PB87-229688
Investigation of Slurry Cutoff Wall
Design and Construction Methods for
Containing Hazardous Wastes
Cincinnati Univ., OH
Prepared for
Environmental Protection Agency, Cincinnati, OH
\ug 87
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NTIS
-------
-------
EPA/600/2-87/063
August 1987
INVESTIGATION OF SLURRY CUTOFF WALL
DESIGN AND CONSTRUCTION METHODS
FOR CONTAINING HAZARDOUS WASTES
by
Richard M. McCandless and Dr. Andrew Bodocsi
Department of Civil and Environmental Engineering
University of Cincinnati
Cincinnati, Ohio 45221
EPA Contract No. 68-03-3210
Work Assignment #07
Technical Project Monitor
Naomi P. Barkley
Land Pollution Control Division
Hazardous Waste Engineering Research Laboratory
Cincinnati, Ohio 45268
HAZARDOUS WASTE ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA/600/2-87/063
2.
3. RECIPIENT'S ACCESSION
PB87 22
SMS
4, TITLE AJJD SU.BTITLE
Investigation of Slurry Cutoff Wall Design and
Construction Methods for Containing Hazardous Wastes
5. REPORT DATE
Auqust 1987
6. PERFORMING ORGANIZATION CODE
7R*charc]
-------
NOTICE
The information in this document has been funded wholly or in part by
the United States Environment Protection Agency under Contract 68-03-3210 to
the Department of Civil & Environmental Engineering, University of Cincinnati.
It has been subject 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.
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and governmental concern about the nation's environment and its effect
on the health and welfare of the American people. The complexity of the
environment and the interplay among its components require a concentrated
and integrated attack upon environmental problems.
The first step in seeking environmental solutions is research and
development to define the problem, measure its impact, and project possible
remedies. Research and development is carried out continually by both indus-
try and governmental agencies concerned with improving the environment. Much
key research and development is handled by EPA's Hazardous Waste Engineering
Research Laboratory. The laboratory develops new and improved technologies
and systems to restore contaminated sites to usefulness; and to promote waste
reduction and recycling. This publication is one of the products of that
research—a vital communications link between the research and the user
community.
This document describes basic geotechnical research into the effective-
ness and reliability of slurry wall barriers using specialized experimental
equipment. It will be useful to engineering and earth science professionals
currently involved in the design or permitting of such containment systems.
For information, please contact the Land Pollution Control Division of
the Hazardous Waste Engineering Research Laboratory.
Thomas R. Hauser, Director
Hazardous Waste Engineering Research Laboratory
-------
ABSTRACT
Specific technical design standards for soil-bentonite slurry trench
cutoff walls used to isolate hazardous wastes have not been established. A
review of current design and construction methods was performed for the
purposes of summarizing current engineering practice, identifying areas of
technical debate, and initiating necessary research to promote the development
of rational standards. The review of current methods was followed by labora-
tory studies using experimental test equipment to study selected components
of model cutoff walls. An instrumented slurry test column was developed and
used to investigate the hydraulic characteristics and importance of bentonite
slurry seals formed on the walls of the cutoff trench during construction.
Also, a slurry wall tank capable of accomodating a 508 mm (20 inch) diameter,
101.6 mm (4 inch) thick circular cutoff wall was used to evaluate the effects
of overburden pressure (vertical consolidation) and hydraulic gradient (hori-
zontal consolidation), and to evaluate the potential for self-remediation of
hydraulic defects ("windows" through the barrier) via in situ consolidation
of the soil-bentonite backfill.
The testing of slurry seals in the column involved the penetration of a
5% bentonite:water slurry into two different sands, the formation of a dif-
ferent type of slurry seal in each case, and the measurement of their hydrau-
lic conductivities based upon the time-rate of flow and the measurement of
internal pore pressure conditions. The distinction between the two types of
seals generally relates to the degree of filtration of hydrated bentonite
particles during slurry penetration. A clean fine sand was used to investi-
gate seals formed by the surface filtration mechanism which produces a thin
dense, surface "filter cake". A medium to coarse sand was used to study the
characteristics of deep slurry penetration and the formation of a rheological
blockage involving limited or no surface filtration of bentonite particles.
Seals representing these two filtration extremes were found to be significant-
ly different in computed breakthrough time for an idealized cutoff wall
(i.e., the time needed for the water to flow from one side of the wall to the
o.ther). Based upon their measured hydraulic conductivities, the breakthrough
time in the medium to coarse sand (rheological blockage seal) was three times
greater than that computed for the fine sand (surface filtration seal, or
"filter cake" ). Ignoring the contribution of the soil-bentonite backfill
itself, however, the effectiveness of the two types of seals considered alone
were not significant and not greatly different, both types having breakthrough
times on the order of two weeks or less. It was therefore concluded that the
effectiveness of a soilbentonite slurry cutoff wall is a function of the
integrity of the backfill alone, that slurry seals cannot be relied upon to
offset the effects of latent defects in the backfill, and that the current
practice of disregarding the slurry seal in cutoff wall design should not be
changed.
IV
-------
Three model cutoff walls were constructed and tested in the slurry wall
tank. Testing of the first wall principally served to identify needed modifi-
cations in equipment and test procedures. The second model wall was permeated
with water under three hydraulic gradients for each of three different ver-
tical surcharge pressures. The average equilibrium hydraulic conductivity of
the model was measured under each set of conditions. The sequential applica-
tion of surcharge and hydraulic flow pressures was interrupted on two occa-
sions, however, by hydrofracture of the model. The exact nature and extent
of damage due to hydrofracture could not be established, and the relative
effects of overburden pressure (vertical consolidation) and hydraulic gradient
(horizontal consolidation) could not be clearly separated. The tests served,
however, to demonstrate that both overburden pressure and hydraulic gradient
have significant and comparable effects on reducing the average conductivity
of the wall. Moreover, water content, unit weight, and vane shear strength
data measured on samples of the soil-bentonite backfill after the test indi-
cated that effective overburden stress decreased somewhat with increasing
depth in the model, most likely due to friction between the backfill and sand
in which the model was constructed. This suggests that in situ consolidation
of backfill in real site walls may be somewhat mitigated by sidewall friction
and points to the need for additional testing at pilot-scale where, because
of the thickness of the wall, the side friction effects would be less pro-
nounced.
The third model wall was intentionally breached by two small slot-like
"windows" representing small pockets of entrapped bentonite slurry in the
backfill immediately after construction. By incrementing surcharge pressure
it was possible to "heal" the windows as evidenced by a return to the prede-
termined baseline hydraulic conductivity of the wall. This suggests that in
situ consolidation of the backfill may serve to eliminate hydraulic defects
in the form of pockets of slurry or micro-cracks related to chemical degrada-
tion. The effective depth of elimination via in situ consolidation cannot be
determined however until the state-of-stress in a soil-bentonite slurry wall
is explicitly known.
Given the primary importance of the backfill itself (as concluded from
testing of slurry seals), the unknown character and frequency of hydraulic
defects in a typical cutoff wall, and the unknown distribution of effective
stress with depth, it is concluded that pilot-scale research utilizing actual
construction equipment and methods is needed.
-------
CONTENTS
Foreword i i i
Abstract iv
Figures viii
Tables xi
Conversion of Units xii
Acknowledgement xiii
1. Introduction 1
2. Conclusions 5
Slurry Seals 5
Model Cutoff Walls 6
3. Recommendations . -.8
Research Needs .' • .. . 8
4. Summary of Current Methods /.' . .10
Design ,'.« . -.10
Specifications .14
Construction 19
QC 20
Performance Monitoring 23
5. Slurry Seals 25
Equipment and Procedures 28
Slurry Test Column 28
Small-Scale Tests 34
Results and Discussion 36
Slurry Test Column 36
Small-Seale Tests 44
Comparative Breakthrough Analysis 47
6. Model Cutoff Walls 51
Equipment Design and Preliminary Testing 54
Methods and Materials 62
Results and Discussion 75
In Situ Consolidation and Hydraulic Conductivity. . . .75
Window Closure 101
7. Quality Assurance 107
References 112
Bibliography 113
Appendices .
A. Procedural Outline for the Operation of the_Slurry Test Column. 117
B. Procedural Outline for the Operation of the'Slurry Wall Tank. . 122
C. Routine Geotechnical Test Results 127
D. Standard Laboratory Procedures (SIP'S). . . 140
Preceding page blank
VI 1
-------
FIGURES
Number Page
1. Grain size distribution of sands used in the investigation
of slurry seals in the slurry test column 27
2. Schematic of the slurry test column system 29
3. Schematic of the basal outflow assembly used in
the slurry test column 29
4. Typical baseline (no slurry seal) pore pressure distributions for
fine (+200) and medium to coarse (+40) sands used in this study..31
5. Typical initial pore pressure distributions after formation
of slurry seals in the a) +40 and b) +200 sands 32
6. Graphical construction used to determine effective seal length
(Ls) and pressure head loss ( Ps)'across the seal for
a) +40 and b) +200 sands 34
7. Schematic of "short" column used to evaluate long-term stability
of surface filtration seals 35
8. Measured initial pore pressure distributions for slurry seals
formed under a driving pressure of 5.0 psi in:
a) +40 sand and b) +200 sand 39
9. Measured pore pressure distribution for rheological blockage
seal in +40 sand after 233 hours (Test 7, Run 6) 40
10. Initial pore pressure distribution for the surface filtration
seal in unsaturated +200 sand (Test 2, Run 1) 41
11. Initial pore pressure distributions measured for surface
filtration seals on the +200 sand under net driving pressure of
a) 5.0 psi and b) 10.0 psi 43
12. Measured volumetric flow rates for surface filtration seals
on a) +60 Ottawa sand and b) replicate sample 45
12. Measured volumetric flow rates for surface filtration seals
on c) +80 Ottawa and d) +100 Ottawa sands 46
vi i i
-------
Number Page
12. Measured volumetric flow rates for surface filtration seals
on +100 Ottawa sand for seal formation/permeation
pressure of 10.0 psi 47
13. Idealized conditions after construction of cutoff walls
in the +40 and +200 sands 48
14. Schematic of the slurry tank system 54
15. Structural components of the slurry wall tank and reaction frame....56
16. Construction details of permeant reservoir
for the slurry wall tank 58
17. Detail of upper surface membrane and hydraulic cutoff for
a) Test Sequence 1 and b) Test Sequence 2;
c) bottom hydraulic cutoff 59
18a. Front elevation of slurry wall tank control panel 60
18b. Side elevation of slurry wall tank control panel 61
19. Loading plan for slurry wall tank Test Sequence 1 65
20. Grain size distribution of soils used in Test Sequence 1 ...68
21. Construction detail of soi 1-bentonite placement trench 70
22. Loading plan for slurry wall frame Test Sequence 2 72
23. Grain size distribution of soils used in Test Sequence 2 73
24. Hydraulic conductivity results for Test Sequence 1 76
25. Post Sequence 1 sampling plan and unit weight test results 77
26. Post Sequence 1 unit weight and water content data as a function
of depth in the model cutoff wall 78
27. Composite of hydraulic conductivity results for Test Sequence 2 81
28. Chronological summary of Sequence 2 testing 82
29. Hydraulic conductivity results for test 2(a) 83
30. Hydraulic conductivity results for test 2(b) 85
31. Hydraulic conductivity results for test 2(d) 86
32. Hydraulic conductivity results for test 2(e) 87
ix
-------
Number Page
33. Hydraulic conductivity results for test 2(f) 89
34. Hydraulic conductivity results for tests 2(g') and 2(g) 90
35. Procedure used to project equilibrium hydraulic conductivities
for tests 2(f) and 2(g) 92
36. Sequence 2 equilibrium hydraulic conductivities as a function
of effective overburden pressure 94
37. Sequence 2 equilibrium hydraulic conductivities as a function
of combined effective overburden (vertical) and hydraulic
(net horizontal) pressure 95
38. Hydraulic conductivity of Sequence 2 soil-bentonite as measured
in rigid-wall (compaction mold) permeameter 96
39. Hydraulic conductivity of Sequence 2 soil-bentonite as measured
in rigid-wall (compaction mold) permeameter 96
40. Post Sequence 2 sampling plan and unit weight test results 98
41. Pos't Sequence 2 unit weight and water content data for soil-
bentonite samples as a function of depth in the model wall 99
42. Post Sequence 2 vane shear strength and water content data
as a function of depth in the soil-bentonite wall ....100
43. Finite element mesh for soil-bentonite slump simulation
and predicted and measured slump mass shapes 103
44. Baseline and window closing hydraulic conductivity results
for test 3(a) 104
45. Details of slot window configuration for test 3(a) 105
46. Percent error in measured hydraulic conductivity in the slurry
wall tank as a function of time interval between volume flow
measurements 109
47. Percent error in measured hydraulic conductivity in the slurry
test column as a function of time interval between volume
fl ow measurements 110
-------
TABLES
Number Page
1. SOURCES OF UNPUBLISHED INFORMATION ON SLURRY WALL TECHNOLOGY 11
2. MEASURED PROPERTIES OF 5% ORDINARY (unaltered) BENTONITE SLURRY 26
3. MAJOR PROOF-TEST MODIFICATIONS - SLURRY TEST COLUMN 30
4. TEST CONDITIONS AND RESULTS OF THE SLURRY TEST COLUMN 37
5. CONDITIONS FOR SMALL-SCALE TESTS 44
6. TEST CONDITIONS FOR TEST SEQUENCES 1 AND 2 52
7. SUMMARY OF PROJECT TESTING - SLURRY WALL TANK 53
8. BASELINE CHARACTERIZATION AND POST-TANK EVALUATION TESTS 53
9. INITIAL EXPERIMENTAL CONFIGURATION AND SUBSEQUENT MODIFICATIONS
TO THE SLURRY WALL TANK 66
10. MEASURED PROPERTIES OF 5% ORDINARY (unaltered)
BENTONITE SLURRY - BATCH 1 69
11. MEASURED PROPERTIES OF 5% ORDINARY (unaltered)
BENTONITE SLURRY - BATCH 2 71
12. FINES CONTENT OF SAND SAMPLES
Sequence 1 Post-Test Evaluation 79
XI
-------
CONVERSION OF UNITS
U.S. customary units are used as the primary basis of measure in most
figures contained in this report. Conversions to other systems of units of
measure will be facilitated by the following table.
Conversion Factors - Metric to English
To Obtain Multiply By
Inches Centimeters 0.39370
•Feet Meters 3.28084
Yards Meters 1.09361
Miles Kilometers 0.62137
Ounces Grains 3.52740 x 10"2
Pounds Kilograms 2.20462
Gallons Liters 0.26417
Fluid ounces Milliliters (cc) 3.38140 x 10'2
Square inches Square centimeters 0.15500
Square feet Square meters 10.76391
Square yards Square meters 1.19599
Cubic inches Milliliters (cc) 6.10237 x 1Q-2
Cubic feet Cubic meters ' 35.31466
Cubic yards Cubic meters 1.30795
Conversion Factors - English to Metric
To Obtain Multiply By
Microns Mils 25.4
Centimeters Inches 2.54
Meters Feet 0.3048
Meters Yards 0.9144
Kilometers Miles 1.60934
Grams Ounces 28.34952
Kilograms Pounds 0.45359
Liters Gallons 3.78541
Milliliters (cc) Fluids ounces 29.57353
Square centimeters Square inches 6.4516
Square meters Square feet 0.09290
Square meters Square yards 0.83613
Milliliters (cc) Cubic inches 16.38706
Cubic meters Cubic feet 2.83168 x 10~2
Cubic meters Cubic yards 0.76455
xi i
-------
ACKNOWLEDGEMENT
This document was prepared by the University of Cincinnati for the United
States Environmental Protection Agency's Containment Branch within the Land
Pollution Control Division of the Hazardous Waste Engineering Research
Laboratory (HWERL) in fulfillment of contract No. 68-03-3210, work assignment
No. 07. The EPA Technical Project Monitor for this work assignment is Ms.
Naomi P. Barkley. Co-principal investigators from the University of Cincin-
of Civil and Environmental Engineering are Richard M.
Andrew Bodocsi. Major contributors include graduate
(Steve) Lin and Douglas Keller. Co-op students Thomas
David Klyce and Steven Liatti assisted in many aspects of
the project. The technical contributions of Mr. Gerard Roberto and Dr. Frank
Weisgerber are appreciated, as is the quality assurance/safety support of
Martha E. Lambert.
The technical contributions of the following individuals are also recog-
nized:
nati's Department
McCandless and Dr.
students Jong Jen
Crawford, Ali Kerr,
Matthew J. Barvenik
Bruce S. Beattie
Richard M. Burke
Nicholas J. Cavalli
Paul Dudko
Jeffrey C. Evans
Kurt J. Guter
Don Hentz
Christopher Jepson
Robert Kingsbury
Richard S. Ladd
F. Barry Newman
Christopher Ryan
Glenn D. Schwartz
Geoffrey Shallard
Enzo Zoratto
GZA (Goldberg, Zoino & Associates)
Federal Bentonite
IT Corporation
ICOS Corporation of America
Woodward-Clyde Consultants
Bucknell University
Granger Land Development Co.
Federal Bentonite
American Colloid Co.
American Colloid Co.
Woodward-Clyde Consultants
GAI Consultants, Inc.
Geo-Con, Inc.
IT Corporation
IT Corporation
IT Corporation
Appreciation is also extended to the
federal, state and industry organizations
related to this report.
numerous other individuals from
who were contacted on matters
XI 1 1
-------
SECTION 1
INTRODUCTION
Slurry trench cutoff walls were first used as seepage control barriers
in the United States in the early 1940's. Since that time, their use has
become more widespread and now includes application as hydraulic barriers to
control the movement of contaminated groundwater from hazardous waste disposal
sites. Specific technical design standards for slurry trench soil-bentonite
cutoff walls have not been established. Each application is unique and
requires site-specific engineering evaluation. Nevertheless, the current
state-of-the-art involves fundamental concepts, performance criteria, and
methods common to all applications. The purposes of this project are three-
fold:
• to compile information on current design and construction methods
• to identify specific research needs to promote the development of
rational design and construction standards
• to perform initial research in selected areas of need
Th'e project has a phased scope, the first phase of which involved review
of published literature on slurry wall technology, interviews with owners,
engineering consultants and construction contractors, and a general assessment
of methods and research needs. Based upon these findings, two subsequent
research phases emphasized laboratory .nodel studies of slurry rfall barriers.
The overall performance of a soil-bentonite cutoff wall as a groundwater
pollution control barrier is influenced by numerous factors in each of the
following general categories:
1) site-specific geological and hydrological conditions including the
type and concentration of the chemical or leachate to which the
barrier is exposed
2) construction-related variables such as the integrity of the key into
an aquiclude at the base of the wall, the presence or absence of
significant as-built defects which function as "windows" (pockets of
entrapped slurry or soil fro.n the walls of the cutoff .trench)
3) the homogeneity, hydraulic conductivity, and long-term integrity of
the soil-bentonite backfill itself
4) variable and sometimes subjective methods in measuring the various
1
-------
4) engineering properties which constitute the basis for design (dif-
ferent methods of measuring hydraulic conductivity, for example)
Clearly, the host of interacting variables which may affect the design
and actual performance of a soil-bentonite cutoff wall are highly situation-
dependent. Moreover, the list of possible research topics is almost as long
as the inferred list of variables. For these reasons it was necessary to focus
research efforts on areas of general interest (common to most applications)
and highest potential significance.
For example, a civil works construction guide specification for soil-
bentonite slurry trench cutoffs recently released by the U. S. Army Corps of
Engineers^) states that:
"The actual permeability of the slurry trench is dependent on both
the filter cake which forms on the sides of the trench and the
soil-bentonite backfill. The contributions of both are dependent
on the relative permeability and thickness of the two materials.
For design purposes, however, it is recommended that the perme-
ability of the slurry trench be based only on the soil-bentonite
backfill material."
This design philosophy reflects the fact that the hydraulic conductivity
of a bentonite slurry seal or "filter cake" (in any of its several forms)
generally is not established, not to mention questions of long-term integrity
under chemical or leachate attack. Information which would warrant the con-
sideration of the filter cake, or construction techniques which would enhance
its contribution, would be highly useful.
Another area of special interest deals with the question of in situ
consolidation of soil-bentonite backfill after placement. If significant
consolidation does occur after construction, the average hydraulic conduc-
tivity of the barrier may be lower than that suggested by the results of
laboratory tests on relatively unconsolidated samples, and the finished
barrier may also have an inherent ability to eliminate minor construction
defects (via consolidation) given sufficient time. Field observations,
however, suggest that limited in situ consolidation occurs (evidenced by
lack of subsidence of the backfill surface) which in turn suggests that
friction between the backfill and the soils comprising the walls of the
cutoff trench and accompanying arching may be significant limiting factors.
The bulk of this report describes the results of basic research into
a few of the questions just discussed. The research objectives were based
upon a general consensus of need within the technical community and involved
specific testing to determine the following for bentonite slurry seals:
* The depth of penetration of slurry or filtered slurry into typical
granular soils
• The hydraulic conductivity of various types of seals ("filter cakes")
derived from slurry penetration and slurry filtration during penetration
into typical granular soils
-------
• The stability of the seals (described above) after initial development
Research detailed herein also focused on the characteristics of soil-
bentonite backfill alone and in combination with slurry seals. Much of the
work involved testing' of model slurry walls using specialized testing equip-
ment (described later) to investigate:
• in situ consolidation and the effect of surcharge loading and hydraulic
gradient on soil-bentonite hydraulic conductivity
• The feasibility of "window" closure within a soil-bentonite wall due to
overburden consolidation pressures.
To facilitate investigation of these experimental objectives, two experi-
mental systems were designed, fabricated, and performance tested as part of
this study.
An instrumented slurry test column (hereinafter referred to as "column")
was developed to study various bentonite slurry seals formed on the walls of
the cutoff trench during construction. The system consists of an acrylic
test column equipped with probes to measure in situ pore pressure after slurry
penetration and formation of a slurry seal in different sands. Pressure
monitoring during permeation produced data on the depth of slurry penetration,
the hydraulic conductivity of the overall seal, the characteristics of smaller
zones of different hydraulic conductivity within the seal, and changes in
these features as a function of time under varying hydraulic pressure condi-
tions. Details of the column system and the tests conducted are presented in
Section 5.
A second experimental system was developed to study model soil-bentonite
cutoff walls under simulated field hydraulic and surcharge loading conditions.
Testing involving the slurry wall tank ("tank") is documented in Section 6 of
this report. The tank system accomodates circular cutoff walls roughly 559 mm
(22 inches) in height, 101.6 to 152.4 mm (4 to 6 inches) thick, and up to
609.6 mm (24 inches) in diameter. The tank is of stainless steel construction
to allow for the use of selected chemical permeants and employs a pneumatic
bladder system to vertically confine and consolidate the model wall during
permeation in the horizontal direction.
Finally, additional experimental work on soil-bentonites and bentonite
slurry seals was concurrently performed under a separate work assignment
entitled "Quick Indicator Tests to Characterize Bentonite Type"(l). This work
focused on the characterization of several bentonite types for the purpose of
jobsite type-distinction and addressed issues such as:
• The effect of selected chemicals on the long-term hydraulic conduc-
tivity of typical surface filtration seals.
* The effect of bentonite slurry concentration (percent bentonite) on
the hydraulic conductivity of typical seals permeated by water and
selected chemicals
-------
• The effect of chemical pre-treatment of sodium bentonite during
manufacture (different vendor "types") on the hydraulic conduc-
tivity of surface filtration seals.
0 The effect of bentonite and other selected mix design parameters on
the hydraulic conductivity of a "standard" soil-bentonite.
• The performance of various bentonite "types" commonly specified for
use
The study involved numerous hydraulic conductivity tests employing deion-
ized water and various concentrations of acetone and methanol. Results of
the study are not incorporated herein except by reference.U)
-------
SECTION 2
CONCLUSIONS
SLURRY SEALS
• Based upon results of limited testing involving slurry seals, or "filter
cakes", formed in unsaturated sands, it appears that the hydraulic conductiv-
ity of such seals is not measureably different from comparable seals formed
on the same sand in a saturated condition. In other words, there is no
performance distinction to be made between slurry seals formed above and
below the groundwater table in uniform deposits of sands of the types tested.
• For seals ("filter cakes") formed on fine sands by the surface filtra-
tion mechanism, test results suggest that: 1) the hydraulic conductivity of
a seal is inversely proportional to the prevailing hydraulic head under which
the seal forms, 2) the thickness of the slurry seal is a function of seal
formation time only, and 3) as follows from 1) and 2), the density of the
seal is proportional to the prevailing hydraulic head under which the seal
forms. Considered together, these conclusions suggest that the thickness of a
surface filtration seal on the walls of a cutoff trench is constant over the
depth of the trench, but the density increases and the hydraulic conductivity
of the same seal decreases with increasing depth in the trench.
• Based upon the measured hydraulic conductivity and thickness of surface
filtration seals and rheological blockage seals formed in samples of +200
(sieve size) sand and +40 sand, respectively, the effectiveness of the seals
considered alone (i.e. without a soil-bentonite backfill) are not greatly
different, and are not significant in terms of overall containment time
(giving typical field breakthrough times of 12.2 days and 9.4 days for the
+200 sand and +40 sand, respectively). This implies that the seals adjacent
to defects in the soil-bentonite backfill will not serve as effective cutoffs.
* Based upon the measured properties of the two types of seals, and assum-
ing there are no hydraulic defects (windows) through the soil-bentonite
backfill, the effectiveness of a soil-bentonite cutoff wall constructed in a
deposit of +40 sand in terms of breakthrough time is three times as great as
that formed in a deposit of +200 sand (93.5 years vs. 31.0 years). As a
comparison, the typical soil-bentonite wall alone would have a breakthrough
time of 28.2 years.
* In summary, the slurry seal can contribute significantly to the effec-
tiveness of the slurry cutoff wall, especially if it is a rheological blockage
type seal (as in the +40 sand). However, this is true only if the soil-bento-
nite backfill has no defects (windows). Near hydraulic defects in the soil-
-------
bentonite backfill, neither the backfill alone, nor it in combination with
either type of slurry seal will function as an effective cutoff.
MODEL CUTOFF WALLS
The performance of three model soil-bentonite walls was investigated in
the laboratory using a specially designed slurry wall tank. The testing
program comprised two hydraulic conductivity test sequences and a single
window closing test. The former were performed mainly to evaluate the effects
of different overburden pressures and net hydraulic pressures (or gradients)
on the hydraulic conductivity of soil-bentonite walls. The latter experiment
was designed to investigate the capability of overburden pressures to close
windows (slurry pockets or shrinkage cracks) inside soil-bentonite walls.
The following conclusions are drawn from the results of this study:
• The slurry wall tank may be successfully applied to the study of the
hydraulic aspects of soil-bentonite slurry walls under various stress condi-
tions. However, the tank is not of sufficient size to realistically model
conventional construction methods and the type and extent of construction-
related as-built defects that may occur in practice. The frequency character-
istics and impact of such defects, as well as the means to remediate them,
therefore remain unknown.
• The hydraulic conductivities of the model soil-bentonite walls, as meas-
ured in the slurry wall tank, were found to be slightly lower than those
obtained using rigid-wall permeameters and the same soil-bentonite mixture.
The two sets of test results differed by a factor of about four. This may be
attributed to the following factors: (1) the model soil-bentonite wall was
bounded by a slurry seal on both sides, whereas the rigid-wall samples had a
seal on the surface only; (2) there was always an overburden pressure applied
to the soil-bentonite wall, but none to the rigid-wall samples; (3) minor
sidewall leakage of permeant for the rigid-wall samples is suspected, whereas
none was possible in the model soil-bentonite walls.
The comparable hydraulic conductivities suggest that results from the
rigid-wall permeameter, and especially from the triaxial type permeameter,
are quite representative of the hydraulic conductivities of full-scale slurry-
walls. This is contrary to the data published on the relationship between
field and laboratory hydraulic conductivities of compacted clay soil liners,
where ratios in conductivities reaching up to the thousands have been reported.
The good agreement in laboratory and field conductivities for soil-bentonite
is attributed to the homogeneity of the material.
* The average hydraulic conductivity of the model soil-bentonite walls was
observed to decrease both as a function of increased overburden pressure
(causing vertical consolidation), and increased hydraulic gradient (causing
horizontal consolidation), as well as their combined effect.
" Hydrofracture, or a rupture.in the soil-bentonite wall may be induced in
the subsurface at locations where the hydraulic driving pressure exceeds the
-------
total vertical overburden pressure. This phenomenon was evidenced in two
separate tests in the form of excessive inflow and outflow of permeant after
incrementing the hydraulic driving pressure. Although the applied surcharge
pressure at the top of the wall in these cases was higher than the hydraulic
pressure, it was not effective over the full depth of the wall resulting in
general hydrofracture (presumably near the base of the wall).
• Density, water content and vane shear strength data measured on samples
from the soil-bentonite walls after testing, all indicate that consolidation
did take place due to the applied surcharge and hydraulic seepage stresses.
In all cases, the degree of consolidation decreased somewhat with increasing
depth in the wall indicating that friction between the soil-bentonite and
sand may be a limiting factor in the vertical propagation of overburden
stress.
• Visual inspection of the model soil-bentonite wall after Sequence 1 tests
tests revealed minor entrapments of sand that presumably fell from the trench
walls during the construction of the model. Given the scale of the experiment
and the level of care exercised in building the model, this phenomenon may be
even more prevalent in real-site walls. Visual examination did not, however,
reveal any entrapped pockets of slurry.
• The window closing test performed on the third model soil-bentonite wall
demonstrates that artificial slot windows can be effectively closed and the
average hydraulic conductivity of the wall returned to the baseline (no
window) value with the application of surcharge pressures. The success of
this test implies that the real-site impact of as-built defects, such as
entrapped slurry or shrinkage cracks due to chemical degradation of the
backfill will be minimized or eliminated to the extent the wall consolidates
under its own self-weight, or under any applied surface surcharge pressure
such as that due to an overlying earth embankment, or under the lateral
pressure due to the in situ soil, or a combination of the above.
-------
SECTION 3
RECOMMENDATIONS
RESEARCH NEEDS
Although the findings presented in the Conclusions Section serve to
advance our understanding of soil-bentonite slurry walls, they also serve to
define several specific areas where additional research is needed. In gen-
eral , the types of studies that are needed can be divided into two categories;
laboratory or bench-scale work and field or pilot-scale work.
Where possible, additional laboratory work should be tailored to derive
maximum benefit from the experimental systems described in this study and
which are now on-hand. Principal among these is the slurry wall tank.
Experiments that should be performed using this system include:
• Hydraulic conductivity tests involving the permeation of the model soil-
bentonite wall with organic chemicals and leachates. Since the informa-
tion yielded from these tests are essential to the understanding of the
performance of soil-bentonite'cutoff wall exposed to organic liquids, it
is recommended that such tests be initiated as soon as possible. Of
essential interest is the ability of overburden pressures to restore the
baseline hydraulic conductivity of a model wall that has developed
shrinkage cracks due to the effects of leachates or chemicals.
* Trial elimination of model windows in the form of discrete pockets of
entrapped bentonite-water slurry. Evans et al.(2) consider this form of
window to have a high likelihood of occurrence in a typical slurry wall.
The experiment would generally involve creating a slurry pocket within a
model wall of known baseline attributes (presumably by injection),
followed by systematic application of surcharge and hydraulic pressures
until the baseline conductivity of the wall was reestablished.
It is recommended that experiments be performed to investigate the factors
controlling the dissipation of vertical stress in a model trench and the
mechanisms by which various types of windows may develop within soil-bentonite
backfill. Both objectives could be addressed using relatively inexpensive
equipment.
A better understanding of the state-of-stress in a model soil-bentonite
wall could be achieved by performing several bench-scale experiments using an
instrumented "friction column". The general approach would be to measure the
-------
involves application of a known load at the top of the confined column
of soil-bentonite and monitoring the pore pressure history and effective
vertical load at the base of the column for several different column heights.
The difference between the applied load and the measured load at equilibrium
would represent the amount of load dissipated due to arching or sidewall
friction. Given sufficient trials, it should be possible to establish the
distribution of vertical effective stress as a function of depth (column
height). Work using the friction column would be supplemented by numerous
direct shear tests to establish friction coefficients to describe the boundary
mechanics between dissimilar soils (soil-bentonite backfill and sand or clay).
It-is recommended that a new soil-bentonite placement flume be
designed and fabricated for the investigation of window development mecha-
nisms. Limited work of this type has been performed in connection with this
study, but primarily for the purpose of refining technique associated with
the construction of model walls in the slurry wall tank. The new flume would
have dimensions of roughly 1.22 m x 3.05 m x 0.46 m (4 ft x 10 ft x 1.5 ft)
and would be used to investigate the development and as-built distribution of
slurry pockets as well as the impact of trench-bottom configuration and sedi-
iments on the integrity of soil-bentonite backfill after placement. The
experiments could use dyed slurry to facilitate the recognition of windows in
the backfill.
Any list of recommendations for additional slurry wall research must
include the testing of full-scale model walls. The foregoing laboratory
studies are a necessary first approximation in answering the research ques-
tions described. They serve to identify many of the primary variables and
their relationships but they typically cannot account for all factors. For
example, the question "how thoroughly can a contractor mix soil-bentonite
using conventional equipment" can only be answered with confidence at field-
scale using actual construction equipment. The same principle holds for the
questions of as-built defects and the stateof-stress in a soil-bentonite
slurry wall. Until these questions are addressed both at laboratory and field
scale, it is possbible that we will not find wholly satisfactory answers. In
this context, it would seem that a pilot-scale slurry wall test facility is
more than justified. The cost of such a test facility might be high, but
this cost would be more than offset if materials and methods were refined to
the point that the need for performance monitoring or additional remedial
work at real sites (after the installation of a slurry wall) became the excep-
tion rather than the rule.
-------
SECTION 4
SUMMARY OF CURRENT METHODS
The initial phase of this study involved a survey of current design and
construction methods which form the basis of present slurry cutoff wall
technology. The survey involved review of published literature on the sub-
ject, interviews with selected vendors and professional design and construc-
tion engineers specilizing in slurry wall applications, and visits to three
slurry wall construction sites. Table 1 is a listing of the vendors, con-
tractors and consultants who were interviewed and the slurry wall installa-
tions visited during the first phase of this investigation. Written and
unpublished oral communications on the subject are summarized under the
headings of Design, Specification, Construction, QA/QC, and Performance Mon-
itoring. For the purpose of emphasis, areas of practice under each heading
which are most variable and therefore most in need of standardization have been
underlined.
DESIGN
Design of a slurry cutoff wall system is a multifaceted undertaking
involving consideration of geological, geotechnical , and hydrological condi-
tions; development of conceptual options; review of options for technical
requirements; input from contractors on constructibility; economic analysis;
and site investigation of sufficient detail to permit a sound evaluation.
Detailed consideration of these general design activities is beyond the scope
of this study. The key issue in most situations is the design of a suitable
soil-bentonite backfill, part of which involves compatibility testing under
assumed service conditions. This area of practice appears to be the most
variable among the published sources and parties interviewed.
Unless on-site soils are suitable for use as backfill, a well-graded
blend of on-site and borrow soils or an all borrow soil is selected. Labora-
tory tests that may be performed at this stage include: grain size analysis,
Atterberg limits, water content, and possibly moisture-density (compaction).
Upon selection of a suitable backfill material, several different percentages
of bentonite, both ordinary and chemically resistant, are mixed with the
selected material. Permeability tests are then performed using water as the
permeant. These tests are continued until the flow reaches equilibrium after
which one or more mixes are selected and subjected to permeation with the
"worst case" leachate to determine compatibility.
Most designers and contractors interviewed indicated that based on their
experience, few if any soil-bentonite samples tested had exhibited any
10
-------
TABLE 1. SOURCES OF UNPUBLISHED INFORMATION ON SLURRY WALL TECHNOLOGY
BENTONITE SUPPLIERS
American Colloid Co., Skokie, IL
Robert Kingsbury
Christopher Jepson
Federal Bentonite, Montgomery, IL
Bruce S. Beattie
Don Hentz
SLURRY WALL CONTRACTORS
Geo-Con, Inc., Pittsburgh, PA
Chris Ryan
IT Corporation, Pittsburgh, PA
Geoff Shallard
Glenn D. Schwartz
Richard M. Burke
ICOS Corporation of America, New York, NY
Nicholas J. Cavalli
DESIGNERS AND/OR
QA/QC CONSULTANTS FOR
SLURRY WALL INSTALLATIONS
GAI .Consultants, Inc., Monroeville, PA
F. Barry Newman
GZA (Goldberg, Zoino & Associates, Inc.)
Upper Newton Falls, MA
Matthew J. Barvenik
Woodward-Clyde Consultants
Plymouth Meeting, PA
Jeffrey C. Evans
Woodward-Clyde Consultants
Clifton, NJ
Richard S. Ladd
Paul Dudko
SLURRY WALL INSTALLATIONS
Granger Land Development Co.,
Municipal Waste Disposal Site,
Lansing, MI
Kurt J. Guter
GE Waste Site, Glen Falls, NY
Enzo Zoratto
Municipal Landfill, Kingsland, N.J.
Nicholas Cavalli
11
-------
increased permeability or reaction with the leachates used. It was suggested
that the low concentration of chemicals in most leachates may be the reason.
In general, the threshold concentration above which the leachate-will affect
the soil-bentonite usually is not attained.
The consultants differ in their approach to permeability testing. One
firm performs triaxial permeability tests with both a closed system (prevents
consolidation) and the standard open system (induces consolidation). Two use
standard triaxial permeability testing with either 76.2 mm (3 inch) or 101.6
mm (4 inch) diameter specimens of 50.8 mm to 101.6 mm (2 to 4 inch)lengths
and rigid-wall permeameters consisting of a modified API Filter Press with
76.2 mm (3 inch) diameter by 50.8 mm (2 inch) high specimens. Another firm
uses triaxial, compaction mold, or Harvard miniature permeameters, giving
preference to the latter because of the reduced time of testing. The Harvard
miniature uses a 25.4 mm (1 inch) diameter by 50.8 mm (2 inch) long sample.
Hydraulic gradients used by various consultants for performance and
compatibility testing ranged from 30 to as high as 150 mm/mm. The number of
pore volumes, which are passed through a sample, varied from a fraction to as
much as three. This, of course, means testing time from a few days to a few
months. Some consultants measure both inflow and outflow, others outflow
only.
Some consultants and contractors design the soil-bentonite to meet the
minimum 1 x 10"^ cm/sec hydraulic conductivity criteria, whereas others target
lower permeability values such as 1 to 5 x 10 8 cm/sec in the design testing.
All consultants and contractors addressed the need for "hydrologic
balancing", that is, either zero gradient or a very low gradient toward the
waste containment area, thereby preventing (or at least minimizing) leachate
contact with the soil-bentonite backfill. Either a drainage trench or a
well-extraction system, plus effective capping of the site are required to
create such a condition. A recognized advantage of hydrologic balancing is to
reduce the risk of undetected hydrofracture of the wall below the ground
surface. In general, since the stress conditions within the backfill and
along the backfil1-trench wall interface are not known, the potential for win-
dow window development via hydrofracture under the anticipated hydraulic
gradient is also an unknown; therefore minimize gradient. Such is not always
possible, however, especially where remediation in the form of pumping and
treatment of contaminated groundwater would be involved. In short, the
designer does not have the information necessary to permit rational considera-
tion of the potential for hydrofracture.
The lack of data on the "state of stress" in a cutoff wall also affects
the selection of appropriate hydraulic gradients and confining stress for
permeability testing and other design parameters.
In terms of containment potential, wall thickness typically is not a
major design consideration (standard widths of 635 to 914.4 mm (25 to 36
inches) usually are adequate) except as it relates to hydrofracture potential.
Such is not established, however, so the adequacy of customary widths is not
known in cases where a hydrologic balance is not maintained over the life of
12
-------
the barrier.
In the case of laboratory permeability testing for design (or QA/QC) pur-
poses, the general consensus is that the stress condition or testing should
model the prevailing stress conditions near the top of the wall. This is
presumably where the least amount of in situ consolidation of the backfill
due to effective overburden stresses would occur, and therefore also the zone
of the highest long-term permeability of the barrier after construction.
Whether or not the backfill actually undergoes significant in situ consolida-
tion is not known. Moreover, the effect of test gradient generally is appre-
ciated, but not always addressed in design testing. Assuming a fair hydro-
logic balance, test gradients should typically be small but were found to
range up to 150 mm/mm for the sake of test expedience as reported earlier.
Clearly, permeability data from soil-bentonite samples which have been artifi-
cially consolidated under such high gradients have questionable relevance to
most as-built conditions.
Some consultants and contractors prefer a low bentonite content (1/2% to
2%) within the soil-bentonite backfill, whereas others prefer higher contents
3% to 5%). On at least two investigated projects, optimum bentonite contents
were determined and then increased by an additional two percent of dry bento-
nite to account for any inadvertent losses during construction. Some labora-
tory test results indicate that additional bentonite above the optimum amount
can cause a slight increase in permeability.In cases of questionable lea-
chate-backfill compatibility, this could greatly detract from the long-term
performance of the barrier. In fact, some consultants and contractors prefer
to reduce backfill hydraulic conductivity by the addition of native plastic
fines because they believe the long-term performance with leachates is better
than if additional bentonite were used (an additional benefit with the use of
native fines is reduced cost). Another detrimental effect of higher bentonite
content could be greater consolidation potential of the S/B backfill.
A number of the consultants and contractors indicated a preference for
use of contaminated trench spoils for the backfill for two reasons: 1) Any
chemical interaction between the trench spoils and the bentonite will occur
during mixing. Although the initial permeability of the backfill may be
slightly higher, there will be less long-term degradation caused by chemical-
ly similar leachate, and 2) Use of these spoils often is cost effective,
since borrow soils need not be purchased.
Each consultant and contractor preferred to use unaltered Wyoming bento-
nite meeting API 13A for use in slurry trench cutoffs unless the compati-
bility testing program indicated that chemically-resistant bentonite would
provide the best performance. One reason is that polymers added to "peptize"
bentonites are biodegradeable and may enhance reactivity and degradation by
chemicals. Another reason is that approximately two times as much chemically
resistant bentonite is required to achieve the desired slurry properties
during construction and the unit cost is two times that of unaltered bentonite.
This yields a total cost factor of about four relative to unaltered bentonite.
Of those interviewed, no consultant or contractor considers the contribu-
tion of the filter cake to the overall conductivity of the wall. Neglecting
13
-------
the filter cake hopefully affords some additional factor of safety under the
low hydraulic gradients present at most sites where cutoffs are installed.
Another essential element in the design phase is to establish an
appropriate method to key into a soil or clay aquielude or into the bedrock.
A key into clay isrelatively straightforward and mainly dependent on con-
struction technique and QA/QC. To key into bedrock (weathered, fractured, or
sound) is much more problematic.
Hydro!ogic modeling is essential to a good slurry trench design so that
a determination can be made of potential hydraulic gradients to which the
wall may be subjected under various hydrologic conditions. Based on this
model, the need for hydrologic balancing can be determined and, if required,
the design of an appropriate drainage or extraction system can be completed.
SPECIFICATIONS
A well-conceived, thorough, and clearly written specification document is
essential to the success of any construction project since it constitutes a
summary of design requirements and may even prescribe specific methods to
achieve the desired end product. In general, specifications for soil-bento-
nite cutoff wall construction are similar in most respects, but normally
contain special features or requirements reflecting site-specific conditions.
r
Assuming a qualified contractor, the consensus among designers is that
an effective soil-bentonite cutoff can be installed, by the slurry trench
method utilizing either a performance specification or a material and methods
specification. Contractors tend to favor a performance specification since
it generally permits wider latitude in the selection of materials and methods
and therefore holds greater economic potential. Designers on the other hand
tend to prefer a materials and methods approach since it is more conservative
and allows for greater control of critical aspects of the job. The optimum
approach probably lies somewhere between these extremes, namely, a performance
specification with certain limiting materials criteria within which the
contractor can utilize his expertise and ingenuity. The specifics of such,
especially in terms of a model or "standard" specification, is the subject
of considerable controversy among designers, contractors, and product vendors.
For example, all parties agree that the most critical issue to be addressed
in the specifications is the design mix—its components and method of batching
and mixi ng.
There seems to be no general concensus regarding the type of bentonite to
be used. Based upon discussion with those interviewed and a review of
available project specifications, the debate appears to be split about 50-50
concerning the use of unaltered sodium-cation Wyoming bentonite as opposed
to a chemically modified version of the same material.
Within the bentonite industry there appears to be a basic distinction be-
tween chemically treated or "resistant" bentonites and chemically enchanced
or "peptized" bentonites. The . chemicallyresistant types generally resist
degradation via cation substitution. However, the properties of these benton-
14
-------
ites in slurry form usually are suppressed relative to those of unaltered
bentonite, requiring that more chemically treated bentonite be used to achieve
specified slurry properties. In contrtast, a chemically peptized bentonite
generally is more "active" than unaltered bentonite, requiring less bentonite
to achieve the desired effect. Most designers favor unaltered bentonite
because of undocumented long-term performanceof thechemicallypeptized
types. The viscosity and other specified properties of the chemically pep-
tized bentonites in slurry form can be made to match those of unaltered
bentonite, while having a net bentonite content which may be lower by as much
as fifty percent. Moerover one general type known as polymer extended bento-
nite derives its enhanced behavior via more complete chemically-induced hy-
dration. Resulting attributes such as greater swell potential are desireable
unless exposure to in situ chemicals or leachates eventually will collapse
the bentonite producing a greater loss of effectiveness (due to net bentonite
deficiency) relative to the use of the unaltered type.
Contractors favor the use of unaltered or chemically peptized bentonite,
generally for economic reasons (chemically resistant types more costly by a
factor of up to four).
The preference of vendors tends to lie with the chemically resistant
bentonites. The reason appears to be a combination of relative economics and
greatest long-term effectiveness of the barrier. Clearly, if a vendor can
furnish more of a "resistant" product which represents a smaller portion of a
finite supply, and at a higher unit cost, he is ahead. Most vendors, however,
stand behind their products in terms of a limited warranty which typically
guarantees the specified performance so long as the barrier is not exposed to
certain chemicals or critical concentrations thereof.
In summary, the question of which bentonite to use and how much benton-
ite to specify is unresolved.Initialresultsfrom arelatedresearch pro-
ject^)) suggest that although there may be as much as an order of magnitude
difference in the permeability of a "standard" soil-bentonite depending upon
the type of bentonite used (unaltered demonstrated the highest permeability
followed by chemically resistant and chemically peptized, the most significant
factor was bentonite concentration (up to 3 orders of magnitude reduction for
change in concentration from 1% to 5%, for all three types). One specifica-
tion reviewed during this investigation was written .from this point of view,
i.e., not restricting the use of any particular bentonite type, but rather,
specifying only that the bentonite content of the backfill be greater than 5
percent, period.
Economic considerations aside, it is the opinion of the authors that the
use of chemically peptized bentonite products is considerably more risky than
the use of the other two types, and that chemically resistant products (which
serve to increase net bentonite concentration for the same slurry properties)
should not be restricted, so long as short-term permeability after several
pore volumes of flow is not adversely affected by the leachate applied.
Beyond this, it is recommended that cutoff wall specifications clearly define
the minimum acceptable bentonite concentration of the backfill in-place.
A comprehensive specification for a soil-bentonite cutoff wall should
15
-------
also clearly prescribe other job requirements ranging from qualification of
the personnel involved to the details of the quality assurance program.
Discussing details of these numerous other areas of specification is beyond
the scope of this study. The following listing which appears as the table of
contents in the recent Civil Works Construction Guide Specification for
Soil-Bentonite Slurry Trench Cutoffs prepared by the Corps of Engineers (3),
is provided for reference purposes. The intent of this guide specification
is to facilitate the preparation of specifications by a designer who under-
stands that "the requirements for the bentonite, backfill, and construction
procedure are highly dependent on the intended purpose of the cutoff wall and
the environment in which it is to be used."
CIVIL WORKS CONSTRUCTION
GUIDE SPECIFICATION
FOR
SOIL-BENTONITE SLURRY TRENCH CUTOFFS
TABLE OF CONTENTS
Paragraph
1. SCOPE
2. APPLICABLE PUBLICATIONS
2.1 American Petroleum Institute (API) Standard
Specifications
2.2 American Society for Testig and Materials (ASTM)
Standards
2.3 Corps of Engineers Manuals
2.4 Military Standards
3. GEOTECHNICAL SITE CONDITIONS
3.1 Exploratory Borings
3.2 . Subsurface Conditions
3.3 Groundwater
3.4 Embankment Conditions
4. DEFINITIONS
4.1 Slurry Trench
4.2 Slurry Method of Excavation
4.3 Bentonite
4.4 Slurry
4.5 Backfill
4.6 Groundwater Level
4.7 Working Surface
4.8 Impervious Stratum
4.9 Slurry Trench Specialist
5. SUBMITTALS
5.1 Schedule and Sequences of Operations
5.2 Layout of Operations
5.3 Contractor's Qualifications
16
-------
5.4 Slurry Trench Specialists' Qualifications
5.5 Slurry Trench Construction Method and Equipment
5.6 Blast Plan
5.7 Bentonite Certification
5.8 Gradation of all Imported Borrow Material
5.9 Backfill Gradation
(Materials Prior to Mixing with Slurry)
5.10 Mix Design
5.11 Equipment and Procedure to Obtain Samples
5.12 Excavation and Backfill Soundings
5.13 Quality Control Testing Equipment and Procedures
5.14 Results of all Contractor Quality Control (CQC)
Tests and Measurements
5.15 Equipment and Procedure to Obtain Undisturbed
Record Samples of Completed Slurry Trench
6. QUALIFICATIONS FOR SLURRY TRENCH CONSTRUCTION
6.1 Contractor
6.2 Slurry Trench Specialist
7. SUBSURFACE INVESTIGATIONS
7.1 General
7.2 Drilling
7.3 Sampling
7.4 Survey
8. INSTURMENTATION MONITORING
8.1 Piezometers
9. MATERIALS
9.1 Bentonite
9.2 Water
9.3 Bentonite slurry
9.3.1 Initial Bentonite Slurry Mixture
9.3.2 Trench Bentonite Slurry Mixture
9.3.3 Additional Bentonite
9.4 Additives
9.5 Backfill
10. EQUIPMENT
10.1 Trench Excavation
10.2 Mixing and Placing Slurry
10.3 Cleaning of Slurry
10.4 Preparation of Trench Bottom
10.5 Mixing and Placing Backfill
11. SLURRY TRENCH CONSTRUCTION
11.1 General
11.2 Working Surface
11.3 Excavation
11.3.1 Bedrock Excavation
17
-------
11.3.2 Blasting
11.4 Placement of Slurry
11.5 Excavated Material
11.6 Backfilling Trench in Case of High Water
11.7 Stability
11.8 Treatment of Trench Bottom
11.8.1 Treat of Rock Surfaces
11.8.2 Cleaning
11.8.3 Sampling
11.9 Backfilling
11.9.1 Mixing Areas
11.9.2 Mixing
11.9.3 Placing
11.9.4 Mixing and Placing During Cold Weather
11.10 Treatment of Top Slurry Trench
11.12 Records and Controls
12. QUALITY CONTROL
12.1 Bentonite
12.2 Water
12.3 Slurry Properties
12.4 Excavation and Backfill Soundings
12.4.1 Elevation of Top Impervious Stratum
12.4.2 Elevation of Bottom ofExcavation
12.4.3 Elevation of Bottom Prior to Backfilling
12.4.4 Profile of Backfill Slope
12.5 Backfill Properties
12.5.1 Slump Tests
12.5.2 Backfill Permeability Determinations
12.6 Samples of Impervious Statum
12.7 Records
12.7.1 As-Built Profile
12.7.2 Results
12.7.3 Bentonite Slurry Mix
12.7.4 Construction Log
13. QUALITY ASSURANCE
14. MEASUREMENT
15. PAYMENT
18
-------
CONSTRUCTION
Slurry trench cutoff wall construction is primarily an art and is there-
fore very dependent on the experience and expertise of the equipment operators
and supervisors. When a contractor cannot bring experienced operators to a
project for whatever reason (most often local union requirements) then the
quality of the installation is completely dependent upon the expertise and
training skill of the supervision personnel. Although some construction
operations are more easily learned than others (a dozer operator can be more
easily trained than a backhoe or clamshell operator), there typically is some
proficiency lag whenever a new operator is "broken in". As a result, the
initial part of a soil-bentonite cutoff may be of somewhat lower quality than
subsequent sections of- the wall. In cases such as this it might be more
appropriate to initiate construction on the up-gradient side of a site. Other
key elements of construction as addressed by those interviewed were:
1. The trench must be excavated so that it is continuous from start to
finish. It must be excavated to an appropriate depth, either into a
clay aquiclude or soft bedrock or onto the surface of competent bed-
rock that has been sealed via grouting or other techniques.
2. When a backhoe is used to excavate the trench, it should remove in
situ soils in continous 152.4 to 304.8 mm (6 to 12 inch) horizontal
layers. In this manner, the soil is peeled off and can be inspected
to determine soil type at all locations. This method is especially
effective in establishing the depth to the aquiclude and the nature
of materials.
3. A continuous clamshell operation, immediately in front of the backfill
toe is necessary to remove pockets of slurry mixed with coarser
materials or chunks of .soil from the trench sidewall (cave-in below
slurry level) or those which may fall from the backhoe or clamshell
bucket. Such pockets could ultimately lead to windows of more
permeable zones through the wall.
4. The slurry must be sufficiently viscous, with the level maintained
near the ground surface to maintain the stability of the trench walls.
The viscous slurry also serves to suspend sand or finer materials
that have fallen off the side of the trench.
5. The backfill should be batched and mixed in accordance wi.th the
specifications, either alongside the trench using a dozer, or at a
central batching area consisting of concrete slab and mixing equip-
ment such as a rubber-tired, pipeline auger backfiller. It is
essential that all large clods of clay be broken down and that
the sand, clay, and bentonite be mixed thoroughly until the mixture
becomes a homogeneous mass of 101.8 to 152.4 mm (4 to 6 inch) slump.
6. Most importantly, the backfill should intially be placed through the
slurry by tremie pipe or clam bucket. Once the backfill surface
"daylights" above the .slurry surface at the starting point of the
excavation, the backfill can be pushed by dozer or dumped by truck
19
-------
onto the exposed backfill surface immediately behind the crest of the
backfill slope. According to Evans et. al.(2), when the backfill
is placed in this manner its weight causes shallow shear displace-
ment within the backfill which produces a gradual forward advance-
ment of the sloped backfill surface along the slurry trench.
7. Batching and mixing the backfill is a major potential source of error.
There is the potential that the homogeneity of different batches of
backfill may vary greatly or that the gradation of trench spoils
might change along the alignment of the wall. Another potential
problem is that significant bentonite losses (up to 15% or 20%) can
occur during spreading of dry bentonite. Such losses can occur as a
result of wind, uneven distribution due to windrows in the spread-out
backfill, and losses to the soil underlying the batching area.
8. Depending on the condition of the in situ soil deposits, desanding
of the slurry may or may not be required. Most often it is. When
desanding is required, an airlift is typically located near the toe
of the advancing backfill slope and the slurry is desanded in a
cyclone separator or other machine.
9. Trench depth must be measured continuously during excavation by the
equipment operator utilizing the construction equipment and checked
by QC personnel using a tape measure or a sounding rod device.
10. Greater caution ordinarily is needed to verify the continuity of a
deep excavation completed with a clamshell than a shallow excavation
completed with a backhoe.
11. Determination of the depth and continuity of the key into bedrock is
difficult to verify where the rock surface is highly irregular and
the overburden is dense and of similar origin. In situations where
the rock is soft, uniform and easily distinguished from the overbur-
den, little difficulty is encountered.
Are the methods and considerations described in the preceding paragraphs
sufficient to insure the adequacy of a soil-bentonite cutoff wall? The answer
is generally not known, especially in light of limited or no long-term perfor-
mance monitoring data for completed walls.
Based upon experimental results from part of this study utilizing the
slurry test column (discussed later), it appears that bentonite slurry seals
can contribute significantly to the overall containment potential of a cutoff
wall, but only if the backfill itself is free from significant defects. The
greatest potential contribution appears to be in the case of slurry penetra-
tion without significant filtration.
QUALITY CONTROL (QC)
Ideally, QC monitoring should be provided during slurry trench construc-
tion to ensure trench integrity and the adequacy of placement to guard against
20
-------
discontinuities or "windows". QC personnel typically check slurry viscosity
in the holding pond and slurry viscosity and density in the trench. The
depth of the trench is typically checked and the materials being excavated
from the trench are inspected to ensure that the in situ materials remain
consistent and to verify that the key is completed into appropriate material.
The depth of the soil-bentonite backfill is checked at close intervals at
least once a day to maintain an accurate profile of the slope of the soil-
bentonite. These depth measurements along the slope of the backfill and
between the backfill toe and face of excavation also represent control on the
continuity of the soil-bentonite wall. The adequacy of such methods to ascer-
tain that no caved-in materials from trench walls or pockets of settled
solids are present within the key is not known.
Some contractors perform daily permeability tests on the backfill from
the batching area(s). These are accomplished using a modified API filter
press to test a 50.8 mm (2 inch) high by 76.2 mm (3 inch) diameter sample
using water. The test is usually completed overnight. During the design
phase a relationship is developed between the field permeability determined
using the API filter press and the design permeability and compatibility
testing results.
The contractor may provide the QC utilizing his own forces, but this
approach is counter-productive, at least in principle. Alternatively, he may
retain a qualified geotechnical consultant or testing laboratory. It is prob-
ably most appropriate that the owner provide a QC representative to perform
the required testing and inspection on a full time basis. Ideally, this
service is best provided by two persons: one to observe and make measure-
ments continously; the other to sample, perform tests, and assist with obser-
vations and measuremnts. The intent of QC can only be achieved if the persons
performing the QC are experienced with slurry trench cutoff construction. A
well-equipped field laboratory is essential.
Overall wall integrity, assuming appropriate design, is dependent pri-
marily on two factors: 1) trench continuity including key into aquiclude,
and 2) the soil-bentonite backfill mix. QC carried out as above on a project
appropriately designed and assuming a qualified contractor will provide
reasonable assurance of overall wall integrity.
The following is a listing of the most important items to be inspected
and/or tested during slurry trench construction:
-depth to top of slurry
-slurry viscosity at the top, middle, and bottom of trench
-slurry viscosity at the holding pond
-slurry density at the top, middle, and bottom of trench
-API filtrate loss tests on the holding pond slurry
-chemistry of slurry mixing water, especially if it is groundwater
21
-------
-sand content of slurry in trench
-gradation of trench spoils and borrow
-fines content and Atterberg limits of the backfill
-uniformity of dry bentonite application
-batching and mixing of the backfill
-slump of backfill
-unit weight of backfill
-location and method of placing backfill
-depth to top of aquielude at minimum 3.05 m (10 ft) spacing
-key into aquiclude
-daily profile of backfill
-trench continuity
-daily permeabilities of backfill from batching area
-verification testing of undisturbed samples from completed wall as job
progresses
-quality of bentonite received using various tests
The last item listed (bentonite quality) is the subject of much concern
considering the previously described controversy over what type of bentonite
to use. Several consultants interviewed agreed that most often, one does not
know whether unaltered or chemically modified bentonite is being supplied to
the jobsite. One consultant claimed that he knew that unauthorized substitu-
tion occurred regularly.
Part of the research project on bentonite characterization previously
referenced dealt with the identification of bentonite type for QC monitor-
ing purposes.'1) Eight different products comprising three type categories
(unaltered, polymer protected or extended, and chemically treated) were sub-
jected to numerous "quick indicator" tests to evaluate the type-distinction
potential of the various test procedures. Based upon a statistical analysis
of results, it was determined that simple test procedures commonly used at
the jobsite to monitor gel strength and apparent viscosity of the bentonite
slurry could also be used to distinguish between the specific bentonites
tested in the study. Despite the high level of confidence associated with the
identification of these particular bentonites, results apply only to a limited
sampling of the population of commercial bentonite products and cannot be
extended beyond the scope of the referenced study without substantial addi-
22
-------
tional testing. It does, however, appear that a general type-distinction
tool could be developed for QC use.
To summarize, there is considerable room for error during the process of
slurry trench cutoff wall construction. For this reason, a high level QC
program is warranted. It is essential that the QC personnel have considerable
experience in slurry trench cutoff construction techniques and a comprehensive
awareness of the design intent. If only one person is providing the QC, and
that person concentrates on performing various tests only, then the QC program
cannot be considered adequate since there will be room to doubt trench conti-
nuity, the quality of the soil-bentonite backfill mix, and numerous other
details of construction.
PERFORMANCE MONITORING
Every consultant and contractor interviewed expressed the need for
compilation of performance data for slurry trench soil-bentonite walls used
for containment at hazardous waste sites. In short, most of the present uncer-
tainties associated with design, specification,construction and QA/QC will
remain unresolved until adequate performance data is available.
The reason for the lack of such information at this time is twofold:
high cost to generate performance data and the potential liability if it is
demonstrated that the barrier does not perform as specified. Although under-
standable, such reasons cannot be allowed to compromise efforts to protect
public health and safety; an equitable yet adequate approach must be devel-
oped.
The scope of any reasonable performance monitoring program should involve
at least the following:
• The installation of observation wells and the periodic performance of
pumping tests to measure the average as-built hydraulic conductivity
of the cutoff wall.
• Periodic comparisons of baseline (pre-construction) and post-
construction groundwater quality at key downgradient locations.
Where the cutoff wall will be subjected to significant loads, say from a
landfill to be built on or adjacent to the wall, then performance monitoring
should also include settlement plates and inclinometers to measure vertical
and horizontal deformations over the life of the cutoff wall.
Even with an effort of such scope, however, it may only be possible to
demonstrate general adequacy or the lack thereof. In cases of deficiency,
the specific cause or causes and therefore the scope of necessary remediation
would likely remain speculative to some (high) degree.
Perhaps more basic research is a large part of the answer. The efforts
described herein reflect a recognized need for such, but constitute only a
small portion of the necessary scope of work. In particular, there is a need
23
-------
to generate performance data which is not influenced by scale or other factors
necessarily connected with research in a laboratory environment. The best
approach certainly would be the development of a pilot-scale system which
would permit testing and evaluation of a slurry trench cutoff wall which was
constructed using customary equipment and methods. The cost of such a system
might be high, but it would be more than offset if materials and methods
were refined to the point that the need for performance monitoring at real
sites became the exception rather than the rule.
24
-------
SECTION 5
SLURRY SEALS
The purpose of this aspect of experimental work was to determine the
performance characteristics of slurry seals formed on the walls of the cut-
off trench during construction. In currrent design practice, the contribu-
tion of a slurry seal to the long-term integrity of the soil-bentonite barrier
is not known, and, therefore not considered. The research effort was designed
to determine slurry penetration, seal thickness and hydraulic conductivity,
and the change in these parameters as a function of time and penetration
depth in various sands.
According to Xanthakos(4), a bentonite seal is formed on the walls
of a cutoff trench via one of three modes:
1. Surface filtration occurs in fine soils when a surface "filter cake"
seal is formed as hydrated bentonite particles are brought together
in the soil voids. The result is a dense packing of material,
allowing only limited slurry penetration into the soil.
2. Deep filtration occurs in medium to coarse soil. In this case,
penetration may extend from a few centimeters to several meters.
Eventually, dense packing of filtered bentonite particles in the
zone just adjacent to the surface produces a seal which stops further
slurry penetration.
3. Rheological blocking, occurs in cases where the slurry flows directly
into formations such as coarse sand and gravel until it is restrained
by friction and its own shear (gel) strength.
In general, the type of seal that will form in porous media is primarily a
function of the grain size distribution of the material. An attempt was made
to investigate each type of slurry seal in terms of relative effectiveness
and longevity as an element of a hydrologic barrier under assumed seal
formation conditions.
All tests involved slurry seals derived from the penetration of a
"standard" 5 percent bentonite:water slurry (weight:volume basis).
The slurry was driven into the test sands under controlled pressure
conditions (seal formation pressure), for a standard period of five hours.
Seals formed in this manner were then permeated by water under variable
hydraulic pressures sometimes different than the seal formation pressure.
Testing also comprised both saturated and unsaturated sands to model
25
-------
conditions below and above the groundwater table, respectively.
The slurry batching procedure used throughout the study is presented
as standard laboratory procedure (SLP) No. S-01 in Appendix D. Table 2 pre-
sents data on several slurry parameters measured on the different batches of
slurry prepared for testing.
TABLE 2. MEASURED PROPERTIES OF 5% ORDINARY (unaltered) BENTONITE SLURRY
Batch
PARAMETER
Test
Total unit weight
(kg/m3)
API filtrate loss
(ml)
Marsh funnel
(sec)
pH
Specific conductivity
(mv)
Bentonite content
U)
1
1
1034.9
14.5
45.0
9.02
.799
5.11
2
2,3,4,5,6
1033.3
13.8
45.1
8.73
.909
5.02
3
7,8
1033.3
15.9
43.2
9.40
.892
5.17
4
9
1034.9
17.7
42.6
9.46
. .741
5.08
Five different gradations of sand were selected for testing. A clean
medium to fine sand identified herein as the "+200 sand" (retained on the no.
200 sieve) was used to study the surface filtration (filter cake) type of
slurry seal in the slurry test column. This sand is predominantly fine, of
roughly uniform size (no. 40 to no. 50 sieve size), with about 25 percent
medium sand by weight as shown in Figure 1.
Three other sands identified as +60, +50, and +100 were also used to
study the surface filtration mechanism, but in smaller columns similar in
design to a conventional rigid-wall permeameter. The "short" columns were
employed to allow for long-term testing of surface filtration seals on a wider
range of sands without interrupting or replacing other scheduled tests in the
large instrumented column. Each of these three sands was processed to achieve
a relatively narrow range of grain sizes. The +60 sand, for example, com-
prised only sand passing the no. 40 sieve and retained on the no. 60 sieve.
Likewise, the +80 sand comprised material passing the no. 60 sieve' and
retained on the no. 80; and so-on.
The last sand tested was a clean medium to coarse sand prepared to
investigate deep filtration and rheological block seals in the large column.
The gradation comprised roughly 75% medium sand and 25% coarse sand as shown
on Figure 1. All material was retained on the no. 40 sieve ("+40 sand").
26
-------
ro
--j
100
90
80
70
5
£ 60
V
oo
| 50
i 40
UJ
o.
30
20
10
0
5C
U.S. STANDARD SIEVE OPENING IN INCHES U.S. STANDARD SIEVE NUMBERS HYDROMETER
6 43 2 I'/i 1 ]/< Vt H 3 4 6 B 10 1416 20 30 40 50 70 100140 200
-
-
-
1 1
1
-•
...
--
-
—
1
-f
—
i i i
-
k
'
\
^
40 SAND
--
-
—
Oh1 — >•[+).,
V
\
~ —
\
A
\
" \
—
i
\
t
--
\
A
1
f
\
i
I
\
b
-
f
\
t
i
\
\
L
1 '
+
1
200 SAN
\
\
\
i.
0 100 50 10 5 1 0.5
GRAIN SIZE MILLIMETERS
COtBlfS
SAMPlf NO.
CBAVfl
COARSf | fINI
nivo" MUM
D
--
• -
-
0
10
20
30 >-
o
£
oc
UJ
10 1
U
60 5
IU
70 *
80
90
100
O.I 0.05 0.01 0.005 0.001
SAND
• COAHSC |
ClASSIMCAMON
GRADATION CURVES
MEDIUM flNE
NAl W
'. U
n
pi
SILT OR CLAY
3210-07
SLURRY TEST COLUMN
AM*
IOIINC NO.
0*H
Figure 1. Grain size distribution of sands used in the
investigation of slurry seals in the slurry
test column.
-------
EQUIPMENT AND PROCEDURES
Slurry Test Column
A schematic of the slurry test column system is shown as Figure 2.
The system consists of a transparent 1.83 m (6 foot) long column into which up
to forty-eight pore pressure probes can be installed after the soil medium is
in place. Probes are spaced along the column in a descending helix pattern
starting at the soil-slurry interface (about one 0.3 m (1 foot) below top of
column). The helix pattern permits a close vertical depth spacing of 3.175 mm
(1/8 inch) over the first 50.8 mm (2 inches) of the column (sixteen probes).
Below this, vertical spacing is gradually increased to a maximum of 50.8 mm
(2 inches) over the bottom 914.4 mm (3 feet) of the column (eighteen probes
in this interval).
The probes permit measurement of the in situ pore pressure before and
after slurry penetration using one of four differential pressure transducers.
Each transducer is connected to a single pressure manifold which is valved to
serve twelve probes. Digital readouts can be switched to the appropriate mani-
fold depending upon the probe selected for observation.
The system also incorporates spring-suspended inflow (head) and outflow
(tail) permeant reservoirs to achieve constant-head test conditions. A
centrally located instrument panel was used to control slurry penetration and
sample permeation via a regulated compressed air source. Figure 3 is a
schematic of the basal outflow assembly which also shows the lowermost (typi-
cal) pore pressure probe. Specific test procedures for the slurry test
column are presented in Appendix A.
The underlying principle for all column work was to accurately model in
situ hydraulic conditions that affect the formation and performance of a
slurry seal. During the course of the testing program it became necessary to
modify equipment and procedures in order to provide for or control:
* hydraulic head (pressure) loss due to frictional resistance in tub-
ing and fittings at high flow rates
* complete de-airing of the test sands prior to determination of base-
line hydraulic conductivity
• unobstructed free-draining outflow from the base of the column
• increased permeant reservoir capacity in order to achieve an equili-
brium flow condition without test interruption
• maintenance of constant-head flow conditions during a test
• reserve slurry capacity for test cases involving deep slurry penetra-
tion
• introduction of slurry after determination of baseline hydraulic con-
ductivity without system depressurization
28
-------
VACUUM
SUSPENSION
SYSTEM
INFLOW
PERMEANT
RESERVOIR
OUTFLOW
PERMEANT
RESERVOIR
ia~n / MANIFOLDS
n D*l'l7crT?'S"pq"
00
CONTROL PANEL
J
SAMPLE
SLURRY
RESERVOIR
SUPPLY
Figure 2. Schematic of the slurry test column system.
x-^200 SCREEN
vr
Tn
•;. TEST SAMPLE (SAND)
-v-
-*O*-RING
PORE PRESSURE
PROBE
COLUMN WALL -
FUTER STONE
ALUMINIUM BASEPLATE^
Figure 3. Schematic of the basal outflow assembly used
in the slurry test column
29
-------
• removal of excess (non-penetrated) slurry after seal formation without
system depressurization
• reduction of probe concentration effects on sample homogeneity near
the center of the sample
• reduced clogging of basal filters due to migration of fines during
the test
The above list includes only the major necessary equipment modifications
related to the proof-testing phase of the work. Table 3 lists specific
modifications made to the column system during the course of this study.
Procedural modifications which relate to these equipment changes are not
specifically itemized, but are incorporated as part of the final test proce-
dures in Appendix A.
In all cases, hydraulic conductivity data were calculated from several
parameters measured during the test. These parameters included the pressure
differential between any two pore pressure probes, the physical distance
between the probes, and the volume flow-rate through the sample (discharge
per unit time). Pore pressure measurements were taken at various times
during permeation. Each set of pore pressure probe readings constitutes a
single "run".
TABLE 3. MAJOR PROOF-TEST MODIFICATIONS - SLURRY TEST COLUMN
* Redesigned basal filter/collector assembly (3)*
* Installation of second permeant reservoir and bypass plumbing
* Installation of pressurized slurry reservoir
* Spring suspension of primary and secondary permeant reservoirs
* Installation of compressed C02 deairing system
• Installation of additional pore pressure probes (2)
• Shortened pore pressure probe length
*Number in parentheses indicates number of times the reported modifica-
tion was m'ade.
Figure 4 shows typical baseline pore pressure distributions (probe
readings) during steady flow for the +40 and +200 sands under roughly equiva-
lent hydraulic gradients. In each case, the relatively uniform pressure
distribution indicates a nearly constant rate of head loss through the sample
prior to the introduction of slurry. The uniform pressure distribution also
30
-------
reflects a high degree of sample homogeneity. After development of a slurry
seal, the steady flow pore pressure distribution for the +200 sand might
appear as shown in Figure 5b. The high rate of head loss over the 6.35 mm
(0.25 inch) distance from 12.7 to 19.05 mm (0.5 to 0.75 inches)sample depth
defines the location, thickness, and gradient across the seal, from which its
hydraulic conductivity can be computed.
The steep decline in pore pressure shown in Figure 5b is typical of the
surface filtration or surface "filter cake" type of slurry seal. The inden-
tification of seal thickness and pressure differential in this case are reason-
ably straightforward. The situation is somewhat more complex where signifi-
cant slurry penetration is involved. Figure 5a represents a typical pore
pressure distribution observed for deep slurry penetration in the +40 sand.
The more gradual rate of head loss through the sample represents gelled or
semi-gelled slurry in the voids of the sand as opposed to the thin dense seal
observed on the surface of the +200 sand.
BASELINE DISTRIBUTIONS: -t-*O * -+-2OO SAND
TOP OF SAMPLE
12 16 2O
COLUMN DEPTH (IN)
2B
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
Figure 4. Typical Baseline (no slurry seal) pore pressure
distributions for fine (+200) and medium to
coarse (+40) sands used in this study.
31
-------
12
POST PENETRATION DISTRIBUTIONS *4O SAND
(a)
1 1
1O H !•••••••! l
8
7
•
S
4
TOP OF SAMPLE
2 4
COLUMN DEPTH (IN)
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
POST PENETRATION DISTRIBUT1ONs*2OO SAND
K-TOP OF SAMPLE
COLUMN DEPTH (IN)
Figure 5. Typical initial pore pressure distributions after formation
of slurry seals in the a) +40 and b) +200 sands.
32
-------
In order to standardize the interpretation of pore pressure plots for
the different sands tested in the column, a graphical construction technique
shown in Figures 6a and 6b was used. In both cases, Ls is the length of
the seal, and A?S "is the pressure drop across the seal. With these two pa-
rameters, the hydraulic gradient across the seal is computed as:
= 27.72 (lVP.y)
where: Px = total pore pressure at probe x, kPa or psi (top of seal)
Py = total pore pressure at probe y, kPa or psi (bottom of seal)
Ls = distance between probe x and probe y, mm or inches
For steady-state flow conditions (constant discharge rate Q, and a con-
stant sample cross-sectional area A), the permeability of the slurry seal was
then calculated as:
With pore pressure data for each of the probes, it was also possible to
compute the permeability between any two pore pressure probes within the
overall slurry seal.
Small-Scale Tests
The experimental system for the long-term investigation of 'surface
filtration seals in various sands was much simpler in design than the slurry
test column. The system comprised several modified rigid wall permeameters
fabricated from transparent acrylic tubing. A schematic of a typical "short"
column is presented as Figure 7.
The major difference between this equipment and the large column was the
absence of pressure transducers and probes to measure pore pressure dissipa-
tion through the slurry seal and changes in the pressure distribution over
time. Moreover, without pore pressure data to accurately define the thickness
of the seal and the amount of pore pressure dissipated across the seal, it was
not possible to compute a hydraulic conductivity value for the seal. Instead,
seal effectiveness over the duration of the tests was determined simply by
measuring volume flow rate.
33
-------
SEAL THICKNESS &i HEAD LOSS
Z 4
COLUMN DEPTH (IN)
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
SEAL THICKNESS &c. HEAD LOSS
(b)
HEAD LOSS THROUGH
SLURRY SEAL
COLUMN DEPTH (IN)
Figure 6. Graphical construction used to determine effective seal
length (Ls) and pressure head loss (APS) across the
seal for a) +40 and b) +200 sands
34
-------
Bleed Valve
Gasket
1* PVC Plate
2 «—Inflow
Typical
'I _ Surface Seal
12'
Acrylic Tube
Porous Stone
J—»Outf low
note: I inch = 25.4 mm
Figure 7. Schematic of "short" column used to evaluate
long-term stability of surface filtration seals.
35
-------
RESULTS AND DISCUSSION
Slurry Test Column
The general test conditions modeled for each sand studied are presented
in Table 4. A single test involved several permeability runs. A run is
defined as a set of pore pressure measurements and the corresponding time and
flow measurements. Although numerous tests involving multiple runs were
attempted, only those considered valid in describing the range of results
observed are included.
The first page of Table 4 presents the type of sand tested and condition
(saturated or unsaturated); the baseline or initial hydraulic conductivity of
the sand K^, prior to slurry penetration; the slurry driving (seal formation)
pressure and the seal permeation pressure; and the slurry seal formation
time. In addition, the total slurry penetration and the hydraulic conductiv-
ity runs for each test are also presented.
A total of seven valid tests were performed on the two sands used in the
slurry test column. Test Nos. 1 and 7 performed on the +40 sand, exhibited
the "rheological blocking" case. Tests Nos. 2 through 6, performed on +200
sand, demonstrated the "surface filtration" case although a minor amount of
penetrated slurry was observed in Tests Nos. 2 and 3.
Measured parameters for the various slurry seals developed in each test
case appear on the second page of Table 4. The measured slurry seal thickness
(Ls), head pressure loss accross the seal (APg), hydraulic conductivity of
the.js.e.a-1 "tKs) , and cumulative permeation time are reported along with the
ratio of baseline (sand alone) and slurry seal hydraulic conductivity
A review of the data indicates that the major differences between the
various tests were the type of sand, the slurry seal formation pressure, and
the permeation pressure for the various runs after initial seal formation.
These variations in conditions produced differences in total slurry penetra-
tion, seal thickness, and the performance of the seals under different perme-
ation pressures.
Differences such as the much greater slurry penetration into the +40
sand (Test Nos. 1 and 5) than into the +200 sand for the same driving pressure
were anticipated. Figure 8 illustrates the difference. Slurry penetration
to a depth of 1.22 m (48 inches) was visually observed in the +40 sand for a
variable driving pressure gradually incremented up to 34.48 kPa (5 psi). The
+200 sand exhibited no slurry penetration for the same driving pressure.
The +40 sand used in Test No. 1 displayed a gradual pore pressure .drop
over the length of the sample as shown in Figure 8a. A surface seal .was not
formed on the top of the +40 sand. The depth of penetration was limited by
the friction that developed between the slurry and the sand grains. Once
the slurry driving pressure was balanced by frictional resistance, the pene-
tration ceased allowing the slurry to gel within the voids of the sand.
Conductivity ratios reported in Table 4 for this test indicate a general
deterioration of the seal over the first 33 hours, followed by dramatic
36
-------
TABLE 4. TEST CONDITIONS AND RESULTS OF THE SLURRY TEST COLUMN
TEST
NO.
1
2
3
4
5
6
7
MATERIAL
&
CONDITION
+40
Saturated
+200
Unsaturated
+200
Saturated
r
+200
Saturated
+200
Saturated
+200
Saturated
+40
Saturated
BASELINE SAND
HYD. COND.
Kb
(cm/sec x 10"7)
500.000
240,000
170,000
264,000
204,000
157,000
782,000
SEAL
FORMATION/
PERMEATION
PRESSURE
(psi)
2.0-5.0/
1.0-2.5,10.0
5.0/
5.0-10.0
5.0/
5.0-10.0
5.0/
2.5
5.0/
5.0
10. 0/
10.0
1.35/
1.35
SEAL
FORMATION
TIME
(hrs)
20.0
5.0
5.0
5.0
5.0
5.0
5.0
24.0
TOTAL
SLURRY
PENETRATION
(In)
42-
48
0.0
0.0
0.0
0.25
2.5
2.5
3.0
3.0
3.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
52.0
(full depth)
RUNS
REPORTED
1
2
3
4
1
2
3
4
5
6
7
8
9
1
2
3
4
5
6
7
3
9
1
2
3
4
5
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
6
11
15
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
37
-------
TABLE 4. (continued)
TEST -
RUN
1-1
1-2
1-3
1-4
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
3-9
4-1
4-2
4-3
4-4
4-5
5-1
5-2
5-3
5-4
5-5
5-6
5-7
6-1
6-2
6-3
6-4
6-5
6-6
6-7
7-1
7-6
7-11
7-17
CUMULATIVE
PERMEATION
TIME
(hours)
2.25
8.0
33.0
52.0
2.57
3.57
4.57
21.07
21.57
22.07
22.57
23.07
26.07
1.36
2.36
3.36
4.86
19.11
21.36
24.36
27.86
45.36
3.25
69.25
447.50
543.50 •
733.50
24
48
96
120
144
168
192
26.8
48
72
96
120
144
167
15.5
233.0
544.25
830.0
SEAL
THICKNESS
Ls
(in.)
42.5
42.5
42 5
42.5
0.125
0.125
0.125
0.625
1.625
1.630
2.125
2.125
2.125
0.125
0.125
0.123
0.138
0.130
0.130
0.175
0.175
0.130
..-0.2.5
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
34.0
38.3
34.8
34.8
HEAD LOSS
ACROSS
SEAL
A PC
(PS*)
1.99
2.09
2.14
8.11
6.1
6.0
5.9
4.2
6.1
9.1
9.1
9.1
5.5
5.0
5.4
5.4
5.6
7.2
8.6
8.6
8.5
8.4
2.1
2.8
2.8
2.6
3.0
5.0
4.8
4.2
4.8
5.4
4.3
5.0 _j
• 9.9
3.7
9.5
9.4
9.5
9.4
9.1
1.5
1.6
1.4
1.6
SEAL
HYD. CONO.
KS
(era. sec x 10-7)
100.0
122.2
171.3
647.9
0.19
0.39
0.20
508.0
1720.0
5450.0
10200.0
10400.0
38300.0
no meas.
flow
0.23
0.2S
0.11
0.14
0.081
0.071
0.072
0.035
1.155
0.376
0.299
0.304
27.48
0.231
0.207
0.197
0.172
0.134
0.207
0.168
0.150
0.155
0.140 '
0.141
0.140
0.142
0.114
8101.0
5565.0
2823.0
1742.0
CONDUCTIVITY
RATIO
KbAs
5000
4090
2920
770
1.256,490
617,570
1,214,500
470
140
44
24
23
6
.
716,740
593,390
1,527,250
1,161,010
2,066,900
2,346,470
2,319,470
4,760.900
228,530
701.490
882,030
869,040
9,610
881,970
986,940
1,036,060
1,183,980
1,109,900
986,940
1,214,290
1,048,200
1,013,620
1,123,600
1,110,560
1,117,680
1,105,370
1,374,900
100 '
140
280
450
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
38
-------
TEST 1 — 4
+4O SANOl SATURATED
(a)
V
TOP OF SAMPLE
2O
60
COLUMN DEPTH (IN)
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
TEST 5 - 1
+2OO SANDS SATURATED
(b)
1
\
s
1
•*• TO
I
i
1
P OF SA
MPLE-
COLUMN DEPTH (IN)
Figure 8. Measured initial pore pressure distributions for slurry seals
formed under a driving pressure of 5.0 psi in: a) +40 sand
and b) +200 sands.
39
-------
deterioration after the permeation pressure was increased to 68.95 kPa
(10.0) psi.
Test No. 7 was also performed in saturated +40 sand, but under reduced
seal formation and permeation pressures. The pore pressure plot for Run 6 of
this test is shown in Figure 9. The approach in this case was to establish
the seal using the formation pressure that would produce a full-depth rheo-
logical blockage within the fixed length of the column. A seal formation
pressure of 13.79 kPa (2.0 psi), equivalent to that used in Test No. 1, was
initially applied. This pressure produced penetration in excess of the
column length (slurry discharge observed from base of column), being too
great to allow a Theological blockage to form. The formation pressure was
then incrementally reduced until slurry penetration ceased. This threshold
pressure is reported as the seal formation pressure for this test on the first
sheet of Table 4 (9.31 kPa = 1.35 psi). Differences between this pressure/
penetration depth relationship and that observed for Test No. 1 are believed
to be due to the much greater baseline permeability (Kt>) of the +40 sand in
Test No. 7 as compared to Test No. 1.
Several permeability runs were performed at a permeation pressure equal
to the threshold seal formation pressure of 9.31 kPa (1.35 psi). Results
were very different from those observed in Test No. 1 in two respects.
TEST 7 — 6
•HO SAND: SATURATED
8O
COLUMN DEPTH. INCHES
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
Figure 9. Measured pore pressure distribution for Theological
blockage seal in +40 sand after 233 hours (Test 7, Run 6).
40
-------
First, the hydraulic conductivity of the seal (Ks) was much greater (factor
of 81 initially), and second, there was no indication of seal deterioration,
(increase in hydraulic conductivity) even with extended testing time relative
to Test No. 1 (830 hrs. vs. 52 hrs.). The hydraulic conductivity of the seal
in Test No. 7, although high relative to test No. 1, was gradually reduced
over the duration of the test by a factor of almost 5.
Test No. 2 was performed on unsaturated +200 sand. The sample was
initially saturated and the baseline permeability determined. The sample was
then drained prior to slurry introduction and seal formation according to
standard procedure. Figure 10 presents the pore pressure plot for Test No. 2,
Run 1. As in other tests performed on the +200 sand, only a surface filtra-
tion seal was observed. The surface seal in this test was eroded from beneath
in less than a day due to piping at the soil interface. The rate of seal
deterioration after the initial breach increased as the seal permeation pres-
sure was gradually increased to 68.95 kPa (10.0 psi).
Test No. 3, performed on saturated +200 sand involved non-penetrated
residual slurry which was allowed to remain in the column after the surface
seal had formed and during the various permeability test runs. The excess
slurry progressively increased the thickness of the surface seal during water
TEST 2 - 1
+2OO SAND; UNSATURATED
V)
a.
Ui
at
I
u
cc
0.
UJ
ac.
o
a.
1 _
•7
I
1
1 -
I
1
1
L
_ •"
~~S — |
TOP
1 D
OF
-a — E
SAMPLE
a — B.
~~B 1
— a —
B-
1
O.-i O.8 1.2 1.6 2
COLUMN DEPTH (IN)
2.4
2.8
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
Figure. 10 Initial pore pressure distribution for the surface
filtration seal in unsaturated +200 sand (Test 2, Run 1)
41
-------
permeation, producing a decreasing flow rate throughout the test. The accu-
mulation of gelled slurry was visibly different in color from the initial
seal, presumably due to differences in density. Seal thickness values (Ls)
reported in Table 4 are for the initial (dense) seal. The permeability of
the initial seal (Ks) decreased under the initial permeation head of 34.48 kPa
(5 psi) and decreased at an accelerated rate after the permeation head was
doubled. In contrast to results for Test No. 2, no indication of seal dete-
rioration was observed over the two day test period. This is most likely
attributable to the residual slurry which had the effect of healing incipient
breaches as they developed.
Test No. 4 was also performed on saturated +200 sand using the residual
slurry approach, but at a permeation pressure that was half of the slurry
driving pressure. These conditions represent the field case where the perme-
ation head on the seal due to groundwater conditions is less than the seal
formation head during cutoff wall construction. The hydraulic conductivity
of this seal decreased gradually over the first 23 days of the test, followed
by relatively rapid deterioration over the ensuing seven days. The test was
terminated after a total of 30 days due to cracking of the seal thought to be
caused by minor pressure fluctuation in the column.
Excess residual slurry was not involved in Test No's. 5 and 6 on saturated
+200 sand. In Test No. 5, the seal formation pressure and permeation pressure
were both 34.48 kPa (5.0 psi). In Test No. 6, these values were increased to
68.95 kPa (10.0 psi). Typical pore pressure plots for both tests are present-
ed in Figure 11. Although the seal formation and permeation pressures differ
between the two tests, the seal formation time was held constant at 5.0 hours
in each case. This resulted in surface seals of the same total thickness,
but of different permeabilities. Table 4 reports seal hydraulic conductivity
(Ks) values for both tests. The value for the final run of Test No. 6 is
roughly 47% lower than the final value for Test No. 5. This reduction appears
to be related to the doubling of seal formation pressure in Test No. 6, its
lower initial baseline permeability relative to Test No. 5 (about 30% lower),
or some combination of both effects. Based upon these data seal thickness is
a function of formation time, whereas, the hydraulic conductivity of the seal
is inversely porportional to both seal formation pressure and the initial
baseline hydraulic conductivity of the sand. (Similar relationships were
observed for the +40 sand in Test No's. 1 and 7.)
Although conductivity ratios reported in Table 4 suggest that only two
of the five tests on the +200 sand suffered deterioration of the seal during
permeation, all test except Test No. 3 eventually failed. The failures
appeared to be of an erosion (piping) nature, originating at the interface
between sand and the seal. These failures are believed to be due to slight
pressure changes within the system caused by temperature changes and/or
supply pressure changes from day to night and vice-versa. Such pressure
fluctuations would cause the sand and the acrylic column to expand and con-
tract by different amounts. Such disturbance would cause micro-cracks in the
seal and subsequent widening of the cracks via erosion. Failures of this
nature are believed to be equipment and scale-related and should not occur in
the field unless the soil-bentonite wall undergoes extreme lateral consolida-
tion which would subject the slurry seal to bending stresses.
42
-------
L
IS
12
11
1O
a
a
7
a
a
TEST 5 - 1
+2OO SANOl SATURATED
(a)
OF SAMPLE
COLUMN DEPTH (IN)
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
TEST 6 - 1
•4-200 SAND: SATURATED
(b)
TOP OF SAMPLE
COLUMN DEPTH (IN)
Figure 11. Initial pore pressure distributions measured for
surface filtration seals on the +200 sand under net
driving pressure of a) 5.0 psi and b) 10.0 psi.
43
-------
Small Scale Tests
The test conditions modeled for the sands studied in the rigid wall
permeameters are outlined in Table 5. Ottawa Sand of different gradations
(+60, +80, and +100) was tested under seal formation and permeation pressures
of 34.48 and 68.95 kPa (5 and 10 psi). The same general procedures and the
same 5% slurry used in the slurry test column were also used in the small
scale tests. The major difference between these tests and those performed in
the column was the absence of pressure measurements to accompany each run.
Instead, a run was defined as a volume flow rate measurement made with a
buret. Evaluations of relative seal effectiveness were made by charting the
increase or decrease in flow as a function of time.
The slurry was colored with dye to aid in observing slurry penetration
and changes in the condition of the seal during the test. The primary objec-
tive of this portion of the testing was to determine if the repeated erosion
(piping) failures observed in the column were system-related by observing the
long-term behavior of similar seals in conventional permeameters. Accord-
ingly, the small-scale tests were conducted over a time period of more than
six months.
TABLE 5. CONDITIONS FOR SMALL-SCALE TESTS
Test
1
2
3
4
5
Gradation
+60
+60
+80
+100
+100
Condition
Sat.
Sat.
Sat.
Sat.
Sat.
Baseline
K[j
(cm/sec)
4.1 x lO'3
4.2 x 10-3
3.4 x lO'3
3.3 x 10-3
3.1 x ID'3
Seal Formation
Pressure
(kPa/psi)
34. 48/ 5.0
34. 48/ 5.0
34.487 5.0
34. 48/ 5.0
68.95/10.0
Seal Permeation
Pressure
(kPa/psi)
• 34. 48/ 5.0
34.487 5.0
34. 437 5.0
34. 487 5.0
63.95/10.0
An abbreviated record of volume flow-rate versus time (first one or two
months only) is presented for each of the five tests in Figure 12(a) through
12(e). In all cases, initial seal formation was accompanied by 25.4 mm
(1 inch) or more of penetration of relatively dilute (filtered) slurry. Re-
sidual slurry present after initial seal formation was removed in all cases
so as not to artificially prolong the life of the surface seal during the
tests.
During permeation, the zone of penetrated dilute slurry typically was
observed to detach itself from the surface seal and move downward through
each sample as a discrete slug, eventually emptying into the collection buret.
There was no noticeable disturbance of or change in the surface seal during
the tests. This observation was supported by the volume flow rate measurements
44
-------
(a)
0
0
UJ
I-
J
IL
111
2
3
J
o
10
-3
SAMPLE No. 1
PRESSURE: 5.O PS!
SAND: -+-6O OTTAWA
1O
-4
1O IS 2O 25 3O 35
RUN TIME (DAYS)
0)
\
0
0
UJ
h
3.
O
J
u.
UJ
z
D
O
1O
(b)
10
-3
SAMPLE No. 2
PRESSURE: 5.O PSI
SAND: -i-SO OTTAWA
10
1O 15 20 25 30 35
HUN TIME (DAYS)
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
Figure 12. Measured volumetric flow rates for surface filtration
seals on a) +60 Ottawa sand and b) replicate sample.
45
-------
B , _
\ 10
0
0
111
3
U.
HI
D
O
1O
(c)
SAMPLE No. 3
PRESSURE: 5.O PSI
SAND: -t-8O OTTAWA
10 15 20 25 30 35
RUN TIME (DAYS)
09
N
U
U
UJ
2
O
1L
UJ
O
10
-2
(d)
10
-3
10
^o
<%O<£> O/
oooo 4
SAMPLE No. 4
PRESSURE: 5.O PSI
SAND: -MOO OTTAWA
1O
20
3O 4O
RUN TIME (DAYS)
50
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
Figure 12. Measured volumetric flow rates for surface filtration
seals on c) +80 Ottawa and d) +100 Ottawa sands.
46
-------
0
u
UJ
h
3
U.
UJ
D
J
O
.-2
(e)
.-3
10
SAMPLE No. 5
PRESSURE: 1 O.O PSI
SAND: -MOO OTTAWA
Figure 12.
10 15 SO 25 3O 35 4O
RUN TIME (DAYS)
note: 1 psi = 6.895 kPa
Measured volumetric flow rate for surface filtration
seal on +100 Ottawa sand for seal formation/permeation
pressure of 10.0 psi.
which typically achieved equlibrium in a short period of time and remained
constant throughout the test period. Based upon these data, it is concluded
that the frequent cracking and piping failure of slurry seals tested in the
slurry test column was due to system-related pressure fluctuations as describ-
ed earlier and that slurry seals permeated in situ at or below the seal forma-
tion pressure are stable in the long-term.
Comparative Breakthrough Analysis
Presented in this section is a comparison of breakthrough times (time
for the first drop of permeant to pass through the barrier) based on test
results for the surface filtration and the Theological blocking types of
slurry seals evaluated in the slurry test column.
Figure 13 is a schematic of two typical soil-bentonite walls, showing
the expected zone of slurry penetration and seal formation in the +40 and
+200 sands. Deep slurry penetration accompanied by Theological blockage
occurs in the +40 sand, whereas a surface filtration seal is shown for the
+200 sand. In both schematics, the soil-bentonite backfill is assumed to be
the same, having a hydraulic conductivity of 1.0 x 10~? cm/sec. The depth of
slurry penetration and the hydraulic conductivity of the seal in each case
are based upon data in Table 4 for Test No. 7, Run 6 (+40) and Test No. 5,
Run 7, (+200). These two cases' were selected as the basis of comparison
47
-------
03
+•40 SAND TEST 7-6
S/B BACKFILL
ZONE OF
PENETRATED
SLURRY
UU-36"—L-40"
k. =5.57 x 10 cm/sec
9
-7
k /b=1-0 * 10 cm/sec
+200 SAND TEST 5—7
S/B BACKFILL
V;
Ah
SURFACE
FILTER CAKE,
0.25" I 35.5"| 0.25"
k =1.68 x 108cm/sec
9
k ,u=1-0 x 10 cm/sec
s/b
note: 1 inch = 25.4 mm
Figure 13. Idealized conditions after construction of
cutoff walls in the +40 and +200 sands.
-------
since they represent the conservative end of the range of test results for
each sand and also represent the observed conditions at roughly equivalent
cumulative permeation times (233 hrs. versus 192 hrs.).
For comparative purposes, the same pressure head (Ah) is used in each
case to compute a breakthrough time. The overall thickness of the barrier in
the +40 sand is 2.95 m (116 inches) with a weighted average or effective
hydraulic conductivity k = 3.22 x 10 cm/sec for flow perpendicular to the
direction of slurry seal/soil-bentonite stratification. Note that the thick-
ness of the barrier in the +40 sand is based upon a slurry penetration depth
of 1.016 m (40 inches) on each side of the trench instead of the total pene-
tration depth of 1.32 m (52 inches) reported for this test in Table 4. The
reason is that of the 1.32 m (52 inches) of total penetration, the pore
pressure plot for this run indicates that only the upper 1.016 m (40 inches)
of the slurry seal were effective in dissipating the applied head pressure
(see Figure 9).
Assuming an in-service head differential of 0.914 m (3.0 ft) across each
barrier, breakthrough times (T^) are computed as follows:
Velocity in length per unit time is equal to hydraulic conductivity times
gradient or:
v ~ ki
from which;
L/T = k Ah/L
L = Tk Ah/L
L2 = Tk Ah
T = L2/(k Ah)
For the +40 sand:
Tb = (thickness)2/(kp x Ah)
= (294.6 cm)2/(3.22 x IQ-f
cm/sec x 91.4 cm)
= 93.5 years
For the +200 sand:
Tb = (91.4 cm)2/(9.36 x 10'8 cm/sec x 91.4 cm)
= 31.0 years
The effectiveness of the wall in the +200 sand based upon a breakthrough
criterion would be about one-third as much as that of a similar wall construct-
ed in a deposit of +40 sand. If the surface seal in the +200 sand was removed
during the soi 1 -bentonite backfilling phase of construction, the difference
in breakthrough times would be slightly greater. Moreover, the demonstrated
relationship applies to the entire submerged portion of the wall regardless
of depth because the net head differential (Ah) is everywhere the same below
the prevailing groundwater table. The only factor which might modify the
reported relationship would be a change in the actual hydraulic conductivity
of the two types of seals as a function of depth. Presumably, however, such
differences would remain proportional at all depths due to the equivalent net
head producing the slurry seal in each case.
49
-------
There is no question that a difference in relative effectiveness by a
factor of three is significant. The influence of the soil-bentonite backfill
itself on this relationship is, however, of particular interest.
For the soil-bentonite alone without any slurry seal:
Tb = (90.2 cm)2/(l x 1(T7 cm/sec x 91.4 cm)
= 28.2 years
If we consider the thickness of the two slurry seals and their measured
hydraulic conductivities alone (no soil-bentonite backfill), the respective
breakthrough times are:
for the +40 sand;
(80 in. x 2.54 cm/in)2
'b - 5.57 x 10-* cm/sec x 91.4 cm
= 9.4 days
for the +200 sand;
(0.5 in. x 2.54 cm/in)2
>8 x 10'
12.2 days
Tb - 1.68 x 10'8 cm/sec x 91.4 cm
These data demonstrate that the effectiveness of the two slurry seals
themselves are not greatly different, suggesting that the soil-bentonite
backfill regulates the effectiveness of the slurry seal when the seal is
viewed as a part of the complete barrier. In other words, under steady-state
conditions, the backfill serves to regulate the velocity of flow through the
whole barrier and the slurry seal serves to extend the zone of influence of
the backfill, by an amount equal to the thickness of the seal. The much
greater breakthrough time in the case of the +40 sand is a result of the
greatly extended zone of influence of the backfill.
The above discussion suggests that any practical means of promoting
slurry penetration into the walls of the cutoff trench prior to backfilling
would be worthwhile, since a deep penetration seal would be secure (not
subject to damage or removal during construction) and would greatly extend the
zone of influence of the backfill. Although such is considered to be true,
it should be emphasized that the key to this relationship is the backfill
itself and that the slurry seal, regardless of its thickness or type, ,will
be of limited consequence where hydraulic defects (windows) exist through the
soil-bentonite backfill. In other words, the deep penetration seal greatly
enhances the cutoff capabilities of a backfill without defects, but will pro-
vide limited help near any defects.
50
-------
SECTION 6
MODEL CUTOFF WALLS
The effectiveness of a soil-bentonite cutoff wall as a hydraulic barrier
may be measured by its in situ hydraulic conductivity. Generally, the lower
the hydraulic conductivity of the soil-bentonite backfill in the wall, the
more effective the cutoff wall will be. However, the ability of the wall to
function as a viable hydraulic barrier does not depend solely on the initial
hydraulic conductivity of the soil-bentonite but other factors as well.
Several potential variables that could influence the effectiveness of a cutoff
wall are: wall thickness, soil-bentonite composition (soil type, bentonite
content and slump), bentonite type, type of permeant, overburden pressure, and
the hydraulic gradient across the wall. In addition, even a wel1-designed
soil-bentonite cutoff wall, and one which was built with a high level of
quality control, cannot be guaranteed to perform perfectly. Construction
defects such as inhomogeneity due to improperly mixed backfill, or pockets of
trench-wall soils or slurry entrapped in the backfill during construction may
result in "windows" in the wall. A great number of small windows may also
develop due to the shrinkage and cracking of the backfill as a result of long-
term exposure to chemical permeants.
In this testing program a laboratory-scale test tank (slurry wall tank)
was built to investigate the effects of overburden pressure and hydraulic
gradient on the hydraulic conductivity of model soil-bentonite walls, using
deionized water as permeant. In addition, the potential . for closing or
healing model windows in the form of slurry pockets or large cracks by con-
trolled surcharge pressure was also tested.
The testing program started with a preliminary testing phase. Its
purpose was to evaluate the basic functioning of the tank system and was
conducted by filling the tank with sand only. As a result of this testing,
several changes in equipment, procedures and test soils were made. The first
test series involving a soil-bentonite wall is termed "Sequence 1", in which
two test cases were completed. After the Sequence 1 tests, substantive
changes in materials and methods were again made and a new and independent
test sequence, termed "Sequence 2", was undertaken. Table 6 identifies the
primary variables and test conditions for both test sequences. Following
Sequence 2 tests, a window-closing tests was conducted on the same soil-
bentonite wall that was used in Sequence 2 permeability tests.
Although indicative of project focus, Table 6 and the window-closing
tests do not fully reflect the scope of the research effort. Namely,
in support of the primary tests' completed, considerable laboratory testing
was performed to identify and design suitable test materials and determine
51
-------
TABLE 6. TEST CONDITIONS FOR TEST SEQUENCES 1 AND 2
T
E
S
T
S
E
Q
U
E
N
C
E
#1
T
E
S
T
S
E
Q
U
E
N
C
E
#2
CONSTANTS
Wall thickness = 4 in.
S/B backfill design:
C.H. #2 soil
2% ordinary
bentonite
7" slump
In situ soi 1 :
Unwashed fine sand
Permeant: H20
Wall thickness = 4 in.
S/B backfill design:
C.H. #2 soil
1% ordinary
bentonite
7" slump
In situ soil : +200
fine sand (washed)
Permeant: HgO
TEST
NO.
(a)
(b)
(a)
(b)
(c)
(d)
(e)
(f)
(g1)
(9)
VARIABLES
OVERBURDEN
PRESSURE
ON S/B WALL
(psi)
4.5
4.5
6
6
-
12
12
12
12
24
HYDRAULIC
PRESSURE
APPLIED
(psi)
3
6
3
6
-
3
6
12
3
3
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
their engineering properties, both before and after use in the tank. Tables
7 and 8 document the number and type of characterization tests performed
specifically in support of research on model cutoff walls.
52
-------
TABLE 7. SUMMARY OF PROJECT TESTING-SLURRY WALL TANK
TEST SEQUENCE
NO. OF
PLANNED
PRIMARY
TESTS
NO. OF
COMPLETED
PRIMARY
TESTS
NO. OF
BASELINE
CHARACTERIZATION
TESTS
NO. OF
POST-TANK
EVALUATION
TESTS
1
2
9
9
2
6
54
85
20
49
TABLE 8. BASELINE CHARACTERIZATION AND POST-TANK EVALUATION TESTS
TEST ITEMS
Sieve analysis
Atterberg limits
Specific gravity
Hydraulic conductivity:
rigid wall
flexible wall
Water content
Consol idation
Compaction
Slurry penetration
Slurry properties
Unit weight
Sand placement density
Vane Shear Strength
Slump
BASELINE CHARACTERI
TEST SEQUENCE
1
9
3
1
-
2
26
1
2
1
6
1
2
-
-
ZATION
2
12
4
1
6
3
40
1
2
4
4
1
1
-
6
POST-TANK EVALUATION
TEST
1
8
_
-
-
-
6
_
-
-
-
6
-
-
-
SEQUENCE
2
-
'
-
-
14
_
-
-
-
6
-
29
-'
Subtotal
54
85
20
49
53
-------
EQUIPMENT DESIGN AND PRELIMINARY TESTING
Design
The main objectives of this experimental study were to evaluate the in
situ hydraulic conductivity of laboratory-scale soil-bentonite cutoff walls
under simulated field overburden and hydraulic conditions, and to investigate
the potential for closing of artificially introduced "windows" in such walls
by simulated field overburden pressures.
To achieve these aims, the experimental tank system comprised four main
components: a container within which a model soil-bentonite cutoff wall could
be constructed using simulated field construction methods, a loading device
which would permit the application of simulated field overburden pressures on
the soil elements of the model, a pressurized permeant supply system which
would allow permeation of deionized water through the model soil-bentonite
wall at simulated field hydraulic gradients, and devices which would allow
the measurement of flow across the soil-bentonite cutoff wall. A schematic
of the tank system which illustrates the main components is presented as
Figure 14.
PERMEANT RESERVOIR
PORE PRESSURE
(typ.) J S (L i £Ht
PVC MEMBRANE
V8A I \ I BEARING!
SURCHARGE
BLADDER
SOIL-BENTONITE
BACKFILL
DRAINAGE PANEL
DRAIN
AIR PRESSURE LINE
PERMEANT FLOW LINE
Figure 14. Schematic of the slurry wall tank system.
54
-------
Container—
In the early stage 'of design, two different shapes of containers, a
rectangular box and a circular tank, were proposed and evaluated. The
rectangular box would allow the construction of a straight model cutoff wall
inside the box, while the use of the circular tank would allow the construction
of an annular wall. The straight wall would be easier to construct, but it
would introduce discontinuities at each of its ends. These could result in
end friction and end leakage which would complicate the analysis and affect
the validity of test results. On the other hand, an annular model wall
is more difficult to construct due to the small radius of curvature of the
wall, but its use eliminates the discontinuity problem and the wall is a
truer representation of the field conditions. For these reasons, a circular
tank was chosen as the container for the model soil-bentonite cutoff wall.
With respect to specific requirements, the tank would have to be large
enough to permit the construction of a soil-bentonite wall of at least 101.6
mm (4 inches) thick and still allow sufficient space for model in situ soils
to surround it and provide lateral support to it. However, the tank would
also have to be small enough for easy handling and good accessibility. To
meet these criteria, a circular tank 1.12 m (44 inches) in diameter, 3.51m
(138 inches) in perimeter, and 0.71 m (28 inches) in height was chosen. Thus
within the tank, a 101.6 or 152.4 mm (4 or 6 inch) thick annular soil-benton-
ite wall would be sandwiched between a 304.8 mm (12 inch) diameter center
core of sand and a 304.8 or 254.0 mm (12 or 10 inch) thick outer ring of
sand.
The tank body would have to be made of chemically resistant material to
allow future testing with selected chemicals. It also would have to withstand
a hoop stress of approximately 620.6 kPa (90 psi), the result of the combina-
tion of a design effective overburden pressure of 413.7 kPa (60 psi) and a
hydrualic pressure of 206.9 kPa (30 psi). The body of the tank was made of
1.59 mm (1/16 inch) thick stainless steel and was reinforced by 9.5 mm x 76.2
mm (3/8 inch x 3 inch) metal bands at the top and at mid-height. Unlike the
thin tank wall, the tank bottom plate and the tank lid had to resist bending
and shear stresses. The magnitude of these stresses were dependent on the
layout of the reaction frame. In order to reduce these stresses, several
steel stringers were placed between both the lid and the bottom plate, and
the wide flange beams of the reaction frame. In order to develop the design
stress of 620.6 kPa (90 psi), both plates were 4.76 mm (3/16 inch) thick.
Reaction Frame—
The layout and the design of the reaction frame was simple but functional .
Its purpose was to hold the tank lid down against uplift pressures exerted by
the soil loading system, and to provide an evenly distributed support system
for both the lid and bottom plate of the tank. Figure 15 shows this frame in
both its side and top views. Both the stringers and the wideflange beams
were assumed to be simply supported. The stringers were designed for the uni-
formly distributed loading from the lid, while the beams were designed for the
point loads transferred from the stringers. The reaction forces from the
beams were developed by eight threaded bars and which tie the upper and lower
beam systems together.
55
-------
1
note: 1 inch = 25.4 mm; 1 psi = 6.895 kPa
Figure 15. Structural components of the slurry
wall tank and reaction frame.
Pneumatic Loading System--
Simulated overburden pressure was provided by a pneumatic loading system
which consisted of a set of three industrial grade innertubes (bladders) that
exerted the desired pressures against the soils in the tank and reacted
against the lid of the tank. This loading system had to provide adjustable,
yet stable pressures to the soils in the range of 3.45 to 413.7 kPa (0.5 to
60 psi). The location of the three bladders in cross section is shown in
Figure 14.
Fluid Transmission System—
The fluid transmission system was designed to facilitate the flow of the
permeant across the model soi1-bentonite wall with control on the gradient of
the flow. The system was equipped with measuring devices to record both
inflow and outflow. The transmission system consisted of two pressurized
permeant reservoirs, connecting tubes and valves, and an outflow collecting
and measuring system. Figure 16 shows the details of a typical permeant
56
-------
reservoir. It consists of a 152.4 mm (6 inch) ID by 660.4 mm (26 inch) tall
PVC (polyvinylchloride) tube with two 31.75 mm (1.25 inch) thick by 196.9 mm
(7.75 inch) square PVC endplates. The plates were tied together with four
15.9 mm (0.625 inch) diameter threaded bars. In the top plate there were two
connections, one for permeant inlet, the other for pressurized air inlet.
Prior to testing, the reservoir was filled up to the desired level with the
permeant through the permeant inlet, which was then closed. Air pressure of
the desired magnitude was then introduced through the air inlet to drive the
permeant into the center core of the model in the tank. The permeant reser-
voir was equipped with a transparent side tube connected to the reservoir at
both, top and bottom, to allow the reading of the permeant levels inside the
reservoir. These readings were then used to compute the inflow quantities
and the hydraulic conductivity of the soil-bentonite wall. The PVC reservoirs
served well with deionized water as the permeant but may need to be replaced
along with the nylon tubing and fittings depending upon the type of chemical
used in future testing. The control valves were made of stainless steel.
The outflow collecting and measuring devices were also made of nylon,
but contained some brass fittings which may need to be changed for certain
future chemical permeants. The valves were all stainless steel.
Surface Membrane and Hydraulic Cutoff—
One of the most important components of the slurry wall tank system was
the surface membrane and two hydraulic cutoffs, one at the top and one at the
bottom of the tank. The surface membrane served as an interface between the
polystyrene loading blocks and the various soil elements of the model and,
when loaded, also functioned to hydraulically isolate the different elements
of the model during testing. In order to achieve a high level of hydraulic
isolation, the surface membrane was forced into the surface of the soil-
bentonite wall to form a v-shaped vertical cutoff for Sequence 1 testing.
This configuration was not entirely satisfactory so the approach was modified
for Sequence 2 and subsequent tests. Figure 17a shows the folded membrane
cutoff used in Sequence 1 and the modified membrane/cutoff system used in
Sequence 2 and subsequent tests (Figure 17b). The modified cutoff consisted
of three concentric semi-rigid PVC rings which were bonded to the PVC membrane.
This triple cutoff was held in place by the applied surcharge pressure during
permeability testing and effectively prevented leakage of permeant across the
top of the soil-bentonite wall.
The bottom cutoff used in both test sequences is shown in Figure 17c.
This cutoff also consisted of two concentric PVC rings (similar to those used
on the modified surface configuration) which were bonded to the stainless
steel bottom of the slurry wall tank. These bottom cutoffs served to hydrau-
lically isolate the bottom of the model wall and also helped maintain the
proper spacing of two large PVC slip forms used in the construction of the
wall.
Control Panel--
The control panel was primarily designed to facilitate the operation and
control of the pressure systems. It housed all necessary pressure gauges and
regulators for the operation of the tank system. Such operations included
the application of overburden pressures to the soil-bentonite wall, the model
57
-------
Air Inlet Permeant Inlet
Sld
tube
ft rfl
I.
I,
^, '
J
\'
a i
nil Mn
/ /
/ 6 — /
7 « '
/ '
/ '
i j
I j
ft «f
!! 1.2!
I ^—
0.28;
23
1.2
! I •
5i/8Hole
23.5 NOTE: All dimensions are In Inches
28
[—13/18
Permeant outlet 5/8 Thread bar
SECTION
6.125-
6.125
BOTTOM PLATE
note: 1 inch = 25.4 mm
Figure 16. Construction details of permeant
reservoir for the slurry wall tank.
58
-------
(a) UPPER CUT-OFF - TEST SEQUENCE .'10.1
EDUCING BUSHING
RUBBER WASHER
STAINLESS STEEL
LOADING PLATE
SAND
S/B
^PERFORATED PIPE CAP
CORE SANO
-v
S/B
SAND
(b) UPPER CUT-OFF - TEST SEQUENCE MO.2
SANO
S/B
REDUCING BUSHING
•0-RING
STAINLESS STEEL
LOADING PLATE
^PERFORATED PIPE CAP
CORE SAND
S/B
SANO
( 0 ) BOTTOM CUT-OFF
.-•.'
S/B
i
TANK B
* CCHE SANO '\ ' '-.':
PERFORATED PIPE CAP-*
mr
STAINLESS
STEEL PIPE.
'•::-. ;~ ^ ••.-•-. •:•-
1 It
S/8
X
,\
PVC CUT-OFF /
BAFFLE -•'
(GLUED TO TANK)
HPT FEMALE ADAPTER
Figure 17. Detail of upper surface membrane and hydraulic
cuttoff for a) Test Sequence 1 and b) Test
Sequence 2; c) bottom hydraulic cutoff.
59
-------
soil in the center core and the outer ring, and the application of driving
.pressure to the permeant reservoir. The control panel comprised a plexiglass
panel supported by a wooden frame as depicted in Figure 18a and 18b.
1
"1
2
3.
P
J
0
1
f>
I
4
2
5
5 v
I
(='
|3
\
-
Q
1
r l
- \
r' a -«d
8
f^
A Q:
•7!
' 8 -
^ C
}\
>A Q:
4,
•T-j
M
j \
i
•*!
Ei
a.
?4
9/^v^3
/ i -^ ^j
vt
O Cr
Q3B Q2B Q1B
i
!
*
^^
J
L|
-
f
b
— 5 —
-100.
60
— reg
i 0<
•^
(se
s
gauge number
(see Appendix B)
note: All dimensions are in inches;
1 inch = 25.4 mm; 1 psi = 6.895 kPa.
Figure 18a. Front elevation of slurry wall tank control panel
60
-------
Figure 18b.
Oak Wood (2x4)
note: 1 inch = 25.4 mm
Side elevation of slurry wall tank control panel
Preliminary Testing
The purpose of preliminary testing was to demonstrate the adequacy of
the tank system for the intended study, and to formulate, evaluate and refine
soil preparation, placement and testing procedures.
Specifically, preliminary testing involved the proof-testing of the
components of the tank system to:
(a) determine the functioning, adequacy, and accuracy of the fluid transmis-
sion system, including the inflow and outflow measuring devices
(b) check if the selected pneumatic bladders (innertubes) exerted' the
required design pressures on the model soil in the tank
(c) calibrate the control panel instrumentation and to determine the accuracy
of the regulators and gauges used
(d) develop a data recording format and procedures to be used for project
testing
(e) establish test schedules and necessary safety procedures.
61
-------
For the preliminary test the entire tank was filled with the selected
model sand only (no soil-bentonite wall was constructed) to determine its
hydraulic conductivity as measured in the tank. Test results were then
compared to those obtained from standard permeameters to demonstrate the
adequacy of the design of the overburden pressure system and the fluid
transmission system, including the inflow and outflow measuring devices.
Specifically, a fine-to-medium sand, containing approximately 2.5% fines
(minus No. 200 sieve size) was selected for use as the model in situ soil.
This sand was first moistened with water, then placed into the tank in loose
152.4 mm (6 inch) lifts and compacted with a vibrating plate compactor. This
filling processs continued until the desired sample height of 558.8 mm (22
inches) was reached. The bulk unit weight of this sand sample was pre-esti-
mated by compacting the sand into a small calibration vessel of known volume.
From the Standard Proctor Compaction Procedure (ASTM D-698) the bulk wet unit
weight of the compacted sand was estimated to be 1682.1 Kg/m-* (105 Ibs/ft-^).
A polyethylene membrane sheet was placed over the surface of the compacted
sand to provide an upper impervious boundary.
Before the start of permeation, a nominal overburden pressure was applied
to the sand and held constant during the test. Oeionized water as the permeant
was then introduced under a chosen head pressure into the center of the tank.
Both inflow and outflow were monitored for several days. Based on the applied
hydraulic head, the inflow and outflow measurements, and an assumed simplified
percolation path and sample area, a conductivity of 1 x 10"^ cm/sec was com-
puted. This value was comparable with those obtained from rigid wall perme-
ability equipment used to test the same sand at the same density.
This test demonstrated that the design of the tank system was basically
sound, that the overburden pressure was properly applied and the fluid
transmission system, including the inflow and outflow, measuring devices,
functioned as intended.
METHODS AND MATERIALS
This section describes the methods and materials used in two hydraulic
conductivity test sequences, referred to as Test Sequences 1 and 2, and in
the single window-closing test performed during this study.
The slurry wall tank system was set up according to the schematic shown
in Figure 14. Except for the tank itself and the components of the reaction
frame that were manufactured under subcontract, all components of the tank
system were fabricated and assembled in-house at the Center Hill Research
Facility.
As described earlier, the objectives of the experimental work were,
first, to evaluate the effects of selected primary variables on the hydraulic
conductivity of .soil-bentonite cutoff walls, and second, to investigate the
potential for the remediation of selected types of artificial defects (win-
dows) in such walls.
To accomplish the first objective, overburden pressure and net hydraulic
62
-------
gradient were selected as the primary variables. For each of three overburden
pressures, three hydraulic gradients were applied, resulting in a total of
nine separate tests within each test sequence. The test plan also involved
baseline characterization and post-tank evaluation tests of the materials
selected for testing. Such data was necessary for the full interpretation of
hydraulic conductivity data generated using the slurry wall tank. The origi-
nal plan additionally incorporated "barrier enhancement" tests wherein dif-
ferent soil-bentonite additives were to be evaluated for improving barrier
effectiveness. Each test involving the primary variables listed above,
however, required extended test time to achieve equilibrium hydraulic condi-
tions, thus precluding the performance of any barrier enhancement tests.
The general experimental approach in all tests involving the two primary
variables consisted of the following basic steps:
1. construct an annular soil-bentonite cutoff wall of uniform thickness
sandwiched between a model in situ sand core and outer ring of sand
2. consolidate the finished wall by the application of a predetermined
surface surcharge load (pneumatic pressure)
3. introduce permeant under a predetermined hydraulic pressure into the sand
core encircled by the soil-bentonite wall
4. measure the equilibrium permeability of the wall in the horizontal
direction under the prevailing surcharge and hydraulic gradient by
measuring both inflow and outflow at prescribed time intervals
5. increase the hydraulic gradient, re-establish equilibrium flow, and
measure the new equilibrium permeability as in step 4 (typically three
cycles)
6. increment the surcharge pressure to consolidate the wall to a higher degree
followed by a repetition of steps 3 through 5 before going to step 6
again (three cycles)
Although testing procedures differed in minor detail in some cases due to
the experimental nature of the tank system, the general test approach remained
the same. A more complete listing of step-by-step model construction and test-
ing procedures is included in Appendix B.
Test Sequence 1
Test Sequence 1 involved the construction and testing of the first model
soil-bentonite wall. Equipment- and experimental procedures were modified
several times during this phase with the result that this test sequence should
properly be considered as an extension of the proof-testing effort.
The intent of this test sequence was to determine the hydraulic conduc-
tivity of a model soil-bentonite wall for a range of three hydraulic gradients
(i) under each of three different surcharge pressures. Effective surcharge
pressures of 31.03, 62.06 and 124.11 kPa (4.5, 9, and 18 psi) were planned,
with hydraulic gradients of 20.8, 41.6 and 83.2 sequentially applied after
full consolidation under each surcharge pressure.
The desired surcharge (vertical consolidation) pressures on the soil-
bentonite wall were achieved by maintaining the overburden pressure at a
63
-------
higher level than the prevailing hydraulic pressure throughout the test.
Figure 19 is a loading plan for Sequence 1 tests l(a) through l(j) (nine
cases). In each case, the effective consolidation pressure shown on the sand
and soil-bentonite elements of the model is the reported overburden gauge
pressure minus the hydraulic guage pressure shown for the same test. For
example, in order to maintain a constant net effective consolidation pressure
of 4.5 psi on the soil-bentonite wall while incrementing the hydraulic pres-
sure (and therefore gradient) in tests l(a), (b), (c), it was necessary to
increment the overburden pressure to the values shown. The same is also true
for tests l(d) through l(f) and test l(g) through l(j). In this manner, the
effect of hydraulic gradient could be observed for the three consolidation
states associated with the three different effective consolidation pressures.
Due to progressive clogging (siltation) of the tank outflow ports,
permeant leakage across the cutoff at the top of the soil-bentonite wall, and
minor leakage at the base of the tank itself, Sequence 1 was terminated after
test l(b) to allow for repairs and needed design modifications. The specific
modifications made in preparation for Test Sequence 2 are summarized in
Table 9.
For both Sequence 1 and Sequence 2 tests (described later), the hydraulic
conductivity of the soil-bentonite wall was computed on the basis of inflow
measurements using the falling head equation as follows:
K = d'L Ln jlfl.
A-t hj_
where: K = hydraulic conductivity (cm/sec)
a = cross-sectional area of permeant reservoir (cm^)
L = length of sample (cm)
A = cross-sectional area of sample perpindicular
to the direction of flow (cm^)
h0 = initial total hydraulic head (cm)
hi - final total hydraulic head (cm)
t = time interval between two readings (sec)
For the slurry wall tank, the following substitutions are made:
a = 186.6 cm2 (cross-sectional area of the permeant
reservoir)
L = 10.16 cm (4 inches) (thickness of soil-bentonite wall)
A = 5513 cirr (circumferential surface area along
centerline of the circular soil-
bentonite wall)
Substituting these values into the above equation, the permeability
equation for the soil-bentonite wall may be reduced to:
K (cm/sec) = 9.55 x 10'5 -^—r Ln
t(hrs)
64
-------
01
Ul
Overburden (c)
Gauge
Pressure (b)
(psi)
(a)
Hydraulic (c)
Gauge
Pressure (b)
(psi)
(a)
Effective
Consol idation
Pressure
(psi )
3 10.5
3 7.5
3 6
15 (f)
9 (e)
6 (d)
(0) (6) 12 (f)
(0) (3) 6 (e)
(0) (1.5) 3 (d)
3 4.5 3
i
' SAND S/B
. • - •
* •* " * 1
.SAND'
6 15
6 12
6 10.5
(0) (6)
(0) (3)
(0) (1.5)
6 9
SAND . S/B
18 (j)
12 (h)
9 (g)
12 (j)
6 (h)
3 (g)
6
CORE
SAND
12 24
12 21
12 19.5
(0) (6)
(0) (3)
(0) (1.5)
12 18
SAND .' S/B
24
18 Calculated*
Hydraulic
15 Gradient
(D
12 83.2
6 41.6
3 20.8
12
:ORE
SAND
TANK
NOTE: Effective consolidation pressure = overburden gauge pressure minus
hydraulic gauge pressure. *Head loss through core and outer ring of
sand assumed to be zero. 1 psi = 25.4 mm.
Figure 19. Loading plan for slurry wall tank Test Sequence 1.
-------
TABLE 9.
INITIAL EXPERIMENTAL CONFIGURATION AND SUBSEQUENT
MODIFICATIONS TO THE SLURRY WALL TANK
SEQUENCE 1
SEQUENCE 2
In situ sand
Soi 1-bentonite
Wall construction
Surface contain-
ment membrane
Top of wall flow
cutoff barrier
Bottom of wall
flow cutoff
barrier
Drainage
Surcharge loading
configuration:
a) inner core
b) wall
c) outer ring
Pore pressure
measurement of
in situ sand
unwashed brick sand
Center Hill #2 soil,
2% ordinary bentonite
PVC slip-form with unpres-
surized steel backfill
tremie pipe (some hand
placement of backfill
required)
translucent flexible poly-
ethylene sheeting (1-a);
transparent PVC membrane
(1-b)
folded polyethylene sheet
forced into soil-bentonite
with rigid PVC ring (1-a);
PVC triple ring assembly
glued to surface PVC membrane
(1-b)
stainless steel baffle pan
at midpoint in base of wall
six 3/8" I.D. ports on
exterior of tank with
interior fabric-covered,
slotted PVC collectors
combination stainless steel
and semi-rigid PVC block
bearing plate
pneumatic bladder only
stainless steel bearing
plate ring
none
washed brick sand (+200
material only)
Center Hill #2 soil,
1% ordinary bentonite
PVC slip-form with pres-
surized plexiglas backfill
tremie pipe
transparent PVC
PVC triple ring assembly
glued to surface PVC mem-
brane
concentric PVC rings glued
to tank bottom along inside
and outside of wall
six 3/8" I.D. ports on exte-
rior of tank connected to con-
tinuous l"-thick interior
fabric-covered drainage panel
around perimeter of tank
combination stainless steel and
rigid polystyrene block bearing
plate
semi-rigid polystyrene bearing
plate ring
semi-rigid polystyrene bearing
plate ring
polyethylene tubing probes
immediately outside wall at
depths of 1/3 and 2/3 wall
height "_
note: 1 inch = 25.4 mm
66
-------
Soils--
The soils used in Sequence 1 tests were as follows:
In situ sand - fine brick sand containing 2.5% silt and clay
(-200 sieve size)
Slurry - 5% concentration (dry weight basis) of ordinary
(unaltered) bentonite
Soil-bentonite - 2% concentration (dry weight basis) of ordinary
backfill (unaltered) bentonite in Center Hill No. 2 Soil
(67% brick sand, 33% Center Hill clay of moderate
plasticity)
In situ sand—In order to clearly evaluate soil-bentonite wall perfor-
mance, the hydrauic conductivity of the model in situ sand should be greater
than that of the soil-bentonite backfill by several orders of magnitude. Also,
slurry penetration into the sand during cutoff wall construction should be
limited to a shallow surface filtration seal, and the sand should be reason-
ably stable during excavation of the slurry trench.
Based on these criteria, a locally produced fine brick sand containing
approximately 2.5% silt and clay fines was chosen as the model in situ sand.
Figure 20 shows the grain size distribution of this sand (dashed line). The
hydraulic conductivity of the sand, as measured using a conventional rigid-
wall permeameter, was 1 x 10"^ cm/sec. This value is at least four orders of
magnitude greater than that of the designed soil-bentonite backfill.
Slurry penetration tests using a modified rigid wall permeameter were
also performed on this sand to confirm that a thin surface filtration seal
would develop. The permeameter was modified by using an acrylic tube to
permit visual observation of slurry penetration. A 5% slurry was driven into
the sand under a nominal driving pressure of 1 psi, producing surface filtra-
tion with virtually no slurry penetration.
Slurry—The primary function of a bentonite-water slurry is to stabi-
lize the walls of the slurry trench before and during the placement of the
soil-bentonite. Such stabilization is achieved by the formation of a slurry
seal on the walls of the trench and the hydrostatic pressure of the slurry on
that seal. The slurry used for this purpose should therefore exhibit suffi-
cient viscosity, and shear strength, with only minimal filtrate loss. A 5%
to 7% bentonite concentration in water is typical in current practice; a 5%
bentonite concentration was used in this project.
An ordinary (unaltered) light gray colored commercial bentonite was used
for Sequence 1 testing. Table 10 presents data on the engineering properties
of the slurry prepared using this bentonite.
The slurry was batched in accordance with attached standard laboratory
procedures (Appendix D) and stored in a high humidity environment for a
minimum of seven days prior to use.
67
-------
UNIVERSITY OF CINCINNATI
CENTER MILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
oo
U.S. JIAK
100, 4 4 J
W
3
* *°
| 90
91 .«
0 "
£
0
4
1 1
JOARD SI[V( OrtMIHO IM IMCItfS U.S. MA HOARD $I|V| NUMKU HTDMOMCICt
7 IVS 1 U h H 3 4 * • 10 I«I6 70 JO 40 50 70 tOOI40 100
1
• BUSHELMAN MEDIUM
- - ESTE SAND
0 OTTAWA SAND (F-9
CENTER HILL CLAY
- CENTER HILL #2 S
1 1 1
SAHD
5)
OIL MIX."
500 100 30 10
tot.
IA-,iMU
IU» o* wrm
OtAvfl
COATlf | »Ht
< [ T
5
^v~r j^ij ji
V
V
-\
V
\
v
\
^ J
V
\
1 '
M
IV.
L
s
y
V
i
w
i
v
1
-1
~Y
L
\
i •
X
1
|
\
\
\
1
*x
"**^
-i—
1
1
1
1
K. \
M
\
\ i
j
t-.
d
v
i
^
^
S
\
\
\
\
»
^
\
""*
^
} 1 0.) O.I 6.05 O.OI 0.005 O.O
GIAIM IK( MnilMtKIS
I«HD
toull «
-------
Table 10. MEASURED PROPERTIES OF 5% ORDINARY (unaltered)
BENTONITE SLURRY - BATCH 1
Unit weight 1120 kg/m3
(69.9 lb/ft3)
API filtrate loss 14.5 ml
Marsh funnel viscosity 49.6 sec
Plastic viscosity 15 cp
Apparent viscosity 25 cp
10-sec gel strength 0.059 kg/m2
(1.2 lb/100 ft2)
10-min gel strength 0.039 kg/m2
(0.81b/100 ft2)
Soil-bentonite backfill—The soil-bentonite backfill used in this Se-
quence was designed to have .a target hydraulic conductivity of 1 x 10'7
cm/sec under nominal levels of consolidation and low hydraulic gradients.
According to D'Appolonia(5)t a typical backfill should contain on the
order of 20% fines and be well-graded to achieve this low value of hydraulic
conductivity. After testing several locally available soils, it was concluded
that a mixture of several soil components would be required to achieve the
desired gradation and baseline hydraulic conductivity. Figure 20 presents
grain-size distribution curves for the various components used to produce the
Center Hill No. 2 standard soil, and a combined distribution curve for the
resulting mixture.
As shown, this soil mixture was composed of 50% medium sand, 20% "F-95"
Ottawa sand, 10% fine brick sand, and 20% Center Hill clay. The soil mixture
was further mixed with 2% ordinary (unaltered) Wyoming bentonite and suffi-
cient .deionized water to produce a 7" slump (water content of 26%, dry
weight basis). The backfill was batched according to Center Hill SLP S-04A
(Appendix D) and stored in a high humidity environment for a minimum of seven
days prior to use.
Test Sequence 2
As reported in Table 6, Sequence 2 involved the use of a different soil-
bentonite and model in situ sand as well as revised surcharge and hydraulic
loading conditions (results discussed hereinafter are therefore not directly
comparable with those for Sequence 1).
The bentonite content for the soil-bentonite used in Sequence 2 was
reduced from 2% to 1%. This change was made to expedite project testing by
reducing testing time required to establish equilibrium flow under each set
of test conditions. The unwashed brick sand used as the model in situ soil
in Sequence 1 was replaced with washed (+200) brick sand for Sequence 2 tests.
69
-------
This change, together with the installation of a perimeter drainage panel
(Hydroway Drain by Monsanto), was primarily aimed at reducing the amount of
hydraulic head lost through the sand so that design test'gradients across the
soil-bentonite wall could be achieved. To monitor the effectiveness of
in reducing incidental head loss, two pore pressure probes
right outside the soil-bentonite wall at depths of 1/3 and 2/3
these changes
were installed
of wall height.
Prior to Sequence 2 testing, additional slurry penetration tests were
performed on a variety of sands. Washed brick sand was found to be most
suitable, since slurry penetration under a nominal hydraulic pressure was
on the order of 38.1 to 50.8 mm (1.5 to 2 inches) but not deep enough to clog
the newly installed fabric-covered perimeter drainage panel. Processing
(washing) of brick sand for use in the tank involved a repeated manual agita-
tion, settling, and decanting procedure. Over one ton of processed brick
sand was required.
New backfill placement methods used in the construction of the soil-
bentonite wall were tested using a plexiglass soil-bentonite placement trench
(Figure 21) and a plexiglas tremie pipe. The trench was used to observe the
slurry displacement process during backfilling. The tremie pipe coul.d be
pressurized, thus insuring continuous flow; a condition which previously was
not possible (Sequence 1'tests) due to the combined effects of friction along
the wall of the tremie pipe and the shear strength of the soil-bentonite.
I
jl
1
98
DOUBLE
THICKNESS
.____^__.
N— ^
^ J.
2
5
ELEVATION
SECTION
-.12.5 k-
4 1 i
U— 24.5 c/c-
J (TYP.)
—|~12. 5-
I 1 l»
1 1 1.5
1 I 1 U 1
PLAN
ALL MATERIALS
V PLATE PLEXIGLAS,
DOUBLE THICKNESS
WHERE INDICATED
ALL DIMENSIONS IN INCHES
note: 1 inch = 25.4 mm
Figure 21. Construction detail of soil—bentonite placement trench.
70
-------
Figure 22 is the proposed loading plan for Sequence 2 tests. Note
that effective consolidation pressures were increased from 31.03, 62.06 and
124.11 kPa (4.5, 9 and 18 psi) used in Sequence 1 to 41.37, 82.74 and 165.48
kPa (6, 12 and 24 psi). This change allowed for an increase in the ratio of
soil-bentonite:sand effective consolidation pressure from 1.5 to 2.0 in all
cases. This change was made to reduce the possibility of permeant leakage
beneath the surface membrane at the interior interface between'the soil-ben-
tonite wall and the model in situ sand (a situation which might have developed
in test 1 (c) had it been performed). This change was also intended to show
the pronounced effect of larger surcharge pressures on the hydraulic conduc-
tivity of the soil-bentonite wall.
Sequence 2 baseline soil characterization work paralleled that for
Sequence 1. Soil-bentonite was prepared by addition of 1% bentonite to the
Center Hill No. 2 standard soil shown in Figure 23. A water content of 23%
and a slump of seven inches were measured for this batch. Table 11 presents
the pertinent properties measured on a second batch of 5% ordinary (unaltered)
bentoniteiwater slurry prepared for Sequence 2 use. Note that several para-
meters measured on batch 1 slurry are not reported for batch 2. This is
because the rotational viscometer used to determine plastic and apparent
viscosity was being modified for use on a different project and was not
available for batch 2 characterization work.
TABLE 11. MEASURED PROPERTIES OF 5% ORDINARY (unaltered)
BENTONITE SLURRY-BATCH 2
Unit weight 1036.7kg/m3
(64.67 lb/ft3)
API filtrate loss 14.2 ml
4
Marsh funnel viscosity 41.9 sec
pH 8.60
The last significant design modification for Sequence 2 testing involved
the use of polystyrene bearing plates to transfer pneumatic surcharge pres-
sures to the various elements of the model. Polystyrene was selected because
of its ability to conform to minor irregularities on the surface of the soil
while maintaining sufficient rigidity to distribute the loads evenly.
Window Closing Test
After the completion of Sequence 2 testing, a window closing test was
conducted to observe if artificially created openings in a model slurry wall
could be closed by applied surcharge pressure. Two window slots were used to
71
-------
Overburden (c) 3 12 15
Gauge
Pressure (h) 3 99
(psi)
(a) 3 7.5 6
Hydraulic (c) (n) (6) 12
Gauge
Pressure (b) (0) (3) 6
(psi)
(a) (0) (1.5) 3
Effective
Consol idatlon
Pressure 3 6 3
(n-i\
IPS 1 1 ......... . ....
;•'•'•'•'••. -~ .'CORE
• SAND S/B .SANE
•'•' • .' .'• '•'• £?Z.T • •'•'.'•
(
(f) 6 18 18
(e) 6 15 12
(d) 6 13.5 9
(f) (0) (6) 12
(e) (0) (3) 6
(d) (0) (1.5) 3
6 126
SAND '. S/B ^[j
'•'„•'•'.'.'•'. .'.•. .'.'
L TANK (L
U) 12 3
(h) 12 2
(g) 12 25.
(j) (0) (
(h) (0) (
(g) (0) (1.
12 2
• . . . * " • ' *
' SAND ' S
:••.'•'•' •/•'.'•.'• "
0 ?A
7 18
5 15
6) 12
3) fi
5) 3
4 12
-— . * ." •
~ CORE
'* SAND
™ ^ * . • • ."
C.
NOTE: Effective consolidation p-ressure = overburden gauge pressure minus hydraulic gauge pressure
*Head loss through core and outer ring of sand assumed to be zero. 1 psi = 6.895 kPa
Figure 22. Loading plan for slurry wall frame Test Sequence 2.
-------
UNIVERSITY OF CINCINNATI
CENTER MILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
US. 11*1
HWW-, * 4
9Q -
I-
$ „"
o 40
•
t
0 _
O
A
n
1
NOAID SUV! OflKNO M
I IM r H H H
—
1
1 1 1
»KI«S U.S. SIANOAtO SMVt NUMtin HTOIOMIIII
J 4 • • 10 U 14 10 10 40 50 7O MOI4O 1OO
1
I1
^
\
' rr '
\
\\
\ si
N
BUSHELMAN MEDIUM SAND
CENTER HILL NO. 2
MASHED +200 ESTE SAND
-
—
100 100 10 ID 1
O (AIM Sill M
IMMIMI
CO*
HIV CNMnH
OUVII
COAOM | t»«
•i
?
1
. i _.
A
\-
\
^
|
I
X
I
V
\
v
•N
1 1
4
X
\
k
%v
01
MUWHM
01
IANO
COUM | MH04 | n«
C\Alt*«AnOM
mi wu
U
n
ORAOAIION aims
n
.
—
—
—
0.01 0.01 O.001 • 00
MI CM a*t
0
10
10
.0 5
U
"I
"I
*° £
/o *
(0
to
too
01
- HKUC. 3210-07 SLURRY WALL TANK
_ AM*
•»»
Figure 23. Grain size distribution of soils used in Test Sequence 2.
-------
simulate potentially entrapped slurry or smaller shrinkage cracks that could
develop due to the leaching of chemicals through the wall.
A new slurry wall was built for the window closing test. The soil mate-
rials used in the new wall as well as the tank system itself were identical
to that used in the Sequence 2 tests. The new wall was preconsolidated under
an effective overburden pressure of 41.37 kPa (6 psi), while it was permeated
with water at a hydraulic pressure of 20.69 kPa (3 psi). This was continued
until an equilibrium flow condition was reached. Once this equilibrium was
reached, first the hydraulic pressure, then the overburden pressure were
carefully dropped to zero. Next, the lid of the tank and the loading bladders
and plates were removed, and the PVC membrane was rolled back to expose the
sand in the outer ring. To access the soil-bentonite wall, two 152.4 mm (6
inch) wide by 152.4 mm (6 inch) deep excavations were made in the outer sand
ring. The excavations were located diagonally across from each other, along
radii that were positioned 45 and 225 clockwise from the top drainport of the
tank. The windows were then created by pushing a steel bar of uniform cross-
section horizontally through the wall at a depth of 127.0 mm (5 inches) below
the top of the wall. The slot openings were then covered by 76.2 mm (3 inch)
square fabric-covered wire mesh before backfilling the outer ring sand.
Fabric-covered wire mesh was also inserted on the inside face of the wall at
the positons and depth of the window slots. They were placed vertically
through a cut in the PVC cover sheet above each slot location and along the
vertical inside interface between the surface of the wall and the inner core
of sand. The cuts in the PVC membrane were then patched followed by reassem-
bly of the other components of the tank system. Details related to the
location, dimensions and construction of the two slot windows are shown in
Figure 45 in the following section of this report.
The window closing test itself comprised staged incrementation of effec-
tive overburden and hydraulic pressures up to a maximum of 13.79 and 20.69 kPa
(2.0 and 3.0 psi), respectively. During this time, several hydraulic conduc-
tivity measurements were taken to document the closure of the window slots.
The test concluded with disassembly of the tank, re-excavation at the location
of the two window slots, and visual inspection of conditions as described in
the following section of this report.
74
-------
RESULTS AND DISCUSSION
In Situ Consolidation and Hydraulic Conductivity
Test Sequence 1—
This sequence comprised two hydraulic conductivity tests, designated as
tests l(a) and l(b).
Test l(a) was the first test of a model soil-bentonite wall in the tank.
It was conducted over a period of about two weeks with an effective overburden
(consolidation) pressure of 31.03 kPa (4.5 psi) on the wall and a hydraulic
gradient of 20.8. Figure 24 shows the measured hydraulic conductivity as a
function of time for this test. An apparent equilibrium hydraulic conductiv-
ity of about 2.7 x 10~? cm/sec was measured. This value is not considered
reliable due to several difficulties encountered during the test. The most
significant of these was permeant leakage across the top of the wall, between
the upper surface of the wall and PVC membrane, due to the stiffness (lack of
conformity) of the surface membrane and cut-off system.
Other significant problems observed during this initial test included:.
1) an indeterminant loss of head pressure (reduction of hydraulic gradient)
by virtue of permeant flow through the in situ sand elements of the model,
2) clogged outflow ports on the wall of the tank due to migration and build-up
of fines from within the sand 3) the ponding of water over the PVC surface
membrane (also due to the clogging of outflow ports) which served to produce
a net back-pressure on the soil-bentonite wall thereby further reducing the
effective hydraulic gradient.
Together, these adverse developments had an indeterminant net effect on
test results. It is believed, however, that the hydraulic conductivity mea-
sured for the soil-bentonite wall during this test was considerably higher
than the true value.
Test l(b) was conducted over a period of 36 days with an effective
overburden (consolidation) pressure of 31.03 kPa (4.5 psi) on the same soil-
bentonite wal 1 used in test l(a). In this case, however, the hydraulic
gradient was doubled to a value of 41.6. Results for this test are also
presented in Figure 24. Although many of the problems associated with test
l(a) remained unresolved during this test, it was undertaken primarily to
gain additional experience with the tank and to check the performance of a
redesigned surface membrane and cut-off system (see Figure 17). Although the
measured equilibrium hydraulic conductivity of 3.4 x 10"8 cm/sec is considered
to be unreliable (for many of the same reasons as test l(a) results), the
general pattern is similar to that of test l(a). The eightfold decrease in
permeability relative to test l(a) is believed to be due mostly to the elimi-
nation of permeant leakage across the top of the wall. A portion of the
observed decrease may be attributable to the increased hydraulic gradient
which produced additional horizontal consolidation of the soil-bentonite wall.
Due to the testing difficulties described and the development of local-
ized corrosion and leakage around the bottom of the tank near the end of test
l(b), Sequence 1 testing was discontinued to permit minor repair (re-welding)
75
-------
10
-5
3 10
-6
10
-7
TEST CONDITIONS
S/B mix:
In-situ Soil:
Permeant:
Effective Overburden
Pressure on S/B Wall:
Hydraulic Pressure:
Hydraulic Gradient:
2.7 x 10
-7
3.4 x 10"8
3.4 x 10 ° NV Kb)
C.H. t2 Soil and 2X
bentonite
unwashed Este sand
200
400 600
TIME (HOUR)
300
1000
note: 1 psi = 6.895 kPa
Figure 24. Hydraulic'conduct!vity results for Test Sequence 1.
of the tank and allow for incorporation of the design modifications previously
described and summarized in Table 9.
Post-Test Evaluation—
Post-test evaluation at the end of Sequence 1 testing consisted of
selective sampling and laboratory testing of soi1-bentonite and in situ sand.
Undisturbed tube samples of soil-bentonite samples and bulk (manually exca-
vated) samples of sand were taken at the locations and depths indicated in
Figure 25. Results of unit weight and water content measurements for soil-
bentonite samples are plotted in Figure 26. Observed general trends for
samples from the soil-bentonite wall are as follows:
* water content increased with increased sample depth
76
-------
TANK PLAN AND
PROFILE
(CONTROL PANEL |
TOP OUTLET-*J
PORT s^~*
/
/
/
/
(
,
o V
^BOTTOM N
S^
-N
\OUTLET
\PORT
??'
^.
— .
>>»
pr
/
/
/
/
•—<
J°
*s
> —
1
1
4"
— ,
\
\
>°
.-*
J»
X,
X.
\
\
,0) 1
/
/
S
19"
v6
r
^r"-!
r
/
/
'
'
'
. —
SAMPLE
NUMBER
T-5
T-l
T-2
T-3
T-4
TO
S-l
S-2
S-3
S-4
S-5
S-6
S-7
S-8
SAMPLE LOCATION
PLAN
r , 8
8", 90°
8", 130°
8", 90
8", 180°
8", 90°
O.I lOQ0
16", 90°
16", 180°
0", 0°
16", 90°
16", 180°
0", 0°
16", 0°
16", 135°
DEPTH
z
3-7/8"
4"
9-1/2"
8-7/8"
19-3/4
iqn
0"-2"
1/2"-
4.5"
2.5"-
6"
7.5"-
11.5"
7.5"-
11"
7.5"-
12"
17.5"-
20"
16.75'
-19.5"
SOIL
DESCRPTN.
TUBE
1 V/// /A
V/ / / /A
\S / // / A
V////A
\S / / / / A
P / / J
BULK SAND
II
II
II
II
II
II
PROPERTY
WET
UNIT WT
Ib/ft3
129.2
130.7'
127.5
129.5
121.4
125 7
WATER
CONTENT
18.7
18.5
21.3
19.7
30.3
20 1
DRY
UNIT WT
108.8
110.3
105.1
108.2
93.2
104 7
note: 1 inch = 25.4 mm; 1 lb/ft3 = 16.02 Kg/m3
Figure 25. Post Sequence 1 sampling plan and unit weight test results.
• water contents of samples aligned with tank ports (in -drainage
path) were higher than for samples at the same elevation which were
not so aligned
* dry unit weight decreased with increased depth
• dry unit weight of samples not aligned with drain ports exceeded
initial dry unit weight (net consolidation occurred)
* dry unit weight of samples aligned with drain ports exceeded the
initial baseline dry un.it weight above a depth of about 0.43 m
(17 inches)
77
-------
CO
WATER CONTENT
ur
15 20 25 30
DESCRIPTION
OF SOIL
MET UNIT WEIGHT
(Ib/ft3)
120 125 130 135
DRY UNIT WEIGHT
(Ib/ft3)
90 100 110
(Initial)
Samples from
drain port
1ine (top)
Broken Line — —
0
Samples 90
away from
drain port
1 ine
BOTTOM
CUT-OFF
note: 1 inch = 25.4 mm; 1 lb/ft3 = 16.02 Ky/m3
Figure 26. Post Sequence 1 unit weight and water content data
as a function of depth in the model cutoff wall.
-------
These trends demonstrate that consolidation of the soil-bentonite wall
did occur, but possibly not to the same degree over the full depth of the wall,
presumably due to some dissipation of effective vertical consolidation stress
with depth. In the case of sample T-4 (Figure 25), it appears that a net
decrease in dry density may have occurred over a limited portion of the wall
in the vicinity of the bottom cut-off barrier. This condition may reflect
localized hydrofracture at the base of the wall.
Results of sieve analysis of bulk sand samples are not presented in the
form of comparative gradation curves, since changes in the grain size
distribution as a result of Sequence 1 testing were very small and would not
be apparent if presented in that form. Table 12 reports the total percentage
of fines (minus 200 sieve size) measured for bulk sand samples. Considering
the inherent accuracy of the sieve analysis method, the trends are not
conclusive, but it appears that:
• the percent fines increased relative to initial values for all sand
samples, including those taken from the center core of the model
• a slight increase in percent fines with increased sample depth may be
indicative of concentration of flow over the lower portions of the wall
• there is no significant difference attributal to sample alignment
with outlet ports for samples at roughly the same elevation
A straightforward explanation for these observations is not possible.
Although the evidence for migration of fines during permeation is not strong,
it is believed that this phenomenon likely did occur and may have involved
limited removal and transport of clay from the soil-bentonite wall itself.
TABLE 12. FINES CONTENT OF SAND SAMPLES
Sequence 1 Post-Test Evaluation
SAMPLE LOCATION
SAMPLE
DEPTH
0" - 4.5"
2.5" - 6"
7.5" - 11.5"
16-3/4" - 20"
OUTER RING
SAMPLES ALIGNED WITH
TOP OUTLET PORT
(Si) 2.8%
(S4) 3.2%
(S8) 3.3%
SAMPLES NOT ALIGNED
WITH TOP OUTLET PORT
(S2) 2.9%
(S5) 3.2%
(S7) 3.3%
CENTER
CORE
(S3) 3.3%
(S6) 3.6%
NOTES: 1. Sample depth is measured from the upper surface of the specimen
2. The pre-test fines content of the sand was 2.5%
3. 1 inch =25.4 mm
79
-------
Test Sequence 2—
Sequence 2 tests were conducted on a 1% soil-bentonite wall over a period
of six months. During this time six tests involving staged incrementation of
overburden and hydraulic head pressure were completed. Figure 27 is a summary
of results of the six tests. All curves exhibit similar characteristics.
They all start with high hydraulic conductivities, drop quite rapidly in the
early stages of the tests, then decrease gradually until equilibrium hydrau-
lic conductivities are reached. Test 2(f) is an exception to this trend,
since it had to be discontinued before reaching equilibrium due to rupture of
the pneumatic surcharge bladder.
Figure 28 presents a chronological summary of events of Sequence 2 tests.
In the typical test the wall started with a high initial hydraulic conductivity
represented by an open triangle, then dropped off and gradually approached
the final equilibrium value designated by an open circle. The incidences of
hydrofracture and ruptured surcharge bladder are indicated by solid triangles
and circles, respectively. As shown, four tests, 2(a), 2(b), 2(d), and 2(e),
were completed as planned. Test 2(c) was not attempted due to apparent
hydrofracture. Instead, test 2(d) involving a higher overburden (consoli-
dation) pressure was initiated in an attempt to heal the hydrofracture. Test
2(f) was also interrupted by the rupture of the surcharge bladder which
resulted in a second hydrofracture of the soil-bentonite wall. Both occur-
ences of hydrofracture damaged the wall.
To estimate the extent of damage, test 2(g') was conducted at overburden
and hydraulic pressures matching those of test 2(d). As can be seen, the
results of test 2(g') did not match those of test 2(d), having an equilibrium
hydraulic conductivity almost an order of magnitude higher. These data
clearly indicate that permanent damage or structural changes occurred in the
soil-bentonite wall as a result of hydrofracture. Consequently, test 2(g) is
somewhat unrelated to the other Sequence 2 tests, in that it was conducted on
a damaged wall. Nevertheless, results for test 2(g) do illustrate that a
doubling of surcharge pressure from 82.74 to 165.48 kPa (12 to 24 psi) can
effectively reduce the conductivity of the wall.
Aside from damage and alterations in the wall due to hydrofracture, the
results of Sequence 2 tests are considered to be qualitatively valid due to
improved equipment design, the use of a free-draining sand, improved soil
placement techniques, and better testing procedures. One such improvement
involved the use of in situ pore pressure probes at two depths immediately
outside (downstream) of the soil-bentonite wall to measure tail pressures.
A more detailed account of each test illustrated on Figure 27, along
with an interpretation of test results is presented in the following para-
graphs.
Test 2(a) was conducted at an effective overburden (consolidation)
pressure of 41.37 kPa (6 psi) and a net hydraulic pressure of 20.69 kPa
(3 psi) which represents a hydraulic gradient of 20.8. As shown in Figure 29,
the wall had an initial hydraulic conductivity of 2.3 x 10"^ cm/sec which
quickly dropped off and then gradually approached an equilibrium value of
7.4 x 10'8 crn/sec.
80
-------
E
u
>
H
O
O
z
O
O
O
V
-5
:
-6
=
-7
1 0
1 n — . —
-9
10
ii
k
\fv
\\
\\
\\
TEi
20
2(t
2(c
2(e
2(1
2(c
^N^^
%-.-— 2(f)
TEST SEQUENCE 2
Effective
Overburden Hydraulic Gradient
>T Pressure Pressure
(psi) (psi)
i) • 6 3 20.8
)) 6 6 41.6
1) 12 3 20.8
0 12 6 41.6
r) 12 12 83.2
) 24 3 20^8
2(a)
\
\ -"V
— — 2(d)
V^
2(g)
'
/2(e)
200 400 600 800
TIME (HOURS)
1000
1200
note: 1 psi = 6.895 kPa
Figure 27. Compos.ite of hydraulic conductivity
results for Test Sequence 2.
81-
-------
2CO
: 1 2 / 1 2 2 ( g )
24/3
1 0
30
90 120
TIME (Days)
1 50
180
V In it I a I K
O Equilibrium K
V Breakthrough
• Ruptured Surcharge Bladder
Q Projected Equilibrium K
Figure 28. Chronological summary of Sequence 2 testing.
82
-------
-5
10 -
0 -n
S 106-
e
o
O
O
o
O
o 10 -
TEST 2(a)
Effective Overburden
Pressure
Hydraulic (Permeation)
Pressure (psi)
Hydraulic Gradient
3
20.8
--8
10 -
200
400
600
TIME (HOURS)
note: 1 psi = 6.895 kPa
Figure 29. Hydraulic conductivity results for test 2(a).
83
-------
For test 2(b) the effective overburden pressure was maintained identical
to that for test 2(a) while the net hydraulic pressure was doubled to 41.37
kPa (6 psi). The increase in hydraulic pressure caused an intial jump in
hydraulic conductivity of the wall to about 5.1 x 10~7 cm/sec (see Figure 30).
Thereafter, the conductivity decreased rapidly, followed by a more gradual
rate of decrease to an equilibrium value of about 5.3 x 10~8 cm/sec. This
test primarily shows that an increase of hydraulic pressure (or gradient) will
cause horizontal consolidation of the wall and a corresponding reduction of
the equilibrium hydraulic conductivity. The doubling of hydraulic pressure
from 20.69 to 41.37 kPa (3 to 6 psi) at constant effective overburden pressure
of 41.37 kPa (6 psi) produced a 28% decrease in the equilibrium hydraulic
conductivity for test 2(b).
Test 2(c) was an aborted test and is not shown on Figures 27 and 28.
Conditions for this test comprised an effective overburden pressure of 41.37
kPa (6 psi) on the soil-bentonite wall (same as for tests 2(a) and 2(b)), and
a hydraulic pressure of 82.74 kPa (12 psi); an additional 41.37 kPa (6 psi)
relative to test 2(b). As the hydraulic pressure was gradually incremented
to about 68.95 kPa (10 psi), however, the wall suffered apparent hydrofracture
as evidenced by a marked increase in inflow and outflow yielding an apparent
hydraulic conductivity of 1 x 10~5 cm/sec (nearly 200 times greater than that
of test 2(b)). It is believed that the hydrofracture probably occurred near
the bottom of the soil-bentonite wall at points where the applied hydraulic
pressure exceeded the effective overburden stress. Although the test was
quickly terminated to mininmize the damage, the hydrofracture probably influ-
enced the results of all subsequent tests. Such can be observed in the
results of test 2(d).
Test 2(d) involved reconsolidation of the soil-bentohite wall at an
overburden pressure of 82.74 kPa (12 psi) followed by permeation of the wall
under a hydraulic pressure of 20.69 kPa (3 psi). The wall was allowed to
fully consolidate under the higher surcharge pressure prior to the start of
permeation. As shown in Figure 31, the initial hydraulic conductivity of the
wall was about 1 x 10~6 cm/sec. This initial value exceeded that initially
measured for test 2(b) despite full consolidation under >the 12 psi surcharge
pressure. It is believed that the intial valued for test 2(d) is somewhat
higher than it would have been without the hydrofracture in test 2(c).
Despite the hydrofracture, however, the net decrease in equilibrium hydraulic
conductivity from test 2(b) to that of test 2(d) (41.37 kPa (6 psi) increase
in effective overburden pressure and 20.69 (3 psi) decrease in hydraulic
pressure) amounted to an additional 38%.
Arguments regarding the validity of test results obtained after hydro-
fracture can of course be made for all tests in Sequence 2. However, even-
though the results of test 2(d) and subsequent tests may not be absolutely
correct, they are still considered to be qualitatively right.
Test 2(e) was conducted at an effective overburden pressure of 82.74 kPa
(12 psi) (unchanged from test 2(d)) and with a hydraulic pressure increased
from 20.69 to 41.37 (3 to 6 psi). Results are shown in Figure 32. Under
these test conditions, the hydraulic conductivity of the wall jumped to an
initial value of 4 x 10"^ cm/sec, and with time decreased and stabilized at an
84
-------
10-5-
"» -6
! 10*-
>
K
>
O
3
O
z
o
o
o
0 10 -
TEST 2(b)
Effective Overburden
Pressure (psi)
Hydraulic (Permeation)
Pressure (psi)
Hydraulic Gradient
6
41.6
io8-
200
400
600
TIME (HOURS)
note: 1 psi = 6.895 kPa
Figure 30. Hydraulic conductivity results for test 2(b).
85
-------
-6
10 -
o
>
o
a
z
o
o
I"'-
X
TEST 2(d)
Effective overburden
pressure (psi) : 12
Hydraulic (permeation)
pressure (psi) : 3
Hydraulic gradient : 20.8
-8
10 -
3.3 x 10
200
400
eoo
TIME (HOURS)
note: 1 psi = 6.895 kPa
Figure 31. Hydraulic conductivity results for test 2(d).
86
-------
o
**
>
a
z
o
u
i
Effective Overburden
Pressure (psi)
Hydraulic (Permeation)
Pressure (psi)
200
400
600
TIME (HOURS)
800
1000
1200
note: 1 psi = 6.895 kPa
Figure 32. Hydraulic conductivity results for test 2(e).
87
-------
equilibrium hydraulic conductivity of about 1.3 x 10~8 cm/sec; an additional
61% decrease with respect to the equilibrium hydraulic conductivity of test
2(d).
For test 2(f) the effective overburden pressure was maintained at 82.74
kPa (12 psi), while the hydraulic gradient was further increased from 41.37
to 82.74 kPa (6 to 12 psi). The 41.37 kPa (6 psi) increase in hydraulic
pressure caused a large initial jump in the hydraulic conductivity of the
wall to a value of 2 x 10"^ cm/sec as shown in Figure 33. The test proceeded
without complication until the surcharge pressure bladder above the soil-ben-
tonite wall ruptured. This release of surcharge pressure caused a second
hydrofracture of the soil-bentonite wall and a corresponding large increase
in flow (not shown on Figure 33.) At this point, test 2(f) was terminated.
The rupture occurred overnight, however,
resulting in the emptying of the permeant reservoir and exposure of the wall
to the pressurized air. Once again, the structure of the wall was apparently
disturbed as evidenced by an apparent increase in hydraulic conductivity.
In order to evaluate the damage test 2(g') was initiated after the rup-
tured bladder was replaced. This test was conducted with the same effective
overburden pressure (12 psi) and hydraulic pressure (3 psi) as test 2(d). It
was postulated that under these pressure conditions the soil-bentonite wall
of test 2(g') should achieve an equilibrium hydraulic conductivity comparable
to that of test 2(d) if it was not permanently damaged. Test results appear
in Figure 34. After 14 days of testing an apparent equilibrium hydraulic
conductivity of 2 x 10"^ cm/sec was measured for the wall. This value was
nearly one order of magnitude higher than that measured for test 2(d), indicat-
ing permanent damage to the wall.
Even though the properties of the wall had changed, and results from
further testing would not be directly comparable with preceding test results,
it was decided to conduct one additional test to determine if a large effec-
tive overburden pressure would serve to heal the damage due to hydrofracture.
To this end, test 2(g) was initiated at a hydraulic pressure of 3 psi, and an
effective overburden pressure of 24 psi. Results for this test also appear
in Figure 34.
Discussion—The main objective of tests 2(a) through 2(g) was to evalu-
ate the effect of consolidation stresses on the hydraulic conductivity of a
model soil-bentonite wall. In situ consolidation of the soil-bentonite back-
fill occurs in both the vertical and horizontal directions.
Vertical consolidation of the backfill is related to the effective change
in vertical stress within the soil-bentonite as a result of a change in
surcharge pressure applied at the surface of the wall. This effect was
modeled in Sequence 2 by incrementing the overburden pressure. Horizontal
consolidation of the backfill occurs in response to the applied horizontal
stress; in this case the seepage force exerted on the wall by permeant flow
under the applied hydraulic gradient. This factor was modeled in Sequence 2
by the application of various hydraulic head pressures.
In the field, both vertical and horizontal consolidation occur simul-
88
-------
-5
10 -
,.--
TEST 2(f)
Effective Overburden
Pressure (psi) : 12
Hydraulic (Permeation)
Pressure (psi) : 12
Hydraulic Gradient : 83.2
u
13
a
z
o
o
-------
-5
>•»
O
o
i io"e
o —
> -
H _
>
CONDUCT!
I I
O
_1
1 -/
> -
r
;
-8
(
I I 1 1 . \ 1
TEST 2(g') e o
TEST 2(g) -a
\ To Be 8 x 10 (Not Shown)
x a
\ 0. £r ».
fe... .---°^ —
"'•-0--'* Nsx _._X3
-Q
1 1 1 1 1 1
> 200 400 600
_
-
-
-
TIME (HOURS)
note: 1 psi = 6.895 kPa
Figure 34. Hydraulic conductivity results for tests 2 (g ) and 2 (g)
90
-------
taneously. The equilibrium conductivity values reported for Sequence 2 tests
may therefore be regarded as representing the combined effect of both vertical
(surcharge) and horizontal (hydraulic gradient) consolidation stresses for
the test conditions applied.
Except for test 2(g), the data suggest a logical trend of decreasing
hydraulic conductivity as a function of either increasing surcharge pressure
(compare results of tests 2(b) and 2(ej~)or increasing hydraulic gradient
(compare results of test 2(a) and 2(b), 2(d), 2(e), 2(f)). Although the
observed trend is logical, the data fail to reflect the correct magnitude of
change in hydraulic conductivity between successive tests in several cases.
For example, the actual equilibrium hydraulic conductivity for test 2(d) (and
therefore also for all subsequent tests) is lower than reported. As described
earlier, the reason is that hydrofracture permanently changed the properties
of the wall, thus artificially offsetting groups of data measured before
hydrofracture from other groups of data measured after hydrofracture. In
terms of groups of internally consistent data then, the following two sets of
compatible test may be recognized: 2(a),(b) and 2(d),(e).
Figure 28 (chronological summary) includes projected values of the
equilibrium hydraulic conductivity for tests 2(f) and 2(g) which were
interrupted prior to achieving equilibrium flow conditions. These data
reflect an attempt to estimate the final equilibrium value in each case as if
hydrofracture had never occurred. The reason for doing so in the case of
test set 2(d),(e) was to complete the set for the purpose of more fully
evaluating the effect of increasing hydraulic gradient.
The projection procedure used to estimate the equilibrium values for
tests 2(f) and 2(g) is subjective, but is based upon the" similar trends
observed for all data in tests 2(d),(e), and (f) (up to the point of
hydrofracture). The procedure essentially involves 'the use of a graphical
approach to predict the equilibrium value for a test case assumed to be
accurate (2(e)); validation of the procedure on a second test case also
assumed to be valid (2(d)); and then extension of the procedure to the test
case in question (2(f)). Figure 35 illustrates the procedure used to estimate
the equilibrium value for test 2(f). The technique reflects the following
inherent assumptions:
* No crushing of the soi1-bentonite particles would take place, .under
a confining pressure of 82.74 kPa (12 psi)
• The soil-bentonite wall deforms elastically in the early stage of
permeation. The total elastic strain (deformation) produced in
each test case is proportional to the magnitude of the incremental
hydraulic pressure applied
* The time required to produce total elastic strain for each test
case is also proportional to the magnitude of the incremental
hydraulic pressure applied. in each case, however, the elastic
strain rate is assumed to be the same
91
-------
1 0-
-e
1 o-
•x.
•
>
o
3
O
z
o
o
oc
Q
> 5-
X
PROJECTION TECHNIQUE
\ ^— parallel tangents
(const, strain rate)
\
-8
BEST FIT CURVE FOR OBSERVED RAW DATA
2 (f )
2 . 5 x 1 0~
1 0-
- 9
0 200 400 600 800 1000 1200
TIME (Hours)
Figure 35. Procedure used to project equilibrium hydraulic
conductivities for tests 2(f) and 2(g).
92
-------
The equilibrium hydraulic conductivities estimated using this technique,
as well as those measured for the other tests which were not interrupted, are
plotted as a function of effective overburden pressure in Figure 36. The
data presented in this manner do not present an entirely logical picture.
For example, a comparison of data for tests 2(a),(b) with that for tests
2(d),(e) suggests that the same amount of change in hydraulic pressure (20.69
kPa to 41.37 kPa in each case) produces a greater amount of change in permea-
bility at the higher level of effective overburden pressure. Logically, the
doubling of hydraulic pressure between tests 2(d) and (e) should have produced
a lesser net effect than the same change for tests 2(a), and (b), especially
since the wall had already been exposed to a hydraulic pressure of 41.37 kPa
(6 psi) in test 2(b). Clearly the test results do not lend themselves to
straightforward interpretation but they do suggest that:
* Test set 2(d),(e),(f), exhibits greater changes in hydraulic conduc-
tivity for the same change in hydraulic pressure reflecting the recon-
solidation or "healing" of a soil-bentonite wall that had been damaged
(hydrofractured) to an undetermined extent.
• The change in equilibrium hydraulic conductivity due to either a unit
change in hydraulic head pressure or a unit change in surcharge
pressure can be on the same order of magnitude, i.e., the effect of
horizontal (gradient-induced) consolidation can be as large as the
effect of vertical (surcharge) consolidation for a comparable pressure
change.
The equilibrium values shown in Figure 36 represent the combined effect
of vertical and horizontal consolidation. If plotted in terms of combined
(total) vertical and horizontal effective stress with the corresponding
premeability data on an arithmetic scale, these same data would appear as
shown in Figure 37. Presented in this manner, the plot very closely resembles
a typical void ratio versus pressure plot obtained from a conventional one-
dimensional consolidation test. The interpretation of the data is also the
same in that the curve illustrates that beyond a certain point the rate of
change in permeability (or void ratio) greatly diminishes for a unit change
in total stress. In other words, the material becomes more difficult to
consolidate the more consolidated it is. Although the two sets of data (2(a),
(b) and 2(d), (e),(f)) are not strictly compatible as explained earlier,
viewed in this fashion they do reflect a logical and expected trend.
Figure 36 also contains hydraulic conductivity data for two conventional
premeability tests on the same soil-bentonite used in the Sequence 2 model
cutoff wall. The tests were performed in rigid-wall (compaction mold)
permeameters after consolidating the soil-bentonite under an effective pres-
sure of 34.48 kPa (5.0 psi). In order to model conditions in the tank as
closely as possible, both samples comprised soil-bentonite over a layer of
+200 sand on which a thin surface filtration slurry seal had been formed.
The tests were run in excess of 40 days at hydraulic gradients comparable to
that used in test 2(a) of Figure 36. Results are presented in Figures 38 and
39.
As shown in Figure 36, the results for test 2(a) and the two conventional
93
-------
-e
_
—
*•* ^
O
CD -7
^ ~~
E _
*
1-
> io'8
H- —
O _
0 —
z
o
o
o
~ -9
10
-------
100 —
E
o
t 60 -
U
0
z
o
U
O 40 -
OC
a
20 -
10 —
1 —
V
\
\
V
(b)
(d)
— — A
(O
I
a
I
15
I
21
12 15 18
COMBINED VERTICAL AND HORIZONTAL
EFFECTIVE CONSOLIDATION PRESSURE. p«l
24
note: 1 psi = 6.895 kPa
Figure 37. Sequence 2 equilibrium hydraulic conductivities as
a function of combined effective overburden (vertical)
and hydraulic (net horizontal) pressure.
95
-------
0)
\
E
u
H
H
J
H
ID
<
UJ
Z
LT
UJ
£L
1O
1O
-6
10
-7
10
-a
Figure
TOTAU TEST TIME - 49.72 DAYS
. WATER (to 49.78 days)
_4.2. _
i=26
a 10 12 14
FL.OW (Pore Volumes)'
Hydraulic conductivity of Sequence 2 soil-bentonite
as measured in rigid-wall (compaction mold) permeameter.
to
\
E
U
I-
H
J
H
ffl
<
UJ
Z
tr
UJ
Q.
10
-s
10
-6
10
-7
10
TOTAL. TEST TIME - 43.66 DAYS
WATER (to 43.66 (Jays)
sample disturbed
2.1
i=13
5 6 7 ' B
FLOW (Pone Volumes)
Figure 39. Hydraulic conductivity of Sequence 2 soil-bentonite
as measured in rigid-wall (compaction mold) permeameter.
96
-------
tests are in good agreement. The differences may be attributable to several
factors.including different sample preconsolidation pressures (34.48 vs.
41.37 kPa (5.0 vs. 6.0 psi)) and unquantified sidewall leakage in the rigid-
wall cells which is not present in the tank.
By virtue of scale and design, results in the tank should be regarded as
more reliable than those for smaller conventional tests, ideally approaching
the in situ values of full-scale slurry walls. If such is assumed, the
apparent agreement observed is in stark contrast to laboratory and field
comparisons for clay liners reported by Day and Daniel(6) which indicated
that laboratory test results may be unconservative by up to three orders of
magnitude.
The primary reason for good agreement between results for the conven-
tional permeameters and the tank appears to be the high degree of homogeneity
of the soil-bentonite.
The actual as-built homogeneity of a soil-bentonite cutoff wall is not
known. Given the general nature of its construction, however, it is reasonable
to assume that overall, a soil-bentonite wall is more homogeneous than a
typical clay liner, and that the good agreement observed between the different
test methods used in this study may be explained on that basis.
Post-Test Evaluation—
As with Sequence 1 tests, several water content and unit weight tests
were performed on samples from the second soil-bentonite wall after the
completion of Sequence 2 work. In addition, vane shear tests were performed
at several depths to measure the shear strength profile of the soil-bentonite
wall after vertical and horizontal consolidation under the various loading
conditions and the two incidences of hydrofracture associated with Sequence 2
testing. Sequence 2 density and water content data are presented in Figure
40 along with a schematic of the model wall for orientation purposes. These
same data are plotted as a function of depth in the model wall in Figure 41.
Vane shear strength and additional water content data are presented in Figure
42.
Recall that test 2(g) was conducted after the hydrofracture associated
with test 2(f), and that the equilibrium hydraulic conductivity measured for
test 2(g) was higher than that for 2(f) despite the doubling of effective
overburden stress (82.74 to 165.48 kPa (12 to 24 psi)). For this reason, post-
test results for the unit weight, water content, and vane shear tests are
considered to be valid only in a qualitative way.
Nevertheless, the data for these tests considered together, as well as
independently, demonstrate that the effective overburden stress applied at
the surface dissipated to some degree with depth, with the net result that in
situ dry density and vane shear strength are highest near the top of the wall
(highest degree of consolidation) and in situ water content is highest at the
base of the wall (lowest degree of consolidation).
97
-------
CO
TANK PLAN AND
PROFILE
(CONTROL PANEL |
TOP OUTLET-^
PORT />^~*
{ &
^B
1
' 1 i i
OTTOM V.
\OUTLET
\PORT
^—- <
•
^
i — .
-x
M°
~is
* — -
. . »n
z
1 — .
\
^
,0.0) 1
/
/
L|
r
T
22
I
SAMPLE
NUMBER
T-l
T-2
T-3
T-4
T-5
T-6
B-l
B-2
B-3
B-4
B-5
B-6
B-7
'B-8
B-9
B-10
SAMPLE LOCATION
PLAN
r e
8" 135°
8" 315°
8" 135°
8" 315°
8" 135°
8" 315°
16" 270°
16" 180°
11" 225°
11" 45°
16" 270°
16" 180°
11" 225°
11" 45°
16" 270°
16" 180°
DEPTH
z
m-
17"
17"
23"
23"
10"
10"
11"
11"
17"
17"
17"
17"
22"
22"
SOIL
DESCRPTN.
y///////i
V///////1
I////////I
V///////1
Vlllllll\
V///////1
Bulk sand
Bulk sand
Penetrated
Penetrate<
Bulk sand
Bulk sand
Penetratec
Penetrates
Bulk sand
Bulk sand
PROPERTY
WET
UNIT WT
Ib/ft3
137.06
135.40
135.28
135.24
135.13
134.26
I zone
l zone
1 zone
zone
WATER
CONTENT
X
15.3
15.1
15.3
15.5
16.6
17.0
DRY
UNIT WT
Ib/ftJ
118.87
117.64
117.33
117.08
115.89
114.76
note: 1 inch = 25.4 mm; 1 lb/ft3 = 16.02 Kg/m3
Figure 40. Post Sequence 2 sampling plan and unit weight test results.
-------
VO
note: 1 inch = 25.4 mm; 1 lb/ft3 = 16.U2 Kg/m3
Figure 41. Post Sequence 2 unit weight and water content data for
soil-bentonite samples as a function of depth in the model wall
-------
Depth
(In.)
ft _
11 A —
Vane Shear Strength (Ib/ft3)
100 160 200 2SO 300 380
/
^0*
^
(.'.
^^
D~.
I
I
I
^T
(3
L**~
\*~
Q
X
vr.
<* \
\
^'
•— .
x"
i
/
bX
. c
/
c
*
Jl
Water Content (%)
14 16 18 20
Ts
\v
\
\
1
\
V
\
^
note: 1 inch = 25.4 mm; 1 Ib/ft2 = 0.048 kPa
Figure 42. Post Sequence 2 vane shear strength and water content
data as a function of depth in the soil-bentonite wall
100
-------
Window Closure
Soil-Bentonite Modeling--
A limited investigation into the stress-strain (deformational) charac-
teristics of a typical soil-bentonite was undertaken to supplement the fore-
going hydraulic research on model soil-bentonite walls. This work involved
the modeling of stress-strain properties for the evaluation of "window"
closing in the tank by the application of surcharge pressures.
The probability of "window" in a field-scale slurry wall, whether
representing an as-built condition related to in situ materials or faulty
construction technique, or resulting from long-term chemical attack, is not
known. Presumably, undetected subsurface windows might develop due to soil-
bentonite entrapment of slurry or entrapment of sloughed trench wall materials
during the backfilling operation. According to Evans et al,(2) the probabil-
ity of such entrapments along the surface of the advancing soil-bentonite "mud
wave" is high. Moreover, limited evidence of these types of "windows" was
observed during post-test evaluation of the Sequence 1 soil-bentonite wall.
While "windows" of entrapped granular soils probably would have to be
grouted to be closed, slurry entrapments and micro-cracks due to chemical
degradation of the backfill may be closed by in situ consolidation due to
surcharge pressure. An experimental study of the feasibility of closing an
"artificial window" was undertaken as part of this project and its results
are presented in the next subsection of this report.
Paralleling the experimental work using the tank, a numerical analysis
was undertaken to develop a finite element model of soil-bentonite behavior
which could be used to predict soil-bentonite deformation under various
surcharge pressures. Specifically, knowing the modulus of elasticity (E) and
the Poisson's ratio (^-} of the soil-bentonite, and the location and size of
the "window" in the wall, the surcharge pressure needed for closing the window
may be estimated by a finite element analysis.
In order to determine the appropriate E and^cc values to use in the
"window closing" analysis of an unconsolidated soil-bentonite, a series of
parallel slump tests, and counterpart finite element analyses were undertaken.
First, slump tests were performed on the 1% soil-bentonite used in Sequence
2 tests and the shape of the slumped soil mass was measured. Next, the
deformation of the soil mass was modeled immediately after the removal of the
slump cone, using a modified version of a finite element program developed by
Dr. Frank Weisgerber (UC Department of Civil and Environmental Engineering)
to study the stress-strain relationships of TENsion-WEAK materials. The
modifications were made to model the deformation of the cone-shaped soil-
bentonite mass under its own weight.
i
The general scope of the analytical approach using TENWEAK was as follows:
* develop a finite element coordinate mesh representing the undeformed
initial shape of the slump cone soil mass.
101
-------
• Use the known total unit weight and assume an initial Young's modulus,
E, and Poisson's rations for the soil-bentonite.
e compute soil-bentonite displacements due to the incremental applica-
tion of self-weight.
* compare the computed deformations to the measured slump displacements.
If they do not agree, assign new E and-^<- values and repeat the last
two steps as necessary to achieve agreement between the physical
system and the numerical model.
The finite element mesh used as well as the predicted and measured slump
results are shown on Figure 43. Comparison of the computed and actual deformed
shapes for the 1% soil-bentonite used in Sequence 2 tests is very good, with
only a minor discrepancy observed at the base of the soil-bentonite mass.
This discrepancy is explained in terms of basal friction between the soil-
bentonite and bottom metal plate of the slump test apparatus which was not
included in the numerical analysis. Such friction will be modeled in future
analyses to reduce the predicted basal spreading to that actually measured.
Window Closing Test--
As explained before, the purpose of the window closing test was to
demonstrate that surcharge, or overburden pressures can effectively close a
window in'a model soil-bentonite wall. If this can be adequately demonstrated,
then it is expected that entrapped slurry can be squeezed out of real-site
slurry walls perhaps even by in situ overburden pressure. Also, chemically-
induced shrinkage cracks can be closed by the same overburden pressure.
After Sequence 2 testing a new wall for the window closing test was
constructed. Before a "window" was constructed in the wall, the new wall was
preconsolidated by an effective overburden pressure of 41.37 kPa (6 psi) and
a hydraulic pressure of 20.69 kPa (3 psi) (gradient of 20.8). Baseline (no
window) results for test 3(a) are shown on Figure 44. The hydraulic conduc-
tivity versus time curve has the typical shape of the tests of Sequence 2.
Immediately after applying the pressures the hydraulic conductivity of the
wall was measured to be 2 x 10~6 cm/sec. With elapsed time the hydraulic
conductivity decreased, rapidly at first and more gradually thereafter, until
it reached an equilibrium hydraulic conductivity of 3.2 x 10~9 cm/sec.
-After measuring the baseline hydraulic conductivity as described above,
the pressures were reduced to zero, the tank was disassembled, and two small
slot windows were made at the locations and depths shown in Figure 45. The
tank was then reassembled and window closing test 3(a) was started.
An effective overburden pressure of 6.895 kPa (1.0 psi), and a hydraulic
pressure of 6.895 kPa (1.0 psi) were initially applied. During the,fourth
day of testing these pressures were increased to 10.34 and 13.79 kPa (1.5 and
2.0 psi), respectively. After the fifth day of testing the pressures were
raised to 13.79 and 20.60 kPa (2.0 and 3.0 psi), respectively. The curve
also shown in Figure 44 exhibited an initial rapid rise due to the combined
effects of the window openings' and the hydraulic pressure increments.
102
-------
14
12
Young's Modulus (E)
Poisson Ratio
Total Slump
Mesh Element
Node Point
NUMERICAL ASSUMPTIONS
1.72 kPa (0.25 psi)
0.49
177.8 nun (7 inches)
PREOICTEU SLUMP SHAPE
(Numerical)
OBSEKVED SLUMP SHAPE*
(Slump Test)
Coincides with predicted slump above element 2
Figure 43. Finite element mesh for soi1-bentonite slump simulation
and predicted and measured slump mass shapes.
103
-------
o
0
o
HYDRAU
-t
-e
-
-
-9
\
I
1 1 1 1 1 1
TEST 3(a)
T
iU— window Closing Curv«
i
|[ Initial : 1.0/1.0 pal
fi* : 1.S/2.0 pal
li Final : 2.0/3.0 pal
Pr
li
•— Baaellne Curve |j
6.0/3.0 |^
1 i
l 1
l 1
\ ^\
Nj ' ^
\ 1 ' *
^ — ^ A \ 3.2^x 10
\ \
\ o— .— &-—•&-&
\
Window Closing Teat Initiated
I I I I I I
-
_
-
-
400
800
1200
TIME (HOURS)
1600
2000
2400
note: 1 psi = 6.895 kPa
Figure 44. Baseline and window closing hydraulic
conductivity results for test 3(a).
104
-------
•
__£
i
i
SAND
nt
v—
7 v
8/B
CORE SAND ':
WINDOW SLOTS
: '• ^/1
"tit
/ty
8/B
.•'
SAND
FABRIC-WIRE MESH B ft\
ASSEMBLY "
'k,,--J I
1.88
SAND PIT
Figure 45. Details of slot window configuration for test 3(a).
Some of the jumps in hydraulic conductivity in the initial portion of the
window closing test were recognized as being due to the increments in hydraulic
pressure. The other cause, of course, was the presence of the window slots.
The "baseline" equilibrium hydraulic conductivity of the slurry wall
with the windows was difficult to establish. Given the baseline (pre-window)
consolidation of the wall under an effective overburden pressure of 41.37 kPa
(6 psi), it was presumed safe to increment the overburden pressure to about
half that value and still be able to measure an apparent permeability of the
wall with the two slot windows.. Surprisingly, the two windows started to
close under an effective overburden pressure of only 13.79 kPa (2 psi). It
105
-------
was therefore not possible to measure the exact apparent permeability before
the start of window closure or the specific threshold value of surcharge
pressure necessary to initiate window closure (somewhere between 10.34 and
13.79 kPa). Test results do, however, conslusively show that the artificially
created slots were completely closed by an applied 13.79 kPa (2.0 psi) effec-
tive overburden pressure. Under the pressure conditions of 13.79 kPa (2 psi)
effective overburden pressure and 20.69 kPa (3.0 psi) hydraulic pressure, the
equilibrium hydraulic conductivity was 2 x 10~9 cm/sec, or very close to the
aforementioned baseline conductivity of 3.2 x 10~9 cm/sec.
Following the hydraulic conductivity test, the slurry-wall tank was
disassembled and the outer ring of sand was excavated in the vicinity of the
previous window slots in order to visually inspect the slots. There was no
visual evidence of any slot left in the wall, the windows were completely
shut. However, when probed with a tapered steel bar there seemed to be
slightly softer spots at the locations of the previous window slots. But
with respect to hydraulic conductivity the windows were effectively shut.
The success of this window closing test has important ramifications for
real-site slurry walls. It means that the effective overburden pressure in
the wall, which may be considerable, may close slurry windows and may even
close a multitude of minute shrinkage cracks that may have developed in a
wall due to the effects of chemical leachates.
Future testing will involve revised procedures to accurately measure the
"baseline" conductivity of the wall and the threshold surcharge pressure
necessary to initiate closure of the window slots. For example, in this test
sequence a better procedure would have been to maintain the initial hydraulic
pressure of 6.895 kPa (1.0 psi) throughout the test, and let the hydraulic
conductivity come to equilibirum at the effective overburden pressure of 6.895
kPa (1.0 psi). This hydraulic conductivity may then have been called the
conductivity of the wall with the window slots. Following that, the overburden
pressure would have been increased to 10.34 kPa (1.5 psi) without incrementing
the hydraulic pressure and again an equilibrium hydraulic conductivity would
have been established. This increase in pressure would have been expected to
begin to close the window slots and begin to decrease the hydraulic conduc-
tivity of the wall. This procedure would have been continued until the
conductivity of the wall dropped to the equilibrium conductivity of 3.2 x
10~9 cm/sec, as established in test 3(a).
106
-------
SECTION 7
QUALITY ASSURANCE
GENERAL
The research described herein is supported by an approved Quality
Assurance Project Plan dated May 2, 1985, and a single addendum thereto.
Both documents were reviewed in accordance with the guidelines provided in
the "QA Procedures for HWERL, August 1985", and the "Interim Guidelines and
Specifications for Preparing QA Project Plans, QAMS 005/80".
The overall QA objective for this project was to provide data quality at
(EPA) Level 3:
"Projects intended to evaluate and select basic options or to perform
preliminary assessments of unexplored areas require QA levels to pro-
duce scientifically but not legally defensible conclusions. The data
must be valid in a relative but not necessarily an absolute way and
must support the technical choices made."
Five specific considerations that frame the QA effort in general and
that were addressed in the above documents are briefly described in the
following paragraphs.
Accuracy
Civil engineering research often involves measurement of various para-
meters on construction materials whose composition and properties are well
known or established within specific limits (steel, concrete, etc.). Such is
not the case in the study of earth materials and their response under various
environmental conditions. Results from standard geotechnical tests may vary
widely, even when the soil appears to be uniform. As such, there exist no
scientific standards or true values against which one could estimate the
accuracy of the routine soil characterization tests and hydraulic conductivity
measurements performed during this study.
As reported in Section 12.1 of the QAPP, the target accuracy level -for
measurements of hydraulic conductivity on this project was +3%.
This value was based upon a target hydraulic conductivity of 1.0 x 10"?
cm/sec. Where measured conductivities were an order of magnitude lower, for
example, it was impractical to achieve the same target level, due to the
excessive time interval required and the limited capacity of the permeant
reservoir. Such error magnification is commonly associated with the measure-
ment of progressively smaller levels of hydraulic conductivity using conven-
107
-------
tional equipment. For the tank and column, the accuracy of measurement at
hydraulic conductivities less than 1.0 x 10~' cm/sec was in proportion to the
target level of + 3% at 1.0 x 10~7 cm/sec. Considering all measurable sources
of error, hydraulic conductivity measurements involving in the slurry wall
tank, as presented in Section 6, were at a level of _+ 12% or better. The
accuracy of similar measurements involving the slurry test column was vari-
able, ranging from less than 1% (high permeabilities), to more than 100% (very
low permeabilities).
Precision
No specific industry precision criteria for the various test parameters
exist. Throughout the study, precision considerations for both the tank and
column were addressed in terms of the expected margin of error based upon
confidence intervals around a typical measurement of discharge through, or
hydraulic gradient across, the test specimen. In other words, "precision" as
used herein was based upon optimizing the measurement systems and defining
the expected margin of error associated therewith.
Equipment proof-testing modifications referenced in Sections 5 and 6 were
made for the purpose of optimizing the respective measurement systems to the
point where valid test results could be obtained. In both cases, permeant
flow measurements were determined visually in a sight tube as a function of
test time. Because of known and constant geometry of the permeant reservoir
a known and constant error was associated with each flow measurement. This
error was determined by calibration for both the tank and column. In each
case, it was possible to limit the experimental error by applying the known
volumetric error over a sufficiently large total volume of flow. In other
words, it was possible to limit experimental error by controlling the time
(and, therefore, total discharge) between outflow measurements. Figures 46
and 47 present accuracy curves for various assumed "true" conductivity values
for the tank and column, respectively.
Sources of error which were not directly measureable included random
operator error, variations in sample processing and preparation between
different tests, and minor variations in environmental conditions within the
laboratory. The effect of extraneous sources of error is considered minimal,
however, due to the level of effort developing standard laboratory procedures
for each system during the proof-testing phase.
Completeness, Representativeness, Comparability
These three considerations generally address the validity of research
results on a limited sampling of a population in terms of applicability to
the population as a whole. The materials used and test conditions applied
during the course of this study were designed to be representative of those
which normally exist at field-scale. Although the degree of completeness,
representativeness and comparability of test results is not known, they are
judged to be high.
An on-site Quality Assurance Technical Systems Audit was performed at
the Center Hill Facility on May 1 and 2, 1985. Recommendations resulting from
108
-------
TOO
Q.
O
O
0£
ce
80
60
40
20
HYDRAULIC GRADIENT = 20.8
1x10
cm/sec
8 16 24 32 40
TIME INTERVAL FOR VOLUME FLOW RATE MEASUREMENTS (HOURS)
NOTE: 1. The above curves 'apply to the case of hydraulic pressure
20.69 kPa (3 psi).
2. For hydraulic pressure of 41.37 and 82.74 kPa (6 and 12 psi),
the percent error is 1/2 and 1/4, respectively, of that shown
for each permeability curve at the same time interval..
Figure 46. Percent error in measured hydraulic conductivity in the
slurry wall tank as a function of time interval between
volume flow measurements.
109
-------
HYDRAULIC GRADIENT = 550
o
LU
O
o
.
O
on
cc.
o
UJ
D-
4 6 8 10 12 14
TIME INTERVAL FOR VOLUME FLOW RATE MEASUREMENTS (HOURS)
16
NOTE: 1. The above curves apply to the case of a filter cake thickness
of 6.35 mm (0.25 inches) and a pressure differential across the
cake (Ap) of 34.48 kPa (5.0 psl).
2. ForAp = 68.95 kPa (10.0 psi), error is 1/2 of that shown.
3. Holding other variables constant, % error is directly propor-
tional to cake thickness.
Figure 47. Percent error in measured hydraulic conductivity
in the slurry test column as a function of time
interval between volume flow measurements.
110
-------
this project are summarized as follows:
• Where possible, test equipment should be modified to permit
direct data recording onto hard copy in order to minimize
subjectivity of observations.
* All results should be recorded in ink and all test forms
should be initialed by the analyst conducting the test.
* Samples stored in the moisture room should be identified
with water-resistant labels.
* The inherent precision of the data should be reflected in
terms of the reporting of significant figures.
Due to the nature of the tank and column equipment and budgetary
limitations associated in general with the project, it was not practical to
automate either system for direct data recording onto hard copy. Otherwise,
the above recommendations were implemented for the research described herein.
Ill
-------
REFERENCES
1. McCandless, R.M. and Bodocsi, A. Quick Indicator Tests to Characterize
Bentonite Type. Draft Final Report Contract No. 68-03-3210; Work Assign-
ment No. 08, USEPA, HWERL, AWBERC. Cincinati, Ohio, August 1986.
2. Evans, Jeffrey C., Lennon, Gerard P., and Witmer, Kevin A. Analysis of
Soil Bentonite Backfill Placement in Slurry Walls, in: Proceedings of
the 6th National Conference on Management of Uncontrolled Hazardous Waste
Sites, November 4-6, 1985, Washington, D.C., pp. 357-361.
3. U.S. Army Corps of Engineers. Civi1 Works G u i d e S p e c i f i c a t i o n For Soil-
Bentonite Slurry Trench Cutoffs"CW-02Z14, May 1985.
4. Xanthakos, Petros. Slurry Walls. McGraw-Hill Book Company. New York,
New Vork, 1979.
5. D'Appolonia, D. J. Slurry Trench Cut-off Walls for Hazardous Waste
Isolation. Technical Paper. Engineered Construction International, Inc.,
Pittsburgh, Pennsylvania, April 1980.
6. Day, Steven R. and Daniel, David E. Hydraulic Conductivity of Two Proto-
type Clay Li ners. Journal of the Geotechnical Engineering Division.
ASCE, Vol. Ill, No. 8, August 1985. pp. 957-970.
112
-------
BIBLIOGRAPHY
AHher, George R., Evans, Jeffrey C., Witmer, Kevin A., and Fang, Hsai-Yang.
Inorganic Permeant Effects Upon Bentonite. ASTM STP No. 874, "Hydraulic
Barriers in Soil and Rock", July 1985, pp. 64-77.
American Colloid Co. The Role of Base Bentonite in Polymer-Treated Soil
Sealants. Skokie, Illinois.
American Petroleum Institute. Glossary of Drilling Fluid and Associated
Terms. API Bulletin Oil, 2nd Ed. Dallas, Texas, May 1979.
American Petroleum Institute. Oil Well Drilling Fluid Materials. API Spec.
ISA, 10 Ed. Dallas, Texas, April 15, 1984.
American Petroleum Institute. The Rheology of Oil Well Drilling Fluid.
Bulletin, API Spec 13D, 1st Ed. Dallas, Texas, August, 1980.
American Petroleum Institute. Standard Procedure for Field Testing Drilling
Fluids. API Recommended Practice, API RP 13B, 10th Ed. Dallas, Texas,
June 1, 1984.
American Society for Testing and Materials. Soil and Rock: Building Stones.
in: Annual Book of ASTM Standards, Vol. 04.08, Section 4. PhiIdelphia
Pennsylvania, 1984.
Anderson, David C. Effects of Organic Solvents on Clay Soil - Contaminant
Resistant Bentonite Slurry Mixtures. K. W. Brown and Associates, Inc.
College Station, Texas, 1983.
Ayers, John E., Lager, David C., and Barvenik, Matthew J. The First EPA
Superfund Cut-off Wall: Design and Specifications. Presented at the
Third National Symposium on Aquifer Restoration and Groundwater Monitor-
ing, 1983.
Barvenik, M. J., Hadge, W. E., and Golberg, D. T. Quality Control of Hydrau-
lic Conductivity and Bentonite Co'ntent During Soi 1/Bentonite Cutoff Wall
Construction, in: Land Disposal of Hazardous Waste. Proceedings of
the Eleventh Annual Research Symposium, Cincinnati, Ohio, April 1984.
EPA/600/9-85/013, pp. 66-79.
Black, C. A. Methods of Soil Analysis. American Society of Agronomy, Madi-
son, Wisconsin, 1965.
Bowles, Joseph E. Engineering Properties of Soils and Their Measurement.
McGraw-Hill, New York, 1978.
113
-------
Bowles, J. E. Foundation Analysis and Design. 3rd Edition. McGraw-Hill Book
Company, New York, 1982.
Boyes, R. G. H. Structural and Cut-Off Diaphragm Walls. Applied Science
Publishers, Ltd., London, England, 1975.
Brown, K. W., Thomas, J. C., and Green, J. W. Permeability of Compacted Soils
to Solvent Mixtures and Petroleum Products in: Land Disposal of
Hazardous Waste. Proceedings of the Tenth Annual Research Symposium,
Ft. Mitchell, Kentucky, April 3-5, 1984. EPA-600/9-84-007. pp. 124-137.
Chemical Rubber Publishing Co. CRC Handbook of Chemistry and Physics,
65th Edition. Boca Raton, Florida, 1984.
Cement Bentonite Slurry Wall Saves Time, Money as Tailing Dam Cut-Off.
Engineering News-Record, December 2, 1976.
Davis, Ken E. and Herring, Marvin C. Laboratory Evaluation of Slurry Wall
Materials of Construction to Prevent Contamination of Groundwater from
Organic Constituents.Presented at:The Seventh Annual National
Groundwater Quality Symposium, Las Vegas, Nevada, September 26-28, 1984.
Dresser Industries, Inc. Drilling Fluids Engineering Manual. Magcobar Divi-
sion, Oilfield Products Group, Houston, Texas, 1977.
Evans, Jeffrey C., Fang, Hsai-Yang, and Kugelman, Irwin J., Containment of
Hazardous Materials with Soil-Bentonite Slurry Walls in: Proceedings of
the 6th National Conference on the Management of Uncontrolled Hazardous
Waste Sites, November 4-6, 1985, Washington, D.C., pp. 369-373.
Evans, Jeffrey C. and Fang, Hsai-Yang. Geotechnical Aspects of the Design
and Construction of Waste Containment Systems. Proceedings of the
National Conference on the Management of Uncontrolled Hazardous Waste
Sites, November 1982.
Foreman, D. E. and Daniel, D. E. Effects of Hydraulic Gradient and Method
of Testing on The Hydraulic Conductivity of Compacted Clay to Water,
Methanol, and Heptane. EPA-600/9-84-007. pp. 138-144.
Gelled Bentonite Produces Low-Cost Retaining Walls. Engineering News-Record,
January 1, 1976.
Green, William J., Lee, Fred G., Jones, Anne R., and Palit, Ted. Interaction
of Clay Soils with Water and Organic Solvents: Implications for the
Disposal of Hazardous Wastes. Environmental Science Technology, Vol,
17, No. 5, 1983. pp. 278-282.
Harter, Robert D. Reactions of Minerals with Organic Compounds in the Soil.
in: Minerals in Soil Environments. Soil Science Society of America,
Madison, Wisconsin, 1977. pp. 709-739.
114
-------
Hughes, John. A Method for the Evaluation of Bentonite as Soil Sealants for
the Control of Highly Contaminated Industrial Wastes.Presented at:
Purdue University Industrial Pollution Conference, May 12, 1977.
JRB Associates. Slurry Trench Construction for Pollution Migration Control
EPA-540/2-84-001. U.S. Environmental Protection Agency, Municipal
Environmental Research Laboratory, Cincinnati, Ohio, February 1984.
Low, Philip. Nature and properties of Water in Montmorillonite-Water Systems.
Soil Science Society of America Journal, Vol. 43, No. 5, pp. 651 - 658,
1979.
McNeal, B. L. Prediction of the Effect of Mixed-Salt Solutions on Soil
Hydraulic Conductivity. Soil Sci. Soc. Amer. Proc. 32: pp. 190-193,
1968.
Millet, R. A. and Perez, J. Y. Current USA Practice: Slurry Wall Specifica-
tions. Journal of the Geotechnical Engineering Division. Vol. 107:GT8,
August 1981, pp. 1041-1056.
Morrison, Allen. Arresting a Toxic Plume. Civil Engineering Magazine,
ASCE, August, 1983.
Rowel 1, D. L., Payne, D., and Ahmad, N. The Effect of the Concentration
and Movement of Solutions on the Swelling, Dispersion, and Movement
of Clay in Saline and Alkali Soils.J. Soil Science, 20 (1):pp. 197-
188, 1969.
Spooner, Philip, et al. Slurry Trench Construction for Pollution Migration
Control. EPA-540/2-84-001, U.S. Environmental Protection Agency,
Cincinnati, Ohio, February 1984.
Spooner, P. A., Wetzel, R. S., and Grube, W. E. (JRB Associates, McLean, Vir-
ginia, 22101). Pollution Migration Cut-Offs Using Slurry Trench Con-
struction. National Conference on Management of Uncontrolled Hazardous
Waste Sites, Washington, D. D., November 29 - December 1, 1982. pp.
191-198.
Tallard, Gilbert. Slurry Trenches for Containing Hazardous Wastes. Civil
Engineering Magazine, February, 1984.
Theng, B. K. G. Clay-Polymer Interactions: Summary and Perspective. Clay
and Clay Minerals, Vol. 30, No. 1, 1982. pp. 1-10.
U.S. Army Corps of Engineers. Engineering and Design Laboratory Soils Test-
ing. EM 1110-1-1906. Waterways Experiment Station, 1980.
United States Pharmacopeia, The National Formulary. 20th Revision. U.S.
Pharmacopial Convention Inc., Rockville, Md. 1980.
115
-------
Van Olphen, H. and Fripiat, J. J. Data Handbook for Clay Minerals and Other
Non-Metallic Minerals. Pergamon Press, Oxford, 1979.
Ward, Linda M. Close-up on Cleanup at Li pan' Waste Site. Hazardous
Materials & Waste Management Magazine. May-June, 1984. pp. 40-42.
116
-------
APPENDIX A
PROCEDURAL OUTLINE FOR THE OPERATION OF THE SLURRY TEST COLUMN
Typical sample preparation and testing procedures for the slurry test
column are outlined in this section. The following information is intended
primarily to document the methods used during this study and to provide a
reference for use by those who may use the column in future research. Refer-
ence is made to the slurry test column schematic which is preprduced herein
as Figure Al.
1. The column should be clean and free of foreign matter. This is especially
important at. the column ends where contact is made with the end plates.
The end cap 0-rings should be intact (not torn, separated or cracked).
If they do not fit in the groove for any reason, they should be refabri-
cated. The 0-ring groove and surrounding area should be free of sand
particles, grit, excess silicone, etc.
2. The column and bottom end plate (containing the basal filter) should be
assembled. A layer of pea gravel should be placed around the end plate
filter. A circular 200-mesh screen screen should be placed on top of the
gravel to prevent the migration of sand into the gravel filter.
3. Place the top cap on the column and secure it in place with the four
all-thread rods. Initialize the load ring to zero. Place the load ring
under the column to determine the empty weight of the column. Use a
piece of plywood on the floor and a PVC cap to protect the three-way
valve on the bottom end plate.
SAND PLACEMENT IN COLUMN
1. Make sure the solid pvc brace collar on the side of the column is secure
and remove the top plate.
2. Using an inverted "slump cone" as a funnel, place the air-dried, (uniform
moisture content) sand into the column as uniformly as possible. The
sand should be blended and thoroughly mixed prior to placement.
Several (four to five) moisture content determinations for the sand
should be run as the sand is being placed. Continue to place sand into
the column until the level of the sand is five to eight inches above that
of the top probe port.
3. Record the final dial reading on the load ring and compute the total
column weight. The wet weight (Wwet) of the sand in the column is equal
to the final column weight minus the initial (empty) column weight. Mark
117
-------
VACUUM
SUSPENSION
SYSTEM
INFLOW
PERMEANT
RESERVOIR
OUTFLOW
PERMEANT
RESERVOIR
SLURRY
RESERVOIR
SUPPLY
Figure Al. Schematic of slurry test column system.
and record the initial sand height before saturation with
SAMPLE DEAIRING, PERMEATION AND CONSOLIDATION
the dry sand to displace the entrapped air within
s connected to the bottom of the samole via the
Slowly bubble C02 into „ . _ _._r
the sample. The C02 is connected to the bottom of the sample via the
three-way valve at the bottom of the base plate.
Open the three-way valve to the C0£ line.
Vent the top of the sample by opening one of the ports on the top plate.
Turn the C02 tank on and turn the C02 tank regulator in the clockwise
direction until 1 psi registers on the gauge.
Bubble C02 through the sample for ten minutes.
Back the regulator off slowly. Turn off the C02 tank and close the
three-way valve at the bottom of the sample.
a.
b.
c.
d.
e.
Slowly saturate the sand sample with deionized (deaired) normalized
water. The sample saturation should start at the bottom and proceed
toward the top of the column. Connect the water line to the bottom
of the column and Tank 1 via the three-way valve.
a. Fill Tank 1 with deionized, deaired, normalized water.
b. Turn the three-way valve on Tank 1 toward column base.
c. Crack open the three-way valve at the column base toward the water
tank (opposite direction'of the C02) to permit slow filling of the
118
-------
column.
d. Water should flow slowly under its own head to fill the column. Once
the water heads in the Tank and column have equalized, Tank 1 can be
placed under nominal pressure (1 to 2 psi) to complete the saturation
process.
e. Saturation of the sample should be very slow to ensure all trapped C02
is dissolved into the water. The entire process may take as long
as eight hours. The rate of rise of the water level should not
exceed 1/4" per minute.
PORE PRESSURE PROBE DEAIRING AS THE SAMPLE IS SATURATING
To unclog and deair the pore pressure probe lines:
1. Submerge all probes in a pail of warm water.
2. Connect the auxiliary pressure line to the manifold bleed line.
3. One by one, blow out all the lines. A line is clear when bubbles appear
in the pail of water. The probes may require disassembly to remove sand
or gelled slurry prior to deairing.
4. When all lines are clear, re-attach the bleed line to the bleed reservoir.
5. Fill the bleed reservoir with deionized, deaired, normalized water. Con-
nect the pressure line to the cap on the reservoir. One by one, force
water through the manifold system and probes at a pressure of 1.0 psi or
less. Insert the water-filled probes into the column just above the
rising water level as the column is being saturated. Force all remaining
air out of the probe and probe tubing at that time.
NOTE: Be careful not to let the bleed reservoir run out of water. This
will force air into the manifold system, the line, and ultimately
into the sample.
FLUSHING CARBONIC ACID
Once all the probes are in place and the sample saturated, continue to
flush the column to remove all remaining carbonic acid from the sample.
Three pore volumes are considered sufficient to remove the dissolved
C02- Three pore volumes are approximately equal to 30,000 cc or 30 liters of
fluid. Let water flow out the top of the column until the 30 liters have been
collected.
BASE LINE K-VALUE DETERMINATION (KD)
The test, consisting of multiple runs, is performed as follows:
1. Fill Water Tank 2 with water and connect to column top plate.
2. Empty Water Tank 1 and connect to column base plate.
119
-------
3. Calibrate or check calibration of the pressure transducers.
4. Determine a head and tail pressure calibration to give the appropriate
"AP," depending on the experimental design. Since the baseline k-value
will correspond to the slurry penetration value of two different pressure
differentials, it is best to run the higher W P test first. For example,
pick head = 30 psi and tail pressure = 20 psi ( AP = 10) for first test,
and for second test, pick head = 30 psi and tail = 25 psi ( AP = 5).
These values need only be approximate because they will have to be
adjusted during the test to compensate for friction losses in the lines.
5. Once the pressures are set for each tank, open the valves. The head
and tail pressures will change due to flow friction losses. Allow the
pressures to stabilize, and then regulate them to design test levels.
(This may not be possible with large flow velocities.)
6. Note and mark the water level in the head tank (ht) and simultaneously
start the time clock. Read and record pressures at various locations.
Select two probes a distance "L" apart and record their stabilized pres-
sures. These pressures should remain unchanged throughout the baseline k
test. Pick pressure locations near the top and bottom of the sample.
7. When the head tank runs low, this signifies the end of the test. Stop
the timer and close the valves. Note and mark the water level in the
head tank
8. Calculations: Baseline Coefficient of Permeability k^
In all tests, the primary parameters needed to comp'ute hydraulic
conductivity between any two pore pressure probes are:
Q : steady state outflow of permeant per unit time; cm^/sec
P]_: total in situ pore pressure reading at arbitrary probe 1, psi
?2'- total in situ pore pressure reading at arbitrary probe 2, psi
L: vertical distance between probe 1 and 2
A: cross-sectional area of soil in column; cm^
With these parameters, steady state equilibrium hydraulic conductivity
(k) is computed as:
* iA
where: . ,. . Pi-P? AP
i = zone gradient = —Lj—^— = —j-—
For the general case, with pressures converted to inches of water column,
lengths in inches, and the elevation of the lower probe (?2) as a datum:
• _ (Pi X 2.31 X.12) + L - (P? X 2.31 X 12 ) + 0
1 - i
120
-------
which is reduced to:
(Pi - P?) X 2.31 X 12 . .
L + L ~
27.72 (A P)
L L
where: A P = difference in pore pressure probe readings, psi
9. Replicate the k^, test using a A P equal to the slurry driving pressure to
be used to test the slurry seals. The two k& values computed should be
relatively close to one another.
CONSOLIDATION AND VOID RATIO CALCULATION
Passing three pore volumes of fluid through the sample under a A P higher
than planned for program testing should consolidate the sample to a stable
height and void ratio.
To determine the void ratio:
1. Determine the height of the consolidated sample in inches.
2. Given: Wwe^ and w determined while filling the column.
The weight of the solids "Ws" can be calculated knowing "w", the
moisture content of the sand placed in the column, as:
The dry unit weight of the sand is:
ii Us
Wd = v
where tf _ * (.52)
v - 4 • ns
and hs is the final sample height in inches
From the above: the void ratio is calculated as:
= Gs YW _
wd
where e = void ratio
Gs = specific gravity of the sand = 2.65 to 2.68*
*Compute void ratio based upon Gs = 2.65 and Gs = 2.68 unless
Gs is explicitly known by virtue of direct test. Report com-
puted range for e (e = 0.65 to 0.69, etc.)
w = unit weight of water (62.4 pcf)
121
-------
APPENDIX B
PROCEDURAL OUTLINE FOR THE OPERATION OF THE SLURRY WALL TANK
The following procedures represent those used in Sequence 2 tests.
This step-by-step outline is for a typical straightforward test sequence
involving a single overburden pressure and a single hydraulic pressure. Test
conditions normally are varied per the applicable expeirmental design, with
corresponding changes in test procedure as required. Reference is made to
the slurry wall tank schematic which is reproduced herein as Figure 81.
1. Cut a new PVC membrane and attach three PVC baffle rings and two PVC
disks per tank detail. (Should be done before starting to build wall).
2. Place the inner and outer cylindrical slip forms in the tank and wedge
them into the bottom baffles of the corresponding diameters.
3. Place moist in situ soil in six inch loose lifts into the center core and
outer ring and compacr to a total height of twenty-two inches. With each
lift of soil, siphon bentonite slurry from the container into the annular
space between the two slip forms. -The slurry level should be kept as
close as possible to the elevations of the insitu sand at all times.
4. Alternately, lift the outer slip form (larger diameter) and inner slip
form (smaller diameter) slowly and icrementally, allowing sufficient time
between each lift for the slurry to penetrate into the insitu sand and
form a filter cake. If in situ soild does not have sufficient binder to
assure center core stability, the center slip form should be lifted in
conjunction with the soil-bentonite backfilling procedure as follows:
a. A temporary support or guide device should be set up above the center
slip form to restrict lateral movement while it is being lifted.
b. Raise the center slip form to a height of about five inches (two
inches above the bottom baffle) and allow twenty minutes (or longer)
for initial filter cake formation. Using the pressurized tremie
pipe, place soil-bentonite over the entire wall area to a depth of
about 4 1/2 inches, moving the pipe around the circumference of the
tank as needed. During the process of backfilling, the displaced
slurry is siphoned into a holding container, taking care not to allow
the level of slurry to fall below the top of the trench.
c. Raise center slip form three more inches and move tremie pipe 180°
around the perimeter of the tank to the location opposite to that
initially used. Allow at least twenty minutes for filter cake forma-
tion before backfilling.' Steps (a), (b), (c), should be repeated
until the soil-bentonite has reached its design height.
122
-------
PERMEANT RESERVOIR
(Ty C MEMBRANE T"
BEARING1
PLATE
V8A
DRAIN
AIR PRESSURE LINE
PERMEANT FLOW LINE
Figure Bl. Schematic of slurry wall tank system.
d. Finally, displaced s.lurry is removed from the upper surface of the
in situ soil, and the upper surface of the soil-bentonite wall is
leveled off manually using a trowel!.
5. Position the membrane/cutoff assembly over the surface of the soil elements
with the baffle rings directly over the surface of soil-bentonite wall and
slowly press the baffle rings down into the soil-bentonite wall. Punch
small holes in the membrane between the baffle rings to allow air to
escape (if necessary). These holes should then be patched using PVC
cement after all air bubbles have been removed and the soil-bentonite has
fully contacted the membrane.
6. Position the three polystrene load distribution plates over the membrane,
center core, and outer ring.
7. Connect the air bladder between the center load plate and tank lid.
8. Lower the tank lid to cover the tank. (The three pneumatic bladders
should be pre-glued on the tank lid to facilitate alignment.)
9. Assemble' and secure the lid to the upper components of the tank reaction
frame.
10. Necessary line connections are made from the tank to the control panel
123
-------
and to both reservoirs.
11. Perform the test sequence as follows:
A. Saturation process
1. Fill both reservoirs with deionized water with valves V9A and V9B
open.
2. Slowly saturate the in situ core using reservoir 1: valves VI, V6,
V7, and V9 open; gauge G2B at 1.5 psi pressure.
3. Slowly saturate the outer in situ ring of sand using reservoir 2:
valves V2, V4A, V4B, V5A, and V8B open; gauge G2B at 1.5 psi pres-
sure. Make a temporary connection between VI (drain) and V5A
(tank) to allow flow of water from reservoir 2 to outer in situ
core.
4. Monitor outflow from valves V7, V4A, and V4B. When outflow begins,
allow thirty minutes to deair the soil before closing both water
supply valves (VI and V2).
B. Consolidation process
r
1. Based upon consolidation test data, consolidate the soil-bentonite
wall to at least 90% of primary consolidation before the start of
permeation. Set G38 to test consolidation test pressure. Keep
V7, V4A, and V4B open to permit drainage during the consolidation
process. •
2. Next, consoldiate the core and outer ring. Set gauges 2A (center
core overburden) and 3A (outer ring overburden) to test consolida-
tion pressure per the applicable experimental design.
3. Continue consolidation until full primary consolidation of the
soil-bentonite wall has been achieved. Close valves V7, V4A, and
V4B.
C. Permeating Process
1. Fill both reservoirs with water or leachate per applicable experi-
mental design.
2. To minimize the elastic deformation of soil-bentonite wall and
thus to expedite the equilibration process, permeant is introduced
stepwise into the center core at an incremental hydraulic pressure
of 20% of the designed hydraulic pressure (per test series) per
. . each application. Before the application of each incremental
hydraulic pressure, counterbalance pressure should be added to the
effective overburden pressure on the center in situ core and the
soil-bentonite wall respectively, to negate the buoyant effect.
The detailed operation procedures are described below.
124
-------
a. Add a counterbalance pressure equal in amount to the incremen-
tal hydraulic pressure to gauge 3A (center core overburden).
taneously add one half amount of the same incremental hydrau-
lic pressure to the soil-bentonite wall using gauge 38.
b. Allow at least one hour for the pneumatic bladders to become
fully inflated under these pressures.
c. Apply the incremental hydraulic pressure to the reservoir
being used. Open valves VI, V6, and V7 if reservoir 1 is
used (V2, V6, and V7 for reservoir 2) to direct flow into top
and bottom of the center in situ core.
d. Repeat step (a) thru setp (c) until the full desired hydraulic
pressure is reached. It is advisable to exercise step 3 and
step 4 (following) during this loading process..
3. Record date, time, gauge pressures, valve postions and reservoir
level on test mode log sheet.
4. Monitor system, taking inflow and outlfow readings periodically
per QA/QC accuracy/precision requirements. To limit measure
error and to keep the applied hydraulic pressure approximately
constant, readings should be taken when the water level of the
active reservoir has dropped between 5 and 20 centimeters.
Switch reservoirs after each reading.
5. Compute hydraulic conductivity of soil-bentonite wall based on
inflow reading (outflow for checking only).
6. Continue system monitoring and computation of hydraulic conductiv-
ity until equilibrium flow conditions are established.
7. Depressurize, disassemble, and clean out tank for next test or
excavate, sample and document conditions per applicable experimen-
tal design.
8. Check all pressure gauge calibrations prior to initiation of next
test sequence.
125
-------
TABLE Bl TEST TANK COMPONENT SPECIFICATION
COMPONENTS
TANK
Tank body
Tank Lid
SUPPORT SYSTEM
W beam
S strinyer
Threaded bar
Thread nuts
Bearing plate
Timer
AIR PRESSURE
SYSTEM
Inner Tube
Pressure
gauge
Regulator
FLUID FLOW SYSTEM
Reservoi r
3-way valve
On-off valve
Tube
Connector
Reducing Bushing
CONTROL PANEL
MISCELLANEOUS
PVC membrane
Loading plate
SPECIFICATIONS
QUANTITY
1
1
4
12
8
32
16
2
1
1
1
4
4
6
2
10
4
4
200'
8
2
1
1
1
1
2
DIMENSION
D x H x t
44"x23"xl/16"
45"x3/16"
W8x28 - 6 '2"
S 3x7.5
1-1/8" - 5'
2-1/8"
6"x3"xl/4"
8"x8"x5'
10" x22"
13"
6"
4-1/2" lOOpsi
2-1/2" lOOpsi
range 0-160psi
6"x26"xl/3"
1/4"
1/4"
3/8"
2'x5'xl/2"
46"
I0.2'xOD.12"
12-1/2,19-1/2
20-1/2,41-1/2
12"& 20"
MATERIAL
Stainless
Steel
S.S.316
C.S.(A-36)
C.S.(A-36)
C.S.(A316)
C.S.(A316)
C.S.(A316)
Oak Wood
Trailer tube
Regular
Regular
Test gauge
Reg. gauge
Fairchild
PVC
S.S.
S.S.
Brass
Nylo-Seal
Nylo-Seal
S.S.
Plexi-glas
gage 10
Polystyro-
foam
REMARKS
Two 3" x 3/8" strip
metal plates are
wound around the
tank at top and
middle portion.
Two different sizes
of collar are welded
to the bottom of
lid.
C.S.: Carbon Steel
Length varies from
3' to 4'.
Accuracy: 0.2 psi
Accuracy: 0.5 'psi
Oak wood support.
Three PVC rings
attached
126
-------
APPENDIX C
ROUTINE GEOTECHNICAL TEST RESULTS
127
-------
ro
oo
§2
o
100
vo
80
70
i
| 60
>•
E in
z 30
c
§«
m
IU
ft.
JO
JO
10
0
30
U.S. STANDARD SIEVE OPENING IN INCHES US STANDARD SIEVE NUMBERS HYDROMETER
4 4 ] t tV> 1 H Yl H 3 4 » « 10 1416 20 30 40 30 70 WO 140 700
-
-
-
-
-
-
1 1 1
1
1 1 1
1
1
IV
K) 100 30 10 3
GRAIN SUE M
COi
SMMIINO
:UMPOS .
nr) 1 °"Am
1 count IM
iilv 0*01 rm
st
^
^
s
1 1
0.9
ailMETHS
i
'«.
-
V
-
--
v
s
s
k
V
\
A
1
.
O.I 0.03 0.01
iANO
COAIU
OASSVKAIION
33%
67%
^I£_Sj
"TR. H
\n
AY
GRADATION CURVES
MtowM | nut
NAIWX
u
31
n
15
n
18
.. _
.
I :
i
*
-
—
^
^
fcV
0 003 0.0
Ml Ol ClAY
0
10
20
30 5
o
»!
5ol
8
«o £
s
ro *
10
90
100
01
~o.c, 3210-07/08
CTR HILL SOIL #1
AtfA ~-
KMMO^
o —
».n 8/21/84
Figure C-l. Grain size distribution, Center Hill No. 1 Soil
-------
UNIVERSITY OF CINCINNATI
CENTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
U5
148
139
in
in
»-
i*.
=3
U
o= 120
UJ
o.
•
fS
O9
_J
1 Ml
»—
as
ta
U
g no
^
i
IQK
100
95
90
8]
1 ; '!,.'!! \
•*-*-, ^
1 H
I \ \
^
m
, V
— VA
. i- V
^
t=4
1 — 1-\
1
1
K
a
i\
^
fp^
^^
— \ — \ — \
^
^
^
• '- • f - - '
/
/ *-*
\
1
) 1!
h '
t^
V\^
rV\*
MfA^
4rW
t^m'
s\\\
~vv^
b^
\=**
==±^
V
2t
COMPACTION TEST
PROJECT 3210-07/08
JOB NUMBER — DATE 9/21 & 9/24/84
SAMPLE NUM'BE'R TS AKUASEAL
DEPTH
TESTED BY RWK CALCULATED BY RWK
CHECKED B7 ^~
MAY novnCKKTTY ^n\ ^A^
LBS/CU.FT. 115.3 113.2
OPTIMUM W.C.% 13.9 15.2
NATURAL U.C.%
HAMMER WT..LBS 5.5
DROP, IN. 12
NO. LAYERS 3
NO. SLOWS/LAYER 25
OIA. MOLD IN. 4.0
HEIGHT MOLD IN. 4.5
VOL.. MOLD CU.FT. 1/30
COMPACTIVE
EFFORT. FT. LBS/CU.FT.
SIMPLE CLASSIFICATION
CENTER HILL #1 SOIL:
33% ESTE SAND/67% CENTER HILL CLAY
KEY
'* o BENTONITE ADDED IN SLURRY
J* FORM
^ *
-\? A BENTONITE ADDED IN POWDER
W* FORM
r^A
*^r\
\ \ \
\. »,\ . .\
\'\ \
\ \ \
?rW\
— Vt\
'\\\
^V>A
=:2^^
s. \ \
\ \. \
'\\ \
* \ \
r \ \|\
1 — • — \ X \
\ i\ \
^ j ^ \
\ \\
' \\ \
\ -\---X
! ^^
i V X
\
X
Z5 30 35
RATER CONTENT - PERCENT
Figure C-2. Standard Procter test results for 1% soil-bentonite
129
-------
UNIVERSITY OF CINCINNATI
CENTER HILL SOLID AMP HAZARDOUS WASTE RESEARCH FACILITY
us
no
135
130
tfe
a
u
a, 120
v»
1 its
t—
o
•
>*
a MO
tos
too
99
90
as
l..ln,vl
•' i'' 1
-L- — l—l
' \ " 1 \
JvA
Y ^ .^
— V \ \
— \ \ \
— vv^
. .\ \
' .'\'\
• r-V1
:-, \
-
i b— r
=\
k •
HFVA
M^F^
H^
N
0 1
1
V
=\
^y ,
fF^
\xV
^P^\ ;\
•^ ^V-\j
^^ \
^
5 2
COMPAaiON TEST
PROJECT 3210-07/08
JOB NUMBER — DATE 1
SAMPLE NUMBED 2* AtCWAs^AL
3EPTH —
TESTED BY RUK CALCULATED BY
:HECKED BV
MAX DRV DENSITY fnl (a)
LBS/CU.FT. 114.9 111.8
DPTIMUM W.C.i 15.0 16.4
WTURAL U.C.S — —
HAMMER WT..LBS S.5
HOP. IN. 12
NO. LAYERS 3
HO. BLOWS/LAVEft 25
J1A. MULD IN. 4.U
1EIGHT MOLD IN. 4.5
/OL..MOLO CU.FT. W30
COMPACTIVE
EFFORT.FT.LBS/CU.FT.
SIMPLE CLASSIFICATION
CENTER HILL #1 SOIL:
331 ESTE SAND/67S CENTER HILl
KEY .
0 BENTONITE ADOEt
'$ SLURRY FORM
-------
UNIVERSITY OF CINCINNATI
CENTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
0.1 0.2 0.3 0.kO.J
1.0
0.9
0.8
0.7
o-0.6
1
JO. 5
0.4
U . 0
0.2
\
C
.1 0.2 0.3 OA
E
>
«•
XJ
Pr
Typ. ,* ap^i.. ARTIFICIAL
DIM 2.5 u.
Bt . 48 la.
Orarturdca Pnimm, PQ ?/*^ t
Pneouol. Pmnm, p, 0 T/«» ft
C.*...l. I=d«, Ce 0.119
ci»..iri=.tt«.
U.
PL
Cr 0.030
». 2.7*
"»
tamrfci
C.H. Soil No. 1
1
tc
»
•
•
1
•••or*! f, '
a 3 k
U
N
<
f
^
V
2 3 *
t/H ft
J
^
•i
*
H
n
5
Smtart TM«
u»t*r CoBtort, »0
Void tetlo, «a
* 9
Onr Duwtty. r4
12:34 >
0.56
92.96 *
-- is/ft
10 20 as
10
20 2)
Aft«TTt.t
•f
•f
«f
)
16.42 *
0.324
134.3 *
n^t 3210-07/08
An*
tarlnc So.
n
9«^»l« He. "
n»t»
CONSOLIDATION TEST REPORT
*ASSUMED
Reproduced from
best available copy.
Figure C-4. Consolidation test results for Center Hill No. 1 Soil
131
-------
UNIVERSITY OF CINCINNATI
CENTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
0.1
1.0
n Q
0.8
0.7
,-0.6
a
JO. 5
0.4
0.3
0.2
0
.1
0.2 0.3 0.1
(
• Jt
1.
s
*.
»a»
t^
s
• I
1
U
(>
^
•
\
• •
2 1
ri
C^
i±
\
^
• i
k
!ti
i^
5
cs
1 M
!i
•
tJ
0.2 o.jo.kaj i a 3 k j
tnteim, t, t/iq ft
tw<**~i~ "ARTIFICIAL
OUa 2.5 u
. it 1.0 i..
Orcrburdra Pntran, 90 " T/n ft
Ptveeual. Fr»iun, ya U T/«j ft
GavtvulOB lafas, Ce 0.251
CUMlTloattc.
u. 36
n. 13
Cr 0.058
a. 2.7*
D10
«-*• 1% SOIL-BENTONITE
C.H. No. 2 SOIL
Biten T*rt
u»t«r Gamut, v.
Told tatla, «9
tetmttcB, S0
Bnr Dnittr, T4
36.12 <
1.08
97.50 *
— i»/«
10
10
» 2?
20 2)
W^r?.«
•f
•r
»f
j
11.75 *
0.52
51.01 »
n^. 3210-07/08
A~ —
Bona« So.
teptb
n
SB^B!« Bo.
Dat, 12/84
CONSQUDATION TEST REPORT
*ASSUMED
Figure C-5. Consolidation test results for 1% soil-bentonite
132
-------
BHIVEKSm OF CINCINNATI
CENTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
0
1.0
0.9
0.8
0.7
0-0.6
a
fO.5
0.4
0.3
0.2
0
1 0.2 0.3 O.kaj
(
>>
s
s
1
N
.1 0.2 0.3 o.k a;
»*
T^of8p«i». ARTIFICIAL
DIM 2.5 la.
R 1-0 te.
0«*rbux4*a Pnuun, po 0 T/«q ft
PneoBMl. Pmiun, ;e 0 T/«q ft
c.,™.!. I**,, ce 0 . 304
Cl««i1flc«tla»
u. 42
PL 16
Cr 0.040
o. 2.7*
Dio —
M-*. 2% SOIL-BENTONITE
C.H. No. 2 SOIL
1
r
>
Si
a
^
• •
1
•»ur*, p.
a J *
I
S
(k>
s
•
^
•
>
4
)
2 3 *
T/«q ft
5
S
s
^
*)
5
Btfon T»rt
««t»r Caaunt, »a
Told Mtls, •.
o
* o
Cry Dn*lt7, 74
37.06 %
1.04
89.1 #
-- lb/ft
10 20 aj
10
20 25
After T«re
"f
•f
sf
3
19.9 *
.634
85.5 »
^.ct 3210-07/08
AIM --
Barta* !»o. —
Drpth
B.
9«d* So. — —
DM*
CONSOLIDATION TEST REPORT
"ASSUMED
Figure C-6. Consolidation test results for 2% soil-bentonite
133
-------
UNIVERSITY OF CINCINNATI
CENTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
0.1 0.2 0.3 O.kO,?
.3
1.2
1.1
1.0
0.9
lo.s
•a
£
0.7
0.6
0.5
0.4
0
)•
•
Ei
.1
k (
>
• ••
)
\
M
S
•
*>«.
^
!9
u
!•
• •
0.2 o.jouaj
h
*»«•»—• ARTIFICIAL
m.- 2.5 ta.
R 1 ia.
Ovtrtwten Pr>»m, Po 0 T/KJ ft
PraeoBMl. Pn«un, p0 0 t/«q ft
Capmilea Indu, Cc 0 . 303
Clu.lflc.tia.
U. 44
R. 17
Cr 0.033
o. 2.7*
Dio —
»«t. 3% SOIL-BENTONITE
C.H. No. 2 SOIL
1
1
><*
^
•
k
^
mm
a 3*
v
t \
\
\
• 1
1 23
r»»«ur», p, t/«q ft
\
A
"(>
*
5
s
N
i •
s[
=^>
3
Sftan T«rt
Vuar CAUnt, »o
7ot« tette. .0
a»tw»ti«, s.
Dry Dtt«lt7. r4
42.27 %
1.34
86.75 *
— in/ft
u 20 aj
10
20 2?
After Trat
'f
V
8f
3
20.92 *
0.66
85.58 *
Pr»J.« 3210-07/08
AIM
Bort^.0. —
Drpth
E.
Sc^niA So* — ™*
„«. 12/27/84
CONSOUDATION TEST REPORT
*ASSUMED
Figure C-7. Consolidation test results for 3% soil-bentonite
134
-------
uurvEBsm OF CINCINNATI
COTTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
1.10
1.00
, Pe 0 ?/*4 ft
Pneouol. Pr»««oi», pc 0 ?/•% ft
rratim«iao Zadu, C. 0.378
g...tfi»tioc
iii
PL
^
u
.-*
^
-(
N
^
^
'
U3 20 25
^
^
S
c
E
CM
u
II
010
0.05
0.0
) 10 20 8}
DS1
Brfar.T.«
*««• Cont«Bt, v0
Told IlKtta, «0
SMtntlCB. 30
teT Dt»ltT. 7d
24.'9 -*
1.156
101.0 *
iVft3
Aft«.S«t
"f
•f
8f
23.8 *
0.826
100* *
C9 = Cr 0.048
°. 2.7
°10 —
*-**• Remolded C.H.#2
soil with
2% bentonite
fc^., 3210-07/03
*». —
Borto« So.
QrpUl
SI
3-pl.So. —
n*. 5/85
CONSOLIDATION TEST REPORT
•"Assumed
Figure C-8(a). Consolidation test results for remolded 2% soil-hentonite
135
-------
UNIVERSITY OF CINCINNATI
CENTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
0* " "
10
20
30
GO
LoJ
0
1—4
« 40
o
DEFORMATION IN 1
_g § S g ' ' S
Jf ,
h1
^- —
>-— =
u
=3
|^N,
•M
"\ f •
-<)..
"( ^ •
T 7 •
T^ »
i
"~ 1*"
i
•OT<: WMOIKUIOI
1
TM
11 S •
' ^* *^
i( \_^
^^
••;>
^
rr w i
... tcr
5C
* ' L
"^
^
^
A-
—
•^
t
(X^
*t
1
' *fi
«
» • Mi a »o me ax
1 1 llllllllll
S L
-" •«
-I..S
T
-
t50=4.7min
^
—I
1
( -r r
*;•-.
"-.
"urr —
* ^ ?
: ;i^*
tcQ=3.3 min
F
SO*1'8
U '
^•11
Ii
= 0.96
min
)L,
s
min
"f-;i "
*~i
icvwa
m T/ie rr.
1 1
>^«
•Y
picn
^e
m^
5
n
i
1
"TT=*
50=16ir
^
^Y
~
-i>,.
v
J=
,5..
i
pirbr'T
>-
-w> J
-t . «
nn •
.ra^ A ^*
v tt ~\*
-Ul._/
1
1 j!
j •
7 ;
a
d i !i
0.25 TSF
1
lit
1 1 1 1 in
0.50 TSF
""iTl
i
1
1.0 TSF
I
*=n
•7)
n
2.0 TSF
i ]
U
1
1 !
i
I j
iii in
4.0 TSF
! 1
i
8.0 TSF
1 '
LI «J uit 1 <• a •<••• mo raoo ••
TIME IN MINUTES
r^n 3210-07/08: 2% SOIL-BENTONITE; C.H.#2 SOIL/ENVIROGEL 20C
u»«
lortoc to. ~ ~ !
u.1. to. —
n """
CONSOUOATION TEST-TIME CURVES
Figure C - 8(b).
Consolidation test time curves for
remolded 2% soil-bentonite
136
-------
0
0.64}
0.60
0.56
0.52
0-0.48
I
1 0.44
0.40
0.36
0
.1 0.2 0.3 O.Ik 0.5 1 2 3 k 5 10 20 25
*s
s
hv
's,
«
1
'\
1
1
0-
•^
^
»••
s
M>
V
>
1
1
1
i
-rC
S,
)'
^
s
1
— *•
1,
k
)"
S,
S
1
I
•«;
S
.J
5
1
T
,
v
f==!
4j
1
s
•**•
SJ
y
Hi
>
^«
.1 0.2 0.3 o.u as l 2 3 ik
PrMiure, p, T/»q ft
Type of 3p«cmu ARTIFICIAL
Warn 2.5 la.
Ht 0.516 la.
Overburden Preiiure, Po T/tq ft
Preconjol. Prs«»ure, pe Q T/iq. ft
Cospreiiloa Ladex, Ce 0.102
Clanlfleatlon SM-SC
u 20
PL 14.8
c. 2.70
Dio 0.08 nun
Hemrt. Remolded C.H. 32 Soil
with IS
Bentonite
^
^
Ss
--
•0
.
.c
CM
u
II
0.40
0.20
0
5 10 20 25
Before Tact
Water Content, vo
Void Ratio, «Q
Saturation, 3g
Dry Denilty, 7.
23:6 >
0.638
100 >
After Tcrt
"f
«.
6f
101.5 "/«3|
15.1 ^
0.396
100 ^
P^J66* 3210-07
Area
Boricg Ho.
Septn
n
Saople Ho.
fete
CONSOLIDATION TEST REPORT
Figure C-9(a). Consolidation test results for remolded 1% soil-bentonite.
137
-------
UNIVERSITY OF CINCINNATI
CENTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
0.1
1000
Note: Numbers beside curves
are pressures (T/sq. ft
10 100
Time (minutes)
1000
l]tn 3210-07: 1% SOIL-BENTONITE; C.H. i»2 SOIL
CONSOUDAT10N TCST-TIMC CURVES
Fiyure C - 9(b). Consolidation test time curves for
remolded 1% soil-bentonite
138
-------
UNIVERSITY OF CINC1HNATI
CEJTER HILL SOLID AND HAZARDOUS WASTE RESEARCH FACILITY
ATTERBERG LIMITS DETERMlHATIOH
#3210-07
S2
LIOUIO LIMIT,
21
54 q? I 70 01
Kl 0« PIT 1011, • TH( | 4fl in
C1 17
Sq 07 I SI tjg
u.-j.j 27= |
| 4 a*;
s s«;
7 In
17 .74 1 17 on
OQ
T>
,(j-.».;| 16.J6 I 21.12
26.32
27.06 I
as
••Mi CD«TC«T. xf-
»•>
17.1
18.3
18.5
21.6
23.4
d7
i\
I ->i
18 I
14.8
5.2
15 .20
lumber of blows
igi tuns •
0« t I 1011 • IHt
TO • i
17 ?a
30 q?
or o T ion • Ti»e
33.13
41 02
38 97
»!•.?,
32.05
32.06
j. it c* »>»' JDH. t,fa.-i.j 5.J9 I 3 97
63!
15.5
14.2
14.7
•!:••:•.•.:< l^^S
Material: IS 5/B mixture CH ?2 soil
Fiyure C - 10. Atterbery Limits of 1% soi1-bentonite
139
-------
APPENDIX D
STANDARD LABORATORY PROCEDURES
140
-------
PROCEDURE LIST
TEST SLP NO.
Slurry Mixing Procedure (Powdered) S-01
Slurry Mixing Procedure (Hydrated Bentonite Paste) S-02
Slurry Re-Mixing After Hydration S-03
2% Soil/Bentonite Batching Procedure (20 Ib) S-04A
Specific Conductance QC Check QC-01
Laboratory SIump Equivalent EX-01
Modi fied Swel 1 Test EX-02
Filter Cake Hydraulic Conductivity: API Filter Press EX-03
Soil-Bentonite Hydraulic Conductivity: API Filter Press ....EX-04
Water Content G-01
Specific Gravity of Solids, Gs G-02
Grain Size Analysis G-03
Free Swel 1 G-04
Unit Weight G-05
SI ump Test G-06
pH Measurement E-01
Slurry Conductance E-02
Speci f i c Conductance E-03
Slurry Density (Mud Weight) 1-01
Marsh Funnel Vi scosity 1-02
Rotational Viscometer 1-03
API Fi 1 trate Loss 1-04
141
-------
SLP No. S-01
Revision 2-062085
Sheet 1 of 2
SLURRY MIXIN3 PROCEDURE
(Powdered Bentonite)
EQUIPMENT:
electric stirrer and stand
2000 ml beaker
stopwatch (accurate to 0.1 sec)
PROCEDURE:
balance (accurate to O.Olg)
1000 ml graduated cylinder
spatula
To mix a 5% bentonite slurry (weight.-volume basis):
1. a. A 5% bentonite slurry is defined as 100.0 grams oven-dry bentonite
in 2000 ml total liquid volume (TLV) of water or water:chemical
mixture (per applicable experimental design):
100.0 g
2000.0 ml TLV
= 0.05 g/ml = 5%
NOTE: Concentrations other than 5% or slurry volumes other than
2000 ml defined on same basis.
b. Bentonite for testing purposes should never be oven-dried; an
"equivalent dry weight" (EDW) is determined on the basis of the
moisture content of a representative sample as described below.
2. Component Proportioning:
a. Determine water content (w%) of air-dry bentonite per SLP No. G-01 .
b. Determine EDW of air-dry bentonite as : EDW = 100.0 (1 + W%).
c. Determine volume of water (Vw) in EDW sample as Vw = 100 (w£)
(weight = volume assumed).
d. Determine additional liquid volume (ALV) required for mixing as:
ALV = 2000 - Vw.
Example: a. w% = 0.11 (decimal basis)
b. EDW = 100.0 (1 + wX) = 100.0 (1.11) = 111.0 g
c. Vw = 100 (w%) = 100 (0.11) = 11.0 g
d. ALV = 2000 - Vv
2000 - 11.0 = 1989 ml
3. Decant the ALV of deionized water, selected test chemical, or water:
chemical mixture into 2000 ml beaker.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
142
-------
SIP No. S-01
Revision 2-062085
Sheet 2 of 2
4. Position lab stirrer and stand with paddle blade about 1/2" to 1" off
bottom of beaker and slightly off center of beaker^ .
5. Start lab stirrer and add bentonite to water gradually over a three-
minute period. Dispense bentonite directly into vortex created in
beaked.
6. After all bentonite is added, continue stirring to a total mixing
of five minutes.
7. Decant slurry and visually check beaker for significant bentonite re-
sidue or lumps. If significant bentonite remains in the beaker,
repeat steps 5-7.
8. Store slurry undisturbed in moisture room (use storage time per
applicable experimental design).
FOOTNOTES:
1 This paddle height facilitates mixing, and being off-centered decreases
the chances of bentonite being lost to stirrer rod and paddle blade.
2 Bentonite can be added by convenient means, i.e., via spatula, a clean,
dry funnel, etc. Adding the bentonita into the vortex facilitates mix-
ing and decreases the chance of bentonite sticking on stirrer rod and
paddle blade.
3 At this time, the paddle blade may be centered in the beaker, if desired.
The total mixing time should be doubled for granular bentonite to enhance
mixing.
REFERENCE: Center Hill in-house procedure
143
-------
SLP No. S-02
Revision 1-032585
Sheet 1 of 2
SLURRY MIXING PROCEDURE
(Hydrated Sentonite Paste)
EQUIPMENT:
electric stirrer and stand balance (accurate to O.Olg)
2000 ml beaker 1000 ml graduate cylinder
stopwatch or timer spatula
2000 ml Nalgene container 500 ml wash bottle
PROCEDURE:
NOTE: This procedure for preparation of weight: volume chemical-bentoni te
slurries using water-hydra ted bentonite paste. Mixing should be
done in Hazardous Waste Lab (one-pass air, room 130) using appli-
cable safety equipment and procedures.
1. Determine water content of selected bentonite paste per SLP No. G-01
(if not pre-deter mined by at least four separate measurements).
2. Based on water content results, weigh out quantity of paste equivalent
to 100 grams dry bentonite
3. Based on water content results, compute quantity of "available" water
in bentonite paste sample (W|j).
4. Determine total water volume CAj) and total chemical volume (Cj) re-
quired for 2000 ml final volume of desired slurry.
Example: test slurry = 5% bentonite in 252 acetone. Wr- = 2000
ml x 0.75 (proportion of water) = 1500 ml. Cj = 2000
ml x 0.25 = 500 ml.
5. Suotract quantity of water in paste sample C^} from total water
required (Wj) to give volume of mixing water (Wm).
6. Decant roughly three-fourths of computed Wm into 2000 ml beaker, pour
remainder of Wni into clean, dry 500 ml wash bottle.
7. Position lab stirrer and stand with paddle blade about 1/2" to 1" off
bottom of beaker and slightly off center of beaker.
3. Start stirrer and slowly add equivalent dry bentonite quantity (3d)
using spatula.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Was-ta Research Facility
144
-------
SLP No. S-02
Ravi si on 1-032535
Sheet 2 of 2
9. After Bj has been added to the extent practical, carefully wash weigh
boat and spatula witii remainder of mixing water volume (Wm, step 6),
taking care that all washings are directed into the 2000 ml beaker.
10. Decant remainder of mixing water (Wm, if any) into 2000 ml beaker and
start timer.
11. Continue stirring for a total of 10 minutes; add total chemical volume
(Cj) during final two minutes (to minimize evaporative losses).
12. Decant slurry and visually check beaker for significant bentonite
residue or lumps. If significant bentonite remains in beaker, repeat
mixing (step 11).
13. Decant thoroughly mixed slurry into Nalgene storage container, place
in moisture room and leave undisturbed for 24 hours (or other storage/
hydration time per applicable experimental design).
REFERENCE: Center Hill in-house procedure.
145
-------
SIP No. S-03
Revision 1-032585
Sheet 1 of 1
SLURRY RE-MIXING AFTER HYDRATION
EQUIPMENT:
electric stirrer and stand
2000 ml beaker
hydrated slurry mixture
stopwatch (accurate to 0.1 sec)
PROCEDURE:
To re-mix a bentonite slurry after hydration:
1. Shake slurry container to suspend any sediment.
2. Decant slurry into clean, dry 2000 ml beaker. Step 1 may have to be
repeated intermittantly.
3. Position lab stirrer, stand, and beaker containing slurry with paddle
blade 1/2" - 1" off bottom of beaker per SLP No. 01.
4. Start lab stirrer and mix slurry for at least five minutes. Greater
mixing time may be required if slurry has gelled during storage.
Mixing time should be identical for all slurries of the same bentonite
concentration.
REFERENCE: Center Hill in-house procedure
CENTER IHILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
146
-------
SLP No. S-04 A
Revision 0-052385
Page 1 of 1
2% SOIL/BENTONITE BATCHING PROCEDURE (20 lb.)
EQUIPMENT:
scale (accurate to .1 lb.)
mixer
1000 ml and 100 ml graduated cylinders
spatula
PROCEDURE:
1. Determine moisture content of bentonite (wb) per SLP No. G-01 .
2. Weigh 20 Ibs. of previously mixed Center Hill #2 soil (50% Bushelman
medium sand, 20% F-95 Ottawa sand, 20% Center Hill clay, 10% Este
sand).
3. Weigh equivalent of .4 Ibs. of bentonite [Wb = .4(1+ wb)]. The weight
of water in the bentonite is Wwb = Wb - .4.
4. Thoroughly blend soil and bentonite in mixer for five minutes.
b. To achieve 26% water content in soil-bentonite, add (5.2 - Wwb) Ibs. of
deionized water (1 liter = 2.2 Ibs.).
6. Test actual moisture content per SLP No. G-01.
7. Label and store sample in moisture room.
REFERENCE: Center Hill in-house procedure
= CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
147
-------
SLP No. QC-01
Revision 0-011386
Sheet 1 of 3
SPECIFIC CONDUCTANCE OC CHECK
Two OC samples will be run whenever any data is generated. These OC
check samples will be analyzed before any samples are analyzed. If the
expected results are not obtained, no samples will be run until the
problem can be identified and corrected.
OC SAMPLE PREPARATION
Place approxiately 2 g of anhyHrous KC1 in a snail beaker (100 ml)
in a 103°C oven for at least 24 hours. Remove from oven and place in
desiccator for 1/2 hour to cool to room temperature. Dissolve 745.6
mg of the KC1 in the purest laboratory water available and dilute to
1000 ml in a volumetric flask. Transfer to a glass-stoppered glass
bottle and clearly nark as follows:
0.0100N KC1
745.6 ng/1
conductivity = 1413 uu/cn
DATE:
INITIALS
This OC sample may be stored and used until either insufficient
volume remains OR the sanple is over 3 months old.
Prepare a second QC sanple by carefully pipeting 10 ml of the
.0100N KCl into the purest laboratory water available and dilute to
•1000 ml in a volumetric flask. Clearly mark this flask as follows:
0.0001M KCl
conductivity = 14.94 uy/cn
DATE: INITIALS:
This sample must be prepared fresh with each analysis.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
148
-------
SLP No. QC-01
Revision 0-011386
Sheet 7. of 3
Using the specific conductance SLP and the associated data sheet,
measure the conductivity of both OC sanples. If either one of the QC
samples fails to give the expected results, the analysis must be stopped
and corrected before any data can be generated.
Acceptable Range
1413 yy/cm 1398 - 1428
1413*1% neter accuracy +1 for possible conductivity of water
14.94 uZJ/cm +4 meter accuracy +1 for possible conductivity of water
10 - 20
149
-------
SLP No. QC-01
Revision 0-011386
Sheet 3 of 3
UNIVERSITY OF CINCINNATI, CENTER HILL RESEARCH FACILITY
DEIONIZEO WATER CHECK
DATE
INITIALS
TEMP (°C)
CONDUCTANCE ( umhos )
(uncompensated)
CONDUCTANCE ( unhos )
(temperature conp.)
ALL CHECKS DONE WITH A TEMPERATURE COEFFICIENT SETTING OF 2.0
' RANGE SWITCH SETTING OF 2umhos
150
-------
SLP No. EX-01
Revision 0.090385
Sheet 1 of 3
LABORATORY SLUMP EQUIVALENT
EQUIPMENT:
slump cone cross bar
rod ruler
funnel 10W40 spray oil
soil pan large mixer
calipers stop watch
sand cone hand scoop
rigid mold compaction base water wash bottle
2 wing nuts
SLURRY MIXING PROCEDURE:
5% slurry - Follow standard mixing procedure, SLP No. S-01 .
10% slurry - Follow standard mixing procedure, but double the mixing time.
NOTE: Allow slurry to hydrate for 24 hours prior to using, or other
time interval per applicable experimental design.
S/B MIXING PROCEDURE:
1. Add 30 pounds of dry C.H. #1 soil (67% clay, 33% sand) to mixing bowl.
2. Add the required amount of slurry to the soil over a 3-5 minute
interval while the mixer is running at speed #1.
3. Determine amount of water needed to approximate desired slump. Add
water by filling wash bottles and squirting the water into the center
of the mixing bowl. Allow S/B to mix for at least an additional 15
minutes after all of the H£0 has been added.
4. After 15 minutes remove mixing blow from mixer. With a spatula,
scrape the sides and the bottom of the bowl to loosen any dry soil
that may be sticking to the inside of the bowl.
5. Allow S/B to mix for an additional 15 minutes.
6. Once mixing is completed, allow sample to hydrate in moisture room
for a length of time per the applicable experimental design.
CENTER 'HILL
United States Environmental Protection Agency
Solid and Hazardous Was'ta Research Facility
151
-------
SLP No. EX-01
Revision 0.090385
Sheet 2 of 3
SLUMP EQUIVALENT PROCEDURE:
1. Invert small sand cone and set it on top of the bottom side of a clean
moisture tin.
2. Spray a light coating of oil inside of the cone.
3. With a spatula fill the cone to 1/3 its volume. Rather than rodding
the sample, hold the cone down and apply a slight pressure to the sam-
ple with the broad side of the spatula. Repeat this procedure two
more times. Strike off excess S/B with spatula blade.
4. Dampen compaction mold base with M20 and place the center of the base
over the bottom of the cone. Turn mold base and cone over and slide
moisture tin off. Make sure cone is full.
5. Screw wing nuts onto mold base and slide crossbar down to the top of
the cone.
6. Adjust wingnuts so the bar rests on top of the cone.
7. Remove bar.
8. Slide cone straight up with no torsional motion. Remove cone within
2-3 seconds.
9. Replace crossbar and measure slump with depth gauge.
NOTE: Be sure to subtract the thickness of the depth crossbar from
the gauge reading to determine slump. When using depth gauge
to determine slump, take »5 readings and average them to
determine slump.
NOTES:
• For best results, try to obtain 2-,4-,6- and 8-inch slumps for the 1,
3, and 5 percent S/B mixtures. Do not attempt to predetermine water
contents because 1, 3, and 5 percent S/B all have different W% for iden-
tical slumps.
• Immediately after running slump test, put the leftover S/B in a con-
tainer with a lid and store it in a moisture room.
• For the 24-hour and 1-week tests, make sure the S/B is mixed thoroughly
before running the slump tests, since the water tends to separate from
the S/B over time.
• Each time a set of slump tests are done (3 slump cone tests and 3 sand
cone tests), at least 3 water content tests should be taken. This is
done to verify that the water content of a particular sample is remain-
ing constant.
152
-------
SLP No. EX-01
Revision 0.090385
Sheet 3 of 3
• After every slump trial the cone and soil pan must be washed off.
After every trial with the sand cone, the cone and base must be washed.
• After running a sand cone trial, discard the S/B used in that trial so
that the oil does not contaminate the S/B used for future tests.
• To determine W%, average the W% from both the slump test and sand cone
tests.
• To determine slump, average the three slumps.
153
-------
SLP No. EX-02
Revision 1.051986
Sheet 1 of 2
MODIFIED SWELL TEST
EQUIPMENT:
consolidometer frame
consolidometer dial gauge
floating ring consolidometer
consolidation ring
2 porous stones
2 filter papers
balance accurate to 0.01 g
fume hood (or equivalent)
PROCEDURE:
250 ml beaker
8" evaporation dish
metal spatula
airtight container
weigh boats
timer
mortar
pestle
1. Obtain approximately 40 grams of "air dry" bentonite powder.
2. Determine the water content of approximately 20 grams of the ben-
tonite powder per SLP No. G-01.
3. Place one porous stone in the consolidometer. Fit the consolidation
ring around the stone. Place one filter paper within the ring on the
stone (trim paper to fit if necessary).
4. Weigh out 20 grams "equivalent dry weight" of the remaining bentonite
powder in a tared weigh boat and spoon into the consolidation ring.
Level with metal spatula.
5. To level sample completely, drop floating ring assembly five times
from a height of approximately 1/2 inch.
6. Place the other filter paper (trimmed) within the ring on top of the
leveled sample. Then place the second porous stone on top of the
filter paper, alignment hole up.
7. Place the consolidometer on the rigid base of the consolidation
frame. Center the consolidometer such that the tip of the dial
gauge stem will rest in the alignment hole of the top porous stone
when lowered. Record initial dial gauge reading.
8. Fill the consolidometer with deionized water just to the top of the
consolidation ring. Start the timer and record initial dial reading
when the bentonite starts to swell as indicated by initial movement
of the dial gauge.
9. Take a reading once every hour for seven hours.
CENTER HILL
United States Envlronmpntal Protection Agency
Solid and Hazardous Waste Research Facility
154
-------
SLP No. EX-02
Revision 1.051936
Sheet 2 of 2
10. With every reading, check the water level and adjust, if necessary,
so that it remains at the top of the consolidation ring.
11. After seven hours, extract the sample from the ring and inspect the
porous stone and ring for signs of "binding" or other problems which
might affect swell values.
12. Plot raw data as inches swell versus time in hours.
13. Repeat steps 1-12 for the same bentonite, modifying step 1 for the
chemical "modification" tests as follows:
a. Place 40.0 g "equivalent weight" bentonite in a 250 ml beaker.
Add 50 ml chemical and mix thoroughly using metal spatula
(additional chemical may be added if needed to completely
saturate the bentonite, but the total volume of chemical per
test should be held constant if multiple tests are planned).
b. Pour bentonite/chemical slurry into 8" evaporation dish and place
in fume hood until all chemical has volatilized (usually 24
hours).
c. After chemical volatilization, obtain 20.0 g of "volatilized"
bentonite powder for test. P.ulverize dry "volatilized" bentonite
lumps as necessary using a mortar and pestle.
NOTE: In some cases, chemical may not completely volatilize even
after extended fume hood drying. In such cases, discard
unvolatilized bentonite (wet or pasty consistency) and proceed
if there,remains sufficient "volatilized" bentonite.
ALSO: It has been demonstrated that the rate of swell and the total
7-hour swell value of a typical (unaltered) bentonite in
baseline (water only) tests is a function of the initial "air
dry" water content. For comparisons between different ben-
tonites, therefore, it is necessary to precondition the
various bentonites to the same initial "air dry" water content
before testing. This is most readily accomplished by drying
the relatively "wet" bentonites down to the "air dry" water
content of the driest sample to be tested. Drying should be
accomplished using an evaporation dish in a low humidity
environment. (Do NOT use SLP No. G-01.) In any case, the
"air dry" water content of the different bentonite samples
should not vary by more than 1 percentage point. An initial
"standardized" air dry water content under 10.0% is recom-
mended.
155
-------
SLP No. EX-03
Revision 0-061986
Sheet 1 of 2
FILTER CAKE HYDRAULIC CONDUCTIVITY: API FILTER PRESS
EQUIPMENT:
standard API filter presses filter paper (Dresser-Magcobar
25 ml graduated cylinders 2-189 or equivalent)
dial or vernier calipers spatula
timer
PROCEDURE:
A. Prepare bentonite slurry per SLP S-02.
B. Allow slurry to hydrate undisturbed for 24 hours (or other, per
applicable experimental design) in moisture room.
C. (OPTIONAL) Re-mix slurry per SLP S-03 if necessary to achieve
uniform mix.
1. Assemble filter press cell in the following order: base cap, rubber
gasket, screen, filter paper, rubber gasket, and filtrate cylinder.
2. Fill the cylinder with slurry to witnin 1/4" of the top and place
in filter press frame.
3. Position top cap (connected to pressure source) with rubber gasket
in place on the filter press cel.l and secure with T-screw.
4. Place a clean, dry 25 ml graduated cylinder under the filtrate drain
on the support stand.
5. Close the in-line valve, open the cell pressure valve, close the cell
bleed valve, and adjust the regulator to 100 psi.
6. Open the in-line valve and start timer.
7. (OPTIONAL) Record the volume of filtrate collected at 7.5 minutes
elapsed time. (This amount doubled should be a close estimate of the
30-minute filtrate volume.)
8. Record the volume of filtrate collected at 30 minutes total elapsed
time.
9. Close in-line valve, open bleed valve to remove system pressure, and
close the filter press cell pressure valve.
10. Loosen the T-screw, remove the cell top cap, remove cell from frame
and decant slurry.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
156
-------
SLP No. EX-03
Revision 0-061986
Sheet 2 of 2
11. Fill the cell with the aqueous organic permeant to within 1/4" of the
top and place in filter press frame with 0-30 psi regulator.
12. Reposition top cap on the filter press cell and secure cell with
T-screws.
13. Place a clean, dry 25 ml graduated cylinder under the filtrate drain
on the support stand.
14. Open the cell pressure valve with the cell bleed valve open and adjust
regulator to the desired pressure.
15. Close cell bleed valve and commence timing.
16. Periodically measure and record effluent volume and elapsed time,
refilling cell with permeant if necessary.
17. After desired testing period, measure and record final effluent
volume and elapsed time.
18. Close the cell pressure valve, open bleed valve to remove system
pressure, and shut off pressure at (wall) regulator.
19. Loosen the T-screw, remove cell from the frame, and decant permeant
into approved container.
20. Disassemble filter cell and remove filter paper on which filter cake
has been formed.
21. Remove excess slurry from filter cake with a gentle stream of water,
then measure and record filter cake thickness (mm).
22. Calculate and record the permeability (k) of the filter cake using
Darcy's equation (Q = kiA).
157
-------
SLP NO. EX-04
Revision 0-031286
Sheet 1 of 3
SOIL-BENTONITE HYDRAULIC CONDUCTIVITY: API FILTER PRESS
EQUIPMENT:
Standard API Filter Press Machined Plexiglass Discs
Graduated Cylinders Consolidation frame & boxes
Bentonite Paste Spatulatas
Medium Sand Timer/Clock
Vernier Calipers
Scale (accurate to .lg)
PROCEDURE:
A. Prepare soil-Dentonite backfill per SLP S-04 A.
1. With vernier calipers, measure the diameter of the API filter press
cell and calculate its cross-sectional area.
2. Calculate the surcharge weight necessary to consolidate sample at
5 psi.*
3. Assemble the filter press in the following order: oase cap, rubber
gasket, screen, rubber gasket and filter press cell.
4. Measure the distance from the top of the filter press cell, to the
screen on the base cap
5. Remove the filter press cell and apply a tnin (approx. 1/32") film of
fully hydrated bentonite paste to the inner wall of the cell, extending
about 2-1/4" up the side of the cell.
6. Reassemble the filter press cell.
7. Carefully place a thin layer (1/16" +) of -medium sand on the screen,
spreading and "rodding" the sand to ensure good contact between the
sand and the cell wall.
8. With the plexiglass disc, tamp the sand to level, remove large void
spaces, and consolidate.
9. Measure the distance to the top of the sand ([>£) with the calipers.
10. Calculate the thickness of the sand layer (Hsan
-------
12. Weigh the cell + paste + sand to nearest O.lg.
13. Carefully spoon the soil-bentonite into the cell in thin lifts approxi-
mately 1/2" thick.
14. Spread and "rod" the soil-bentonite to achieve good backfill-cell contact
and minimize artificial void space (air bubbles).
15. Tamp the soil-bentonite with the plexiglass disc to level it, and further
minimize artificial void space.
16. Measure the distance to the top of the soil-bentonite (03) and calculate
the height of the soil-bentonite layer as:
HS-b - 02 - °3
17. Repeat steps 12-15 until the soil-bentonite is approximately 2" deep.
18. Weigh the cell + paste + sand + backfill, and compute the weight of the
soil-bentonite (re. step 12).
19. Place the soil-bentonite filled cell in the consolidation frame.
20. Fill the consolidation box with enough sand to allow for 5 psi* consoli-
dation, and place on the API cell using the plexiglass load disc, ball
bearing and plexiglass rod to support the consolidation box.
21. Record the initial scale reading on the side of the consolidation box
when the weight initially contacts the sample.
Note: Steps 20 and 21 generally require coordination of activities
between two assistants.
22. Consolidate the sample for 43 hours.*
23. After consolidation, record the final scale reading, calculate the total
consolidation and remove the consolidation box.
24. Remove the plexiglass disc from the cell.
25. Measure and record the post-consolidation height of the soil-bentonite
for use in computing hydraulic gradient.
26. Weigh the cell and sample.
27. Place the cell on the filter press frame, fill with the desired permeant,
attach the cell cap, open the air source and valves and begin permeating at 2
psi.*
28. Approximately 24 hours* later, increase the pressure to 3 psi.*
*Per applicable experimental design
159
-------
29. After 24 hours*, Increase the pressure to 4 psi*. recording the date and
time when permeating at 4 psi* began.
30. Periodically measure and record the effluent volume (Q) and elapsed time,
refilling the cell with permeant as necessary.
31. Calculate hydraulic conductivity (k) at each reading as: k=Q/iA; using
individual* or cumulative* filtrate and time values.
32. After the desired testing period*, or attainment of study state (equili-
brium) conditions*, or passage of the specified minimum pore volumes*,
calculate and record a final hydraulic conductivity value.
33. Close the cell pressure valve, open the bleed valve to remove the system
pressure, and close the in-line valve.
34. Remove the cell from the frame and decant any unused permeant into an
approved container.
35. Weigh the sample and cell and determine the final total weight of the
soil-bentonite sample.
36. Extract the soil-bentonite sample and take a representative sample for a
moisture content determination, per SLP G-01.
37. Dispose of the remaining soil-bentonite per applicable disposal procedure.
*Per applicable experimental design
REF: Center Hill in-house procedure
160
-------
SLP No. G-01
Revision 0-013085
Sheet 1 of I
WATER CONTENT
EQUIPMENT:
balance (accurate to O.Olg)
drying oven
dessicator
moisture tins
PROCEDURE:
1. Weigh moisture tin; record tare weight in grams (T).
2. Weigh wet soil plus tin; record in grams (W).
3. Oven dry overnight to a constant weight at a temperature of 105°C
± 3°-
4. Allow tin to cool in dessicator for 15 minutes. Weigh and record
in grams (D).
CALCULATIONS:
Percent Moisture = (W - D) x 100
(D - T)
SPECIAL NOTE:
For soils treated with flammable chemicals, sample should be air-
dried in fume hood 24-hours before oven drying (Step 3).
REFERENCE: API RP 13A, Tenth Edition, Sec. 4.6/Center 'Hill in-house.
CENTER IHILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
161
-------
SLP No. G-02
Revision 0-021385
Sheet 1 of 2
SPECIFIC GRAVITY OF SOLIDS, Gs
EQUIPMENT:
500 ml volumetric flask
vacuum pump and 1/4" tubing
soil drying oven
balance (sensitive to 0.01 g)
thermometer (accurate to 0.1°C)
evaporating dish
water bath (sink)
PROCEDURE:
A. Flask Calibration: all specific gravity tests
1. To a clean flask, add de-aired distilled water at room temperature
until the bottom of the meniscus is at the calibration mark on the
flask.
2. Carefully dry the outside of the bottle and the inside of the neck
above the water surface.
3. Measure water temperature to 0.1°C at several positions in flask.
If water temperature varies with elevation, agitate flask to promote
mixing and temperature equilibration. Record equilibrated tempera-
ture.
4. Weigh bottle plus water to 0.01 g; record.
5. Repeat Steps 1-4 at about 4°C below room temperature (water bath
cooling prior to Step 2) and about 4°C above room temperature (warm
water bath or gentle heating over flame). In each case, add or
remove water to the calibration mark before measuring temperature
and weight.
6. Plot calibration curve consisting of at least three weight/tempera-
ture measurements ; store calibration curve on file for future
reference.
B. Specific Gravity: Typical Soils
1. Determine water content of soil per SLP No. G-01.
2.
3.
Add quantity of moist or damp soil equivalent to oven-dried weight
of 50 grams (cohesive soil) or 150 grams (cohesionless soil) to
evaporating dish and sufficient de-aired, deionized water to form
a slurry.
Transfer'slurry to 500 ml flask. Using wash bottle, wash evaporating
dish to insure collection of all soil. Add de-aired, distilled
water until flask is about half full. Slowly agitate (mix) con-
tents for 3 minutes.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
162
-------
SLP No. G-02
Revision 0-021335
Sheet 2 of 2
4. Connect flask to vacuum line and evacuate flask for 4 hours (cohesive
soils) or 1 hour (cohesionless soils), gently agitating the flask
several times during this procedure to promote removal of entrapped
air.
5. After proper evacuation, fill flask with de-aired, deionized water
until the bottom of the meniscus is coincident with the calibration
line on the neck of the flask.
6. Thoroughly dry the outside of the flask and remove moisture on the
inside of the neck using a paper towel. Take care not to contact
meniscus with towel.
7. Weigh flask and contents to nearest O.Olg. Record on data sheet,
line 4.
8. Agitate suspension to assure uniform temperature and measure temper-
ature of suspension to nearest 0.1°C with thermometer at mid-depth
of flask.
9. Carefully transfer all flask contents to large evaporting dish. Use
wash bottle to ensure collection of all soil.
10. Oven-dry sample to a constant weight at 110°C +3C, allow soil to cool
at room temperature in dessicator, weigh the dry soil to 0.01 g.
Record as weight of dry soil, Ws, on data sheet, line 10.
11. Obtain weight of "flask plus water" at the temperature measured in
Step 8. Record on data sheet, line 6.
12. Obtain water relative density and temperature correction factor
(« ) from Corps of Engineers Laboratory Soils Testing Manual, Table
IV-1, page IV-8.
13. Compute specific gravity, Gs, per data sheet as:
Gs =
ws
Ww
C. Specific Gravity: Bentonites
The following modifications of standard procedure apply to bentonites
only. Steps 1-13 for typical soils apply, except as follows:
2. Use 15.0 grams equivalent dry weight; add bentonite slowly to promote
thorough mixing.
4. Evacuate flask overnight to achieve adequate removal of entrapped
air and allow for bentonite hydration.
163
-------
SLP No. G-03
Revision 1-032585
Sheet 1 of 7
GRAIN SIZE ANALYSIS
(Mechanical , Hydrometer, Combined)
EQUIPMENT:
U.S. Standard Sieves: Nos. 4, 10, 20, 40, 100, 200, pan
mortar and pestle
balance (sensitive to 0.1 g)
sedimentation cylinders (1000 ml, graduated)
hydrometer: No. 151H or equivalent for general soils testing
No. 14 331 58 for bentonite slurry
soil dispersion mixer and baffeled mixing cup
soil dispersant: NaP03 (Calgon)
thermometer (sensitive to 0.1°C)
timer and stopwatch
wash bottle
GENERAL OUTLINE:
This SLP is intended to apply to a mechanical analysis (gravel and
sand, e.g. coarser than No. 200 sieve), a hydrometer analysis (silt and
clay, e.g. finer than No. 200 sieve), or a combined total analysis (gravel-
sand-silt-clay mixtures) . With minor modification as noted herein, the
procedure may also be applied to bentonite slurries of low concentration
(generally 1% bentonite).
For the sake of procedural consistency and clarity, this SLP has been
developed in conjunction with the applicable data forms. Test results are
reported on a separate grain size distribution plot (standard C.H. form).
The test procedure with corresponding data forms is separated into
four parts:
Part 1. Mechanical Analysis Procedures
Part 2. Hydrometer Analysis Procedures
Part 3. Worksheets
Part 4. Appendix - Calibrations arid Constants
CENTER IHILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
164
-------
SLP No. G-03
Revision 1-032585
Sheet 2 of 7
PART 1
A. Mechanical Analysis:
1. Obtain approximately 500 g of representative soil sample; air dry
sample and pulverize silt and clay lumps using mortar and rubber-
tipped pestle.
2. Carefully wash sample on clean No. 200 sieve; place +200 fraction
in oven to dry overnight.
3. Determine weight of washed, oven-dried +200 sample; enter as "Wj"
on Sheet 7.
4. Assemble selected sieve stack, plus pan; pour sample on top sieve;
cover and secure sieve stack in shaker; shake for 10 minutes at
medium speed.
5. Place large aluminum pan on triple beam balance; determine tare
weight, grams; enter tare weight in Sections 1 and 2, sheet 7.
6. Starting with top sieve (largest screen opening), carefully empty
contents into aluminum pan, recording the cumulative weight total
for each successive sieve. Data should be recorded in "cumulative
weight retained" column of Sheet 7, Sections 1 and 2, as total
weight including pan, followed by the net soil weight (e.g.
120.0/20.0 where pan tare = 100.Og).
7. For each sieve, compute cumulative % retained as the cumulative net
weight retained, divided by 1^3.
8. For each sieve, subtract cumulative % retained from 1.00; enter
result as cumulative % finer.
9. Plot cumulative % finer results on standard C.H. grain size dis-
tribution form, noting all pertinent sample classification and
identification information.
PART 2
B. Hydrometer Analysis: Typical Soils and Liquids
1. Determine composite correction for temperature, meniscus, defloc-
culant and chemical per Appendix Sheets A-l and A-2 (if not al-
ready performed and on file for the test conditions selected).
2. Calibrate hydrometer to determine depth to center of bulb as per
Appendix Sheet A-4 (if not already performed and on file for the
hydrometer to be used).
165
-------
SLP No. G-03
Revision 1-032585
Sheet 3 of 7
3. Prepare fresh 4% deflocculant solution by mixing 40.0 grams de-
flocculant with 960 ml of deionized water.
4. Oven-dry soil sample, collect -10 sieve fraction, and weigh out
exactly 50.0 g. Thoroughly mix with 125 ml of 4% deflocculant
solution, cover and allow to stand 24 hours.
5. Transfer mixture into soil dispersion cup and add deionized water
until cup is about two-thirds full; mix for three minutes.
6. Decant mixture into 1000 ml sedimentation cylinder, taking care
to wash all sediment from cup with wash bottle; add additional
deionized water up to 1000 ml mark.
7. Prepare control cylinder consisting of 125 ml deflocculant solu-
tion and 875 ml deionized water in 1000 ml graduate.
8. Place hydrometer and thermometer in control cylinder and allow
control cylinder and sedimentation cylinder temperatures to
stabilize and equilibrate.
9. Place palm of hand tightly over top of sedimentation cylinder and
repeatedly agitate by turning cylinder upside down until all sedi-
ment is suspended (typically one minute).
10. Place sedimentation cylinder on counter at start of test (time
"zero"), immediately insert hydrometer and take readings at 30
sec, and 1, 2, 3, 4 minutes total elapsed time. Record on hydro-
meter worksheet, Sheet .
11. Repeat Steps 8 and 9 until agreement is obtained for readings
during the first four minutes; take temperature of suspended
solution.
12. Repeat Step 8. Take only 4-minute readings. Continue test, tak-
ing readings at 8, 15, 30, .and 60 minutes and 2, 4, 8, 16, 32,
and 64 hours, total elapsed time, recording temperature with each
hydrometer reading.
13. For each reading, enter the following on the worksheet:
Cc - composite correction for test conditions (to be estab-
lished prior to test per Appendix sheets A-l and A-2)
K - composite viscosity, temperature and specific gravity
correction, from Table I (Appendix Sheet A-3)
L - depth value to center of hydrometer bulb, cm, from
hydrometer calibration (Appendix Sheet A-4).
14. Compute particle diameter in millimeters (D) and percent sus-
pended (Wj%), record on worksheet and Sheet 7, Section 3.
166
-------
SLP No. G-03
Revision 1-032585
Sheet 4 of 7
15. After all readings have been taken, agitate sediment as in Step
8; pour over No. 200 sieve. Wash bottle repeatedly as necessary
to insure collection of all sediment from cylinder.
16. Wash material retained on NO. 200 sieve (if any) into aluminum
weigh pan, oven dry for 24 hours at 105°C +3C.
17. Sieve +200 fraction in No. 4, 10, 20, 40, 100, 200 sieves per
Step A-4 (mechanical analysis), compute cumulative % retained and
cumulative % finer per mechanical analysis procedure, using 50.0
grams as total original soil weight.
18. Plot cumulative % finer results on standard C.H. grain size dis-
tribution form, noting all pertinent sample classification and
identification information.
C. Hydrometer Analysis: Bentonite and/or Chemicals
The following modifications of the "typical" hydrometer procedure
are required for testing bentonites or other testing involving liquids
other than deionized water.
a. Use 10.0 grams bentonite in Step B-4.
b. Prior to Step B-4, transfer mixture into graduate cylinder and add
additional deionized water up to a total volume of proportion to
1000ml of final bentonite/chemical mix (e.g. if test calls for
75% acetone slurry, add water to 250 ml mark during this step).
Then transfer mixture into soil dispersion cup, mix for three
minutes and decant into sedimentation cylinder. Transfer and mix
in several steps if necessary due to large volume required when
low chemical concentrations are being made.
c. When entire bentonite-deflocculant-deionized water mixture has
been decanted into sedimentation cylinder, add selected chemical
up to 1000 ml mark (instead of additional water per Step B-6)
using appropriate safety procedures. Cap cylinder with parafilm
to minimize evaporative losses.
d. Conduct test per remaining "typical" procedures B-7 thru B-18.
167
-------
PART 3.
COMBINED ANALYSIS WORKSHEET
SLP No. G-03
Revision 1-032585
Sheet 5 of 7
PROJECT NO
TEST TYPE
SAMPLE NO.
DATE
TEST NO.
ANALYST
I.
Ill,
SAMPLE PREPARATION Gross Sample: Grate soil sample (if clayey) or break
up lumps and clods with hands. Air-dry gross sample, pulverize with mortar
and rubber pestle. Separate on #10 sieve, work through by hand. Retain
both fractions.
II. +10 FRACTION
Pan No.
Wt. oven dry soil & pan
tare
Wt. oven dry soil
-10 FRACTION
Determine gross air dry.weight.
Back-figure gross "oven" dry
weight using moisture content
data determined in Step IVb.
Pan No. (sieve pan)
Wt. air dry soil & pan*
tare
Wt. air dry soil
Wt. "oven" dry soil
Obtain representative sample from
remainder of r!0 fraction. Use
Sartorius balance to determine
moisture content per SLP No.
G-01. Calculate "oven" dry
weights in III & IVa.
MOISTURE CONTENT
Where:
"oven"
dry = wt. air dry soil
1.0 + moi St. cont.
g
g|
Tin
wt.
wt.
wt.
wt.
Moi
No.
air
oven
dry
dry
soil
soil
& tin*
I tin
tare
water
dry
sture
soil
content:
|0.
g
g
g
g
g
1
W2
IV.
a.
-10 FRACTION: HYDROMETER ANALYSIS
Split out representative portion
from gross -10 fraction after
Step III. (Use -50 g for clays,
-100 g for sandy soil s.)
Pan No.
Wt. air dry soil & pan*
tare
Wt. air dry soil
Wt. "oven" dry soil used
in hydrometer test
g
9 W0
g
g
*Minimize time between these determinations
168
-------
HYDROMETER ANALYSIS WORKSHEET
SLP No. G-03
Revision 1-032585
Sheet 6 of 7
Project No_._
Analyst
Sample No.
Test No.
Sample Description^
Remarks:
Test Type
Date
Hyd. No.
INITIAL DATA
deflocculant
fluid medium
specific gravity, Gs
Wt. soil used, W0:
(from sheet 5)
ELAPSED
TIME TEMP
DATE TIME (MIN) READ °C Cr R & Cr
0.5
1
2
3
4
8
15
30
60
120
SEE NOTE
L
K
D
WT
RMK
NOTES: D = K
WT% =
100 Gs
Wn
Gs-1 (R + C)
L from Appendix Sheet A-3, K from Appendix Sheet A-4
Test Type: X -10 fraction of total sample (combined)
Y -10 size = total sample (combined)
I -200 size = total sample (hydrometer only)
169
-------
GENERAL GRAIN SIZE ANALYSIS WORKSHEET
Project No.
Test Type
Sample No.
Date
SLP No. G-03
Revision 1-032585
Sheet 7 of 7
Test No._
Analyst "
Wy: Gross dry wt. total sample = \
SR: Sample ratio = W?/WT =
(combined analysis only; SR =
^i + W? = WT = + =
/ s
1.0 for M or H)
g
M = Mechanical, H = Hydrometer, C = Combined Analysis
.M or C
1 . +10 fraction
of gross dry
sample.
pan no.
tare g
M or H or C
-10 fract. of
2. mech. anal . or
+200 fraction
after hydro, test.
pan no.
tare g
Wn: 9
H or C
3.
Hydrometer Test
Results
SIEVE
NO.
.
-
-
-
-
-
-
4
10
20
40
100
200
DIAME
IN
3
2
1-1/2
1
3/4
1/2
3/8
.187
.079
.0331
.0165
.006
.003
ELAPSED
TIME MIN.
0.50
1 .0
2.0
5.0
15.0
30.0
60.0
120.0
:TER
MM
76.1
50.8
38.1
25.4
19.0
12.7
9.5
4.76
2.0
.841
.420
.149
.075
/
/
i
DIAMETER
MM
A
CUM. WT.
RETAINED
A/WT
CUM. %
RETAINED
1 Ml . . , ._ o llJ—
(H or C) —
B
-*B/W0
D
ws
C
CUM. %
FINER
CUM. %
FINER
X
C x SR
X
D x SR
-SUMMARY PLOT-
170
-------
SLP No. G-03
Revision No.
Sheet A-1
1-032535
PART 4. APPENDIX
Composite Correction
GRAIN SIZE ANALYSIS
General: A separate calibration is required for each hydrometer/test liquid
combination (water or chemical:water soil slurries). Calibration
is performed to determine composite correction (Cc) for combined
effects of meniscus, deflocculant, chemical and temperature. All
Cc values should be rounded to the nearest tenth and subtracted
from the raw hydrometer readings on the Hydrometer Analysis Work-
sheet.
Typical Calibration Set-up:
1.000 '""— >
1000 ml *
SPECIFIC GRAVITY
OF FLUID MEASURED
AT CENTER OF BULB
: . - . . _ ..
DISTILLED H.O _^*
•" 'WITH 125cc2
-- . OF 4% DEFLOCCULANT
SOLUTION
t
i
i
i
/
'1
/
x
• «-
s
}
/
'/
/
-HYDROMETER 151 H
OF MENISCUS = 1.037
-•• -"-
171
-------
SLP No. G-03
Revision 1-032585
Sheet A-2
Typical Calibration Curve:
264
o
— 5 25
o
. r-j
BC
U— « 24 44
u
a
o
as
o
>>
1
23-.
22-n
21 T
20 —
19-.
DISTILLED WATER
DATA POINT FOR
CALIBRATION
SET-UP ON
SHEET A-l
HYDROMETER 151 H
2.6 2.8 3:0 3.2 3.4 3J6
COMPOSITE CORRECTION
3.8 U.Q
172
-------
SLP No. G-03
Revision 1-032585
Sheet A-3
TABLE 1. VALUES OF K FOR USE IN EFFECTIVE DIAMETER COMPUTATION, SHEET 6.
UNIT WEIIillT OF SOU. SOLIDS (if/cm I
Temp.
CO
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
2.50
0.0151
0.0149
0.0148
0.0145
0.0143
0.0141
0.0140
0.0138
0.0137
0.0135
0.0133
0.0132
0.0130
0.0129
0.0128
2.5.1
0.0148
0.0146
0.0144
0.0143
0.0141
0.0139
0.0137
0.0136
0.0134
0.0133
0.0131
0.0130
0.0128
0.0127
0.0126
2.60
0.0146
0.0144
0.0142
0.0140
0.0139
0.0137
0.0135
0.0134
0.0132
0.0131
0.0129
0.0128
0.0126
0.0125
0.0124
•-'.65
0.0144
0.0142
0.0140
0.0138
0.0137
0.0135
0.0133
0.0132
0.0130
0.0129
0.0127
0.0126
0.0124
0.0123
0.0122
2.70
0.0141
0.0140
0.01.38
0.0136
0.0134
0.0133
0.0131
0.0130
0.0128
0.0127
0.0125
0.0124
0.0123
0.0121
0.0120
2.T5
0.0139
0.01.18
0.0136
0.0134
0.0133
0.0131
0.0129
0.0128
0.0126
0.0125
0.0124
0.0122
0.0121
0.0120
0.0118
2.SO
0.0137
0.0 US
0.0134
0.0132
0.0131
0.0129
0.0128
0.0126
0.0125
0.0123
0.0122
0.0120
0.0119
0.0118
0.0117
2.S5
0.0136
0.0134
0.0132
0.0131
0.0129
0.0127
0.0126
0.0124
0.0123
0.0122
0.0120
0.0119
0.0117
0.0116
0.0 1 1 5
173
-------
SLP No. G-03
Revision 1-032585
Sheet A-4
Hydrometer Calibration
General: Different hydrometers are required for the hydrometer analysis,
depending upon the range of specific gravity likely to be en-
countered for a given test. The calibration below is for hydro-
meter 151 H (standard soils testing) and Fischer hydrometer 14
331 56 (1% bentonite slurries) and is used to determine the
effective length (L) to the center of the hydrometer bulb for
calculation of effective diameter (0) per the Hydrometer Analysis
Worksheet (Sheet 6). If a different hydrometer is used, a separ-
ate effective depth calibration will be required per Corps of
Engineers Laboratory Soils Testing Manual, Appendix V, Section
3-C.
-18 -
16 -
14 -
12
-10 -
—8 —
• 6 -
4 -
2 -
- 0 -
FISCHER HYDROMETER
14 331 5B
(HYDROMETER READING -1.000) x 1000
i
10
I
15
i
20
i
25
I
30
35
i
40
I
45
50
174
-------
SLP No. G-04
Revision 0-013085
Sheet 1 of 1
FREE SWELL
EQUIPMENT:
TOO ml graduated cylinder
parafilm
balance (accurate to O.Olg)
weighing boats
thermometer (accurate to 0.5°C)
PROCEDURE:
1. Determine water content of "dry" bentonite per SLP No. G-01.
2. Fill graduated cylinder with 100 ml deionized water.
3. Weigh out 2.0 grams bentonite (dry weight basis) and sprinkle onto
water surface in cylinder until surface is covered.
4. Add the entire 2.0 grams over a period of 30 minutes, letting each
successive layer wet and settle.
5. Cover graduated cylinder with parafilm and allow to hydrate.
6. Measure volume of swelled bentonite (and temperature of water above
bentonite) at 2 hours and 24 hours total elapsed time. Record as
swell volume (ml) at corresponding time and temperature.
7. Discard bentonite and liquid per applicable safety procedures.
REFERENCE: U.S. Pharmacopoeia, 20th Revision/Center Hill in-house.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
175
-------
SLP NO. G-05
Revision No. 0-022885
Sheet 1 of 2
UNIT WEIGHT
(Undisturbed Tube Sample)
EQUIPMENT:
steel sample tube balance sensitive to O.Olg
steel sample rod moisture content tin
steel extrusion rod drying oven
aluminum sample tray soil knife
dial calipers flat-bottom sloop
PROCEDURE:
1. Select steel sample tube, determine tare weight to 0.01 g; record.
Measure tube I.D. and length to 0.001", compute tube volume to 0.001
cm3; record.
2. Prepare flat surface area on soil to be sampled using soil knife
or flat bottom scoop (surface can be any incliniation per applicable
experimental design, e.g. vertical vs. horizontal tube sample).
3. Affix tube to sample rod using rod pin, place tube perpendicular to
fresh sample face and press tube into soil using steel sample rod
and slow, continuous motion. Depth of penetration should be less
than or equal to the tube length (e.g. do not over-penetrate and
thereby disturb sample).
4. Allow tube to remain undisturbed for five minutes, then rotate
sample rod and tube to shear soil at leading edge of tube, and slow
extract sample rod and full tube. If soil is too soft to be recovered
intact, it will be necessary to manually excavate around the tube and
use a wide blade knife or flat bottom scoop to hold the sample in
place during tube extraction.
5. Trim soil flush with sample tube, clean all soil from outside of tube,
weight tube with soil to 0.01 g; record.
6. Using extrusion rod, slowly extrude sample toward rear (pin end) of
tube. Receive sample in sample tray as it is extruded.
7. Carefully examine surface of extruded sample for disturbance due
to gravel or coarse sand materials or general voids which would
invalidate calculated sample volume. If sample is badly disturbed,
discard and re-sample.
8. Slice sample lengthwise and inspect per Step 7.
9. Select representative portion of sample for water content determina-
tion per SLP No. G-01.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
176
-------
SLP No. G-05
Revision Mo. 0-022885
Sheet 2 of 2
10. Calculate wet unit weight as: wet sample weight, g
sample volume,
Calculate dry unit weight as: wet unit weight
water content
REFERENCE: Center Hill in-house procedure
-------
SLP No. G-06
Revision 0.090385
Sheet 1 of 1
SLUMP TEST
EQUIPMENT:
slump cone
tamping rod
SAMPLE PREPARATION:
Prepare slump sample per slurry and soil-bentonite mixing procedures,
if applicable.
PROCEDURE:
1. Dampen slump cone and the bottom of the soil pan.
2. Hold cone in place by standing on the two foot pegs.
3. Fill cone 1/3 of its volume («=3 scoops).
4. Rod each layer with 25 strokes of the tamping rod, making sure that
each layer is rodded evenly. Rod the bottom layer throughout its
depth. Rod the second layer and the top layer each throughout their
depths, so that the strokes just penetrate into the underlying layer.
5. In filling and rodding the top layer, heap the S/B above the mold
before rodding is started. If the rodding operation results in sub-
sidence of the S/B below the top edge of the mold, add additional
S/B to keep an excess of S/B above the top of the mold at all times.
After tne top layer has been rodded, strike off the surface of the
S/B with the tamping rod.
6. Remove the mold immediately from the concrete by raising it care-
fully in a vertical direction. Raise the mold a distance of 12 inches
in 5+2 seconds by a steady upward lift with no lateral or torsional
motion. Complete test without interruption.
7. Immediately measure the slump by determining the vertical difference
between the top of the mold and the displaced original center of the
top surface of the specimen.
8. Record slump in terms of inches.
CENTER IMLL
United States Environmental Protection Agency
Solid and Hazardous Wast* Research Facility
178
-------
SLP No. E-01
Revision 1-032585
Sheet 1 of 1
pH MEASUREMENT
(Using Auto Control)
EQUIPMENT:
Corning pH meter 140
pH probe
test liquid
100 ml beaker
MEASUREMENT PROCEDURE:
1. Rinse electrodes with test liquid or deionized water.
2. Immerse electrodes in test liquid in a 100 ml beaker and press "AUTO"
and "=".
3. Record the pH reading after stable endpoint is reached. (The "pH" and
"A" LEDs will flash until a stable endpoint is reached, then the dis-
play will freeze and the LEDs will stop flashing and stay on.)
4. Recant the slurry for future use.
CALIBRATION (1-POINT):
1. Press mode to select pH mode ("pH" LED will flash).
2. Rinse electrodes with pH 7.00 buffer or deionized water.
3. Immerse electrodes in pH 7.00 buffer; press "CAL" until the 7 LED
flashes and then press "AUTO". The auto "A" and 7 LEDs will flash un-
til a stable endpoint is reached, then they will stop flashing and
stay on. The display will read 7.00.)
CALIBRATION (2-POINT):
1. Perform steps 1 thru 3 in 1-point calibration procedure.
2. Select second buffer, pH 4.00 or pH 10.01.
3. Rinse electrodes with appropriate second buffer or deionized water.
4. Immerse electrodes in the second buffer and press "CAL". The "A" and
4/10 LEDs will flash. When a stable endpoint is reached, the LFDs
will stop flashing and stay on. Display will read 4.00 or 10.01,
depending on the second buffer used.
CENTER IjlUL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
179
-------
SLP No. E-02
Revision 1-032585
Sheet 1 of 2
SLURRY CONDUCTANCE
EQUIPMENT;
conductivity meter temperature probe
conductivity cell 250 ml test slurry
250 ml beaker
PROCEDURE:
1. Pour test slurry into beaker to 250 ml mark.
2. Set conductivity meter function switch to temperature compensated
conductance position.
3. Immerse conductivity cell into test slurry. The electrodes must be
submerged and the electrode chamber free of trapped air. Tap cell
to remove any bubble and dip it into the slurry several times to
assure complete wetting.
4. Connect conductivity cell leads to meter terminals.
5. Rotate temperature compensation control to desired percentage/°C
setting, plug in temperature probe, and place into test slurry.
6. Rotate range switch to lowest range position which gives a within-
range meter reading. (An over-range value is indicated by a "1"
followed by blanks. An under-range is indicated by a small letter
"u". Readings may be in error when operating in the under-range
condition. On the 0.1-2 micromho range, allow extra time to stabi-
lize.)
7. Record stabilized conductance value (c).
8. Set the function switch to temperature position, read and record
displayed temperature in °C.
9. Record conductance as micromho per % bentonite concentration.
CONDUCTIVITY = C x K
1 1
RESISTIVITY = T~ 7 K = r~x~T
K = cell constant = 1.0/cm or 100/m
CENTER .HILL
United States Environmental Protection Agency
Solid and Hazardous Wast* Research Facility
180
-------
SLP No. E-02
Revision 1-032585
Sheet 2 of 2
NOTES:
1. After use, clean conductivity cell with foaming acid tile cleaner
or with a 1:1 solution of isopropyl alcohol 10 N HC1. Rinse cell
with distilled water, allow to air dry and store in deionized water.
2. When testing water or chemical-based bentonite .slurries, temperature
compensated conductance reference values should first be obtained on
the corresponding water or chemical solution having no bentonite.
Bentonite slurry specific conductance is recorded as the difference
in conductance between the slurry and the reference standard.
181
-------
SLP No. E-03
Revision 1-012186
Sheet 1 of 3
SPECIFIC CONDUCTANCE
EQUIPMENT:
conductivity meter temperature probe
conductivity cell 250 ml test sample
250 ml beaker
PROCEDURE:
1. Pour test sanple into beaker to 250 ml mark.
2. Set conductivity meter function switch to temperature compensated
conductance position.
3. Rinse conductivity cell with test sample. Immerse conductivity cell
into test sample. The electrodes must he submerged and the electrode
chamber free of trapped air. Make sure the slot in the top of the
probe is submerged and sealed with sample. Tap cell to remove any
bubble and dip it into the sample several times to assure complete
wetting. Probe must be held upright to obtain accurate reading.
4. Make sure conductivity cell leads are connected to meter terminals.
5. Rotate temperature compensation control to desired percentage/°C
setting (this will be 1.95 unless stated stated otherwise) and record
on data sheet. Plug in temperature probe and place into test sample.
6. Rotate range switch to lowest range position which gives a within-
range meter reading. (An over-range value is indicated by a "1"
followed by blanks. An under-range is indicated by a small letter
"u". Readings may be in error when operating in the under-range
condition. On the 0.1 - 2 micromho range, allow extra time to
stabilize.)
7. Record stabilized conductance value (c) on specific conductance data
sheet.
8. Set the function switch to temperature position, read and record
displayed temperature in °C.
9. Set the function switch to read and record stabilized uncompensated
conductance value.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous-Wast* Research Facility
1*2
-------
SLP No. E-03
Revision 1-012136
Sheet 2 of 3
10. Record the range switch setting hy entering the number corresponding
to the appropriate range:
Range Number
11. Record conductance as micromho/cm.
CONDUCTIVITY = C x K
1
RESISTIVITY = C ' K = C x K
K = cell constant = 1.0/cm or 100/m
NOTES:
1. After use, clean conductivity cell with foaming acid tile cleaner
or with a 1:1 solution of isopropyl alcohol 10 N HC1. Rinse cell
with distilled water, allow to air dry and store in deionized water.
?. When testing water or chemical-based bentonite slurries, temperature
compensated conductance reference values should first be obtained on
the corresponding water or chemical solution having no bentonite.
Bentonite slurry specific conductance is recorded as the difference
in conductance between the slurry and the reference standard.
183
-------
SPECIFIC CONDUCTANCE DATA SHEET
PROJECT:_
DATE:
TECHNICIAN:
SLP No. E-03
Revision 1-012136
Sheet 3 of 3
SAMPLE
in
READING
WITH
CflMP
READING
W/0
COMP
TEMP
°C
RANGE NO.
1) 2 u« 4) 2 tno
2) 20 vO 5) 20 mo
3) 200 u« 6) 200 mO
1
1
1
1
TEMP
COEF
'
CONDUCTANCE
wO
Cell Constant = 1.0/cn
If range 4, 5, or 6 is used, multiply reading with comp by 1000 to obtain
conductance in ut5 /en
Conductance (uU /en) = Reading with conp (gU ) x Cell constant (cm)
Conductance (yU /en) = Reading with conp (ntJ) x Cell constant (cm) x 1DOO
184
-------
SLP No. 1-01
Revision 0-013085
Sheet 1 of 2
SLURRY DENSITY
(Mud Weight)
EQUIPMENT:
standard API mud balance
thermometer (accurate to 0.5°C)
PROCEDURE:
A. Prepare bentonite slurry per SLP No. S-01 or S-02 per applicable
experimental design.
B. Allow slurry to hydrate undisturbed for 24 hours (or other, per
applicable experimental design) in moisture room.
C. (OPTIONAL) Re-mix slurry per SLP No. S-03.
1. Set instrument base on flat, level surface.
2. Measure and record slurry temperature in °C.
3. Fill clean, dry cup with slurry to be tested. Insert cup lid and
rotate until firmly seated, expelling excess slurry and trapped air
through orifice in lid.
4. Carefully wash or wipe the outside of the cup clean and dry.
5. Place balance arm on base with knife edge on fulcrum.
6. Move rider until graduated balance arm is level, as indicated by the
level vial on the beam.
7. Read density in lb/ft3 along the inside edge of the rider (closest to
mud cup).
8. Report slurry density in pounds per cubic foot (Ib/ft3).
9. Replicate test as required by experimental design.
10. Return slurry to storage for future test (24-hour, 7-day) if called
for by experimental design. Otherwise, discard slurry per applicable
safety procedures.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
185
-------
SLP No. 1-01
Revision 0-013085
Sheet 2 of 2
CALIBRATION:
a. Calibration is carried out using the above procedure, substituting
water for slurry. The level bubble should be centered when the rider
is set at the water calibrator mark on graduated beam.
b. If the beam is not balanced, adjust by adding or removing sand from
the wel1 at the end of the beam.
REFERENCE: API RP 13B, Tenth Edition, Sec. 1.1-5.
186
-------
SLP No. 1-02
Revision 0-013085
Sheet 1 of 1
MARSH FUNNEL VISCOSITY
EQUIPMENT:.
standard API Marsh funnel viscometer
graduated measuring cup (1 quart)
stopwatch (accurate to 0.1 sec)
thermometer (accurate to 0.5°C)
PROCEDURE:
1. Prepare bentonite slurry per SLP No. S-01 or S-02 per applicable
experimental design.
2. Allow slurry to hydrate undisturbed for 24 hours (or other, per
applicable experimental design) in moisture room.
3. Hold clean, dry funnel upright with finger covering the orifice.
Pour slurry into funnel until slurry level reaches the bottom of the
screen.
4. Remove finger from orifice and simultaneously start stopwatch. Mea-
sure time required for slurry to reach the 1-quart level in the cup.
Record time to nearest 0.1 second.
5. Measure and record slurry temperature in °C.
6. Report viscosity in Marsh seconds at measured temperature.
7. Replicate test as required by experimental design.
8. Return slurry to storage for future test (24-hour, 7-day) if called
for by experimental design. Otherwise, discard slurry per applicable
safety procedures.
REFERENCE: API RP 13 B, Tenth Edition, Sec. 2.1-2.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
187
-------
SLP No. 1-03
Revision 0-013085
Sheet 1 of 2
ROTATIONAL VISCOMETER
EQUIPMENT:
N.L. Baroid rheometer model 286
rotational viscometer
stopwatch (accurate to 0.1 sec)
thermometer (accurate to 0.5°C)
PROCEDURE:
1. Prepare bentonite slurry per SLP No. S-01 or S-02 per applicable
experimental design.
2. Assemble rheometer by installing shield, bob and rotor.
3. .Record slurry temperature, °C.
4. Pour slurry into the cup up to the scribed line on the inside surface.
Position cup on the viscometer using pins on cup bottom and holes in
rheometer base plate. Immerse sleeve into slurry to the scribed line.
5. Record zero (at rest) indicator dial reading, if any.
6. Stir slurry at 600 rpm and record stabilized indicator dial value.
7. Shift rheometer to 300 rpm and record stabilized indicator dial value.
8. Stir slurry for 10 seconds at high speed (in excess of 600 rpm).
9. Turn rpm selector knob to "off". Allow slurry to stand undisturbed
for 10 seconds after the rotor has stopped.
10. Switch rpm selector knob to GEL position (3 rpm). Observe the maxi-
mum (peak) indicator dial value after starting rotation. Record
peak value as 10-second gel strength in lb/100 ft2.
11. Repeat steps 7, 8, and 9, allowing slurry to sit undisturbed for 10
minutes (step 8). Record peak value as 10-minute gel strength in
lb/100 ft2.
12. Return slurry to storage for future test (24-hour, 7-day) if called
for by experimental design. Otherwise, discard slurry per applicable
safety procedures.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Waste Research Facility
188
-------
SLP No. 1-03
Revision 0-013085
Sheet 2 of 2
CALCULATIONS:
Apparent Viscosity (cP) = 600 rpm reading
2
Plastic Viscosity (cP) = (600 rpm reading) - (300 rpm reading)
Yield Point (lb/100ft2) = (300 rpm reading) - plastic viscosity
CALIBRATION:
Per prescribed OA/QC frequency, determine 300 rpm and 600 rpm reading
for deionized water at 21°C. Compute % difference for each relative to
calibration standard of (300) and (600). Record in calibration
log book. If result exceeds , notify principal investigator.
REFERENCE: API RP 138, Tenth Edition, Sec. 2.4-6
189
-------
SLP No. 1-04
Revision 1-032585
Sheet 1 of 2
API FILTRATE LOSS
EQUIPMENT:
standard API filter press stopwatch (accurate to 0.1 sec)
25 ml graduated cylinders filter paper, Dresser -
dial calipers Magcobar 2-189 or equivalent
250 ml beaker spatula
PROCEDURE:
A. Prepare bentonite slurry per SLP No. S-01 or S-02 per applicable
experimental design.
B. Allow slurry to hydrate undisturbed for 24 hours (or other, per
applicable experimental design) in moisture room.
C. (OPTIONAL) Re-mix slurry per SLP No. S-03 if necessary to
achieve uniform mix.
1. Assemble filter press cell in the following order: base cap, rubber
gasket, screen, filter paper, rubber gasket and filtrate cylinder.
2. Fill the cylinder with slurry to within 1/4" of the top and place in
filter press frame.
3. Position top cap (connected to the pressure source) on the filter
press cell, making sure the rubber gasket is in place. Secure cell
with T-screw.
4. Place a clean, dry 25 ml graduated cylinder under the filtrate drain
on the support stand.
5. Open the cell pressure valve, close the cell bleed valve and the in-
line valve, and adjust the regulator to 100 psi.
6. Open the in-line valve and start stopwatch.
7. Record the volume of filtrate collected (ml) at 7.5 minutes total
elapsed time. (This amount doubled should constitute an estimate of
the total 30-minute filtrate volume.)
8. Record the volume of filtrate collected (ml) at 30.0 minutes total
elapsed time.
9. Shut off pressure at the regulator, open bleed valve to remove system
pressure, and close the filter press cell valve.
CENTER HILL
United States Environmental Protection Agency
Solid and Hazardous Watt* Research Facility
190
-------
SLP No. 1-04
Revision 1-032585
Sheet 2 of 2
10. Loosen the T-screw, remove the cell top cap, remove cell from the
frame and decant slurry into beaker.
11. Disassemble filter cell and carefully remove filter paper on which
the filter cake has been formed.
12. Carefully remove excess slurry from filter cake using edge of spatula.
13. Measure and record thickness (mm) at four points on the filter
cake using dial calipers.
14. Average four measurements, subtract thickness of filter paper, and
record.
15. Determine moisture content of filter cake per SLP No. G-01 .
16. Discard filtrate and unfiltered slurry per applicable safety proce-
dure.
REFERENCE: API RP 13B, Tenth Edition, Sec. 3.1-3.
191
------- |