WATER POLLUTION CONTROL RESEARCH SERIES • 14010 EFK 06/72
USE OF LATEX AS  A SOIL SEALANT
TO CONTROL ACID  MINE  DRAINAGE
    U.S. ENVIRONMENTAL PROTECTION AGENCY

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters.  They provide a central source of
information on the research, development, and demonstration
activities in the water research program of the Environmental
Protection Agency, through inhouse research and grants and
contracts with Federal, State, and local agencies, research
institutions, and industrial organizations.

Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, DC  20460.

-------
                USE OF LATEX AS A SOIL  SEALANT

                TO CONTROL ACID MINE DRAINAGE
                        UNIROYAL, INC.
                        Research Center
                    Wayne,  New Jersey 07^70
                            for the
                 COMMONWEALTH OF PENNSYLVANIA
           DEPARTMENT OF ENVIRONMENTAL RESOURCES

                               and

               OFFICE OF RESEARCH & MONITORING
               ENVIRONMENTAL PROTECTION AGENCY
                       GRANT 11+010 EFK
                          June 1972
Forsalo by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 • Price $1.00

-------
                EPA Review Notice
This report has been reviewed "by the Office of
Research and Monitoring, EPA, and approved for
publication.  Approval does not signify that the
contents necessarily reflect the views and poli-
cies of the Environmental Protection Agency, nor
does mention of trade names or commercial products
constitute endorsement or recommendation for use.
                         11

-------
                             ABSTRACT
A study was made to test the feasibility of using latex as a soil
sealant to prevent water seepage into subterranean abandoned mines.

A variety of latexes were screened in laboratory tests using recon-
structed soil columns.  The most promising latex (an SBR. rubber latex)
was then field tested on selected 1/4 acre plots near Lanse,
Pennsylvania.  In general the field tests confirmed the laboratory
finding that latex does reduce the permeability of soil to water.
However, the economics are not attractive and most of the latex is
deposited in the top foot of soil where it is subject to damage by
microbiological attack, frost and surface vegetation.
                                  111

-------
                               CONTENTS
Section

I           CONCLUSIONS

II          RECOMMENDATIONS

III         INTRODUCTION

IV          MATERIALS AND METHODS

V           SOIL CHARACTERISTICS AND NATURAL FIELD CONDITIONS

VI          EXPERIMENTAL

VII         RESULTS OF LABORATORY PERMEABILITY AND ELUTION
            STUDIES

VIII        FIELD PERMEABILITY STUDIES

IX          LATEX IRRIGATION AND EVALUATION AT LANSE

X           LATEX DISTRIBUTION IN SOIL

XI          SOIL BOX EXPERIMENT

XII         COSTS

XIII        ACKNOWLEDGEMENTS

XIV         REFERENCES

XV          GLOSSARY

            APPENDIX
3

5

9

15

31


35

41

47

63

67

75

77

79

81

83

-------
                               FIGURES

No.                                                               Page

1       Topographic map of site at Lanse,  Pennsylvania              7

2       Topographic map of site over the abandoned Kato mine
        near Snowshoe,  Pennsylvania                                 8

3       Unassembled Uniroyal hollow auger                          14

4       USDA soil texture triangle for Lanse soil samples          19

5       USDA soil texture triangle for Kato soil samples           20

6       Soil temperature at Lanse, Pa. at five depths              29

7       Irrigation system for application of latex at Lanse
        site                                                       48

8       Effect of latex on delta soil moisture                     53

9       Effect of latex on delta soil moisture                     54

10      Distribution of SBR in Lanse soil                           64
                                  VI

-------
                                TABLES

No.                                                              Page

I       Sealants Evaluated                                       10-11

II      Particle Size Distribution of Soil Samples from
        Lanse Site                                               16-17

III     Particle Size Distribution of Soil Samples from
        Kato Site                                                IS

IV      Physical Properties of Soils at the Lanse and Kato
        Sites                                                    21

V       Mineral Analysis of Upper Slope Composite Soil Samples
        at Lanse Site                                            22

VI      Chemical Properties and Atterberg Limits of Soils at
        Lanse and Kato Sites                                     24

VII     Water Tables and Elevations at Lanse Site                25

VIII    Hydraulic Head in Piezometer Pipes at Lanse Site         26

IX      Rainfall Data from November 1968 to October 1970         30

X       Addition of Sealants to "Undisturbed" Soil in
        Shrink Tubing                                            33

XI      Penetrating Sealants for A Horizon Soil                  36

XII     Penetrating Sealants for A Horizon Soil                  37

XIII    Penetrating Sealants for B Profile                       38

XIV     Inorganic Chemicals as Soil Sealants                     40

XV      Falling Head Percolation Tests                           42-43

XVI     Darcy Hydraulic Conductivity (K) of Field Permeameters   44

XVII    Falling Head Percolation Tests at Lanse                  45

XVIII   Surface Seal Tests on Spoil Bank at Lanse                46

XIX     Latex Application Schedule at Lanse                      49

XX      Soil Moisture Data                                       52
                                   vii

-------
                           TABL2S (Contd.)

No.                                                              page
XXI     Soil Moisture Data                                       55

XXII    Gravimetric 7» Soil Moisture of Samples Taken
        August 31, 1970                                          56

XXIII   Permeability Tests at Lanse on September 16, 1970        57

XXIV    Permeability Tests at Lanse on October 12,  1970          58

XXV     Permeability Tests at Lanse on October 13,  1970          59

XXVI    Permeability Tests at Lanse on May 5, 1971               60

XXVII   Comparison of Permeability Efficiency with Time          61

XXVIII  Distribution of Tagged J-1405 SBR in Kato Soil (Wt. %)   65

XXIX    Percent Latex in Kato Soil                               66

XXX     Latex Experiments in Laboratory Reconstructed Soil
        Column of Ogdensburg, N. J. Soil                         69

XXXI    Soil Box Water Additions, Soil Tension and Resistance
        Data                                                     70-71

XXXII   Soil Box Permeability Tests                              73

XXXIII  Irrigation Cost at Lanse                                 75
                                 Vlll

-------
                              SECTION I

                             CONCLUSIONS

In laboratory tests using reconstructed soil columns, rubber latex
showed good sealing efficiency when applied at a rate equivalent to
4000-5000 pounds per acre.  Results of field testing, however, were
inconclusive.

The ideal situation in which latex would coagulate in a narrow zone
two to three feet below the surface by reacting with acidic or metallic
constituents of the soil was not attained.  Rather, the latex was
deposited progressively as it passed through the soil (most of it
remained in the top foot), the rate of deposition being dependent on
soil structure and composition as well as on properties of the latex
such as particle size and emulsion stability.  Latex stability appears
to be a more critical property than latex particle size in controlling
penetration.  Addition of excess anionic or nonionic surfactants to
latex improved its penetration into the soil.

A styrene-butadiene (SBR) rubber latex of 2400 * average particle size
with high mechanical stability gave the best balance of penetrating
ability and sealing capacity in laboratory tests and this was confirmed
in tests on a 5 x 5 foot test plot at Lanse, Pennsylvania.  This latex
was subsequently applied to three one-quarter-acre plots at Lanse by
sprinkler irrigation, and compared to adjacent areas that were sprinkled
with water only.  Soil moisture measurements did not enable us to judge
the effectiveness of the seal because differences between sampling
positions were greater than differences between treated and untreated
sections, but permeability tests showed that water flow was much slower
(90-99% reduction) within the top ten inches of soil in the latex-
treated sections.  This sealing effectiveness was reduced after winter-
ing.  There was no deleterious effect on the vegetation as a result of
sprinkling the areas with latex.

A laboratory study conducted in a 5 x 5 x 5 foot box containing uniform
reconstructed soil from a New Jersey source suggests that latex pene-
tration into the soil in the field occurs mainly through natural macro
openings and not through the capillaries of the soil structure.

Dilute solutions of ammonium hydroxide or sodium carbonate (with no
latex present) were found to seal effectively in laboratory soil
columns as well as in field applications.  However, since these chemicals
are water soluble, the seal is temporary.

Raw material costs of latex at the application rates used in the field
testing are in the order of $1000 per acre.  Equipment and operating

-------
costs are $200-500 per acre depending on size of area and availability
of suitable water.

An auger type sampling device was developed for obtaining "undisturbed"
samples of soil.

An analytical method to measure latex in soil was developed.

-------
                              SECTION II

                            RECOMMENDATIONS

The generally disappointing results obtained in the present study,
together with the high material costs involved, preclude any general
recommendation for a continuation of the work.

One recommendation arises, however, from some observations on the
application of latex to spoil banks.  It was found that dilute latex
used on a slope of spoil bank material at an application rate of
400 Ibs per acre (dry basis) stabilized the soil and made it much
more resistant to washout by rainfall.  It is recommended that a
study be conducted to use latex for temporary stabilization and/or
sealing of spoil banks before and during reclamation procedures, on
coal slurry ponds and coal refuse piles.

-------
                                SECTION III

                                INTRODUCTION

Acid pollution from abandoned mines causes serious contamination of
major streams in coal mining areas.  In Pennsylvania alone it is esti-
mated that 3000 miles of streams are devoid of fish and plant life
because of coal mine effluent.   The pollution arises because water
and oxygen react with iron sulfide in the mine cavity to produce sul-
furic acid and iron sulfate, which later find their way back into
surface streams.

Two general approaches to the problem are possible:  (1) prevent the
acid-forming reaction from taking place, (2) treat the acid mine water
after it is formed so that it will not be harmful to the ecology when
discharged into streams and other surface waters.  The work described
in this report belongs in the first category - it was based on the
assumption that acid formation in the mine cavity can be prevented by
keeping water (one of the reactants) out of the mine.  This was to be
accomplished by forming a waterproof seal over the mine cavity which
would prevent the seepage of surface water into the mine.

It was planned to form such a seal at or near the surface of the
ground by applying latex to the soil.  Ideally a dilute latex would be
applied to the soil and would penetrate to a depth of two to three feet
below the surface where it would be coagulated by reacting with a con-
stituent of the soil or by other means.  A seal below the surface of
the soil would have several advantages over a surface seal - it would
not be subject to destruction by traffic, frost or vegetation and,
more important, it would allow the use of the top soil for many agri-
cultural purposes.

A preliminary feasibility experiment was performed prior to application
for the present grant.  It showed that it was indeed possible to form a
sub-surface seal by means of latex.  A core of soil was obtained
locally (Wayne, N. J.) by driving a 6.5 cm inside diameter steel pipe
about 100 cm long into subsoil for a distance of 71 cm.  The soil was
retained within the pipe for this experiment.  A part of the free space
above the soil in the pipe was charged with water, and the seepage rate
of the soil after it had become saturated was 15 ml/min under a total
head (wet soil plus supernatant column of water) of 92 cm.  After the
supernatant water had drained completely, 100 ml of 57, rubber latex
was added and allowed to seep into the soil.  This was followed by
additional water and the seepage rate was found to be 2 ml/min under
the same head.  This seepage rate remained unchanged when remeasured
on the following day.  Since there was no visible evidence of the
latex in the collected drain water it is presumed to have been
retained by the soil.

-------
The work described in this report includes:

         Evaluation of the characteristics of the soils of interest,

         Testing of sealing effectiveness of latexes and other
         materials in reconstructed soil columns in the laboratory,

         Study of field conditions (permeability, etc.) at the
         selected test site, and corroboration of laboratory
         results by small scale field tests,

         Application of the latex sealant of choice to the selected
         test site and evaluation of the seal so obtained,

         Soil box experiments with a uniform soil to interpret
         penetration and permeability behavior of soil under
         controlled conditions,

         Projection of cost based on the test results.

The original plan was to conduct all the field tests on a twenty acre
plot in Lanse, Pennsylvania (called Lanse site throughout this report).
However, easement complications and soil conditions required that the
scope of treatment be reduced to three plots of about one-quarter acre
each.  An alternative site was chosen on State Forest lands northeast
of Snowshoe, Pennsylvania, above the abandoned Kato mine  (called Kato
site throughout this report).  Some preliminary testing was carried
out at the latter site but the projected large-scale field test was
not performed.  Topographical maps of the two sites are shown in
Figures 1 and 2.  Sampling sites referred to throughout this report
are shown on these maps.

-------
                                     Figure  1

                             TOPOGRAPHIC  MAP  OF SITE
                                           at

                              LANSE,  PENNSYLVANIA

                                       Explanation

                                   0    HOLE WITH WATER
                                   •    STAKE
                                  O     TREE  OR STAND
                                                 IOO f«t
                                   CONTOUR  INTERVAL-2f«»t

                                  PLANE TABLED BY  FRANK CARUCCIO
                                  AND JAKE TOLSMA FOft UNIROYAL INC.
                                  PLANIMETRY WITHIN 20 «««t
                                  TDPOORAPHY WITHIN 0.« foot
T POLC
                                                      O
                           ROAD

-------
                   I  I I  I I I
                                                       n
              400    800
                 FEET
                         I ZOO
Fig. 2 - Topographic  map of site over the abandoned
         Kato mine  near  Snowshoe,  Pennsylvania

-------
                                  SECTION  IV

                           MATERIALS AND  METHODS

Only  limited work has been reported in  the literature about the use of
latex in  soil.  Unisol  912'3  (a 9:1 oil-latex emulsion made by
International Synthetic Rubber Company) and Phillips Petroset^ geo-
technic emulsions are surface soil stabilizers which penetrate a
maximum of two inches,  but do not reduce  permeability.  Diamond Alkali
found  that styrene-butadiene rubber latex as a one percent latex-soil
mixture reduced permeability to 0.7 from  21 feet per year under a head
of twenty feet of brine.  We have found no literature reference for
use of latex as sealant at depth  in the soil, although asphalt has
been  applied at depth using civil engineering methods"'''.

Most  of the materials investigated as soil sealants in the present
work were latexes, although some  studies  were also conducted with a
variety of inorganic and organic chemicals.  Of the latexes tested,
most were rubber but a  few were plastic,  e.g. polyvinyl chloride.

Sealants examined in this investigation are listed individually in
Table I with supplier, code number,  and other pertinent information.
The materials consisted of the following  general classes with addi-
tional details as indicated:

     1.  Anionic latexes:

         a.   styrene-butadiene copolymers and terpolymers (SBR)

         b.   butadiene-acrylonitrile copolymer (Paracrils^)

         c.   polyvinyl chloride homopolymers  and copolymers (PVC)
         d.   ethyler.e-propylcne terpolymer

         e.   vinyl  acetate homopolymers and copolymers (VA)

         f.   acrylics (acrylate copolymers)

         g.   2-chlorobutadiene homopolymer (chloroprene)  (prepared
             specially as  cationic material coagulated by soil).


     2.  Water  Soluble polymers:

        a.   polyvinyl ether
        b.   methylcellulose

        c.   polyethyleneimines

        d.   polyacrylic acid  homopolymer

        e.   copolymers  and terpolymers  with acrylic acid  types
             (water soluble as  sodium or ammonium salts)

        f.   alginates

        g.  monomers  (for polymerization  in soil, e.g. acrylamide)

-------
                                            TABLE I
                                       Sealants Evaluated
Supplier




Naugatuck Chem.
Naugatuck Chem.
Uniroyal - R .C
Borden


Code
J1405
J1896
J1925
J2620
J2714
J2752
J2768
J3471
J3595
J3902
Pyratex
-
BJMM
BJLT
Marvinol® 8-78-1
5402
R-8438-20C
Royalene® 502
GM-38
GM-39
GM-45
2134
2140
2153
2158
2635


Type Polymer
SBR
SBR - mod .
SBR - mod .
Paracril
SBR
SBR - mod .
SBR
SBR
SBR
SBR
SBR - mod.
Natural Rubber
Paracril®
Paracril®
PVC
PVC
SBR - acrylic
EPDM
EPDM - parafin extended
EPDM - Aroclor 1242 extended
EPDM - oil + clay extended
VA
VA homo; car boxy la ted
VA ; acrylic
VA
PVC

Part.
Size
a
1100
1500
1400
800
1800
2000
1000
2400
1300
3000
800
(4000+)
-
2000
500-1500
-
(3500+)



-
2000
2000-3000
4000
-


7.
Solids
41
50
46
41
51
47
46
49
50
75
41
50
25
27
35
38
16
55



45
46
55
55
55
Surface
Tension
Dynes /cm.
Cone. 57,
52.5
44.2
64.3
50.4
39.2
36.5
72.2
35.7
77.
38.
56.5
52.
51.9
45.8

43.
33.

34.

42.
40.
42.
44.
37.


10.6
9.6
9.6
10.3
11.4
8.9
9.0
9.8
9.5
10.5
10.3

10.2
7.5
3.8
3.1




3.5
6-7
4.5-5.5
4.6
7.

-------
                                       TABLE I  (cont'd.)
Supplier
Dow

Goodrich
Paisley
duPont
Code
870
874
2679X6
H-9052-0
Neoprene Latex
   950
 (Water Soluble Polymers)
Dow Methocel      HG
Montrek
Goodrich
General Analine
 & Film Gantrez
 12
 600
 K714
 M155
 (Wgter Dispersible Organics)
 Phillips  Petroset SB-1
 Ciba Uvitex       EEC (cone.)
Type Polymer
PVC vinylchloride
PVC vinylidenechloride
Acrylic-self thickening
Vinylacetate
 (Cationic surface of
 40 ammonium type)
Hydroxypropyl methyl
cellulose
Polyethyleneimine
 Polyacrylic acid
 Vinylmethylether
                  Solprene GEO
                  Chem. brightener

Part .
Size
I


400-800
M.W.


7,
Solids
(50)
50
48.5
50.
50.
Surface
Tension
Dynes /cm.
Cone. 57.
36.
34-37
45.
46.

12,000 100
40,000-60,000 34
200,000


8.
8.
8.8
4.5
                                        10

-------
     3.  Inorganic dispersions with and without additives

         a.  expandable clays (montmorillonite type)

         b.  non-expanding clays (kaolin type)

         c.  finely divided silica (cabosil)


     4.  Water soluble inorganics
         a.  sodium carbonate

         b.  ammonium hydroxide

         c.  sodium hexametaphosphate.


A variety of chemicals and latexes were screened at concentrations of
about 2-1/2 to 5% for their ability to (a) penetrate soil and  (b)
reduce the percolation of water through it.  Initial laboratory experi-
ments were made with reconstructed soil samples.  Effectiveness of the
sealant was measured by change in percolation rate.  Percolate showing
turbidity was evaporated to measure the amount of eluted latex.

To be useful for the present purpose, a latex should be able to pene-
trate into the soil (ideally, 2 or 3  feet) before forming a seal.
Latex which reacts immediately with the soil and forms a seal  at the
surface is not satisfactory because the seal would be subject  to
destruction by mechanical forces (traffic, natural movement of ground
due to freezing and thawing, growth of vegetation, etc.) and chemical
action (atmospheric oxidation, etc.).  The amount of latex eluted
through a 4- or 8-inch deep column of soil gives some indication of
its penetrating ability.  In other experiments (referred to herein as
"skinning" tests) the extent of sealant penetration was measured by
removing soil from the top of the laboratory column in 1.5 to  2.0 cm
increments and measuring the corresponding percolation rates.

Some of the more promising sealants were  then evaluated in the field
by one or more of the following procedures:

     1.  Permeability tests were performed in situ by pressing
         eight-inch-diameter, twelve-inch-long, stove pipes into
         the ground eight to ten inches and measuring the time of
         disappearance of a measured  amount of water.  The thin
         walls of the stove pipes made compaction minimal; stones
         in some cases created deleterious wall effects.  Alter-
         natively eight-inch-diameter permeameter tubes thirty
         inches long and 1/4 inch thick with a sharp edge at the
         lower end were driven into the soil twenty-four inches
         to measure permeability in situ.  Less than five per-
         cent compaction occurred.
                                    12

-------
      2.  Sealant was applied to test areas of five by five feet.
          Several weeks later core samples were removed in four-
          inch increments to depths of 8 and 16 inches by means
          of a thin-walled sampling tube.  Samples were examined
          for presence of sealant.  The hole that had been cored
          was confined by introducing an empty sampling tube of
          the same diameter and then pressing it to a depth one
          inch greater than the original sample.  Percolation
          rates were measured in these tubes by the falling
          head method.

 Soil density measurements were made by obtaining the net weight of soil
 in a given volume.   Typical values for reconstructed columns were about
 1.40 to 1.45 after draining several days and about 1.34 to 1.38 after
 air drying about a week.   Soil heights in the column sometimes varied
 0.5 to 1.0 cm because of preferential redistribution of top-soil on
 adding make-up water.  Less disturbance was evident when a Marriott
 feed bottle was used to maintain a constant hydraulic head.

 Ideally,  soil samples tested for permeability in the laboratory should
 be identical in physical  condition to the soil  in the field.   A com-
 mercial split-tube  sampler,  having a  heavy wall to contain the split
 tube,  produced soil  samples that were severely  compacted.   Thin wall
 sampling tubes were  much  better in this respect but noticeable com-
 paction occurred if  samples longer than four inches were cored.   Addi-
 tional  compaction occurred  when the soil was removed from the tube.

 A  jack-type hollow auger  modified from a description of similar equip-
 ment by Hayden and Heineman  was built to take  longer samples three
 inches  in diameter.   It  features an outer sleeve  containing  a peripheral
 spiral  which,  when turned,  forces  itself into the soil,  and  an inner
 sleeve  that contains  either  a metal split sleeve  or a Plexiglas  tube
 which  is  held  stationary  while  filling with soil.   (See  Figure 3)
 However,  wall  effects were  evident  when permeability tests were  per-
 formed  in the  Plexiglas tubes.   Undisturbed  soil  cores  are  impossible
 to  obtain if  stones or strong roots are  present.   Because  of  these
 difficulties most of  the  laboratory permeability  tests were  performed
 on  reconstructed  soil columns.

One, three, and  five  foot deep  piezometer  pipes were  placed at the
Lanse and Kato  field  sites.   The water  from  the ground seeps  into the
 tubes and is a measure of the depth of  the  permanent  or  perched water-
 table at  that  point.
                                    13

-------
Fig. 3 - UNASSEMBLED UNIROYAL HOLLOW AUGER
      (WITH PLEXIGLASS AND SPLIT-TUBE LINER)

-------
                               SECTION V

            SOIL CHARACTERISTICS AND NATURAL FI^LD CONDITIONS
This section contains a summary of our investigation of the soil con-
ditions (type,  porosity, moisture content, etc.) at the two sites
(Lanse and Kato, Pennsylvania) selected for field testing of latex as
a soil sealant.  Other natural conditions such as rainfall, frostline,
depth of the water table,  etc., were also measured and recorded for
possible later use in assessing the effect of the latex seal on the
total environment.

The numbered sites referred to throughout this report are shown on the
maps of the Lanse and Kato sites (Figures 1 and 2).  KA, KB and KG are
positions at the Kato site - all other numbered positions refer to the
Lanse site.

Two other terms used throughout the report which should be explained
are  'horizon" and ''profile".  These are synonymous terms for soil
layer, A horizon  (or A profile) referring to topsoil and B horizon
(or B profile) referring to subsoil.

The properties that control the flow of water through the soil are of
special interest because these will also govern the passage of a
sealant into and  through the soil.  Normally the particle size distri-
bution is a good measure of soil permeability, which is directly pro-
portional to sand content and  inversely proportional to silt and/or
clay content.  Tables II and III show  the percentages of gravel or
stones, sand, silt and clay of Lanse and Kato soil samples.  According
to the United States Department of Agriculture classification, the
gravel is ignored ami the residue separated into sand,  silt, and clay
fractions.  Figures 4 and 5 indicate by means of the U3DA soil texture
triangle that the variability of composition of Lanse soil ranges from
sandy loam in the creekbed to clay loam.  The variation is even greater
among soil samples from Kato.

Torosity of soil, also known as percent voids, is another  indication  of
how readily water will  travel  through  soils.  Soil density usually
increases and porosity decreases xvith  depth, and as expected,  the
porosity of the  top soil at Kato, a forest, is exceptionally porous
and open (see Table IV).

The pore diameter and length of soil capillaries is extremely variable,
and its distribution in the field is impossible to determine.  In
addition to the  capillaries, soil also contains many macro cracks,
crevices, and passages around  stones and roots.  All these affect the
flow rate of water through soil.  i'oiseuille ' s law states that the rate
of flow of a liquid through a  narrow tube is proportional to the
4th power of the  radius of the  tube.   Tore  size normally decreases with
depth and in soil containing appreciable amounts of silt and clay this
is especially  true.  Thus, deeper soils should be less  permeable and
such was the case at Lanse and Kato.


                                  15

-------
                                TABLE II

       Particle  Size  Distribution of Soil Samples fron Lanse Site
       Depth  Increment (inches):
 Soil  Texture of Upper Slope Composite
  % Gravel  (+2  mm)
  7o Sand  (+0.05 -  2 mm)
  7o Silt  (+0.005 -  0.05  mm)
  % Clay  (-0.005 mm)

 U.S.D.A.  Classification
 (without  gravel)
  % Sand
  % Silt
  7o Clay
0-8
8-16
16-24   24-32
Soil Texture of Creekbed Composite
  70 Gravel
  70 Sand
  7o Silt
  70 Clay

U.S.D.A. Classification
(without gravel)
  7o Sand
  7o Silt
  7o Clay
Soil Texture of Pine Woods Composite
  7» Gravel
  % Sand
  70 Silt
  7o Clay

U.S.D.A. Classification
(without gravel)
  7o Sand
  % Silt
  7, Clay
6.3
35.7
32.3
25.7
39.7
31.4
28.9
Sandy
Clay
Loam
4.1
47.3
35.3
13.3
49.3
36.8
13.9
Loam

10.0
35.7
40.8
13.5
39.7
45.3
15.0
Loam
7.7
34.3
31.8
26.2
38.8
31.2
30.0
Clay
Loam

12.6
50.5
25.1
11.8
57.8
28.7
13.5
Sandy
Loam
6.8
34.5
40.7
18.0
37.0
43.6
19.3
Loam
13.6
36.4
25.0
25.0
43.4
26.6
30.0
Clay
Loam

11.1
56.5
22.5
9.9
63.5
25.3
11.2
Sandy
Loam
13.4
32.3
35.3
19.0
37.3
40.8
20.9
Loam
21.5
36.3
21.0
20.2
48.2
25.0
26.8
Clay
Loam

32.7
40.8
19.2
7.3
60.6
28.6
10.8
Sandy
Loam
17.3
39.3
31.6
11.8
47.5
38.2
14.3
Loam
                                  16

-------
                           TABLE II  (Cont.)
      Depth Increment (inches)
Soil Texture at Site 03A
  7, Gravel
  % Sand
  % Silt
  7o Clay

U.S.D.A. Classification
(without gravel)
  7o Sand
  7o Silt
  % Clay
Soil Texture at Site 3C
  "I, Gravel
  7o Sand
  7. Silt
  7, Clay

U.S.D.A. Classification
(without gravel)
  70 Sand
  7. Silt
  7. Clay
0-8
8-16
16-24   24-32
5.1
30.3
48.6
16.0
31.9
51.3
16.8
Silt
Loam
4.4
28.4
42.2
25.0
29.7
44.1
26.2
Loam

12.1
27.8
44.1
16.0
31.6
50.2
18.2
Silt
Loam
8.0
22.3
39.0
30.7
24.2
42.5
33.3
Clay
Loam
36.8
19.3
31.5
12.4
30.5
49.9
19.6
Silt
Loam
1.5
20.1
42.4
36.0
20.4
43.0
36.6
Clay
Loam
61.1
10.2
19.5
9.2
26.3
50.1
23.6
Silt
Loam
5.8
23.2
38.4
32.6
24.6
40.8
34.6
Clay
Loam
                                   17

-------
                               TABLE III

      Particle  Size  Distribution of Soil Samples from Kato Site
        Depth  Increment  (inches):
Soil Texture of I
-------
xllOO
                                                 SILT
                                                   CLAY
                                                    LOAM	\
    \/
  SANDY
   CLAY
     LOAM
                                                                Selected
                                                                   Samples

                                                                +  0"-8"
                                                                a  8"-16"
                                                                o  |6"-24"
                                                                *  24"-32"
                                                                Lonse	

                                                                I. Upper Slope
                                                                2. Creekbed
                                                                3. Pinewoods
                                                                4. 03A
                                                                5. 3C
               60            40            20
                 Percent Sand
Fig. 4 - USDA soil texture triangle for Lanse soil samples
0 B

-------
       20
A 100
                                                                      -I-  0" - 8"
                                                                      a  8" - 16
                                                                      o  16"-24"
                                                                      A  24"-32"
                                                                      Koto
                                                                       I. KA
                                                                      2. KB
                                                                      3. KC
                                                    SILTY  /
                                                     CLAY
                                                       LOAM
                                                                               II
80
60            40
  Percent Sand
20
              Fig. 5 - USDA soil texture triangle for Kato soil samples

-------
              TABLE IV

Physical Properties of Soils at the
        Lanse and Kato Sites
         Depth Increment (inches):    2-4       4-8

Soil 2C cored January 29,  1969 (Lanse)

  7« Moisture
  Sample volume, cc.
  Dried wt., g.
  Bulk density D^, g/cc
  Total porosity, ST  = 100
      Pp = 2.65 g/cc
         Depth Increment (inches):

Soil KA cored July 8, 1969 (Kato)

  % Moisture
  Sample volume, cc
  Dried wt., g.
  Bulk density, g/cc
  Total porosity

Soil KB cored July 8, 1969 (Kato)

  7. Moisture
  Sample volume, cc
  Dried wt., g.
  Bulk density, g./cc.
  Total porosity

Soil KG cored July 8, 1969 (Kato)
  7» Moisture
  Sample volume, cc.
  Dried wt., g.
  Bulk density, g./cc.
  Total porosity
         Depth Increment  (inches)
Soil  3D cored November  6, 1969  (L

  7. Moisture
  Sample volume, cc.
  Dried wt., g.
  Bulk density, g./cc.
  Total porosity
                                      8-12
12-16
nse)
32.3
64.33
71.0
1.10
58.5
0-8
21.6
411.6
334.6
0.812
69.3
22.2
411.6
323.8
0.785
70.4
23.3
411.6
286.5
0.698
73.7
0-4
nse)
29.8
205.8
188.2
0.914
65.5
24.5
205.9
261.9
1.27
52.0
8-16
14.0
411.6
603.3
1.46
44.9
14.3
411.6
525.8
1.28
51.7
16.3
411.6
558.9
1.36
48.3
4-8
26.1
205.8
212.7
1.03
61.0
20.8
205.9
256.2
1.24
53.2
16-24
15.5
411.6
596.6
1.45
45.3
11.5
411.6
59.09
1.44
45.6
13.5
411.6
517.0
1.26
52.5
8-12
23.6
205.8
245.8
1.19
55.1
                                                  9.3
                                                205.9
                                                323,
                                                  1
                                                 40
    5
    57
    7
                                                24-32
                                                 12.5
                                                411.6
                                                550.0
                                                  1.34
                                                 49.5
                                                 13.5
                                                411.6
                                                677.6
                                                  1.64
                                                 38.0
                                                 15.7
                                                411.6
                                                609.5
                                                  1.48
                                                 44.1
                                                12-16
                                                 23.7
                                                205.8
                                                310.4
                                                  1.51
                                                 43.0
                   21

-------
Under saturated conditions, soil water is under no tension but is
subject to gravitational forces only.  Saturated flow of water in soil
is therefore normally downward, unless an impervious lower layer
induces lateral flow.  This downward gravitational force will carry
latex down to the depth of small capillaries where particle blockage
can occur.

Elemental analysis of Lanse soil (Table V) combined with crystal-
lographic examination revealed that the major component of < 2 ram
fraction of the soil is quartz.  The relatively high aluminum and iron
concentration indicates the presence of layered hydrated clays.
                               TABLE V
       Mineral Analysis of Upper Slope Composite Soil Samples*
                            at Lanse Site

                Depth Increment (inches):     0-8      8-16     16-24
   A12°3
   Fe203

   Mn02

   MgO
   GaO
   Ti02
   Na20
   K20

   CuO
   NiO

   CoO
   Ignition Loss  at 900°C  -  %
       '•Analyzed  at  Penn.  State  University  Minerals  Composition
       Laboratories except  for  Cu,  Ni,  and Co.
Composition (%")
79.5
10.3
6.25
0.09
0.35
0.17
0.82
0.41
1.60
0.003
0.004
0.008
6.96
75.5
13.8
6.16
0.05
0.48
0.10
0.83
0.38
2.09
0.003
0.004
0.002
6.25
75.5
14.8
5.77
0.07
0.60
0.09
0.86
0.38
2.18
0.005
0.006
0.003
5.90
                                  22

-------
Dr. Frank Caruccio, our consultant on this project, examined soil
samples at sites 03A (Lanse) and KA (Kato) at four depth increments
using an optical microscope with polarized light to differentiate
between quartz clay and hydrated layered silicate clays.  Lanse 03A
soil was found to contain a relatively high amount of clay resembling
montmorillonite in structure whereas the Kato KA soil has less
hydrated clay of different shape that suggests a micaceous structure.
The quartz clay particles were stained with iron compounds.  X-ray
diffraction of clay separated from 03A soil at 24-32 in. depth con-
firmed the presence of montmorillonite.  Cation exchange capacity (CEC)
values ranging from 6 to 13 meq/100 g of soil (Table VI) also prove
the presence of reactive minerals.  Chemically active mineral groups
in the soil interact with latex and other additives.  Physical inter-
action with sealant particles can occur by adsorption and surface
effects.

It is common among soils developed in the temperate zones to have a
lower pH at a two or three foot depth and it was hoped to make use of
this difference to effect the coagulation of latex, since latex
stability is very dependent on pH.  We looked for this effect in the
soil of the Lanse and Kato areas but as can be seen in Table VI the
pH variation with depth is very small.  At sites KB and KG, areas
where rotting oak leaves cover the ground, the pH is generally lower.

The Atterberg limits measure the consistency of the soil containing
enough water to make a smooth paste.  The liquid limit  (L.L.) or upper
plastic limit is the percent moisture at which soil becomes semi-fluid.
The plastic limit  (P.L.) is the percent moisture at which the soil
crumbles when it is rolled to a 1/8 inch thick thread.  The plastic
index (P.I.) is the difference between the liquid and plastic limits.
All three are directly related to the amount of clay in the soil.
Because we added the soil sealant by an irrigation method, the water
holding capacity is of interest.  Except for the creekbed soil, which
is relatively low  in clay, all the samples have the desirable positive
plastic index.

Lanse Site

The accompanying topographic map  (Figure 1) shows  the experimental test
area at Lanse, Pennsylvania.  Brown's Run is a dry creekbed during the
summer but a flowing stream in the winter and spring.  The area has
four distinct topographic features that account for the variability  in
soil properties.  Along part of the creekbed is a boggy area at flood-
plain level, there are slightly sloping upper plain areas on each side
of the stream, there is a stand of pine trees at the northwest corner
of the property, and spoil banks  from previous strip mining operations
both to the northeast and northwest (not shown on map).

From December to June, when the stream was flowing  (or  frozen), there
was a perched water table which we measured occasionally  (see Table VII)
                                  23

-------
                               TABLE VI

             Chemical Properties and Atterberg Limits of
                    Soils at Lanse and Kato Sites
                 Depth Increment  (inches)   0-8   8-16   16-24   24-32
pH
  Upper slope composite (J.anse)
  Pinewoods composite (Lanse)
  Creekbed composite (Lanse)
  03A  (Lanse)
  3C   (Lanse)
  KA (Kato)
  KB (Kato)
  KC (Kato)

Cation Exchange Capacity  (meq/100 g)

  03 A
  KA
  KB

Atterberg Limits (7, moisture)

  Upper slope composite L.L.*
                        P.L.*
                        P.I.*
  Pinewoods composite   L.L.
                        P.L.
                        P.I.
                        L.L.
                        P.L.
                        P.I.
                        L.L.
Creekbed composite
3C
  03A
  KA
  KB
  KC
                          I.
                      P.L
                      P.I
                      L.L
                      P.L
                      P
                      L.L.
                      P.L.
                      P.I.
                      L.L.
                      P.L.
                      P.I.
                      L.L.
                      P.L.
                      P.I.
5
5
5
5
6
6
4
5
10
9
7
30
26
4
27
26
1
29
26
3
31
26
5
30
26
4
30
25
5
-
-
-
33
31
2
.9
.5
.7
.6
.2
.4
.8
.2
.4
.9
.8
























5
5
6
5
6
6
4
4
7
6
5
27
22
5
23
20
3
22
21
1
31
25
6
26
20
6
28
21
7
22
18
4
29
23
6
.7
.6
.0
.5
.3
.1
.8
.8
.0
.9
.5
























5
5
5
5
6
5
4
4
9
6
4
28
22
6
23
18
5
19
20
-1
35
29
6
26
21
5
31
21
10
22
18
4
30
23
7
.3
.5
.8
.3
.2
.6
.9
.8
.9
.3
.6
























                                                                5.8
                                                                5.5
                                                                6.0
                                                                5.9
                                                                5.0
                                                                4.9
                                                                7.4
                                                               11.6
                                                               12.8
26
21
 5
22
18
 4
18
.19
-1
35
26
 9
27
21
 6
33
22
11
26
18
 8
35
24
11
* L.L. = liquid limit; P.L. = plastic limit; P.I. - plasticity index.
                                  24

-------
                              TABLE VII

              Water Tables and Elevations at Lanse Site
Hole Elevation
Position (ft)
1A
IB
2A
2B
2G
3A
3B
3G
4B
4C(N)
4C(S)
5C
Creekbed S.
Creekbed N.
03A
1565
1562
1570
1565
1563
1571
1568
1570
1572
1570
1570
1572
1558
1570
1574
12/3/68
-
2
-
17
5
23
11.5
17
11
-
-
-


•»
12/11/68
-
-
21
16.5
26
21
19.5
19
29
18.5
20


—
4/7/69
3
0
5
13
4.5
20
12
15
9
25
13
-



4/11/69 5/2/69
3
2
8
17
8.5 14
22.5
17
18 20
11.5 17
28 28.5
14.5 19
17.5 21



The relative water table depth and position elevations show that the
water flows laterally through the soil toward the stream bed.  The data
from piezometer pipes at site 4B confirms that the perched water table
is caused by a relatively impermeable layer between three and five feet
from the surface because, while the perched water table is evident in
uncast holes, the hydraulic head is much higher at three feet depth
than at five feet (Table VIII).

We had expected a soil of moderate permeability from the soil texture,
but the presence of a perched water table indicated low permeability,
so we asked Mr. William L. Braker, soil scientist from the USDA, Soil
Conservation Service, Clearfield Office, to examine and define the
soil for us.  His report (condensed) with definitions follows:

     "A deep soil (which all of these are), is one in which depth to
     hard bedrock is greater than forty inches.  Soil permeability
     rates as defined by USDA, Soil Conservation Service, are as
                                   25

-------
                              TABLE  VIII
           Hydraulic Head in Piezometer  Pipes  at  Lanse Site
Position 4B
Position 4CN
      Depth:

  Date

  12/11/68
  1/29/69
  2/27/69
  A/11/69
  5/5/69
  5/23/69
  6/12/69
  7/7/69
  1/29/70
  2/12/70
  2/27/70
  3/18/70
  3/30/70
  4/20/70
  5/12/70
  6/1/70
  6/12/70
  6/20/70
  6/25/70

(placed  4/11/69)

  5/5/69
  5/21/69
  6/12/69
  1/29/70
  2/27/70
  3/18/70
  3/31/70
  4/30/70
  5/12/70
  6/12/70
                                     1  ft
0
0
0
0
0
0
0
0
3.5
0
0
0
2.5
1
0
0
0
0
0
                                     0
                                     0
                                     0
                                     0
                                     0
                                     0
                                     0
                                     0
                                     0
                                     0
24
19
16
23
16
11
 2
 0
31
27
26
23.
30
26
10,
22
 1,
 8.
15
                         5 ft.
0
0
5
2.5
3
3
4
2.5
4
7
4.5
4
4.5
3.5
4
5.5
6
5.5
5
.5
.75
0
2.5
5
2
4
2
0
0
0
.25
0
0.5
5.5
5
6
3.5
2
3
                                 26

-------
     follows:

       Descriptive  Term  Range in Inches/Hour  Feet/Year

       Slow                 less than 0.2       < 146

       Moderately slow      0.2 - 0.63          146 - 460

       Moderate             0.63- 2.0           460 - 1460

       Moderately rapid     2.0 - 6.3          1460 - 4600

       Rapid                 more than 6.3       > 4600

     "The primary reason for the slow and moderately slow permeabil-
     ties are  the fragipans in the subsoils of the Ernest and Brink-
     erton soils; and is due to the position in the landscape of the
     Atkins  soil.   Ernest is the most extensive soil on this plot.
     The Atkins is the floodplain soil along the creek.  Brinkerton
     is the wetter upland soil on the southwest side of the plot.

     "Ernest are deep, moderately well to somewhat poorly drained
     soils of the uplands.  They have developed in loamy, colluvial
     material  derived from shale and sandstone bedrock.  These
     nearly level to moderately steep soils have a moderately slowly
     permeable fragipan subsoil.  The water table normally rises to
     within 12 and 18 inches of the surface during wet periods of
     the year.

     "Atkins are deep, poorly and somevrtiat poorly drained flood
     plain soils.  They have developed in loamy alluvial sediments
     eroded from higher residual soils derived from gray and brown
     shale and sandstone.  These nearly level soils have a moder-
     ately slowly permeable subsoil.  The water table normally
     rises to the surface during much of the year.  Most use
     problems are related to flooding, the high water table, and
     to the moderately slowly permeable subsoil.

     "Fragipan is a subsurface soil structure that seems to be
     cemented when dry, but is fragile when moist, has a relatively
     high bulk density and is slowly or very slowly permeable to
     water.  Fragipans may be a few inches to several feet thick."

The Brinkerton soil does not concern the problem because we were unable
to obtain an easement for the property in the southwest corner of the
plot.  The two plots to which latex was applied are Atkins soils.  The
spoil bank is not a soil in the real sense of the term.

Freezing and thawing changes the structure of soil because of the expan-
sion of ice lenses in the soil   .  To determine how deeply the ground
freezes on the site in the winter, resistance thermometers were placed
at 2-inch and 1, 2, 3, and 4 foot depths at site 4B at Lanse and soil
                                   27

-------
temperatures were measured one or two  times  a week  (see  Figure  6).  At
no time did the soil temperature go below 32°F  at the  one-foot  depth.
From periodic observations in the field  the  frost line was determined
to be about eight to ten inches deep.  The saturated condition  of  the
soil and presence of a water table probably  limited the  frost-line
depth. The implication is that the soil  structure is affected only  to
a depth of 8 to 10 inches.  The mean monthly air temperature is
included for comparison with the soil  temperatures.

A rain gage was obtained from and set  up by  the Pennsylvania Depart-
ment of Forests and Waters.  Data in Table IX compares rainfall rates
at Lanse, Clarence (which is near the  Kato site), and  Philipsburg
FAA-AP.  The totals show the variability of  rainfall within a thirty-
mile span.

Kato Site

We began to investigate the Kato site  in June, 1969.   The forested
mountain top of about 90 acres,  of which 70-75 acres are State  Forest
Land,  overlies the abandoned Kato deep mine.  Since it is surrounded
by an unrestored high wall and spoil banks,  vehicular  access to the
area is available at only one place from the southeast corner.  The
only water close to the site is acid mine water from drainage holes
at the base of the mountain.

Piezometer pipes were placed at one,  three,  and five foot depths about
three hundred feet from the high walls at the south (KA), northeast
(KB),  and northwest (ICC) border of the State Forest land.  No water
was found in these pipes.  Water table measurements were in uncased
holes made when coring soil samples contained water.
                                   28

-------
 82
 78
 74
 70
 66
 62
 58
54
50
46
42
 38
34
30
26
     24.2
                Monthly Mean Air Temp. °F
24.3\  3O.O\46.8 . 5^/  i 5^.^ , 6^7 , 65.3 , Sfitf , 472
  Dec.  Jan.   Feb.  Mar.   Apr.   May   June  July  Aug.  Sept.  Oct.   Nov.  Dec.
  1968  1969
                    Fig.  6 - Soil temperature at Lanse, Pa.  at five depths

-------
                               TABLE  IX

           Rainfall Data from November 1968  to  October  1970
Philipsburg, Pa .
November 1968
December
January 1969
February
March
April
May
June
July
August
September
October
Annual Total (inches)
November
December
January 1970
February
March
April
May
June
July
August
September
October
Annual Total
4.01
2.93
1.47
0.67
1.31
3.81
3.91
5.92
6.26
3.14
1.82
2.36
37.61
4.38
3.82
1.06
3.56
2.39
4.36
3.30
5.02
6.17
6.17
2.40
4.38
45.73
Clarence. Pa
4.29
3.32
1.33
0.78
1.23
2.35
2.59
4.55
5.23
3.92
1.34
2.80
33.73
3.59
6.44
1.09
3.61
2.66
4.08
4.07
3.27
6.78
6.78
1.67
3.95
44.14
1 9
Lanse, Pa.''


1.35
0.30
1.45
2.85
2.10
2.90
4.70
2.20
2.35
1.35
21.55
(10 months)
2.80
2.70
2.15
1.30
2.40
3.45
2.95
2.45
5.15
5.15
1.45
4.20
35.10
(1)  Data from U.S. Dept.  of  Commerce, ESSA, Environmental Data Service,
(2)  Data obtained  from raingauge  on the site.
                                 30

-------
                              SECTION VI

                             EXPERIMENTAL

LABORATORY PROCEDURES

Reconstructed Soil Samples

Most laboratory sealant studies were carried out in 2.0 inch I.D.
Plexiglas tubes of 9.0 inches height containing 4.0 to 4.5 inches
reconstructed soil from an 0-8 or 8-16 inch soil horizon at the
Lanse site.  (When soil was used from the Kato site, the top soil
horizon was approximately 0-4 inches.  The 4-8 inch horizon was some-
what comparable to the Lanse 8-16 inch horizon.)  A 20 mesh Saran or
steel screen was placed over the rubber stopper with a bottom drain to
a 100 ml measuring cylinder.  In some earlier experiments a 1/2 inch
bed of 1/4 inch pea-gravel was then added, but in later experiments
this was eliminated as it showed no particular utility.  Field soil
samples collected by a shovel were stored in sealed containers and
sieved through a 4 mesh screen just before use.  (Moisture content
was typically about 25% for A horizon soil and about 20% for B horizon
soil.)  The sieved soil minus roots was then added  in increments of
about 1 1/4 inches.  Compaction was effected by dropping a weighted
ramrod (1-7/8 inches O.D.) five times from a height of two inches.  If
about twenty tamps were used or if the soil columns were rapped vigor-
ously while water-logged, compaction was high with  no detectable per-
colation.  Freezing such samples would then usually cause very slow
percolation, e.g. 1-2 ml/day.

Using this as a standard procedure, percolation flows were usually
stablized after two to three days to rates of about 100 to 200 ml/hour.
Percolate was recycled carefully to minimize disturbance of the  soil
surface.

When percolation flows were about 100 to 200 ml/hour, various additives
(sealants) were introduced to the moist column  in  two to three incre-
ments to maintain a slight head.  Additive concentrations were usually
2-1/2% with some at 5% in earlier experiments.  Typical use of 40 ml
of  2-1/2% additive corresponds to 4400 Ibs/acre.

When no more additive was present as supernatent liquid, water washes
of  about 15 ml were  then used several times  followed by special  addi-
tives as needed.  More water washes were  then used  to maintain a head
of  about 1/2 inch.  Percolate was collected  in  100  ml cylinders  to
measure percolation rates in ml/hr.  Turbid  percolate was evaporated
to  determine amount of eluted latex.  Percolate was sometimes checked
for pH by Pan-pH Indicator Paper9»^° reading in 0.5 units.   Surface
tension measurements were sometimes made  by  a DuNouy tensiometer10  in
experiments using a  surfactant  for  pretreatment.
                                    31

-------
 Clear  percolate  wash  liquid  from latex  applications  was  usually
 recycled  carefully  but washes  of soluble  sealants  were discarded,  using
 fresh  tap-water  as  make-up.

 Since  the hydrostatic head above the  soil  was  usually about  1/2  inch
 average,  Darcy values (K) were calculated  by the equation
                                   h A

              where K = hydraulic conductivity  (ft/day)

                    L = length of column  (ft)

                    h = effective head  (ft) = L + y^

                    Q = ft  /day effluent

                    A = cross-sectional area
                    — = hydraulic gradient
                    LI
Percolation flows of 1.00 ml/hour thus correspond to a Darcy value of
0.0346 ft/day or 12.6 ft/year where L = 4 inches and h = 4.5 inches.
(Darcy values in this report have been converted to feet/year to avoid
fractional numbers.)

Reduction of percolation is frequently used as a measure of sealant
efficiency in this report and is defined by the equation


                      % Reduction = a ~ b x 100
                                      a

               where a = ml/hour before use of sealant

                     b = ml/hour after use of sealant
"Undisturbed" Soil Samples

Although most of the laboratory work with sealants was done with recon-
structed soil samples, some'undisturbed"field cores were also tested
for percolation characteristics before and after use of additives (See
Table X).

Containment of the soil samples for this work was accomplished by one
of two procedures:  1) coating with a cold-molding polysulfide-'-^' H
and 2) coating with a polyolefin shrink tubing* '   .
                                  32

-------
                              TABLE X

     Addition of Sealants to "Undisturbed" Soil in Shrink Tubing

                 Sll    S12     S14    S3     S4     S8      S9
Soil Source
Depth (in.)
Sealant
  Ib/acre
  hr to add
Surface Film
KB
0-8
NH,
4000
29
No
KB
8-16
VA
2000
29
Yes
KA
8-16
NH4
4000
29
No
2D
0-8
VA
750
29
Some
2D
8-16
VA
1150
29
Yes
4B
8-16
NH4
2700
29
No
3A
0-8
3471
6200
-
Low
                             Darcy Permeability (Ft/Yr)
Start
6/19-6/26/68
6/26-7/3
Dried out for
one month
8/2-8/9
8/9-8/14
8/14-8/27
8/27-9/5
9/5-9/15
9/15-9/20

   Ave.
   Efficiency
31.
0.
0.
2
73
317
2.9
0.73
2.6
15.
0.
*
8
73
98
0.73
.73
.73
2.
,
»
2
73
73
2.
•
f
6
37
73
12.


8-3.7
-
-
1.3
3.5
2.4
3.3

2.9

2.7
91
2.1
2.1
1.5
2.0
2.0
2.9

2.1
28
8.7*
1.8
1.8
1.5
 .74
 .55

1.3
92
0.41
 .96
 .92
 .96
Low
.30
.66
.84
.34
.96
.67

.63
 71
,73
.0
,81
.89
.91
.44

.80
 70
 .34
 .60
 .74
 .93
1.1
 .64

 .72
  70
      * S14 Application of 10,000 Ib/acre Na_C03 as 5% solution
        VA  = Borden vinyl acetate #2140
        NH^ = Gale on NH3; applied as 2% NH4OH
        S9  = Naugatuck SBR J3471 to thin-walled sampling
              tube for one month before transfer
              Cores of 6.7 inches height; no eluted latex
                                    33

-------
 Perma-Flex CMC Blak-Tufy (a polysulfide)  gave excellent adherence to
 the soil with no wall effects.   The three components were mixed as
 directed and a coating brushed  on the vertical soil core resting on a
 1/2 inch bed of pea-gravel  above a rubber stopper with a hole for a
 drain.   (Modeling clay was  used as a seal around the gravel and masking
 tape was used to extend the core about 3/4 inch.)  The soil column was
 placed  in a circular container  and then liquid mix was added to the
 height  of masking tape.  Setting occurred in about one hour.  The
 following day the firm soil column was used for percolation studies with
 a head  of about 1/2 inch water.

 The shrink tubing procedure was recently  described by Bondurant  .  The
"undisturbed"soil core was placed inside the tubing above a 1/2 inch bed
 of pea  gravel supported by  a one-hole rubber stopper.  Shrinkage was
 effected by a heat gun.  No appreciable wall effects were evident when
 tested  with fluorescent additives.   Shrink tubing was not useful with
 reconstructed soil samples.

 Soil Characterization

'Soil characteristics were determined by standard methods wherever
 applicable.   Samples were prepared  according to ASTM Method D421-58
 Particle size distribution  was  determined by the method for "Grain
 Size Analysis of Soils",  D422-6316  using  Apparatus A.   Soil particle
 specific gravity was determined by  method D854-58  .  Oven dry mois-
 ture content was measured according to procedure D2216-63T  .   Liquid
 and plastic limits were determined  by methods  D423-61  * and D424-5920
 respectively.   The cation exchange  capacity (CEC)  was  analyzed by the
 ammonium acetate method described by Bear  .   Mr.  Norman Suhr,  assis-
 tant director of the Mineral  Constitution Laboratories at Pennsylvania
 State University,  analyzed  heavy metals content of three soil  samples
 by atomic absorption,  and sodium and potassium by flame photometry.
 Our analytical  laboratory made  elemental  analyses  by means of  emission
 spectroscopy.   The pH was measured  on soil  samples that were slurried
with an equal weight of distilled water.

FIELD PROCEDURE

Measuring the effectiveness of  latex as a soil  sealant when applied  to
 test plots  in the  field was not straightforward.   In the present work,
since the test  area  was small compared  to the  size of  the underground
mine, the quantity and  quality^of the mine  effluent could not  be
expected  to  change  appreciably.  Thus our method  of estimating sealing
effectiveness did  not  involve monitoring  of mine  effluent,  but con-
sisted  of measuring  soil  moisture under the treated area at depths to
five  feet and comparing the data to  those from an adjacent (untreated)
area.   If the seal was  effective, moisture  levels  would be lower under-
.neath the treated  area,  although it  was recognized that lateral  move-
ment of ground water could  have an effect on the results.
                                  34

-------
                            SECTION VII

       RESULTS OF LABORATORY PERMEABILITY AND ELUTION STUDIES

Sealants for A Horizon Soil

The twelve latexes listed in Table XI penetrated A horizon soil as
shown by elution through a laboratory soil column and gave good
sealing efficiency.  See details in Table XII.  High elution through
laboratory soil columns of 4 to 6 inches depth is desirable because
it predicts penetration of latex to a greater depth in the field.

No definite correlation was evident between penetrating ability and
particle size, surface tension, or pH.  Most of the materials, however,
were characterized by low surface tensions in the 36 to 53 dynes/cm
range while particle sizes were usually in the intermediate range of
about 1500 to 2500 L

Naugatex J-3471 (a highly crosslinked styrene-butadiene copolymer) was
the most promising of this group for stability, penetrating ability,
and sealant efficiency.  It also showed utility with JB profile soil.

Many of the other latexes tested in this  investigation showed good
sealing efficiency but were judged unsatisfactory because they gave
low penetration or formed surface films with A horizon top soil.  In
general, latexes containing cationic surfactants were not satisfactory
because they  formed a heavy surface film.  This is due to neutraliza-
tion of the positive latex charge by the  negatively charged soil  par-
ticles.

Sealants for  B Horizon Soil

With reconstructed soil  columns, only a few materials gave appreciable
penetration of B horizon soil.  Most of the sealant  is believed  dis-
tributed through the column with only a fraction appearing as  free
latex.  Part  of the sealant was sometimes evident as a surface film,
presumably resulting from agglomeration of latex particles.  Latexes
showing penetrating and  sealing activity  were:  Goodrich  2679X,
Borden 2140,  Borden 2134, Paisley 71-9052-0,  Naugatuck 3471 and
Naugatuck R-8438-20C.  See Table XIII  for further details.

Latex Stability

Laboratory soil column experiments  demonstrated  the  need  for  greater
penetration of latex into  the  soil.   Indeed  if  the chemical  stability
of a latex is defensive,  surface  films  result.

Chemical  stability can be  determined  in one  of  two ways by adding a
coagulating  salt,  such as  calcium chloride or aluminum  sulfate.   One
way  is  to measure  how much salt  is  needed to  form a  bulk  precipitate;
                                   35

-------
                                             TABLE XI




                             Penetrating Sealants for A Horizon Soil




                                                                Latex Properties
Supplier
Naugatuck



Dow

Borden




Code
J1896
J1925
J2752
J3471
870
874
2134
2140
2153
2158
2635
Type
SBR-mod .
SBR -mod .
SBR-mod.
SBR
PVC
PVC-Vinylidene
Chloride
Vinyl Acetate (VA)
VA copolymer
Vinyl Acrylic
Vinyl Acetate
PVC Copolymer +
%
Latex
E luted
29
17
27
74
50
50
18
32
47
27
43
Particle Surface
Size Tension
A Dvne/cm.
1500
1400
2000
2400



2000
2000-3000
4000

53
64
37
36
36
36
42
40
42
44
37
- _EM_
10.6
9.6
8.9
9.8
8.
8.
3.5
7.2
5.
4-6
7.
Sealing
Efficiency
m
89
81
79
80
78
54
86
94
95
94
88
                           plasticizer
Goodrich
2679x6
Acrylic
33
45
8.8
79

-------
                                                   TABLE XII

                                     Penetrating Sealants for A Horizon Soil

                                                      Percolation
       Sealant

       Latexes
        Naugatuck
OJ
      Incremental Additions

      Naugatuck

      Naugatuck
        + TEPA Coagulant
        Dow
        Dow
        Borden
        Goodrich
        Acrylamide + Aram, persulfate
           R - re-used column.
1896
1925
1925
2752
2752
2768
2768
3471
3471
3471
3471
3471

3471
870
874
VA 2140
  2140
PVC 2635
VA 2134
VA 2153
VA 2158
2679X6


Lb/Acre
5500
5500
+3300 R
5500
+3300 R
5500
+3300 R
5500
+3300 R
5500
5500
5500
5500
4400
5500
5500
5500
5500
6600
8800
4400
4400
5500
4400
Rate •

Start
710
925
180
610
145
775
310
520
101
1610
530
1060

2240
1060
848
252
1040
1320
1240
1065
1240
228
721
- ft/vr
After
Sealing
77
180
26
145
52
310
153
104
41
14
42
5.5

177
230

88
64
164
173
56
78
48
218
Efficiency 70
of
Sealant
89
81
86
79
96
60
50
80
62
99
92
99

91
78
54
65
94
88
86
95
94
79
70
Latex
E luted
29
17
-
27
-
33
-
74
-
25
22
0

20
50
50
18
32
43
67
47
27
33
34
Soil
Source
2C
2C

2C

2C

2C

KB
KB
KB

03A
2C
2C
03A
2C
KB
KB
KB
KB
03A
2C
Surface
Film
Mod.
Low

Low

Low

Low

Mod.
Low
Mod.

Low
Low
Low
Low
Low
Some
No


No
No

-------
oo
                                                    TABLE XIII
                                        Penetrating Sealants for B_ Profile
Percolation
Rate - ft./yr.

Sealant
Latexes
Goodrich

Borden



Naugatuck



Paisley

Solutions of
Goodrich
+ Tergitol
Dow Montrek

NH4OH

Code
2679x6
2679x6
2140
2134
2134
2134
3471
3471
R-8438-20C
N32C03
71-9052-0

Polymers and plonomers
K714
15-S-12
12 |
600J
600

Sod. hexametaphosphate
Hydrazine


Lb . /Acre
3500
5500
4800
2750 R
1670
1670 R
1670
4400
44001
5400J
5400
4400

2800
31
4400
3100 R
4400
19400
Froze
4400
6600 R

Start
472
570
665
88
1010
91
12
55
1680
1720
1510

202

335
265
204
478
1
1260
1700
After
Sealing
15
25
10
2
91
20
1.4
11
3.8
3.4
94

16

265
28
138
18
3
10
3.0
Efficiency %
of Latex
Sealant
97
95
98
97
91
78
89
80
99
98
87

93

22
90
32
99.8
97
97
Soil
Eluted Source
9
16
13
73
0
0
0 Sk.
12
0 Sk.
0 Sk.
34
Inches
Penetrated
(1-2)





(1-2)
OA3
2C
OA3
KB
KB

KB
KB
KB
KB
KB





OA3
KB
KB
Surface
Film
Some
Some
Some
Low
Low

Low
Some
None
None
None








        Sk = skinned  (removed  top  half  inch of soil).
R
re-used column

-------
a second way is to add a very small amount of coagulant to a dilute
latex solution and measure the optical density at 700 rag, in a Gary
spectrophotometer to determine how much the particles have grown by
insipient coagulation.  Both these methods were used on a series of
latexes with and without added anionic or nonionic surfactants.

It was found that:

1)  The addition of surfactant before adding electrolyte increased
    chemical stability,

2)  Tergitol 15-S-12  (a nonionic) was more effective than Aquarex G
    and Nacconol 90 (anionics), and

3)  800 A particle size latex required more electrolyte than 400 A
    latex to cause instability.

These results confirm the behavior of latex in soil columns where the
highest elution has occurred when latex was stabilized with excess
Tergitol 15-S-12.

To further corroborate these observations with respect to reaction
with the soil, a series of experiments was carried out by shaking 10 g
air-dried Lanse soil  for 0.5 to 1.5 hours with 20 ml 2 1/2% latex by
itself and with 10 parts added surfactants.  Samples were then  settled
overnight, centrifuged for separation of  soil and coagulated latex,
and decanted for measurement of solids in the supernatant liquid.
The data show that stability of latex to  soil is increased by  the
addition of surfactants, particularly Tergitol.  (The latter is more
effective with 10 parts than with 5 parts per 100 parts rubber.)

Although this vigorous shaking of latex and soil in a bottle is a
more drastic condition than latex percolation, the results show that
the addition of surfactants to the latex  will reduce coagulation by
chemicals in Lanse soil,  i.e., increase the penetration of the  latex
into the soil.

Water  Soluble  Inorganics  as Soil  Sealants

Dilute  aqueous ammonium hydroxide  (2  1/27.) was found to be an  effective
soil sealant  (see Table XIV).  This effect was checked a number of
times  with both A and B profile  soil  from Lanse  or  the Research Center.

Use of 2-1/2  or 5% aqueous  sodium carbonate was  also found to  be a
very effective  sealant for  all classes of soil tested  (Table XIV).
The mechanism of  this sealant  action  is believed to  involve  swelling
resulting from montmorillonite type clays.
                                  39

-------
                              TABLE XIV

                 Inorganic  Chemicals as Soil  Sealants
Sealant  Lb/Acre
NH4OH

NH^OH
                   Percolation
                   Rate  -  ft/yr
                          After
                   Start  Sealing
4400
11000
11000
605
1580
1420
8
1.6
6
Efficiency
of
Sealant

99
99.9
99.5
                                              7o Latex
                                              Eluted
                                               1-2
                                               1-2
Soil
Source

03A
RC
RC
Surface
Film

None

None

None
Ammonium  carbonate  and  bicarbonate were  found  to be  ineffective.

The effectiveness of  ammonium  hydroxide  and  sodium carbonate as soil
sealants  is of  interest but  is of very limited value because of the
water  solubility of the reagents and  the resulting temporary nature
of the  seal.

Work by Agey °  reports  effective sealant activities with sodium car-
bonate  as well  as a variety  of lithium salts.  With a western  type of
soil, Agey did  not  obtain effective sealant  action with ammonium
hydroxide, however.  Letey^O has reported that soil and sand can be
made water repellent by treatment with ammonium hydroxide.  He also
noted  that a certain  type of humic acid produced by microbiological
action  was a very effective  sealant in the form of Fe^""*" or Al '~"H~
salts.  Phillips-*!  reports that mixtures of  sodium humate with alkali
metal  carbonates or polyphosphates are effective soil sealants.
            09
Work of PuriJ^  in India notes  that calcium saloids in clays are quan-
titatively converted to sodium saloids by action of sodium carbonate.
The fine clay particles produced fill up interstices and make  the
soils  impervious to water.   This technique was used on a 13 mile
length  of canals in India to reduce seepage very effectively.

jjlays as Sealants

Since commercial bentonite clays cost about  two cents per pound,
Montmorillonite BP  was  investigated as a sealant in very dilute (0.25
to 0.50%) dispersion and as  an extender in several latexes.  The clay
was kept suspended  by the addition of a surfactant.  Pore blockage
occurred, but most  seals were made at the surface and in no case was
penetration greater than two inches,  in laboratory soil columns.  It
was concluded that  no further work was warranted.
                                  40

-------
                          SECTION VIII

                   FIELD PERMEABILITY STUDIES

Natural Permeability

To determine the relative permeability of the soil a series of percola-
tion tests were performed at the three locations of interest at Lanse,
i.e., 3C, 03A and strip bank, and KA and KB at Kato.  Soil cores were
removed to 2, 8, 16, 24 and 32 inch depths with the thin wall sampler
tube, each core separated from the next by about fifteen feet.  Then
two-foot-long thin-wall tubes were inserted into the 2, 8, and 16 inch
deep holes and  three-foot long tubes were inserted into the 24 and 36
inch deep holes.  Each tube was tamped down about one inch further to
assure a bottom seal; the tubes were filled with water and the rate of
fall was measured.
                                                                      O £
The  test is  similar to those done for septic  tank percolation studies  ,
but  is not quantitatively related to Darcy's  coefficient of hydraulic
conductivity because the water can flow in three dimensions from  the
bottom of the tube.  The measurements are dependent on the length of
test, since  the rate of fall is a function of the head.  Therefore, it
should be noted that the 24  and 32 inch deep  tubes should give faster
percolation  rates because of the one foot greater head.  At locations
03A  and 3C the  data show a sharp decrease in  percolation at increasing
depth, Table XV.  However, the percolation at 03A and 32 inch depth is
much higher  than at 3C (0.2  vs. 0.01 inches/hour); the fragipan layer
must be deeper  at the 3C location since this  area does experience a
perched water table.  On the strip bank the rates of  falling head are
randomly variable,  appearing to be as much dependent  on  the particular
location of  insertion as on  depth of placement.  For  example, water
drained  from the 16 inch tube as  fast as  it was  added, probably  into  a
sub-surface  cavity  or crevice.

At Kato  the  tubes were more  difficult to  place because of  interference
of tree  roots.  For this reason  2  inch deep  tubes were not  inserted.
These  short  tests at Kato were  inconclusive,  varying  extensively  when
the  tubes were  moved.

Field  permeability  tests were  also  performed  in  two  (previously
described)  eight-inch diameter  permeater  tubes  at Lanse  and  one  at
Kato.  Because  the  soil  is  confined  inside  the  tube,  hydraulic  con-
ductivity  can be  calculated according to  Darcy's flow equation.

The  high values reported in Table XVI experienced  when the tubes  were
rewetted from a dry surface condition,  enable us to conclude that
 infiltration is faster  than the saturated permeability of the soil.
 In  contrast to  the  falling head percolation tests,  these permeability
results  do not  show a high variation between locations.
                                    41

-------
    5/19 -  5/21



    Location

    Lanse 3C
    Lanse  03A
•P-
S3
    Strip bank
    (Lanse)
                                                        TABLE XV


                                             Falling Head Percolation Tests
Depth:
Test *
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7
8
1
2
3
4
5
6
7

2"
Time
10
3
4
5
5
5
5
6
43
15
13
10
15
1


20
1
3
4
1


min.
min.
min.
min.
min.
min.
min.
min.
min.
min.
.5 min.
. min
min
hr


min.
hr
.1 hrs
.4 hrs
hr



Rate*
9.75
480
360
300
300
300
300
240
25.
34.
33.3
17.2
20.
13.


5.6
4.25
1.9
1.5
0.5


8"
Time
10 min.
14 min.
5.3 hrs
14 hrs
16.3 hrs
8.8 hrs
14.4 hrs
l.hr
44 min.
12 min.
12
10
15 min.
7.9 hrs
1 hr.

20 min.
1 hr
3.1 hrs.
4.4 hrs
15 hrs
1 hr


Rate
2.4
5.9
3.5
4.3
0.65
0.54
0.51
0.38
23.
27.5
19.8
9.4
6
2.7
2.25

10.5
8.5
5.26
5.5
1.23
3.13

16"
Time
24.3 hrs
15.1 hrs
5.3 hrs
2.6 hrs
16.3 hrs
8.8 hrs
14.4 hrs
l.hr
11 min.
10 min.
2.7 hrs
10 min.
15 min.
7.9 hrs
15.4 hrs
1 hr
drains

Rate
0.78
0.78
0.81
1.05
0.70
0.70
0.61
0.36
33.4
24.
6.
9.
6.
2.4
1.46
0.63

immediately




-





24"
Time
47 hrs
39.4 hrs
14.3 hrs
440 hrs




46 min.
24.6 hrs
16 hrs
3 hrs
17 hrs
7.9 hrs
15.4 hrs
1 hr
20 min.
20 min.
17.9 hrs
4.3 hrs
3.1 hrs
15 hrs
1 hr

Rate
0.01
0.019
0.017
0.0125




0.5
0.42
0.66
0.83
0.73
0.81
0.72
0.63
40.
7.1
1.9
3.2
3.7
2.0
3.63
32"
Time
47 hrs
39.5 hrs
14.3 hrs
440 hrs




41 min.
24.5 hrs
16 hrs
3.1 hrs
17 hrs
8 hrs
15.4 hrs
1 hr
20 min.
1 hr
3.9 hrs
4.3 hrs
15 hrs
1 hr


Rate
0.008
0.011
0.009
0.0054




0.18
0.14
0.18
0.20
0.26
0.38
0.43
0.38
11.2
8.25
2.80
2.60
1.80
2.88

                    * Rate is given in inches/hour.

-------
                      TABLE XV  (cont'd.)

Location
Lanse 3C


Lanse 03A


Kato KA

Kato KB

Depth:
Test #
1
2
3
1
2
3
1
2
1
2
JL!
Time
40 min.
10 min.
-
10 min.
20 min.
10 min.
2 min.
12 min.
10 min.
10 min.

Rate*
3.2
3.8
-
30
23
13
180
117
6
1.12
16"
Time
25 min.
16.2 hrs.
24.6 hrs.
20 min.
10 min.
24.2 hrs.
5 min.
10 min.
10 min.
45 min.

Rate
0.20
0.24
0.47
43
3.4
0.62
5.25
1.25
2.25
2.5
24"
Time
585 hrs.
16.7 hrs.
25.2 hrs.
2 hrs.
19 hrs.
24.3 hrs.
10 min.
42 min.
10 min.
23.2 hrs.

Rate
0.023
0.052
0.044
0.38
0.38
0.47
6
3.1
1.12
0.04
32"
Time
585 hrs.
16.7 hrs.
25.2 hrs.
2 hrs.
19 hrs.
24.3 hrs.
19.8 hrs.
—
11 min.
64 min.

Rate
0.009
0.026
0.029
0.25
0.35
0.53
0.025
-
25
3
*Rate is given in inches/hour.

-------
Date

 6/12
 7/8
 7/9
 9/4
 9/5
 9/19
10/21
10/22
10/23
11/7
11/8
11/20
11/21
                               TABLE XVI

         Darcy Hydraulic Conductivity (K)  of Field Permeameters
Time of Test
   (Hrs.)
    24
    17.
    24.
    15.
    24.
    24.
     3.
    24.
    24.
    19.
    24.
   288.
    24.
                        Lanse Site
K (3C)
(ft./yr.)
7.67
20.8
7.67
12.0
7.67
*56.9
*52.2
11.3
3.65
nil
nil
1.83
3.65
K (4B)
(ft./yr.)
17.2


29.6
11.3

*104.
19.0
3.65
4.75
3.65
2.56
3.65
                                                            Kato Site
                                                             K  (KA)
                                                            (ft./yr.)
                                                              38.3
      *dry
 Effect  of  Sealant Applied  in  the  Field

 Ammonium hydroxide, sodium carbonate, and Naugatex  J-3471  latex were
 found to be effective sealants  in laboratory  columns.  Field  tests  of
 these sealants vrere made by sprinkling dilute solutions of these  com-
 pounds  on  three  five-by-five  foot plots  separated about fifteen feet
 from each  other.  In each  case, directly after the  application of the
 sealant solution, a water  flush was sprinkled over  the area to abet
 penetration into the soil.

 Two or  three weeks later,  soil  cores were dug in 4-inch increments  to
 8 inch  and 16 inch depths, making two holes into which two-foot-long
 thin-wall  tubes were inserted.  Concurrently  8 inch and 16 inch holes
 were dug and thin-wall tubes were inserted into an  adjacent untreated
 area.   Subsequently, falling head percolation rates were measured.
 The results of the percolation  tests are shown in Table XVII.  The
 treatments appear to have  been  more effective at 16 inches deep than
 at 8 inches.  In the case  of the  ammonium hydroxide tests,  the effec-
 tiveness at the 16 inch depth was somewhat delayed  and short  lived.
Although the test time is  shorter for the sodium carbonate and latex
 treatments, the 16 inch deep reduction in percolation is significant.
 Comparing the untreated area with the treated area  results for
November 8 only, the percent reduction in percolation is 96.6% for
 the sodium carbonate and 98.9%  for the latex  treated area.
                                  44

-------
                                            TABLE XVII

                              Falling Head Percolation  Tests at Lanse
                             (Units: Time - minutes, Rate - in./hr.)
 Depth:


 8/1/69
10/21



10/22


10/23


11/6




11/7



11/8
NH^OH Treated
8"
Time
10
5
5
30
10
15
30
15
15
15
15
Rate
32.25
25.5
22.5
15.0
16.25
6
5
4
4.5
4.5
4.
16"
Time
5
5
15
15
15


15
15
15
15
Rate
12
12
7.38
9.5
9.5


8.5
8.
8.5
8.
Untreated
8"
Time
10
5
5
30
10
15
30
15
15
15
15
NH4OH Treated
16
15
15
-
.
-
-
-
-
-
8
10
10
.
-
-
-
-
-
-
-
-
-
60
66
74
.
-
.
_
-
-
-
125
113
108
.
-
-
-
-
-
-
-
-
-
60
240
-
.
1183
120
60
1457
92
30
30
60
60
.
-
1077
60
60
60
1301
48
72
60
0.625
0.375
-
.
0.2
0.25
0.375
0.19
0.16
0.50
2.5
0.63
0.25
.
-
0.38
0.75
0.75
0.75
0.9
1.9
2.0
2.0
15
15
15
-
.
-
-
-
-
-
10
10
10
40
60
-
-
-
-
-
-
-
-
Rate
43.5
37.5
36.25
23.5
35.5
21.5
20.75
22.5
22.
22.
21.5
16"
Time
5
5
15
15
15


15
15
15
15
Rate
186.
190.
80.
68.
57.


47-
43.
46.
41.
Untreated
47
41
42
-
-
-
.
-
-
.
20
18
18
17
16
-
-
-
-
-
-
-
-
15
15
29
30
30
90
60
30
30
30
10
10
10
40
60
15
15
30
60
60
60
60
-
58
54
35
34
23
13
10
17
19
18
26
20
20
21
18
18
19
16
14
8
8
8
-
8"
Time











Na
15
15
15
15
.
.
-
-
-
.
10
10
10
40
60
-
-
-
-
.
.
-
-
Rate











2co3
44
45
54
54
-
-
-
-
-
-
28
24
24
20
20
-
-
-
-
-
.
-
-
Time











16" 8" 16"
Rate Time Rate Time











Rate











Treated utex 34n
15
15
115
241
1188
120
60
30
30
30
30
60
60
-
-
1075
60
60
60
1480
60
60
60
0.5
0.5
0.25
0.25
0.15
0.12
0.25
0.50
0.75
0.75
0.13
0.25
0.03
-
-
0.20 15 14 60
0.19 15 12 60
0.19 30 12 60
0.38 60 10
0.26 55 4 1360
0.19 60 3.5 60
0.31 60 3.3 60
0.31 - - 60















0.38
0.25
0.13
-
0.12
0.06
0.06
0.12

-------
  Spoil Banks

  The  soil  aggregate system  that composes  the  spoil banks  is not defin-
  able because  of  the  intermixture of soil,  sub-soil  and bedrock com-
  ponents,  with resultant large voids.   If penetrating sealants are not
  effective, it may be practical to seal these restored stripped areas
  with a  surface seal.  Therefore, to test surface seals on the spoil
  bank at Lanse, five  solutions which had  previously  given effective
  surface seals in laboratory soil columns were applied in field tests.
  Five 8-inch diameter galvanized stove  pipe sections 8 inches long were
  pushed  into the  rocky soil 1 to 1 1/2  inches deep on the spoil bank in
  Location  4D.  The following sealants were  tested:

    GM-39:      800 ml. of  2 1/2% EPDM  latex  containing 40 parts of
                Aroclor 124210 per 100  parts  of rubber.

    GM-38:      800 ml. of  2 1/2% EPDM  latex  containing 40 parts of
                paraffin soap per 100 parts of rubber.

    PR-10:      800 ml. of  5% solution  of Phillips PR-10 sealant, a
                latex-oil mixture.^

    BP-clay:    3500 ml. of water containing  80 g. of montmorillonite
                BP clay and 40 g. of sodium-hexametaphosphate.

    Water
    Repellant:  250 ml. of water containing 2.5 g. of nonylphenoxyacetic
                acid dissolved in 20 g. of 28% NH4OH, followed by 250
                ml. of 0.17. HC1.

  These short tests indicate that PR-10  is a poor sealant,  GM-38 and
  GM-39 are intermediate and montmorillonite clay and nonylphenoxyacetic
  acid appear to have sealed well enough to be of further interest.
  See  Table XVIII.   The acid became more effective after it was allowed
  to dry.

                             TABLE XVIII
              Surface Seal Tests on Spoil Bank at Lanse
      [Time (in minutes) to Dryness of 500 ml Increment of Water]

                                    PR 10       BP-Clay       Repellant

                                       7          105            103
                                      62           -              61

10/23          6323             21
              35           31           6           30            100

11/7          28           33           2          112            110
              60           37           3           -

11/8          59          135           3          223            575
                                   46

-------
                              SECTION IX

                LATEX IRRIGATION AND EVALUATION AT LANSE
It was decided to demonstrate the use of latex as a soil sealant by
irrigating three small plots at Lanse coded 03A on the west of Brown's
run, 3C on the east of the stream, and SB,  an area on the spoil bank.
Earlier plans to sprinkle latex on twenty acres of the side were
abandoned because of easement difficulties.  On the basis of results
from laboratory soil columns and small plot experiments at Lanse,
Naugatex J-3471 was the latex of choice.  The irrigation system con-
sisted of plastic pipe (two inch size) for the main line from the
water meter to the irrigation area because of lower price and ease of
handling.  Because the municipal water supply was limited to 20 gpm,
low-volume 14V-LA-TNT Rainbow sprinklers were specified with low angle
20° 3/32 inch nozzles to reduce wind effects.  At the expected nozzle
pressure of 50 psig, each sprinkler covers a 56 foot diameter area at
a delivery rate of 1.64 gpm.  Each of the three test areas was covered
with  twelve sprinklers, six used with water alone as a  control area
and six used with latex and subsequent water spraying.  One-quarter-
turn  inlet valves were positioned on the main lines leading to each
area.  Flow to the treated and untreated areas was set  at 10 gpm by
Dole  flow control valves in the main entry lateral to each of  the  six
sprinklers.

To  avoid the use of dilution tanks, a Moyno pump  operated by a 1/2 HP,
1140  rpm 5:1 reducing drive was used  to deliver  1/2 gpm 48% latex  for
mixing with 10  gpm water.  A 3500 watt  gasoline  generator  provided
power for pump  and  light.  Latex  was  applied  at  night to take  advantage
of  the lower  evaporation and higher  humidity.

Neutron  gauge  access  tubes were  placed  in  orthogonal  arrangement in the
areas that were  irrigated, with  four  tubes in the section  that were
sprinkled with latex  and  four  tubes  in  the untreated  irrigation section
in  each  of  the three  locations.   Each tube,  five and  one-half  feet long
and sealed  at one  end with a plug welded  in place,  was inserted five
feet  into  the ground  in a  cored hole.

 In each  irrigation area,  the left section of six sprinklers spread the
dilute latex while the right section sprinkled the same amount of
water to serve as a control plot.  As shown in Figure 7 each sprinkler
 has a 56-foot sprinkling diameter at 50 psig design water pressure.
The sprinklers are 30 feet apart on the laterals and 35 feet apart
 between the laterals.  The access tubes (coded AT on Figure 7) were
 placed in comparable spots on the two sections of each area,  so that
 four soil moisture measurements on the two sections could be compared.
 Bouyoucos resistance blocks (BB) were also placed in areas 03A and 3C
 at one,  two and three foot depths, but not on the spoil bank area.
                                   47

-------
oo
                 00
                  •


                  •~J

                  I
              3   H
              (D   H.
              m  cw
              en
                  O
                  3
                  w
                  rt
                  ro
                  3
                  I-1'
                  O
                  H-
                  O
                  3
                  to
                  rt

                  X

                  Co
                                                      aatex
                                                      ump


                                              Fig. 7 - Irrigation system for application of  latex  at Lanse site

                                                       AT -  access tube;  BB - Bouyoucos block;  FCV -  flow control valve

-------
A service contract was arranged with Pennsylvania State University for
thirteen once-a-week neutron gauge measurements of. 25 access tubes at
4 depth intervals, extending from June 10 to September 10, 1970.  The
information was supplied as volume percent water content.

On June 17-18, 18-19, and 19-20 (1970) latex was applied to plots 03A,
3C and the spoil bank (SB), respectively.  Table XIX shows the time
schedule and the quantities of water and latex irrigation to each area.
Latex applications in Areas 03A and 3C were apparently successful, but
on the spoil bank appreciable puddling occurred with some latex run-off
into the control area.  A portion of the SB area had experienced prior
compaction and settling; this part of the area showed the greatest run-
off.

                              TABLE XIX

                 Latex Application Schedule at Lanse
Plot
Date

Pre-Irrigation
  Start
  Time
  Gallons of water

Latex Irrigation
  Start
  Time
  Gallons of water
  Gallons of Latex
     (47.8% T.S.)
  Latex (solids)
Post Irrigation
  Start
  End
  Time
  Gallons of water
Total water (gal.)
            (inches)
    03A
6/17-18/70
4:00 p.m.
5 hr.
4,645
9:00 p.m.
12 hr.
12,485
275

1,052*
12:00 N
 2:40 p.m.
2 hr. 40 rain.
2,230
19,370
1.59
    3C
                                   SB
6/18-19/70
4:10 p.m.
4 hr. 35 rain.
4,000
8:45 p.m.
9 hr.
9,700
275

1,052*
5:52 a.m.
2:05 p.m.
8 hr. 13 min.
7,200
20,900
1.71
6/19-20/70
9:15 p.m.
1 hr. 37 min.
1,430
10:52 p.m.
10 hr. 30 min.
10,990
275

1,052*
9:25 a.m.
3:30 p.m.
6 hr. 5 rain.
4,810
17,230
1.41
     *Equivalent to 4,650 Ibs/acre

The purpose of the water irrigation before latex application was to
provide a wetting front in the soil, whereas the subsequent post-
irrigation was needed to rinse the latex from the vegetation, and to
flush the latex deeper into the soil.  In addition, 5,000 gallons of
                                   49

-------
 irrigation water were applied on June 20-21 to Area  03A, June  22-23
 to 3C, and June 23-24 to SB to continue the downward movement  of mois-
 ture  into the soil.  Almost half an inch of rain  fell on June  21 which
 aided in keeping the ground surface high in moisture.

 There was no difference in the appearance of the vegetation  between  the
 latex-treated and control areas at any time after application.  The
 grass and legumes were thicker and greener than the  year before and  the
 weeds were higher, probably because of the wetter condition  provided by
 the irrigation.  No deleterious effects to the trees on the  spoil bank
 has been observed.

 Analysis of Soil Moisture Data

 It was anticipated that soil moisture measurements taken at  intervals
 of one foot depths would be a valid way to determine the effectiveness
 of the latex as a soil sealant.  Neutron gauge measurements  reported
 by Dr. L. T. Kardos, Environmental Scientist, Pennsylvania State
 University, as volume percent soil moisture were analyzed in two
 groups, the first including two weeks before and five weeks  after
 application, and the second involving data taken the subsequent seven
 week period.  The data was treated for analysis of variance  and tests
 of significance.

 The analysis divided the sources of variation into the following four
 factors:

  1.  Left versus right section of each plot.

  2.  Depth at which soil moisture was measured, that is 1,  2, 3 and
      4 feet.

  3.  a)  Week-to-week variation.

      b)  Alternatively,  pre-treated weeks versus post-treated weeks.

  4.  Access tube position.  There are four positions in each section.

 Each plot was analyzed separately for group 1 data as follows;

  Analysis A.  All 224 measurements were utilitzed (2 sections x 4
               depths x 7 weeks x 4 positions), in which factor 3b
               (2 levels) was considered.

  Analysis B.  Pretreatment data only were considered (2 sections x
               4 depths x 2 weeks x 4 positions) in  which factor 3a
               compares two weeks under untreated condition.

  Analysis C.  Post-treatment data only were considered (2 sections
               x 4 depths x 5 weeks x 4 positions) for the first five
               weeks after treatment.   Factor 1 above is now a measure
               of treatment significance.

From these three analyses,  variances of the main factors and their
                                  50

-------
interactions were calculated.

By comparing the difference based on the latex application to the error
variance (the sum of all variances not attributed to treatment), the
effect of treatment can be evaluated.  On this basis Analysis B deter-
mines that there is a difference between the sections in areas 03A and
3C prior to treatment but not for the spoil bank.  Analysis C deter-
mines that there is also a significant difference between the treated
and control sections after treatment for 3C and 03A areas but not for
the spoil bank.

Analysis A answers the question whether the difference between the
treated and control sections was significantly larger after treatment.
The answer is negative for all three plots based on analyses of vari-
ance.  However, the averaged delta value of the spoil bank (SB) after
treatment does show a significantly larger difference at the one-foot
level (Table XX, t = 5.88).  Figure 8 shows graphically that delta
moisture increased after latex application and remained high through
the next four weeks.

Table XX explains why Analysis A gives no treatment significance.
Although there are wide differences between the control and test sec-
tions after treatment, these same differences are evident prior to
treatment as well.  Table XX shows the soil mositure averages of the
four positions in each section prior to (two weeks), and after  (five
weeks) the latex was applied to the test area.  The standard deviation
of the means (S.D.) is also given for each set of values.  The  relative-
ly high standard deviations reflect the wide fluctuation of values  from
position more than from week-to-week variation.  The delta value is the
average volume percent soil moisture of the test area minus the control
for each depth.

In Figure 8 the delta values are plotted for each of the first  seven
times soil moisture was measured.  It can be seen that except for the
one-foot depth in area SB, the delta difference before and after latex
treatment is marginal.

The fact that the delta soil moisture decreases with depth at 3C but  is
a maximum at four feet in  area 03A indicates that soil moisture differ-
ences are characteristic of section differences.  The "Student-t-
Distribution" demonstrates that many of the values  show very low signif-
icance.

For the group 2 data, the  difference (delta) in soil moisture between
the average (4 positions)  of the  treated section and the average of the
control section for the three areas of  interest over the period of
July 30 to September  17, 1970, inclusive is shown in Figure  9.
Table XXI shows that  the soil moisture  and delta soil moisture  average
of the seven measurements  differ  little from the first  five post-
treatment data.
                                   51

-------
                              TABLE XX
Soil Moisture Data
Depth % Moisture % Moisture
(feet) Control* S.D. Test Area** S.D.
Plot
1
2
3
4
Plot
1
2
3
4
Plot
1
2
3
4
Plot
1
2
3
4
Plot
1
2
3
4
Plot
1
2
3
4
3C - Prior
23.67
33.85
35.45
33.22
3C - After
29.44
36.21
35.71
33.56
03A - Prior
26.27
28.71
26.01
25.09
03 A - After
to Treatment:
1.108
.973
1.056
2.117
Treatment:
.680
.480
.691
1.555
to Treatment:
.618
.319
.710
.529
Treatment:
30.20 .453
29.06
27.26
26.57
.448
.741
.408
SB (Spoil Bank) - Prior to
18.75
15.07
14.41
15.25
SB - After
17.86
16.98
15.02
15.73
.544
.737
1.120
.903
Treatment;
.143
.584
.852
.545

25.99
31.67
32.42
27.04

30.94
32.73
32.89
26.78

28.16
28.67
25.02
32.65

33.71
30.02
27.34
37.57
Treatment:
18.39
15.40
15.64
16.67

21.10
15.54
14.74
16.87

1.623
3.094
3.201
1.351

.899
1.716
1.759
.856

.025
.983
.434
2.473

.892
.683
1.020
1.906

.583
1.085
1.145
.983

.533
.824
.808
.755
Delta

2.31
-2.18
-3.02
-6.19

1.50
-3.48
-2.82
-6.78

1.89
-.04
-.99
7.56

3.51
.96
.08
11.00

-.36
.33
1.23
1.42

3.24
-1.45
-.28
1.13
S.D.

1.965
3.244
3.370
2.512

1.127
1.782
1.890
1.775

2.117
1.034
.832
2.529

1.000
.817
1.261
1.950

.797
1.311
1.602
1.335

.552
1.010
1.175
.931
t-dist
fDeltaN
\sT~J

1.17
0.67
0.90
2.46

1.33
1.95
1.49
3.82

0.89
0
1.19
3.00

3.51
1.18
0
5.65

0.45
0.25
0.77
1.06

5.88
1.44
0.24
1.21
*  Average of 4 positions, 2 weeks
** Average of 4 positions, 5 weeks
S.D. is the standard deviation.
                                     52

-------
      6/10
            6/17
6/24
                                            7/10
                  7/17
                                                                V
                                                                   Area 03A
                 After latex
                 application
         7/22 (1970)
      6/10
             6/17
6/24
         7/10
7/17     7/22 (1970)
H
2
o
eg
AJ
 4

 2

 0

-2 .

-4
                                                                    Spoil
                                                                    Bank
      6/10
            6/17
6/24
7/1      7/10
Test date
7/17
7/22 (1970)
           Fig.  8 - Effect of latex on delta soil moisture,  i.e.,  difference
                    In soil moisture between treated and control areas.
                                      53

-------
 O
 to
 a
 4J
 ft


 &
                                                                                Area 03A
      9/30
8/5
8/12     8/19
8/26     9/2
                    9/7   (1970)
       7/30
         8/12
         8/19
8/26
9/2
                                                                                Area 3C
                                                  9/9       9/7    (1970)
9
u
CO
•o1   °
CO

a   ~
u  -2
       7/30
8/5
8/12
         9/2
          9/9
                     8/19      8/26

                        Test date

Fig. 9 - Effect of latex on delta soil moisture, i.e., difference

         in soil moisture between treated and control areas
9/17   (1970)
                                           54

-------
                              TABLE XXI

                          Soil Moisture Data

              Seven weeks combined - 7/30 to 9/17 (1970)
Depth
(feet)
% Moisture % Moisture
Control Area S.D. Test Area
S.D.
Delta
S.D.
t-dist.
(Delta}
VS.D. 1
Plot 3-C
1
2
3
4
28
34
35
33
.31
.88
.04
.02
.653
.294
.444
1.278
29
32
31
26
.54
.37
.32
.15
.836
1,422
1.467
.656
1
-2
-3
-6
.23
.51
.72
.88
1
1
1
1
.061
.452
.532
.436
1.16
1.73
2.42
4.80
                              Plot 03-A
1
2
3
4
28
28
26
26
.91
.79
.47
.04
.552
.235
.517
.341
32
28
27
37
.03
.92
.14
.16
.783
.600
.855
1.359
3


11
.12
.14
.66
.12
.958
.644
.999
1.401
3.
0.
0.
7.
26
22
67
94
                               Plot SB
1
2
3
4
16.86
16.21
14.94
15.27
.197
.461
.636
.486
19.82
15.27
14.36
16.78
.453
.694
.632
.608
2.97
- .94
- .58
1.51
.494
.833
.897
.778
6.01
1.13
0.65
1.94
Because the results of the in-depth soil moisture were inconclusive, we
determined gravimetric soil moisture to a depth of one foot and com-
pared treated and control sections at areas 03A and 3C.  As in neutron
gauge moisture measurements, position variations were high.  Moisture
decreased with depth, except where the soil was saturated with free
water or was especially stony as noted in Table XXII.  In area 03A the
soil from the treated section was wetter than the control, whereas the
reverse was true at 3C (see average values and L-U values in Table XXII),
Evapotranspiration by the vegetation is probably not a factor because
it rained the night before the samples were taken.  Again, section
differences appear to be dominant.

Since soil moisture did not serve as a measure of sealing effectiveness,
permeability tests were done on several occasions.  Twelve 8-inch
diameter by 12-inch long stove pipes were pressed into the ground ten
inches, three in comparable positions in each section of Areas 03A
and 3C.  After sprinkling each area with 12,000 gallons of water,
                                   55

-------
                              TABLE XXII

     Gravimetric % Soil Mositure of Samples Taken August 31, 1970

                       Latex Treated  (L)          Untreated Control (U)
Depth, in.
03A



Av.
Av. A (L-U)
3C




Av.
Av. A (L-U)
0-4
38.2
41.8
33.6
28.8
35.6
5.6
22.4
26.9
25.8
28.2

25.8
-3.7
4-8
26.6
36. 51
30.0
26.2
29.8
2.8
17.4
23.1
23.0
20.5

21.0
-1.8
8-12
23.8
30.91
34. 31
20.7
27.4
5.0
10. 12
22.0
23.4
21.0

19.1
-2.6
0-4
32.4
29.1
31.2
27.4
30.0

27.3
33.3
28.7
31.3
26.6
29.5

4-8
26.4
26.5
27.2
28.0
27.0

26.2
30.4
20. 72
12. 72
23.9
22.8

8-12
22.6
23.0
21.0
22.8
22.4

23.8
26.8
22.7
16.3
19.0
21.7

      Free water in soil
     o
      Soil sample was stony
permeability tests were done on September 16, 1970 in the stove pipes
by adding 500 ml increments of water and measuring the time to dis-
appearance of free water from the tubes.  Specific permeability was
calculated by dividing the infiltration in cubic centimeters per
minute (500 cc/t (min.)) by the surface area (324 cm2) of the tube.
The permeability is synonomous with Darcy's hydraulic conductivity K,
cm/rain.  These numbers are recorded in Table XXIII as cc/min. cm , as
well as percent efficiency, as defined by the permeabilities of the
control minus the latex treated tubes divided by the control times 100.
The data demonstrate a consistent reduction of permeability in the
experimental sections of 90 to 99% efficiency,  even though there is
significant variability between tubes within each section.

On October 12,  1970 when the tests were repeated, the efficiency
remained high except for position 03A-1.  For some unexplained reason
the 03A-1 tube in the treated area became quite permeable while infil-
tration of the control at 03A-1 and 2 decreased appreciably.  The 03A
                                  56

-------
                                 TABLE XXIII
Location
03A-1



03A-2



03A-3


3C-1



3C-2



3C-3



Overall

Time
fain.)
65
245
120

>480



80
200
125
250



10
20
22
20
44
80
73
75
average
Latex Treated CL)
Permeability**
(cc. /min. -era?) Average
0.0238
0.0063
0.0128 0.0143

< 0.0032 < 0.0032



0:0193
0.0077
0.0123 0.0131
0.0062 0.0062



0.154
0.077
0.070
0.070 0.095
0.0350
0.0193
0.0212
0.0206 0.0240
0.0260

Time
fain.)
3
6
4
5
2
15
11
16
1
1
1
5
21
19
26
1
2
2
2
1
2
2
2

Control (C)
Permeability**
(cc./min .-cm?)
0.512
0.257
0.386
0.308
0.771
0.103
0.140
0.096
1.54
1.54
1.54
0.308
0.0735
0.0810
0.0594
1.54
0.771
0.771
0.771
1.54
0.771
0.771
0.771


Average Efficiency*


0.366 95. 7X



0.278 >98.8X


1.54 99. OX



0.130 95. OX



0.963 90. OX



0.963 97. 5X
0.707 96. 4X
*  Efficiency - ^ '  L
                  C
** Permeability
                                         57

-------
                                 TABLE XXIV


              Permeability Tests at Lanse on October 12. 1970
Latex Treated
Location
03A-1


03A-2 ,
03A-3


3C-1


3C-2


3C-3


Overall
(L)
Time Permeability** Time
(min.) (cc. /min. -cm.) Averaee (rain.)
7
12
13
contained
free J
water
158
176

165
204

26
46
36
21
35
45
average
0.220
0.129
0.119
nil
0.0098
0.0088

0.0093
0.0076

0.0594
0.0335
0.0429
0.0734
0.0440
0.0343

34
20
0.156 19
70
100
80
1
0.0093 1
1
6
0.0085 23
24
2
3
0.0453 2
2
2
0.0506 2
0.0450
Control (C)
Permeability**
(cc.Anin.-cm^) Average Efficiency*
0.0454
0.0771
0.0812
0.0220
0.0154
0.0193
1.54
1.54
1.54
0.2570
0.0670
0.0643
0.771
0.514
0.771
0.771
0.771
0.771



0.0679 negative
0.0189 ~ 1007.


1.54 99.37.


0.1294 93.27.


0.685 93.47.


0.771 93. 9%
0.535 91.57.
*  Efficiency •
                C - L
   Permeability • /;5^3 .
              '   t (min.)
                                         58

-------
treated section was obviously vetter than the adjacent control section
with small water puddles in low places and in tube 03A-2.  These data
are shown in Table XXIV.

To determine the permeability at ten inches depth the tubes were
removed from the ground with the enclosed soil and then reinserted in
the holes without the soil.  When this was done in holes 03A-2 and 3
water flowed into the holes from the thoroughly saturated soil.  We
may conclude from this that under saturated conditions the sealant
induces lateral flow at a higher level in the soil.  However, the
effect of the latex sealant on permeability from the ten-inch depth
down is much poorer than in the top ten  inches as the data in Table XXV
show.
                              TABLE  XXV

           Permeability Tests at Lanse on October 13,  1970
               (10 Inches  of  top soil removed from tubes)

        Latex  Treated  (L)	     	Control (C)	
                 Perm**                        Perm**
        Time     (cc/                  Time    (cc/                 Eff.*
Loc.    (min.)    min. cm )    Aye       (min.)    min. en/)     Ave      (%)
03A-1      6      0.257                   4      0.386
           9      0.172      0.215        11      0.140     0.263    18.3
03A     Free water  entered hole          10      0.154
        when soil was  excavated

                                         26      0.059     0.107
03A-3        Same as 03A-2            >233     >0.0066    0.0066

3C-1    >250     <0.0062                  6      0.257
                                         50      0.0309    0.144   >95.5

3C-2       20      0.0771                  6      0.257
          120      0.0151     0.0461      15      0.103     0.180    74.4
3C-3       34      0.0454                 16      0.0963
           57      0.0271     0.0363      85      0.0182    0.0573   36.7

Overall average             0.0759                        0.1263   39.8

                     C -  L
       * E
Efficiency
                        C
                         1  SAT
      ** Permeability =   '  ,\
                    3   t (min.)
                                   59

-------
                 TABLE XXVI



Permeability Testa  at Lanae on May 5. 1971
Latex Treated (L)
Location
03A-1


03A-2



03A-3



03A Average
3C-1




3C-2




3C-3




3C Average
Overall Average
* V wet
Time(t)
(min.)
46
85
93
20
27
24
26
10
33
26
28

2
6
8
9
10
7
14
17
18
16
6
12
16
18
21



Permeability
cc . /min . /car
0.0336
0.0182
0.0166
0.0772
0.0571
0.0644
0.0594
0.154
0.0468
0.0594
0.0551

0.772
0.257
0.193
0.172
0.154
0.221
0.110
0.0910
0.0858
0.0965
0.257
0.129
0.0965
0.0858
0.0735

0.1380
_ fC - L\ ,A,
Average Time
(min.)
35
106
0.0228 84
1
2
1
0.0645 1
5
6
4
0.0813 6
0.0562
1
3
4
4
0.310 3
1
3
3
3
0.121 2
1
2
2
3
0.1284 2
0.186

1 AA T»^_ 	 _UJ
Control (C)
Permeability*'
cc . /min . /car
0.0441
0.0146
0.0184
1.54
0.772
1.54
1.54
0.309
0.257
0.386
0.257

1.54
0.515
0.386
0.386
0.515
1.54
0.515
0.515
0.515
0.772
1.54
0.772
0.772
0.515
0.772

0.668
1.543

* Average I
Efficiency


0.0257 11



1.368 95



0.302 73
0.5652 91




0.668 54




0.771 84




0.874 85
0.771 75
80

                       60

-------
The permeability was still generally lower in the treated than in the
control areas in May 1971 (see Table XXVI).  For example in Area 03A
the average sealing efficiency for the three locations was 917e, while
in Area 3C the sealing efficiency was 75/i.

In Table XXVII we compare the permeability efficiency as a function of
time from September 1970 to May 1971.  The results indicate that seal-
ant qualities are retained but at a reduced efficiency after wintering.
The overall sealing efficiency for the 03A and 3C has been reduced from
96 to 80 percent from September 16, 1970  to May 5, 1971.  The reduc-
tion in sealing effectiveness may be caused by the action of freezing
and thawing, and/or by increased porosity of the soil by root growth
and the action of living organisms - insects, worms, etc.  It should
be recalled that most of the latex has been found within the top
twelve inches of soil.

                             TABLE XXVII
           Comparison of Permeability Efficiency with Time
Location
  03A-1
  03A-2
  03A-3

  3C-1
  3C-2
  3C-3

  Overall  average for
  03A and  3C Areas*            96                 92            80

     *Based on mean of all the permeability measurements


 This confirms our conclusion (previously drawn from soil moisture data)
 that area differences are at least as large as differences caused by
 latex  treatment.  Referring to Table XXVI,  it may be pointed out that
 one inch  of water (2.54 cm) at 0.0562 cc/min. cm* permeability requires
 7 1/2  hours to infiltrate, whereas at 0.186 cc/min. cm2 one inch of
 water  will infiltrate in 2 hours.

                                    2
 Calculation:   97
99
95
90
98
% Efficiency on
1970 October 12, 1970
negative
~100
99
93
93
94

May 5, 1971
11
95
73
54
84
85
                                   61

-------
                                   e\
              f-, or, 2.54 cc/min. cnr    ..,,-,.      0 ,      , ..
              @ 3C 	  =  137  mm.  » 2 hrs.,  17 mm.
                   0.186 cc/min.  cm2

The calculated infiltration times  are  shorter than the actual observed
infiltration rates of the rain because measurements in the stove pipes
are affected by wall and disturbance factors  as a result of placing
the pipes into the soil.
                                  62

-------
                              SECTION X

                      LATEX DISTRIBUTION IN SOIL
Methods
The optimum situation in sealing the soil with latex would be to
locate the sealant at the depth where natural permeability is lowest,
and to narrow distribution of the latex to a film.  To ascertain how
closely we were achieving these objectives it was necessary to devise
a method to measure the concentration of latex in soil.  A number of
methods were explored and two were used in actual field tests.

SBR latex tagged with C   was prepared and used  to determine the dis-
tribution of polymer in the soil.  The latex was made by copolymer-
izing a mixture of radioactive and normal styrene with butadiene^in
bottles by conventional methods.  Average particle size was 600 A.
Diluted aliquots had 1.40 microcuries radioactivity per gram of latex.
Tagged latex was mixed with soil in a range of proportions from 0.01
to 0.25% by weight to develop a calibration curve.  Analyses of soil
in the laboratory and from the field were determined by comparison
with this master curve.

Styrene-containing polymers have a characteristic peak at  700 cm"1 but
crosslinked SBR is not soluble in solvents normally used  in infra-red
spectroscopy, e.g., carbon disulfide.  We developed a method  for  solu-
bilizing the crosslinked polymer by stirring overnight at  room  temper-
ature a soil sample in chloroform with t-butyl hydroperoxide  and
osmium tetroxide.  The filtered chloroform residue was dried  to con-
stant weight, thoroughly mixed and pressed into  a KBr pellet, and
concentration was determined by measuring the  absorption  at 700 cm"1.
Details are given in an Appendix.

Results

Concurrent with the sprinkler  irrigation of  areas  03A  and 3C  at' Lanse,
C1^  tagged latex was applied  to  soil  confined  in eight-inch diameter
stove pipes twelve  inches  long which  were pressed  into  the ground ten
inches.  In another experiment at  locations  3A at Lanse,  and  KA and  KB
at Kato, a mixture  of  160  ml  of  tagged  latex and 480 ml  of regular
latex was applied  followed by a  comparable water addition.  Sample
cores were dug with a  one-half inch pipe  cover on July  14 and August 20,
1970 and on May 5,  1971.   Samples  were  oven-dried,  ground, sieved
through 40 mesh screen and the radioactivity was measured.  Similarly,
cores were dug at Kato on  October  14,  1970  and May 3,  1971.   Figure  10
is  a histogram of  the  results of the  tests  at Lanse.

Figure 10 shows  that 1)  the concentration decreased most in the first
month,  and  2)  it  decreased most  in the  top twelve inches.  Interest-
 ingly,  the  shape  of the histograms remained similar for each location.
For example,  all  three times of  test have the highest concentration
                                   63

-------
.05
                               Area 3C
          48     12    16    20    24
                              Area 3A
    0     4     8     12    16    20    24
         Soil core depth  (inches)
    Fig.  10 - Distribution of SBR in Lanse soil
I   I
                                                         July  14,  1970
                                                        Aug.  20,  1970


                                                    JNl May 5,  1971
                                64

-------
at 4-8 inches in the 3C area,  and at 0-4 inches in the 03A and 3A
areas.  The higher values on May 5, 1971 of 03A and 3A at 16-24 inches
may indicate some mobility of the SBR, although sensitivity of the
analysis at concentrations below 0.01% is not good enough to draw this
conclusion categorically.

The distribution of C1^ tagged J-1405 SBR at Kato as shown in
Table XXVIII indicates very little change between October 14, 1970 and
May 3, 1971 soil samples.

                             TABLE XXVIII

        Distribution of Tagged J-1405 SBR in Kato Soil (Wt. 7.)

Depth (in.)    KA      10/14/70   5/3/71         KB    10/14/70  5/3/71

0-4                     0.024      0.018                  0.027     0.023

4-8                     0.015      0.019                  0.045     0.043

8-12                    0.009      0.006                  0.016     0.012

12-16                   0.003      0.007                  0.004     0.010

16-20                   0.000      0.007                  0.000     0.006

20-24                              0.008                           0.006
 The  distribution of  different average particle diameter latexes
 applied in eight-inch diameter tubes at locations KA and KB in Kato
 was  investigated by  the infra-red analysis of oxidized residues.
 Latex B-2301-02 is 400 *  average particle diameter,  ANJ-27-100K latex
 is 800 *  and J-1405  is 1200 ft in particle size.  The objective of this
 test was  to determine the effect of latex particle size on the distri-
 bution of latex in  the soil.   The results are shown in Table XXIX as
 percent rubber in soil and as the percent of the total found to a
 depth of  twenty-four inches.   There is general similarity among the
 three latexes of different particle sizes.  However, there is sig-
 nificant difference  between locations KA and KB.  The highest con-
 centration at KA is  within the top four inches, whereas at KB the
 concentration is maximum at the four to eight inch depth.  We have
 reported that soil  texture at KB is coarser than at KA.  It may be
 concluded that soil  conditions have a major influence on how the
 latex is distributed within the soil.

 From the point of view of a material balance, these results are not
 quantitative, accounting for only 20-50% of the latex applied.  Since
 the results are obtained against standards done in triplicate, this
 discrepancy is explained by losses along the walls of the tubes and
 lateral dispersion from the bottom of  the tube walls.  The accuracy
 of the method is as good as the accuracy of the standards, which
 had less than ± 5% deviation.  The precision of the method is very
 good as demonstrated by comparing Samples J-1405-1 and  2 of  the table.
                                   65

-------
        TABLE XXIX
Percent Latex in Kato Soil
Depth (inches) 0-4
KA-J-1405
KA-J-1405 (dup.)
KB- J- 1405
KB-J-1405 (dup.)
KA-B2301-02
KB-B2301-02
KA-ANJ-27-100K
KB-ANJ-27-100K
.043
.043
.048

.032
.019
.027
.022
4-8 8-12 12-16 16-20
.023
.023
.125
.117
.017
.101
.022
.035
Percent of Total Latex
Depth J-1405-1
0-4 50.5
4-8 27.1
8-12 13.0
12-16 9.4
16-20
20-24
J-1405-2 B2301-02
50.5 49.2
28.3 26.2
11.8 7.7
9.4 1.5
9.2
6.2
.011 .008 nil
.010 .008
.017 .004 .002

.005 .001 .006
.009 .002 .004
.012 .008 .003
.013 .007 .003
for Each Depth
ANJ-27-100K J-1405
37
30
16
11
4
1
.0 24.2
.2 63.2
.4 8.6
.0 2.0
.1 1.0
.3 1.0
20-24
nil
-
.002

.004
.003
.001








X stones

B2301-02
13.8
73.1
6.5
1.5
2.9
2.2

ANJ-27-100K
27.5
43.8
16.3
8.7
3.7
stones

-------
                              SECTION XI

                         SOIL BOX EXPERIMENT
The extrapolation of results from reconstructed laboratory soil
columns to field conditions is difficult because of the limitations of
laboratory experiments.  Permeability and elution tests performed in
the laboratory differed from the natural state in that the soil was
sieved through 1/4 inch screen to remove large stones and roots, its
density was controlled by tamping the soil into two-inch I.D. tubes
instead of normal weathering, and water and latex were applied with a
constant head of liquid.  Tube wall effects are important in inter-
preting data.  Conversely, interpretation of field results were con-
founded by non-homogeneities within the sample tested, a soil profile
with gradation of permeability and density with depth, and uncontrolled
weather conditions and seasonal changes that affected the results.
Nevertheless, both laboratory and field tests showed that a  specific
latex did perform as a soil  sealant.

In view of these facts a  soil box experiment was designed to overcome
some of the disadvantages associated with the small laboratory  soil
columns and the field  experiments.  Two wood soil boxes 5 ft x  5  ft
x 5 ft were built in the  greenhouse at the Uniroyal Research Center.
They were lined with plastic PVC sheeting provided with tubes every
six inches of depth on one side of each box to measure water removal
rates under saturated  conditions.  Tensiometer gauges were placed at
depths of 6, 12, 18, 24,  36  and 48 inches and Bouyoucos blocs and
resistance thermometers were positioned at 6,  24  and  48 inch depths.
Each box was filled with  a uniform sub-soil obtained  from Ogdensburg,
N. J.  The texture of  this  soil was determined  to be  a  sandy loam by
the ASTM Hydrometer method,  as  shown below:

                  Texture of Ogdensburg, N. J.  Soil
                                            %          USDA

             Gravel  (+2 mm)                  11.6

             Sand  (-1-0.05 mm -2 mm)          48.0        54.1

             Silt  (+0.005  mm -0.05 mm)      30.4       34.2

             Clay  (-0.005  mm)               10.0       11.7

                USDA.  texture classification (without gravel)
                     is sandy loam

The  boxes were thoroughly wetted by sprinkling to abet settling and
 allowed  to dry by draining to check seepage  ports and functioning of
 tensiometers.
                                   67

-------
 In preparation for  applying latex in the  soil  box experiments,  twelve
 laboratory soil columns  of  Ogdensburg soil  were reconstructed to 1)
 test the  permeability  behavior of this particular soil  and 2) to test
 effects of different particle  size latexes  and levels and methods of
 adding surfactant.  These tests were carried out by first measuring
 water permeability  for two  weeks and then measuring permeability for an
 additional three weeks after applying 40  ml of 2 1/2% latex (1  g solids)
 The results are shown  in Table XXX.   The  data  show that:

 a.   Although the soil  column permeability was  stable for  four days
     before adding the  latex (the "before  latex" value is  an average
     of 4  days  results), continued wetting  caused further reduction
     in permeability (Columns 2,  6,  11).

 b.   The three latexes  which had no surfactant  added formed surface
     films  (Columns  1,  3, 10).   Surface films cannot be  observed
     with ANJ-27-100K latex  but their presence  is indicated by the
     non-wetting behavior of the surface when the first  few drops
     of water are applied to the column soil (for example  see
     Column 10).

 c.   Surface films were also observed in columns 4 and 5 where 10
     phr of Tergitol 15-S-12 was applied to  the soil  before the
     latex;  however, some latex was  eluted.

 d.  ANJ-27-100K latex  seems to become  less  effective in sealing
     efficiency  with passage of time  (Columns 7,  10,  12).

 e.  Of those  columns treated with  latex to  which 5  phr Tergitol
    was added,  only J-1405  latex was eluted through  the column
     (Column 8).  This  is probably due  to  the higher  stabilization
    of the  J-1405 latex.  It should  be noted that percent  reduc-
    tion in permeability is  based on the  63.9% latex remaining
     in the  column.

 Naugatex J-1405  latex  -t-  5 phr  Tergitol 15-S-12 surfactant  was selected
 for the soil box experiment because  it had  the best  balance of elution
 and permeability with  no surface film.  This latex was ordered from
 Uniroyal Chemical for  the experiment but was received as Naugatex
 J-2758, a change in designation  only of Naugatex J-1405.

On May 25,  1971, Soil  Box #1 was sprinkled  with eleven-and-one-half
 gallons of  2 1/2% solids J-2758  latex containing five phr  of  addi-
 tional Tergitol 15-S-12  surfactant.  This is equivalent to 4000  Ibs.
of rubber per acre.   Before the  latex was applied the soil was pre-
wetted with eight gallons of water and after application of the  latex
eighteen gallons of water was  sprinkled on  the  soil  in one gallon
 increments on the same date.  The latex was  readily washed in and  left
no surface residue.   Concurrently, twenty gallons of water were
sprinkled on Soil Box #2 as a control.  Table  XXXI is a record of  the
                                  68

-------
                                                                 TABLE XXX



                             Latex Experiments tn Laboratory Reconstructed Soil Columns  of Ogdensburg. H. J. Soil
VD
Columns t
Latex
Surfactant added
latex added (g)
Latex eluted (X)
Surface film
Before Latex
After Latex:
lat week
2nd week
3rd week
1
J-1405
no
0.8
0
evident
85

0.7
1.4
1.9
2
none
no
0


107

88
69
63
3
12301-
02
no
1.0
0
evident
83

0.5
0.8
0.6
4 5
J-1405 B2301-
02
t 10 phr ^
on aoil
1.0 1.0
36.4 34.9
evident evident
Permeability in
69 59

6.8 16
5.7 19
3.2 23
6
none
no
0


ColuB
78

64
53
46
ABU -27-
100K
10 phr
on aoil
1.0
41.6

ras
57

13
22
28
8
9
J-1405 B2301-
02
<_ 5
in
1.0
36.1

75

39
34
33
phr _^
latex
1.0
0

78

32
31
34
10
AHJ-27-
100K
no
1.0
0

82

1.2
6.8
17
11
none
no
0


97

76
63
56
12
ASJ-27-
100K
5 phr in
latex
1.0
0

75

2.6
1.9
21
            1 teduction
                          Before - After
lat week
2nd week
3rd week
99.1
97.5
97.9
17.8
35.5
41.1
99.4
99.0
99.2
90.2
91.7
95.3
73.0
67.9
61.0
18.0
32.1
41.0
77.2
61.4
50.9
58.0
54.7
56.0
59.0
60.3
56.5
98.5
91.9
79.2
21.6
35.0
42.2
96.4
97.5
72.0
                     * Columns are 2" in diameter and contain a 6" column of soil.

-------
                               TABU XXXI

         Soil Box Water Additions. Soil Tension and Resistance Data
Date
5/24
5/251
5/26
5/27
5/28
5/29
6/1
6/2
6/3
6/4
6/5
6/7
6/8
6/9
6/10
6/11
6/12
6/14
6/15
6/16
6/17
6/18
6/21
6/22
6/23
6/24
6/25
6/26
6/28
6/30
7/2
7/6
Water
Added Tens ioraeters
(gal.) 6"
8 72
18 0
7 0
9 0
6 0
1 0
0 0
0 3
0 5
0 11
13
19
31
47
59
64
622
60
60
60
59
58
8 O3
12 50
7 8
13 2
10 0
2
21
40
78
-
12"
38
41
2
0
0
0
2
4
4
8
8
10
11
15
16
19
33
35
41
43
50
54
68
60
13
6
3
5
9
12
22
72
18"
Box
30
31
3
0
0
0
3
5
6
8
7
10
10
12
10
10
15
16
16
20
25
26
40
26
10
5
4
5
8
8
13
50
(Centibars)
24"
#1
30
25
7
2
3
5
5
6
6
8
8
12
10
10
10
10
12
12
12
16
17
17
28
26
16
8
8
8
9
9
12
40
13
11
10
6
5
6
5
6
6
8
8
11
11
11
11
10
10
10
10
12
13
10
13
13
12
11
10
9
10
10
10
14
48"
12
11
11
10
10
10
10
10
10
10
11
12
12
12
12
11
11
11
11
13
14
11
12
12
11
10
10
10
11
11
11
11
Bouyoucos Block
(Resistance, ft)
930
620
660
690
690
-
700
680
680
-
-
630
620
630
640
650
690
750
790
-
760
820
820
770
740
640

600
630
640
820
2'
500
500
450
460
470
-
480
460
470
-
-
430
420
410
410
410
410
420
440
-
440
420
430
420
420
410

400
410
400
410
4'
360
360
360
360
360
-
360
360
360
-
-
340
330
330
330
330
330
340
340
-
340
340
340
330
330
320

320
320
320
320
1 - 11 1/2 gal.  of J-2758  latex was  sprinkled on Soil of Box #1 at 2 1/2% Total Solids
2 - At readings  between 65  and 85  tenslometers may begin to take in air, making
    subsequent readings lower .
3 - Refilled with water
                                        70

-------
                        TABL£ XXXI (Cont.)
Soil Box Water Additions, Soil Tension and
Date
5/24
5/25
5/26
5/27
5/28
5/29
6/1
6/2
6/3
6/4
6/5
6/7
6/8
6/9
6/10
6/11
6/12
6/14
6/15
6/16
6/17
6/18
6/21
6/22
6/23
6/24
6/25
6/26
6/28
6/30
7/2
7/6
Water Box #2 (Control)
Added T ens lometers (Centibars)
(gal.) 6J1
1
20 0
7 0
9 0
4 0
1 0
3
8
10
15
18
35
43
52
64
74
80
82
82
84
86
831
8 77
12 76
7 5
13 0
10 0
2
9
15
28
72
12"
-
16
6
3
2
0
3
7
6
10
10
14
24
15(?)
17
18
22
24
24
24
26
28
34
30
17
7
7
6
7
8
9
26
18"
-
13
10
4
2
3
5
7
7
10
10
12
13
12
13
13
15
16
16
17
20
18
22
22
19
14
7
8
8
8
10
15
24"
-
10
9
7
7
5
5
6
6
7
7
11
2A
-
-
105
12
12
11
12
14
12
5*
-
-
-
-
-
-
-
-
_
36"
-
13
11
12
11
10
9
11
11
12
13
16
15
14
14
13
14
14
13
15
16
14
14
14
15
14
14
14
14
14
14
4
Resistance Data
Bouyoucos Blocks
(Resistance, fl)
6"
960
590
610
600
610
-
620

660


610
640
680
720
750

940
1060
1110
1360

2040
1410
630
490
470

550
590
610
930
-21
460
460
460
450
450
-
450

440


440
400
400
400
400

400
400
420
410

400
400
390
390
390

360
370
370
370
Jtl
250
250
250
250
250
-
250

250


240
240
240
240
240

240
240
250
250

250
250
250
250
250

250
250
260
260
4 - Tensiometer water leaked out  of tube.
5 - Refilled Tensiometer.
                                    71

-------
amounts of water applied and  the responses of  the  tensiometers  and
Bouyoucos resistance blocks to  the  increase  in soil  moisture  for  each
box.  Water was sprinkled in  one gallon  increments  intermittently to
each box for several days to  May 29  to keep  the surface wet,  and  then
each box was allowed to dry.  No significant differences  in drying
rates nor water elution were  observed between  the  two  soil boxes.

A second wetting cycle was begun on  June  21 and concluded on  June 25,
a total of fifty gallons of water being added  to each  box.  Again dur-
ing the drying cycle the drying and  elution rates were similar.
Because of its location, Soil Box #1 received  at least an hour  more
exposure to the sun each day; this probably accounts for  the  somewhat
higher tensiometer readings from June 28  to July 7.  In Soil  Box  #2
the 48 inch tensiometer gave  no readings  because of  a  water leak  and
the 24 inch tensiometer gave  intermittently spurious results.

Permeability tests were performed in the  soil  boxes  by means  of eight-
inch diameter, twelve inch long, stove pipes pressed into the soil
eight inches deep.  These tests, shown in Table XXXII, indicate that
the latex did not act as a sealant.  On the other hand, in laboratory
soil column tests with this soil, J-2758  latex did show sealing action.
The explanation of these anomalous results may be that the drying
cycles caused shrinkage that  opened  the latex-containing  soil to  its
natural porosity.  High daily temperatures in  June caused rapid and
severe drying of the soil at  and near the surface.

On June 23 samples of soil from Soil Box  #1 were cored in four-inch
sections to 36 inches depth, were oven dried,  sieved through  a  40 mesh
sieve and oxidized, and analyzed for styrene-butadiene rubber (SBR).
Ninety-two percent of the SBR (0.107% in  soil)  was found  in the top
four inches, eight percent (0.009% in soil) was found  in  the  four-to-
eight inch depth; none was detected below this  depth.  The Ogdensburg
soil used in this experiment  is a rather  open  soil, characterized as a
sandy loam,  so penetration was expected to be  better,  i.e. deeper,
than the above values.   The poorer penetration  of the  latex in  the soil
box as compared to field application at Lanse  suggests that much  of the
latex in the field penetrated into the soil by  means of macro-cracks,
worm holes,  etc.,  and not through the capillaries of the  soil.  The
lack of such macro-paths in the soil box  and the method of packing the
soil in the box (tamping every two to three inches of  soil with a five-
foot long 2x4 board)  evidently restricted the downward  flow of  latex.
                                  72

-------
                               TABLE XXXII
Date
July 16/?7/
July 19
 July  22
 July 23
Soil Box Permeability Tests
Box #1
Locat ion
1


1




2





3




Time
(min.)
6
7
8
8
8
8
8
9
5
5
5
5
6
5.5
10
11
11
10
10.5
Permeability
(cc./min./cm*)
0.257
0.221
0.193
0.193
0.193
0.193
0.193
0.163
0.308
0.308
0.308
0.308
0.258
0.280
0.154
0.141
0.141
0.154
0.147
                                                       Box t2 (Control)
Time
(min.)
8
8
9
9
10
10.5
10
10.5
9
10
10
10
10
12
5
8
8
8
8
Permeability
(cc./min./cnr)
0.193
0.193
0.163
0.163
0.154
0.147
0.154
0.147
0.163
0.154
0.154
0.154
0.154
0.129
0.308
0.193
0.193
0.193
0.193
      Average
7.7
0.216
9.1
                                                                 0.174
   500 cc. water was placed on soil in 8" diameter stovepipe 12" long
   pressed into soil 8" deep, and time recorded to disappearance of
   water.
                                     73

-------
                             SECTION XII

                                COSTS


Table XXXIII itemizes the cost associated with the installation of an
irrigation system and the application of latex and water in the three
experimental areas at Lanse.  There were no external labor charges (non-
Research Center) except for the installation of the water meter.  Cost
of pipe and fitting are high because long lines were required at each
of the three different locations; these costs were minimized by using
plastic instead of metal pipe.  Fifteen of the seventeen drums of latex
received were used, five drums for each area.

                             TABLE XXXIII

                      Irrigation Cost at Lanse
Equipment;
  Moyno pump                                                $  355.00
  Moyno pump (lease spare)  15% of purchase price                53.25
  Pipe and  fittings                                            900.00
  Flow control valves                                           22.00
  Sprinkler heads                                               156.60

  Garden  hose  (3/4" heavy)  300  ft                               75.60
                                                              1,562.45

 Raw Materials:

  Latex  J-3471,  3500 Ibs  @ $0.34/lb  (dry basis)             1,190.00
  Water,  75,000  gal.  -  June 1  to July 1                          42.50

 Other Charges

  Water  meter installation                                       70.71
  Deposit on water meter   $85.00
  U-Haul rental  trailer                                         26.26

  Neutron gauge  measurements 6/10-9/10 by
      P.S.U.  personnel                                        1,836.00


 In  tank-car lots latex would cost about $0.25/lb  so that raw material
 cost for a 4000 Ib /acre  application would be in the range of $1,000
 per acre.  Equipment costs for setting up and irrigation are sensitive
 to  size of area to be irrigated, design of system, and availability
 and cost of water.  A realistic range is estimated to be between
 $200-500/acre.
                                    75

-------
                             SECTION  XIII

                           ACKNOWLEDGMENTS
This report was prepared by Jacob Tolsma  and  Arnold N. Johnson of the
Uniroyal Research Center, Wayne,  New Jersey.

The valuable assistance and suggestions of the project officers,
Mr. Ronald D. Hill, Dr. David R.  Maneval  and  Mr.  John J. Buscavage
is gratefully acknowledged.

The authors are grateful to the following Uniroyal personnel:
Mr. Daniel Shichman, manager of Engineering Research, and  Dr. Emmanuel
G. Kontos, manager of the Polymer Physics Research for their  direction,
encouragement and suggestions regarding the experimental work,  and  Mr.
Mark Olson for the design of the hollow auger soil sampler.

We also wish to gratefully acknowledge the efforts and contributions
of  ideas proposed by our  consultants, Dr. Frank T. Caruccio,  assistant
professor  of geology at  State University College, New Paltz,  New York,
and Dr. Pa Ho Hsu, associate professor of soil chemistry,  Rutgers
University, New Brunswick, New Jersey.
                                   77

-------
                              SECTION XIV

                              REFERENCES

1.  ''Abandoned Pits Problem: Mine Acid  Clean-Steam Block", Pittsburgh
    Press, Harrisburg Bureau 2/30/69.

2.  "Chemical Conditioners Upgrade Soil  Quality", News Feature Chem.
    Eng., October 21, 1968 p. 66, 68, 70.

3.  Trade Bulletin, "Soil Stabilization",  International  Synthetic
    Rubber Co., Ltd., England.

4.  Phillips  Petroset Emulsions.  Rubber Age,  July 1969  p. 118.

5.  W. T. Gooding,  et al., "Soil  Sealing Chemicals and Techniques",
    U. S. Dept. of  Interior, Res. and Dev.  Report #381,  p. 19.

6.  L. M. Ellsperman, W.  R. Morrison, "Asphaltic Membranes for Water
    Seepage  Control", ACS Symposium  on  New Uses for  Asphalt,  Atlantic
    City, Preprint  JL3 #4.  Division  of  Petroleum Chemistry pp.  C133-163.

7.  C. M. Hansen  "The Use of Asphalt to Increase Water Holding Capacity
    of Droughty Soils",  Ibid  pp.  Cl64-169.

 8.  C. W. Hayden  and Wm.  H. Heineman,  "A Hand-Operated  Undisturbed Core
    Sampler",  Soil  Science ^06  #2,153-6 (1968).

 9.  Supplied by Pfaltz  & Bauer, Inc., 126-04 Northern Blvd.,  Flushing,
    New  York.

10.  Mention of commercial products  does not imply endorsement by the
     Federal Water Pollution Control  Administration.

11.   Perma-Flex CMC Blak-Tufy,  The Perma-Flex Mold Company,
     1919 E.  Livingston Ave.,  Columbus,  Ohio  43209.

12.   H. L. Dalis,  Inc.,  Long Island City, New York
     Size FT-221-2 for 1-7/8 inch diameter cores  (thin wall samples)
     Size FT-221-3 for 2-3/4 inch diameter cores  (hollow auger)

13.   J. A. Bondurant.   Soil Science 107, No. 1,  70 January (1969).

14.   Dow Chemical Company, Midland,  Michigan.

15.   Procedures for Testing Soils, 4th  ed. Dec.  1964.  Published by
     Amer. Soc. for Testing Materials Committee D18,  Phila.,  Pa. p 76.

16.  Ibid. p. 95.

17.  Ibid. p. 92.
                                      79

-------
 18.   Procedures  for Testing Soils,  4th ed.  Dec.  1964.   Published by
      Amer.  Soc.  for Testing Materials Committee D18,  Phila.,  Pa. p. 107.

 19.   Ibid.  p.  109.

 20.   Ibid.  p.  85.

 21.   Firman E. Bear,  "Chemistry of the Soil",  Reinhold Publ.  Company
      (1964)  p. 497.

 22.   Helmut Kohnke,  Soil  Physics,  McGraw Hill  (1968)  p.  129-131.

 23.   W.  T.  Gooding,  Feasibility Study of Chemical Sealing of  Soils,
      Research  and Development  progress Report  No. 226,  U.  S.  Dept.
      of  the  Interior, June  1967.

 24.   A polyvinyl methyl ether  supplied by General Aniline & Film as
      a viscous 50%  aqueous  solution.

 25.   B.  A.  Hunter,  and B. Von  Schmeling,  U.  S.  Patent  No.  3,411,939
      to  Uniroyal, Inc., Nov. 19,  1968.

 26.   Rubber  Age, July 1969,  pp.  118.

 27.   R.  H.  Karol, Soils and Soil  Engineering,  Prentice-Hall,  Inc.
      ch  17  (1960).

 28.   E.  Higashimura  et al.,  U.  S.  Patent 3,417,567, Ex.  5  Mitsubishi
      Rayon  Co., Ltd., Dec.  24,  1968.

 29.   W.  W. Agey, Reduction  of  Seepage Losses from Canals by Chemical
      Sealants, Report #7584, U. S.  Dept.  of  Interior,  Bu.  of  Mines
      (1965).

 30.   J. Letey, Soil Sc. Soc. Amer.  Proc., 33 No.  1 149  (Jan.,  Feb.  1969)
      J.  Letey, Soil Sc. 93  149-153  (1962).

 31.   K. G. Phillips et al., U. S. Patent  No. 3,379,014  to  Nalco  Chemical
      Co., April 23,  1968.

 32.  A. N. Puri,  Soils, Their Physics  and Chemistry, Reinhold  Publish-
      ing Corp., New York,  N. Y. p.  94-99  (1949).

33.  J. F. Parr,  Soil Science 107 No.  2 94 (Feb.  1969).

34.  J. B. Decoste,  Ind. Eng. Chem. Prod. Res. Development, _7  (4)
      (Dec. 1968).

35.  Yasunori Nishijima,  Reports on Progress in Polymer  Physics  in
     Japan VIII 139  (1965).

36.  David E. Hill,  Percolation Testing for Septic Tank  Drainage,
     Bulletin #678,  Connecticut Agricultural Expt. Sta., New Haven  (1966).

                                    80

-------
                              SECTION XV

                               GLOSSARY

A Horizon - the surface layer of a mineral soil having maximum biolog-
     ical activation; commonly referred to as topsoil.

A Profile - the topsoil layer (synonymous with A Horizon)

Atterberg limits - a measure of the workability or consistency of the
     soil as affected by the water content.  The limits are defined by
     the water contents required to produce specified degrees of con-
     sistency that are measured in the  laboratory  (see liquid limit,
     plastic limit and plastic index).

B Horizon - the subsoil below the A Horizon topsoil.

Bouyoucos block - an  instrument for measuring  the  water content of  soil
     based on dielectic conductivity.

B Profile - the subsoil  (synonymous with  B Horizon).

Cation exchange capacity  (CEC) - the  sum   of  the  chemically exchange-
     able cations of  a  soil.

Darcy  Value  (K) - a  coefficient of  permeability to correlate effects  of
     column height,  head  and  diameter
                                 K -
                                 K   hA

      where K «• hydraulic conductivity (ft/day)
            L = length of soil column (ft)
            h = effective head (ft)
            A = cross sectional area (ft^)
            Q = quantity discharged (ft^/day)

 Fragipan - a type of subsurface soil structure having a relatively high
      bulk density.  It is cement-like when dry and very slowly perme-
      able to water.

 Latex - an aqueous dispersion of finely divided rubber or plastic
      materials of medium to high molecular weight.  Stability is con-
      trolled by use of surfactants, typically a mixture of anionic and
      non-ionic materials.

 Liquid limit - the water content  (%) at which soil becomes semi-fluid,
      like softened butter, as measured by a  standard ASTM procedure.
                                     81

-------
                           GLOSSARY (Contd.)

 Mariott feed bottle -  a device for maintaining a constant liquid level
      above a soil  column during percolation  testing.

 Montraorillonite -  a specific  type of clay soil (hydrous aluminum
      silicate).

 Permeability -  a measure of the readiness with which  the soil  permits
      the passage of water through a unit cross-section.

 Permeameter tube - a thin-walled open-ended  tube used to perform in
      situ measurements of soil permeability.   The tube is pressed
      into the soil for part of its length, a  measured amount of
      water is placed in the tube and the time required for the
      disappearance of  this water is taken as  a measure of soil
      permeability.   Eight-inch-diameter,  twelve-inch-long stove
      pipes,  inserted eight inches into  the soil,  were used as
      permeameter tubes for much of the  present work.

 Piezometer tube  -  a tube placed in the  ground to establish the
      location of the water table.   An open-ended tube is driven into
      the ground, then  withdrawn,  the soil core removed,  and the empty
      tube reinserted in the hole.

 Plastic  index -  the difference between  liquid and plastic limits.   It
      gives an indication of the "clayeyness"  or plasticity of  a soil
      and is  widely used in engineering  classification for soils.

 Plastic  limit -  the water  content (70) at which soil begins to  crumble
      on  being rolled into  a thread 1/8  inch in diameter.   It represents
      the lowest  water  content  at which  soil can be deformed readily
      without cracking.

 Porosity -  the ratio of volume of voids  to the total  volume of a given
      mass  of soil.

 PVC - a  homopolymer of  vinyl chloride.

 SBR - a  rubber copolymer of styrene  and  butadiene.

 Surfactant -  a chemical  which  will  reduce the surface tension  of
      aqueous  solutions.  There are  three  classes,  namely,  anionic,
      cationic, and  non-ionic.

Tensiometer  - an instrument combining a manometer  and a  porous  membrane
      for measuring  soil  water  suction.

VA -  a homopolymer  of vinyl acetate.
                                  82

-------
                               APPENDIX

   ANALYTICAL METHOD TO DETERMINE  STYRENE -BUTADIENE  RUBBER IN  SOIL


Sample Preparation

Soil cores are cut in the field in 4-inch depth increments, dried,
ground and sieved through a 40 mesh screen and weighed.

Polymer Oxidation and Separation from Soil

Crosslinked styrene -butadiene rubber (SBR) is insoluble in most sol-
vents.  The purpose of oxidizing the SBR in the soil is to render it
soluble in chloroform.  Twenty grams of soil containing SBR is weighed
into a 125 ml  erlenmeyer flask.  Fifty milliliters (ml ) of chloroform,
15 ml  of t-butyl hydroperoxide, and 2.5 ml  of 17. osmium tetroxide in
benzene are added.  A one -inch magnetic stirrer is placed in the flask,
the flask is stoppered and the contents are stirred overnight on a mag-
netic stirrer at room temperature.  Overnight stirring has proved ade-
quate to give reproducible results .

The oxidized chloroform soluble portion  is separated from  the soil by
filtering through a #1 Whatman filter  into a 500 ml  vacuum flask.  The
soil  is washed four  or  five  times with additional chloroform.  The
chloroform filtrate  is  air dried by evaporating the chloroform in a
small beaker  in  a hood  and subsequently  dried to constant weight in a
vacuum dessicator.

Concurrent with  the  analysis of samples  of unknown  SBR concentration,
a  known amount  (for  this work 15 mg  was  used)  of the  same SBR in 20 g
of a  soilblank should  be run as a  standard in  triplicate.

Infrared  Analysis

A  portion of  the oxidized,  thoroughly  dried  residue is ground with  and
pressed  into  a KBr  pellet  and the  characteristic  peak  at 700  cm"1  is
measured.  The  equivalent  optical  density (O.D.)  of the  standard  is
measured  and  a  specific O.D. per  mg   is  calculated.

Then  the  mg.  of  SBR in the sample soil is calculated  as  follows:

 (O.D.  of sample) (total MR  residue)  + O.D^ Qf standard  = mg   of SBR
        mg  of sample in pellet         mg                 £  sample

              ma  of SBR in sample       =
                  mg  of soil
                                   83

-------
Comments

It has been found that a very dark or black residue indicates that the
oxidation has been too severe and invariably gives low results.  It is
therefore important to do the drying at room temperature as well as
the oxidations.

The soil oxidation products contribute to the residue weight at 0-4"
and 4-8" depths, but do not appear to influence the optical density
of the styrene peak at 700 cm"*-.

The oxidized residue is partially insoluble in carbon disulfide, making
necessary the imbibation of solid residue into a well mixed KBr pellet,
rather than the  less tedious method of infrared analysis in solution.
                                  84

-------
1

5
,4 cr r.s.s ion Number
2

.Snbj'fc / I- it-lit & Group
0 5G
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Uniroyal, Inc.
              Research Center
              Wayne, New  Jersey  07470
   Title
        Use of Latex as  a  Soil Sealant to Control Acid  Mine Drainage
1 0 Aulhorfs)
Tolsma, Jacob
Johnson, Arnold N.
1
-------