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

-------