possible strength reduction. The general findings follow:
(a) Relatively minor flow reduction occurred in all types of geotextiles evaluated. The
reduction values varied from 10% to 20%. Note that these amounts were distinctly less
than those that occurred in most of the previously described aerobic tests. It is felt that
sediment clogging did not form since flow was not occurring during the incubation
periods. Furthermore the absence of a soil column had a dramatic (but quantitatively
unknown) effect on improving the flow rates.
(b) All of the exhumed geotextiles had heavy biological growth that could be easily seen
and felt.
(c) Informative scanning electron micrographs taken at various times of incubation were
compared with the as-received geotextiles. After 3 months of incubation, complete
growth around the individual fibers or growth in clusters could generally be seen, recall
Figures 4 and 5. Although difficult to quantify, the amount of growth was clearly
related to the time of immersion.
(d) The micrographs also revealed that the biological growth was easily removed from the
fiber's surface. There appeared to be no fixity or attachment of the biofilm clusters to
the fibers.
(e) The above observation was corroborated by various strength tests performed on the
geotextiles after immersion. Within the statistical limits of testing there was no strength
reduction over the 12 month period. This suggests that for these leachates, biological
degradation of geotextiles is not a problem. As a result, Phase II studies did not
include the polymer degradation concern.
3-2 Phase nfa> - Improved Flow Rate Columns and Remediation Attempts
This phase of the project, which lasted 24 months, between September 1988 and August
1990 used improved flow rate measuring systems. This type of improved system has been
developed into a test method procedure and subsequently adopted by the American Society for
Testing and Materials as a Standard Test Method (ASTMD1987-91, 'Test Method for Biological
Clogging of Geotextile or Soil/Geotextile Filters"). The flow rate measuring column is a 100 mm
14
-------
(4.0 in.) diameter fixed wall permeameter which can function in either a variable or falling head
test mode. Ninety-six (96) of these test devices were used under the following set of conditions;
• four different geotextile filters,
• without sand and with sand above geotextile filters,
• aerobic and anaerobic conditions, and
• six landfill leachates.
Continuous flow rate testing on the 96 permeameters was monitored for 6 months. Once trends
were established, a series of different remediation procedures were attempted which lasted for
approximately 14 months.
The following comments apply to the first 6 months of flow testing, i.e., before the first
remediation was attempted.
(a) The columns with sand above the geotextiles clogged considerably more than those
with the geotextile alone, i.e., 23% of the flow was retained for sand/geotextile
columns versus 34% flow retained for geotextile columns. Note that if the heat bonded
nonwoven fabrics were eliminated from the geotextile group, the flow rate retained by
the geotextile group would be 45%. This suggests that geotextiles can clog less than
natural sand filters.
(b) Of the four geotextiles evaluated, the highest retained flow rate was achieved with the
lightweight needle punched nonwoven (38%), with the heavyweight needle punched
nonwoven (34%) and woven monofilament (32%) slightly behind. The nonwoven
heat bonded geotextile had the lowest retained flow of only 10% after 6 months of
evaluation.
(c) Of the various landfill leachate types, the lowest retained flow rate resulted from use of
the NJ-4 (14%) and DE-3 (17%) leachates. Recall from Table 3 that these are the
leachates with the highest TS and BOD concentrations. The other four landfill
leachates and their percentages of flow retained after 6 months of testing were PA-6
(26%), MD-5 (29%), PA-1 (38%), and NY-2 (41%).
After the initial 6 months of flow rate testing confirmed the results of the Phase I study,
several remediation procedures of the flow columns were attempted. The first remediation, a
leachate backflush, improved flow rate but to varying amounts between the different columns.
After 4 months of resumed flow testing, the flow rates decreased and allowed for a second
remediation. This remediation used a water backflush. Upon resumption of flow the flow rates
increased, but over the next 5 months they gradually decreased.
The third remediation utilized a nitrogen gas backflush. It improved flow rates, but 3
months later they were once again reduced. The fourth, and last, remediation was a forward
vacuum extraction, which only nominally improved flow rates when it was performed.
Thereafter, the flow rate again decreased. The overall average behavior of the 96 columns is
shown in Figure 6. It visually describes the decreasing flow rate trends between remediation
attempts and the rapid increase in flow rates immediately following remediation. Individual filter
15
-------
100
80
1
-
60
40
20
o
1
0 2 4 6 8 10 12 14 16 18 20 22 24
Time (Months)
Figure 6 - Average Response of 96 Flow Rate Columns from Phase H(a) Activites
16
-------
types and their respective behavior for all 96 combinations are found in Reference 7.
To quantitatively assess the overall performance of the remediation attempts and their
relative performances in contrast to one another, the data were analyzed with respect to their
percent of flow rate improvement. Within each combination, however, there were decided
differences. For example:
(a) Backflushing of geotextiles by themselves was more efficient than backflushing of
geotextile/sand systems. The average recovery efficiencies were 29% and 13%,
respectively.
(b) With sand overlying a geotextile there was no measurable difference from one type of
geotextile to another.
(c) With only a geotextile, remediation was most effective with the woven monofilament
geotextiles (38% recovery efficiency), slightly less effective with the nonwoven needle
punched lightweight (31%) and the heavyweight (30%) geotextiles, and relatively
ineffective with the nonwoven heat bonded geotextiles (16%).
(d) When sand was placed over the geotextile, there was no difference between anaerobic
and aerobic remediation schemes.
(e) With only a geotextile, remediation was slightly better under anaerobic conditions than
with aerobic conditions.
(f) When sand was placed over the geotextile, the remediation recovery efficiency rankings
were: water > nitrogen > leachate > vacuum
(g) With only a geotextile, the remediation recovery efficiency rankings were:
water > leachate > nitrogen > vacuum
3.3 Phase UCti) - Biocide Treated Geosynthetics
Because of the relatively large flow rate decreases observed during the course of this study,
an investigation into the use of biocides in the flow system was undertaken. This was done under
the assumption that the biocide would kill the microorganisms that come into contact with it and
that the nonviable (i.e., dead) matter would pass through the system in much the same way that
fine particles or sediment moves through any other filtration/drainage system. Because it was
believed that the biocide should be introduced on a long-term basis rather than as one bulk dose,
the biocide was added to the polymer compound during manufacture of the selected geonets or
geotextiles. The reasoning was that the biocide would gradually release via molecular diffusion
through the polymer structure and migrate to the surface of the ribs or fibers over a long period of
time.
From the results of these biocide treated geosynthetics, the location of the biocide vis-a.-vis
the initial formation of a biofilm layer was felt to be critical. This was confirmed at the end of the
tests after solidifying the test columns with epoxy and cutting them apart. Clearly, the biofilm
layer was occurring at the top of the sand column some 50 mm (2 in.) above the biocide-treated
geosynthetics. Although there may have been some flow rate improvement due to high
concentrations of biocide, it was very subtle (at best) and was masked by the inherent scatter in the
17
-------
test data. There was essentially no difference between flow rates in anaerobic versus aerobic
conditions.
These findings led to additional tests without sand above the biocide-treated geosynthetics
that forced the leachate to interface directly with the biocide. Rather than use a single type of
geotextile, three different types were utilized. The opening sizes varied from 0.15 mm (nonwoven
needle punched style), to 0.21 mm (a woven monofilament) to 0.42 mm (another woven
monofilament). Quite clearly, the flow rates through the largest opening size geotextiles, i.e., the
0.42 mm, were the highest. This suggests that microorganisms (dead or alive) must be able to
pass through the system. Whenever these microorganisms reside on or within the small pores of
the filter, partial, or even excessive, clogging is possible.
3-4 Summary of the Previous Investigation
A simulated field-oriented project concerning biological clogging of landfill drainage
systems was focused on geotextile filter clogging. Six different MSW landfill leachates were
used. The filter was singled out (versus the geonet drain, drainage stone or perforated pipe) since
it had the smallest openings and was likely to become clogged before other components.
Geotextiles were emphasized because they are generally used for this particular application.
Phase I results reoriented the initial goals since the granular soils covering the filters were
clogging before the underlying geotextiles. Furthermore, sediment and/or particulates were a
major factor in flow rate reductions, which appeared to be synergistic with the biological clogging.
Clearly, partial filter clogging was occurring with a gradual reduction of flow rate over time.
These trends were common to all six landfill leachates used. All of the landfills were domestic
municipal solid waste facilities; but their waste stream, volume of waste deposited, and liquid
management schemes differed. Recognized early in this Phase I activity was that remediation
attempts would be a necessary part of the overall study, but the Phase I experimental setup could
not accommodate such activities. New and different test devices would be needed if such attempts
were to be made. Some conclusions, however, were drawn from Phase I activities.
• Filter clogging (as indicated by flow rate reductions) over the 12 month test period varied
widely, the range being between 10% and the limit of the test devices.
• A geotextile filter must be relatively open in its pore structure if it is to limit the amount of
clogging, i.e., the geotextile must be capable of passing the sediment or particulates
along with the associated microorganisms into the down-gradient drainage system.
• The polymer type (polypropylene, polyester or polyethylene) comprising the geotextile
fibers appears to be a nonissue.
• Both anaerobic and aerobic conditions promote clogging; the relative amounts, however,
were not capable of being identified because of differing setups.
• The mechanical strength of the geotextiles was not adversely affected by the 12 month
exposure to the various leachates. This finding, coupled with numerous micrographs
which showed no chemical attachment of bacteria clusters to the fibers, led to the
conclusion that biological degradation of polymeric based geotextiles does not occur.
18
-------
Phase II(a) of the study saw the development of a new and vastly improved test device for
flow rate evaluation. The 100 mm (4.0 in.) diameter flow columns developed during the project
have the following capabilities.
• All types of cross sections can be evaluated: geotextiles by themselves, soil/geotextile
systems, soil/geotextile/geonet systems, or soil/geotextile/gravel systems.
• Anaerobic or aerobic conditions can be maintained.
• Flow rates can be evaluated using falling head or constant head measurements.
• The flow columns are small and portable. Therefore, they can be stored indoors and
taken to a site for evaluation, or stored at the site, or even stored within the leachate
storage tank or sump.
• Various methods for remediation of clogged systems can be evaluated.
• The test columns and their measurement protocol have been adopted as an ASTM Test
Method under the designation of D1987-91.
• The test columns and their contained materials can be solidified by epoxy and cut in half
to visually observe the conditions existing within the cross section.
• Since all parts of the columns consist of PVC plumbing and swimming pool accessories,
they are readily available, easily sealed by chemical wipes, and inexpensive.
The following conclusions were reached from this Phase n(a) study.
• Flow rate reductions were similar to the results of Phase I, and the conclusions drawn
earlier were substantiated.
• If geotextile and/or soil filters are to be used in leachate collection systems, they should
have sufficiently open voids to pass the sediment or particulates along with the
microorganisms contained in the leachate into the downstream drainage system.
• The limiting or equilibrium flow rate retained must exceed the site specific design
requirement. If flow rates over time are not adequate, remediation is necessary. It was
found that the water backfiush technique gave the best results (35% improvement),
nitrogen gas backfiush (23%), and leachate backfiush (17%) methods were next. The
vacuum extraction was the least effective; it provided only nominal improvement (2%).
• The periodicity of backflushing to open up a clogged or semi-clogged filter system
appears to be approximately 6 months.
Incorporating biocides into the geotextile (or geonet) polymer structure to keep the flow
system open was Phase H(b) of the study. The concept was to add various amounts of a time-
released biocide into the polymer compound as the product was manufactured—biocide that would
essentially diffuse to the surface of the fibers during its service life. On contact, the biocides
would kill the viable microorganisms in the leachate. In the tests that were conducted on 16
separately built flow rate measuring devices, some experimental evidence indicated that 2% and
4% biocide was partially effective. The remains of of the dead bacteria must, however, be
permitted to pass through the system, and this apparently could not happen for these particular test
setups. Thus, the idea of a very open filter system was further substantiated. While not
19
-------
successful within the context of this setup, the suggested use of biocides continues. Thus
Appendix "A" of this report presents the results of the biocide experiments of this precursor study.
3-5 Recommendations of the Previous Investigation
Based on the major findings of this project, namely,
• under continuous flow of landfill leachate, a gradually decreasing flow rate occurred for
all types of filters (soil or geotextile) and eventually reached an equilibrium value,
• the equilibrium value of flow rate varied according to the type of filter, the type of
leachate, and the hydraulic gradient, and
• the equilibrium flow rate for any given filter system must be compared with the design
required flow rate to ultimately assess the adequacy of the filter's design,
it was felt that the following recommendations be considered regarding geotextile and soil filters
placed over different types of leachate collection drains.
(a) Leachate collection systems at landfills that are decommissioned or exhumed for other
reasons should be investigated in light of the results of this study.
(b) Design criteria should be developed that consider the amount and type of
microorganisms and sediment present in the leachate along with conventional issues
such as hydraulic gradient and type of filter.
(c) This particular project should be followed by another effort aimed at a larger variety of
geotextile filters along with design guidance to predict the field performance of those
existing sites which are exhumed.
20
-------
4. Overview of This Project
Based upon the recommendations of the project just described, a second U.S. EPA funded
study entitled "Leachate Clogging Assessment of Geotextile and Soil Landfill Filters," was
conducted. Its designation was CR-819371 and the project period extended from March 1, 1992
through February 28, 1995. The remainder of this report focuses on this second prnjarf' The
project consisted of field, laboratory, computer modeling and design oriented tasks. Figure 7 is a
flow chart of the project and the interrelated tasks. The major tasks of the project were the
following:
« Field exhuming of sites-pf-opportunity which came about in the 18-month time frame
between the two projects and extended into the beginning of this second project.
• Long-term laboratory tests of a wide range of geotextile filters and two sand filters using
ASTM D1987 permeameters of the type developed in the earlier study.
• Utilization of the U.S. EPA sponsored HELP computer model for design input
parameters for the field exhumed sites.
• Formulation of a new methodology for the design of filters used with landfill leachate
collection systems.
• Posing a challenge to the newly developed design formulation with the results of the
field exhumed sites-of-opportunity where "ground truth" had been established.
• Forming final summaries and conclusions for both of the projects described in this
report, i.e, CR-814965 and CR-819371.
21
-------
Very Fast
Flow Rates
1.
Overview/Concerns
based on
Previous EPA Project
i
Field Exhuming
S ites-of-Opportunity
Laboratory Study
using ASTMD1987
Permeameters
±
Fast
Flow Rates
Determination of
^ allow
for Site Specific
Flow Rates
Determination of
kreqd
using HELP Model
Formation of the
Drainage Correction Factor
(using drain geometry)
Development of a
Design Methodology
Intermediate
Flow Rates
J
Substantiation
of
Design Method
Summary and Conclusions
Figure 7 - Flow Chart for This Project
22
-------
5. Field Exhuming of Leachate Collection Systems
This section describes the field exhuming of actual solid waste landfill leachate collection
systems with focus on the geotextile filter and its performance [8]. All sites were obtained as
"sites-of-opportunity," whereby permission was granted by the owner/operator of the facility. By
request of the owner/operators, the description of the sites is not identified as to the location and
other details unimportant to this study.
5.1 Field Exhuming Details
Four (4) landfill leachate collection systems were exhumed during the course of this study.
The leachate collection systems had been in service for up to 11 years at the time of their
exhuming. The design of each system, the reason for exhuming, the findings and performance
levels were quite different for each site. Table 4 counterpoints these differences. Each site is
detailed in regard to specific conditions in the following sections. It is relevant to note that all four
sites were in the northeast region of the USA, thus temperature, precipitation, and general climatic
conditions were reasonably similar to one another.
Table 4 - Overview of the Leachate Collection Sites Exhumed
Site
1
2
3
4
Waste Type
domestic and
light industrial
domestic and
light industrial
industrial solids
and sludge
domestic and
rural
Construction
1981
1985
1990
1976
Liquid
Management
Scheme
leachate
recycling
leachate
recycling
leachate
withdrawal
leachate
recycling
Reason for
Exhuming
no flow in the
collection system
leachate seeps through
the landfill cover
no flow in the
collection system
no flow of methane
into extraction wells
Critical
Element in
Drainage System
geotextile filter
drain location.
geotextile filter
geotextile filter
5.2 Field Exhumed Site #1
The solid waste in this landfill was a mixture of domestic and light industrial wastes. The
facility was constructed in 1981. The plan and cross-sectional views are shown in Figures 8 and
9, respectively. Within one year after the cell was complete, leachate flow rates began to
significantly decrease. Since the liquid management scheme was a form of leachate recycling, the
situation was of concern with respect to the lack of leachate for reinjection and the high leachate
heads being built up and imposed on the liner. Furthermore, leachate seeps began appearing
through the landfill cover (which was a compacted clay liner) which resulted in leachate running
down the side slopes of the landfill.
The original design called for a drain along the low elevation side of the facility intended to
intercept leachate flowing down gradient on the geomembrane as seen in Figure 9. The facility
was constructed accordingly. Once in the drainage system, the leachate flowed to a sump for
23
-------
(a) Aerial Photographs of Site #1 Showing Exhumed Area of Cell "A".
Cell "A" is 11 ha (27 acres) in Area and the Exhumed Area was
Rectangular in Shape Measuring 30 m x 91 m (100 ft x 300 ft).
| Geotextile Socked 100 mm (4 in.) PVC Pipe
y Gravel Trench f
1.2 m (4.0 ft)
Concrete Sump
(b) Plan View of Site #1 Leachate Collection System. The Leachate Collection
System was Embedded in a Sand Layer and Consisted of a Geotextile
Socked 100 mm (4.0 in.) Perforated PVC Pipe Within a Gravel Trench.
Figure 8 - Plan Views of Site #1
24
-------
Geotextile
Socked Pipe Drain
of Site #1
/ /
/ / / /
v>*
XV* A
X A *
A A A
A^£ A
*/>.
A 3KA
A A A
-
Buffer
Sand
-
(a) Cross-Sectional View of Cell "A" of Site #1 Showing the Location
of the Leachate Collection System with Respect to the Other
Components of the Cell.
Geomembrane {
100 mm (4.0 in.)
Perforated PVC Pipe
Geotextile Sock {
(b) Cross-Section Detail of the Leachate Collection System of Site #1
Showing the Arrangement of the Geotextile Filter Socking the
Perforated Pipe.
Figure 9 - Cross-Sectional Views of Site #1
25
-------
collection and removal. The drainage system included a 100 mm (4 in.) diameter SDR 41
perforated PVC drainage pipe wrapped (or "socked") with a nonwoven heat bonded geotextile.
This socked pipe was embedded in 600 mm (24 in.) of crushed stone. The drainage system fed
into a sump within a concrete standpipe which was located at the lowest elevation of the landfill
cell.
Due to the nature and magnitude of the clogging of this drainage system, a large section of
the landfill was exhumed. A total area of 19,000 m3 (25,000 yd3) of solid waste was removed to
uncover the leachate collection system. Upon exhuming the collection system, numerous
problems were found to be contributing to the malfunctioning of the leachate collection system.
The most obvious shortcoming of the system was the geotextile encapsulating the perforated pipe.
The geotextile was excessively clogged with particulates as well as biomass. The fouled geotextile
was no longer allowing leachate to flow into the perforated pipe. Permittivity tests on the
geotextile confirmed these findings as the original value of 1.1 sec-1 decreased to 8.2 x ICh4
sec-l. (Permittivity of a geotextile is defined as the permeability divided by its thickness and is a
commonly used geotextile value used in permeability and flow rate calculations).
A second problem was the 100 mm (4 in.) PVC drainage pipe used in this installation. It
was crushed in several locations and noticeably damaged in others. This SDR 41 pipe was
significantly underdesigned on the basis of strength. The damage most probably occurred at the
time of installation of the pipe due to stresses imposed by construction equipment. It was noted
that the pipe, although crushed, was still able to transmit leachate. This was apparent from
leachate removal records and the fact that leachate flowed into the perforated pipe when the
excessively clogged geotextile was stripped away from its outer surface.
The shale aggregate which surrounded the socked pipe was a third limiting factor. This
collection stone was' classified as a poorly graded gravel (GP) as defined by the Unified Soil
Classification system. Its gradation corresponded to that of an AASHTO#57 stone which was
from 6 to 30 mm (0.25 to 1.25 in.) particle size. Upon exhuming, this gravel was filled with fines
and biomass to the point that it had agglomerated together forming "biorock" as shown in Figure
10(b). In its current state the gravel permeability had been reduced from 25 cm/sec to 1.2 x 10-2
cm/sec. The presence of fines and biomass were the result of years of filtering leachate.
The fourth component of this system which deserved attention was the protection layer of
soil in which the leachate collection drain was embedded. The primary purpose of the protection
soil was to insure that the waste did not come into direct contact with the geomembrane. The
secondary purpose of the protection soil was to provide a granular media for the leachate to be
transmitted down gradient to the drainage trench. The protection soil was a well graded sand
(SW) as classified by the Unified Soil Classification system. Its gradation was that of AASHTO
#10. Upon exhuming this 450 mm (18 in.) layer of protection soil layer, a heavily contaminated
lense of fines and biomass was observed. In addition there were several regions of the protection
soil that were heavily marbled with black and orange ochre staining. This staining was due to
26
-------
(a) Photograph of the Leachate Collection
System of Site #1. Note that Even Though
the Pipe was Crushed it Still could
Convey Leachate.
(b) Photograph of Several Components
of the Drain of Site #1. Note that
the Gravel had Agglomerated Into
"Biorock" and That the Geotextile
was Heavily Strained and Clogged
with Sediment and Biomass.
(c) Photograph of the Geotextile Socking and Perforated Pipe of Site #1.
Note that the Geotextile was Bound Around the Pipe by String Ties
Placed at 1 m (3 ft.) Intervals. This Technique Allowed for Fines to
Infiltrate Through the Flap Edge. In Addition Note that the Intensity
of "Biofouling" was Directly Above the Perforations hi the Pipe.
Figure 10 - Photographs of Site #1
27
-------
metals (particularly iron) and organics in the leachate. It was apparent from observing the standing
sidewall of solid waste that the majority of the leachate was draining on top of the protection soil
layer and not within it. This indicated that the protection sand was no longer allowing leachate to
enter and be drained. It should be pointed out that the protection sand may never have exhibited
the required transmissivity for this application and is the reason why most current facilities require
drainage gravel in this application rather than sand, or no protection layer whatsoever. The
drainage soil must be capable of limiting the head on the liner to a maximum of 300 mm ( 12 in.)
under current federal regulations and some state regulations and private facilities require 0. 1 to 1 .0
cm/sec as the minimum permeability.
A sample of the leachate was taken from this site and analyzed with the following results-
•COD = 3 1,000 mg/1
• TS = 28,000 mg/1
•BOD5= 27,000 mg/1
In comparison to the leachates described in Table 3, this was a relatively harsh MSW leachate with
high organic content and relatively high solids content. The nature of the leachate was due in part
to the practice of leachate recirculation. Recirculation of leachate was intended to induce
degradation of the waste mass. It must be recognized, that the practice of recirculating leachate can
contribute to the clogging of geotextile and natural soil filter layers at this, and any other site where
the technique is utilized.
Upon making observations in the field, all of the materials were sampled and brought to the
laboratory for testing and evaluation, see Table 5. Sample S 1 was of the sand layer underlying the
waste and around the drainage gravel of the leachate collection system. These protection soil
samples were evaluated for their permeability in constant head permeability tests, per ASTM
D2434. The sand was evaluated as close as possible to its in-situ conditions, which was labeled
the SI (a) value. The soil was then thoroughly washed to obtain the original, as-placed value.
This material was called S 1 (b). As seen in Table 5(a), the resulting differences were remarkable.
During the 10 year period of operation the permeability of the sand decreased from 4.3 x 10-2
cm/sec to 1.6 x 10~5 cm/sec, i.e., a reduction of over three orders of magnitude.
Table 5 - Results of Laboratory Tests on Materials of Sites #1
(a) Permeability Results of Natural Soil Materials
Number
SI (a)
Sl(b)
S2(a)
S2(b)
Description
Field Sample of Biofouled Sand
Cleaned Sand
Field Sample of Biofouled Angular Gravel
Cleaned Angular Gravel
Permeability, k (cm/sec)
1.6 x 10-5
4.3 x lO-2
1.2x10-2
25
28
-------
Table 5 - (continued)
(b) Permittivity Results of Geotextiles
(The thickness of the heat bonded nonwoven geotextile was 0.038 cm)
Number
GTl(a)
GTl(b)
Description
Nonwoven Heat Bonded Geotextile
Cleaned Nonwoven Heat Bonded Geotextile
Permittivity
(sec-1)
8.2 x 10-5
1.1
Permeability
(cm/sec)
1.8 x 10-4
4.2 x 10~2
Sample S2 consisted of the angular gravel taken from the leachate collection system. This
gravel was located between the protection sand and the geotextile around the perforated collection
pipe. The sample was also evaluated for its permeability per ASTM D2434. As with the previous
sample the soil was thoroughly washed to obtain the original (as placed) value. Seen in Table 5(a)
is that the average permeability coefficient decreased from 25 cm/sec to 1.2 x 10-25 which is again
over three orders of magnitude decrease.
Gradation curves were generated for all soils in the test program. There was very little
difference between the washed samples and the exhumed field samples. This indicated that the
mass of the biomass was insignificant with respect to the overall sample's mass; however, its
impact on permeability was quite another matter.
In addition to testing the sand and gravel, the geotextile of the leachate collection system
was tested for its cross plane flow characteristics, see Table 5(b). Sample GT1 was taken from
the geotextile which was socked around the perforated pipe in the leachate collection system. The
test utilized was the geotextile permittivity test which, as noted previously, is defined as the
permeability divided by the geotextile's thickness, in units of sec-1. The ASTM designation for
the permittivity test is D4491. This test was performed on the exhumed geotextile, labeled
GTl(a), and then on the washed sampled, labeled GTl(b). Table 5(b) indicates thatthere was a
three order of magnitude decrease in permittivity between the exhumed and cleaned samples.
5.3 Field Exhumed Site #2
Site #1 and Site #2 were at the same landfill and were exhumed at the same time. They are
differentiated because they were installed five years apart, they were constructed of different
materials, they were configured very differently, they were in different locations and ultimately
they were installed for different purposes.
As a relief system to the excessively clogged drain in Site #1 just described, an auxiliary
underdrain was installed as shown in plan and cross sectional views in Figures 11 and 12,
respectively. Note that the drain was placed as an auxiliary retrofit near the top of the
geomembrane lined slope where it intersects the cover soil. The drain was installed to intercept
leachate seeps on the sidewall of the landfill. Apparently the drain remediated the problem of side
wall seeps but left a considerable depth of leachate in the facility, approximately 4.6 m (15 ft.).
29
-------
(a) Aerial Photograph of Site #2 Showing Exhumed Area of Cell "A"
Note that There were Two Drains at This One Location. The Drain
- £'%s%>
1.2 m (4.0 ft)
Concrete Sump
(b) Plan View of Site #3. The Drain Consisted of a Geotextile Wrapped
Trench with a Perforated Pipe Filled with Rounded Gravel. The Drain
°f ^ Landfm at ^ ^t6186011011 of the
Figure 1 1 - Aerial and Plan of Site #2
30
-------
^v-;-:;::;;|!||
^v
A A A
A A A
A A A
A A A
•p>:
A T^A
*****
X
~2* " • —
Trench wrapped
rlrain nf Qito &9
Uiaill \Jl Ollty TTt
(a) Cross-Section View of Site #2 Showing the Location of the Drain
Near the Anchor Trench. This Drain was a Retrofit to the Original
Design and Intended to Remediate the Problem of Side Wall Seeps
Geotextile
Trench
Wrap
A A
k A
A A
k A
A A
A
Gravel \<<
I JL * A JLJUL* 'K'.
100 mm (4 in.)
Perforated
HOPE Pipe
(b) Close-Up Cross-Section of the Drain of Site #2 Showing the
Arrangement of the Geotextile Lining the Trench. The Drain
had a Rectangular Cross-Section with Dimensions of
600 x 1200 mm (24 x 48 in.).
Figure 12- Aerial and Plan of Site #2
31
-------
This head was directly imposed on the liner at the base of the landfill.
Upon exhuming Site #2, see Figure 13, a rectangular drainage trench measuring
approximately 600 mm (24 in.) wide and 1200 mm (48 in.) high was located. The drainage trench
included a 100 mm (4 in.) diameter SDR 30 perforated HDPE drainage pipe. The geotextile
wrapped trench drain was filled with rounded quartz gravel of particle sizes ranging from 6 to 18
mm (0.25 to 0.74 in.). The geotextile used at this site was a 7 percent open area woven
monofilament geotextile. The drain fed into a sump where the leachate was periodically removed.
The drain was seen to be functioning reasonably well. As can be seen in Table 6(a), the
permeability (per ASTM D2434) of the in-situ rounded gravel is 28 cm/sec which was slightly less
than its cleaned (assumed to be original) value. In addition, the woven monofilament geotextile
experienced only a nominal decrease in permittivity. As can be seen in Table 6(b) the geotextile
decreased from its as-received permittivity (as per ASTM D4491) of 0.9 sec-1 to an in situ
permittivity of 0.33 sec-1, which was considered to be only a nominal decrease.
A sample of the leachate from the site was taken and analyzed with the following results:
•pH = 7.5
• COD = 10,000 mg/1
• TS = 3,000 mg/1
•BOD5 = 7,500 mg/1
This is an average municipal solid waste landfill leachate (recall Table 3) with organic content in
the high thousands and solids content in the low thousands.
The shortcoming of the drain was its location with respect to the elevation of the liner of the
landfill. The invert of the perforated pipe In this drain was 4.6 m (15 ft.) above the liner. This
means that the drain only intercepted leachate from the upper two thirds of the landfill and could
not alleviate the problem of leachate head on the liner. However, the drain did eliminate the side
wall seeps.
Table 6 - Results of Laboratory Tests on Materials of Sites #2
(a) Permeability Results of Natural Materials
Number
S3(a)
S3(b)
Description
Field Sample of Biofouled Rounded Gravel
Cleaned Rounded Gravel
Permeability, k (cm/sec)
28
53
(b) - Permittivity Results of Geotextiles
(The thickness of the woven monofilament geotextile was 0.041 cm)
Number
GT2(a)
GT2(b)
Description
Woven Monofilament Geotextile
Cleaned Woven Monofilament Geotextile
Permittivity
(sec-1)
0.33
0.9
Permeability
(cm/sec)
1.35 x lO-2
3.69 x 10~2
32
-------
B«rnHr AwfJ.The Lafge Majority of the Waste was Exhumed Via a
Backhoe and Disposed of in an Adjacent Active Facility. The Entire Volume of the Waste
Exhumed Amounted to Approximately 19,000 m3 (25,000 yd3)
** Drain- ^ Drain Difficult to exhume due
Gravel and the Saturated Condition of the Waste
no-w - °f Site#2' Note ^ me Geotextile was Stained
not Clogged. This Woven Geotextile was still Functioning after 3 years of service.
Figure 13 - Photographs of Site #2
33
-------
5.4 Field Exhumed Site #3
Site #3 was a relatively small industrial landfill 240 by 120 m (800 by 400 ft.) in size with
9.1 m (30 ft.) of waste depth. A plan view of the site is shown in Figure 14(a) and an isometric
diagram in Figure 14(b). This facility was a double lined landfill with a leak detection system
between the primary and secondary liners. Only the primary geomembrane and its overlying
leachate collection system is shown in Figure 14(b) since it was the focus of our investigation.
This industrial waste facility co-disposed industrial plant refuse, slurried fines and lime
stabilized waste. It will be shown that the slurried fines were the problem materials in this
situation. The slurried fines were seventy percent less than a #100 sieve (0.15 mm) and forty-five
percent less than the #200 sieve (0.074 mm). Both of these values were less than the apparent
opening size of the geotextile used at this site which was equal to #80 sieve (0.19 mm).
From visual observations, it was determined that the leachate collection system was not
functioning as intended, for example:
• The leachate in the drainage gravel was under pressure as measured by a piezometer in
the area of the sump. This indicated that the leachate entered the drainage gravel at high
elevations but was unable to enter the pipe network at lower elevations.
• There were several areas of ponded rain water on the surface of the solid waste. With
tune, the ponds grew and required surface pumping.
• The amount of leachate collected in the sump was much less than that anticipated by the
operators on the basis of other similar cells at the facility.
Of the three observations listed above, the first one focused attention on the problem
element of the design. By observing the leachate stained pea gravel of the leachate collection
system it was assumed that the leachate entered the stone drainage system through the upper
geotextile. The geotextile wrapped around the 100 mm (4.0 in.) perforated pipe however, was
excessively clogged. This put the leachate in the drainage stone under hydraulic pressure causing
the liquid to build pressure in the area of the sump due to the elevation head at this point. The
photographs of Figure 15 show some of the conditions that existed at the site and the condition of
the geotextile that socked the perforated pipe. As with Site #1 described previously, the primary
cause of the malfunctioning leachate collection system was the geotextile filter wrapped directly
around the perforated pipe. This type of "socked pipe" design for leachate collection systems
appears to be a flawed design particularly when the pipe is of the smooth (i.e., non-profiled) type.
A series of three permittivity tests per ASTM D4491 were conducted on the geotextiles at
this site. As seen in Table 7 both the upper geotextile and the geotextile socking the perforated
pipe were evaluated. A sample of geotextile was also washed clean of particulates and then
retested. This procedure gave a basis for comparison by approximating the as-received
permittivity of the geotextile. Due to the nonuniformity of the clogging and the small size of the
sample taken in the field, it was decided to condition a sample of geotextile in the laboratory to
simulate field conditions before testing. This geotextile was conditioned with site specific slurried
fine waste for 6 months prior to testing for permittivity.
34
-------
Fine Grained
Slurried Waste
•120 m
(400 tt)
Figure 14(a) - Plan View of Site #3 Showing the Dimensions of the Cell as Well as the
Exhumed Area with Respect to the Different Waste Disposal Locations.
j j I I 4- i Leachate
Ind. Waste Slurry
Geotextile
Pea Gravel
100 mm HOPE Pipe
Geotextile
rr 2.0 mm HOPE GM
Secondary System
Figure 14(b) - Isometric View of Site #3 Showing the Geotextile Filter Separating
the Slurried Waste from the Pea Gravel and a Second Geotextile
Filter Beneath the Pea Gravel Which also Socked the Perforated Pipe.
35
-------
Industrial Process.
A Large MaJ°rity of the Waste Consisted of Slurried Fines from
of Site #3> A 6 m by 6 m (2°ft by 20 ft) Pit
(C) Kh^ai?h ?K ^ G!ptextile Torn Away from the Perforated Pipe. It was observed that
Leachate in the Pea Gravel Drained into the Sump After the Geotextile Sock w^Removed.
Figure 15 - Photographs of Site #3
36
-------
Table 7 - Permittivity Test Results for Site 3 (ASTM D-4491)
(The thickness of the needled punched nonwoven geotextile was 0."274 cm)
Number
1
2
3
Description
Cleaned geotextile
Upper geotextile
Geotextile sock*
Permittivity
(sec-i)
1.82
3.14x 10-2
1.57 x 10~5
Permeability
(cm/sec)
0.5
8.6 x 10-3
4.3 x 10~6
i o—~w~**»^ wwiiutuwuvu wiLu uic Hue Murncci waste for six
months and then tested.
The results of Table 7 show a two order of magnitude decrease in permeability of the upper
geotextile over the as-received geotextile. Also, there was a five order of magnitude decrease in
permeability of the geotextile used in socking the perforated pipe over the as-received geotextile
As will be discussed later, these test results show that the upper geotextile is much less clogged
than the geotextile socking the perforated pipe.
In addition, grain size analyses were conducted on the waste and granular drainage media
of the site. These results are shown in Figure 16.
Finally, an apparent opening size test was conducted on the 330 g/m2 (10 oz/yd2) needle
punched nonwoven geotextile. A value of #80 sieve size (0.19 mm) was obtained.
In addition to characterizing the waste as far as its particle size, a sample of the landfill
leachate taken from Site #3 was analyzed and resulted in the following values-
•pH = 9.9
• COD = 3,000 mg/1
• TS = 12,000 mg/1
*BOD5 = 1,000 mg/1
As expected, this industrial waste did not have much organic matter which was indicated, by
relatively low readings of COD and BOD5, recall Table 3. In contrast, the value of 12,000 rng/1
of total solids (TS) was quite high. The clogging at this site was felt to be due primarily to the
paniculate clogging by the sludge rather than biofouling from microorganisms.
As can be seen from the site photograph of Figure 15(a), the slurried turbid waste was at a
very high moisture content (greater than 40%). Figure 15(b) shows the actual exhuming process.
Approximately 9.1 m (30 ft) from the sump, a pit was dug through the waste, protection soil
layer, upper geotextile, pea gravel and finally reached the geotextile socking the drainage pipe.
Upon going through the different layers, the geotextile socking the pipe was identified as the layer
inhibiting flow of leachate into the drainage pipe. Once the geotextile sock was removed from
around the drainage pipe the system flowed freely.
The geotextile above the pea gravel and below the protection sand was the same type of
geotextile that was socking the drainage pipe. It was important to observe that the upper geotextile
37
-------
Gravtl
-
i
Sand
Coars*
2 ?
§ I
to Pm€
Silt
U S standard $i*v« sues
3 4 1 8
l- S 2 f
Clay
'I 5
e o
Gram diameter, mm
Q — <«
?o 5
o a
!!
Figure 16 - Grain Size Distribution of Soils and Sludge at Site #3
38
-------
appeared to be functioning while the geotextile socking the drainage pipe was excessively clogged.
This was known because the pea gravel was relatively clean and full of leachate. The reason was
felt to be due to the different demands on the two filters. The upper geotextile was still functioning
while the geotextile surrounding the drainage pipe was not functioning because all of the landfill
leachate had to flow through the perforations in the drainage pipe. At each of the 12 mm (0.5 in.)
diameter perforations in the pipe, the 500 mm2 (0.78 in.2) of geotextile which filtered the leachate
had clogged with particulates and the leachate was subsequently backing up in the cell.
Conversely, the upper geotextile covered the entire base, or footprint, of the cell. Its available area
for flow was enormous compared to the geotextile socking the holes in the drainage pipe. Hence,
the available area for flow through the upper geotextile was far in excess over that of the geotextile
socking the perforated holes in the drainage pipe.
5.5 Field Exhumed Site #4
The solid waste in this landfill was a mixture of domestic, light industrial, and sewage
treatment plant sludge. The facility began to receive waste in 1976. The landfill utilized the
concept of leachate recirculation via injection wells. In so doing leachate was reintroduced into the
waste mass through a series of wells. This particular liquid management scheme was intended to
decompose the solid waste into harmless constituent end products over the operating life of the
facility.
This field site was different from those previously exhumed in that a leachate collection
system at the base of a landfill was not being exhumed. In this case the investigation was on the
geotextile filter around well casings from a methane gas extraction system. The methane produced
by this solid waste landfill was being collected for use as fuel for electric power generation. The
methane was extracted by 23 m (75 ft) deep wells at the top of the landfill which were on a 30 m
( 100 ft) rectangular spacing. The wells were made from 100 mm (4.0 in.) perforated schedule 40
PVC pipe with a needle punched nonwoven geotextile surrounding them acting as filters. The
geotextiles were attached to the well casing via duct tape spaced every 900 mm (36 in.). A plan
view of the landfill is shown in Figure 17(a) and a typical extraction well is shown in isometric
form in Figure 17(b). The gas from these wells was collected, cleaned, condensed and it was then
used as fuel to run a turbine which powered a generator and produced electricity.
The gas extracted from the landfill at a vacuum of 7 to20kPa(l to3 Ib/sq. in.) of mercury
was essentially saturated. This is one of the drawbacks of performing both leachate recirculation
and gas extraction in close proximity to one another. The gas wells at such sites are inevitably
going to attract leachate. The leachate entering the wells was very strong (compared to the sites
listed in to Table 3) and had the following characteristics:
• COD = 24,000 mg/1
• TS = 9,000 mg/1
•BOD5 = 1 1,000 mg/1
39
-------
300 m
(1,000 ft)
Figure 17(a) - Plan View of Site #4 Showing the Configuration and Dimensions of
the Cells. The Exhumed Wells Were in the Area of the Oldest Cells
and Very Close to the Methane Condensing and Cleaning Plant.
Well Head
"Tit
k'^* *
Cover Soil
Geotextile
Geomembrane
Compacted Clay
Geotextile Socked
100 mm (4.0 in.) PVC Pipe
Figure 17(b) - Isometric View of Site #4 Showing the Geotextile Filter Socking the
Perforated Pipe Which was Used as Extraction Wells.
40
-------
Over the last year of service, approximately 30% of the extraction wells stopped producing
recoverable amounts of methane. The operator tried to increase the vacuum on the wells;
however, this resulted in drawing air into the well head. The result was a decrease in the methane
content of the extracted gas which made it useless as fuel.
The malfunctioning wells were extracted by a small crane as shown in Figure 18. All-
thread was drilled through the pipe and slings were attached to it. The well casing was pulled by
extending the boom and allowing the hoist to compensate for the inclination of the crane.
Field samples of the geotextile filter were retrieved and brought back to the laboratory for
testing. The samples of geotextile were retrieved from 3.0, 7.5 and 15 m (10, 25 and 50 ft)
depths into the well. The samples at the two lower elevations were black with organic material and
caked with fine sediment. There were particularly large deposits in those areas where the
geotextile covered the pipe perforations. The sample retrieved from the upper elevation was fouled
like the lower two samples, although not to such a great extent. Table 8 shows the results of
permittivity tests conducted on the three exhumed geotextile samples as well as a geotextile sample
which was washed clean. The cleaned geotextile was tested to give an estimate of the as-received
permittivity of the needle punched nonwoven geotextile. The average thickness of the four
geotextiles tested was 0.22 cm. This value was used in the conversion of permittivity to
permeability.
Table 8 - Permittivity Test Results of the Geotextiles of Site #4
(The thickness of the nonwoven needle punched geotextile was 0.22 cm)
Description
Washed Geotextile
Geotextile from 3.0 m (10 ft) depth
Geotextile from 7.5 m (25 ft) depth
Geotextile from 15 m (50 ft) depth
Permittivity,
(sec-i)
1.1
1.7 x 10-2
7.3 x 10~5
3.4 x 10-4
Permeability,
(cm/sec)
2.3 x 10-1
3.8 x lO-3
1.6 x 10-5
7.5 x 10-5
Note that there was a decrease in permeability by as much as four orders of magnitude.
This decrease in permeability substantiated the lack of methane recoverability from the geotextile
socked gas extraction wells. As proof that the landfill was still producing recoverable amounts of
methane, new wells of a different design were placed in the same location as the old ones and
significant amounts of methane emerged. It also should be mentioned that the permittivity of the
geotextile at the 3.0 m (10 ft) depth was not as low as the geotextile permittivities at the
7.5 m (25 ft) and 15 m (50 ft) depths because the geotextile did not encounter as much leachate at
the upper elevations in the landfill.
Due to the poor performance of the extraction wells the owner of the facility changed the
well design over time. The casing was changed from drilled to slotted well screens which
increased the surface area of the pipe penetrations. The geotextile sock was eliminated from the
41
-------
(a) Photograph of the Top of the Oldest Landfill Cell. The Photograph shows the Head of a
Typical Extraction Well and the Methane Condensing and Cleaning Plant.
™ 11 o i j IT o • ,°f Site M- A Crane was used to remove the Wells so that Additional
Wells Could be Set in the Same Locations Utilizing the Existing Manifold System.
(c) The Geotextile Tom Away from the Perforated Pipe. Note the Helvy Layer of Fines and
Biomass on the Geotextile Filter which was Limiting How of Methane into the Pipe.
Figure 18 - Photographs of Site #4
42
-------
design completely. The elimination of the filter was a result of excessive clogging of the
geotextile. Last, but not least, the diameter of the augured hole was increased from 200 mm (8
in.) to 300 mm (12 in.). This increase in diameter allowed for an annulus of gravel to surround
the perforated pipe and also insured that the gravel extended for the entire depth of the well.
5.6 Summary of Field Exhuming Study
The exhuming of leachate collection system filters was critically important to the overall
thrust of the project. While sparse in number, the four sites provide ground truth and a target for
the design method development to follow. At each of the four sites we acted as technical
observers and performed the following tasks:
• Observed and recorded (via reports, photographs and video) the exhuming processes.
• Sampled the various components (leachate, geotextiles and soils).
• Repaired geotextiles and soil components of the leachate collection system.
• Conducted required tests on the various samples (natural soils and geotextiles) taken
from the field.
« Cleaned and tested the various materials as close as possible to their as-received
condition and compared the results to the in-situ values.
In addition to these tasks we hand excavated the final thickness of solid waste and buffer soil
overlying the leachate collection filter so as not to cause removal damage. Bulk excavation of the
solid waste with construction equipment was performed by the facility operator or by an outside
contractor.
The experience of these four sites allows us to conclude that geotextile socking of
perforated pipe for the application of leachate collection filters of solid waste landfills simply does
not work. As seen in the different location sketches of Figure 19, the positioning of a geotextile as
a pipe sock, trench wrap or aerial filter is significant with respect to its long term performance.
The ratio of the leachate to be filtered to the available open area of the downstream drain is a critical
design detail. It should not be a surprise that an aerial filter will function in an application where
the same type of filter material placed around a perforated pipe will fail. All of the excessively
clogged sites that were exhumed were socked pipe filters. It should also be noted that all of the
pipes exhumed were of the solid wall type. It is possible that profiled or corrugated pipe would
have performed better, but this is not known to be the case.
In addition to the configuration of the filter with respect to the drain, a major consideration
hi the performance of the drain is the type of leachate being filtered. If the leachate had high total
suspended solids and high organic content, a filter criteria which is based on water as the permeant
will probably not be suitable. See Table 9 for a comparison of leachates from these four exhumed
case histories. In cases where the leachate contained high amounts of organics and solids, the
available filtration area of the filter and probably the opening size of the geotextile filter must be
increased to encourage transmission of the suspended matter through the filter rather than causing
excessive clogging. Recognize that in so doing, the suspended materials will move into the
drainage pipe network. Thus, the drainage pipe network may have to be set at a high gradient
43
-------
0
O°!
o
geomembrane
(a) Socked Pipe
L°
01
geomembrane
(b) Trench Wrap
O-
o 0
o
geotextile
~/~^
oQ
o ° °o 9-\ ° o ° °
o OOO o OVJOo OOO
^ i nn o
-------
and/or be available for periodically flushing and cleaning to insure its effectiveness.
Table 9 - Summary of the Leachate Characteristics of the Exhumed Field Sites
Site
No.
2
3
4
Landfill
Type
Municipal
Municipal
Industrial
Municipal
PH
6.9
7.5
9.9
6.1
COD
(mg/1)
31,000
10,000
3,000
24,000
TS.
(mg/1)
28,000
3,000
12,000
9,000
BOD5
(mg/1)
27,000
7,500
1,000
11,000
Additional summarizing details for the four exhumed sites are given in Table 10. It is clear
from these particular sites that geotextile filters of leachate collection systems can become
excessively clogged with fine particles and/or microorganisms. Three of the four sites had
excessively clogged geotextile filters. All three of the excessively clogged geotextile filters were
designed as filters socking perforated solid wall pipe. It is obvious from these findings that the
practice of socking perforated pipe for the application of leachate collection filters in solid waste
landfills is not good landfill practice and is fraught with problems.
The task of designing a filter at the base of a landfill facility for long-term transmission of
leachate is difficult. Leachate, while varying tremendously in its characteristics, is invariably laden
with suspended solids, dissolved solids and microorganisms as was seen in Table 9. The filter
being the component with the smallest opening sizes is the first to encounter this leachate The'
choices are to either pass the sediment and microorganisms or collect them entirely or partially.
The first choice risks downstream drain clogging, the second excessive filter clogging For
geotextile filters this decision must be made relatively quick in comparison to a soil filters due to
their thinness. As described earlier, the possible preferred mechanism for geotextile filtration is by
depth filtration [5], but many geotextiles used are quite thin. One way of assessing the potential
behavior of such filters is the ASTM D1987-91 test method. This test is the subject of the
following laboratory portion of this project.
Table 10 - Overview of Exhumed Leachate Collection Systems
Site
No.
1
2
3
4
Waste
Type
domestic and light
industrial
domestic and light
industrial
industrial solids and
sludge
domestic and rural
Age Upon
Exhuming
10
6
0.5
6
Liquid
Management
Scheme
leachate
recycling
leachate
recycling
leachate
withdrawal
leachate
recycling
Performance
Upon
Exhuming
excessively
clogged
marginally
clogged
excessively
clogged
excessively
clogged
Critical
Element in
Drainage System
geotextile filter
drain location
geotextile filter
geotextile filter
,
45
-------
6. Laboratory Investigations for Allowable Flow Rate
This section describes the laboratory test setup, materials used and experimental results of
flow rate (permeability) tests which constituted the laboratory testing portion of the project [9].
The experimental design consisted of 48 separate flow columns with four replicate sets of 12
different columns (10 with geotextiles and 2 with sands). The four replicate sets were each
permeated with the following liquids:
« water (as the control liquid)
« leachate"D"
• leachate "P"
• leachate"L"
Furthermore, the permeation rate through each of the columns was varied between very high, high
and intermediate; thus 144 long term flow tests were conducted. All tests were continued until
flow equilibrium or excessive clogging was reached. By determining the equilibrium permeability
of each filter at three higher-than-typical field flow rates, the resulting curve can be back
extrapolated to obtain the allowable permeability of the filter in question at a site specific flow rate:
Thus these tests represented accelerated flow rate tests to determine site specific allowable flow
rates of the various filters under investigation.
6.1 Test Setup
The laboratory test set-up for this project was aimed at determining the relative potential for
paniculate and microorganism clogging of geotextile and soil filters as they were permeated with
leachate from various municipal solid waste landfills. The test method involved measuring the
permeability of the filter over an extended period of time to determine the decrease in permeability
caused by the accumulation of particulates and microorganisms. The permeabilities were
measured under constant head conditions initially and as flow rates decreased, the permeability
measurements were changed to falling head conditions.
Before discussing the specific test setup, details of the test method will be described.. A
geotextile filter specimen or soil/geotextile composite specimen was positioned in a flow column
(permeameter) so that leachate could permeate perpendicular to its plane. The setup provided great
liberty in selecting the test conditions for which the column was exposed for the duration of the
test. Test specimens could be exposed to either partial wetting or full saturation. Partial wetting
resulted in aerobic conditions. Full saturation resulted in anaerobic conditions. It should be noted
that all of the tests in this study were conducted under fully saturated conditions. This decision
was made on the basis that the base of a municipal solid waste landfill is generally considered to be
anaerobic.
The main element of the test setup was the flow column or permeameter. The terms flow
column and permeameter will be used interchangeably in this report. The flow column conformed
to the specification of the ASTM D1987-91 Test Method. Each device held a specific geotextile
specimen on a cylindrical mount of 100 mm (4.0 in.) inside diameter. The geotextile was
temporarily attached to the mount by a model glue. The geotextile was then permanently bonded
46
-------
to the ring with a soldering iron or hot scissors. Finally, the specimen mount was seated against a
containment ring which was a 100 mm (4.0 in.) coupling with PVC solvent cement. For these
tests, the bond had to last for up to one year in the presence of leachate and in some cases with soil
overburden.
Upper and lower tube sections are then joined to the containment ring. A support gravel of
37 mm (1.5 in.) size was placed under the geotextile. To do so, the upper tube, coupling and
lower tube assemblage was inverted and the geotextile was supported on a cylindrical stump which
fit into the lower assembly. Rounded quartz gravel filling the lower tube was then placed on the
geotextile. The lower cap was then bonded to the lower tube thus encapsulating the gravel in the
lower half of the flow column. The column was inverted to its proper orientation and was ready to
be threaded onto a 32 mm (1.25 in.) male adapter of the recirculation system. Some of the gravel
used was small enough in size to fit through the 1.25 in. threaded opening at the base of the flow
column. To ensure that this would not happen a stainless steel nail was drilled into the male
adapter connecting the flow column to the recirculation system. This effectively halved the
opening and held the gravel in the flow column while not restricting flow. It should be noted that
the upper end cap was not yet been bonded to the upper 100 mm (4.0 in.) tube. This afforded a
last chance to inspect the geotextile containment ring seal as well as the ability to carefully place an
optional soil layer above the geotextile. The controlled densification of the soil would not be
possible with the upper end cap in place. The final step in the assemblage of the flow column was
to bond the upper end cap to the upper tube.
Figure 20(a) shows a schematic diagram of the system, while a photograph of the flow
column is shown in both component and assembled form in Figure 20(b).
Once the flow column was assembled it was connected to a leachate recirculation system.
The recirculation system was designed to run continuously and uniformly for at least one year. It
was constructed entirely of plastic materials thus ensuring that buildup of rust would not occur.
The recirculation system consisted of piping network connecting twelve flow columns in
series to a reservoir recirculation system. The system was capable of supplying approximately
95 I/day (25 gal/day) of leachate to each column. As can be seen from the cross section of Figure
21(a) and the photograph of Figure 21(b) the network consisted of two valves, a threaded"!1", a
union, an inlet manifold and an outlet manifold for each flow column.
The valves in the system regulated the flow through the columns as well as provided a
means of isolating any one column from the rest of the set of twelve during the flow reading
periods. The threaded "T" provided access to the flow columns so that a telescoping "L" tube
assembly could be attached to the columns for permeability testing. As can be seen in Figure
22(b), the "L" tube assembly could be made from different diameter tubing and could be utilized in
both the constant and falling head measurement modes. The union was only included in the piping
network to connect and disconnect the columns from the recirculation system at the conclusion of
the test series. Without the unions the recirculation system could not be reused.
The inlet and outlet manifold systems consisted of a 3.0 m (10 ft) piece of PVC pipe tapped
47
-------
flow
upper end cap
100mm
I Cupper
100mm
lower end cap
tube
geotextile test specimen
containment ring
lower end tube
support gravel
100mm «*-
a) Cross section of
permeameter
b) Photograph of Components
and Total Assembly
Figure 20 - How Columns Developed During This Study and Adopted by
ASTM as D-1987 Test Method
48
-------
Inlet
Manifolds
GT
ASTM
D-1987
Flow
Columns
GT
Outlet ^
Manifolds^
a) Cross section of flow
column racks showing
two typical permeameters
b) Photograph of flow column rack
showing twelve permeameters
Figure 21 - Flow Columns and Support System with Associated Pipe Network
49
-------
Supply
Reservoir
Return
Reservoir
CX3 /InFet
Port
Manifold
Outlet
Port
Wooden
Support
Stand
Counter
a) Cross section of recirculation
reservoir system
Figure 22 - Reservoir Recirculation System
b) Photograph of recirculation
reservoir system
50
-------
and fitted to the twelve separate test cylinders to accommodate the leachate flow. One end of the
manifold was attached to the reservoir and the other end was fitted with a 100 mm (4.0 in.)
diameter elbow and standpipe to act as an air vent for the system. The final component of the test
setup was the reservoir system for the recirculation of leachate to the flow columns. As seen in
Figure 22(a) and (b), the system consisted of a series of three reservoirs. The supply and return
reservoirs each held 190 litre (50 gal) of leachate whereas the pump reservoir held only 57 litre (15
gal) of leachate. The supply reservoir was connected to the inlet manifold while the outlet
manifold was connected to the return reservoir. An elevation head of 300 mm (12 in.) was
imparted on the two reservoirs by separating the supply and the return overflows by this exact
amount. This differential in elevation head was the driving force of the entire system. The supply
reservoir was filled by way of the 57 litre (15 gal) pump reservoir contained within a wooden
stand and upon which the other two reservoirs were located. The pump was a centrifugal pump
equipped with a automatic float switch which triggered a charge of 57 litre (5 gal) of leachate
through a pipeline up into the supply reservoir. The submersible centrifugal pump was also
equipped with a solid state counting module which accumulated the number of times the pump was
triggered over the course of a day. Knowing this number afforded a means to calculate the
average amount of leachate that passed through each column.
The entire system of four 12-column units, with associated leachate reservoirs is shown in
plan view in Figure 23. The set-up was totally contained within an isolated room in a laboratory
separated from other activities. The room was well ventilated to ensure that methane, carbon
monoxide and sulfur dioxide did not accumulate. In addition, standard temperature and humidity
were maintained in the room for the duration of the test program.
6.2 Test Procedure
Due to the wide range of flow rates that were to be encountered in using the flow columns
just described, three different testing procedures were used to obtain permeability values. The
constant head test was used to determine permeabilities in the range of 1 cm/sec to 0.01 cm/sec.
Falling head tests, utilizing a 34 mm (1.33 in.) diameter tube, were used for determining
permeabilities in the range of 0.01 cm/sec to 1(H cm/sec. Permeability values less than 1(H
cm/sec were also obtained via the falling head procedure;'however, in this case a 9.7 mm (0.38
in.) diameter tube was used for the measurements due to better sensitivity and control.
Since the procedure for running the different leachate flow streams are identical, the general
procedure shared by all the flow columns will be described and then the differences involved with
running the constant head and falling head tests will be addressed. Additional details are found in
the thesis of G. R. Koerner [10].
The flow columns were periodically tested for permeability at various intervals of leachate
throughput. Initially, tests were conducted every day. This frequent interval of testing was due to
the fact that during the first stage of the test the various materials being permeated were undergoing
considerable change. After a month of testing, the testing frequency decreased to once a week. In
51
-------
3 £
s «
2
2
CO
m —
2 2
JI
o Z g
2 X
E- < cu 2'
a
£:22ONO--cNcn-3-inTf^r?
fifnenTj-T^T^-Tj-Tj-^f
I •§ U^'vo
^ S CN CS
rj ^ en rr in" vo" r~-' oo" ON" o" —* CN CN CN
•*__ .-H—,^^,..,,.,^^ ^ ^ ^
oo ON" 2 H ^J
— CN
oo ON o — CN cn
CN CN en m en m
vo
vo
m
I
us
00
O
-------
testing the 48 flow columns, the three sets of leachates were utilized first and then the water
(control) was evaluated. This practice allowed the room to be washed clean of leachate at the end
of the day with water.
The following sequence was the procedure for running a series of 48 permeability tests.
1. The valves leading from the supply and return tanks were closed.
2. Leachate was released to the elevation of the threaded "T" connection by opening the
main drainage port of the system. Note that at no time during the permeability
measurement process was the geotextile or geotextile/soil specimen unsaturated. This
was important to verify because of the premise that the system was functioning under
anaerobic conditions.
3. The valves upstream and downstream of the twelve flow columns in a series were
closed.
4. Outlets were placed on the discharge end of the system.
5. The individual column permeabilities were then measured. Typically, the test run was
started at columns 12, 24, 36 and 48 and work progressed down.the rack. See Figure
23 for the ordering of the columns along the rack. This insured that the two soil filled
columns were tested last. Back pressure trapped in the system would cause the sandy
soils to become unstable and this obviously would disturb the soil matrix. For this
reason, 5 minutes was allowed to pass before testing any column after it was
disconnected from the recirculation system.
6. The upstream and downstream valves were opened for a single flow column to be
tested.
7. The plug was removed from the threaded "T".
8. The "L" tube assembly was attached to the threaded'T".
9. Aconstant head of the permeant was regulated into the "L"tube assembly maintaining
a constant hydraulic head of 50 mm (2.0 in.). This value was selected because it: was
the prescribed head for measuring geotextile permeability according to ASTM D4491.
10. Upon maintaining a constant head for 1 minute, 1000 cc of liquid was collected from
the outlet, the time for which was recorded.
11. After the time was confirmed from a second run, and was in agreement with the first
reading, the test was complete. Caution was taken so as not to have excessive
hydraulic head over the geotextile specimen at this stage in the testing sequence. The
danger here was of flushing the geotextile free of particulates and biomass leading to
erroneous results.
12. It was felt to be good practice to check the outlet works for zero datum after the test
was complete. This is done by waiting a few minutes after the test was over and
checking if the residual water level was zero after the system reached equilibrium. If
the test did not return to the zero datum mark, the test was rerun.
13. Upon confirming the result, the test was over and the next flow column in the series
53
-------
was evaluated. To do so, the upstream and downstream valves of the flow column
just tested were closed. It was important to isolate one column at a time during this
phase of the testing process.
14. As described previously, the flow columns decreased in permeability over time. As
the columns fell below 0.01 cm/sec, a falling head test was conducted in the same "L"
tube assembly as with the constant head test. The only difference in the procedure was
that the flow rate was so slow that a higher head and longer time was required to
expedite data collection. Therefore the "L" tube assembly was filled to a elevation
head of 900 mm (36 in.) above the geotextile and allowed to fall to 450 mm (18 in.)
above the geotextile. Time was recorded during which the 450 mm (18 in.) charge of
leachate traveled through the flow column. This procedure did not require a known
quantity of leachate to be collected from the outlet works. Only the time for the 450 m
(18 in.) charge of leachate to pass through the geotextile was needed for the required
permeability calculation. '
15. During the progress of permeability reduction, the flow columns were again in need to
shorten the time interval for testing. When the permeability of the columns was less
than ICT4 cm/sec, falling head tests lasted over 30 min. The easiest way to shorten
this testing time was to decrease the volume of permeant passing through the column.
Hence the diameter of the standpipe in the "L" tube assembly was reduced from 34
mm (1.33 in.) to 9.7 mm (0.38 in.). The measurements were conducted at 300 mm
(12 in.) of head independent of the diameter of the standpipe.
16. After the twelve flow columns in a series had been completed, the columns to the
recirculation system were reconnected. The supply and return reservoirs were again
connected to the inlet and outlet manifolds and all valves were opened slightly except
for the valve leading to the supply reservoir. Note that the valves were opened slightly
and not opened completely. This allowed the system to back saturate slowly, purging
the air out of the supply manifold stack. It was important not to back saturate the
columns quickly. The process took at least 30 minutes to perform. If one was too
quick, the risk of backflushing the entire twelve flow columns was taken. This would
lead to irregular and false results.
17. After the entire system was saturated again, the valves were set to regulate 95 I/day (25
gal/day), or other targeted value, through each column. The system was then left to
recirculate until the next flow measurements were made.
18. Upon completing a series of 12 measurements, the next set of 12 were evaluated with
a second leachate, then the third set of 12 with the third leachate and finally the last set
of 12 with water as the permeant.
54
-------
6-3 Leachate. Filters and Flow Rates Used in Tests '
As illustrated in Figure 23 there were four identical flow rate systems in the study. Water
(as the control) was used in one series and three different municipal solid waste leachates were
used in the others. The leachates were retrieved fresh from their respective landfills every two
weeks. Over the entire test period (lasting from 120 to 365 days), a significant change in the
leachate characteristics was observed. What was originally intended to be three very distinct and
different leachates, ended up being three quite similar leachates with only slightly different
characteristics. The average values over the test period are listed in Table 11. Here it is seen that
the BOD5 and TSS values (the two preferred indicator values for filter clogging assessment) were
quite close to one another.
The various filters used were both natural soils (sands) and geotextiles. The two soils
were Ottawa sand and concrete sand, each in the medium sand size range, the difference being the
particle shape and gradation. Ottawa sand was rounded, while concrete sand (being quarried) was
angular. Furthermore, the Ottawa sand was poorly graded while the concrete sand was well
graded. The Ottawa sand was approximately an order of magnitude higher in its permeability and
was classified as an "SP" soil by the Unified Classification System. The concrete sand was
classified as "SW".
Table 11 - Description of Average Leachate Characteristics Evaluated in this Study
Permeanent Leachate Landfill pH COD BOD5 Chlorides TSS TDS~
Management Age (mg/1) (mg/1) ' (mg/1) (mg/1) (mg/1)
Water (Used n/anTa7X) 20" 7 5 n~~ n—
as Control)
Leachate "D" Recycled 10 6.4 3500 2000 1000 300 4600
through the
landfill
Leachate "P" Continuously 5 7.1 4000 2500 1500 500 4000
removed and
pumped to a
treatment facility
Leachate "L" Continuously 2 6.7 3000 2100 3500 600 5900
removed and
pumped to a
treatment facility
n/a = not applicable ~ ~ ~ ~~~~~
55
-------
The ten geotextiles used in this study had properties shown in Table 12. It was felt that this
selection resulted in a relatively complete assortment of geotextile filters used as leachate collection
filters at solid waste landfills. All ten of these geotextiles are commercially available and most (if
not all) are currently being used as primary leachate collection filters in landfill applications.
Table 12 - Description of Geotextile Characteristics Evaluated in This Study
Description
N32W
N14W
N7W
A 10W
N22NW/W
H4NPNW
H8NPNW
H16NPNW
T4HBNW
P6NPNW
Polymer Structure
Type*
PP
PP
PP
PP
PE
PET
PET
PET
PP
PP
woven
woven
woven
woven
nonwoven
and woven
needled
nonwoven
needled
nonwoven
needled
nonwoven
heat bonded
nonwoven
needled
nonwoven
Mass per POA
Unit Area (%)
g/m2 oz/yd2
200 6
270 8
200 6
250 7.5
740 22
130 4
270 8
540 16
120 3.5
220 6.5
32
14
7
10
n/a
n/a
n/a
n/a
n/a
n/a
Apparent Opening Size
(mm) (in.) (sieve #)
0.71
0.81
0.41
0.64
6.32
1.12
2.44
4.65
0.38
1.98
0.028
0.032
0.016
0.025
0.249
0.044
0.096
0.183
0.015
0.078
n/a
n/a
n/a
n/a
50
70
80
100
90
70
Permittivity Permeability
(sec"1) (cm/sec)
4.8
2.4
0.4
1.0 ,
2.4
2.1
1.5
0.5
0.6
1.6
0.34
0.19
0.016
0.064
1.50
0.24
0.37
0.23
0.023
0.32
Regarding the flow quantity passing through the various filters considerable thought and
deliberation went into the selection process. Knowing that flow rates of approximately 18,700
1/ha-day (2000 gal/acre-day) are typical for New York State landfills [11], it would take years to
begin to see measurable decreases in permeability in the experimental setup just described. For
example, the above stated flow rate for a 100 mm (4.0 in.) diameter permeameter is a minisule
0.015 I/day (0.004 gal/day) which was too low for measurement tests of this type. Thus it was
decided to accelerate the testing by greatly increasing the flow rate. Three accelerated flow rate
values were selected. By performing all of the 48 tests at each flow rate, the data could be plotted
and back extrapolated to the lower field anticipated value. The entire setup was changed with new
geotextile and sand filters after each flow rate cycle was completed. The flow rates selected were
those given in Table 13.
56
-------
Table 13 - Flow Rate Values Used for Accelerated Testing
Classification
very high
high
intermediate
Flow Rate (Actual)
I/week
95
3.8
0.95
gal/week
25
1.0
0.25
Flow Rate (Equivalent)
1/ha-day
17,000,000
660,000
170,000
gal/acre-day
1,800,000
71,000
18,000
Thus the number of long term permeability tests that were performed was increased by a factor of
three. The experimental design included four permeants, twelve filters and three flow rates for a
total of 144 long term permeability tests.
6.4 Test Results from Very High Flow Rates
The set of figures presented in this section are the permeability results of the 48 flow
columns at the "very high" flow rate. Note that each set of 12 repeat columns were permeated
with four leachates; water (as the control) and leachates "L", "P" and "D" as described previously
in Table 11 at the very high flow rate of 95 I/week (25 gal/week).
Figure 24 presents the results of the two sand soil filters which were evaluated. In both
cases, the water control was relatively constant, with the permeability of the Ottawa sand being an
order of magnitude higher than the concrete sand. This was understandable since the Ottawa sand
has round and uniform particles versus the angular and more well graded particles of the concrete
sand. The variations in shape and size distribution of the concrete sand allowed for greater
packing and densification. Therefore, a tighter matrix and a more tortuous path for the liquid to
follow was created.
The two sands also behaved differently to the permeation of leachate. When contrasting
the two sets of results, the concrete sand was seen to exhibit an immediate decrease in permeability
while the Ottawa sand exhibited a few days of near "water like" flow before gradually decreasing
in permeability. This can be attributed to the difference in void space of the two soils. Apparently
the concrete sand trapped suspended particles from the onset of the test while the Ottawa sand did
not trap suspended particulates until after an initial flow period.
In comparing the shape of the two soil curves, a similar gradual decrease in permeability
was noticed. When compared to the geotextile response curves it was seen that the soils take
much longer to come to equilibrium than do geotextiles. This is due that the soils being 100 mm
(4.0 in.) thick while the geotextiles range in thickness from 0.38 to 6.3 mm (0.015 to 0.249 in.).
This is an illustration of depth versus cake filtration as described hi Section 2.
In comparing the shape of the two soil leachate response curves to one another it was seen
that the Ottawa sand reached equilibrium faster than the concrete sand. This was due to the fact
that the Ottawa sand consisted of uniform particles with uniform pores. Ottawa sand came to
equilibrium in 70 days while concrete sand took 100 days to reach its residual value. Hence the
57
-------
Ottawa Sand Fast
Water
Leachate V
Leachate T
Leachate '0*
0 20 40 60 80 100 120 140 160 180 200
TIME (days)
10
Concrete Sand Fast
Water
Leachate T_-
Leachate "P"
Leachate tr
0 20 40 60 80 100 120 140 160 180 200
TIME (days)
Figure 24 - Sand Response Curves After 120 Days of Leachate Permeation
58
-------
Ottawa sand was uniformly decreasing in porosity while the concrete sand was very sporadic in its
behavior. '
Figure 25 presents the data from three woven, monofilament, polypropylene geotextile
filters having different percent open areas, e.g., 7%, 14% and 32%. The geotextiles were
different with regard to yarn construction, not weave. A plain weave was utilized in all cases for
these woven geotextiles. The N 7 W geotextile was manufactured with calendered fibers while the
N 14 W and the N 32 W geotextiles were made with noncalendered fibers having a rounded cross
section instead of a flattened tape. The variation between the N 14 W and the N 32 W was that the
N 14 W geotextile utilized a multifilament yarn. In addition to the obvious difference in percent
open area(POA), the shape of the opening also changed with the selection of the different yarns.
The N 14 W geotextile exhibited a three dimensional quality where the other woven geotextiles
were very flat. Although there was not much thickness added to the geotextile it did allow flow in
the bias of the material. This aided in the performance of the filter as well as formed an irregular
surface for filter cake formation. Hence, filter performance was improved.
It should be noted that in all cases the woven geotextiles took 4 to 8 days for measurable
permeability decrease to occur. This time corresponded to 380 to 770 litre (100 to 200 gal) of
leachate passing through the geotextile filter. This suggested that the openings in the woven
geotextiles were too large to immediately catcher trap particulates. Note that during this time the
downstream drainage media was accommodating the particulates that were carried into the drainage
system.
As far as the overall performance of the group of three woven geotextiles was concerned,
the N7W geotextile performed poorly with respect to the N14W and N32W geotextiles. This
indicated that the POA of a woven geotextile was the limiting characteristic which distinguishes
between different performances. It was felt that the minimum POA for acceptable woven
geotextiles was approximately 10%.
There was no discernible difference between the various leachates for any of the woven
geotextiles. This was not surprising for at this time all three leachates had quite similar
characteristics. As can be seen in Table 11, the leachates were all of different ages and individual
properties; however their combined effect yielded similar responses for this method of testing. In
retrospect, the intent was to expose the geotextiles to four very different permeants; however, the
experiment materialized into testing the filters with water and three quite similar leachates.
Figure 26 presents the data from three nonwoven, needlepunched, polyester, geotextile
filters. They were 130, 270 and 540 gm/m2 (4, 8, and 16oz/yd2) mass per unit area (often called
"weight"). The water control tests remained quite constant throughout the 120 day testing
duration. In contrast, decreases in permeability from leachate testing were observed immediately.
This was indicative of particles being trapped in the relatively thick geotextiles from the inception
of the test. All of the geotextiles in this group were constructed of 10 denier polyester fibers
which are fine in comparison to the other geotextile fibers used in this study. The porosity of
59
-------
N 7 W
Water
Uachate T
Leachate
Leachate "D"
20 40 60 80 100 120 140 160 180 200
Time (days)
N 14 W
Water.
Leachate f
Leachate "P"
Leachate "D"
20 40
60 80 100 120 140 160 180 200
Time (days)
N 32 W
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure 25 - Woven Response Curves After 120 Days of Leachate Permeation
60
-------
H 4 NPNW
Water
Leachate "I/
Ueactiate "P"
Leachate "D"
20 40 60 80 100 120 140 160 180 200
Time (days)
H 8 NPNW
Water
Leachate 1°
leachate
Leachate *D*
20 40 60 80 100 120 140 160 180 200
Time (days)
H 16 NPNW
Water
Leachate 1."
Leachate 'P-
Leachate "D"
20 40 60 80 100 120 140 160 180 200
Time (days)
Figure 26 - Nonwoven Needlepunched Curves After 120 Days of Leachate Permeation
61
-------
these geotextiles varied due to the nature of the spunbonding process and mass per unit area by
which they were made.
From the curves on Figure 26 it is apparent that the higher weight geotextiles outperform
the lighter weight geotextiles. This is a result of the thickness of the geotextile. From Darcy's
equation we note that;
Q = kiA (1)
V/t = k(Ah/L)A (2)
where
Q = flow rate .
k = permeability
i = hydraulic gradient
A = cross sectional area
V = volume of fluid passed
t =time
Ah = change in hydraulic head
L = thickness of the specimen
Solving for k we find that;
k = (VL)/[(A)(Ah)(t)] ' (3)
Knowing that Y A, and Ah are constants, the permeability (k) is seen to be directly proportional to
the thickness (L) of the specimen. As the thickness increases it influences the gradient "Ah/L" in a
similar manner and the result is a higher permeability. This confirms that small differences in
thickness will greatly influence the calculated permeability value. It is for the reason that geotextile
permittivity is often used in lieu of geotextile permeability.
All the geotextiles in this group had equilibrium values of approximately 1 x 10-5 cm/sec
with no significant differences on the basis of mass per unit area. When compared to the other
geotextiles in the study these materials where at the low end of the performance scale.
Figure 27 (upper curve) presents the results of T 4 HBNW, a nonwoven, heat bonded,
polypropylene geotextile. The water control was consistent, but the leachate curves throughout the
test continued to decrease. At 120 days (the conclusion of the tests) they were less than 7 x 10-6
cm/sec which was low in comparison to other filters. It is conceivable that this value could have
decreased further and resulted in an excessively clogged filter.
Figure 27 (lower curve) presents results for a nonwoven, needlepunched, polypropylene
geotextile filter of 200 g/m2 (6 oz/yd2) mass per unit area. The water control is constant, and the
trend in permeability decreased gradually to a limiting value around 1 x 10-5 cm/sec. Once again,
the type of leachate did not make a discernible difference. The values are seen to be similar to the
set of curves shown in Figure 26 for polyester fabrics. Thus fabric structure, rather than polymer
62
-------
T 4 HBNW
Water
Leachate f
Leachate "P
Leachate 'D'
20 40 60 80 100 120 140 160 180 200
Time (days)
P 6 NPNW
Water
Leachate t"
Leachate *P
Leachate "D"
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure 27 - Nonwoven Curves After 120 Days of Leachate Permeation
63
-------
type, appeared to be the more relevant issue in permeability testing. This geotextile had much
greater variability in its structure due to a more random laydown process. Therefore this product
had a range of opening sizes instead of a near constant opening size as with the series of polyester
geotextiles.
Figure 28 presents two "special" geotextiles which are uniquely targeted to the filtration
market by their respective manufacturers. Figure 28 (upper curve) gives the permeability results
of A 10 W which was a woven, monofilament, polypropylene, geotextile. The warp and weft
fibers were constructed of greatly dissimilar yarns. The warp fibers were rounded monofilaments
while the weft fibers were crimped staple yarns. The concept behind the construction of this
geotextile was to weave these two dissimilar materials into an open mesh. Since the staple yam
was bulky it also provided the geotextile with openings on the bias and some thickness.
Like the other tests in the study the water control was well behaved for this geotextile. The
curves for the leachate tests quickly decreased to 1 x 10-5 cm/sec (within 45 dayg) and ^
gradually decreased further until the terminal permeability was somewhat less than 1 x 10-5
cm/sec. There was no discernible difference between the different types of leachate.
Figure 28 (lower curve) is of the N 22 NW/W geotextile. This was the most uniquely
different geotextile in the entire study. The geotextile was comprised of many staple polyethylene
fibers supported on an open woven polyethylene mesh. The designation of NW/W was assigned
to this geotextile because it was configured from both nonwoven and woven yarns. The staple
fibers in the matrix consisted of a 500 denier fiber which gave the geotextile a high initial
permeability as well as a low apparent opening size (AOS). This geotextile behaved best of the
nonwoven geotextiles and exhibited an average equilibrium permeability of 3.5 x 10-4 cm/sec,
irrespective of the type of leachate. However; the N22 W/NW geotextile did take 10 days to show
any decrease in flow. This behavior was indicative of a filter with relatively large openings The
response of the N22 NW/W geotextile was encouraging for it brings some additional possibilities
of the utilization of geotextiles into this application other than simply opening up the pore
structure.
6-5 Test Results from High and Intermediate Flow Rates
Asecond set of 48 flow columns were constructed using a completely new set the same 12
filters as given in Table 12. They were permeated by the same (as near as possible) four permeants
as given in Table 11. For these tests, however, the flow rate was reduced from 95 I/week (25
gal/week) to 3.8 l/week(1.0 gal/week). It was designated as a "high" flow rate, recall Table 13.
While the behavior was generally similar to that shown in Figures 24 to 28, there were two
important differences. First, the time to equilibrium was slightly longer, and second the
equilibrium permeability was considerably higher. The complete set of curves are given in
Appendix "B" and accompanying text is found in Reference 10.
A third, and final, set of 48 flow columns was constructed in the same manner (again a
completely new set of filters was used) with the flow rate being further reduced from 3.8 I/week
64
-------
A 10 W
Water
Leachale T_'
Leachate *P°
Leachate "D"
20 40 60 80 100 120 140 160 180 200
Time (days)
N22NW/W
Water
Leachate 1*
Leachale "P"
Leachate "D-
o 10
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure 28 - Response Curves of Experimental Geotextiles After 120 Days of Leachate Permeation
65
-------
(1.0 gal/week) to 0.95 I/week (0.25 gal/week). These tests were designated as an "intermediate"
flow rate, recall Table 13. Again, similar behavior was observed to the higher flow rates, with
equilibrium times being still longer and equilibrium permeability values becoming still higher. The
curves for this intermediate flow rate are also given in Appendix "B" with accompanying text in
Reference 10.
6.6 Comparison of Results Using Different Flow Rates
With equilibrium permeability values for all twelve filters (2 soils and 10 geotextiles) at
very high, high and intermediate flow rates, a characterization and comparison of the results can be
made. Table 14 gives the transition time, i.e., the onset of equilibrium permeability, and the
equilibrium permeability value itself for the 12 filters at the three different leachate flow rates. As
seen in the table there are considerable differences in behavior between the various combinations
of filters and leachate conditions. Some observations follow:
• There was the tendency for all three leachates to reach similar transition times and
equilibrium flow rates. Thus the generalized table represents average values for the three
leachates.
• The water permeant used as the control was worthwhile but did not give any succinct
information except that leachate behaves very differently from water.
• The concept of an accelerated flow rate test seems reasonable on the basis of results of
these tests. By simply passing greater quantities of leachate through the filter,
progressively shorter transition times to flow equilibrium were required and
progressively lower equilibrium permeability values resulted.
• Comparing the different filter responses to one another was difficult except for those
which perform best and poorest. The N22 NW/W geotextile resulted in higher
equilibrium permeability than any other geotextile or soil filter. The N32W was also very
high in its equilibrium permeabaity, with a large number of geotextiles and Ottawa sand
close behind. Conversely, the A10W and T4HBNW geotextiles resulted in the lowest
equilibrium permeabilities.
6-7 Summary of Laboratory Permeability Testing
Presented in this section was the long term permeability testing of four permeants, on 12
different filters at three accelerated flow rates. The resulting 144 tests took 120 to 300 days to
reach equilibrium. The results of the very high flow rates are included in this section, while the
high and intermediate flow rate results are given in Appendix "B".
For all 144 columns, the water flow tests acted as an good control liquid. The permeability
data remained quite constant and, furthermore, gave an indication of the type of fluctuation in the
ASTM D1987 flow columns. The stability of the test results was encouraging, as was the
functioning of this entire experimental test set-up.
Insofar as the leachate flow tests are concerned to obtain allowable permeability values for
a given filter, it took anywhere from 25 to 100 days for the test results to reach equilibrium, or to a
point where equilibrium could be estimated. The equilibrium flow rates for the various types of
66
-------
o
u.
|
^
0
u.
1
'•€
C,
03
C
T*
1
ftS
^
o
tt.
•
x:
Ml
£
a
a;
^
_o
•Si
EC
3
3
E_
* C
3 ^
m
1
C™* ^-s
.2 K
'S «— '
C
2
ii'
tu
u
-< ___^
II
3
|
^
(d) Composite
Nonwoven/V
Geotextile Fil
N22NW/\
67
-------
filters are also given in Table 14 in the order of highest to lowest values of permeability for each
filter group. Here it is seen that the range of permeabilities is from 7 x 10~4 to 4 x 10-6 crn/Sec,
i.e., greater than 100 times different from highest to lowest. Clearly, the selection of a specific
type of filter is important in the performance of a landfill leachate collection system. It should once
again be mentioned that the permeability variation as a function of the type of leachate was not
significant. This was probably due to the fact that over time, the leachates from these three
different landfills came close together, particularly insofar as sediment loading is concerned.
Using the data of Table 14 we are in a position to graph the results of each filter at the three
different flow rates evaluated. This behavior is shown in Figure 29. Figure 29(a) illustrates the
behavior of the 2 sand soils, Figure 29(b) the 4 woven geotextiles and Figure 29(c) the 6
nonwoven geotextiles. The composite nonwoven/woven geotextile is included in this group.
All 12 filters are shown together in Figure 29(d) where one can observe a trend toward a
somewhat leveling of the curves as the flow rate decreases. It is suggested that by back
extrapolation to a site specific flow rate, an allowable permeability for a particular filter can be
obtained. This value of allowable permeability, or "kallow", for a given filter is the essential value
that will be used in the design method to follow.
68
-------
Permeability (cm/sec)
> o o 5 o o o
04 "» *• w K> i. o
^^^
"^-
^*S
Q • Ollawa
• " Concret
--a
N*
Sand 1
e Sand 1
iu f ' - f f
105 106 io7 108
Flow Rate (1/ha-day)
Figure 29(a) - Flow Rate Curves for Sand Soils
10*" 10
Flow Rate (I/ha-day)
Figure 29(b) -Flow Rate Curves for Woven Geotextiles
69
-------
4>
t/3
"s
U
ty
.Q
«
V
I
10
10° IO7
Flow Rate (1/ha-day)
H4NPNW
H8NPNW
H16NPNW
T4HBNW
P6NPNW
N22NW/W
10'
Figure 29(c) - Flow Rate Curves for Nonwoven Geotextiles
10*
10
-1
10
-2,
JS
2
I
10
-•*.!
io-5i
10
-6.
10-
io6 io7
Flow Rate (I/ha-day)
Ottawa Sand
Concrete Sand
N7W
N14W
1 N32W
A10W
H4NPNW
H8NPNW
H16NPNW
T4HBNW
P6NPNW
N22NW/W
rrrf
10*
Figure 29(d) - Master Curves for all Twelve Filter Evaluated (2 sands and 10 geotextiles)
and Estimated Upper and Lower Bounds for Extrapolation Back to
Typical Site Specific Flow Rate Values
70
-------
7. Modeling for Required Flow Rates
In order to compare the allowable permeability of a given filter to a site-specific required
permeabihty an analytic design model is necessary. This section uses the "Hydrologic Evaluation
-deldevelopedby Schroeder, eta, ,12] for the purpose of
7-i Overview of the HELP MnHPl
The HELP model was developed by the U.S. Army Corps of Engineers for the U S
runemal Protection Agency. The primary purpose of the model was to enable a comparison
of landfill design alternatives as judged by an assessment of liquid flow through the solid waste
material. The model can be characterized as a quasi-two dimensional, deterministic, liquid-routing
model for solving liquid balance situations. The liquid can be leachate or water depending on
whether it flows through solid waste or soil.
The model accepts weather, waste, soil and site specific design data and uses solution
techniques that prunarily accountfor theeffects of surface storage, infiltration, evapotranspiration
and percolation. Figure 30 is a conceptual design of the HELP model simulation process
Rainfall/Snow
HHJ
Interception I
Snow J
Evaporation
Transpiration
ion I I
Runoff
Snow
Plant/Vegetation
Evaporation
Filter (Soil or GT)
Depth of Head
Figure 30 - HELP Model Simulation Process
Results are expressed as daily, monthly, annual and long-term peak or average liquid quantities at
any point or location in the landfill. Version 3 of the model developed in 1994 was used for (his
investigation, Schroeder, etal. [13]. acuiorims
71
-------
The required permeability was obtained by imputing the site specific weather, waste, soil
and geometric data and sequentially varying the filter permeability. The permeability value was
varied from 1.0 to 1 to 1(T7 cm/sec. By tracking the peak daily discharge response with respect to
the varying filter permeability a threshold was established. The threshold condition was
established when the filter began to limit the flow into the underlying drainage layer. This
threshold was defined as the value of k,^.
7-2 Use of the HELP Model for Exhumed Sites
Using the site specific configuration of the four exhumed landfills described in Section 5,
coupled with the hydrologic data for the closest city to the site, the HELP model was used to
generate the required permeability of the filter located above the leachate collection drainage layer.
Note that the model does not preferentially distinguish between a geotextile or a natural soil
material. Some of the more relevant input data for the four exhumed sites is given in Table 15.
Here it is seen that many geometric values are required in addition to the hydrologic data. Also
important are characteristics of the waste insofar as its moisture content and density are concerned.
Thus the model has a large number of input parameters. This situation is heightened for our
exhumed sites since they are unknown in regard to some of their details. For this reason default
values were used in some situations.
Table 15 - Input Data of Exhumed sites for Use in HELP Model to
Obtain the Required Filter Permeability
Site
No.
1
2
3
4
Cell Area.
(ha)
2.8
2.8
2.9
5.6
(Acre)
7
7
7.3
13.8
Base Slope
(%)
1.5
1.5
2,0
1.5
Pipe Spacing
(m)
61
61
61
31
(ft)
200
200
200
100
K
Drainage Stone
(cm/sec)
0.01
0.3
0.3
0.3
The curves generated to obtain the required geotextile permeability for each of the four
exhumedsites are given in Figure 31. The results for Site 1, shown in Figure 31 (a), indicate that
the peak daily discharge decreases significantly when the filter permeability fails below 1 x 10-5
cm/sec. All waste heights produce similar trends. Below this value of permeability, the filter is
essentially starving the underlying drainage layer and it (the filter) becomes the controlling
material. Thus the required permeability of the filter for this site is at least 1 x 10~5 cm/sec. The
situation is similar for Site 2 with the required permeability again being at least 1 x 10~5 cm/sec,
see Figure 31 (b). The results for Site 3, shown in Figure 31 (c), indicates a slightly different trend
with the break in the curves being somewhat higher than the first two sites. The required
72
-------
1U
^
•n 105-
W HJ
o> !
tf 1f)4.
« 10 .j
w ;
® in3-
>» •
® m2-
^ 1U ;
m1-
j
/
S
ITE
1
o 30
m waste
i waste
m waste
m waste
10'8 10'7 10'6 lO'5 1(T4 10'3 1(T2 ID'1 10°
Filter Permeability (cm/sec)
(a) - Site No. 1
£
53
5
I
10'
10-
10J
i
-------
tf
SB
WJ
Q
>>
CQ
1.5 m waste
3 m waste
15m waste
30 m waste
10
1(T8 1(T7 1(T6 10'5 1CT4 1CT3
Filter Permeability (cm/sec)
(c) - Site No. 3
10
-1 ir>0
Q
10'
10-
10'
10'
10J
1
A
/
sr
i
x
^
PE 4
1 1
t— r
• — 3rr
* 30
m waste
waste
m waste
7i waste
10'8 KT7 10'6 10'5 10'4 10'3 10'2 10'1 10°
Filter Permeability (cm/sec)
(d)-Site No. 4
Figure 31 - (Continued)
74
-------
permeability of the filter material is estimated to be at least 5x10-5 cm/sec. Lastly, the results for
Site 4, shown in Figure 31(d), again give a clear break in the curves at 1 x 10-5 cm/sec? which is
the minimum required filter permeability. In the design to follow we will use the above listed
values as the required permeability of the filters without increasing them in any arbitrary manner,
e.g., by a partial factor-of-safety. The factor-of-safety will be included in a traditional manner in
the next section.
It should be noted that all of the curves are quite similar with required filter permeability
values between 1 x 10-5 and 5 x 10_5 cm/sec This is not particularly surprising since ^ four
landfills were in the northeastern part of the United States and share many similar site specific
conditions. Also to be mentioned is that Sites 1 and 2 were at the same location, only with slightly
different geometric configurations.
7-3 Summary of HELP Model Utilization
As illustrated by the use of the HELP model in this section, design related values of
required permeability of the filters at the exhumed sites have been obtained. They are as follows:
Site No. 1: 1^= 1 x lO-5 cm/sec ' .
Site No. 2: k^ = 1 x 10-5 cm/sec
Site No. 3: k^ = 5 x 10~5 cm/sec
Site No. 4: k^ = 1 x 10~5 cm/sec
Also as mentioned, the similarity was somewhat expected but nonetheless the HELP model was
further investigated. The sensitivity of the model to numerous input parameters was evaluated
with the results given in Appendix "C". The parameters under investigation were the following:
• Site location.
• Thickness of the lateral drainage layer.
8 Slope of the base of the lateral drainage layer.
• Evaporation depth.
• Runoff curve number.
• Permeability of the lateral drainage layer.
• Height of the waste material.
• Permeability of the waste material.
These parameters were investigated by observing their sensitivity on the peak daily discharge of
the drainage system and (most importantly for the purposes of this study) the permeability of the
filter. The entire parametric study is presented in Appendix "C" with the relevant conclusions.
The HELP model is a useful tool for comparison of design alternatives. It is used by all
landfill regulators in the United States. Hence (by default), it must be used by the consulting
engineering community and landfill owner communities as well. The model does not seem to be
extremely sensitive to subtle changes to weather, soil or site specific design data. However, it
75
-------
does appear to yield logical response trends to systematically varied input parameters.
The results generated from the HELP model tend to be conservative. This implies that
estimates of leachate generated at the base of a landfill generally overpredict the amount that is
actually collected. Confidence in the HELP model is greatly increased when it is calibrated with
actual field data. This can be obtained from estimates of leachate generation rates in the open
literature or from the adjacent cell in a landfill expansion situation. Version 3 of the HELP model
is very user-friendly and is a powerful tool to access the hydrologic performance of landfills.
76
-------
8. Design Method and Substantiation
At this point we are in position to not only develop a design methodology (utilizing both
the laboratory and design sections of the report), but also to substantiate the design method 'using
the results of the field exhumed sites. Hence, the results of Sections 5, 6 and 7 will be brought
together in this section, per the project flow chart of Figure 7.
8>1 Design Methodology and the Drainage Correction Factor
A general method for designing all engineering materials is the formulation of a factor-of-
safety (FS). It is the ratio of an allowable material property to that of the site specific required
property. The value must be greater than one, since it is necessary to compensate for any
uncertainties in the testing and/or design processes. While the actual target value is site specific
and ultimately the decision of the designer, it is felt that the recommended value for a landfill filter
design should be considerably in excess of one, e.g., ten or higher.
The value of factor-of-safety for filtration is formulated by comparing the allowable filter
property for the candidate material with the required filter property for the specific site under
consideration. For the case of filter designs involving permeability this ratio is expresed as
follows:
reqd
where
FS = factor-of-safety
kallow = allowable permeability
kreqd = required permeability
The allowable permeability is determined from laboratory testing as described in Section 6. More
specifically, Figure 29 illustrated its evaluation for the twelve filter materials and their respective
leachates. The required permeability is generally obtained using the HELP model as was
described in Section 7 for the four sites which were exhumed.
In dealing with geotextile filters, an equivalent term for permeability is the permittivity,
' Y'. It is defined as the traditional permeability value "k" divided by the geotextile thickness "t":
JtV
V =-
t (5)
Thus an equivalent equation for factor-of-safety is as follows:
Callow
FS=
'
77
-------
where
FS = factor-of-safety
Vaiiow = allowable geotextile permittivity
V reqd = required geotextile permittivity
Consistency with design methods in all engineering materials (including geosynthetics) suggests
that these equations be used for filters of leachate collection systems, which is the topic of this
study. Unfortunately, the field exhuming efforts described in Section 5 suggest the need for a
modification. Field Sites 1, 3 and 4 were all cases where the geotextile filter was wrapped directly
around perforated drainage pipes. Thus only a small portion of the geotextile was available for
flow since it was limited to those small areas directly adjacent to the holes in the pipe. To a lesser
extent field Site No. 2 also had a flow area less than the entire footprint of the landfill cell. It will
also have to be suitably accommodated. As a result of such restrictions in available drainage area
in all four of the field exhumed cases, a compensating term for drainage areas less than the full
footprint of the landfill cell must be formulated. The term we have selected is a drainage correction
factor, or "DCF', which is formulated into Equations 4 and 6 as follows:
FS = allow
k .xDCF
reqd
or
FS =
where
DCF = drainage correction factor
The value of DCF is defined as the ratio of the entire landfill or cell area (i.e., the footprint)
divided by the actual flow area that is available beneath the filter(s) for drainage. The DCF is site
specific and determined via a geometric ratio taking into account the area of influence with respect
to the size and configuration of the drain.
Expressed as an equation, the DCF is as follows:
DCF
where
DCF = drainage correction factor
App = footprint area of the landfill or cell
Aj-jg = area of actual downstream drainage system
Four typical field situations have been encountered and are illustrated in Figure 32.
78
-------
Geotextile
tm • m
Perforated Pipe~]>
Geomembrane
(a) Areal Filter
DCF = 1.0
(b) Trench Wrap
DCF =10 to 40
| Geomembranej
(c) Socked Corrugated Pipe
DCF = 64to260
I Perforated Pipe |
I Geomembranq
(d) Socked Smooth WaU Pipe
DCF = 7,500 to 24,400
Figure 32 - Various Leachate Collection Filter Configurations with the Associated
Range of Drainage Correction Factors
79
-------
(a) For an areal filter placed over the entire footprint of the landfill or cell, the DCF is equal
to 1.0 since the footprint area and drain system area are equal, see Figure 32(a). Note
that this could be considered conventional design practice which would eliminate the
need for a DCF term.
(b) For a geotextile wrapped gravel, or trench wrap, the DCF varies from 10 to 40
depending upon the spacing of,the drains and the size of the drainage gravel cross
sectional area, see Figure 32(b). Note that the gravel must be sized appropriately with
respect to the holes or slots in the perforated pipe and vice versa, but this is
conventional design, see Cedegren [14].
(c) A geotextile wrapped around, a corrugated pipe with slots in the valleys of the
corrugations results in typical DCF values of 60 to 260, see Figure 32(c). The value
depends on the pipe spacing and the diameter of the pipes. Approximately 50% of the
surface area of the pipe is available for leachate flow using corrugated pipe, i.e., the
area above the valleys of the corrugations. This value dividedinto the landfill footprint
area is the numeric value of the DCF.
(d) A geotextUe wrapped around smooth wall pipe with holes perforated in it results in
DCF values of 7,500 to 24,000, see Figure 32(d). The value depends on the pipe
spacing and the number and diameter of holes in the pipe. Since very little area is
available for flow, .the resulting DCF values are extremely large.
Table 16 gives some additional insight into the influence of the different designs illustrated
m Figure 32 and the impact of drain spacing, drain size, hole size and number of holes on the
numeric value of DCF.
Table 16 - Selected Values of Drainage Correction Factors (DCF) for Use in Calculating the
Factor-of-Safety of a Leachate Collection Filter &
Drain
Configuration
(a) Areal Coverage
(b) geotextile
wrapped around
gravel (i.e., a
trench wrap)
(c) geotextile around
corrugated pipe
(i.e., socked
pipe)
(d) geotextile around
smooth wall
pipe (i.e. socked
pipe)
Drain
Spacing
(m)
n/a
15
30
45
60
15
30
45
60
15
30
45
60
(K)
n/a
50
100
150
200
50
100
150
200
50
100
150
200
Drain
Size
(mm)
n/a
450 x 300
450 x 300
450 x 300
450 x 300
150
150
150
150
150
150
150
150
(in.)
n/a
18 x!2
18 x!2
18x12
18 x!2
6.0
6.0
6.0
6.0
6.0
6.0
6.0
6.0
Hole
Size
(mm)
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
12
12
12
12
(in.)
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
0.5
0.5
0.5
0.5
Number
of Holes
(perm)
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
1.8
1.8
1.8
1.8
note: u/a = not applicable — ~ ' '
(per ft)
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
6
6
6
6
Drain
Correction
Factor
(DCF)
1
10
20
30
40
60
130
190
260
7,500
12,000
18,000
24,000
80
-------
8.2 Substantiation of Design Methodology
The implication of the high DCF values in Table 16 on the resulting factor-of-safety values
in Equations 7 or 8 is obvious. The numeric impact will be illustrated on the basis of the field
exhumed sites that were described in Section 5.
Table 17 gives a summary of the findings from the four exhumed sites. Included are kallow
values from the laboratory testing of the same type of geotextile used at each site, and k^ value!
for the site-specific hydrologic and waste conditions from the HELP model. Also included in
Table 17 is the DCF for the geometry of the geotextile filters that were actually used at each of the
exhumed field sites.
Using the design formula recommended in Equation 7, it is seen that the calculated factor-
of-safety values are drastically different for the four different sites. Clearly, the three sites that
showed evidence of excessive clogging, i.e., Sites 1, 3 and 4, could readily have been predicted
as failures (FS =0.0003, 0.18 and 0.53, respectively) and major modifications should have been
made in the design stage. Also, Site No. 2 was still functioning and was substantiated using this
formulation although its factor-of-safety is certainly not excessive (i.e., 7.1) in light of the
required long-term performance of the filter.
Obviously, the factor-of-safety must be greater than one, and the higher the value the more
conservative is the design. As far as the selection of a recommended value for the factor-of-
safety, the decision is really that of the design engineer. Of course, it must be approved by the
reviewer of the permit application as being acceptable for the particular site in question Baised
upon observations at the the four exhumed sites it appears that the factor-of-safety should have
been in excess of ten (10) since the sites are either practicing leachate recycling or are accepting
sludge as part of the waste stream. This issue as well as providing design guidance as to index
properties of specific geotextile filters will be given in the conclusions and recommendations
section.
Table 17 - Corroboration of the Newly Modified Factor-of-Safety Equation (Equation 7)
as Applied to Four Exhumed Field Sites
Site
1
2
3
4
-™ — a^— •.
Observed
performance
Terrible
Good
Terrible
Poor
l^allow
(cm/s)
6 x 10-4
1 x 10-2
9x10-3
9xlO-3
kreqd
(cm/s)
1x10-5
1 x 10-5
5 x 10-5
1x10-5
Value of
DCF
24,000
140
990
1,700
Calculated
factor-of-safety
value
0.0003
.7.1
0.18
0.53
Predicted
performance
Failure
Acceptable
Failure
Failure
81
-------
9. Conclusions and Recommendations
The project around which this report has been written focused on various types of landfill
leachate collection filters. It was felt that filters were the most likely component of a leachate
collection system to become excessively clogged. Such excessive clogging would eventually
starve the downstream drain (gravel or geonet) and pipe system of leachate, thereby rendering the
liquids management practice ineffective. Current options of liquids management practices in this
regard are either leachate withdrawal on demand or leachate recycling.
It should be clearly recognized that the design of the filter of a landfill leachate collection
system is considerably more challenging than filter design in transportation and geotechnical
engineering applications. This is due to the nature of the liquid (leachate versus water), the
required lifetime before excessive clogging occurs (generally decades) and the filter's
inaccessibility after the waste is placed (cost of exhuming and environmental concerns). The
situation is exacerbated if leachate recycling is the liquid management strategy at the landfill site.
With these thoughts in mind and having the benefit of an earlier EPA financed study to bear
upon, a five-part study has been reported herein. The individual parts, and their respective
conclusions, were as follows.
(a) The field exhuming work described in Section 5 was particularly rewarding. It gave insight
as to "ground truth" and although only four sites were exhumed they set the tone for much of
the work to follow:
• In the three sites with little to no leachate flowing out of the collection pipes, a geotextile
filter was the culprit with clear indications of excessive clogging.
• In all three cases, the geotextile filter was wrapped directly around the perforated removal
pipe, i.e., it was so-called "socked pipe". This configuration must simply cease to be
designed and installed.
• The fourth site exhumed was still functioning. It had a geotextile filter wrapped around a
gravel trench with a perforated pipe within the gravel. This was a more acceptable filter
configuration than was socked pipe.
• The exhuming of these four sites points to what was felt to be the preferred filter design
configuration, i.e., one hi which the filter covered the entire footprint of the landfill or cell.
This was called an "areal filter".
• Either geotextile or soil filters could be used in areal filter configurations, however,
geotextiles will generally be favored due to savings in the initial installed cost and greater
available landfill volume due to reduction in thickness over natural soil filters.
. • Regarding the leachate itself, it appeared as though either high suspended solids or high
microorganism content, or a combination of both, were important insofar as excessive filter
clogging was concerned.
• These two properties of leachate were best quantified by the total suspended solid (TSS)
content and the biochemical oxygen demand (BOD) as characterized by the 5-day
measurement, i.e., BOD5.
82
-------
• Although very subjective, leachates with TSS > 2500 mg/1 and/or BOD5 > 2500 mg/1 were
considered harsh and were of special concern.
(b) Long-termlaboratorypermeability tests described in Section 6 focused on ten geotextile filters
and two soil filters. The tests were performed at constant flow rates using four different
permeants. Water was the control and its response was counterpointed against three different
municipal solid waste leachates. The flow rates were accelerated over typical field flow rates
and were categorized as "very high", "high" and "intermediate". Data for the very high flow
rates were given in Section 6 and for the latter two flow rates were given in Appendix "B".
• The ASTM D1987 test method can be used as a performance test to determine allowable
filter permeability using most types of liquids, including leachate, as the permanent.
• The test can be conducted for geotextile or soil filters, under aerobic or anaerobic
conditions, using falling head or constant head measurements, for as long of a period of
time as is necessary to reach a conclusion about terminal (or equilibrium) permeability.
• The test results can be presented as permeability (k), permittivity (v) or flow rate (q).
• The twelve different filters examined in these tests came to approximate equilibrium
between 120 and 300 days.
• The equilibrium flow rates varied by two orders of magnitude between the different types
of filters investigated.
• The highest equilibrium flow rates were the high denier, thick, nonwoven geotextiles, open
woven geotextiles and sand, i.e., the filter materials with the largest voids.
• Utilizing accelerated quantities of leachate flow (i.e., higher than field anticipated values)
passing through the various filters allowed for back-extrapolation to the leachate quantity
anticipated at a given landfill site. This procedure resulted in a site specific value of
"kalW''Vallow" or "qallow" for the candidate filter.
• The three different municipal solid waste leachates used for these long term laboratory tests
gave remarkably similar flow rate results. This was surprising since one leachate was
selected for its high TSS content, another for its high BOD5 content and the third for a
combined high TSS and high BOD5 content.
• The use of water as a control permanent was important in illustrating that filters for leachate
are considerably more challenging than water with respect to excessive clogging. The use
of water as a control also signified that the laboratory test setups were properly
functioning.
(c) An effort was made to utilize the HELP model developed under EPA sponsorship. Such
modeling was necessary from a design perspective since filter design requires a predicted or
required, flow value of Ty", «Vieqd» or "q^" for utilization in arriving at a factor-of-safety
value. In this part of the study (described in Section 7 and presented as a parametric study in
Appendix "C") the following conclusions were reached.
83
-------
• For no waste in the cell and the drainage system acting as a dewatering system, flow rates
decreased with decreasing the drainage layer permeability from 1.0 to 1 x 10-3 cm/s, i.e.,
the filter was of no consequence.
• Furthermore, the flow rate was not altered until the filter permeability became less than
approximately 1 x 10~5 cm/s.
• As waste was placed in the facility, drainage layer permeabilities from 1 .0 to 1 x 10-3 crn/s
only moderately decreased flow rates from the collection system.
• Furthermore, for filter permeabilities from 1.0 to 1 x 10-5 cm/Sj tne flow rate exiting from
the system only decreased moderately.
• For filter permeabilities less than 1 x 10~5 cm/s, however, the drainage system flow rates
decreased rapidly.
• For the "no filter" case presented in Appendix "D", the HELP model indicated that the
drainage layer permeability directly influenced the amount of leachate drained for the no
waste condition. For accumulated waste, the flow of leachate did not vary greatly for drain
layer permeabilities between 1.0 and 1 x 10~5 cm/s. Lower values, however, began to
rapidly decrease the exiting flow rates.
• The HELP model was felt to be successful in predicting the required flow rates from the
four exhumed field sites for subsequent analysis in the design model substantiation.
(d) The design method formulation and substantiation part of the study described in Section 8
was successful in utilizing each of the previous sections and arriving at a final design
formulation. Furthermore, the findings of the field exhumed sites were used to challenge the
validity of the design model. More specifically, the following items were concluded.
• The traditional factor-of-safety model of comparing allowable permeabili ty with required
permeability was not appropriate if the site specific drain configuration limited the available
flow area to less than that of the landfill or cell footprint area.
• In light of the above statement a drainage correction factor, "DCF", was included in the
denominator of the traditional factor-of-safety formulation, i.e.,
rjc _ allowable
(7)
V ' '
k .
required
• Values of DCF varied enormously. They ranged from 1.0, where the filter covered the
entire footprint of the landfill or cell, to 24,000 for the case of geotextile socked smooth
wall perforated pipe.
• Other filter configuration strategies resulted in DCF values that were between these two
extremes.
• The implication of high DCF values on the calculated value of filter factor-of-safety was
obvious; high DCF values resulted in disastrously low factors-of-safety values.
84
-------
• These low factor-of-safety values were associated with geotextile socked pipe of both the
smooth wall and corrugated wall variations. Neither type should be used for the filter of a
leachate collection system.
• The above formulated design equation was utilized to predict the factor-of-safety of the
filters in the four sites that were exhumed. A direct correspondence to the observed
behavior was seen to exist. For the three sites where leachate was not flowing, the
predicted factor-of-safety values were 0.0003, 0.18 and 0.53, respectively. All are
obviously unacceptable. For the one site where leachate was still flowing the predicted
factor-of-safety was 7.1. This value was felt to be marginally acceptable.
• While the "ground truth" data of the four exhumed sites was sparse, we feel that the design
formulation embodied in Equation? was substantiated and should be used for all landfill
leachate collection system filters.
(e) Long-term laboratory permeability tests were performed on the "no filter" design strategy as
described in Appendix "D". in this strategy, "select" solid waste was placed directly on
drainage soil (gravel or sand) with no filter used whatsoever. Eight long term "no filter
tests", 4 on gravels and 4 on sands, were conducted for up to 1000 days. Flow rates though
the permeameters were typical of leachate field rates, i.e., these tests were not accelerated
flow rate tests. The following conclusions were reached.
• For the gravel tests, the solid waste above the drainage layer became the controlling flow
material. The system permeability was essentially constant for the duration of the tests.
• For the sand tests, the solid waste again dominated the flow behavior. However, there
was a trend that the system permeability was decreasing over time which was due to
paniculate and/or microorganisms clogging of the sand drainage material.
• It was felt that if the "no filter" strategy is used, the drainage material should be gravel with
a permeability of 1.0 cm/sec, or higher.
• Furthermore, if a "no filter" strategy was proposed and the leachate had high TSS and/or
high BOD5 (e.g., either value greater than 2,500 mg/1), then long-term laboratory tests
should have been conducted to substantiate the design feasibility.
Insofar as recommendations from this study are concerned, it is felt that the optimal
strategy for the filter of a leachate collection system is to place a geotextile over the entire footprint
of the landfill. In this way, the drainage correction factor (DCF) in Equation 7 is 1.0 and all of the
filters we have investigated during the course of this study result in acceptably high long term flow
rates for the sites that were evaluated. Even further, we can suggest that for municipal solid waste
leachates where the TSS and BOD5 are not excessively high, the types of geotextile filters listed in
Table 18 can be recommended.
85
-------
Table 18
Recommended Geotextile Filtered) for use with Relatively Mild Landfill Leachates
Which Have Low TSS and Low BOD5 Values, e.g., Less than 2500 mg/l<2>
Type of
Geotextile
(a) Woven MonofllnmpntO)
mass per unit area,
g/sq. m (oz/sq. yd.)
percent open area,
grab tensile strength,
N (Ib)
trapezoidal tear strength,
N(lb)
puncture strength,
N(lb)
burst strength,
kPa (Ib/sq. in.)
(b) Nonwoven Needle Punched*3)
mass per unit area,
g/sq. m (oz/sq. yd.)
apparent opening size,
mm (sieve size)
grab tensile strength,
N(lb)
trapezoidal tear strength,
N (Ib)
puncture strength,
N(lb)
burst strength,
kPa (Ib/sq. in.)
Granular Soil Protection
Layer Over Filter
170 (5.0)
10
1100 (250)
250 (55)
400 (90)
2700 (390)
200 (6.0)
0.212 (#70)
700 (160)
250 (55)
250 (55)
1300 (190)
Select Waste(4) Placed
Directly Over Filter
200 (6.0)
in
1U
1400 (310)
350 (80)
500 (110)
3500 (510)
270 (8.0)
0.212 (#70)
900 (200)
350 (80)
350 (80)
1700 (250)
Notes. 1. Laboratory test data and the rp.nni«itp ri^cion ma,, ^«..,u :« i .*•*,.
_—™,^_, ^ov v,«vn tmu UK, icajmaiic ueMgn may result m less <
than listed in the table. Properly designed they are acceptable
Low TSS and BOD5 refers to < 2500 mg/1, for higher values of TSS and/or BOI>5,
the procedures and details given in this report should be followed
J^J o UeS/AStfentgth ,Usted in ?e above table m in approximate agreement of the
Class 2 and Class 1 values per the proposed AASHTO M288 specification for
transportation facilities in the high and very high survivability ratings, respectively
Select waste cannot contain any hard or coarse material which can damage the
geotextile. For hard or coarse waste the strength requirements of the geotextile
must be increased at the discretion of the design engineer
86
-------
10. References
1. Christopher, B. R. and Fischer, G. R., "Geotextile Filtration, Principles, Practices and
Problems," J. Geotextiles and Geomembranes, 11(4-6), 1992, pp. 337-354.
2. Christopher, B. R. and Holtz, R. D., "Geotextile Engineering Manual," FHWA Rep No
FHWA-TS-86/203/1044P, Federal Highway Administration (FHWA), 1985, Washington,
3. Koerner, R. M., Koerner, G. R., Fahim, A. K.and Wilson-Fahmy, R. F., "Long-Term
Performance of Geosynthetics in Drainage Applications," NCHRP Rept. 367, National
Academy Press, 1994, 54 pgs.
4. Chian, E. S. K. and deWalle, F. B., "Sanitary LandfillLeachate and Their Treatment " J
Environ. Eng. Div., ASCE, Vol. 102, No. EE2, April 1976, pp. 411-431.
5. Heerten, G., "A Contribution to the Dimensioning Analogies for Grain Filters and Geotextile
Filters," Proc. Geofilters 92, Karlsruhe, Germany, Oct. 20-22, 1992, pp. 210-216.
6. Driscoll, F. G., Groundwater and Wells. Johnson Division, St. Paul, MN, 1986.
7. Koerner, R. M. and Koerner, G. R., "Landfill Leachate Clogging of Geotextile (and Soil)
Filters," EPA/600/2-91/025, July 1991, available through NTIS as Report No. PB91-
213660.
8. Koerner, G. R., Koerner, R. M. and Martin, J. P., "Field Performance of Leachate
Collection Systems and Design Implications," Proceedings of the 31st SWANA Conference,
San Jose, CA, 1993, p. 365-380.
9. Koerner, G. R., Koerner, R. M. and Martin, J. P., "Design of Landfill Leachate Collection
Filters," Journal of Geotechnical Engineering, ASCE, Vol. 120, No. 10, October 1994 pp
1792-1803. ' *v'
10. Koerner, G. R., Performance Evaluation of Geotextile Filters used in Leachate Collection
Systems of Solid Waste landfills. Ph.D. Thesis, Drexel University, Philadelphia, PA, June,
11. Phaneuf, R. J., "Landfill Bioreactor Design and Operation," Proc. Landfill Bioreactor Design
and Operation, Mar. 23-24, 1995, U.S. EPA (to be published).
12. Schroeder,P. R., Lloyd, C. M., Zappi, P. A. and Aziz,N. M., 'The Hydrologic Evaluation
of Landfill Performance (HELP) Model," EPA/600/R-94/168a, September, 1994.
13. Schroeder, P. R., Dozier, T. S., Zappi, P. A., McEnroe,'B. M., Sjostrom, J. W. and
Peyton, R. L., "The Hydrologic Evaluation of Landfill Performance (HELP) Model:
Engineering Documentation for Version 3," EPA/600/9-94/xxx, U.S. Environmental
Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH, 1994.
14. Cedegren, H. R., Seepage. Drainage and Flow Nets. J. Wiley & Sons, NY, 1967.
15. , Standard Specification for Geotextiles, M288-96 (draft), AASHTO, Transportation
Research Board, Washington, DC, February 2, 1995.
87
-------
16. Oweis, I. S., et al., "Hydraulic Characteristics of Municipal Refuse," Journal of Geotechnical
Engineering, ASCE, Vol. 116, No. 4, April, 1990, pp. 53.
17. Koerner, G. R. and Koerner, R. M., "Long Term Permeability of Granular Drainage Media "
Proceedings Workshop on Leachate Recycling, D. Reinhart, Ed., March 1995 (in press).
18. Daniel, D. E. and Koerner, R. M., "Quality Assurance and Quality Control for Waste
Containment Facilities," U. S. EPA, Technical Guidance Document CA#CR-815546-01-0
Cincinnati, OH, 1993.
88
-------
APPENDIX "A"
BEHAVIOR OF BIOCIDE TREATED GEOSYNTHETICS
A-l Introduction
In light of the concern over excessive clogging of leachate collection filters, an attempt at
using biocides in the various flow systems was undertaken. This was done under the assumption
that the biocide would kill the microorganisms that came into contact with it and that the non-viable
(i.e., dead) matter would pass through the system in much the same way that fine particles or
sediment moves through any other drainage system. Furthermore, the introduction of the biocide
was felt to be best achieved when delivered on a long-term basis rather than as one bulk dosage.
Thus, biocide was added to the polymer compound during fabrication of selected geonets or
geotextiles. The reasoning for this approach was that the biocide would time release, via molecular
diffusion, through the polymer structure and migrate to the surface of the ribs or filaments over a
long period of time. If the approach was seen to be of value, calculations could then be made as to
the long-term time release behavior. This Appendix to the report describes our attempts to increase
the flow rates of landfill leachate filters and drainage systems using biocide treated geonets and
geotextiles.
A-2 Type of Biocide
The biocide used in this study is Vmyzene® SB-1 PR manufactured by Morton Thiokol,
Inc. of Danvers, Massachusetts. Vmyzene SB-1 PR is a concentrate of 10, 10' -
oxybisphenoxarsine (OBPA) in a polypropylene resin carrier. The product, is supplied as a
homogeneous solid in pelletized form measuring approximately 3.5 mm by 2.5 mm. It is
recommended for use in polyolefins and other polymeric compositions requiring preservation
against fungal and bacterial deterioration. The manufacturer states that "low levels of Vinyzene
SB-1 EEA (a similar product but in an ethylene acrylic acid copolymer resin carrier) will provide
long term preservation against fungal and bacterial attack and will help prevent surface growth,
permanent staining, embrittlement and premature product failure...Vinyzene can be incorporated
into the polymer compound at any convenient stage of the manufacturing process. The product can
be fed into an extrusion operation in much the same way as pelletized color concentrates."
In 1976, EPA placed OBPA on its list of suspect pesticides that might be hazardous to
human health. EPA's review of animal and other studies on OBPA, however, indicated that it was
not as hazardous as originally suspected. On May 4, 1979 the U.S. Environmental Protection
Agency decided that the pesticide, OBPA, which is used in a wide variety of plastic consumer
products to protect them from fungal and bacterial damage, did not pose a threat to human health or
the environment if used in accordance with label instructions. This decision meant that OBPA was
restored to its former place on EPA's list of currently registered pesticides.
Materials containing OBPA include swimming pool liners, wall coatings, vinyl roofs on
cars, marine upholstery, awnings, industrial fabrics, and caulking for tubs, sinks, weatherstripping
and gutter repair. The EPA registration number for Vinyzene® is 2829-115 and Morton Thiokol's
89
-------
patent number is 4,086,297.
Some selected physical and chemical properties of 10, lO'-oxybisphenoxarsine (OBPA)
as follows.
are
1 Molecular Formula:
1 Molecular Weight:
Structural Formula:
C24HI6As203
502.2
1.40-1.42
White to off-white crystalline solid
185 - 186°C.
< 10"6 torr @ 21°C
• Specific Gravity:
•Appearance:
• Melting Point:
• Vapor Pressure:
< IQr6 torr @ 100°C
1.0 x 10~3 torr @ 150°C
3.0 x 10~3 torr @ 200°C
8.5 x 10-3 torr @ 250°C
• Thermal Decomposition Range: 300-380°C
•Solubility: 5ppminH20
2.75 gm/100 gm of 95% ethanol
2.30 gm/100 gm of isopropanol
2.78 gm/100 gm of xylene
A'3 Incorporation of the Biocide into Different Geosvnthetics
The biocide was shipped to the respective geosynmetic fabrication facilities for inclusion into
the candidate geonet or geotextile. After the dosage was decided upon (it varied from 1 to 8% by
weight), it was added to the standard compound, suitably mixed and extruded into ribs (for
90
-------
geonets) or filaments (for geotextiles).
In the first series of tests, either 1, 2 or 4% biocide levels were introduced into the
compound to produce a 6.3 mm (0.25 in.) thick, high density polyethylene (HOPE) geonec. For
control purposes, the same type of geonet was produced without the addition of any biocide. The
cross section of the columns for these tests consisted (from the top down) of sand/geotextile/
geonet/gravel and they were coded as Series "A", see Table A-1. Tests were conducted both
saturated at all times (hence anaerobic) and allowed to air dry between readings (thus aerobic). All
tests in this series were run for 444 days.
The second series of tests, i.e. Series "B", utilized the biocide in a geotextile and did not use
a geonet. The cross section consisted of sand/geotextile/gravel. The geotextile was a nonwoven
needle punched polypropylene and it contained either 2% or 4% biocide. The biocide was
introduced at the fabrication facility along with the manufacturers standard compound of resin,
carbon black (or other antidegradant) and processing package. As seen in Table A-l, there were
also geotextiles included with no biocide to act as control materials. Tests were conducted under
both constantly saturated conditions (hence anaerobic) and intermittently saturated, then air dried
conditions (thus aerobic). All tests in this series were performed for 444 days.
Test Series "C" consisted of biocide treated geotextiles and no geonets; but unlike the
previous series, three different types of geotextiles were evaluated. The geotextiles were
nonwoven needle punched (as before) and also two types of woven monofilament geotextiles with
different opening sizes, see Table A-l. The tests were also different in that gravel was used above
the geotextile instead of sand. Thus the flow column consisted of gravel/geotextile/gravel, with the
geotextiles treated with 2, 4 or 8% biocide. Again, the biocide was introduced at the
manufacturing facility. In this series, which lasted 121 days, all tests were kept saturated, thus
anaerobic.
A-4 Field Testing and Evaluation Procedures
Flow rate testing for each of the columns with biocide treated geosynthetics utilized the 100
mm (4.0 in.) diameter incubation and test permeameters described in the main body of the report.
There were 8 columns in Series "A", 8 columns in Series "B" and 16 columns in Series "C". All
permeameters were made and conducted according to the ASTM D1987 test method. Series "A"
and "B" were evaluated over a 444 day duration and Series "C" were evaluated for 121 days.
All of the tests in this biocide study used leachate from the same municipal solid waste
landfill site. This particular leachate had the highest concentration of COD, TS and BOD5 of the
six landfill leachates which were evaluated during the course of the project. The approximate
properties of the leachate were as follows:
• pH = 5.8
• COD = 40,000 mg/1
• TS = 17,000 mg/1
• BOD5 = 24,000 mg/1
91
-------
•o
.•s
i
s
,p
I
I
ia
en
g
a
o
T3
I
3l
C/5 t>
PQ
^
3 .a is .H is .y. is a •- •- o 0.-a o -^ «
« 'S « '?5 o is o 2 oolSIc'oIS'oIS
JitP^rO ^"rrOOi-Oiio
.
-
C
=«-<=<
uiiiiii laiiiiii
OO— '— •
oooooooo
OOOOOO OOOOOOOO
OOOOOOOO OOOOOOOO
H &* 9^ ^ 9" 5* ^ ^ cu a. Cu a. a. cu a. a.
uo-p-fca.^^)!, D-a.a.^DHa.D.a,
.?;zzzzzzz zzzzzzzz
zzzzzzzz zzzzzzzz
ccccccc cccccccc
C/5COCOCOCOOOC/3CO C/3&OCOCOCOC/3COC/3
z z <— i «N z z
««<<<
6666
o
a
b8
_«.*J
o o —
"8
•o
H « ^>p p on
CJ *J *J «
••a § § § P
*^ -^* r-i .—i *-»
•^«§8g
jgco P ^ % c 0
s O | g 1^,^
92
-------
T3
C
•a
G
O
U
_
i
i>
•§
*"*
o
*5
I
^ 5
on «
S —
w .2
o CQ
u
0 -5
0.
u
4> T3
>< O
f «
O o
o.
CO U
O 2r
£ o
:§ ^
^ _
8 .§?
o
1 s
Is S.
^5 ^^
« £•<
O
flj
si
00 <
^ •
50
C
•3
U
.O.O.O.H.O.H.O.O
HI"! HI"!
bccccccc
22222222
OOOOOOOO
• i i i i . i ,
O O inn
o o o' o o' o o" o
x & S: fc fc §: fe &
zzzzzzzz
S 2 2 2 2 1 2 2
OOOOOOOO
zzzzzzzz
ooc^44cioi
uuuuuuuo
O O O CJ
15 15 15 IS
c o o o
CJ O 1> OJ
c c c c
2222
OOOO
i i . i
O O CM -
-------
Fresh leachate was used for each test since it was taken directly out of the sump at the low
elevation of the landfill or from the nearest underground storage tank.
The tests were of the falling head variety which measured the time of flight for a high head
of leachate to reach a predetermined lower value. The protocol for the test itself is included in the
ASTM D1987 test method. Calculations allowed for the determination of a "system" hydraulic
conductivity, or permeability value, which was in units of cm/sec. Note, however, that the
permeabiliy being measured was the permeability of the composite system including all
components which may retard flow. In this Appendix, the "permeability" value will be used on a
comparative basis, with the original value being the highest that the system could possibly achieve.
A-5 Results of Test Series "A"
As indicated in Table A-1, this test series consisted of a sand/geotextile/geonet/gravel cross
section with the biocide having been introduced into the geonet during its manufacture. The
biocide levels were at 0, 1, 2 and 4% and tests were conducted under both anaerobic and aerobic
conditions. ,
The system permeability results for test Series "A" are shown in Figures A-l and A-2
representing anaerobic and aerobic conditions, respectively. Separate curves are presented for the
control and each biocide level in the geonet. Comparison of these two figures indicates that there is
essentially no difference in the flow characteristics from the anaerobic to the aerobic state. Within
the curves of each figure a nominal improvement in permeability from using 2 or 4% biocide in the
geonet was evidenced at the conclusion of the 444 day test period. However, because the
improvement in flow is nominal at the end of testing and flow improvement is not evidenced
throughout the entire testing period, statistical variation in the data may influence the behavioral
trends. Our general feeling was that using biocide in the geonet was simply not logical since the
.flow rate in the geonet was relatively high. Thus the biocide probably did not have adequate
residence time to be effective.
A-6 Results of Test Series "Bj'
As indicated in Table A-l, Test Series "B" consisted of a sand/geotextile/gravel cross section
with the biocide having been introduced into the geotextile. The biocide levels were at 0, 2, and
4% and tests were conducted under anaerobic and aerobic conditions. The rational for this change
from the previous test series was that flow in the geotextile would be much lower than in the
geonet due to its significantly smaller void spaces. The decreased flow rate in the geotextile would
possibly allow for the biocide to have a greater contact time with the microorganisms in the leachate
and hence be more effective.
Figures A-3 and A-4 provide a comparison of anaerobic and aerobic conditions. A replicate
for the geotextile with 0% biocide was provided for each condition and the data was averaged for
plotting. A comparison of these figures reveals little difference in flow characteristics from
anaerobic to aerobic conditions. This same trend was seen previously with the geonet tests.
Generally, the geotextile with 4% biocide provided slightly higher flow rates with the exception of
the anaerobic conditions in which the geotextile clogged severely beyond 400 days. As with Series
94
-------
0.7
100
200 300
Time (dags)
400
500
Figure A-1 - Effect of Geonet Biocide Content on System
Permeability under Anaerobic Conditions
(Test Series "A")
95
-------
100
200 300
Time (days)
400
500
Figure A-2 - Effect of Geonet Biocide Content on System
Permeability under Aerobic Conditions
(Test Series "A")
96
-------
0.0
100
200 300
Time (days)
400
500
Figure A-3 - Effect of Geotextile Biocide Content on System
Permeability under Anaerobic Conditions
(Test Series "B")
97
-------
0.7
0.0
100
200 300
Time (days)
400
500
Figure A-4 - Effect of Geotextile Biocide Content on System
Permeability under Aerobic Conditions
(Test Series "B")
98
-------
"A" tests, statistical variation played a significant role. Two test specimens in Series "B", BO-AN1
and B4-A, were resin set and dissected to.visually determine the extent of clogging' In each
specimen it appeared as though the Ottawa sand clogged within the upper 25 to 50 mm (1.0 to 2.0
in.) of the specimen. The upper layer of soil was completely bonded together while the soil
beneath it and above the geotextile was loose. The biofilm apparently did not reach the level of the
geotextile indicating that either the biocide was too far from the biofilm itself or that the grain size
distribution of the sand was sufficiently small to create its own clogged layer.
A-7 Results of Test Series "C"
After evaluating the behavior of the flow column trends of Test Series "A" and "B", it was
apparent that a clearly defined flow improvement resulting from biocide activity was not being
observed. It was considered likely that the biofilm layer was occurring in the upper portion of the
sand, hence the biocide in the geonet (Series "A") and in the geotextile (Series "B") was too far
away from the clogged layer to be effective. Thus, it Became necessary to assemble an additional
series of 16 columns without overlying sand, which w;is the thrust of Series "C".
Series "C" columns consisted of a cross section of gravel/geotextile/gravel. The gravel was
25 to 37 mm (1.0 to 1.5 in.) in size and was negligible insofar as retarding flow was concerned
The geotextiles were treated with varying amounts of t iocide, from 0 to 8% (recall Table A-l) and
all columns were evaluated in the anaerobic condition This latter decision was made since there
was little difference in the anaerobic and aerobic flow rates in the previous tests and anaerobic
conditions were felt to better simulate landfill leachatf conditions. The geotextiles in this series
varied considerably. Those used were the following:
• nonwoven needle-punched with an opening size of 0.15 mm
• woven monofilament with a opening size of 0.21 mm
• woven monofilament with an opening size of 0.42 mm
The first part of Series "C" tests consisted of the nonwoven needle punched polypropylene
geotextile with 0,2,4 and 8% biocide within the fabric. A replicate set was constructed so that the
values used in graphing are the average of two data sets. The permeability behavior under varying
biocide contents is displayed in Figure A-5. There appeared to be little difference in flow at the
onset of testing, however, there was an improvement in flow with 8% biocide at the completion of
testing 121 days later. The use of 8% biocide, however, may affect the strength characteristics of
the geotextile and currently EPA has restricted biocide content of this type in other media to 4%.
As with the other test series, statistical scatter was significant.
In the second part of Series "C" testing, a different manufacturer's nonwoven needle
punched polypropylene geotextile with 0,2 and 4% bioc ide was used. A replicate was constructed
for the control, i.e., 0% biocide, and graphs were plotted using the average of the data sets To
compare the two different products, Figure A-6 is presented. In the first month of testing there
was little difference in flow rates. As the test progressed the 2 and 4% biocide geotextiles tended
to give better flow rates with a large improvement iiL flow at 121 days. However, statistical
scattering and the short duration of the testing were concerns with respect to the significance of the
99
-------
C1-0-AN
* C1-2-AN
• C1-4-AN
o C1-8-AN
0.0
50 75
Time (days)
100
125
Figure A-5 - Nonwoven Needle-Punched Effect of Geotextile
Biocide Content on System Permeability
(Test Series "C")
100
-------
3.0
0.0
25
50 75
Time (days)
100
125
Figure A-6 - Comparison of Two Npnwoven Needle-Punched
Geotextiles with Varying Biocide Content
(Test Series "C")
101
-------
data.
Two woven monofilament polypropylene geotextiles were used for the third part of the
Series "C" tests. The apparent opening sizes and other relevant test conditions are provided in
Table A-l. Each geotextile was tested with 2 and 4% biocide content. A control with 0% biocide,
was not used for this test set. To compare the effects of opening size on clogging, Figure A-7 was
prepared. From the graph it was seen that the larger opening size geotextile provided a measurable
increase in flow rate with the 4% biocide content giving better results in general. The smaller
opening size geotextile with 4% biocide excessively clogged two months into the test. The four
samples in this series were then epoxy resin set and dissected in the same manner as the Series B
specimens. While difficult to observe visually, it was obvious that the larger opening size (0.42
mm) allowed more epoxy to flow through the geotextile indicating that it was indeed providing
better flow than the 0.21 mm opening size geotextile.
A-8 Conclusions of the Bincide Study
From the results of the Series "A" tests (biocide in geonets) and Series "B" tests (biocide in
geotextiles) it was concluded that the location of the biocide vis-a-vis the initial formation of a
biofllm layer is critical. This conclusion was tentatively reached during the conducting of these
tests. It was confirmed at the termination of the 444 day tests after setting the test columns with
epoxy and cutting them apart. Clearly the biofilm layer was occurring at the top of the sand
column some 50 to 75 mm (2 to 3 in.) above the biocide treated geosynthetics, recall Figure A-5.
While there may have been some flow rate improvement due to high concentrations of biocide, it
was very subtile (at best) and was masked by the inherent scatter in the test data. There was
essentially no difference between flow rates in anaerobic versus aerobic conditions.
These findings led to Series "C" tests which contained no sand above the biocide treated
geotextile and forced the leachate to interface directly with the biocide. Rather than use a single
type of geotextile, three different types of geotextiles were utilized. They had opening sizes
varying from 0.15 mm (the nonwoven needle punched styles used in test Series "B"), to 0.21 mm
(a woven monofilament), to 0.42 mm (another woven monofilament). Quite clearly, the flow rates
through the largest opening size geotextiles, i.e. the 0.42 mm, were the highest. This suggested
that microorganisms (dead or alive) must be able to pass through the system. Whenever these
microorganisms reside on, or within, the small pores of a filter (either natural soil or a geotextile)
there was a possibility of partial, or even excessive, clogging.
102
-------
3.0
0.0
25
50 75
Time (days)
100
125
Figure A-7 - Comparison of Woven Monofilament Geotextile
Opening Size with Varying Biocide Content
(Test Series "C")
103
-------
APPENDIX "B"
PERMEABILITY RESULTS FOR "HIGH" AND
"INTERMEDIATE" FLOW RATES
(NOTE THAT THE "VERY HIGH" FLOW RATE
RESULTS ARE GIVEN IN THE BODY OF THE
REPORT IN SECTION 6)
104
-------
Ottawa Sand
Water
Leachate f
Leachale "P"
Leactiate '0*
20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-l - Permeability Response Curves for Ottawa Sand Permeated with 4 Fluids at
a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
Figure B-2 -
Concrete ganfl
Water
Leacfcate V
Leactiate "P-
Leachate 'D'
20 40 60 80 100 120 140 160 180 200
Timo (days)
105
-------
N 7 W
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-3 - Permeability Response Curves for Geotextile Filter N 7 W Permeated with
4 Fluids at a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
N 14 W
tio"
10'
»
Q) ..--
0. 10
Water
Leachate f
Leachate"P"
Leachate "D-
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-4 - Permeability Response Curves for Geotextile Filter N 14 W Permeated with
4 Fluids at a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
N 32 W
10'
* 10"5
o> fi
CL 10'6
h
T
M
,,„
rrm
-• —
•
-• —
ar«^
sj
i
Water
Leachate T"
Leachate "P*
Leachate *D*
4- 1
-B-B-«
V
T
\^
*•»-»
•»*-•
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-5 - Permeability Response Curves for Geotextile Filter N 32 W Permeated with
4 Fluids at a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
106
-------
H 4 NP
20 40 60 80 100 120 140 160 180 200
Time {days)
Figure B-6 - PermeabUity Curves for Geotextile Filter H 4 NPNW Permeated with
4 Fluids at a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
H 8 NPNW
Water
Leachale t"
Leachate -P-
Leachale -o*
-------
T 4 HBNW
Water
Leachate f
Leachate "P*
Leachate '
I
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-9 - Permeability Curves for Geotextile Filter T 4 HBNW Permeated with
4 Fluids at a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
o
0>
•2?
o
10'
10
-2
10'
= 1C'
I
10-
10*
P 6 NPNW
Water
Lsachate *L*
Leachate "P"
LaachateTT
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-10 - Penneability Curves for Geotextile Filter P 6 NPNW Permeated with
4 Fluids at a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
1Q8
-------
A 10 W
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-l 1 - Permeability Curves for Geotextile Filter A 10 W Permeated with 4 Fluids
at a Fast Flow Rate of 3.8 I/week (1.0 gal/week)
,0-3
CO 4
* 10'4
I 10"5
N 22 NW/W
•U4-
Water
Leachate f
Leachate "P-
Leachate '0*
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-12 - Permeability Curves for Geotextile Filter N 22 NW/W Permeated with
4 Fluids at a Fast How Rate of 3.8 I/week (1.0 gal/week)
109
-------
Ottawa Sand
20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-13 - Permeability Response Curves for Ottawa Sand Permeated with
4 Fluids at a Fast Flow Rate of 0.95 I/week (0.25 gal/week)
J3
S 10-
Concrete Sand
Water
Uachatet*
LeachateV
Loachate "D"
0 20 40 60 80 100 120 140 160 180 200
Time (cm/sec)
Figure B-14 - Permeability Response Curves for Concrete Sand Permeated with 4 Fluids
at a Fast How Rate of 0.95 I/week (0.25 gal/week)
110
-------
N 7 W
20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-15 - Permeability Response Curves for Geotextile Filter N 7 W Permeated with
4 Fluids at a Fast Flow Rate of 0.95 I/week (0.25 gal/week)
N 14
•«• 10' 1
Q>
CO
'"- ....-2
E 10 z!
a
10"3|
"5
2 ID'S-
E ;
| 10'6:
—a —
— • —
•
Watei
Leacf
Leacf
Leacti
tateV
ate'P-
iate *D*
—
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-16 - Permeability Response Curves for Geotextile Filter N 14 W Permeated with
4 Fluids at a Fast Flow Rate of 0.95 I/week (0.25 gal/week)
N 32 W
Water
LeachateT."
Leachatef-
Leachale *D*
20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-17 - Permeability Response Curves for Geotextile Filter N 32 W Permeated with
4 Fluids at a Fast Flow Rate of 0.95 I/week (0.25 gal/week)
111
-------
0
E
1C'
10'
H 4 NPNW
Water
Leachate f
Leachate "P*
Leachate *O*
0 20 40 60 80 100 120 140 160 180 Too
Time (days)
Figure B-18 - Permeability Curves for Geotextile Filter H 4 NPNW Permeated with
4 Fluids at a Fast Flow Rate of 0.95 I/week (0.25 gal/week)
H 8 NPNW
Water
Leachate V
Leachate f
Leachate "D*
20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-19 - Permeability Curves for Geotextile Filter H 8 NPNW Permeated with
4 Fluids at a Fast Flow Rate of 0.95 I/week (0.25 gal/week)
H 16 NPNW
Water
Leachate T'
Leachate "P"
Leachate "O*
20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-20 - Permeability Curves for Geotextile Filter H 16 NPNW Permeated with
4 Fluids at a Fast Flow Rate of 0.95 tfweek (0.25 gal/week)
112
-------
T 4 HBNW
Water
Leachate f
Leachate "P"
Leachate "D"
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-21 - Permeability Curves for Geotextile Filter T 4 HBNW Permeated with
4 Fluids at a Fast How Rate of 0.95 I/week (0.25 gal/week)
£10-4
I 10'5
| 10-6
P 6 NPNW
Water
Leachate "L°
Leachate "P*
Leachate TT
20 40 60 60 100 120 140 160 180 200
Time (days)
Figure B-22 - Permeability Curves for Geotextile Filter P 6 NPNW Permeated with
4 Fluids at a Fast Flow Rate of 0.95 I/week (0.25 gal/week)
113
-------
^^ «n * 1
A 10 W
O iu
0
Em * 2^
10
>.
ermeabil
y o t
a> •
r
M '—
^
=»
H^IHHHIH
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-23 - Permeability curves for Geotextile Filter A 10 W Permeated with 4 Fluids
at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week)
N 22 NWAV
trio0 •
g w1-
10*2-
S 10'3-
Z
S io-4-
E
® in'5-
OL 1U n
— »—
B
•
— ^ —
•
=s=J
Wats
Leac
Leac
Leac
*• ii -
T
hate-L
haleT*
hate'D
-P»S=
•
•
0 20 40 60 80 100 120 140 160 180 200
Time (days)
Figure B-24 - Permeability Curves for Geotextile Filter N 22 NW/W Permeated with
4 Fluids at a Intermediate Flow Rate of 0.95 I/week (0.25 gal/week)
114
-------
APPENDIX "C"
HELP MODEL PARAMETRIC ANALYSIS
C-l Overview
The HELP model which is generally used to follow and apportion the liquids flowing into
and through a solid waste landfill mass was briefly described in Section 7. It was utilized to
determine the site specific "k^" values for the four landfills that were exhumed. This data was
then used directly in the design method to substantiate the results of the field exhuming activities.
In addition to this direct use of the model, considerable effort also went into assessing the latest
version of the model insofar as a parametric analysis was concerned. The results of this ancillary
effort are presented in this appendix.
The purpose of the present effort was to perform a parametric evaluation of items that are
particularly important in estimating the leachate flow through a landfill and that need to be defined
as input data in the HELP model. The parametric study was performed to determine the effects of
these major design parameters on the amount of leachate collected at the bottom of a landfill and to
highlight the likely variations in leachate flow rates due to changes in the value of a specific
parameter.
The study examined the general effect of various parameters for two different landfill
configurations. In both cases, final closure of the landfill had not occurred. The first
configuration consisted of the waste material underlain by a lateral drainage layer (without a filter)
and then an impermeable geomembrane. This is the "no filter" scenario. The second
configuration assumed that a filter was placed between the waste material and the lateral drainage
layer. Again, an impermeable geomembrane was beneath the lateral drainage layer. This second
case is particularly important in estimating the required flow rate (permeability) of a filter layer;
either geotextile or natural soil material.
C-2 Site Description
The analysis was carried out for a landfill in Philadelphia, Pennsylvania. The reason
being that all four exhumed landfills described earlier were within a 320 km (200 mile) radius of
Philadelphia. However, to simulate different climatic conditions the investigation was extended to
include both a dry and a wet site. Phoenix, Arizona and Seattle, Washington were arbitrarily
chosen to simulate the dry and wet climates, respectively. For all three cities default climatologic
data for the year 1974 was used. Mean monthly temperatures and precipitation values contained in
the program for 1974 are given below. Note that the HELP model is generally used in standard
units, thus dual units will not be given in this section except for permeability which is in the
customary units of cm/sec.
115
-------
Philadelphia. Pennsylvania
Normal Mean Monthly Temperatures (F)
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
31.2 33.1 41.8 52.9 62.8 71.6 76.5 75.3 68.2 56.5 45.8 35.5
Average Monthly Precipitation Values (inches)
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
2.95 2.14 4.91 2.77 3.21 4.43 2.08 3.83 4.68 1.93 0.81 4.04
Average Annual Precipitation (inches) 37.78
Peak Daily Precipitation (inches) 1.7
Seattle. Washington
Normal Mean Monthly Temperatures (F)
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
39.1 42.8 44.2 48.7 55.0 60.2 64.8 64.1 60.0 52.5 44.8 41.0
Average Monthly Precipitation Values (inches)
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
7.76 3.98 5.83 2.37 1.34 1.24 1.51 0.01 0.2 1.97 8.17 8.33
Average Annual Precipitation (inches ) 4271
Peak Daily Precipitation (inches) 1.43
Phoenix. Arizona
Normal Mean Monthly Temperatures (F)
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
52.3 56.1 60.6 68.0 77.0 86.5 92.3 89.9 84,6 73.4 60.6 53.3
Average Monthly Precipitation Values (inches)
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
0.57 0.02 1.37 0.01 0.00 0.00 0.84 1.15 1.07 2.12 0.44 0.59
Average Annual Precipitation (inches) 8.18
Peak Daily Precipitation (inches) 1 !o4
C-3 Landfill Description
The landfill under consideration was assumed to be a 5 acre cell measuring 330 ft wide by
600 ft long as shown in Figure C-1. Two different leachate collection system configurations were
used; these are shown in Figures C-2(a) and C-2(b). Note that the filter in Figure C-2(b) is
identified as a filter/separator since it is serving in a dual role and is properly identified as such.
The leachate being transmitted within the lateral drainage layer was collected by a perforated feeder
116
-------
Perforated Pipe
Zone of influence of a single feeder pipe
Sump
330ft
T
100 ft
Figure C-l - Plan View of 5 Acre Landfill Cell Showing Perforated
Pipe Network for Leachate Removal
117
-------
Lateral Drainage Layer
_.— Perforated Pipe
2 % slope
••»
Geomembrane
(a) LANDFILL WITH NO GEOTEXTILE FILTER / SEPARATOR
0to 100ft
2ft
Geotextile Filter/Separator
Lateral Drainage Layer
Perforated Pipe
H = 0 to 100 ft
2ft
Geomembrane
(b) LANDFILL WITH A GEOTEXTILE FILTER / SEPARATOR
Figure C-2 - Idealized Cross Section of Leachate Collection and
Removal System Used in the Present Study
118
-------
pipe placed at the bottom of this layer, serving an area equal to 165 ft x 100 ft. The properties of
the different layers shown in each configuration are given in Table C-l.
Table C-l - Description and Soil Properties of the Different Layers
Case 1. Landfill with no geotextile filter/separator
Layer No. Soil Type* (USCS) Permeability (cm/sec)
1 Waste 0.0002
2 Soil 1, Gravel 1.0
3 Soil 21, Barrier IxlO'10
Height (in.)
0 to 1200
24
1
Layer Type"*
1
2
4
* Default soil characteristics given by the HELP model
l^,/ j •— — J f fc*«^*** pr V*. WSA.V**~*.t^.I.A J.U. V WA
Layer type 2 - Lateral drainage layer
Layer type 3 - Barrier soil layer
Layer type 4 - Barrier soil layer with an impermeable liner
Case 2. Landfill with a geotextile filter/separator
Layer No.
1
2
3
4
Soil Type* (USCS)
Waste
Soil 1, Geotextile
Soil 1, Gravel
Soil 21, Barrier
Permeability (cm/sec)
0.0002
0.1 to Ix 10-7
1.0
IxlO-io
Height (in.)
0 to 1200
0.5
24
1
Layer Type**
1
1
2
4
In the analysis it was assumed that the landfill had not reached its permitted height and was
not finally capped. If it were, it would have been necessary to define the fraction of infiltration
that passed through leaks in the geomembrane in the cover of the landfill as well as drainage off of
the surface of the geomembrane. In the present analysis these two factors were not necessary.
C-4 Landfill With No Geotextile Filter/Separator
The analysis was divided into two parts; a sensitivity analysis and a parametric study.
C-4.1 Sensitivity Analysis
The sensitivity of the HELP model to variations in the values of selected input parameters is
illustrated in this subsection. This will aid one to assess the influence of each of these parameters
in controlling the amount of leachate collected in the leachate collection system.
As this study was mainly a sensitivity analysis, it is carried out only for the Philadelphia,
Pennsylvania site. The parameters selected for the sensitivity analysis are given in Table C-2.
119
-------
Table C-2 - Parameters Selected for the Sensitivity Analysis
Parameter
Thickness of the lateral drainage layer (in.)
Slope at the base of the lateral drainage layer (%)
Evaporation depth (in.)
Runoff curve number
6
0.5
6
20
Range
to
to
to
to
36
5
18
100
As previously mentioned, this investigation was concerned with the evaluation of the effect
of different parameters on the amount of leachate gathered at the base of the landfill. Accordingly,
the value of the peak daily discharge from the lateral drainage layer (Q in units of fWday) was
chosen for the current comparison. Most of the sensitivity analyses were carried out for the no
waste condition (where the amount of liquid is a maximum) and for a 10-foot deep waste condition
(the typical thickness of the first lift of waste). The comparison carried out between these two
conditions clarified the role of the waste material in controlling the liquid balance as the presence of
waste resulted in greater retention, higher evaporation values, and lesser amounts of leachate
arriving at the underlying leachate collection system.
(a) Effect of the thickness of the lateral drainage layer
Figure C-3 illustrates the effect of the thickness of the lateral drainage layer, assuming a
permeability of 1.0 cm/sec, on the value of the peak daily discharge, Q. It can be seen that the
value of Q, for the no waste condition, was not affected by the thickness of the lateral drainage
layer for values ranging between 12 and 36 inches. However, for very small layer thicknesses the
value of Q was reduced. For example, for a 6 inch thick layer, Q was reduced by 12%. On the
other hand, when waste was added, the value of Q seemed to be unaffected by the variation in the
thickness of the lateral drainage layer.
(b) Effect of the slope at the base of the lateral drainage layer
To study the effect of the slope at the base of the drainage layer on the values of Q, slope
values ranging between 0.5% and 5.0 % were considered for the no waste and 10 ft waste
conditions. The results plotted in Figure C-4 indicate that in this range of slope values the amount
of peak daily discharge was not affected except for the no waste condition at a very small slope
value of 0.5 %, where Q was reduced by 13.5 %.
(c) Effect of the evaporation depth
In Figure C-5 the effect of the evaporation depth on the value of Q is presented. It is seen
that for the no waste condition, Q had a constant value for evaporation depths ranging between 9
and 15 inches. It should be mentioned here that the default value of the evaporation depth
suggested by the HELP model, for a landfill with a bare ground surface constructed in
Philadelphia; Pennsylvania, is 9 inches. For evaporation depths less than 9 inches Q was slightly
reduced. For example, at the evaporation depth of 6 inches Q was reduced by 10 %. On the other
hand, as the evaporation depth was increased from 15 to 18 inches, the value of Q increased by
120
-------
>, 2500
1
* 2000
"0
o 1500
TO
<5
"1 1000
b
>«
~ 500
(0 •")VJU
Q
^
8 o
t
<
,L
.
, V
No Waste
1 0 ft Was
1
te
6
36
12 24
Lateral Drainage Layer Thickness ( in )
Figure C-3 - Effect of Thickness of Lateral Drainage Layer on Peak
Daily Discharge (No Geotextile)
•^ 3000
•O
2500
,-S
2. 2000
0)
J 1500
«
b 1000
'5
Q 500
CO
CL 0
nr^
: * '
1 Q 1
> 1 .
1 B 1
•— * <
1
1
i — 1
i
«a — No Waste
• — 10 ft Waste
01 2 3 4 5 6
Slope at Base of Lateral Drainage Layer ( % )
Figure C-4 - Effect of the Slope at the Base of the Lateral Drainage Layer
on Peak Daily Discharge (No Geotextile)
121
-------
— 3000
•o
^ 2500
•tr
"o
— 2000
CD
P)
co 1500
o
Q 1000
_>»
'5
0 500
CO
0)
a 0
•
'
I 1
I . — =
-I
^*
JT1
a — NO
* ™ 10
3 6 9 12
Waste — ;
ft Waste
15 18 21
Evaporation Depth ( in )
Figure C-5 - Effect of Evaporation Depth on Peak Daily Discharge
(No Geotextile)
Q
>,
.
s
Q.
2500
2000
1500
1000
500
•
t
i 1
Q NoW
• 5flW
I
aste
aste
• — I Oft Waste
0 20 40 60 8(
3 10
0 12
Runoff Curve Number
Figure C-6 - Effect of Runoff Curve Number on Peak Daily Discharge
(No Geotextile) 6
122
-------
18%. When waste was added, the value of Q did not vary with the variation in the evaporation
depth for values between 6 and 18 inches.
(d) Effect of the runoff curve number
The runoff curve number is a number ranging between 10 and 100 and is used to partition
incoming rainfall or snowmelt between runoff and infiltration. It is required to specify the runoff
curve number at one stage in the program, however, if the landfill is open, the control is passed to
another subroutine where it is required to specify the fraction of the total potential runoff that
actually drains from the surface of the top layer. Accordingly, as this fraction was always given a
zero value in the present study, it was expected that the runoff curve number would have no effect
on the value of Q. This is clear from Figure C-6 where it is shown that Q is independent of the
runoff curve number and has a constant value depending on the waste height.
C-4.2 Parametric Study
This section reports an investigation carried out to clarify the importance of the parameters
given in Table C-3 in controlling the amount of leachate collected in the bottom of an open landfill.
Table C-3 - Parameters Selected for the "No Filter" Parametric Study
Parameter
Permeability of the lateral drainage layer (cm/sec)
Height of the waste material (ft)
Permeability of the waste material (cm/sec)
Symbol
K
H
Range
1.0 to 1 x 10-6
0.0 to 100
0.02 to 0.00002
(a) Effect of the permeability of the lateral drainage layer
Figure C-7 illustrates the relationship between the lateral drainage layer permeability, K,
and the peak daily discharge, Q, for different waste heights. For the no waste condition it can be
seen that Q decreased greatly with the decrease in K for values ranging between 1.0 and O.Q01
cm/sec. At lower permeabilities the rate of the decrease in .Q becomes much smaller. The
relationship between K and Q on a log-log scale could be divided into two straight lines, the slope
of the first line, representing K values greater than 0.02 cm/sec, was much smaller than the slope
of the second line which simulated K values less than 0.02 cm/sec.
When waste was placed above the drainage layer the value of Q was tremendously reduced
even for small waste heights. Clearly the leachate collection system for an unfilled solid waste cell
acts as a dewatering system until waste is placed. For a given waste height, the value of Q was
slightly affected by the variations in the permeability values. The reduction in Q occurs at K
values less than 1 x 1CM cm/sec.
(b) Effect of the height of the waste material
The decrease in the value of the peak daily discharge, Q, due to the increase in the waste
height, H, is plotted in Figure C-8 for landfills with lateral drainage layers of different
123
-------
CO
•
o
CO
O)
CO
o
Q
_>»
ra
Q
to
CL
No Waste
5 ft Waste
10 ft Waste
50 ft Waste
100 ft Waste
.000001 .00001 .0001 .001 .01 .1 1
Lateral Drainage Layer Permeability ( cm/sec )
— 10000 is
.000001 .00001 .0001 .001 .01 .1 1
Lateral Drainage Layer Permeability ( cm/sec )
10
Figure C-7 - Effect of Lateral Drainage Layer Permeability on Peak Daily
Discharge (No Geotextile)
124
-------
2500
CO
TJ
s 2000
oo K
O
•^•^
O>
O
W
b
.>,
'5
Q
CO
Q)
Q.
1500
1000
500
-10
Permeability of Lateral
Drainage Layer
K=1
K=0.1
K=0.01
K=0.001
100
Waste Height ( ft )
-------
permeabilities. The results indicated that the value of Q depended greatly on the height of the
waste, H. The permeability of the lateral drainage layer, K, was less significant. As soon as
waste was placed in the landfill, curves corresponding to different values of K coincided forming a
single curve. As shown in Figure C-8, for H values greater than zero, the relationship between
H and Q plotted on a semi-log scale can be represented by two linear portions. The first portion
corresponds to H values between 5 and 20 feet, while the second portion represents H values
between 20 and 100 feet and has a smaller inclination than the first portion. This indicated that the
rate of the decrease in Q was much greater at the early stages of waste disposal However as
waste accumulated in the landfill, the rate of the reduction in Q became much smaller.
(c) Effect of the permeability of the waste material
The effect of the permeability of the waste material on the amount of the peak daily
discharge.Q, is illustratedin Figure C-9 for different waste heights. As shown in TableC-3 the
value suggested by the default data for the waste permeability is 0.0002 cm/sec. The results
indicated that at small waste heights, e.g., 5 and 10 ft, the values of Q corresponding to waste
permeability values ranging between 0.002 and 0.00002 cm/sec were not significantly altered
For the 5 feet waste condition the variation in Q value was about 16.4%, whereas it was only
13% for the 10 feet of waste condition. However, if the value of the permeability of the waste
material was increased to 0.02 cm/sec, the corresponding values of Q for 5 ft and 10 ft waste
conditions were 71% and 51% higher than those calculated using the default values In the case
of waste heights of 50 to 100ft, the amounts of leachate corresponding to the default permeability
value were very small. In this case, even though the values of Q increased at a relatively high rate
with the increase in the waste permeability, the actual values still remained comparatively low
C-5 Landfill with a Geotextile Filter/Separatnr
A parametric study was performed to assess the variations in the computed leachate values
due to the placement of a geotextile filter/separator layer beneath the solid waste material and above
the lateral drainage layer as illustrated in Figure C-2
-------
5*
T>
0)
CO
u
a
.>.
'55
Q
CO
£
5 ft waste
10ft waste
50ft waste
100 ft waste
50
..00001
.0001 .001 .01
Waste Permeability ( cm/sec )
o
-------
The analysis was divided into two parts. The first part highlighted the effect of the different
parameters given in Table C-4 on the amount of leachate collectedat a landfill in Philadelphia
Pennsylvania. The second part of the investigation aimed at pointing out the variations in the
amounts of leachate that accumulated in different climates. Accordingly, the investigation was
extended to include both a wet and a dry site. Arbitrarily selected were Seattle, Washington and
Phoenix, Arizona, for wet and dry locations respectively. In this second part of the investigation
only the major parameters affecting the amount of leachate were taken into consideration.
C-5.1 Parametric Study for a Landfill in Philadelphia; Pennsylvania
The effect of the different parameters listed in Table C-4 are demonstrated in the following
subsections.
(a) Effect of the permeability of the geotextile layer
The relationships between the geotextile permeability, k, and peak daily discharge Q are
given in Figures C-10 through C-13 for different lateral drainage layer permeabilities and different
waste heights. In the analysis it was assumed that the geotextile layer was a vertical percolation
layer. For each of the lateral drainage layer permeabilities it was noticed that for the no waste
condition, Q had a constant value if the geotextile permeability, k, exceeded 1 x 10-4 cm/sec.
When the geotextile permeability was less than 1 x 1(H, the value of Q was greatly reduced with
any reduction in the value of k.
As waste was added, the value of Q decreased as the height of the waste increased. At any
given height of waste the value of Q was not affected by the decrease in the geotextile permeability
down to a value of 1 x 10-5 cm/sec. If ^ geotextUe permeability was reduced lower than this
value, for example due to progressive clogging, the amount of discharge decreased with the
decrease in the geotextile permeability. As the geotextile permeability decreased to a value of 1 x
10-7, the geotextile became excessively clogged and the amount of leachate passing through it
became extremely small, i.e., it then became a de-facto liner.
It is worth noting that the vertical flow submodel of the HELP program is based on
Darcy' s law which is given by the following equation:
Qv = k(dh/dl) (C.n
where
Qv = rate of vertical flow, in./day.
k = permeability of layer, in./day.
h = gravitational head, in.
1 = length in the direction of flow, in.
Free outfall was assumed from each layer such that dh/dl was set equal to unity (Schroeder
et al., [13]). This assumption was acceptable if the permeability of the profile was constant or
increased with depth. Because this assumption was not fulfilled at the interface between the waste
material (with a permeability of 0.0002 cm/sec) and the geotextile filter (with a permeability which
can be less than 0.0002 cm/sec) a different procedure was employed at the top interface of the
128
-------
-- 2500
t
0)
as
o
to
s
^
CO
CD
a.
No Waste
5 ft Waste
10 ft Waste
50 ft Waste
100 ft Waste
.0000001 .000001
.00001 .0001 .001 .01
Geotextile Permeability { cm/sec )
>«0000
•S
O 1000
CD
O)
co
o
CO
b
>.
CO
100
.0000001 .000001 .00001 .0001 .001 .01 1
Geotextile Permeability ( cm/sec )
Figure C-10 - Effect of Geotextile Permeability on Peak Daily
Discharge (K for Lateral Drainage Layer = 1 cm/sec)
129
-------
2500
No Waste
5 ft Waste
10 ft Waste
50 ft Waste
100 ft Waste
0
.0000001
.000001 .00001 .0001 .001 .01
Geotextile Permeability ( cm/sec )
NOOOO
.01
.0000001 .000001 .00001 .0001 .001
Geotextile Permeability ( cm/sec )
Figure C-l 1 - Effect of Geotextile Permeability on Peak Daily Discharge
(K for Lateral Drainage Layer = 0.1 cm/sec)
130
-------
(0
p.
1000
800
CD 600
CO
o
•J2 400
CD
Q.
No Waste
5 ft Waste
10 ft Waste
50 ft Waste
100 ft Waste
200
.0000001 .000001 .00001 .0001 .001 .01
Geotextile Permeability ( cm/sec )
CD
S>
(0
"I
5
10000
1000
100
1
.0000001 .000001 .00001 .0001
.001
Geotextile Permeability ( cm/sec )
Figure C-12 - Effect of Geotextile Permeability Peak on Daily Discharge
(K for Lateral Drainage Layer = 0.01 cm/sec)
131
-------
^ 200
CO
O
* 50
'55
Q
ca i
<•
100001 .00001 .0001
\. —
O—
1 —
.001
1
1
1
1
i
.01
i
a
• *>
• i
i
1
M
.1
i i
1
Geotexlile Permeability ( cm/sec )
— 1000
^ 100
0)
ra
I
Q
1
Q
|
Q. .0000001 .000001 .00001 .0001 .001 • .01
Geotextile Permeability ( cm/sec )
No Waste
5 ft Waste
10ft
.1
Figure C-13 -Effect of Geotextile Permeability on Daily Discharge
(K for Lateral Drainage Layer = 0.001 cm/sec)
132
-------
geotextile. In this case, it was felt to be more realistic to assume that a geotextile layer of
significantly reduced permeability was acting as a barrier layer. Accordingly, the total head above
it was calculated and the amount of liquid percolating through it is given by the equation :
Op =kp KTH+TO/TC] (C.2)
where
Qp = rate of percolation through the barrier layer.
kp = permeability of the barrier layer.
TH = total head in the profile above the barrier layer.
Tc = thickness of the barrier layer.
Figure C-14 illustrates the variation in the peak daily discharge for a 5 ft waste condition
assuming geotextile permeabilities ranging between 0.1 and 1x10-9 cm/sec. The permeability of
the lateral drainage layer was assigned a value of 1 cm/sec. The geotextile layer was considered to
be either a vertical percolation layer or a barrier layer. The results indicate that the values of the
peak daily discharge assuming the geotextile to be a barrier layer, were much higher than those
corresponding to the geotextile acting as a vertical percolating layer. However, for each case the
amount of Q corresponding to geotextile permeabilities higher than or equal to 1 x 10'5 cm/sec
were almost equal. The amount of leachate decreased with any further decrease in the geotextile
permeability and understandably reached extremely small values at a permeability of 1 x 10'8
cm/sec.
It is.worth pointing out that even though the amount of peak daily discharge was greatly
affected by the leachate accumulation on top of the geotextile layer, the average annual total amount
of leachate was only slightly altered. This is illustrated in Figure C-15 where the relationships
between the geotextile permeabilities and the average annual totals are plotted assuming the
geotextile to act either as a vertical drainage layer or as a barrier layer. The diagram indicates'that
the average annual totals are equal for both assumptions down to a geotextile permeability of 1 x
10-6 cin/sec. If the effect of the head on top of the geotextile layer was neglected, the calculated
amount of leachate appears to be greatly reduced at geotextile permeabilities less than 1 x 10-6
cm/sec. However, as the effect of liquid accumulation on top of the geotextile (due to its low
permeability) was taken into consideration, the average annual amount of leachate was almost the
same for different geotextile permeabilities down to a value of 1 x 10-7 cm/sec. As the geotextile
permeability was reduced to 1 x 10-9 cm/seCj the ieachate could hardly pass through and in this
case the geotextile acted as a barrier material rather than as a filter.
To highlight the effect of the accumulation of leachate on top of the geotextile, the
relationships between geotextile permeability and head on geotextile, peak daily discharge,'and
average annual totals are plotted for different waste heights in Figures C-16, C-17 and C-18,
respectively. The results show that waste heights up to 100 ft have no effect on the amount of
leachate for this situation. In this case the major parameter affecting the amount of leachate is the
133
-------
CO
<0
S>
m
o
w
CO
Q
(0
0)
Q.
Geot. as a vertical percolating layer
Geot. as a barrier
Geotextile Permeability ( cm/sec )
CO
T3
*
O
o>
E»
(0
Q
^>>
'55
Q
CO
&
Geot. as a vertical percolating layer
Geot. as a barrier
io-TTTo-7 10-6 10-5
Geotextile Permeability ( cm/sec )
Figure C-14 - Variation in Peak Daily Discharge Assuming Geotextile to be a
Vertical Percolating Layer or a Barrier Layer for 5 ft Waste
134
-------
•e- 20000
0>
15000
•X 10000
cu
c
0)
0)
5000
Geot. as a vertical percolating layer
Geot. as a barrier
10
^6
10 • 4 10 • ^ 10 2 i
Geotextile Permeability ( cm/sec )
10
Geot as a vertical percolating layer
Geot as a barrier
< riO'lu10"a 10'° 10"
Geotextile Permeability ( cm/sec )
Figure C-15 - Variation in Average Annual Totals Assuming Geotextile to be
a Vertical Percolating Layer or a Barrier Layer for 5 ft Waste
135
-------
X
fl)
o
0>
O
o
T3
CO
V
I
Waste height = 5, 10, 50 & 100 ft
1
-------
geotextile permeability. As the geotextile permeability became less than the waste permeability,
which was assumed to be 0.0002 cm/sec, leachate started to accumulate and the head on top of the
geotextile layer started to increase as indicated in Figure C-16. Also, as the geotextile permeability
was reduced to values lower than the waste permeability, the value of the peak daily discharge
started to decrease as shown in Figure C-17. However, Figure C-18 indicates thatthe variations
in the amount of the average annual rainfall above a geotextile permeability of 1 x 10'8 cm/sec are
small.
(b) Effect of the permeability of the lateral drainage layer
To illustrate the role of the value chosen for the lateral drainage layer permeability on the
amount of leachate, the no waste and the 5 feet waste conditions were taken into consideration.
The results for these two cases are plotted in Figures C-19 and C-20 respectively. Figure C-19
shows that for the no waste condition the value of the peak daily discharge, Q, was reduced as the
permeability of the lateral drainage layer decreased. It is worth pointing out that these results are
all based on the assumption that the geotextile was a vertical percolating layer.
When waste was added on top of the geotextile layer, even at very small thicknesses, the
value of the daily discharge was no longer dependent on the value of the soil permeability as
shown in Figure C-20. It became mainly dependent on the waste height and the geotextile
permeability being less or greater than IxlO-s cm/sec as previously mentioned.
(c) Effect of the height of the waste material
Figures C-21 and C-22 show the effect of the waste height, H , on the value of Q, for
different geotextile permeabilities, k, assuming the geotextile to be a vertical percolating layer. To
illustrate the effect of this parameter two conditions were considered, i.e., lateral drainage layer
permeabilities of 0.1 and 0.01 cm/sec. InFiguresC-21 and C-22 the pattern illustrating the effect
of the waste height on the variations in Q was the same, however, the value of Q was dependent
on the permeability of the lateral drainage layer.
For geotextile permeabilities higher than IxlO-s cm/seethe value of Q depended mainly on
the waste height. This is indicated by the three curves corresponding to geotextile permeabilities
equal to 0.1 ,0.001 and 0.00001 cm/sec as they coincide forming one curve at H values greater
than 5 ft. For geotextile permeability, k, equal to 1 x 10-6 cm/sec the value of Q was slightly less
than that corresponding to geotextiles with higher permeabilities. As the waste height reached
50 ft, the value of Q coincided with that corresponding to geotextiles with higher permeabilities .
When the geotextile permeability reached a value of 1 x 10-7 the waste height had no effect on the
value of Q. This was due to the fact that in this case the geotextile acted as if it were an
impermeable, barrier layer permitting only small amounts of leachate to percolate through it.
(d) Effect of the permeability of the waste material
The changes in the amount of peak daily discharge due to the increase in the waste
permeability are shown in Figure C-23 for the 5 ft waste condition. These results are also based
on the assumption that the geotextile was a vertical percolating layer. The diagram shows thatthe
137
-------
— 2500
^ 2000
o
ca
1
b
_>.
1
03
Q.
0.1
0.01
0.001
0.0001
0.00001
0.000001
.001 .0.1 .1 1
Lateral Drainage Layer Permeability ( cm/sec )
Figure C-19 -Effect of Lateral Drainage Layer and Geotextile Permeability
on Peak Daily Discharge (For No Waste Condition)
— 200
03
Q
•i
100
50
No Gaot. & Geot. Perm.-0.00
1.0001 & 0.00001
Geot. Permeability
0.1
0.01
0.001
0.0001
0.00001
0.000001
.001 .01 .1 1
Lateral Drainage Layer Permeability ( cm/sec )
Figure C-20 - Effect of Lateral Drainage Layer and Geotextile Permeability
on Peak Daily Discharge (For 5 ft Waste Condition)
138 .
-------
CO
o
CO
'5
Q
2500
p 2000
2. 1500
0)
O5
Geotextile Permeability
k=0.1
k=0.001
k=0.00001
k=0.000001
k=0.0000001
1000
10 20 , 30 40 50 60 70
Waste Height (ft )
80 90 100
10000
1000
o>
O)
Geotextile Permeability
k=0.1
k=0.001
k=0.00001
k=0.000001
k=0.0000001
Q
|
-1.0 0 10 20 30 40 50 60 70 80 90 100
Waste Height (ft )
Figure C-21 - Effect of Waste Height on Daily Discharge for Different Geotextile
Permeabilities (K for Lateral Drainage Layer = 0.1 cm/sec)
139
-------
— 1000
•g 900
0)
S>
03
'5
Q
•s
(D
Q.
Geotextile Permeability
k=0.1
k=0.001
k=0.00001
k=0.000001
k=0.0000001
100
Waste Height ( ft )
1000
Geotextile Permeability
k=0.1
k=0.001
k=0.00001
k=0.000001
k=0.0000001
-10 0 10 20 30 40 50 60 70 80 90 100
Waste Height ('ft )
Figure C-22 - Effect of Waste Height on Daily Discharge for Different Geotextile
Permeabilities (K for Lateral Drainage Layer = 0.01 cm/sec)
140
-------
300
O 200
Q
I
JC
ra
100
.0001
.001 .01
Waste Permeability ( cm/sec )
.1
Figure C-23 - Effect of Waste and Geotextile Permeability on Peak Daily Discharge
for 5 ft Waste (K for Lateral Drainage Layer = 1 cm/sec)
141
-------
value of Q was slightly altered due to the changes in the geotextile permeability up to a value of 1 x
10-5 cm/sec for the different waste permeabilities. If the waste permeability increased from
0.0002 to 0.002 cm/sec the increase in the value of Q did not exceed 15.5 %. However, if the
waste permeability was increased to a value of 0.02 cm/sec the increase in the values of Q ranged
from 41% to 71% depending on the geotextile permeability. For geotextile permeabilities greater
than 1 x 10-6 cm/sec the value of Q did not vary with the variation in the waste permeability.
C-5.2 Parametric Study With Geotextiles - Different Climates
So as to highlight the effect of different climatic conditions on the amount of peak daily
discharge, two cities other than Philadelphia were taken into consideration. These were Seattle,
Washington and Phoenix, Arizona which represent wet and dry climates, respectively. Mean
monthly temperatures and average monthly precipitation for the different cities were given in
Section C-2. To illustrate the effect of climatic variations only the major parameters affecting the
amount of leachate were considered. These are the geotextile permeability and the waste height as
shown previously in Table C-4.
(a) Effect of the permeability of the geotextile layer
Figure C-24 illustrates the effect of the geotextile permeability, k, on the amount of peak
daily discharge, Q, for Seattle, Washington. For the no waste condition it is clear that the value of
Q was almost constant for k values higher than IxlO-4 cm/sec. At a value of 1 x 10'5 cm/sec, a
reduction in the value of Q was noticed. As the value of k was reduced to 1 x 10'6 cm/sec the
amount of Q decreased at a very high rate to reach very small values.
As waste accumulated in the landfill, the value of Q decreased with the increase in the waste
height. At any specific waste height, the amount of Q was almost unaffected by the reduction in
geotextile permeability up to a value of 1 x 10-s cm/sec. However, as the geotextile permeability
reached a value of 1 x 10'6 cm/sec, the amount of leachate was greatly reduced. At a geotextile
permeability equal to 1 x 10'7 cm/sec, Q reached an extremely low value.
The relationship between k and Q for Phoenix, Arizona is given in Figure C-25. It is clear
that Q had significant values only before waste was placed in the landfill. In this case the
variations in Q values were of small magnitudes for geotextile permeabilities greater than 1 x 10'5
cm/sec. As the geotextile permeability decreased, the value ofQ diminished.
(b) Effect of the height of the waste material
Figures C-24 and C-25 also show the effect of the waste height, H , on the values of Q. It
is clear that the increase in the height of the waste was one of the major factors that resulted in a
large decrease in the amount of discharge. In Seattle, the values of Q even after the placement of
the waste were still measurable, however, in Phoenix the values of Q become negligible as soon as
waste was placed even at small heights. For geotextile permeabilities between 0.1 and 1 x 10'5
cm/sec the values of the peak daily discharge in Seattle, Washington corresponding to waste
heights of 5, 10, 50, and 100 ft were about 670, 390, 110 and 14 fWday, respectively. On the
142
-------
2500
CO
-
2000
O
CO
.
03
O)
O.
1500
;=; 1000
Q
.>•
8
No Waste
5 ft Waste
10 ft Waste
50ft Waste
100 ft Waste
500
.0000001 .000001 .00001 .0001 .001 .01
Geotextile Permeability ( cm/sec )
.0000001
.000001 .00001 .0001 .001 .01
Geotextile Permeability ( cm/sec )
.1
Figure C-24 - Effect of Geotextile Permeability on Daily Discharge at Seattle,
Washington (K for Lateral Drainage Layer = 1 cm/sec)
143
-------
03
T3
1=
o
o>
u
CO
b
J»>
"5
Q
-------
other hand, in Phoenix, Arizona, the values of the discharge corresponding to 5, 10, 50 and. 100
ft of waste were 4.5, 3.5, 3.3 and 3.3 ft 3/day, respectively .
(c) Effect of the climatic conditions
A comparison between the three different cities (Philadelphia, Seattle and Phoenix) is
shown in Figures C-26 to C-30. For the no waste condition, illustrated in Figure C-26, the
variation in Q values corresponding to the different climates were not extremely large, the values
of Q being 2390, 2070 and 1600 ft 3/day for Philadelphia, Seattle and Phoenix, respectively.
As waste was added, the variation in the values of the peak daily discharge between the
different locations became pronounced. The value of Q for Seattle was much greater than that for
Philadelphia and Phoenix. For example, in Philadelphia and Phoenix the amount of peak daily
discharge after applying 5 feet of waste was only 7 % and 0.3 % of that corresponding to the no
waste condition. On the other hand, in Seattle the amount of Q after adding 5 feet of waste was
32% of that obtained for the no waste condition. The curves plotted in Figures C-27 through C-
30 indicate that the pattern of the relationships between k and Q were almost the same for different
waste heights, even though the discharge magnitudes progressively decreased as waste heights
increased.
C-6 General Comments
This section of the report was concerned with the evaluation of the major parameters
affecting the leachate movement in a landfill as predicted using Version 3 of the "Hydrologic
Evaluation of Landfill Performance" (HELP) computer model. The study examined the general
effects of various input parameters, which must be identified by the user, on the amount of
leachate collected at the base of a .landfill in which the waste had not reached its full height and
final closure of the landfill had not occurred.
The study was divided into two main parts. The first part was a sensitivity analysis to
highlight the effect of some of the input parameters on the calculated output values. The second
part was a parametric study to evaluate the effect of the major parameters which are particularly
important in estimating the percolation through the landfill and hence, need to be well defined by
the user. To achieve this goal two landfill configurations were considered. In the first
configuration the waste was placed directly on top of the lateral drainage layer (i.e., the "no filler"
situation) which was underlain by an impermeable liner. In the second configuration it was
assumed that the waste was placed on top of a geotextile filter/separator which covered the lateral
drainage layer. Again, an impermeable liner was considered to fee placed beneath the drainage
layer.
The leachate movement in the landfill was calculated using the default climatologic data for
Philadelphia, Pennsylvania for the year 1974. To compare the variations in the amount of leachate
due to different climates, both a wet and a dry site were considered. These were arbitrarily
chosen to be Seattle, Washington and Phoenix, Arizona, respectively.
From the results obtained the following comments can be drawn:
145
-------
^ £3UU
"s: 2000
O
co 1500
o>
co
to 1000
Q
c5 500
O
CO
taM
^nr
H
JGC
^
/
/ ,
/
l>
i
1
?
^
*
//
i
u
1
V
m*
-••
T1 ^
i*-
-••
• I
O Seattle;Washington
— •— Philadelphia;Penn.
• Phoenix ;Arizona
Geotextile Permeability ( cm/sec )
Figure C-26 -Effect of Geotextile Permeability on Daily Discharge
For Different Cities (For No Waste Condition)
800
700
600
Q)
a
(B
O
9
a.
Seattla;Washington
Philadelphia;Pann.
Phoenix;Arizona
.0000001 .000001 .00001 .0001 .001 .01
Geotextile Permeability ( cm/sec )
.1
Figure C-27 -Effect of Geotextile Permeability on Daily Discharge for
Different Cities (For 5 ft Waste Condition)
146
-------
>. 400
,5 350
0 30°
^ 250
O>
| 200
I 150
03
* 5°
S n
•
^^
--'
/
1
^
/
/
^-'
^^
f
— Ch— Seattle;Washington
— • — Philadelphia;Penn.
— •— Phoenix-.Arizona
•
.0000001 .000001 .00001
.0001
.001
.01
.1
Geotextile Permeability ( cm/sec )
Figure C-28 -Effect of Geotextile Permeability on Daily Discharge
For Different Cities (For 10 ft Waste Condition)
03
o
_>.
'JO
a
XL
s
a.
Seattle;Washington
Philadelphia;Penn.
Phoenix ;Arizona
.0000001 .000001
.00001 .0001 .001 .01
Geotextile Permeability ( cm/sec )
Figure C-29 -Effect of Geotextile Permeability on Daily Discharge
For Different Cities (For 50 ft Waste Condition)
Soattla;Wasnington
Philadelphiapenn.
Phoenix Arizona
.0000001 .000001 .00001 .0001 .001 .01
Geotextile Permeability ( cm/sec )
.1
Figure C-30 -Effect of Geotextile Permeability on Daily Discharge
For Different Cities (For 100 ft Waste Condition)
147
-------
Case 1. Landfill with No Geotextile Filter / Separator above Drainage Laver
(1) Sensitivity Analysis
The different factors taken into consideration in this part of the study were thickness of the
lateral drainage layer, slope at the base of the lateral drainage layer, evaporation depth and runoff
curve number for a landfill located in Philadelphia, Pennsylvania considering a no waste condition
as well as a 10 ft waste condition.
(i) No waste condition
Before any waste material was deposited in the landfill the model indicated that the above
mentioned parameters have no or very small effect on the amount of leachate, especially within
ranges of practical values.
(a) Thickness of the lateral drainage layer
The results indicated that for thicknesses ranging between 1 and 3 ft, the amount of
leachate, Q, was not altered. However, if the thickness was reduced to 0.5 ft, Q was decreased by
12%. Federal regulations state that the thickness of the lateral drainage layer should not be less
than 1 ft.
(b) Slope at the base of lateral drainage layer
The results show that the calculated value for Q was not affected by slopes ranging
between 1% and 5%. On the other hand, if the slope was as small as 0.5% (as allowed in most
Federal regulations), Q was reduced by 13.5%. Some state regulations recommend thatthe slope
at the base of the lateral drainage layer should not be less than 2%.
(c) Evaporation depth
The default value for the evaporation depth was given as 9 inches. For an evaporation
depth of6in., Qisreducedby 10%. As the evaporation depth increased to 15 in., Qwas equal to
that corresponding to 9 in. If the evaporation depth increased to 18 in., Q increased by 18.5%.
(d) Runoff curve number
In the case of open landfills the runoff curve number had no effect on the values of Q.
In this case, a fraction of the total potential runoff that actually drained from the surface of the
waste layer should be specified by the user as data input.
(ii) 10 ft waste condition
When waste was placed in the landfill, the calculated amounts of leachate were not altered
with variations in either the thickness of the lateral drainage layer, the slope at its base or the
evaporation depth. As previously mentioned, in the case of open landfills the runoff curve number
was not meaningful.
(2) Parametric Study
The parameters chosen for this study were permeability of the lateral drainage layer, height
of the waste material and permeability of the waste material for accumulated waste.
(i) No waste condition
For the no waste condition the amount of Q decreased greatly with the decrease in the
permeability of the lateral drainage layer, K, from 1.0 to 0.001 cm/sec. When K values became
148
-------
equal to or less than 1 x 1(H cm/sec the corresponding Values of Q became very small.
(ii) Waste accumulation
(a) Permeability of the lateral drainage layer
At any specific waste height the decrease in K from 1.0 to 1 x 1O6 cm/sec resulted in
small variations in the corresponding values of Q.
(b) Height of the waste material
The height of the waste material, H, was one of the major parameters affecting the
amount of discharge. As waste was added the values of Q were reduced dramatically compared to
those corresponding to the no waste condition.
(c) Permeability of the waste material
The default value suggested for the permeability of compacted waste in the model is
0.0002 cm/sec. If the waste permeability increased with time up to 0.002 cm/sec, e.g., due to
waste degradation, or decreased to 0.00002 cm/sec, e.g., due to the deposition of the fine waste
particles into the voids between larger waste particles, the change in the value of Q was relatively
small. However, if the waste permeability increased to a value of 0.02 cm/sec the increase in Q
became pronounced.
Case 2. Landfill with a Geotextile Filter / Separator above Drainage Layer
(li) Parametric study for a landfill in Philadelphia, Pennsylvania
The parameters taken into consideration when assuming the presence of a geotextile filter/
separator on top of the lateral drainage layer were permeability of the geotextile layer, permeability
of the lateral drainage layer, height of the waste material and permeability of the waste material..
(ij No waste condition
(a) Permeability of the geotextile layer
The reduction in the geotextile permeability, k, from 0.1 to 1 x 10'4 cm/sec had almost
no effect on the value of Q. However, for k values less than or equal to IxlO'5 cm/sec the values
of Q became significantly low.
(b) Permeability of the lateral drainage layer
For a given geotextile permeability, higher than or equal to 1 x 10-4 cm/sec, the values
of Q decreased greatly with a decrease in the permeability of the lateral drainage layer. However,
for k values less than 1 x 10*4 cm/sec the decrease in the lateral drainage layer permeability did not
affect the values of Q.
(ii) Waste accumulation
(a) Permeability of the geotextile layer
As waste was placed in the landfill the variation in k values between 0.1 and 1 x 1Q-5
cm/sec did not affect the values of Q. However, as the geotextile permeability was reduced to
values lower than lx 1Q-5 cm/sec, e.g., due to sediment and/or biological clogging, the amount of
leachate collected at the bottom of the landfill started to diminish. As the geotextile permeability
149
-------
dropped lower than the solid waste permeability, the geotextile acted more as a barrier layer rather
than as a filter. The values of the peak daily discharge assuming the geotextile to be a barrier layer
were much higher than those corresponding to the geotextile acting as a vertical percolating layer.
However, the average annual totals corresponding to either of the assumptions were the same up
to a geotextile permeability of 1 x 10'6 cm/sec. This is due to the fact that the program allows
only downward flow in a barrier layer. Thus, any leachate accumulating on a barrier layer will
eventually percolate through it. Percolation rate depends upon the head of the liquid above the
base of the barrier layer. As the geotextile permeability reached 1 x 1O7 cm/sec the geotextile filter
was now excessively clogged, allowing only very small amounts of leachate to percolate through
it. Hence, the greatly reduced geotextile permeability became the dominant parameter affecting the
amount of leachate collected within the drainage layer.
(b) Permeability of the lateral drainage layer
For a given waste height and for geotextile permeabilities greater than or equal to 1 x
lO-5 cm/sec, the variations in the permeability of the lateral drainage layer had very small effect on
the amount of leachate. For geotextile permeabilities less than 1 x lO5 cm/sec the variations in the
lateral drainage layer permeability had no influence on the values of Q.
(c) Height of the waste material
For geotextile permeabilities, k, greater or equal to 1 x 10'5 cm/sec the values of Q
were only dependent on the height of the waste material. When k was less than 1 x 10-5 cm/sec
the values of Q became dependent on both the geotextile permeability as well as the waste height.
(d) Permeability of the waste material
At any specific waste height and for geotextile permeabilities higher than or equal to 1
x 10-5 cm/sec the amount of leachate increased with the increase in the waste permeability. Also,
the rate of the increase in Q increased with the increase in the waste permeability. For geotextile
permeabilities less than 1 x 10'5 cm/sec any increase in the waste permeability did not affect the
values of Q.
(2) Parametric study for landfills in different climates
As it would be expected, the amount of leachate collected in a landfill constructed at a wet
site was higher than that collected in a dry site. For different climates, the patterns of the
relationships between the amount of leachate, Q, and the various parameters discussed above were
primarily the same, even though the values of Q differ depending on the variations in the
precipitation and temperature at the different sites.
C-7 Conclusions of the HFJ.P Model Analysis
The conclusions of the parametric and sensitivity study of the HELP model presented in
this appendix follow.
(1) Regarding the situation with no filter layer above the lateral drainage layer
(i) For no waste condition:
150
-------
- The decrease in the permeability of the lateral drainage layer, K, from 0.1 to 1 x 10'3 cm/sec
caused a tremendous decrease in the amount of primary leachate. When K became equal to or less
than 1 x 10-4 cm/sec the amount of leachate became extremely small.
(ii) For accumulated waste:
- At any specific waste height the amount of leachate did not vary for K values ranging between 1
and 1 x 10-5 cm/sec. Any further decrease in the permeability caused a decrease in the amount of
leachate.
(2) Regarding the situation of having a geotextile or natural soil filter layer above the lateral
drainage layer:
(i) For no waste condition:
- The leachate flow decreased with the decrease in the lateral drainage layer permeability from 1 to
1 x 10-3 cm/sec.
- The amount of leachate was not altered with the reduction in the geotextile permeability , k, from
0.1 to 1 x 10-4 cm/sec. For k values equal to or less than 1 x lO'5 cm/sec, however, the leachate
flow became very low. Thus the permeability of a geotextile equal to 1 x 10-5 cm/sec j^^ me
leachate flow for the underlying drainage system. As the geotextile permeability became smaller,
leachate flow rates were proportionally decreased.
(ii) For accumulated waste:
- The decrease in the lateral drainage layer permeability from 1 to 1 x 10'3 cm/sec had a very small
effect on the amount of leachate produced.
- The variation in leachate flow for geotextile permeabilities, k, decreasing from 0.1 to 1 x 10'5
cm/sec was negligible. For k values less than 1 x 10-5 cm/sec ^ ^^ of leachate became v
-------
APPENDIX "D"
THE "NO FILTER" DESIGN SCENARIO
D-l Overview
It is believed by the authors of this report that by using a properly designed geotextile filter
over the entire footprint of a landfill or landfill cell an acceptable long-term strategy for a leachate
collection system in a modem landfill will result. Of course, the proper geotextile must be selected
and the recommended process is embodied in the design formulation that was presented in Section
8. However, some designers are considering the option of "no filter" at all. Indeed, some
facilities are constructed with no filter whatsoever. With the no filter strategy the waste is placed
directly on the drainage material which is either gravel or sand. The logic of this design strategy is
that the filter contains the smallest voids, which are most likely to experience excessive clogging.
If entirely removed, the likelihood of the larger sized drainage material clogging in comparison to
that of the filter is distinctly lessened. Furthermore, by using gravel as the drainage material
instead of sand (and even large size gravel at that) the likelihood becomes progressively smaller.
D-2 Design Considerations
The thought of using no filter in the leachate collection system leads some to question
problems of particle loss into the drainage material. In the no filter scenario, the permeability of
the drainage material must be greater than that of the waste but still be capable of retaining the
relatively large particles of the waste. In this regard, field pump tests conducted at several
different landfills by Oweis [16], suggest that the permeability of municipal solid waste is typically
from 1 x 10~2 to 1 x 10~4 cm/sec.
Unfortunately, this value is too low for rapid and efficient collection and removal of
leachate. Thus the permeability of the drainage soil is increased, often to a particle size resulting in
a permeability of 1.0 cm/sec. This is typical of a quarried gravel of 6 to 45 mm (0.25 to 1.5 in.)
particle size, e.g., AASHTO #57 stone. With such soils the situation of particle loss and the
accompanying clogging of the drainage gravel must be considered.
In the context of the design Equations 7 and 8 in Section 8, the DCF = 1.0 and the factor-
of-safety is formulated accordingly, i.e.,
FS =
reqd
where
= required permeability of the drainage layer
kailow = allowable permeability of the drainage material
The value of "k^" comes from the HELP model, and "k^low" is experimentally obtained. A
series of "no filter" experiments is described in this appendix.
152
-------
D-3 "No Filter" Experiments and Results
The experimental investigation presented in this appendix involved evaluating the
permeability of eight different drainage materials over an extended period of time to determine; (a)
if equilibrium flow rates exist, and (b) what are the respective values [17]. The test method used
for these tests (as in Section 6) was in accordance with ASTMD1987. The tests were conducted
using rigid wall permeameters of 100 mm (4.0 in.) diameter and flow rates were measured under
constant head conditions. The resulting permeability values were calculated using Darcy's formula
at incremental time periods throughout the duration of the tests. Each of the flow columns were
permeated with municipal solid waste leachate at a flow rate of 20,000 1/ha-day (2000 gal/acre-
day). This flow rate was chosen on the basis of a New York State survey of its landfills which
found that average leachate flow rates are in this approximate range, Phaneuf [11]. Thus these
tests are not accelerated flow rate tests in the context of the tests conducted in Section 6, recall
Table 13. All tests were conducted under saturated anaerobic conditions and gas production was
noted in all of the columns.
The experimental design consisted of eight flow columns. Figure D-l(a) shows a
photograph and Figure D-l(b) a schematic diagram of a 4-unit test setup. Four of the
permeameters contained different types of gravel underneath the waste and an, additional four
columns had different types of sand underneath the waste.
The solid waste, as well as the leachate, was obtained from a local municipal solid waste
landfill. The solid waste was excavated out of a region of the landfill that was saturated with
leachate for approximately two years. It should be noted that some of the larger pieces of the
waste were cut to fit inside the flow columns so that they could be compacted into the 100 mm
(4.0 in.) diameter flow permeameters.
Figure D-2 presents the particle size distribution curves of the eight drainage soils selected
for this study. They cover a wide range of particle sizes and gradations. Table D-1 is presented to
further characterize the various soils. Note that all soils are granular with relatively high
permeability values. The reason for selecting these particular soils was that in formulating the
draft EPA Leak Detection Rules, the requirement of a minimum 1.0 cm/sec permeability drainage
material was regularly discussed. The four gravels selected for this study meet this criterion. All
were poorly graded gravels (GP's) under the Unified Classification System. They conformed to
sizes between AASHTO#3 and #57 gravels. The mineral composition of the four soils v/ere
selected on the basis of the most prevalent types of quarried stone. Thus quartz, gneiss,
limestone, and shale gravels were included in the study.
153
-------
(b) Photograph of a Set of Four Long Term Flow Permeameters
T
300 mm Leachate Head
Municipal Solid Waste
Different Granular Soils
(b) Schematic Diagram of a Set of Four Long Term Flow Permeameters
Figure D-l - Long Term Flow Permeameters Used for the Testing of Municipal Solid
Waste Placed Directly Above Granular Sorts Without the Use of Sand
or Geotextile Filters
154
-------
S#7
S#5 | S#8
I S*6
U I tl«l«o*< I I I
tllv< OMHIMCI I INI ^ UlW.HO.IIO III.I NUMI1II
''itfw" I N'iTOJ:fririr-fy:!!:.T
-------
Table D-l - Granular Soil Characteristics for "No Filter" Permeability Study
Column
1
2
3
4
5
6
7
8
Soil
Type
gravel
gravel
gravel
gravel
sand
sand
sand
sand
Mineral
quartz
gneiss
limestone
shale
limestone
quartz
quartz
quartz
Shape
rounded
angular .
angular
angular
angular
angular
angular
rounded
dio
(mm)
17
27
5
8
0.4
0.4
0.2
0.6
dso
(mm)
32
33
16
17
7
6
.9
.9
^0
(mm)
34
36
17
17
9
8
1.4
1.0
CU
2.0 •
1.3
3.4
2.1
22.5
20.0
8.5
1.7
use
Class.
GP
GP
GP
GP
SW
SW
SW
SP
AASHTO
Classfication
3
3
57
57
2A
2A
10
n/a (Ottawa)
dj 0 = grain size of 10 percent finer by weight
ds o = grain size of 50 percent finer by weight
dgo = grain size of 60 percent finer by weight
CU = Coefficient of Uniformity = d6 Q^l, 0
USC = Unified Soil Classification
AASHTO = American Association of State Highways and Transportation Officials
When the Leak Detection Rules finally appeared in the Federal Register in the spring of
1992 the permeability requirement did not increase to 1.0 cm/sec but remained at the previously
regulated value of 0.01 cm/sec. Thus the need arose for an additional four columns using sand as
the drainage soils. Limestone and quartz sandy soils were selected. Both materials were classified
by the Unified Classification System as well graded sands (SW) and are designed as AASHTO
#2A materials. In addition to these two well graded sands, concrete sand and Ottawa sand were
selected to investigate additional alternatives. The concrete sand was classified as a SW in the
Unified Soil Classification system and #10 in the AASHTO system. The Ottawa sand was
classified as poorly graded sand (SP) in the Unified Soil Classification system and does not fall
under any designation in the AASHTO classification system.
Figures D-3(a) and (b) show the graphed results of the system permeability (which
includes both solid waste and granular drainage soil) for the eight permeameters that were
evaluated. The permeameters with gravel were maintained for nearly 1,000 days while the
permeameters with sand were maintained for nearly 800 days. The results of the permeameters
with gravel are shown in Figure D-3(a). They indicate that the solid waste was definitely the
controlling flow material. There was very little difference in system permeability due to particle
shape, gradation or mineral composition of the different gravels. In addition, the system
permeability remained relatively constant for all four gravel columns. The slight rise in
permeability after 200 days was explained as the start of a trend of piping via loss of fines. Since
the permeameters were not recharged with municipal solid waste this could have eventually led to
further increases in permeability.
156
-------
o
0>
to
E
o
CO
o>
E
10
-4,
Soil #1 GP Quartz
Soil #2 GP Gneis
Soil #3 GW Limestone
Soil #4 GW Shale
200 400 600
Time (days)
800
(a) Behavior Using Gravels Beneath Solid Waste
1000
U
CD
CO
E
o
JQ
CO
0)
E
10
-1
10
-2.
10
-3.
10
-4.
Soil #1 SW Limestone
Soil #2 SW Quartz
Soil #3 SW Concrete Sand
Soil #4 SP Ottawa Sand
— • 1- . ,— , ,
200 400 600
Time (days)
(b) Behavior Using Sands Beneath Solid Waste
800
Figure D-3 - Long Term Flow Results as per ASTM D1987
157
-------
The results of the permeameters with sand shown in Figure D-3(b) indicated that the solid
waste was again controlling the flow. The permeability values were in the 0.01 to 0.001 cm/sec
range which was lower than expected for the sands alone. The sands by themselves had a
permeability between 0.05 and 0.10 cm/sec. There was a trend in all of the permeameters with
sand that suggested that the system permeability was decreasing over time. This decrease was not
pronounced or sudden but it was definitely observable.
In addition to the information presented, the solid waste placed directly on drainage soils
was qualitatively analyzed. The flow columns used for this study were made from acrylic which
is transparent. In the case of the gravels it appeared that small amounts of fines were migrating
through the gravel over time. With the sand columns there was staining of the municipal solid
waste-to-sand interface. Additionally, a qualitative observation was made about the columns in
which limestone aggregate was utilized. In all cases where limestone was used it had not
agglomerated together after being permeated with leachate for 1000 days. Thus the limestone used
in these tests had not bound together via a reaction with the leachate. The limestone in this
experiment was of a low carbonate content (less than 5%) and did not react with the leachate to any
appreciable degree. There is concern over this issue [18] and additional research appears to be
warranted in this regard.
D-4 Conclusions of the "No Filter" Study
Landfills are complex and constantly changing bio-reactors. As such, a designer is faced
with the challenge of designing a leachate collection system for changing conditions. Each stage
of the evolution involves different processes, mechanisms, reactions and microorganisms. Some
stages are short-lived and others are nearly permanent. A designer needs to establish the worst-
case scenario, or critical condition, during this evolutionary process and design accordingly.
While the trends of the curves in Figure D-3 are encouraging, it must be cautioned that they are not
accelerated flow rate tests and the 1000 days test duration is a simulated case of only 3 years of
typical field flow rates.
Irrespective of the cautions raised in this section, the no filter strategy is being practiced at
several landfills. In all cases, the waste placed over the drainage media is considered "select
waste". As such, there are no large objects allowed in the select waste. This precaution must be
taken so as to eliminate penetration of the drainage material and possible puncture of the
underlying liner system. It is absolutely critical that the utmost care be taken in the initial
placement of the select waste. The highest level of construction quality control and construction
quality assurance must be exercised. For this reason it may be prudent to increase the thickness of
the drainage layer to compensate for waste intrusion and also for clogging in the upper 1/3 to 1/2
of the thickness of the layer.
158
------- |