'
BEB
United States
Environmental Protection
Agency
Region VIII
1860 Lincoln Street
Denver, Colorado 80295
Solid Waste
-8-EPA A TECHNICAL
PROGRAM REPORT
A REVIEW OF METHODS FOR DETERMINING THE
PERMEABILITY CHARACTERISTICS OF COLORADO
SOILS EXPOSED TO HAZARDOUS WASTE
-------
A TECHNICAL ASSISTANCE PANELS PROGRAM REPORT:
A REVIEW OF METHODS FOR DETERMINING THE PERMEABILITY
CHARACTERISTICS OF COLORADO SOILS EXPOSED
TO HAZARDOUS WASTE
Prepared For:
U.S. Environmental Protection Agency
Region VIII
1860 Lincoln Street
Denver, Colorado 80295
Prepared By:
Fred C. Hart Associates, Inc.
Market Center
1320 17th Street
Denver, Colorado 80202
June, 1981
-------
A REVIEW OF METHODS FOR DETERMINING THE PERMEABILITY
CHARACTERISTICS OF COLORADO SOILS EXPOSED TO HAZARDOUS WASTE
ENVIRONMENTAL PROTECTION AGENCY REGION VIII
Grand Forks
• Great Falls
• BISMARCK
Missoula
• HELENA
• Miles City
• Butte
Billings
Aberdeen
• Sherdian
PIERRE
• Rapid City
Sioux Falls •
Rock Springs
CHEYENNE
• SALT LAKE CITY
• Ft. Collins
• DENVER
• Grand Junction
Green River 9
• Pueblo
i
-------
Public Law 94-580 - October 21, 1976
Technical assistance by personnel teams. 42 USC 6913
RESOURCE RECOVERY AND CONSERVATION PANELS
SEC. 2003. The Administrator shall provide teams of personnel, including
Federal, State, and local employees or contractors (hereinafter referred to as
"Resource Conservation and Recovery Panels") to provide States and local gov-
ernments upon request with technical assistance on solid waste management,
resource recovery, and resource conservation. Such teams shall include techni-
cal, marketing, financial, and institutional specialists, and the services of
such teams shall be provided without charge to States or local governments.
This report has been reviewed by the Project
Officer, EPA, and approved for publication.
Approval does not signify that the contents
necessarily reflect the views and policies of
the Environmental Protection Agency, nor does
mention of trade names or commercial products
constitute endorsement or recommendation for
use.
Project Officer: William Rothenmeyer
-------
TABLE OF CONTENTS
Page
List of Tables iv
List of Figures v
Acknowledgements v"i
I. INTRODUCTION 1
II. IMPACT OF CHEMICALS ON CLAY LINER PERMEABILITY -
REVIEW OF STUDIES 4
A. Background • 4
B. Brief Characterization of Earthen Materials Used as Liners 7
C. Description of Major Studies Identified in the Literature 10
D. Summary and Evaluation of the Studies 27
III. CHARACTERISTICS OF SELECTED COLORADO CLAYSTONES AND
CLAYEY SOILS 29
A. Physical Properties of Clays Relevant to Their
Suitability as Liners 29
B. Front Range Rock Units Suitable as Clay Sources 33
IV. STATE OF COLORADO'S LIST OF PROPOSED CHEMICALS 41
I
V. TEST METHODS FOR EVALUATING COLORADO'S CLAYEY SOILS 45
A. Description of the Tests Identified in the Literature 45
B. Recommendations 48
C. Summary and Conclusions 51
APPENDIX A Coefficients of Permeability for Liquids and Liquid
Mixtures in Contact with Clays 53
APPENDIX B Test Method for the Permeability of Compacted Clayey
Soils (Constant Elevated Pressure Method) 56
REFERENCES 64
-------
LIST OF TABLES
Table Number Title Page Number
1 Variability of Engineering Properties
for the Pierre Formation 35
2 Variability of the Engineering
Properties for the Dawson and
Denver Formations 40
3 Dielectric Constant, Density, and
Viscosity Values Of Selected
Organic Solvents 43
i v
-------
LIST OF FIGURES
Figure Number Title Page Number
1 Unified Soil Classification System 5
2 Lee's Pressure Permeability Column 12
3 Matrecon's Exposure Cell for Soil Liners 24
4 Generalized Geologic Map of the Front Range
Urban Corridor, Colorado 30
5 Composite Strati graphic Section, Front Range
Urban Corridor, Colorado 31
A-l Lee's Pressure Permeability Column 54
B-l Permeameter Chamber Used by MERL 58
v
-------
ACKNOWLEDGEMENTS
This report was written by Marc A. Jewett and M. Raymond Buyce of Fred C.
Hart Associates. Other personnel providing assistance, review, and comment
included John J. Gaudette, Burke P. Lokey, Stephen J. Orzynski, and Alexandra
P. Wright. The assistance and helpful suggestions provided by William Rothen-
meyer and other Environmental Protection Agency, Region VIII personnel is
greatly appreciated. Dr..Kurt W. Brown at Texas A&M University, Dr. G. Fred
Lee at Colorado State University, and Dr. G.W. Gee at Battelle-Pacific North-
west Laboratory all contributed helpful guidance and assistance throughout the
course of the project.
vi
-------
I. INTRODUCTION
During the past several years, interest in the proper land disposal of
hazardous wastes has intensified. Numerous instances of groundwater
contamination resulting from poor disposal practices have stimulated renewed
efforts to properly solve the hazardous waste disposal problem. The Resource
Conservation and Recovery Act (RCRA) of 1976 has set forth, as a primary goal,
the elimination of threats to health resulting from the uncontrolled dumping of
hazardous wastes. Through RCRA, the Environmental Protection Agency (EPA) has
been assigned the task of establishing standards applicable to owners and
operators of facilities which are used to treat, store, or dispose of hazardous
waste.
One form of waste management which has received increased attention within
the regulatory climate involves the impoundment of hazardous wastes in earthen
lagoons or landfills. The use of protective, impermeable liners within the
impoundments is generally considered an acceptable means of minimizing the
potential for groundwater contamination. The liners can be used to minimize
both the seepage of leachate and other liquid waste components from the
impoundment, as well as to prevent the migration of methane and other gases
generated within the disposal area.
A wide range of materials for the containment of wastes is available from
which to choose. Candidate liner materials which have either been utilized in
the past or else are under consideration for use in waste impoundments include
the following:
o natural fine-grained soils, remolded so as to minimize permeability to
waste fluids;
o admixed materials, such as asphalt concrete, soil cement, bituminous
seals, and other additives; and
o synthetic membrane liners, constructed of polymeric rubbers or
plastics.
1
-------
The selection of a particular liner depends on such factors as the avail-
ability and quality of natural on-site lining materials, type and characterist-
ics of the wastes, and the costs associated with the installation and mainte-
nance of a specific type of liner. Generally, because of their low cost and
availability, clayey soils are considered the first alternative for a waste
confinement liner.
Under the auspices of RCRA, the State of Colorado will be involved with
the permitting of hazardous waste disposal facilities. It is anticipated that,
wherever possible, facility developers will wish to utilize earthen lagoons or
containment facilities lined with native soils derived from one of several geo-
logic formations located within the State.
Through the EPA's Technical Assistance Panels Program, the State of
Colorado has requested Federal aid to determine the types of laboratory analy-
ses necessary to evaluate remolded Colorado soils for use in hazardous waste
management facilities. This report presents the results of a review and evalu-
ation of existing methodologies available to determine the flow characteristics
of remolded natural soil materials when exposed to hazardous constituents.
Specifically, the report contains the following:
o A review of studies concerning the impact of different chemicals on
the permeability and physical characteristics of fine-grained soils;
o A summary of the failure mechanisms induced in natural clay liners by
chemicals which could alter the liners' permeability or stability;
o A review of the physical characteristics of the Colorado soils of
interest which may relate to identified failure mechanisms;
o A review of a list of ten chemicals provided by the State for inclu-
sion with the recommended testing program;
o A recommendation of test procedures which can be used to evaluate
Colorado soils for use as liners in hazardous waste disposal facili-
ties.
2
-------
Only through the proper testing of candidate liner materials can rational
decisions regarding the suitability of a particular material be made. It is
the intent of this report to provide the State with a set of suitable testing
protocols whereby the permeability characteristics of candidate soils can be
evaluated. Appropriate decisions can then be made regarding the use of a par-
ticular soil for lining a waste impoundment to contain particular waste materi-
als.
3
-------
II. IMPACT OF CHEMICALS ON CLAY LINER
PERMEABILITY - REVIEW OF STUDIES
A. Background
In their proposed regulations for the disposal of hazardous waste, the EPA
has specified several performance standards for earthen materials used as
barriers to seepage (1). These proposed standards are summarized below:
1) Surface impoundments shall not be used to contain wastes
which are detrimental to materials used as barriers to waste
movement.
2) Hazardous wastes shall be tested for compatibility with the
intended liner materials.
3) Soils used for liners shall:
o be classified under the Unified Soil Classification System,
shown in Figure 1, as CL, CH, SC, or OH;
o allow more than 30 percent passage through a No. 200 sieve;
o have a liquid limit equal to or greater than 30;
o have a plasticity index equal to or greater than 15;
o have a pH of 7.0 or higher;
o have a permeability equal to or less than 1x10"7 m/sec; and
o have a permeability not adversely affected by the waste to
be placed in the impoundment.
4
-------
FIGURE 1
UNIFIED SOIL CLASSIFICATION SYSTEM
FIELD IDENTIFICATION PROCEDURES
(Eheluding particles iorger»thon I inches and bosmg froctions on estimated weights)
GROUP
SYMBOLS
li
TYPICAL NAMES
9
•
«»
•
•»
M
O
o
eg
(ft
1 §
s 5
O -
M t.1 »
5 H &
2 -si -s
O •» M
O
ui e c
S i !
o 1 £
° i
• M
0 -
£
C «
1 -
r 1
o .
a »
o
§* *
S 5 S
0 o
> tt> »
a - »
«n c •* n
J 8 « M
> 2 z °
2 ° c o
3 i 2 3
* * «>
C «- -O
a «
¦c P
""a °
• 5 e
1 - •
2 ¦*
W)
u»
S 8
+
X O v*
:??
hJ -1
.J —'
Wide ronge m gram size ond substantial omounts
of oil intermediate particle sites
GW
Well groded grovels, grovel-sono mixtures,
little or no fines
Predominantly one size or a range of sues
with some intermediate sizes missing
GP
Poorly groded gravels, gravel-sand mnturest
little or no fines
5 s
i ii
^
«fl 2 9
J- u c
u U. « 3
> « °
J S 6
S a °
Non*plostic fines (for identification procedures
see ml below)
GM
Silty gravels, poorly groded grovel-sond-
$iIt mixtures
Piastre fines (for identification procedures
see CL below)
GC
Clayey gravels, poorly graded gravel-sond-
doy mixtures
_ -M
S 5 •
| 5 f ^
- ? -2
Al ® c .
S 5 O •»
s * o ;
c^uo if •
o «.z — -
z o ~ 5
* • 1 3 *
" 5t "2
If If
« M * O
3 2
2 it,
CLEAN SANDS
(Little or no
lines)
Wide ronge in gram sizes and substantial
omounts of all intermediate portide sizes
SW
Well graded sands, gravelly sands, little or
no fines
Predominantly one size or a ronge of sizes with
some intermediate sizes missing
SP
Poorly groded sands, gravelly sonds, little or
no fines
tANOS WITH
FINES
(Appreciable
amount of lines)
Non-plastic fines (for identification procedures
see ml below)
SM
Silty sonds, poorly graded sand-silt mixtures
Plastic fines (tor identification procedures
see cl below)
SC
Clayey sands, poorly graded sand-cloy mixtures
U
• f
5 1
• o
5
o «
O *4
~ 5
• »
5 5 I
8 5§
a a n
i 1 2
Sis
o a •>
w ®
s:
o
£
c
o
£
«
6
a
IDENTIFICATION PROCEDURES ON FRACTION SMALLER THAN No. 40 SIEVE SIZE
SILTS AND CLAYS
Liquid limit
less thon 50
ORY STRENGTH
(CRUSHING
CHARACTERISTICS)
OILATANCY
(REACTION
TO SHAKING)
TOUGHNESS
(CONSISTENCY
NEAR W.ASTIC LIMIT)
None to slight
Quick to slow
None
ML
Inorganic silts ond very fine sonds, rock flour, silty
or cloyey fine sends with slight plasticity
Medium to high
None to very slow
Medium
CL
Inorganic cloys of low to medium plosticity, gravelly
cloys, sandy cloys, silty clays, lean cloys
Slight to medium
Slow
Slight
OL
Organic silts ond organic silt-clays of low
plasticity
2 °
5 _
d | J
O ¦*"
Z T) L
* 3 *
• -s
f J?
J 9
Slight to medium
Slow to none
Slight to medium
MH
Inorganic silts, micaceous or diotomaceous fine
sandy or silty soils, elastic silts
High to very high
None
High
CH
Inorganic clays of high plasticity, fat cloys
Medium to high
None To very slow
Slight to medium
OH
Orgonic cloys of medium ro high plasticity
HIGHLY ORGANIC SOILS
Readily identified by color, odor, spongy feel ond
frequently by fibrous texture
Pt
Peot and other highly orgonic soils
Source: Bureau of Reclamation (53)
5
-------
These proposed performance standards did not appear in the final RCRA
Subtitle C Phase I regulations issued May 19, 1980; it is anticipated, however,
that these, or standards which are very similar, will be published in the Phase
II and III regulations due to be promulgated in the future. Therefore,
although the above proposed performance criteria currently cannot be considered
enforceable standards, they do serve as useful guidelines for the design of
earthen liners.
The thrust of the proposed regulations is to rely on the permeability of
an earthen liner as the primary protection for preventing seepage from a waste
impoundment. The regulations, however, have not specified if the performance
standard regarding permeability applies to water, or to the waste fluids antic-
ipated to be present in the facility. The permeability, and resulting seepage,
values could be quite different depending on the impacts that the waste fluids
have on the structure of the soils. A review of previous studies concerning
the impact of chemical wastes on the flow characteristics of earthen liners has
revealed a general paucity of information on this subject. Typical studies
where the behavior of chemicals in the soil has been discussed relate to aque-
ous systems where water is the dominant fluid. In response to the general lack
of knowledge about the impact of concentrated waste fluids in direct contact
with earthen materials, the EPA has initiated several studies to explore this
topic.
Both the EPA Municipal Environmental Research Lab (MERL) in Cincinnati,
Ohio and the EPA Kerr Environmental Lab in Ada, Oklahoma have recently funded
major studies to investigate the impact of organic chemicals on various liner
materials. MERL is sponsoring ongoing studies on the effect of organic chemi-
cals on clay liners, synthetic membranes, and admix materials. Dr. Kurt Brown
of Texas A&M University is the principal investigator of the impacts of organ-
ics on cl^y liners. Mr. Henry Haxo of Matrecon, Inc., in Oakland, California,
is the lead contact for studies pertaining to membrane liners and admix materi-
als. A design manual on the lining of waste impoundment and disposal faciliti-
es has also been recently published by MERL (2). This document includes work
completed to date by both of the above research groups on their respective pro-
jects.
6
-------
The Kerr Lab in Oklahoma has recently received their final report on the
impact of organic solvents on the integrity of clc(y liners by Dr. G. Fred Lec
of the Department of Civil Engineering at Colorado State University. The re-
port is currently undergoing peer review and will be published by the Kerr Lab
this year. Both the Kerr Lab and MERL projects will be discussed in detail in
Section C.
The emphasis of the recent EPA research activities have been directed
towards RCRA regulated hazardous wastes. Other sponsoring agencies, however,
have similar interests in the impact of various other potentially toxic wastes
on clay liners. Research this past year by the Battelle-Pacific Northwest
Laboratory, sponsored by the Nuclear Regulatory Commission, has been conducted
to evaluate the impact of uranium mill tailings leachate on clay liners under
acidic conditions. Similarly, the geotechnical Engineering Department at
Colorado State University investigated the stability of natural clay liners in
a low pH environment. These studies, and others pertinent to the scope of this
report, will be discussed at length in Section C.
B. Brief Characterization of Earthern Materials Used as Liners
Liners used at waste impoundment sites have the primary purpose of mini-
mizing the migration of waste constituents from the impoundment into the sur-
rounding environs. To fulfill this function, liners operate by two interdepen-
dent mechanisms (2):
o To impede the flow of pollutants and/or a pollutant carrier, usually
water, into the subsoil. The permeability characteristics of the
liner generally determine the rate of flow.
o To absorb or attenuate suspended or dissolved pollutants. This capa-
bility is dependent largely upon the chemical composition of the
1i ner materi al.
Earthen materials typically are more permeable than polymeric membranes,
but usually have a higher attenuative capacity. The thickness of soil liners
can also be varied, and all else being equal, the thicker the liner, the less
7
-------
the "driving force" through the liner. Darcy's Law expresses this relationship
as follows:
q = Ki
Where:
q = rate of travel through the liner (L/T)
K = permeability of the liner (L/T)
i = hydraulic gradient, defined as:
Fluid depth plus liner thickness
(Dimension less )
Liner thickness
By increasing the thickness of the liner, the hydraulic gradient is reduced,
which results in a corresponding decrease in the rate of travel through the
li ner.
Soils which are recompacted to form liners must contain a sizeable portion
of fine grained material. The EPA proposed liner criteria (1) for particle
size specifies that at least 30 percent of the liner materials must be smaller
than a number 200 sieve, which is equivalent to a 0.075 mm grain size. This
corresponds to the silt and clay-size fractions on the ASTM grain size scale
(3). MERL, however, concludes that the percentage of the clay size fraction
(grain size less than .002mm) is the primary concern to achieve proper perform-
ance. MERL recommends that the minimum amount of clay size particles required
in soil to yield a good soil liner is 25-28 percent by weight (2).
The term "clay" is commonly used in the literature as a shorthand for two
separate concepts - clay size and clay mineral. Any particle or grain that is
smaller than 0.002 mm is clay size. A clay mineral is a naturally occurring
chemical compound with a specific range of compositions and a specific type of
layered structure. In general, aluminum, silicon and oxygen atoms form the
layers or sheets and a variety of other ions are sandwiched in between.
In addition to particle size, the chemistry and mineralogy of the clay
size fraction are also important in determining the physical properties of clay
liners. The three clay mineral species most likely to constitute the bulk of
8
-------
the clay size fraction of soils are kaolinite, illite and montmorillonite.
Montmorillonite typically has the highest cation exchange capacity (CEC), the
highest surface area, and the greatest ability to swell upon hydration. Brown
and Anderson (4) and MERL (2) each give an extensive background and review of
the physical properties of the major clay minerals.
Soil density is the other primary factor which can be varied to influence
the flow characteristics of an earthen liner. All other factors being con-
stant, a reduction in the void ratio via compaction of the soil should result
in a lower permeability (2). The reduction of the void ratio promotes two
changes in the soil: a decrease of the effective area available for flow (re-
duction in total pore area) and a reduction of the average pore size. There
are, however, some additional subtle factors which can contradict the statement
that a reduction in permeability is always directly proportional to the maximum
density achievable for a particular soil. For instance, the clay fraction of a
soil is sensitive to changes in water content during recompaction, so that when
soils are compacted wet-of-optimum (optimum moisture being that moisture con-
tent at which maximum density is achieved; wet-of-optimum means that the water
content exceeds this value), a further decrease in permeability can result.
The explanation partially lies in the orientation of the clay particles during
and after compaction. A compacted mass of clay size material will exhibit con-
siderably higher permeability than the grain size of the individual particles
would lead one to expect if the clay size particles form or remain as clumps
(floccules) each containing a multitude of the fine grained particles. Then it
is the size of the floccules and the voids between them that controls the
permeability. When compaction takes place under conditions that permit each
clay size particle to move independently in a fluid medium (particles dis-
persed) permeability can be greatly reduced by their reorientation. Clay com-
pacted dry-of-optimum tends to have an "open", flocculated structure, while the
wet-of-optimum has a dispersed structure (2). Other physical-chemical proper-
ties of a clayey soil, such as the interlayer chemistry of the particle-unit,
and the typical size of agglomerated clay particles, can play significant roles
in determining the permeability values of a candidate soil at a given density.
Those factors, where identified, which can induce an increase in permeability
in an otherwise properly designed earthen liner are discussed more fully under
the individual research projects in Section C.
9
-------
c. Description of Major Studies Identified in the Literature
This section provides a synopsis and brief abstract of each of the major
research efforts identified which explore the effects of various chemicals on
the flow characteristics of clay liners. Each study is discussed separately;
where possible, the general methodology, results, and recommendations are all
included. The identification of failure mechanisms, if available, is also an
integral component of each discussion.
1. G. Fred Lee - Colorado State University
Dr. Lee recently completed work on the impact of organic solvents on
the integrity of clay liners for the EPA's Robert S. Kerr Environmental
Laboratory (5). Three different natural clay soils, each exposed to seven
different organic solvents and water, were studied for changes in perme-
ability. The shrink-swell properties of the clay soils were also investi-
gated following exposure to the solvents. Sorption studies of three sol-
vents, tested with each of the three clay soils, were also performed in
order to assess the general attenuation characteristics of the liners.
Since the scope of this review is to discuss permeability changes and bulk
failure mechanisms induced through the interactions of chemicals with clay
liners, the sorption studies will not be discussed further.
Lee concluded that conventional constant-head permeability tests,
such as those specified by the American Society of Testing Materials (6),
are inadequate for determining the permeability characteristics of clays
when exposed to organic solvents. The constant head technique, generally
used for course-grained materials, is not as reliable as the falling head
method for testing low permeability material such as clays. The falling
head method allows for more accurate measurement of flow rates, since the
quantity of fluid passed through the sample need not be collected and
measured directly. Rather, the volume is calculated from the fall of the
fluid level in the column above the sample. This advantage minimizes the
possibility of significant error resulting from the evaporative losses of
permeating fluids prior to collection and measurement, a factor which is
of upmost importance for flow-restrictive materials such as clays, where
10
-------
the quantity of seepage is so small that great care is necessary to avoid
leaks and evaporation (3). The problem is further compounded by the gen-
erally volatile nature of organic solvents, which allows for even greater
evaporative losses than could occur using distilled water alone. Lee also
concluded that the sealants used in conventional permeameters, although
satisfactory for testing with water, are typically not compatible with or-
ganic solvents.
As a result of the above considerations, Lee recommends the use of
thickwalled Pyrex glass permeameter columns, equipped with inert teflon
joints, for all falling-head permeability tests on clays. Figure 2 shows
the fabricated design of Lee's permeameter. Other important features of
the column include an inlet port for pressurized nitrogen, and a small
diameter standpipe to minimize evaporative losses of highly volatile sol-
vents during testing. Typical flow rates through the column allow testing
to be completed within a 36 day time period; however, for clay-fluid sys-
tems with extremely low flow rates, pressurized nitrogen could be used to
allow for a constant pressure of 10 to 50 pounds per squre inch, thereby
increasing the rate of flow through the column. Procedures for set-up and
execution of the tests are contained in Appendix A.
Using the specially designed glass-teflon permeameter, Lee tested the
permeability of the test clay soils with the following solvents: benzene,
xylene, carbon tetrachloride, trichloroethylene, acetone, methanol, and
glycerol. The results indicate that the ease with which the solvents
permeate the clay-soils can, in some instances, be correlated with several
physical characteristics of the solvent. Lee concluded that the coeffi-
cient of permeability was typically related to the dielectric constant of
the solvent, and also to the solvent's octanol/water partition coeffi-
cient. The dielectric constant is an indicator of a solvent's ability to
dissolve ionic species, and can be viewed as a measure of the hydrophilic
nature of the solvent. The octanol/water partition coefficient also is an
indication of a solvent's compatibility with water. Both of these para-
meters can be readily obtained from standardized chemical reference tables
(7, 8). Lee asserts that his results indicate that the more hydrophilic
the organic solvent, the more rapidly it moves through the clay. In all
11
-------
FIGURE 2
LEE'S PRESSURE PERMEABILITY COLUMN
99 cm
K
•J
A /
( II"')
SOIL
SAMPLE
1 cm
12 cri
High Pressure
Stainless Steel
Valve
-~To Nitrogen
. . * , Tank
Liquid Level
8 mm Graduated Stanpipa
High Pressure Swagelock
Fitting (Teflon)
CI amp
1.8 cm Thick X 10 cm ID
Pyrex Glass
Teflon Gasket
High Pressure Joint
Whatman GF/A Fiberglass
Fi1ter
40 Mesh Stainless Steel Screen
6 cm Pyrex Glass Support
NOT TO SCALE
i To Col 1ecti on
12
-------
cases, however, where the solvent did not cause excessive shrinkage of the
clay (discussed below), the permeabilities of the clay soils to organic
solvents was lower than the permeability to water. Typically, the greater
the hydrophilic tendencies of the solvent (high dielectric constant, low
octanol/water coefficient) the more closely the solvent approached flow
rates achievable with pure water., Typical permeability coefficients were
in the 10~8 to 10"9 cm/sec range.
For five of the seven solvents tested the above correlations did not
hold true. Xylene, benzene, and carbon tetrachloride, three solvents each
having low dielectric constants, would be expected to move quite slowly
through the clays based on their hydrophobic tendencies. Although this
behavior was initially observed, solvent breakthrough occurred shortly
thereafter on one or more of the clays. The phenomenon of solvent break-
through was attributed to the ability of these solvents to induce exces-
sive shrinkage in the clays, eventually leading to the development of
cracks and bulk transmission of fluids. This phenomenon was observed as
early as six days after initiation of the experiment for carbon tetra-
chloride, and as late as 24 days into the testing for xylene. Therefore,
although a solvent may move slowly through a clay, as expected, prior to
the occurrence of shrinkage, there is a principal danger that it will
eventually cause shrinking and cracking with the resultant quick passage
of fluids. The implication of this behavior points to the need for indi-
vidual permeability testing of the candidate liner materials with each
organic solvent expected to be present, rather than strict reliance on in-
direct measurements (such as hydrophobic tendencies) as indicators of sol-
vent behavior.
Acetone and glycerol also generated anomalous results. Acetone is
highly hydrophi1ic, and would be expected to irove quite rapidly through
the liner materials. This was not observed; Lee attributes the behavior
to the possible microbial decomposition of acetone on the clay, with the
resultant production of CO2 and clogging of the pores within the clay with
gas. Similar behavior has also been observed by Poulovassilis (9), who
found that the microbial production of gases produced significant reduc-
tions in soil permeability.
13
-------
Glycerol has a lower octanol/water partition coefficient than water
itself, and would therefore be expected to permeate the clays at a higher
rate than water based on its hydrophilic nature. This also was not ob-
served; Lee attributes glycerol's unusually low permeability to its rela-
tively high viscosity, a property of the fluid which can retard its rate
of movement through porous media such as soil. However, glycerol perme-
ated the test soils much more rapidly than would have been expected based
on a comparison of its viscosity with that of water alone. A number of
investigators have utilized in the past "intrinsic" permeability values,
which take into consideration the viscosity and density of the fluid, to
compare permeability results achievable for fluids other than water (10,
4). The general equation for calculating intrinsic permeability values is
shown below:
Where: k = intrinsic permeability (cm/sec)
q = rate of flow (cm/sec)
i = hydraulic gradient (dimensionless)
n = viscosity of the fluid (gram/cm-sec)
= density of the fluid (gram/cm-*)
g = gravitational constant (cm^/sec)
Lee's results, however, provide empirical evidence that the above
equation does not fully explain the variations in permeability observed
for non-aqueous organic fluids. Viscosity and density are not the only
factors which must be considered when comparing modified permeability
values for solvents which possess different fluid properties. The above
equation should therefore not be construed as being universally
applicable.
Lee also evaluated the permeability of the earthen materials to
mixtures of solvents. The results indicate that solvent mixtures tend to
produce permeability characteristics as if the solvent in the mixture
-------
possessing the highest dielectric constant were acting alone. Hence, the
behavior of the solvent mixture will approximate that behavior which is
achievable with the solvent having the highest dielectric constant.
Although this cannot be accepted as a universal statement, the results
indicate that those organics which induced shrinkage and bulk transmission
of fluid (typically the low dielectric solvents such as xylene, benzene
and carbon tetrachloride) might be co-disposed with a higher dielectric
solvent. This could possibly minimize the chance of c 1 at/ liner failure
resulting from shrinkage.
Concurrent with the column permeability studies, Lee also evaluated
the shrink/swell behavior of the clays when exposed to the organic chemi-
cals. A shrink/swell consolidometer, sensitive to deviations of one
thousandth of an inch, was used to determine volume changes of the cl«*ys
(11). It was found that several chemicals induced measurable volume
changes in the cl«iy soils. Those solvents with high and intermediate
dielectric constants (water and acetone) caused the clays to swell much
more than those with lew dielectric constants (benzene, xylene, carbon
tetrachloride, and trichloroethylene). Xylene and carbon tetrachloride
actually induced shrinkage in the clays, which, as mentioned in the previ-
ous discussion on permeability, can lead to bulk transmission of the
fluids through shrinkage cracks. Benzene induced shrinkage on initial
contact, but then a net swelling was observed subsequently.
Interestingly, Lee also concluded that the montmorillonite content of
the clays was not a factor in predicting the degree of swelling of the
clays when exposed to the organics. Montmoril lonite clay, in pure from,
can swell up to 15 times its original volume (1,500 percent), through the
adsorption of water both within each particle on its interlayer surfaces
with subsequent interlayer expansion and by absorption of water between
the particles (interparticle expansion) (12). Other clay minerals swell
to a lesser extent because absorption of water is limited to the area be-
tween the particles. Lee, however, measured the interlayer distance of
the test clays using X-ray diffraction techniques, and observed that the
interlayer distance did not change following exposure to the organic com-
pounds, even though swelling was observed. Lee concludes that the observ-
15
-------
ed swelling can be attributed entirely to interparticle expansion, rather
than actual interlayer expansion. The percent montmorillonite content
would not, therefore, be correlated with the actual distance between clay
particles except for the fact that montmorillinite clay tends to be consi-
derably finer grained than other clays and therefore has more interpar-
ticle spaces for expansion. Lee believes that this distance would be pri-
marily determined by the charge properties of the solvents, since the
charge properties can enhance or reduce clay particle interactions. Lee
also found that the organic carbon content of the clays was important in
determining the degree of swelling. In general, the higher the organic
carbon content, the greater the percent swell caused by the solvents. Lee
attributes this observation as being the result of Van der Waal's interac-
tions between solvent molecules and the organic substances contained in
the clay.
In summary, Lee has concluded that conventional permeability testing
equipment is ineffective in determining accurate permeability characteris-
tics of clayey soils to organic solvents, due to such interferences as
chemical reactions of the solvent with the permeameter, and the evapora-
tive loss of volatile solvents. The evaporation of test fluids can lead
to inaccurate measurements of flow rates, an extremely critical considera-
tion for flow restrictive materials such as clays. Through the use of a
specially fabricated falling head permeameter, Lee has concluded that the
permeability of candidate liner materials to organics is highly influenced
by the tydrophobic properties of the liquids. However, liner failure was
observed with those solvents which caused shrinkage in the clays, regard-
less of the hydrophobic nature of the solvents. In all cases where
shrinkage did not occur, the clays were less permeable to the organics
than to distilled water. Other factors which are important in evaluating
the permeability characteristics of clays include microbial decomposition
of the solvents which can produce gas and resultant clogging of pores, and
the organic carbon content of the clays which can influence their swell
behavior. Although the permeability of earthen materials can be correlat-
ed to a certain degree with several physical properties of the solvents,
Lee's results indicate that each organic waste must be tested individual-
ly, since the observed correlations are preliminary in nature and excep-
tions occurred with 5 of the 7 organic solvents tested.
16
-------
2. Brown & Anderson - Texas A&M University
Dr. Kurt Brown and co-workers at Texas A&M University are in the pro-
cess of evaluating the effect of organic compounds on clay liner permea-
bility through the funding support of EPA's Municipal Environmental Re-
search Lab (MERL). A recent publication by the Texas A&M researchers (4)
provides the results of a literature review to identify pertinent aspects
of clay liner failures. Although the review indicates a near-complete
lack of knowledge about the possible impact that waste impoundment con-
tents may have on clay liner permeability, several failure mechanisms were
revealed. Those discussed include: clay dissolution, volume changes as a
result of shrink/swell behavior, and clay dispersion.
Clay dissolution can occur in those situations where either strong
acids or bases are present. These harsh environments can promote the
solubilization of various constituents within the clay, such as alumina
and silica. This partial dissolution can lead to increases in permeabili-
ty. Although the source of the extreme pH environments typically arises
from the disposal of acidic or basic wastes, organic acids can also be
produced as by-products of the anaerobic decomposition of organic fluids
in waste impoundments.
Volume changes in the liner can occur under those circumstances which
promote either the shrinkage or the swelling of the clay particles. This
can typically take place when there is a change in the water content of
the clay. The water content can change if an organic leachate displaces
the interlayer water of a clay. A decrease in interlayer spacing, leading
to net shrinkage of the clay, would occur if the fluid lacked the charge
properties associated with water a promoted the retainment of fewer layers
of organic molecules. In general, montmorillonitic soils would be most
susceptable to this type of phenomenon. Although shrinkage is certainly
undesirable since cracking can allow high rates of flow, swelling may also
cause undue effects. Liners composed of particles which swell upon expo-
sure to organics increase in volume, as a whole, with potential loss of
their integrity through heaving. Loss of integrity could also occur if a
swelled liner were to shrink at a later date following displacement of the
imbibed organics with water.
17
-------
Clay dispersion can contribute to liner failure through the phenome-
non of soil piping. Brown 4 co-workers point to several instances of un-
controlled seepage through earthen structures which was the result of
fluid transport in localized channels. Typically, changes in the pore
size distribution of the clays, leading to an increase in the number of
macropores, was found to be associated with the dispersive characteristics
of the clay. Certain organic compounds can promote dispersion in clay
soils, with a resultant increase in the potential for permeability
increases via soil piping. Dispersive clays also lack structural
strength, and this can lead to problems with liner placement and long-term
integrity. The researchers also point out in their studies that although
in many cases a change in clay soil permeability can be readily measured
or observed, the phenomena causing the change are either poorly understood
or else cannot be identified.
Some of Brown's research is reported in the MERL design manual on
liners (2). Appendix III-D of the MERL manual contains the permeability
test method and apparatus which Brown is currently using in his research.
The method uses a constant elevated pressure method to determine permea-
bility. Conventionally available permeameters (13) are used in the method,
with the intent that standardized procedures can be utilized universally
without having to resort to specially fabricated equipment. The permea-
meters are modified slightly so that pressurized air, up to 60 psi, can be
introduced into the chambers to increase the rate of flow through the low
permeability clays. Appendix B to this report contains the full-text de-
scription of the test method, as reproduced from the MERL manual (2).
At the time of this writing, Brown has not published any actual re-
sults of his research. Brown has stated, however, that material will be
available during 1981 (14). Preliminary results will be presented at the
EPA's 7th Annual Conference on the Land Disposal of Hazardous Waste, March
1981, and should also appear in the forthcoming proceedings of the
American Society of Agronomy Conference, Soil Physics Section, held in
December, 1980. The EPA conference proceedings should be published by the
fall of 1981.
18
-------
Discussions with Brown and co-workers (14) have revealed that liner
failure, if it is to occur, will typically take place prior to the passage
of one pore volume of permeant through the standard four inch sample of
material. One pore volume of fluid will generally pass through the mater-
ial in about 21 days, using air pressures of up to 60 psi on the column.
Currently, Brown is investigating the effect of several different pres-
sures on permeability to confirm that artifical variations in the coeffi-
cient of permeability are not being introduced through this action. Brown
also recommends simultaneous experiments with both water and the solvent
of interest in a battery of permeameters, to observe both the net effect
of the solvent on permeability, as well as the variability induced through
the experimental procedure itself. Brown's primary concern in developing
the procedure was to facilitate duplication of the method so that results
obtained by individual research groups can be readily compared.
3. Battel!e-Pacific Northwest Laboratory Study
In June, 1980, Battelle completed a study for the U.S. Nuclear Regu-
latory Commission on the interaction of uranium mill tailings leachate
with clay liners (15). The major concern of the clay liner investigation
was to determine whether or not tailings solution could interact with
earthen liner materials and in time, induce an increase in permeability.
Test procedures were designed to evaluate clay dispersion phenomenon, clay
dissolution, and time dependence of clay permeability following contact of
clay liner materials with an inorganic, aqueous tailings soution having a
pH of 2. A native Wyoming clayey-silt soil, representative of liner
material for a proposed tailings disposal pit, was used for the analysis.
Prior to evaluating the permeability characteristics of the material,
Battelle determined that the soil met the other proposed EPA criteria (1)
for materials to be used as earthen liners.
The Battelle group used a variation of the ASTM sliding weight tamper
method (16) to compact the clay liner material into several permeameter
test cells, each cell being compacted to about 95% of maximum compaction.
The ASTM constant-head permeability test (6) was run on each cell, modifi-
ed slightly so that each cell could be pressurized with air to increase
-------
the rate of flow through the cells. Up to 20 psi of pressure was used.
The flow through the cells was from bottom to top to ensure saturated con-
ditions. Gee (17) has suggested that laboratory grade pressurized nitro-
gen be used in further tests rather than pressurized air, to minimize the
change of oxidation reactions occurring, in the cells. The permeability
tests were run for a minimum of 100 days to a maximum of 250 days; the
outflows from the most permeable columns exceeded 10 pore volumes.
The results of the permeability tests indicate that the permeability
of the liner material was not adversely affected by the tailings leachate
used as a permeant. In no case did the permeability of the material ex-
ceed the EPA proposed permeability criteria of lxl0~7cm/sec, even after
8 months of testing and the passage of 12 pore volumes of permeant. The
average permeability of the material was 7.5xl0*9cm/sec; the range ob-
served in several cells was 5xl0~8cm/sec to 1.5 x 10"^cm/sec. Gener-
ally, an initial decrease in permeability was observed for the samples,
which then leveled-off after approximately 50 days of testing. The ini-
tial decrease may be explained by the precipitation of various inorganic
chemical compound within the clays.
In addition to the permeability tests, Battelle utilized the crumb
test (18) to determine the dispersive characteristics of the liner materi-
al. Clay dispersion is thought to be an important mechanism of failure,
possibly leading to soil piping (4). The results indicated that the clay
was non-dispersive when exposed to the particular test solution.
Battelle also evaluated potential mineralogical changes in the clay
following contact with the acidic tailings solution. X-ray diffraction,
scanning election microscopy, and X-ra^y fluorescence techniques were used
to evaluate whether or not clay dissolution and/or mineralogical changes
were occurring. The results indicate that little dissolution occurred,
and that no drastic or substantial changes in the mineral surface morpho-
logy of the clay was present; the results, however must be viewed as pre-
liminary. The Battelle lab concludes its report with the statement that
within the period of examination, the permeability of the clay material
had not been adversely affected by the acidic tailings solution. In fact,
20
-------
contact with the solution was determined to affect the permeability of the
clay liner beneficially. The decrease in permeability observed increased
the ability of the clay liner material to impede transport of the chemi-
cals by increasing the time available for chemical reactions and increas-
ing the buffering capacity of the soil.
4. Geotechnical Engineering Program Study - Colorado State University
Investigators at Colorado State University have also evaluated the
stability of natural clay liners in a low pH environment (19). Two nat-
ural clay soils were obtained from a test site near Baggs, Wyoming, both
representative of candidate liner materials for a uranium mill tailings
impoundment. Although the samples were located within the same vicinity
of the test site, the two samples differed from one another in terms of
the percentage of montmorillonite present in the clay fraction. Both clay
#1 (yellow clay) and clay #2 (red clay) had a clay content of approximate-
ly 50 percent; however, the yellow clay had approximately twice the mont-
morillonite content as shown below:
CLAY COMPOSITION BY PERCENT OF CLAY-SIZED FRACTION (< 0.002 mm)
Yellow Clay % Red Clay %
Montmorillonite 45 20
II lite 10 15
Vermiculite - 15
Kaolinite 25 35
Quartz 20 15
Both samples were classified as CH soils according to the Unified
Soil Classification System.
Permeability experiments were performed on each of the clays using
falling-head permeameters. Glass-walled permeameters were fabricated to
avoid chemical reactions between the acidic permeant and the permeame-
ters. The clays were compacted to 90$ of the ASTM standard proctor dry
o 1
-------
density (ASTM D-698), transferred to the permeameters and compacted
slightly to affect a seal between the soil and the glass column. Each
clay was saturated and permeated for 21 days with distilled, deaerated
water.
After 21 days, tailings liquor was applied to the columns, and the
permeability experiments continued for an additional 250 days. Typically,
there was a noticeable decrease in permeability following introduction of
the tailings solution; this may be explained by the precipitation of in-
soluble salts on the inner surfaces of the clay.
For the red clay, following the inital decrease in permeability, the
values tended to increase and then level off near the values obtained for
water. At the end of 250 days, the coefficient of permeability was 2xl0-8
cm/sec, somewhat above the initial distilled water value of 7x10"^ cm/
sec. The permeability of the yellow clay exhibited similar trends; how-
ever, the level of increase in permeability, following the initial de-
crease upon application of the tailings solution, was higher. The final
determination after 250 days was about 1.5x10"^ cm/sec. The value for
distilled water was about 1x10*8 cm/sec; therefore, an increase in permea-
bility of about an order of magnitude above that expected for pure water
was observed for the yellow clay.
For both clays, a relatively high amount of scatter in the data
throughout the experiment was observed. For instance, sample outlyers for
the red clay, near the 100 day mark, ranged from 1x10"^ cm/sec to 2x10"?
cm/sean observed two orders of magnitude change over a time period of
several days. The authors present the probable explanation as being par-
tially the result of temperature fluctuations in the laboratory. The ob-
served scatter in the data for this experiment points to the overall pro-
blems which exist in measuring flow rates through materials having coeffi-
cients of permeability of less than 10" ^ cm/sec. More importantly, the
observed scatter indicates that intensive measurements of flow rates must
be made throughout the analysis; single "point" measurements at various
intervals are inadequate to explain whether observed fluctuations in perm-
eability are the result of actual physical changes in the liner material,
22
-------
or are the result of experimental error. The investigations at Colorado
State University (19) clearly demonstrate that flow rates must be measured
at the most frequent interval possible, and statistically analyzed so that
meaningful interpretations of observed changes can be made.
5. Studies by Matrecon, Inc. - Oakland, California
Henry Haxo of Matrecon, Inc. has been sponsored by the EPA Municipal
Environmental Research Laboratory to conduct various studies of the ef-
fects of both hazardous wastes and landfill leachates on a number of
selected lining materials (20-25). Matrecon was also the primary con-
tractor for preparation of the MERL design manual on liners (2). Speci-
fically, Haxo's efforts have focused on impacts of chemical wastes on syn-
thetic membrane liners and admix materials such as soil cements, asphaltic
concretes, and other asphaltic compositions. However, some preliminary
work has been performed with compacted fine-grained soils as well. The
initial intent of the project was to expose the candidate liner materials
to a number of wastes in specially designed test cells for a time period
of 24 months. The test cells would be dismantled at 12 months and 24
months so that the liner materials could be recovered and tested. In all,
144 test cells were utilized.
Since the goal of this report is to provide information on native
soil liners, only Matrecon's results with soils will be discussed. Figure
3 shows the test chamber used by Matrecon to evaluate the soil material.
A twelve-inch thickness of a fine-grained, plastic soil obtained at Mare
Island, California, was used in the test. Five types of hazardous waste,
listed below, were selected for use in the evaluation (17):
o an alkaline sludge;
o a cyclic hydrocarbon sludge;
o lead waste from gasoline tanks;
o oil refinery tank bottom waste; and
o pesticide sludge
-------
FIGURE 3
MATRECON'S EXPOSURE CELL FOR SOIL LINERS
Top Cover
Epoxy
Coated—
Bolt
Flanged Steel
Spacer
Waste Column:
¦11 Gauge Steel
IO"x 15s* 12" High
w/ Welded
2 * Flange
Outlet tube with
Epoxy-cooted
Diaphragm
Waste
Neoprene- Sponge Gasket
Epoxy Grout Ring
ADMIX LINER
Epoxy ond Sand
Coating
Silica
>-Glcss Cloth
Screen-
To
„ Collection
Bog
24
-------
To date, only one soil specimen, that exposed to the oil refinery
waste, has been recovered and tested. After removal of the soil from the
test cell, the permeability of the material was evaluated using a back-
pressure permeameter developed from asphaltic concrete speciments by
Vallerga & Hicks (26). The results from three consecutive trials indicate
permeabilities of 10"® to 10"? cm/sec, which compares quite favorably to
results achieved prior to exposure (25). Further results obtained with
the other soil liner/hazardous waste combinations can be expected to be
published at some point in the future; again however, Haxo's primary in-
tent is to focus on non-earthern liner materials, so the results should be
viewed accordingly.
6. Miscellaneous Studies
This section will attempt to identify other sources of information
contained in the literature which, in addition to the primary studies al-
ready discussed, have a bearing on the impacts of hazardous constituents
on earthen liner materials.
Noyes Data Corporation published in 1980 a text reviewing the charac-
teristics of protective barriers used for containing toxic materials
(27). Although the manual primarily emphasizes design considerations, a
section on the testing of liners is also presented. A review of that par-
ticular section, however, reveals that the material is a summary of the
procedures and results published elsewhere by Haxo (20-25).
Ware et al. (28), present a partial listing of sanitary landfills
and hazardous waste disposal sites where some form of impermeable lining
has been installed. Included in the discussion is an examination of
methods for recovering specimens of liner materials for laboratoi7 testing
which have been installed in a landfill and subjected to actual service
use.
Exxon Research and Engineering Company (29) presents a review of the
characteristics of and constituents contained within seven types of hazar-
dous waste streams. A brief qualitative assessment is also presented on
25
-------
the compatibility of the seven wastes with both natural and synthetic lin-
ing materials.
The MERL design manual (2) discusses a number of factors which can
contribute to soil piping in a clay liner. One of the primary factors is
the dispersive character of the clay. Four tests are listed to determine
the susceptibility of a soil to dispersive erosion. These tests, describ-
ed in detail in a ASTM publication (30), include the pinhole test (a test
of dissolved salts in the pore water of the soil), the crumb test, and the
U.S. Soil Conservation Service dispersion test. The design manual fur-
ther discusses the relationship between a clay's potential for volume
change and its measured Atterberg limit values. Clays with a plasticity
index of 0-30 units are considered to have low swell potential; those with
an index of 30-50 units have a moderate potential, and clays with an index
greater than 50 units have a high swell potential. The presence of organ-
ic cations, however, can block the swell potential of a soil. This char-
acteristic could limit the capacity of a clay to "self-heal" or close *
shrinkage cracks should they develop. Some organic leachates can also
promote swelling; Barrier (31), as referenced in MERL (2), reported swel-
ling of montmorillonite clays as a result of exposure to acetonitrile, xy-
lene, cyclopentane, alcohols, glycols, and ketones (it is interesting to
compare this with Lee's previously mentioned observation that xylene caus-
es shrinkage). MERL also recommends an initial "sensitivity test" be con-
ducted for soils exposed to chemical waste. In this test, the values ob-
tained for the Atterberg limits, using distilled water, are compared
against those achieved with the waste fluid. A major difference between
the two tests will yield a preliminary indication of the sensitivity of
the soil. A soil would be considered "sensitive" if it did not have the
ability to resist changes in physical properties following contact with
the waste fluid. For more details, the reader is referred to the MERL
design manual for discussion of these and other topics.
Williams (32) briefly discusses the role of clay seals in prohibiting
seepage in ponds, lagoons, and landfills. Although permeabilities lower
than 10"? cm/sec are achievable with clays, Williams qualitatively states
that the permeabilities are adversely affected by those industial wastes
26
-------
whose high cation content or low pH induces ion exchange of the sodium
ions within the clays. The exchange of the sodium ions may inhibit or
reverse the swelling process by reducing the clays' ability to adsorb
water; this can increase the permeability of the liner. Williams discuss-
es a number of contaminant resistant bentonitic (montmorillonte-rich)
clays which are commercially available from the American Colloid Company
(33). The bentonites are treated so that upon hydration, the molecular
structure of the product prevents adsorbed water from being removed by low
pH or cation-bearing leachates.
Van Zyl and Caldwell (34) are currently evaluating the efficiency of
a cle\y liner installed to reduce seepage from an impondment containing
highly acidic gypsum tailings. The nine-inch thick liner has limited es-
cape of pollutants to the groundwater sampled in perimeter test wells to
within acceptable EPA limits for fluoride and pH, following ten years of
service life. Although no details are given, the investigators point out
that further evaluations are ongoing.
D. Summary and Evaluation of the Literature
The literature review has revealed the general paucity of information re-
garding the effect of concentrated chemical wastes on earthen liner materials.
However, the passage of RCRA has stimulated the recent funding of several im-
portant research efforts to explore this topic. Many of these studies are on-
going, and therefore, detailed results are not available at this time; however,
the approach and initial experimental procedures have generally been publish-
ed. It was apparent from the literature that the physical and engineering pro-
perties of soils which are important in determining a liner's flow characteris-
tics to water are well documented and are generally easily measured. The same,
however, is not true for non-aqueous waste fluids. Both Fred Lee (5) and Kurt
Brown (4), the principal investigators examining the effect of concentrated
organic wastes on the permeability of of earthen liners, have confirmed that
each waste-liquid clay-liner combination must be individually evaluated, since
at this time general predictive techniques for generic classes of compounds are
not fully developed. Preliminary work does, however, indicate that there is
some correlation between several physical properties of the organic solvents,
27
-------
and their behavior when exposed to candidate clay soil liners. Indirect
measurements that are predictive of permeability behavior, in lieu of time-con-
suming permeability tests, may therefore soon be possible. At this time, how-
ever, carefully controlled permeability experiments are required to successful-
ly evaluate the compatibility of hazardous wastes with respective seepage con-
trol liners.
Section V will more fully discuss the actual testing protocols and sup-
port procedures used by the major research groups to evaluate the permeability
characteristics of candidate earthen materials. Recommendations will then be
made as to the appropriateness of the tests for use by the State of Colorado in
their subsequent evaluations.
28
-------
III. CHARACTERISTICS OF SELECTED COLORADO CLAYSTONES AND CLAYEY SOILS
In Colorado, nearly all of the hazardous waste that is destined for
containment in earthen lagoons is generated in the Front Range Urban Corridor
region. The weathered claystones and clayey soils which are both locally
available and potentially suitable for remolding into liners are derived from
specific portions of the sedimentary rock layers which underlie the area. The
distribution of all the underlying rock units is shown in Figure 4. A hypothe-
tical stack of the sedimentary rock units (a composite columnar section) is
shown in Figure 5 to illustrate their interrelations. The lithologies (rock
types, etc.) of the units that are potential sources of clays are also briefly
described in Figure 5. Those rock units which have been identified as the
sources of potentially suitable clays are the Pierre formation, the Laramie
formation, and the Dawson formation. The Dawson has been subdivided into the
Denver and Arapahoe formations in part of the area, including metropolitan
Denver.
A. Physical Properties of Clays Relevant to Their Suitability as Liners
Some of the physical properties that are considered to be relevant to the
suitability of a clayey material for use as a liner are listed below:
1. Unified Soil Classification:
(CL, CH, SC or OH; USEPA (1))
2. Percentage of Fine-grained Particles:
(30% < 0.074mm; USEPA (1))
(25% < 0.002mm; MERL (2))
3. Engineering Properties:
(liquid limit, plastic limit, plasticity index; USEPA (1))
(potential volume change - shrink/swell; Lee (5))
4. Permeability, not adversely affected by waste:
(1 x 10"^cm/sec. water; USEPA (1))
29
-------
FIGURE 4.
GENERALIZED GEOLOGIC MAP OF THE
FRONT RANGE URBAN CORRIDOR. COLORADO
WYOMING
T
OLLI
rott'Cottlt Rock
C«9leatr«ti co»-
Oovtoa
0«jnr*r>Aropobo«
Poc«stteat aad Cottlt
Ro«k Coagloaorot*
eadlfforoatiotod.
« # »- aJ0'»
*• uOENVER 0 o . 0
• o * r
6" ° o°
'A P'V
Lsroai* FermotUa
V. p ....
r« v. -castle * a •:«
lOCK
;V.j
Vtrstto Foroofioa
osd Triaidfld So»4*
ifoaa
o a ROCK,.
* AS*'«?-*'®,L-B-
. A 9 , J ,
* ° ° sic. * « • - • •
»«- «s« •• „
a<* Q •
!Kd _ a _ « •
r«H
Fa Wilt Soadttoat
COLORAOO
SPRIN091
Pro-Co torado
droop rocfc»
£;&£zH(fS<
Piorro Sbalt
I *' I
Colorado droop
(Gmtrat, Grooaftoro,
CorUlo,Pt Hoyi,ud
Ssoky Hill)
Qooloqy nsdlfM oflor
U.S Soologicol Ssrvoy,
lf39,Q*oi09>e boo •#
Color odo. Seol«:300£00
Source : Hart (12) .
0 10 tO U Rilooitirt
30
-------
(MUM ftaaCalattria Schaal at HUa*.Caa*f*ll>a4
watoalta loailtnfttu taaltaa. float tut* •<
Colotatfa)
d
D
o
Castle Rock
conglomerate
EOCENE
Green Mountain
conglomerate
600'
iiiPiiiir
d
Oawson/
o
Denver-
j.jWAfat'jBV i.r j*
-j
Arapahoe
2
undiff.
1100*
Laramie
400- 600'
Fox Hills
50*120* |
Pierre
8000'
Smoky Hill
430'
Ft Hoys
33-5#
Codell ,
0-25*
Carlile
200*
Greenhorn
60'
Graneros
360'
SUtim J» JUT5TSSZ1 la u Xt|M tram
tlMW W MI'jrtUllM MUUIM.
lUtnoM, ¦< lUynaui a*i|lo«f«u at Umi
T«U« Mwimiti %4uli* flaw* la intt H"l Mil
Mliml «o MM fito af Craa^ Umiili, balav
final aaac of faua-
ttal la umiatmmt »»gxaiaatf taodataaaa la >iMI<
pull lawi baaatety |H<»iUml ta uaaaafanablaf
aail ¦¦paawf alaat 1"2S fra> Calarada Sirlots ca
Toablat (aaarallr oaterua I* vary U|k nail
yacaatlal; laa mil pataatlal la
am tut
FIGURE 5.
COMPOSITE STRATIGRAPHIC SECTION
FRONT RANGE URBAN CORRIDOR,
COLORADO
Source: Hart (12).
-------
5. Clays Naturally Flocculated (deposited under marine conditions) or
Dispersed (fresh water deposition):
(naturally flocculated clays more suitable; Mitchell (35))
6. Clay Mineral Composition of the Clay-Sized Particles:
(percent montmorillonite, MERL (2), Crim (19))
7. Miscellaneous:
(data concerning the structure of composition of the material that
may affect its use as a liner).
In addition to the geologic and mineralogic information, some engineering
data (such as liquid and plastic limits) for portions of the units in specific
areas are listed. A series of geologic engineering studies are on-going by the
United States Geologic Survey (USGS) for the area along the margin of the Front
Range from Boulder to Kassler, Colorado. Engineering maps and data will be
available from the USGS for selected quadrangles as they are completed.
It cannot be overemphasized that site-specific information is necessary.
None of the information provided here should be taken as adequate to approve
the use of a particular material for a remolded liner. The variability within
even the most uniform units is so great that tests should always be performed
on samples of the actual materials to be used.
Another factor that should be considered is the proper selection of the
substrata below a remolded clay liner. One of the failure mechanisms of a clay
liner referred to elsewhere in this report is "piping". A major cause of this
method of failure is the placing of a remolded c 1 ay liner on natural rock with
a considerably higher permeability. To avoid this failure, the permeability of
the substrata rock should also be tested (at least 10-15 percent clcty is
recommended and any weathered zone with jointing due to shrinkage should be
removed).
32
-------
B. Front Range Rock Units Suitable as Clay Sources
1. Pierre Formation
The distribution of the Pierre formation in the Front Range Urban
Corridor region is shown in Figure 4. The Pierre formation consists pri-
marily of dark gray, silty, carbonaceous claystones, shales and silt-
stones. It contains numerous layers of montmorillonite-rich bentonite
(volcanic ash derived) layers. It's thickness varies from 5000 feet in
the southern part of the area to 8000 feet in the northern part. In gen-
eral the formation is somewhat more sandy to the north. Individual clay-
stone intervals are up to 600 feet thick.
Of the formations under consideration here, only the Pierre is made
up of claystone layers that are laterally persistent over large areas
(10's to 100's of miles). Although the formation is very thick, it has
been thoroughly studied. Scott and Cobban (36) have been able to charact-
erize the fossil contents of numerous zones from the bottom to the top of
the unit. Using this fossil control, Gardner (37, 38) was able to estab-
lish subdivisions whose characteristics, particularly the clay mineral
content and engineering properties, area laterally persistent throughout
the Boulder-Denver area. Further work would most likely extend these sub-
divisions over larger areas of the State.
The following data are considered relevant to the suitability of
material derived from the Pierre Formation for use as a clay liner:
o Size Distribution. The Pierre Formation averages 40 to 70 percent
clay-size particles (less than 0.002 mm).
o Clay Minerals. One subunit (about 1000 feet thick in the Boulder-
Denver area) has calcium montmorillonite slightly more abundant than
mixed-layer illite-montmorillonite. In three other subunits the clays
are mixed-layer illite-montmorillonite which is slightly more abundant
than illite. Kaolinite and chlorite are present only in minor
amounts, if at all (Gardner, 37, 38). Bentonite clays are abundant in
the Pierre and are composed predominantly of montmorillonite.
33
-------
o Deposition Environment. In Colorado, the clays of the Pierre should
be in the naturally flocculated structure because most were laid down
in a marine environment.
o Soil Classifications. Soils derived from the Pierre claystones are
designated as CH, CL, MH, or ML in the Unified Soil Classification.
o Gypsum Content. Some claystones in the Pierre have associated gypsum
beds that may have adverse effects on their use as liner material.
o Weathering. The claystone units of the Pierre are commonly weathered
down to 15 feet below the land surface (subtle mineralogical changes
due to weathering have been detected 100 feet down (McGregor, (39)).
Jointing due to dessication is seen down to at least 5 to 10 feet.
The natural permeability of a solid piece of Pierre claystone is
irrelevant to the design of a containment facility sited on the unit
unless the upper jointed zone is removed. Planning of downhole perm-
eability tests should also take this weathered zone into account.
Engineering properties of the Pierre subunits for restricted areas are listed
in Table 1 to illustrate their variability.
2. Laramie Formation
The distribution of the Laramie formation in the Front Range Urban
Corridor region is shown in Figure 4. The Laramie formation consists of
both sandstone and claystone. The sandstones are relatively hard, white
to light tan in color and have abundant feldspar in some places and abun-
dant quartz in others. The claystones are dark gray and carbonaceous.
Thin lignite coal beds are present near the base of the formation. The
unit thickens to the east from 600 feet at Golden to 1000 feet east of
Hyatt Lake. Its lithology is highly variable from place to place accord-
ing to Van Horn (44). Beds in the Laramie are lens-shaped and, although
generalized zones have been traced for many miles, individualized beds
with these zones probably pinch out within a few miles. In the Golden
quadrangle, the lower part of the unit contains subequal amounts of sand-
34
-------
TABLE 1
VARIABILITY OF
THE ENGINEERING PROPERTIES
FOR THE
PIERRE
FORMATION
Potential Volume
Swell
Index
Plastic
Liquid
Plastici
Change Ratings
lbs/per Sq Ft
Limit
Li mi t
Index
Boulder Quadrangle1
3.4 - 7.4
2800
5900
1.3 - 3.4
1150
-
2800
-
-
-
0.7 - 2.8
700
-
2300
-
-
-
Eldorado Springs Quadrangle
2 3.4 - 7.4
2800
_
5900
_
_
1.3 - 3.4
1150
-
4000
-
-
-
0.7 - 2.8
700
-
2300
-
-
-
0.8 - 1.8
800
-
1500
-
-
-
Golden Quadrangle^
4.8 - 8.4
3800
_
7070
_
_
3.5 - 7.7
2850
-
5700
-
-
-
1.5 - 6.2
1300
-
4900
-
-
-
Pueblo Quadrangle4
-
1650
+
-
32
2
-
3500
-
6100
-
41
15
-
2450
-
4500
-
57
9
-
1800
-
2700 +
-
24-43
8-18
-
2600
-
3000
-
37-47
12-21
Sources: ^-Gardner (37).
^Gardner (38).
•^Gardner and Hart (41).
4Scott (42).
35
-------
stone and claystone. Claystone predominates in the upper part and becomes
relatively more abundant to the east across the quadrangle. Coal beds are
present in the lower portion of the unit.
The highly variable nature of this unit makes it unwise to attempt to
project laterally any data derived concerning grain size distributions,
clay mineral percentages, or relative abundances of clay mineral
species. Only data derived from the actual material present on the site
being considered should be used.
The following information is considered relevant to the suitability
of material derived from the Laramie formation for use as cle(y liners:
o Size distribution. Qualitative descriptions suggest about 50 percent
clay-sized particles (less than 0.002 mm) are present in the lower
portion and more than 90 percent in the upper portion.
o CIeo' minerals. Montmorillonite, kaolinite and illite are present.
Guide (43) identified three subunits based on the relative abundances
of the clay mineral types:
Upper - kaolinite dominant; about 1/3 illite.
Middle- montmorillonite dominant; kaolinite and illite subequal.
Lower - kaolinite dominant; minor illite present.
Subsequent work by Van Horn (44) did not fully support these subdivi-
sions. His results indicate that within the upper part of the unit, montmoril-
lonite predominates. Van Horn also indicates that the clays of the Laramie
were laid down in fresh water, and should therefore have a dispersed structure.
Very little data is available on the engineering properties of the Lara-
mie formation. Private consulting firms specializing in soils have site-spec-
ific data on the Laramie that could probably be obtained by contract. The
swelling potential of clays and soils of the Laramie are discussed in Van Horn
(44) and Gardner and Hart (41) for the Golden quadrangle and in Hart (12, 45)
36
-------
for the Front Range Urban Corridor. The only published numerical values are
reported in Gardner and Hart (41) and indicate a range in the potential volume
change rating from 1.5 to 6.2 with swell indices varying from 1300 to 4900
pounds per square foot.
3. Dawson Formation (Subdivided into the Denver and Arapahoe Formations)
The distribution of the Dawson formation (and that of the Denver and
Arapahoe formations) is shown in Figure 4. Where the Dawson formation is sub-
divided the upper part with abundant feldspar and volcanic rock fragments is
called the Denver formation and the lower, more quartz-rich part is called the
Arapahoe formation. The Arapahoe and the lower part of the undifferentiated
Dawson formation is comprised primarily of sandy and silty claystone with
lesser amounts of interbedded quartz-rich sandstone and conglomerate Because
the Arapahoe formation is less commonly exposed, only the upper part of the
undifferentiated Dawson and the Denver formations are considered further here-
in. The Denver formation and the upper part of the undifferentiated Dawson
formation also contain a considerable amount of silty and sandy claystones. In
the interbedded coarser beds of sandstone and conglomerate there is an abund-
ance of feldspar and volcanic rock fragments that readily become clays where
exposed to weathering. Another type of rock that is commonly present in these
units is a sandy to conglomeratic mudstone in which the relatively sparse
coarser material floats in a muddy matrix. The Dawson formation is generally
1100 feet thick. In the Golden quadrangle, the Denver formation has been par-
tially removed by erosion and is about 800 feet thick. In the Englewood quad-
rangle, the thickness of the Denver varies from 270 to 900 feet thick. A large
volume of volcanic debris is characteristic of the Denver. Its lenticular beds
of sediments were deposited by ash falls, mudflows and by fresh water running
over the land. The unit equivalent to the Denver formation south of Golden is
the upper part of the Dawson formation. It, however, contains considerably
less volcanic material.
The following data are considered relevant to the suitability of material
derived from the Dawson and the Denver formations for use as cley liners:
37
-------
o Size Distribution. The variability of the composition of the units
does not make averages meaningful. In the Highlands Range quadrangle
south of Denver, a claystone unit in the Dawson is 4 to 70 percent
clay-sized (less than 0.002 mm). Claystone in the Denver formation is
12 to 55 percent clay-sized.
o Size Frequency (soil particles). The extreme size-frequency variation
of the particles making up the soils derived from the Dawson formation
in the Castle Rock area of Douglas County reported by Soil Conserva-
tion Service (46) reflect the variability of the parent materials.
The percentage of materials less than 0.074 mm in size (silt plus
clay) varies from 7 to 38 percent; the percentage less than 0.002 mm
in size (clay size) varies from 0 to 26 percent.
o Clay Minerals. In the Golden quadrangle, the samples of the Denver
contained abundant montmorillonte and no other types of clay. Dawson
formation samples' from the quadrangle also contain abundant montmor-
ollinte, but kaolinite is also present and its abundance is nearly
equal to the montmorillonite in some samples. IIlite was only present
in one sample. One analyzed sample from a claystone unit of the Daw-
son in the Highlands Ranch quadrangle south of Denver contained 65
percent clay, 2/3rds of which was montmorillonte, nearly l/3rd kaoli-
nite, and minor amounts of illite.
o Deposition Environment. The clays of the Dawson formation and the
Denver formation, which were deposited under fresh water, are likely to
be dispersed and therefore less suitable for remolding as a liner.
The structure of clays of ashfall and mudflow origin are unknown.
o Soils. Soils derived from the claystone of the Denver formation of
the Englewood quadrangle are designated CL and CH (Shroba, (47)). In
the Golden quadrangle, soils formed on the Denver formation are MH,
CH, and CL (Gardner and Hart (41)). Soils in the Castle Rock area of
Douglas County derived from the Dawson formation are designated SC,
SM-SC, SM, and SM-SP (Soil Conservation Service (46)).
38
-------
Engineering properties of portions of the Dawson and Denver formations for
the Golden, Englewood, and Highlands Ranch Quadrangles, and for the Castle Rock
area are listed in Table 2 to illustrate their variability.
39
-------
TABLE 2
VARIABILITY OF THE ENGINEERING PROPERTIES FOR THE DAWSON AND DENVER FORMATIONS
Potential Volume
Change Ratings
Swell Index Plastic
lbs/ per Sq. Ft. Limit
Liquid
Li mi t
Plastici ty
Index
Golden Quadrangle
Denver Formation*
2.8-7.7
2500-6100
Englewood Quandrangl
Denver Formation2
2900-19,000
0- 2500
0- 600
23-40
21-26
NP*-29
45-99
32-53
35
24-59
9-31
NP*-7
o
Highlands Ranch Quadrangle
Dawson Formation-*
Denver Formation-*
Castle Rock Area, Douglas County
Dawson Formation
Soils4
0-5.7
0-3.8
1.2-8.0
1.5-4.5
1.4-9.0
2.8-9.0
0- 4500
0- 3100
1100- 6550
1400- 3500
1250- 9900
2300-11,400
0-36
0-37
25-56
26-45
0-45
28-59
22-28
2- 6
NP*-16
0-51
21-56
37-87
40-65
43-74
51-95
46-51
24-26
NP*-34
0-23
0-23
9-31
3-16
14-66
11-52
*NP = Non-plastic
Sources:
^Gardner and Hart (41).
2Shroba (47).
^Mayberry and Lindwall (48).
4Soil Conservation Service (46).
-------
IV. STATE OF COLORADO'S LIST OF PROPOSED CHEMICALS
The State of Colorado has developed a list of ten chemicals which they
would like to include in their evaluation program. The ten chemicals are:
o Benzene
o 1,2 Dichloroethane
o 1,1,1 Trichloroethane
o 1,1 Dichloroethane
o 1,1 Dichloroethylene
o Methylene chloride
o Perchloroethylene
o Trichloroethylene
o Xylene
o Carbon tetrachloride
The ten chemicals are quite typical of organic solvent wastes requiring
disposal. Of the ten, however, only four have been studied in ar\y of the clay
liner permeability experiments identified in the literature. These four are:
benzene, trichloroethylene, xylene, and carbon tetrachloride. All four were
investigated in the studies conducted by Lee (5).
Lee found that xylene, benzene and carbon tetrachloride all induced liner
failure in at least one of the test clays studied. Carbon tetrachloride had
the greatest tendency to cause clay shrinkage with the resultant transmission
of fluids in bulk. This was observed as early as six days after initiation of
the experiment. Benzene and xylene also induced clay shrinkage, with solvent
breakthrough occuring after eight days and 24 days for benzene and zylene,
respectively. These results indicate that from the state's list of ten chemi-
cal carbon tetrachloride, xylene, and benzene all have a high potential for in-
ducing liner failure. However, since the majority of the other chemicals on
the list have not been tested in previous investigations, it is quite possible
they could induce similar deleterious effects. Because of this distinct possi-
bility, it is impossible at this time to rank the chemicals based on their
potential for inducing liner failure.
41
-------
Lee's initial work also indicated that a clay liner's permeability to an
organic solvent can, in some instances, be correlated with the hydrophilic na-
ture of the solvent, in this case represented by the dielectric constant and
the octanol/water partition coefficient of the solvent. The viscosity and den-
sity of a solvent can also influence the rate of flow of a solvent through
earthen materials as well. The dielectric constant, viscosity, and density of
organic compounds can be obtained from standardized chemical reference tables
(7, 8) while the octanol/water partition coefficient is generally obtained by
empirical observation. Table 3 lists the dielectric constant, density, and
viscosity for the ten chemicals which the State of Colorado would like to in-
clude in their evaluation program.
As can be seen from Table 3, xylene, benzene, and carbon tetrachloride all
have relatively low dielectric constants (2.2-2.5 range) and all three were
implicated by Lee (5) as having a high potential for inducing clay-liner fail-
ure through shrinkage. Lee's empirical evidence therefore supports the general
statement that highly hydrophobic solvents have a relatively high potential for
inducing cley liner failure. This typically occurs via an increase in liner
permeability resulting from the development of shrinkage channels. Solvents
can therefore be transported in bulk through the clay. In this regard, the
solvents which are listed first in Table 3 empirically pose the highest risk of
detrimentally impacting an earthen liner.
It is important to realize, however, that the current state-of-the-art
does not fully support any predictive relationships which may exist between the
permeability of earthen materials and the physical properties of organic sol-
vents. Because the correlation between the permeability characteristics of
test clays and the physical properties of organic chemical compounds is some-
what speculative, it is mandatory at this time that each chemical waste-clay
liner combination be tested individually. Therefore, although certain com-
pounds pose a higher potential risk of inducing liner failure (i.e., those
appearing at the beginning of Table 3), it is important that all of the com-
pounds be tested individually before conclusions regarding liner compatibility
are made.
42
-------
TABLE 3
DIELECTRIC CONSTANT, DENSITY, AND YISCQSITY VALUES
OF SELECTED ORGANIC SOLVENT^
Dielectric Constant Viscosity
Compounds (Dimensionless) Density (g/cm3) (centipoises)
Carbon Tetrachloride
2.24
1.5844
0.845
Benzene
2.28
0.8790
0.649
Perchloroethylene
2.30
1.6064
0.798
Xylene (o-isomer)
2.57
0.8802
0.809
Trichloroethylene
3.42
1.4514
0.532
1, 1 Dichloroethylene
4.602
1.2132
0.358
1, 1, 1, Trichloroethane
7.53
1.3376
0.795
Methylene Chloride
8.93
1.3168
0.393
1, 1, Dichloroethane
10.00
1.1680
0.505
1, 2 Dichloroethane
10.36
1.2531
0.730
NOTE: All values stated at approximate room temperature.
Sources: ^Riddick and Bunger (8).
^Chemical Rubber Co. (7).
43
-------
In evaluating the State's selections, we conclude that the State may wish
to add a highly acidic and a highly basic waste to the list. Extreme pH solu-
tions, such as inorganic acids or caustics, can be potentially detrimental to
earthen liners and may also be quite prevalent in a hazardous waste stream. In
keeping with the general theory of Battelle (12), and Crim and co-workers (19)
that extreme pH solutions in contact with clay liners warrant close scrutiny,
the State may wish to consider the addition of an inorganic acid and/or a base
to the proposed list. Sulfuric acid (H2SO4) and sodium hydroxide (NaOH) would
be two likely candidates that the state may wish to add to the list. Sulfuric
acid is commonly used in the mineral extraction industry (acid-leach process)
while sodium hydroxide is widely used throughout the chemical and petroleum
industries.
44
-------
V. TEST METHODS FOR EVALUATING COLORADO'S CLAYEY SOILS
This section presents the evaluation procedures which are necessary to
determine the flow characteristics of Colorado soils during exposure to
various hazardous wastes. In this section, we have individually listed the
various test procedures used by the major research groups which supplement the
permeability tests. Where available, the standard references are given. We
then make recommendations as to which tests are appropriate for inclusion in a
comprehensive evaluation program. For comparison, the test methods initially
proposed by the EPA to establish soil liner criteria are listed below (1).
Soils used as liners shall:
o be classified under the Unified Soil Classification System as CL,
CH, SC, OH (ASTM D2487-69);
o allow more than 30 percent passage through a Mo. 200 sieve (ASTM D
1140);
o have a liquid limit of 30 or greater (ASTM D423);
o have a plasticity index or 15 or greater (AST D424);
o have a pH of 7.0 or higher (no method given);
o have a permeability of 1x10"? cm/sec or less (ASTM D2434 constant
head test); and
o have a permeability not adversely affected by the waste (no method
gi ven).
A. Description of the Tests Identified in the Literature
1. Studies by G. Fred Lee
In addition to the coluim permeability tests, described in detail in
Appendix A, Lee performed the following analyses:
45
-------
o Clay mineralogy determination (X-Ray diffraction methods of Pierce
& Siegel (49), and Biscayne (50));
o Particle size analysis (Lee (5));
o Moisture-density relationship (Harvard Miniature Compaction
Apparatus - Soiltest, Inc. (13);
o Atterberg limits (AASHO* T-89, T-90);
o Specific gravity (Lee (5));
o Cation exchange capacity (Busenberg and Clemency (51));
o Total carbon analysis (Lee (5));
o Swell properties (consolidometer manufactured by Karol-Warner,
Inc. (11));
o X-Ray diffraction of clays following contact with the organic
wastes (Lee (5));
o Properties of the organic liquids (dielectric constant, octanol/
water partition coefficient).
2. Studies by Kurt Brown
Appendix B contains the methodology, as reproduced from MERL (2), which
is being used by Brown in his current permeability studies. In addition to the
coluirai tests, MERL recommends the following analyses:
* American Association of State Highway Officials
46
-------
o Atterberg limits, using distilled water and the waste liquid of
interest (ASTM D423 and D424);
o Moisture-density relationships (ASTM D698 and D1561);
o Soil strength characteristics (triaxial compression test, direct
shear, vane shear, California bearing ratio);
o Dispersion potential (ASTM (30)).
3. Battelle Study
In addition to their constant head permeability tests, Battelle ran the
following analyses:
o C1 ay texture, fineness, liquid limit, plasticity (ASTM 1978
methods);
o Clay dispersion crumb test (Sherard et al. (18));
o Mineralogical tests (x-ray diffraction analysis, scanning electron
microscopy, x-ray fluorescence);
o Effluent chemistry ( pH, sulfate, aluminum, silicon).
4. Colorado State Geotechnical Engineering Program Study
In addition to the permeability experiments, Crim and co-workers performed
several geochemical tests prior to and after the permeability tests. The texts
were performed in accordance with methods published in the USDA Agricultural
Handbook #60 (52). The tests included:
o Cation exchange capacity (CEC);
o Exchangeable cation concentration (ECC);
o Exchangeable sodium percentage (ESP);
o Lime equivalent as a percentage of CaC03.
47
-------
The clays were then evaluated for moisture-density relationships using
ASTM D-698. Crim and co-workers also monitored the chemistry of the effluent
and the permeant for the following constituents: calcium, magnesium, sodium,
aluminum, iron, and pH.
B. Recommendations
1. Permeability Tests
The literature review has presented two generally acceptable approaches
to determining the permeability characteristies of low-permeability earthen
materials to chemical wastes. The apparatus used by Lee, and discussed in
Appendix A, appears to be somewhat more applicable to determining the per-
meabilities of clays to relatively volatile solvents, since the coluim has
special provisions for minimizing evaporative losses. Furthermore, the column
set-up allows for relatively accurate measurements of flew rates, since it uses
the falling-head technique, thereby eliminating the need for a separate efflu-
ent collection device to measure flow. The major disadvantage is that the
equipment is not readily available and must be fabricated by the investigator.
The equipment used by Brown, discussed in Appendix B, is available from
commercial sources, thereby eliminating the need for in-house fabrication. The
major disadvantage is that the apparatus requires a separate effluent fraction
collector, which could increase the chance for slight evaporative losses of
volatile solvents prior to measurement. The technique, discussed in Appen-
dix B, does take this into consideration; however, the mitigating measures pre-
sented may still not generate as reliable flow measurements as can be achieved
with Lee's falling head technique. Brown's method does have the added advan-
tage of allowing compaction of the sample directly within the permeameter
chamber, rather than requiring a transfer step from a compaction mold to the
permeameter, as is needed with Lee's method. This is important to achieve an
effective seal between the soil sample and the walls of the permeameter column,
thereby reducing the possibility of channelized flow along the walls of the
column. Lee, however, has not alluded to any problems of this nature with his
set-up.
48
-------
In our opinion, either method could be used by the State of Colorado,
since both have individual specific advantages, and both are accepted in
principle by the EPA. It does appear that Lee's method may be more applicable
to extremely volatile solvents, since the method has a better chance of
minimizing evaporative losses. Lee's permeameter will also be mare resistant
to corrosive liquids (those of extreme pH) since it is fabricated from glass.
As discussed in Section IV, however, the State's concern lies in evaluating
organic solvents only, so resistance to extreme pH solutions is not a primary
consideration unless acids or bases are added to the list.
Using either method, reasonable results can generally be obtained within
the time period of a month or less. Brown's recommendation of the passage of
at least one pore volume of permeant appears sound; in typical cases where a
failure was observed, it happened in less than 15 percent passage of the pore
volume. Therefore, passing at least one pore volume of fluid through the
column is considered conservative. Using pressurized nitrogen to increase the
rate of flow through the clay, one pore volume of fluid can permeate the sample
in less than a month; we conclude this is adequate.
2. Supplemental Tests
In addition to the permeability evaluations, we recommend, at a
minimum, the following additional tests be performed on the test soil:
o Determination of soil class using the Unified Soil Classification
System (ASTM D698 and D1561);
o Determination of grain size (STM D1140 and Lee (5));
o Determination of Atterberg limits (ASTM D423 and D424);
o Moisture-density relationships (ASTM D698 and D1561);
o pH.
49
-------
These tests will establish the basic physical properties of the earthen
materials in addition to the permeability characteristics. They will also sat-
isfy those analyses necessary to meet the proposed EPA liner criteria, as sited
by Fung (27). However, if the State has access to some additional equipment,
several further tests would be helpful as they will yield additional, supple-
mentary information. The tests are:
o cl^y minerological analysis by x-ray diffraction (Pierce and Siegel
(49), Biscayne (50));
o clay dispersion tests (ASTM (30));
o additional Atterberg limits analyses employing the waste liquid
rather than distilled water (MERL (2));
o shrink/swell behavior (consolidometer method of Lee (5)).
The x-ray diffraction analyses are highly recommended since they are im-
portant in determining the relative amounts of the three major clay minerals,
kaolinite, montmorillonite, and illite, which actually comprise the clay frac-
tion of the test soil. Battelle (15) has also found x-ray diffraction to be
somewhat useful technique for determining the presence of minerological changes
in the clay following exposure to the acidic test solutions used in the study.
Battelle generally found that the diffraction analyses could be used to deter-
mine changes in the relative abundance of mineral constituents, and to identify
reductions in mineral crystallinity. Neither of these post-exposure mineralo-
gical changes were very extensive, however, under the conditions of the study.
The other three tests listed above can also be considered highly comple-
mentary to the recommended permeability studies. The cl^y dispersion tests,
and the Atterberg limits analyses will aid in determining the "sensitivity" of
the test soil to the chemicals of interest; major changes in the physical pro-
perties and/or behavior of the soil following exposure to the wastes can be
readily observed utilizing these procedures. The shrink/swell consolidometer
studies are important since Lee (5) found that the major cause of clay-liner
failure following exposure to the organic solvents was the excessive shrinkage
50
-------
of the clay. This shrinkage resulted in the quick passage of fluids through
the sample along shrinkage cracks and channels. Since the shrink/well behavior
of the clays can be readily determined in the consolidometer chamber, this test
serves as a useful, supplementary indication of a clay liner's ability to re-
tard the passage of organic fluids. It is therefore recommended that the
shrink/swell consolidometer analysis be performed concurrently with the permea-
bility test.
C. Summary and Conclusions
The crucial issue, in performing any of the permeability tests, is to pay
close attention to those factors identified in the literature which can induce
experimental error. These include temperature fluctuations, inaccurate and/or
non-intensive measurements of flow rates, and volatile loss of permeating sol-
vents. The measurement of flow rates below 10~7 cm/sec is a difficult determi-
nation, and careful technique is required. It is our opinion that the modifi-
cations by Lee and Brown to the standard ASTM permeability determination are
warranted, since the ASTM test is relatively simplistic and intended for more
permeabile materials. Particular attention to the details of the modified test
procedures is required to ensure that reproducible results below 10"? cm/sec
can be achieved.
It is also in the State's best interest to maintain contact with Dr. Kurt
Brown at Texas A&M University, since his work is ongoing and new results may be
available shortly. Helpful suggestions on technique can also be obtained from
Dr. G. Fred Lee of Colorado State, although his research is completed. We also
recommend obtaining a number of documents referenced in this report. At a
minimum, these include:
o the MERL design manual;
o G. Fred Lee's final report to the EPA Kerr Lab in Ada, Oklahoma;
and
o The Battelle-Pacific Northwest Laboratory's final report submitted to
the Nuclear Regulatory Commission.
51
-------
Many of the other documents referenced in this report contain valuable informa-
tion as well, particularly on background material.
Finally, we conclude that the information available to date indicates that
the State's planned efforts to evaluate the permeability characteristics of
native soils is a valid and timely proposal. No general conclusive statements
are yet available about the behavior of earthen liners exposed to chemical
waste. Therefore, comprehensive testing is mandatory at this point in time, so
that meaningful conclusions regarding the use of a particular soil for lining
purposes can be made.
52
-------
APPENDIX A
COEFFICIENTS OF PERMEABILITY FOR LIQUIDS AND LIQUID MIXTURES
IN CONTACT WITH CLAYS*
In this stucjy the coefficients of permeability were determined on remolded
clay using a laboratory apparatus with heavy-duty permeameter columns. (The
colunm is shown in Figure A-l.) These permeameters were designed and built
with Pyrex glass in the University of Texas-Dallas machine shop especially for
this study. All joints in the columns were lined with Teflon to make them
suitable for use with organic solvents. The procedure followed for determining
the coefficient of permeability is described as follows:
About 1000 g to 1200 g of well-ground (pass through a No. 4 sieve) air-
dried sample were weighed and transformed to a mixing pan. The required amount
of deionized water corresponding to the optimum moisture content, was added to
the clay sample. After mixing, the sample was placed in a glass jar, fitted
with Parafilm, and stored overnight. For samples that mixed readily with water
and had low dry strength, it was satisfactory to add water and mix the specimen
immediately prior to testing. It is important that a compacted specimen not be
remixed and used over again. With the mold and collar clamped to the base,
about half of the mixed clay was placed in the mold. The surface was leveled
by pressing lightly with a wooden plunger. The sample was then compacted with
a 5.5 pound (2.5 kg) standard hammer to conform to the standard compaction test
procedures (12,375 feet pound per cubic foot (5.9 x 10^ joules/m^) of compacted
volume. About 10 to 20 percent additional energy was applied depending on the
amount of excess volume of compacted clay. Actual moisture content of the clay
sample was determined separately, by oven drying a duplicate sample at 110°C
for 24 hours. The excess clay in the mold was scraped and leveled carefully up
to the 2 inch (5 cm) mark depth. The molded clay was read|y for testing in the
permeability column, which was then assembled and test fluid introduced. Final
fluid level in the graduated standpipe was adjusted very precisely with the
help of a pipette bulb. Temperature measurements of the fluid were taken with
* Reproduced from Lee (5).
53
-------
FIGURE A-1
LEE'S PRESSURE PERMEABILITY COLUMN
99 cm
SOIL
sample
14 cm
i1 » ¦ .
12 cri
High Pressure
Stainless Steel
Valve
— To Nitrogen
Tank
Liquid Level
8 mm Graduated Stanpipe
High Pressure Swagelock
Fitting (Teflon)
CI amp
1.8 cm Thick X 10 cm ID
Pyrex Glass
Teflon Gasket
High Pressure Joint
Whatman GF/A Fiberglass
FiIter
40 Mesh Stainless Steel Screen
6 cm Pyrex Glass Support
NOT TO SCALE
|To Collection
54
-------
a thermometer immersed in a reagent bottle filled with the fluid. The open end
of the standpipe was firmly covered with two layers of Parafilm. Using this
technique, evaporation of solvent was reduced to an insignificant amount. If
ar\y clay-fluid system had a coefficient of permeability less than 10~10 cm/sec,
the open end of the standpipe could be attached to a pressurized nitrogen tank
and a constant pressure in the range of 10 to 50 psi could be maintained. The
coefficient of permeability 'K' in cm/sec was computed as:
K = QL/AH
where Q is the flow of the percolate in ml per second, L is the height of the
sample in the colunm in cm, A is the cross sectional area of the sample in
square cm, and H is the average head of the fluid medium on the sample in cm.
55
-------
APPENDIX B
TEST METHOD FOR THE PERMEABILITY OF COMPACTED CLAYEY SOILS*
(CONSTANT ELEVATED PRESSURE METHOD)
Introduction
To assess the suitability of compacted clayey soils for the lining of
waste disposal facilities, the primary laboratory measurement is saturated
hydraulic conductivity or permeability. Such a measurement should be made on a
specimen of the soil that has been remolded and compacted in the range of
optimum moisture content to achieve the maximum density possible for a given
compactive effort. STM Method D698 should be use for determining a soil
moisture-density relationship.
Clays compacted at optimum moisture content have the potential to reach
permeability values as low as 10-1^ cm/sec. However, permeability values in
the range of 10"^ cm/sec to 10"^ cm/sec are more probable.
Testing should be continued until the permeability stabilizes, which may
require the passing of one or more pore volumes of a standard 0.01N CaS04
leachate. For these reasons, it is necessary to use increased pressure to in-
crease the hydraulic gradient and reduce the time length of the test. Trapped
air is a common cause for artificially low permeability values. An increase of
pressure reduces air trapped in the core by increasing the weight of gas that
will dissolve in water flowing through the specimen. The higher pressures will
also reduce the volume of the remaining air pockets. Backpressuring is neces-
sary in very low permeability soils to dissolve air in the specimens. The use
of deaerated water is required. The tests must also be performed in a constant
temperature environment.
For a given waste-clay liner combination, there are two fluids that may
alter the permeability of the liner: (1) the fluids present in the waste (pri-
mary leachate) and (2) the fluids generated by water percolating through the
waste (secondary leachate). When only solids are placed in the confinement,
*Reproduced from MERL (2). Information in brackets has been added by Fred C.
Hart Associates, Inc.
56
-------
only secondary leachate need be considered. The permeameter used for testing
the influence of leachate on the permeability of the cley liner must be capable
of safely operating with hazardous materials including industrial solvents,
volatile compounds, corrosive acids, and strong bases.
Soil permeability tests must be carefully performed if they are to be ac-
curate. Leaks, volatile losses, or channel flow along the interface of the
permeameter and soil will greatly affect permeability values.
To avoid channel formation, the clay should be allowed to seat at low
pressure. By allowing two to three inches of standard leachate (0.01 N CaS04
or CaC12) to stand on the core for 24 hours, an effective seat is obtained for
the top few millimeters of the clay core. This thin layer will prevent bulk
flow and the rest of the core should adequately seal when the leachate is forc-
ed into the soil at elevated pressures.
In order to facilitate ease of duplication of the test apparatus, the per-
meameter is based on readily available and easily modified components. Figure
B-l illustrates the modified compaction permeameter which is based on the stan-
dard compaction permeameter that is available through most soil testing supply
houses. All components are in common with the standard permeameter except for
the enlarged fluid chamber, extended studies, and high pressure fittings.
General Comments
For use with fluids other than water, all gaskets should be Teflon. To
avoid leakage around the gaskets, all metal surfaces against which the gaskets
are seated should be wiped clean of grit. All components should withstand con-
tinuous operation at pressures up to 60 psi.
To limit the volume for diffusive mixing of leachate samples after they
have passed through the clciy core, the fluid outlet part should be fitted with
an adapter to a small (1/8 inch inside diameter) Teflon tube. The use of
translucent Teflon at the permeameter outlet provides a convenitent window with
which to monitor the expulsion of entrapped air. Standard leachate should be
57
-------
FIGURE B-1
PERMEAMETER CHAMBER USED BY MERL
—Pressure Intake
§—Pressure Release-
Chamber
Soil
Chamber
Top Plate
—Ring seal
—Extended stud
v/wp&m
L- Porous stone insert
-Base Plate
I/8 inch Teflon tubing
-Outlet to fraction collector
58
-------
passed through the permeameter until there are no air bubbles visible in the
outlet tubing. If soil piping occurs, eluted soil clays will be visible either
clinging to the inside walls of the outlet or as a suspension in the collected
flow samples.
Volatile losses may ocur during sample delivery from the outlet tubing to
the sample bottles in the fraction collector. To limit these volatile losses,
the top of each sample collection container should be fitted with a long stem
funnel and the fraction collector should be placed in an air-tight cooled com-
partment. This is also desirable where the fluids being tested may present a
health hazard to exposed laboratory personnel.
When volatile hazardous chemicals are used, the entire test apparatus
should be fitted into a vented hood. This precaution is insurance against
worker injury in case of a gasket blow out. If testing of several cores simul-
taneously is desired, each pressure in a must have its own cut-off value to
prevent the complete shut down of the test in the case of loss of pressure in a
single specimen. With several specimens producing leachate, it may be desir-
able to have an automatic fraction collector. This is especially useful with
long-term tests. Over the course of a one-month test, it may not be convenient
to change the sample bottles every few hours, twenty-four hours a day.
Calculations
This test is to determine the intrinsic permeability of a compacted clay
by a constant-elevated pressure head method for the flow of any permeant
through compacted clay soils. The equation applicable to the test is Darcy's
Law as modified to normalize permeability values as they are affected by the
permeant's viscosity and density:
k = V n 1
p g A t (L + H)
k = intrinsic permeability (cm2)
V = volume of flow (cm^) in time t
n = permeant viscosity (dyne sec. cm-2)
p = permeant density (g cm~3)
59
-------
g = gravitational constant (cm sec~2)
A = cross-sectional area of flow (cm^)
t = time (sec.)
L = length of soil core (cm)
H = pressure (cm of H£0)
Equipment Requirements
1. Soil crusher (C-2 Laboratory Crusher)
2. Soil grinder (Hewitt Soil Grinder)
3. 2 mm sieve (CB-810 brass sieves)
4. Moisture cans (LT-30 tin sample boxes)
5. Balance capable of weighing 20 kgs. (L-500 heavy duty balance)
6. 105°C drying oven to determine water content of soil samples
7. Compaction molds (CN-405 Standard Compaction Mold)
8. Compaction hammer (CN-4230 Mechanical Compactor)
9. Steel straight edge
10. Permeameter bases and top plates (K-611 Permeameter Adapter)
11. A source of compressed air with a water trap, regulator and pres-
sure meter.
12. A fraction collector with automatic timer for collection of samples
over time (Brinkmann Linear II Fraction Collector with a multiple
distribution head)
13. An air tight, cooled chamber to limit volatile loss of samples dur-
ing and after sampling
14. A vented hood to hold the compaction permeameters and chamber con-
taining the fraction collector. (This is a safety precaution to
limit exposure of laboratory personnel to the hazardous chemicals
used in the studies.)
Note: Equipment in Items 1, 3, 4, 5, 7, 8, and 10 can be obtained from Soil
Test, Inc.; equipment in Item 2 can obtained from B. Hewitt Welding and Re-
pair.
60
-------
Test-Procedure
1. Obtain sufficient clay soil to be tested. Break the soil down to
golf ball size clods and lay them out to air dry.
2. Grind the air dried soil and pass it through a 2 mm sieve.
3. Mix the sieved soil thoroughly, and divide into two lots of equal
weight. Each lot of soil should be placed in air tight containers at
room temperature until the time of use. Each lot should be enough
soil to prepare up to 10 compaction molds (50 kg. will provide for
some spillage losses assuming a mold volume of about 2,000 cm3).
4. Use one lot of the soil to determine the moisture/density relations
of the soil by following the STM Method D-698.
5. Use a second lot of the soil to prepare compaction molds at optimum
moisture content.
6. Fit a valve on top of the permeameter top plate with pressure fit-
tings and connect it to a source of air pressure via copper tubing.
Place a water trap, pressure regulator and pressure gauge in line be-
tween the air pressure source and permeameter. The water trap should
go between the pressure source and regulator to prevent build up of
debris on the membrane in the regulator. The pressure gauge should
be locate between the regulator and a pressure manifold to the perme-
ameters so that the hydraulic head being exerted on the clay cores
mc\y be monitored.
7. Place sufficient volume of the leaching solution in the chamber above
the compacted soil.
8. Apply pressure to force at least one pore volume of the standard
leachate (0.01 N CaS04 or CaCl) through the clc(y cores. After the
intrinsic permeability values are stable and less than lO"1^ cm^
(equivalent to permeability value of 10~5 cm sec-1 at 25°C,) release
the pressure, disassemble the permeameter and examine the core.
61
-------
9. If the clay core has shrunk, it is unsuitable as a clay liner.
10. If the clay has expanded into the upper mold, remove the excess soil
with a straight edge by cutting so as to not smear the cletf surface.
Reweigh the core to determine its density and then remount it on the
permeameter.
11. Repressure the permeameter and pass leachate until the intrinsic
permeability value stabilizes again less than 10~10 cm2 equivalent to
a permeability value of 1.1. x 10"^ cm see"* at 25°C.
12. Remove the remaining standard leachate from eight of thhe fluid cham-
bers and replace it on duplicate cores with each of two wastes or
waste leachates to be tested.
13. If after passage on one pore volume of the various leachates, the in-
trinsic permeability values of the cores are still above 10~10 cm2,
dissasemble the permeameter and re-examine the cores.
14. If the clay core has shrunk, it is unsuitable as a clay liner for
that waste.
15. If the clay core has expanded, repeat step 10 then proceed to step
17.
16. If the clciy has not changed volume, remount it on the parameters.
17. Repressurize the permeameter and pass at least one pore volume of the
standard leachate. If its intrinsic permeability has climbed above
10"10 cm2 (ca. 1.1 x 10"5 cm s"1) the clay is not suitable for
containing the waste. If the intrinsic permeability values measured
on a waste's primary and secondary leachate have consistently stayed
below 10~10 cm2, proceed to Step 18.
18. Examine the translucent Teflon outlet tube for signs of soil particle
migration out of the core. If there is evidence of soil migration,
pass at least one more pore volume to observe if this internal
62
-------
erosion of the core continues. It it continues after the two pore
volumes of standard leachate have passed, the clay is unsuitable for
that waste. If the soil migration stops, at least one pore volume of
the standard leachate should be passed to assure that the core
stabilization is permanent and then proceed to Step 19.
19. If there is no sign of soil migration, depressurize the system and
extrude the clay cores from their molds to examine them for signs of
cracking, internal erosion, soil piping, clay dissolution, structural
changes, or any other difference from the control cores (those having
received only standard leachate).
63
-------
REFERENCES
1. USEPA, Federal Register, December 18, 1978. pp. 58946-59028.
2. USEPA, Lining of Waste Impoundment and Disposal Facilities. Municipal
Environmental Research Laboratory, Cincinnati, Ohio, SW-870, September,
1980.
3. Sowers, G.F., Introductory Soil Mechanics and Foundations: Geotechnical
Engineering, 4th Ed. Macmillan Co., Inc., New York, 1979.
4. Brown, K.W., and Anderson, D., Effect of Organic Chemicals on Clay Liner
Permeability - A Review of the Literature. Proc. of the 6th Annual Re-
search Symposium on the Disposal of Hazardous Wastes, EPA 600/9-80-010,
March, 1980.
5. Lee, G.F., Green, W.J., and Jones, R.A., Impact of Organic Solvents on the
Integrity of Clay Liners for the Industrial Waste Disposal Pits: Implica-
tions for Groundwater Contamination. Final report submitted to the Robert
S. Kerr Environmental Laboratory, Ada Oaklahoma, June, 1979.
6. ASTM, Special Procedures for Testing Soil and Rock for Engineering Purpos-
es. American Society for Testing and Materials, Philadelphia, 1970.
7. The Chemical Rubber Co., Handbook of Chemistry and Physics. 47th ed., CRC
Publishers, Cleveland, 19^i
8. Riddick J.A., and Bunger, W.B., Organic Solvents. Wiley-Interscience, New
York, 1970.
9. Poulovassilis, A., "The Changeability of the Hydraulic Conductivity of
Saturated Soil Samples". Soil Science, 113:81, 1972.
10. Davis, S.N. and DeWiest, J.M., Hydrogeology. John Wiley and Sons, Inc.,
New York, 1966.
11. Karol-Warner, Inc., Highland Park, N.J., Manufacturer.
12. Hart, S.S., Potentially Swelling Soil and Rock in the Front Range Urban
Corridor, Colorado. Colorado Geologic Survey, Inv. Geol. #7, 1974.
13. Soiltest, Inc., Soiltesting Equipment Catalog. Evanston, 111.
14. Brown, K.W., Texas A & M University, Soil and Crop Sciences Dept., College
Station, TX, Personal communication, 1980.
15. Gee, G.W., Cambell, A.C., Sherwood, D.R., Strickert, R.G., and Phillips,
S.J., Interaction of Uranium Mill Tailings Leachate with Soils and Clay
Liners" Battelle-Pacific Northwest Laboratory, prepared for the U.S.
Nuclear Regulatory Commission, NUREG/CR-1494, PNL-3381, June, 1980.
16. ASTM, Annual Book of ASTM Standards, Part 19. American Society for Test-
ing and Materials, Philadelphia, 1978.
64
-------
17,
18,
19,
20
21
22
23
24
25
26
27
28
29
30
31
Gee, G.W., Battelle-Pacific Northwest Laboratory, Personal Communication,
1980.
Sherard, J.L., Dunningan, L.P., and Decker, R.S., Identification and
Nature of Dispersive Soils. American Society of Civil Engineers, Geo-
tech. Eng. Div., 101 (GT4).
Crim, R.G., Shepherd, T., and Nelson, J.D., Stability of Natural Clay
Liners in a Low pH Environment. Proc. of the 2nd Symposium on Uranium
Mill Tailings Management, Colorado State University, Ft. Collins, Colora-
do, November, 1979.
Haxo, H.E., Evaluation of Selected Liners When Exposed to Hazardous
Wastes. Proceedings of the Hazardous Waste Research Symposium, Residual
Management by Land Disposal, EPA 600/9-76-015, July, 1976.
Haxo, H.E., Haxo, R.S., and White, R.M., Liner Materials Exposed to Hazar-
dous and Toxic Sludges. First Interim Report, EPA 6/2-/7-081, June,
W7T
Haxo, H.E., Compatibility of Liners with Leachate. Proceedings of 3rd An-
nual Municipal Solid Waste Research Symposium, University of Missouri,
EPA 600/9-77-026, September, 1977.
Haxo, H.E., Interaction of Selected Lining Materials with Various Hazard-
ous Wastes. Proceedings of the Fourth Annual Research Symposium, Land
Disposal of Hazardous Wastes, EPA 600/9-78-016, August, 1978.
Haxo, H.E., Haxo, R.S., and Kellog, T.F., Liner Materials Exposed to
Municipal Solid Waste Leachate. EPA 600/2-79-038. July, 1979.
Haxo, H.E., Interaction of Selected Lining Materials with Various
Hazardous Wastes-11. Proc. of the bth Annual Symposium on the Disposal of
Hazardous Wastes, EPA 600/9-80-010, March, 1980.
Vallerga, B.A., and Hicks, R.G., "Water Permeability of Asphalt Concrete
Specimens Using Back Pressure Saturation". Toxic Materials, 3 (1), 1968.
Fung, R., ed., Protective Barriers for Containment of Toxic Materials.
Noyes Data Corp., Park Ridge, New Jersey, September, 1980.
Ware, S.A., and Jackson, G.S., Liners for Sanitary Landfills and Chemical
and Hazardous Waste Disposal Sites. EPA 600/9-78-005, 19/8.
Exxon Research and Engineering Company, State-of-the-Art Study of Land
Impoundment Techniques. EPA 600-2-78-196, 1978.
ASTM, Dispersive Clays, Related Piping, and Erosion in Geotechnical Pro-
jects. ASTM Special Technical Publication #623.
Barrier, R.M., Zeolites and Clay Minerals as Solvents and Molecular
Sieves. Academic Press, New York, 1978.
65
-------
32. Williams, R.E., "Control and Prevention of Seepage from Uranium Mill Waste
Disposal Facilities". In: Management, Stabilization and Environmental Im-
pact of Uranium Mill Tailings"! Proceedings of the NEA Seminar, Albuquer-
que, New Mexico, July, 1978.
33. American Colloid Company, Bentonite Soil Sealants for Industrial
Waste-Water Management. Skokie, 111., 60076.
34. Van Zyl, D., and Caldwell, J.A., Efficiency of a Natural Clay Liner for
High Acidity Tailings Impoundment. Proceedings of the First Uranium Mill
Tailings Management Symposium, Colorado State University, Ft. Collins,
Colorado, November 1978.
35. Mitchell, J.K., Fundamentals of Soil Behavior (Series in Soil
Engineering). John Wiley and Sons, New York, 1976.
36. Scott, G.R., and Cotton, W. A., Geologic and Biostratigraphic Map of the
Pierre Shale between Jarre Creek and Loveland, Colorado. U.S. Geologic
Survey Misc. Geol. Inv. Map 1-439, 1965.
37. Gardner, M.E., Preliminary Report on the Engineering Geology of the
Boulder Quadrangle, Boulder, Colorac[o"l USGS Open File Report #1139, 1968.
38 Gardner, M.E., Preliminary Report on the Engineering Geology of the
Eldorado Springs Quadrangle, Boulder and Jefferson Counties, Colorado,
USGS Open File Report #1258, 1969.
39. McGregor, E.E., Personal Communication, 1980.
40. McGregor, E.E., and McDonnough, Bedrock and Surficial Engineering Geologic
Maps of the Littleton Quadrangle7~3efferson, Douglas and Arapahoe
Counties, Colorado^ USGS Open File Report 80-321, 1980.
41. Gardner, M.E., and Hart, S.S., Preliminary Engineering Geologic Map of the
Golden Quadrangle, Jefferson Co., Colorado. USGS, M.F. 308, 1972.
42. Scott, G.R., General Engineering Geology of the Northern Part of Pueblo,
Colorado. USGS Bulletin #1262, 1969.
43. Gude, A.J. Ill, "Clay Minerals of Laramie Formation, Golden, Colorado",
AAPG Bulletin, Volume 34, no. 8, pp. 1699-1717, 1950.
44. Van Horn, R., Geology of the Golden Quadrangle. USGS Prof. Paper 872,
1976.
45. Hart, S.S., Engineering Geologic Map of the Morrison Quadrangle, Jefferson
County, Colorado. USGS Misc. Geol. Inv. Map I-790-H, in press.
46. Larsen, L.S., Soil Survey of Castle Rock Area, Colorado. Soil
Conservation Service, 1974.
47. Shroba, R.R., Geologic Map and Physical Properties of the Surficial and
Bedrock Units of the Englewood Quadrangle, Denver, Arapahoe, and Adams
Counties, Colorado" USGS Map GQ-1524, 1980.
66
-------
48. Mayberry, J.O., Lindwall, R.M., Geologic Map and Engineering Data for the
Hiqhland Ranch Quadrangle, Arapahoe and Douglas Counties, Colorado. USGS
Map M.F. 631, 19/4.
49. Pierce, G.W., andSiegel, F.R., "Quantification and Clay Mineral Studies
of Sediments and Sedimentary Rocks". J. Sedimentary Petrology, 39 1969.
50. Biscayne, P.E., "Mineralogy and Sedimentation of Recent Deep Sea Clay in
the Atlantic Ocean and Adjacent Seas and Oceans", Geological Society of
America Bui 1. No. 74. pp. 803-829, 1965.
51. Busenberg, E., and Clemency, C.V., "Determination of the Cation Exchange
Capacity of Clays and Soils Using an Ammonia Electrode". Clays and Clay
Minerals. Pergamon Press, Great Britain, 21, 1973.
52. U.S. Department of Agriculture, Agricultural Handbook #60, 1954.
53. U.S. Department of Interior, Bureau of Reclamation, Earth Manual. 2nd
ed., U.S. GPO, Washington D.C., 1974.
67
------- |