xvEPA
United States
Environmental Protection
Agency
Environmental Monitoring
Systems Laboratory
P.O. Box 15027
Las Vegas NV89114
EPA-600/7-80-134
June 1980
Research and Development
Monitoring In The
Vadose Zone:
A Review of Technical
Elements and Methods
Interagency
Energy-Environment
Research
and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad categories
were established to facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously planned to foster
technology transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY—ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the effort
funded under the 17-agency Federal Energy/Environment Research and Development
Program. These studies relate to EPA'S mission to protect the public health and welfare
from adverse effects of pollutants associated with energy systems. The goal of the Pro-
gram is to assure the rapid development of domestic energy supplies in an environ-
mentally-compatible manner by providing the necessary environmental data and
control technology. Investigations include analyses of the transport of energy-related
pollutants and their health and ecological effects; assessments of, and development of,
control technologies for energy systems; and integrated assessments of a wide range
of energy-related environmental issues.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161
-------
EPA-600/7-80-134
June 1980
MONITORING IN THE VADOSE ZONE:
A Review of Technical Elements and Methods
By
L.G. Wilson
General Electric Company—TEMPO
Center for Advanced Studies
Santa Barbara, California 93102
Contract No. V-0591-NALX
Project Officer
Leslie G. McMillion
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
U.S. Efovirtmaental Pretsotion
Region 5, Li Wary (5PL-16)
230 S. Ueart»rn Stre*t»
Chlca**, IL 69664
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does men-
tion of trade names or commercial products constitute endorsement or recom-
mendation for use.
-------
FOREWORD
Protection of the environment requires effective regulatory actions based
on sound technical and scientific data. The data must include the quantita-
tive description and linking of pollutant sources, transport mechanisms, in-
teractions, and resulting effects on man and his environment. Because of the
complexities involved, assessment of exposure to specific pollutants in the
environment requires a total systems approach that transcends the media of
air, water, and land. The Environmental Monitoring Systems Laboratory at Las
Vegas contributes to the formation and enhancement of a sound monitoring-data
base for exposure assessment through programs designed to:
• Develop and optimize systems and strategies for monitoring pol-
lutants and their impact on the environment
• Demonstrate new monitoring systems and technologies by applying
them to fulfill special monitoring needs of the Agency's operat-
ing programs.
This report covers the topics of (1) principles of pollutant movement in
the vadose zone (zone of aeration or unsaturated zone), (2) basic chemical
reactions of fluids in the zone, (3) state-of-the-art monitoring techniques,
and (4) the relative advantages and disadvantages of the different monitoring
techniques. Recent intense concern over hazardous waste disposal has indi-
cated the need for instruction on how to monitor in the vadose zone and to
identify the potential gains from the limitations of the methods available for
monitoring in this zone. This report provides technical information needed by
regulating agencies and industrial concerns in dealing with waste disposal
problems. In addition, the basis for future research is provided through
identification of present monitoring limitations.
Further information on this study and the subject of groundwater quality
monitoring in general can be obtained by contacting the Environmental Moni-
toring Systems Laboratory, U.S. Environmental Protection Agency, Las Vegas,
Nevada.
Glenn E. Schweitzer
Director
Environmental Monitoring Systems Laboratory
Las Vegas
m
-------
SUMMARY
The land surface is being utilized increasingly for the disposal of solid
and liquid wastes. Concomitant!y, there has been an increasing awareness of
the need to monitor the movement of pollutants from such wastes into ground-
water. Sampling pump discharge offers a direct approach for monitoring
groundwater. Unfortunately, by the time pollutants appear in well discharge
samples, the groundwater system may have become extensively contaminated. An
alternative approach is to monitor within the vadose zone. Water samples
from this region may presage the quality of water entering the underlying
flow system and hence manifest the ultimate pollution potential. Ameliorative
measures could then be initiated before groundwater contamination becomes
extensive.
In contrast to the advanced state of the art for sampling from the zone
at saturation, vadose zone monitoring is still in its infancy. The discrep-
ancy in technology in part reflects the greater complexity of flow in the va-
dose zone, compared to saturated zone flow, and the associated problem of
obtaining a representative water sample for analysis. The lack of a compre-
hensive methodology for vadose zone monitoring is a serious impediment to
federal and state personnel charged with promulgating or enforcing monitoring
regulations for best management practices in agriculture, for hazardous waste
disposal programs, and for 208 programs. The majority of these individuals
are untrained in the hydraulics of saturated and unsaturated groundwater
flow. Therefore, a need exists for a document presenting a comprehensive re-
view of techniques for monitoring water and pollutant movement in the vadose
zone.
The purpose of this report is to serve as a basic introduction to the
three components of a vadose zone monitoring program. These components are:
(1) changes in storage of water in the vadose zone, (2) the flow rate of water
in the vadose zone, and (3) the temporal and spatial distribution of pollu-
tants in the vadose zone. A brief technical review of each of these compo-
nents is presented to set the stage for a review of associated field methods.
The principles of these methods are briefly discussed, together with some of
the advantages and disadvantages.
Techniques that are reviewed for observing storage changes in the vadose
zone at a waste disposal site include the following:
• Monitoring the spatial distribution of water levels in wells to
delineate the areal thickness of the vadose zone
iv
-------
• The gamma ray attenuation method to characterize bulk density
and water content of vadose zone sediments
• The neutron moderation method for defining the water content
distribution in the vadose zone
• Tensiometers for estimating water content at discrete points in
the vadose zone
• Electrical resistance blocks for estimating water content at
discrete points.
Methods for monitoring water movement (flux) and associated parameters
in the vadose zone include:
• Estimating infiltration rates by infiltrometers and test ponds
• Characterizing the quantity of water moving beneath the soil
zone using the water balance approach
• Determining the direction of unsaturated water movement and as-
sociated hydraulic gradients using tensiometers, psychrometers,
and the neutron moderation method
• Measuring the unsaturated flux of water by adapting laboratory
techniques to the field, using water content profiles, estimat-
ing from suction cup response, and using direct techniques such
as flow meters
• Determining saturated flow in perched groundwater zones using
piezometers and observation wells
• Outlining techniques for determining the saturated hydraulic
conductivity in the soil zone and deeper vadose zone.
Indirect methods for monitoring movement in the vadose zone which are
reviewed include the following:
• The four-electrode method for soil salinity
• The EC probe for monitoring soil salinity
• The four-electrode conductivity cell for observing soil salinity
• The earth resistivity approach for delineating pollution plumes.
Techniques for solids sampling in the vadose zone for determination of
associated pollutants are reviewed.
The following direct techniques for water sampling during unsaturated
flow are described:
-------
• Ceramic-type samplers (suction lysimeters and filter candles)
• Cellulose-acetate hollow-fiber filters
• Membrane filter samplers.
Methods for sampling from shallow perched groundwater zones include:
• Sampling tile drain outflow
• Collection pans and manifolds
• Wells
• Piezometers
• Multilevel samplers
• Groundwater profile samplers.
Sampling from deeper perched groundwater includes:
• Collecting cascading water
• Installing special wells.
VI
-------
CONTENTS
Foreword iii
Summary iv
Figures viii
Tables xi
Abbreviations and Symbols xii
Acknowledgments xiv
Section Page
1 Introduction 1
Importance of Source Monitoring and Monitoring in
in the Vadose Zone: When Should Vadose Zone
Monitoring Be Discounted? 2
Objective of Monitoring in the Vadose Zone 3
Purpose of the Report 4
2 General Features of Vadose Zones 6
3 Preliminary Characterization of the Vadose Zone 10
4 Monitoring Storage Properties of the Vadose Zone 14
Technical Considerations 14
Field Methods for Determining Storage Properties
of the Vadose Zone 25
Utilization of Monitoring Methods for Observing
Storage Capacity Beneath a Pollution Source 42
5 Monitoring Water Movement in the Vadose Zone 43
Technical Review 43
Field Methods for Determining the Rate of Water
Movement in the Vadose Zone 58
6 Monitoring Quality Changes in the Vadose Zone 87
Technical Review 87
Field Methods for Monitoring Pollutant Movement
in the Vadose Zone 103
References 146
Appendix A Conversion Factors 163
Glossary 165
vii
-------
FIGURES
Number Page
1 Schematic representation of the flow paths of a pollutant
from a surface source 3
2 Cross section through the vadose zone and groundwater zone 7
3 Variation of porosity, specific yiejld, and specific
retention with grain size j 17
4 Soil-water characteristic curve for a clayey soil and
sandy soil 20
5 Modified Haines apparatus for obtaining soil-water
characteristic curves 21
6 Soil-water characteristic curves for a sandy soil during
wetting and drying showing the effects of hysteresis 22
7 Schematic representation of porosity, specific yield, and
specific retention . 25
8 Water content distribution above a water table during a
fall (draining) and rise (filling) in the water surface 26
9 Sequence of water content profiles showing the growth and
dissipation of perched groundwater during recharge from
an ephemeral stream 30
10 Equipment and principles of neutron moisture logging 31
11 Dual probe used to measure water content by gamma ray
attenuation 35
12 Schematic representation of tensiometer and section
through the ceramic cup 36
13 Cross section of tensiometer-pressure transducer
assembly 38
14 Thermistor soil cell, meter, and hand auger 40
vm
-------
Number Page
15 Water content profiles of a soil wetted slowly by
sprinkling and a soil continuously ponded 45
16 The effect of initial water content of a soil on
infiltration rates 49
17 Effect of the negative pressure of soil water on the
unsaturated hydraulic conductivity 51
18 Pressure head diagram in a transition from a more
permeable to a less permeable layer 54
19 Relationships among hydraulic conductivity, salt
concentration, and ESP for Pachappa sandy loam 57
20 Distribution of hydraulic heads for three unsaturated
flow cases 62
21 Soil psychrorneter 64
22 Hypothetical water content profiles at several times
during drainage of a soil profile 67
23 Schematic representation of procedure for converting
soil-water pressure values from tensiometer data to
water content values on a water content profile for
a given time 68
24 Sketch of field plot for determining K and J showing
tensiometers and access well 69
25 Water content profiles in hypothetical soil column at
times t = 24 hours and t = 40 hours 71
26 Hydraulic head distribution in three piezometers during
three hypothetical flow cases 80
27 Array of devices for monitoring water movement in the
vadose zone and perched groundwater 86
28 Mobility of Cu, Pb, Be, Zn, Cd, Ni, and Hg in 10 soils 96
29 Mobility of Se, V, As, and Cr in 10 soils 96
30 Wenner four-probe array and associated area of
conductivity measurement 105
31 Cross section of EC probe 107
32 Subsamples from a "dry tube" core sampler 115
ix
-------
Number Page
33 Soil-water sampler 117
34 Vacuum-pressure sampler 118
35 "Hi/Pressure-vacuum soil-water sampler" 120
36 Facilities for sampling irrigation return flow via filter
candles, for research project at Tacna, Arizona 121
37 Clustered suction cup lysimeters in a common borehole 123
38 Membrane filter sampler 126
39 Leachate collector installed at base of sanitary landfill 129
40 Multilevel sampler 131
41 Groundwater profile sampler 133
42 Hand bailer made of Teflon 136
43 Schematic representation of simple air-lift pump showing
principle of operation 137
44 Positive action air-lift pump 139
45 Conceptualized cross section of a well showing cascading
water from perched zone 143
46 Conceptualized monitor well used to sample from a
perched zone 144
-------
TABLES
Number Page
1 Range of Porosity Values in Unconsolidated Deposits on Rocks 17
2 Compilation of Specific Yield Values for Various Materials
in California Valleys 26
3 Typical Hydraulic Conductivity Values 55
4 Factors that Influence the Movement of Viruses in Soil 101
XI
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
atm
bar
CEC
cm
EC
EC,
ECe
ECS
!£•
ESP
ET
I
It
meq
mmhos/cm
PET
pHc
PPT
R
RO
SAR
SP
Unit of pressure, the atmosphere, equivalent to 1.013 x 106
dyne/cm2, or 1.013 bars
Unit of pressure, equivalent to 106 dyne/cm2, or 0.987
atmosphere
cation exchange capacity
centimeters
electrical conductivity
electrical conductivity of a saline soil
conductivity of saturation extract
surface conductance
electrical conductivity of water or soil solution
redox potential
exchangeable sodium percentage
evapotr anspi rati on
irrigation
cumulative infiltration
milliequivalents
mi Hi mhos per centimeter
potential evapotranspiration component of water balance equation
pH adjusted for calcium, magnesium, carbonate bicarbonate
equilibria
precipitation plus irrigation components of water balance equation
r ai nf al 1
runoff component of water balance equation
sodium adsorption ratio
saturation percentage
SYMBOLS
A
a
Dw
dw
Cross-sectional area of aquifer normal to flow; geometric factor
in piezometer equation; factor in Philip's infiltration equation
Empirical factor relating k and 6 in Nielsen-Biggar flux equa-
tion; distance between electrodes in four-probe salinity method
Bulk density
Wet weight bulk density
Density of water
Depth of water applied during surface flooding
Xll
-------
dvz Depth of vadose zone wetted by application of a depth of water c
ft Adjustment factor in four-electrode equation
g Acceleration of gravity
H Total hydraulic head
h Soil-water pressure (or suction)
hcr Critical pressure head of soil for wetting, used in Green-Ampt
equation
Hw Depth at water ponded above soil surface in Green-Ampt equation
J Specific discharge; flux; Darcian flux
JL Flux at depth L in a soil
K Hydraulic conductivity
K(0) Hydraulic conductivity as a function of water content 6
Kd Distribution coefficient
K0 Steady-state hydraulic conductivity
Lf Depth of wetting front in Green-Ampt equation
n Porosity; also volume of mole of water
Pv Volumetric water content of a soil
Pw Moisture percentage of a soil on a dry weight basis
P/P0 Relative humidity
Q Total discharge in Darcy's equation
R^ Measured resistance in ohms
r Effective radius of soil pores
S Sorptivity
S-t Total porosity of a soil or vadose zone region
T Surface tension of a liquid; absolute temperature; tortuosity
T[_ Time lag constant in piezometer equation
t Time
Vj Infiltration rate in Green-Ampt equation
Vj Pore velocity of adsorbed species
Vs Bulk volume of soil
Vw Volume of water in a soil sample
vw Pore velocity of migrating water
v Average linear velocity; J/6
W(j Dry weight of soil
Ww Wet weight of soil
x Height of mercury column in tensiometer manometer
Y0 Distance of sudden rise of water level in piezometer equation
y Distance from soil surface to mercury reference level in
tensiometer
z Depth below land surface to the ceramic cup of a tensiometer
a Contact angle between liquid and solid surfaces
APw Difference in percentage of water between two water contents
Asm Change in soil moisture storage
6 Volumetric water content
¥ Soil-water potential
Pb Particle density of solids
xiii
-------
ACKNOWLEDGMENTS
Dr. Guenton C. Slawson, Jr. and Dr. Lome G. Everett of General
Electric-TEMPO were responsible for the management of the project under which
this report was prepared. The author of the report is Dr. Lome G. Wilson of
the Water Resources Research Center, University of Arizona. Dr. Wilson's ef-
forts were provided under subcontract No. 30167 to the University.
The technical review and support of Mr. Jewell Meyer, Dr. Kenneth D.
Schmidt, and Dr. Richard M. Tinlin are acknowledged. The author is particu-
larly appreciative of the review of Section 5 by Dr. Arthur Warrick, the Uni-
versity of Arizona. His comments and suggestions were of signal value to the
author in his attempt to clarify a rather difficult subject area.
The secretarial/editorial/graphical assistance of the following individ-
uals is gratefully acknowledged: Nancy Svacha, Nancy Becker, Roberta Bowen,
Joelle May, Lorraine Donald, Diane Landis, and Marie Busse.
xiv
-------
SECTION 1
INTRODUCTION
Recent environmental legislation relative to the disposal of wastes has
justifiably stressed the importance of protecting potable groundwater from
waste-associated pollutants. Thus, the Safe Drinking Water Act of 1974 (PL
93-523) calls for the initiation of state programs to protect underground
sources of drinking water. Similarly, the Federal Water Pollution Control Act
of 1972 (PL 92-500) calls for the development of plans in designated areawide
waste-management planning districts in order "... to control the disposal of
pollutants on land or in subsurface water quality." The Resource Conservation
and Recovery Act of 1976 (PL 94-580) recognized that "... open dumping is par-
ticularly harmful to health (and) contaminates drinking water from underground
surface supplies
The potential danger to potable water resulting from improper disposal
of hazardous wastes was highlighted by the recent example of water pollution
near the Love Canal in New York State. However, this case is but one of many
in the United States, and the EPA has compiled over 400 case studies demon-
strating the disastrous results of poor disposal of hazardous wastes (U.S.
EPA, 1978).
In recognition of the need to safeguard groundwater resources, guidelines
emanating from the principal environmental acts call for source and groundwa-
ter monitoring. In particular, the proposed guidelines relative to hazardous
waste disposal (U.S. EPA, 1978) explicitly call for leachate and groundwater
monitoring. Correspondingly, proposed guidelines for the landfill disposal
of solid wastes (U.S. EPA, 1979b) specify that, "A groundwater monitoring
system should be installed for the purpose of detecting the impact of all
landfill disposal facilities which have the potential for discharge to an un-
derground drinking water source." In addition to landfills, other waste dis-
posal operations which may require monitoring include the following: pits,
ponds, lagoons, septic tank areas, dry channels used for effluent disposal,
dry wells, and injection wells. Monitoring irrigation return flow to ground-
water is advisable in irrigated areas, particularly where heavy use of fer-
tilizers and pesticides may cause pollution. Monitoring is an important
consideration during disposal of spoils from coal strip-mining and oil shale
developments (Everett, 1979; Slawson, 1979).
In a multiple-source area, a rating process may be necessary to rank
sources in order of monitoring priority. Rating schemes have been proposed
by LeGrand (1964); Pavoni, Hagerty, and Lee (1972); and Silka and Swearingen
(1978). A systematic approach for rating potential sources and developing
1
-------
source-specific monitoring programs was proposed by Todd et al. (1976). This
approach includes an evaluation of potential pollutants at each source loca-
tion, considers the mobility of pollutants in the site-specific hydrologic
regimen, and examines the potential impact of pollutants on current groundwa-
ter uses. Everett et al. (1976) presented a review of specific techniques and
costs for establishing comprehensive monitoring networks. Similarly, Fenn et
al. (1977) prepared a procedures manual to aid in designing monitoring systems
at solid waste disposal facilities.
IMPORTANCE OF SOURCE MONITORING AND MONITORING IN THE VADOSE
ZONE: WHEN SHOULD VADOSE ZONE MONITORING BE DISCOUNTED?
Figure 1 shows the hypothetical flow paths of a pollutant beneath a waste
disposal site. It is well known that when groundwater underlying such a dis-
posal site begins to manifest pollution by quality changes in pumped water,
the aquifer may have already become badly contaminated. Monitoring the qual-
ity of wastes entering and within a disposal site (i.e., source monitoring),
coupled with monitoring in the vadose zone, offers an "early warning" or
groundwater pollution. Remedial measures may be implemented prior to the on-
set of severe pollution and the associated renovation costs will be eliminated
or reduced. As an example, Wilson and Schmidt (1979) presented a case study
entailing the monitoring of cascading water in an abandoned irrigation well
near a field irrigated with sewage effluent. Groundwater from a nearby pump-
ing well was also monitored. Analysis of cascading water samples showed that
water of very poor quality was in transit through the vadose zone. Further-
more, the quality of cascading water was markedly poorer than that of pumped
groundwater. The important feature of this case study is that if only pumped
groundwater had been used to monitor return flow, it would have been errone-
ously concluded that no pollution was occurring.
In addition to providing an early warning of pollutant movement in
groundwater, vadose zone monitoring may reduce the need for and costs of
groundwater monitoring. In other words, if a vadose zone monitoring program
fails to detect the movement of pollutants, the requirements for groundwater
monitoring may be reduced or largely precluded. The savings in costs for
constructing monitoring wells could be significant.
It is not our intention to proclaim vadose zone monitoring as a panacea
for all hydrogeological conditions and for all waste disposal operations. The
need for and extent of such monitoring should be tailored to site-specific
conditions. For example, if the water table at a given site is relatively
shallow, say within 10 feet of the land surface, vadose zone monitoring may be
minimal. Similarly, if the vadose zone consists of fractured media, flow oc-
curs primarily in channels and the ameliorating interactions of the vadose
zone and water-borne pollutants may be minimal. Monitoring activities for
these two cases could be restricted to source monitoring and groundwater mon-
itoring. Installation of in-source monitoring facilities, including vadose
zone units, may not be recommended for certain sources, such as landfills.
The rationale was presented in the proposed guidelines for landfill disposal
of solid waste (U.S. EPA, 1979b): "In no case should groundwater or leachate
monitoring wells be installed through the bottom of the landfill proper since
such installation could result in creation of a conduit for the direct passage
-------
LAND SURFACE-
TOP SOIL
VADOSE
ZONE
r
POLLUTANT
TRAVEL
WATER LEVEL-
AQUIFER
WELL
DISCHARGE
U-WELL CASING
•PERFORATIONS
BASE OF AQUIFER-
Figure 1. Schematic representation of the flow paths of a pollutant
from a surface source (after Schmidt, 1977).
of landfill leachate into underlying groundwater." Thus, it would be neces-
sary to either install monitoring facilities around the periphery of the land-
fills or use angle-drilling methods to install such facilities below the base
of the lowermost cells.
OBJECTIVE OF MONITORING IN THE VADOSE ZONE
The basic objective of a monitoring program in the vadose zone is to
characterize three properties: (1) storage properties, (2) flow rates, and
(3) spatial and temporal changes in pollutants. Storage properties are im-
portant in that pollutants are placed in temporary storage within the void
space of the vadose zone prior to movement into groundwater. In some cases,
such as in western valleys, the storage space may be sufficient to preclude
movement into productive aquifers. In fact, Winograd (1974) suggested using
the vadose zone for the storage of low-grade radioactive wastes. Mann (1976)
recommended using the storage space of the vadose zone in arid regions to
augment the above-ground storage of tailings ponds. Wilson, Osborne, and
-------
Percious (1968) showed that by locating a disposal pond for blowdown effluent
near an ephemeral stream, periodic recharge events diluted the effluent stored
in the vadose zone.
Information on the flux of water (flux refers to the quantity of water
flowing through a unit area per unit time) may indicate the macroscopic veloc-
ity of water in the vadose zone. Thus, if a piston flow model is assumed, the
arrival time of pollutants at the water table may be determined from estimates
of flux and the thickness of the vadose zone. (As will be shown later, a more
realistic estimate of particle velocity is the average linear velocity, which
accounts for variations in the cross-sectional areas of the flow path and de-
gree of saturation.)
Pollutants may not necessarily travel at the same rate as water because
of attenuation in the vadose zone. Consequently, the travel time of water
represents an upper limit on the mobility of pollutants. (A few pollutants
may move at a higher rate than water because of physical and chemical proper-
ties. For example, based on differences in viscosity, the velocity of gas-
oline should be about 1.5 times that of water (Schwille, 1967).) In
considering rate, it is important to quantify both the infiltration rate at
the ground surface and the percolation rate through the vadose zone.
"Recharge" refers to the movement of water and pollutants across the water
table.
Information on the third element, changes in pollutants during transit
within the vadose zone, is needed to quantify the mobility and spatial dis-
tribution of specific pollutants. The vadose zone may attenuate certain pol-
lutants to the point that they could be excluded as pollution sources. For
example, Runnel!s (1976) demonstrated that soil from Sulfur Spring, New Mex-
ico, has an enormous capacity to remove copper from mill water. The addi-
tional observation that copper removal was irreversible indicated that
thousands of years would be required before groundwater (located at about 100
feet*) at the site would be affected.
PURPOSE OF THE REPORT
Present day monitoring in the vadose zone entails using a mix of methods
developed by soil scientists and hydrogeologists. As a consequence, vadose
zone monitoring requires an interdisciplinary approach. Unfortunately, many
of those charged with designing, implementing, and regulating monitoring op-
erations lack the necessary multi disciplinary skills and, in fact, may not
even be aware that an interdisciplinary approach is required.
The purpose of this report is to provide an introduction to the princi-
ples and techniques of vadose zone monitoring. In the next section, a review
*See Appendix A for conversion to metric units. English units are used in
this report because of their currently common usage in industry and
hydrology-related sciences. In cases where metric units are more appropriate
(e.g., laboratory analyses), the English units are included parenthetically.
-------
of general features of the vadose zone is presented to set the stage for sub-
sequent chapters. The third section reviews preliminary approaches to a va-
dose zone monitoring program. Subsequent sections review the technical
elements and state of the art in present monitoring techniques in each of the
three categories: storage properties in the vadose zone, flow properties in
the vadose zone, and pollutant mobility in the vadose zone. The technical
review in each of these sections is intended to provide the reader with a ba-
sic introduction to the theories governing the selection of specific methods
and to review associated laboratory procedures. The reader who is not inter-
ested in such background information should pass directly to the review of
field methods.
To facilitate comparison of the alternative techniques available for
specific functions, advantages and disadvantages have been presented whenever
possible. Regarding costs, the present rate of inflation precludes quoting
precise values for the techniques described in this report. Instead, the
reader is referred to a report by Everett et al. (1976) which reviews proce-
dures for estimating the costs for many of the techniques under a range of
hydrogeological conditions.
-------
SECTION 2
GENERAL FEATURES OF VADOSE ZONES
The geological profile extending from ground surface to the upper surface
of the principal water-bearing formation is called the vadose zone. As
pointed out by Bouwer (1978), the term "vadose zone" is preferable to the
often-used term "unsaturated zone" because saturated regions are frequently
present in some vadose zones. The term "zone of aeration" is also often used
as a synonym for vadose zone.
Davis and deWiest (1966) subdivided the vadose zone into three regions,
designated as the soil zone, the intermediate vadose zone, and-the capillary
fringe. The surface soil zone is generally recognized as that region which
manifests the effects of weathering of native geological material, together
with the processes of elluviation and illuviation of colloidal materials, to
develop more or less well-developed profiles (Simonson, 1957). The movement
of water in the soil zone occurs mainly in the unsaturated state, i.e., that
state in which the soil water exists under pressures less than atmospheric.
The principal transport mechanisms associated with unsaturated flow are in-
filtration, percolation, redistribution, and evaporation (Klute, 1969). In
some soils, primarily those containing horizons of low permeability, saturated
regions may develop during surface flooding, creating shallow perched water
tables.
The physics of unsaturated soil-water movement has been intensively
studied by soil physicists, agricultural engineers, and microclimatologists.
In fact, a copious amount of literature is available on the subject in peri-
odicals (Journal of the Soil Science Society of America, Soil Science) and
books (Childs, 1969; Kirkham and Powers, 1972; Hillel, 1971). Similarly, a
number of published references are available on the theory of flow in perched
water tables (Luthin, 1957; van Schilfgaarde, 1970). Soil chemists and soil
microbiologists have also attempted to quantify chemical-microbiological
transformations during soil-water movement (Bohn, McNeal, and O'Connor, 1979;
Rhoades and Bernstein, 1971; Dunlap and McNabb, 1973).
Weathered materials of the soil zone may gradually merge with underlying
deposits, which are generally unweathered, comprising the intermediate vadose
zone. In some regions, this zone may be practically nonexistent, the soil
zone merging directly with bed rock. In alluvial deposits of western valleys,
however, this zone may be hundreds of feet thick. Figure 2 shows a geologic
cross section through a vadose zone in an alluvial basin in California. By
the nature of the processes by which such alluvium is laid down, this zone is
unlikely to be uniform throughout, but may contain micro- or macrolenses of
-------
9001-
850
800
750
2 700
I
UJ
_l 650
UJ
600
550
500
.=£=•=«. VADOSE ZONE
HYDROGRAPH
•=JT,7.-*;5aa GROUND-WATER ZONE
I I I I I I I 1 I I I I I I I I I I I I 1 I I 1 I I I I I 1
40 45 50 55 60 65
YEAR
OF AQUIFER ELEVATION 200 FT
Figure 2. Cross section through the vadose zone and groundwater
zone (after Ayers and Branson, 1973).
silts and clays interbedding with gravels and clays. Water in the intermedi-
ate vadose zone may exist primarily in the unsaturated state, and in those
regions receiving little inflow from above, flow velocities may be negligible.
However, perched groundwater may develop in the interfacial deposits of re-
gions containing varying textures. Such perching layers may be hydraulically
connected to ephemeral or perennial stream channels so that, respectively,
temporary or permanent perched water tables may develop. Alternatively, sat-
urated conditions may develop as a result of deep percolation of water from
the soil zone during prolonged surface application of water. Studies by
McWhorter and Brookman (1972) and Wilson (1971) have shown that perching lay-
ers that intercept downward-moving water may transmit the water laterally at
substantial rates. Thus, these layers serve as underground spreading regions
transmitting water laterally away from the overlying source area. Eventually,
water leaks downward from these layers and may intercept a substantial area
of the water table. Because of dilution and mixing below the water table, the
effects of recharge may not be noticeable until a large volume of the aquifer
has been affected.
In contrast to the large number of theoretical studies on water movement
in the soil zone, parallel studies in the intermediate zone have been minimal.
In fact, Meinzer (1942) coined the term "no-man's land of hydrology" to de-
scribe the limited knowledge of this zone. Reasoning from Darcy's equation,
-------
Hall (1955) developed a number of equations to characterize mound (perched
groundwater) development in the intermediate zone. Hall also discussed the
hydraulic energy relationships during lateral flow in perched groundwater.
More recently, Freeze (1969) attempted to describe the continuum of flow be-
tween the soil surface and underlying saturated water bodies.
The base of the vadose zone, the capillary fringe, merges with underlying
saturated deposits of the principal water-bearing formation. This zone is not
characterized as much by the nature of geological materials as by the presence
of water under conditions of saturation or near-saturation. Studies by Luthin
and Day (1955) and Kraijenhoff van deLeur (1962) have shown that both the hy-
draulic conductivity and flux may remain high for some vertical distance in
the capillary fringe, depending on the nature of the materials. In general,
the thickness of the capillary fringe is greater in fine materials than in
coarse deposits. Apparently, few studies have been conducted on flow and
chemical transformations in this zone. Taylor and Luthin (1969) reported on
a computer model to characterize transient flow in this zone and compared re-
sults with data from a sand tank model. Freeze and Cherry (1979) indicated
that oil reaching the water table following leakage from a surface source
flows in a lateral direction within the capillary fringe in close proximity
to the water table. Because oil and water are immiscible, oil does not pene-
trate below the water table, although some dissolution may occur.
The overall thickness of the vadose zone will not necessarily be con-
stant. For example, as a result of recharge at a water table during a waste
disposal operation, a mound may develop throughout the capillary fringe ex-
tending into the intermediate zone. Such mounds have been observed during
recharge studies (e.g., Wilson, 1971) and efforts have been made to quantify
their growth and dissipation (Hantush, 1967; Bouwer, 1978). Pumping in the
groundwater zone also modifies the overall thickness of the zone of aeration,
as apparent on Figure 2.
As already indicated, the state of knowledge of water movement and
chemical-microbiological transformations is greater in the soil zone than
elsewhere in the vadose zone. Fortunately, renovation of an applied wastewa-
ter occurs primarily in the soil zone. This observation is borne out by the
results of the well-known studies of McMichael and McKee (1966), Parizek et
al. (1967), and Sopper and Kardos (1973). These studies indicate that the
soil is essentially a "living filter" effectively reducing certain microbio-
logical, physical, and chemical constituents to safe levels after passage
through a relatively short distance (e.g., Miller, 1973; Thomas, 1973). As a
result of such favorable observations, a certain complacency may have devel-
oped with respect to the need to monitor only in the soil zone. However, in
one reported instance at a disposal operation for cannery wastes, an apparent
breakdown of the living filter occurred because of excessive soil loading with
the sugars, sucrose, and fructose (Meyer, personal communication, 1979).
There is growing concern, moreover, that certain toxic substances, such as
chlorinated hydrocarbons, other refractory organics, and microorganisms, may
escape the soil zone, eventually passing into underlying groundwater (Walker,
1973; Ongerth et al., 1973; Shuval and Gruener, 1973; Robertson, Toussaint,
and Jerque, 1974). For example, when soils are thin and underlain in the in-
termediate zone by fractured rocks, short circuiting of these materials may
8
-------
occur directly into groundwater. Thomas and Phillips (1979) have shown that
for well-structured soils, rapid water movement occurs in macropores between
soil peds. Consequently, entrained pollutants may not have the opportunity
to interact with the bulk of the soil matrix.
Dunlap and McNabb (1973) also pointed out that microbial activity may be
significant in the regions underlying the soil. They recommended that inves-
tigations be conducted to quantify the extent that such activity modifies the
nature of pollutants travelling through the intermediate zone.
The foregoing discussion again emphasizes the need to monitor water move-
ment and water quality throughout the vadose zone. For the soil zone, a large
number of techniques were compiled by Black (1965). Monitoring in the inter-
mediate zone and capillary fringe will require the extension of technology
developed in both the soil zone and in the groundwater zone. Examples are
already available where this approach has been used. For example, Apgar and
Langmuir (1971) successfully used suction cups developed for in-situ sampling
of the soil solution to sample at depths up to 50 feet below a sanitary land-
fill. Meyer, in a personal communication (1979), reported that suction cups
were used to sample from depths greater than 100 feet below land surface at
cannery and rock phosphate disposal sites in California.
In addition to the use of available techniques, new methods will undoubt-
edly be developed as additional insight is gained into the complexity of flow
in the vadose zone.
-------
SECTION 3
PRELIMINARY CHARACTERIZATION OF THE VADOSE ZONE
Prior to initiating a vadose zone monitoring program, an assessment
should be made of existing monitoring efforts and relevant background infor-
mation should be collected. Such a "nonsampl ing" approach will aid in the
selection of methods to offset monitoring gaps and will generally increase
the chances of designing an effective system. Assessing the existing monitor-
ing includes determining sample collection methods for the source, defining
the network of existing monitoring wells, including the locations and types
of wells, and determining methods for sampling groundwater. Background infor-
mation should be obtained on the following items: (1) nature of pollutants,
(2) pollutant source loading, (3) existing groundwater usage and quality, and
(4) a description of the hydrogeologic framework, including infiltration po-
tential and hydraulic/geochemical properties of the vadose zone and groundwa-
ter zone.
The existing source monitoring program could be evaluated by comparing
the ongoing sampling techniques with recommended approaches described in re-
cent published guidelines, such as the report of Huibregtse and Moser (1976).
Similarly, the groundwater monitoring program could be compared with ap-
proaches recommended by Mooij and Rovers (1976), and by Everett et al. (1976).
Approaches for collecting background quality information on a source and ex-
isting groundwater usage and quality have been described in detail by Todd et
al. (1976) and by Fenn et al. (1977). The quality data should be examined for
completeness; for example, there may be gaps in existing analyses. Specific
pollutants should be delineated and ranked.
As a first step in characterizing the hydrogeological framework with par-
ticular emphasis on the vadose zone, all relevant data should be collected.
For example, comprehensive soil survey data may be available through state
agricultural experiment stations, the Soil Conservation Service, or other
agencies. Such surveys generally include information on depths of soil hori-
zons, presence of hardpan, clay pan, color, porosity, structure, texture, or-
ganic matter content, soil pH, and infiltration rates.
For deeper horizons in the lower vadose zone, data may be available from
drillers' well logs, Water Supply Papers of the U.S. Geological Survey, Uni-
versity theses, dissertations, and special reports. In general, well logs by
drilling companies mainly describe conditions below the water table. However,
they should be consulted for possible clues on stratification. In some areas,
the U.S. Geological Survey collects drill cuttings from drilling sites. These
samples are evaluated in the laboratory for particle-size distribution, color,
10
-------
and source of origin. Some universities also collect samples as part of their
hydrology or agricultural engineering programs. Again, the purpose of these
sampling programs is to characterize the saturated zone, but valuable clues
on the vadose zone may also be included (Matlock, Morin, and Posedly, 1976).
Historic trends in groundwater levels should be determined by examining exist-
ing water level contour maps.
In addition to collecting available data, wells in the area of a monitor-
ing site should be used to the fullest extent. For example, static water lev-
els in a network of wells can be plotted to give the water table configuration
and thus an approximate idea of the thickness of the vadose zone. Also, the
presence of cascading water in wells will indicate the possible presence of
perching layers. Finally, abandoned wells, or wells in which the pumps have
been removed for servicing, may be used with borehole geophysical methods
(described later) to provide further clues on the vadose zone.
After exhausting all existing sources for clues on the nature of the va-
dose zone at a site, a test drilling program may be considered necessary to
provide more detailed information. The extent and thoroughness of the program
will depend on the availability of funds. At any rate, care should be exer-
cised in setting up the program to ensure the maximum value of the data. For
example, a careful grid should be laid out for the project area. Also, the
test wells should be established based on the possibility that they may even-
tually be used as observation wells, piezometers, or access wells during an
actual monitoring program.
In general, samples from the soil zone are obtained to: (1) characterize
the average soil texture, water content, or chemistry in increments (say 6-
inch increments); (2) observe the precise depth-distribution of soil texture;
or (3) determine the bulk density or water-release curves of soil increments.
For the first purpose, samples obtained using post-hole augers, screw or
sleeve-type augers, or power-driven augers are useful. For the second pur-
pose, cores are obtained by driving small-diameter tubes into the soil to the
desired depth. For the third purpose, larger-diameter core samplers are used.
These may be hand-driven or forced into the soil by power-driven hydraulic
soil coring equipment. With these methods, cores of a specific volume are
obtained. More information on coring is presented by Blake (1965).
To sample throughout the vadose zone, it may be necessary to construct
deep wells, using standard drilling techniques. Such techniques include jet-
ting, rotary, cable-tool, augering, and air drilling. Of these methods, per-
haps augering, using continuous flights, and air drilling provide the most
usable samples. Problems develop in characterizing the distribution of in-
digenous salt and water content with cable-tool and rotary methods because of
water additions during the drilling process.
One air drilling technique involves driving a double-wall drive tube by a
pile hammer and, concurrently, forcing air under pressure down the annul us of
the pipe. Air and entrained material cut by the bit return to the surface
through the inside pipe. The sample, available continuously, is diverted
into a cyclone sampler where it is bagged for laboratory analyses. According
to the manufacturer, formation changes can be determined within a few
11
-------
centimeters. Furthermore, water seams can be determined immediately. This
feature is advantageous in locating the depth and thickness of perching lay-
ers. Whichever technique is used, samples should be taken in increments
throughout the vadose zone.
Drill cuttings could be examined in the laboratory for particle-size dis-
tribution and other geologic parameters. Knowing the vertical distribution of
grain size in the hole, the location of potential perching zones may be delin-
eated. The collected samples could also be used to estimate storage and
transmissive properties of the vadose zone.
Following construction of a test well or wells, geophysical logging
should be initiated. First, the test hole(s) should be logged. The results
of logging in the test well should be correlated with drill cutting analyses
to delineate stratigraphy. Subsequent logs from other wells could thus be in-
terpreted and the lateral and vertical extent of various layers (e.g., those
favoring perching) defined.
Among the common borehole logging techniques are the spontaneous poten-
tial method, resistance logging, acoustic logging, and nuclear logging. The
spontaneous potential method and resistance logging require an uncased water-
filled well. Acoustic logging is conducted in a fluid-filled cavity and is
not applicable to the vadose zone. Only nuclear logging methods are suitable
for logging in either uncased or cased wells above the water table.
Common nuclear logging techniques include natural gamma logging, gamma-
gamma logging, and neutron logging. The principles of and procedures in ap-
plying these techniques in groundwater investigations are discussed thoroughly
by Keys and MacCary (1971). Briefly, natural gamma logging comprises the de-
tection of natural gamma activity of rocks. Keys and MacCary (1971) list the
following sequence of sedimentary rocks in ascending order of natural gamma
activity: anhydrite, coal, salt, dolomite, sandstone, sandy shale, shale,
organic marine shale, and potash beds. Natural gamma logs are particularly
advantageous in characterizing the vertical extent of sediments and tracing
them laterally from well to well (Keys and MacCary, 1971; Norris, 1972).
Gamma-gamma logs provide a record of the intensity of radiation from a
source in a down-hole probe after it has backscattered and attenuated in the
well and surrounding media (Keys and MacCary, 1971). The down-hole probe con-
tains a source of gamma photons, such as cobalt-60 or cesium-137, and sodium
iodide detector. The principal uses of gamma-gamma logs are to identify the
lithology and to estimate bulk density and porosity of rocks (Keys and
MacCary, 1971).
Neutron logging comprises lowering into a well a down-hole probe contain-
ing a source of high-energy neutrons and a detector of thermalized neutrons.
Neutron loggers provide information on the hydrogen content, and consequently
water content, of sediments within the vadose zone and the porosity of sedi-
ments below the water table. Neutron logs are useful in detecting the pres-
ence of perched groundwater. For example, Keys (1967) used a combination of
gamma logs and neutron logs to delineate a clay perching bed at the National
12
-------
Reactor Testing Station in Idaho. Neutron logging between wells may indicate
the lateral extent of perched groundwater bodies.
13
-------
SECTION 4
MONITORING STORAGE PROPERTIES OF THE VADOSE ZONE
TECHNICAL CONSIDERATIONS
Physical properties of the vadose zone associated with storage of water
include the following: (1) total thickness, (2) porosity, (3) bulk density,
(4) water content, (5) soil-water characteristics, (6) field capacity (spe-
cific retention), (7) specific yield, and (8) fillable porosity. Greater
technical information on each of these items may be found in reference works
by Childs (1969); Hillel (1971); Davis and deWiest (1966); Cooley, Harsh, and
Lewis (1972); Freeze and Cherry (1979); Bouwer (1978); and Brakensiek, Osborn,
and Rawls (1979).
The relationships of each of these properties to storage are reviewed in
this section. Laboratory methods for quantifying the magnitudes of the first
six properties are also presented.
Thickness
The storage capacity of a vadose zone is obviously related to the overall
thickness, i.e., depth from ground surface to the water table. As shown on
Figure 2 , the thickness is not constant but changes in response to pumping or
recharge. The theoretical depth of water that could be stored in a vadose
zone assuming 100-percent saturation is expressed by the following
relationship:
AP,, D.
. _ w b .
where dw = depth of water applied during surface flooding
D^ = bulk density of soil
Dw = density of water
dvz = depth of vadose zone to be wetted
APW = difference in the percentage of water on a mass basis at
saturation and after oven drying (105°C for 24 hours).
14
-------
As an example of the use of this equation to estimate the theoretical
maximum storage capacity of a vadose' zone, consider the following hypothetical
conditions:
1. 100-foot thick vadose zone, uniform and unlayered, i.e., dvz
= 100 feet
2. Bulk density of media = 81.1 Ib/ft3 = 1.3 gm/cm3
3. Density of water =62.4 Ib/ft3 =1.0 gm/cm3
4. Water content of drained vadose zone material = 10 percent (dry
weight basis)
5. Water content of vadose zone material assuming 100-percent sat-
uration = 30 percent (dry weight basis).
Using equation 1, the depth of water, dw, that could be stored in the
100-foot thick hypothetical profile would be:
= 26 feet.
In other words, 26 acre- feet of water per acre of surface area would
theoretically be required to saturate the 100-foot vadose zone. In reality,
the vadose zone would never attain 100-percent saturation, even for an ideal
media, because of entrapped air and because of drainage of the media to field
capacity (see below).
In a multilayered media, the individual bulk densities of the various
layers would need to be considered in using equation 1 to calculate total
storage. Note also that only vertical storage is considered by this relation-
ship. For point and line sources, storage should also be accounted for in a
lateral direction.
Bulk Density
As shown by equation 1, bulk density values are required in order to es-
timate total water storage in the vadose zone. Bulk density represents the
density of a soil or rock, including solids plus voids, after drying (Bouwer,
1978). To determine the bulk density of a material, field cores of a precise
volume are oven-dried at 105°C until constant weight. The bulk density is
the oven-dry mass divided by the sample volume. Blake (1965) and Paetzold
(1979) also reviewed the alternative "clod method" for determining bulk den-
sity of disturbed samples.
Bulk density values range from 1.1 gm/cm3 (68.6
)ils to 1.6 gm/cm3 (100 Ib/ft3) in sandy soils (Hill<
- , „ , - „ . - Ib/ft3) for clay
soils to 1.6 gm/cm3 (100 Ib/ft3) in sandy soils (Millei, 1971).
15
-------
Porosity
As defined by Bouwer (1978), the porosity of a soil or rock material is
the percentage of the total volume of the medium occupied by pores or other
openings. Thus, the total porosity of a soil is a measure of the amount of
water which could be stored under saturation. In actuality, because of en-
trapped air, saturation may not be reached until after a long period of time.
According to Vomicil (1965), total porosity may be calculated by the fol-
lowing expression:
St = 100 (1 - (Db/pb)) (2)
where S^ = total porosity, the percentage of the bulk volume not
occupied by solids
Dt, = bulk density
pb = particle density, approximately 2.65 gm/cm^ (165.4 lb/ft3)
for mineral soil.
Typical values of porosity for representative earth materials are pre-
sented in Table 1 and on Figure 3.
Laboratory estimation of the total porosity of a soil core requires de-
termining the bulk density and particle density. Methods for determining the
bulk density were reviewed above. Particle density is measured precisely
using the pycnometer method (Vomicil, 1965).
Although a knowledge of the total porosity is of value when considering
storage in the vadose zone, also of importance is the pore-size distribution.
For example, sandy soils have primarily large pores of nearly equal size per-
mitting easy drainage (U.S. EPA et al., 1977). However, medium-textured loamy
soils have a greater porosity than sands and also a wider pore-size distribu-
tion. "Thus, more water is held at saturation in soils than in sands and it
is removed much more gradually as matric potential becomes greater" (U.S. EPA
et al., 1977).
The effect of macropores on retention and drainage of water in soils is
receiving renewed attention (Thomas and Phillips, 1979). The effect of macro-
pores is particularly important in well-structured soils. Water movement oc-
curs rapidly in the interstices between soil peds and little storage may occur
within the pores of the soil blocks.
Pore-size distribution is commonly found by means of water characteristic
curves (discussed below).
Water Content
The total porosity of vadose zone materials represents the upper limit
for water storage in the vadose zone. Unless the vadose zone has not been
wetted since deposition, a certain residual amount of water remains in the
16
-------
TABLE 1. RANGE OF POROSITY VALUES IN UNCONSOLIDATED DEPOSITS
ON ROCKS (after Freeze and Cherry, 1979)
Porosity
(percent)
Unconsolidated Deposits
Gravel
Sand
Silt
Clay
Rocks
Fractured basalt
Karst limestone
Sandstone
Limestone, dolomite
Shale
Fractured crystalline rock
Dense crystalline rock
25 to 40
25 to 50
35 to 50
40 to 70
5 to 50
5 to 50
5 to 30
0 to 20
0 to 10
0 to 10
0 to 5
tu
UJ
QL
45
40-
35-
30-
25-
15-
ID-
S'
0-
POROSITY
SPECIFIC RETENTION
(O
-------
pores. Thus, the available storage is diminished by this residual water con-
tent. Water content is expressed on either a mass or a volume basis using the
following expressions presented by Brakensiek, Osborn, and Rawls (1979):
w. - W .
P • --*
w
and
Pv = x 100 (4)
where Pw = moisture percent on a dry weight basis
Ww = wet weight of soil
W
-------
weight basis, and the energy is termed soil-water pressure head, or head (U.S.
EPA et al., 1977). The relationship between soil-water pressure and water
content is called the soil-water characteristic. According to Klute (1969),
the change in water content per unit pressure head change (d6/dh) is called
the water capacity.
As pointed out by the U.S. EPA et al. (1977), "The force by which water
is held in the soil is approximately inversely proportional to the pore diam-
eter. Thus, the larger the pore, the less energy is required to remove water.
As soil dries or drains, water is removed from the larger pores first. The
water remains in the smaller pores because it is held more tightly. Thus, as
soil-water content decreases, soil-water potential increases." Hillel (1971)
pointed out that the amount of water retained at low negative head values de-
pends on capillary effects and pore-size distribution and, consequently, is
strongly affected by soil structure. In contrast, at higher netative heads,
water retention is influenced less by structure and more by texture and spe-
cific surface.
The effect of soil texture on the shape of the soil-water characteristic
curve is shown on Figure 4 where curves are presented for a clay soil and for
a sandy soil. The curve for the clay soil is more gradual and the water con-
tent at a given negative head is greater than for the sandy soil. In the
sandy soil, however, most of the pores are large and, once drained, the amount
of water retained is small (Hillel, 1971).
To obtain soil-water characteristic curves in the laboratory down to neg-
ative pressures of -800 cm (315 inches) of water, the modified Haines method
may be used for individual core samples. This method employs equipment shown
on Figure 5. Soil cores are carefully placed on the fritted glassbead plate
and saturated. After ensuring that the samples are saturated, excess water is
removed and the hanging water column is adjusted through a desired range of
negative heads. The cores are weighed after the soil water has equilibrated
with each successive negative head. At the end of the test, the soil cores
are oven-dried and intermediate masses are converted to volumetric water con-
tent values.
A plot of water content versus applied negative head for a desorption
cycle (obtained via the Haines method) is different from the curve obtained
during a wetting cycle. The difference between wetting and drying curves for
a sand is shown on Figure 6. Note that the water content values at a given
negative head are generally less during the wetting cycle than those during
the drying cycle. Such hysteresis is caused by entrapment of air in the pore
space during wetting. The figure also exhibits "scanning curves" which rep-
resent the change in the water content-negative pressure head relationship if
the soil is suddenly wetted during drainage, or vice-versa.
Topp and Zebchuck (1979) presented a variation of the hanging column
method which permits the simultaneous measurement of soil-water desorption
curves from a number of large-diameter cores. The soil cores contact a "ten-
sion medium" in a tank. The medium comprises glass beads with a narrow pore-
size distribution, permitting high hydraulic conductivity and high air-entry
values. Negative heads up to -100 cm (-40 inches) of water are applied to the
19
-------
cr
UJ
t-
o
o
Q
UJ
X
UJ
tt
en
UJ
a:
a.
UJ
S
z
CLAYEY SOIL
WATER CONTENT (% BY VOLUME )
Figure 4. Soil-water characteristic curve for a clayey soil
and sandy soil (after Mil lei, 1971).
20
-------
BUCHNER(PYREX)-
SOIL-
SINTERED GLASS
PLATE (POROUS)
RUBBER TUBING
SUPPORT ROD-
-ZERO MARK
~ho
-GRADUATED PIPETTE
,0®
,100 cm MARK
Figure 5. Modified Haines apparatus for obtaining soil-water
characteristic curves (after Day, Bolt, and
Anderson, 1967).
21
-------
UNSATURATED
TENSION SATURATED
-200 -100
PRESSURE HEAD, H, (CM OF WATER)
Figure 6. Soil-water characteristic curves for a sandy soil
during wetting and drying showing the effects of
hysteresis (after Freeze and Cherry, 1979).
22
-------
tension medium via a hanging water column. Heads from -100 cm (-40 inches)
to -500 cm (-200 inches) are obtained using a regulated vacuum. After equili-
brating at successive pressures, the cores are weighed. At the end of the
extraction period, the cores are oven-dried and intermediate masses are con-
verted to volumetric water content values.
Inasmuch as the air-entry value of the tension medium for the above meth-
ods is about 0.8 atmosphere of negative pressure, other methods are required
to obtain water content versus head relationships for greater negative head
values. Commonly, the pressure-plate method is employed. Details of the
method are presented by Richards (1965). As described by Paetzold (1979), the
soil sample is wetted and placed in a pressure chamber on a ceramic plate or a
cellulose-acetate membrane. A positive air pressure is applied to the soil
within the chamber causing water to flow from the sample through the membrane.
The air pressure value corresponds to the negative pressure head retaining
water in the sample. The various air-entry values of membranes permit a range
of possible pressures (for example, up to 15 atmospheres for ceramic plates).
After the water content of the sample equilibrates at the applied air pres-
sure, the sample is removed and weighed.
Soil-water characteristic curves may be used to construct the associated
pore-size distribution—water content curves. The relationship between suc-
tion, h, and effective pore-size diameter is approximated by the following
equation:
h = 2T cos a/pgr (5)
where T = surface tension of the liquid
a = contact angle
p = density of liquid
g = acceleration of gravity
r = effective radius.
Field Capacity (Specific Retention)
Field capacity may be defined in a general sense as the volume of water
which a unit volume of soil will retain against the force of gravity during
drainage. The concept of field capacity was developed many years ago by ag-
riculturists concerned with quantifying the amount of water to apply to irri-
gated fields. The original premise was that field capacity is a fixed value
representing the amount of water stored in a soil a certain time after drain-
age has "essentially ceased." By the same token, it is usually assumed that
during recharge (wetting), water movement will not occur until the medium has
been wetted to field capacity. Although these concepts of field capacity are
useful in an applied sense, one should be aware of certain technical limita-
tions. Hillel (1971) discussed such limitations in detail. Briefly, one
limitation is that the simplistic concept of field capacity fails to account
for the dynamic nature of soil-water movement. In particular, drainage does
23
-------
not really cease at field capacity but may continue at a slower rate for a
prolonged period of time. That is, "The redistribution process is in fact
continuous and exhibits no abrupt 'breaks' or static levels. Although its
rate decreases constantly, in the absence of a water table the process con-
tinues and equilibrium is approached, if at all, only after very long periods"
(Hillel, 1971). The modern conception of field capacity is that it is not a
unique soil property; instead, a range of values is possible.
For sandy soils, field capacity may be reached in a few hours. For soils
finer than sandy soils (e.g., sandy loams), 2 or 3 days .may be required to
reach field capacity, and for medium- to fine-textured soils, a week may be
required. For poorly structured clays, the time will be much greater (U.S.
EPA et al., 1977). Approximate values of field capacity (on amass basis)
vary from 4 percent in sands to 45 percent in heavy clays and up to 100 per-
cent or more in organic soils (Hillel, 1971). In terms of matric potentials,
field capacity values for sands range from 0.1 to 0.15 bar (1 bar = 0.9869 at-
mosphere). For medium- to fine-textured soils, the corresponding range is 0.3
to 0.5 bar (U.S. EPA et al., 1977). The value of 0.3 bar is chosen as an av-
erage value.
Knowing the water content values of given soil at field capacity and the
observed water content value at a given time, the depth of water applied at
the land surface to bring the soil to field capacity may be calculated from
equation 1. If the soil is layered, it will be necessary to account for the
sum of the water contents of individual layers (see Brakensiek, Osborn, and
Rawls, 1979, pp 523 and 524).
Among the factors affecting the apparent field capacity are the following
(Hillel, 1971): (1) soil texture, (2) type of clay (e.g., clays predominantly
comprised of the montmorillonite type exhibit a higher water holding capacity
at field capacity), (3) organic matter content (the higher the organic matter
level, the higher will be the field capacity), (4) antecedent water content,
(5) presence of impeding layers, and (6) evapotranspiration. Soil structure
is also an important factor in evaluating field capacity inasmuch as the large
interpedal cracks permit more rapid drainage than the micropores within the
soil blocks.
The water content of a soil sample at 0.3 bar is obtained in the labora-
tory using the pressure membrane method. This method was discussed above in
the section on water characteristics. An alternative method to estimate field
capacity is to assume that field capacity equals one half of the percent of
water content at saturation. That is, Fc = SP/2. Saturation percentage is
measured in the laboratory by determining the number of grams of water to sat-
urate 100 grams of air dry soil (U.S. EPA et al., 1977).
The above discussion relates to the concept of field capacity as employed
by agriculturists. The parallel term used by hydrogeologists is "specific re-
tention," defined as the "quantity of water per unit total volume which will
not drain under the influence of gravity" (Cooley, Harsh, and Lewis, 1972).
Specific retention may be visualized as the water remaining in the dewatered
region of the vadose zone after recession of the water table (Figure 7).
24
-------
IDEALIZED
SOIL
PARTICLES
RETENTION
SPECIFIC
,, YIELD j
POROSITY =
MEASURES THE VOID SPACE
AVAILABLE IN A MATERIAL TO
STORE WATER.
SPECIFIC YIELD +
MEASURES THE WATER
REMOVED BY THE FORCE
OF GRAVITY.
SPECIFIC RETENTION
MEASURES THE AMOUNT
OF WATER RETAINED
IN A MATERIAL.
Figure 7.
Specific Yield
Schematic representation of porosity, specific yield, and
specific retention (after Scott and Scalmanini, 1978).
"Specific yield" is a term employed by hydrogeologists to characterize
storage in an unconfined aquifer. That is, specific yield is "... the volume
of water that an unconfined aquifer releases from storage per unit surface
area of aquifer per unit decline in the water table" (Freeze and Cherry,
1979). Figure 7 shows the conceptual relationship between specific yield and
specific retention. As shown on Figures 3 and 7, the specific yield for a
media equals the porosity value minus the value of specific retention. Repre-
sentative specific yield values for valley sediments in California are shown
on Table 2.
Fillable Porosity
The volume of water that an unconfined aquifer stores during a unit rise
in water table per unit surface area is called the fillable porosity (Bouwer,
1978). As shown on Figure 8, the amount of water placed into storage (fill-
able porosity) during the rise of a water table is less than the corresponding
amount released during drainage (specific yield). The difference in storage
reflects hysteresis caused by air entrapped in pore sequences during the rise
of the water table.
FIELD METHODS FOR DETERMINING STORAGE PROPERTIES
OF THE VADOSE ZONE
As pointed out in a previous section, field methods for determining the
storage properties of the vadose zone are a mixture of techniques developed
by soil scientists and hydrogeologists. Field methods are reviewed for mea-
suring the following characteristics: (1) thickness, (2) porosity and bulk
density, (3) water content, (4) soil-water characteristics, (5) specific re-
tention, and (6) specific yield and fillable porosity.
25
-------
TABLE 2. COMPILATION OF SPECIFIC YIELD VALUES FOR VARIOUS
MATERIALS IN CALIFORNIA VALLEYS (after Cooley, '
Harsh, and Lewis, 1972)
Material
Average
specific yield
(percent)
Clay
Silt
Sandy clay
Fine sand
Medium sand
Coarse sand
Gravelly sand
Fine gravel
Medium gravel
Coarse gravel
2
8
7
21
26
27
25
25
23
22
-250-1
0.2 0.3
WATER CONTENT, 6
0.4
Figure 8. Water content distribution above a water table during a
fall (draining) and rise (filling) in the water surface.
26
-------
Thickness
The thickness of the vadose zone underlying the pollution source may be
estimated directly by determining water levels in the area using either exist-
ing wells or specifically constructed wells. If the area is underlain by
perched groundwater which is a source of potable water, the depth to the
perched water table should be selected as the depth of the vadose zone (Silka
and Swearingen, 1978). For confined leaky aquifers, the depth to the overly-
ing unconfined water table should be selected. For nonleaky confined aqui-
fers, the distance to the base of the uppermost confining layer could be used.
When estimating storage for pollutants, it is advisable to use the minimum
depth to groundwater, determined by examining historic water levels,
Indirect methods have been employed to detect the presence of shallow
perched groundwater, thus providing an estimation of the depth of the vadose
zone. For example, Estes and Simonett (1978) reported on using three imagery
techniques to detect perched groundwater: visible and reflected infrared im-
agery, thermal infrared imagery, and microwave techniques. They evaluated the
potential of these techniques for detecting perched water tables within 5 to
10 feet of the surface in the lower San Joaquin Valley, California. Thermal
infrared imagery appeared to be of greatest value in water table detection be-
cause of unique soil and water thermal characteristics. In particular, the
flux of heat through soils underlain by shallow water tables is different from
that through dry soils. Consequently, a diurnal surface anomaly occurs that
appears warmer at night and cooler during the day than surrounding areas that
do not have perched water tables (Estes and Simonett, 1978).
Bulk Density and Porosity
The bulk density of soil cores or clods obtained from the field may be
determined in the laboratory using techniques described in a previous section.
For in-situ measurements, a technique developed by soil engineers is the ex-
cavation method, also called the sand cone method, described by Blake (1965).
In this method, a hole is excavated to a desired depth. The hole is filled
with sand from a container using a double-cone arrangement. Ottawa sand is
commonly used because the bulk density is accurately known. Thus, knowing the
mass of sand to fill the cavity, the associated volume is determined by cal-
culation. The mass of soil removed from the hole is determined by oven dry-
ing. The bulk density of the soil is then determined by dividing the dry soil
mass by the calculated volume.
Alternatively, a balloon is placed in the cavity and subsequently filled
with water until the cavity is completely filled. The volume of water to fill
the hole is measured. The soil excavated from the hole is oven-dried and
weighed. The dry mass of the soil divided by the volume is equal to the bulk
density.
An indirect method for determining bulk density in-place is the gamma
radiation method. Two variations of the method are used: (1) the two-hole
transmission technique, and (2) the scattering technique. Sources for gamma
photons include cobalt-60 and cesium-137. Sodium iodide detectors are com-
monly used. Details of the two-hole transmission method are described by
27
-------
Blake (1965). Basically, the technique entails lowering probes containing a
source of gamma radiation in one of two parallel tubes (see Figure 11). Si-
multaneously, a probe with a radiation detector is lowered in a second tube
located a fixed distance from the first tube. In this method, the density is
measured in a collimated beam a few centimeters in vertical extent.
Keys and MacCary (1971) discussed the principles of the scattering tech-
nique. In practice, a probe containing both a source and a detector is low-
ered into a well. The source and detector are separated by shielding. In
contrast to the transmission method, the scattering method examines density
in a spherical volume. According to Paetzold (1979), the scattering method
requires a source of higher strength but it is not as accurate as the trans-
mission method. Unfortunately, the double-tube transmission method is pri-
marily suitable for determining bulk density in shallow soils rather than
deeper regions because of the difficulty of installing precisely parallel
tubes. Consequently, the scattering method is used to measure bulk density
in the vadose and groundwater zones.
An alternative technique involves placing a source of gamma photons on
the soil surface and measuring the density in an underlying hemispherical
volume.
Inasmuch as bulk density measured in the field also measures the density
of water present in the medium, a correction is necessary to convert the mea-
sured wet bulk density to a dry weight basis. In particular, a calibration
curve is prepared by measuring count rates in a number of wet soils and sub-
sequently determining the corresponding wet bulk density using laboratory
techniques described above (e.g., the calcium carbide method).
As discussed in the next paragraph, water content values of the soil in
the field are determined at the same time that a count rate is obtained.
Using the calibration curve and knowing the water content, the following re-
lationship is used to determine the dry weight bulk density (Blake, 1965):
Db = Dbw/(l + Pw/100) (6)
where Db is dry weight bulk density and Dbw is wet weight bulk density.
The water content of the soil may be determined by collecting field sam-
ples and oven drying or using the neutron logging method described later.
(When the latter method is employed, water content is determined on a volumet-
ric basis and the above equation must be modified.) Corey, Peterson, and
Wakat (1971) described a method for simultaneously measuring soil density and
water content by means of a dual source containing americium-241 and cesium-
137. Commercial units with dual sources are now available.
Once the bulk density has been determined, the total porosity may be cal-
culated from the dry weight bulk density using the relationship given in
equation 2, i.e.,
St = 100 (1 - (Db/pb)). (2)
28
-------
Water Content by Moisture Logging
For many years, agriculturists have employed the principle of neutron
moderation or thermalization to measure the volumetric water content of soils
in-situ. Recently, the technique has been used by hydrogeologists to monitor
water storage in the intermediate zone, particularly to delineate perching
layers and mounds (Keys, 1967) and also to estimate flow rates. Figure 9 is a
sequence of moisture logs from a well near the Santa Cruz River in Tucson
showing the growth and dissipation of mounds during an ephemeral recharge
event.
Principle of the Neutron Moderation Method--
The method of water content evaluation by neutron moderation depends on
two properties relating to the interaction of neutrons with matter: scatter-
ing and capture (Gardner, 1965). High-energy neutrons emitted from a radio-
active source may be slowed down or thermalized by collisions with atomic
nuclei. The statistical probability for such thermalization depends on the
"scattering cross section" of various nuclei. It turns out that the scatter-
ing cross section of hydrogen causes a greater thermalizing effect on fast
neutrons than occurs with many elements commonly found in soils. This forms
the basis for detecting the concentration of water in a soil (van Bavel,
1963). The second property of interest in the neutron moderation method is
capture of slow neutrons by elements present in the soil with the release of
other nuclear particles or energy. Cadmium and boron have extremely large
capture cross sections compared with hydrogen. The property of energy release
during capture serves as a means of detecting the concentration of slow
neutrons.
In operation, when a source of fast neutrons is lowered into a soil
through a suitable well or casing, a cloud of thermalized neutrons is estab-
lished (Gardner, 1965). This cloud reflects the moderating effect of scat-
tering cross sections of nuclei in the soil mass on fast neutrons. If a
suitable calibration is made to isolate the moderating effects of soil nuclei
other than hydrogen, changes in the volume of the thermalized cloud will re-
flect changes in water content. In general, the wetter the soil, the smaller
the volume of thermalized neutrons. Finally, a detector which relies on cap-
ture of thermalized neutrons may be used, in conjunction with suitable elec-
tronic circuitry, to measure the water content (on a volume basis).
I nstrument ati on~
Instrumentation used to measure water content by neutron thermalization
requires three principal components: (1) a source of fast neutrons, (2) a
detector of slow neutrons, and (3) an instrument to determine the count rate
from the detection equipment (Figure 10).
The first moisture loggers on the market relied on radium-beryllium as a
source of fast neutrons. Because of radiation health problems with radium-
beryllium, bulky shielding was necessary. A common source in units con-
structed today is americiurn-beryllium. Smaller units, such as those used by
agriculturists, contain sources in the millicurie range (e.g., 50, 100, 200
29
-------
DEC 21
DEC 24 DEC 27 DEC 30 JAN 3
MOISTURE CONTENT (VOL/VOL)
to
o
50 10 3O 50 10 30 50 10 30 50 10
JAN 7 JAN 10 JAN 14 JAN 19 MAY 3
MOISTURE CONTENT (VOL/VOL)
10 30 50 10 30 50 10 30 50 10 30 50 10 30 50
Figure 9. Sequence of water content profiles showing the growth and dissipation of perched
groundwater during recharge from an ephemeral stream (after Wilson and DeCook, 1968).
-------
CASING^
LAND SURFACE \
ELECTRONICS
AND
POWER SUPPLY
DETECTOR^
SPACERS -X-
SOURCE
WHEN BERYLLIUM IS .#
BOMBARDED WITH ALPHA :Z.
PARTICLES FROM %>
AMERICIUM, NEUTRONS &
ARE EMITTED :*
AVERAGE ENERGY LOSS V-.
PER COLLISION-
\
Am 02 Be
SOURCE
HYDROGEN
OXYGEN
63 %
12 %
SINGLE-CONDUCTOR
CABLE TO LOGGING
EQUIPMENT
EUTRON
|Q5ev
°R|Mev
NEUTRON MODERATION
EPITHERMAL NEUTRON
I TO 100 ev
NEUTRONS
SECONDARY
GAMMA PHOTONS-ENERGY
• THERMAL V7N CHARACTERISTICS
OF ELEMENT
LV7)
mwy
HYDROGEN
ATOM
^
NEUTRON CAPTURE
Figure 10. Equipment and principles of neutron moisture
logging (after Keys and MacCary, 1971).
31
-------
millicuries) necessitating small-diameter casing (e.g., 2-inch). Larger units
with multicurie sources, such as those used in the petroleum industry for po-
rosity determination, are used to log wells with larger casings.
Detectors of slow neutrons comprise materials with high capture cross
sections (e.g., boron). A charged particle is emitted during capture which
can then be detected by a solid or gaseous counting device (van Bavel, 1963).
The predominant detector is lithium-enriched BF3-
Both the source and detector are located in a cylindrical tool which is
lowered via a cable into access tubing either by hand or motor drive. The
depth of measurement is determined either by graduations on the cable or by
some type of counter. The relative positions of source and detector in the
design of the tool are important. This question was reviewed in detail by van
Bavel (1963).
The pulse emitted during the nuclear reaction resulting from capture is
amplified by a preamplifier within the down-hole tool and sent through the
cable to an above-ground meter for counting. Two portable meters are commonly
used: rate meters or sealers. Gardner (1965) discussed the relative advan-
tages of rate meters and sealers. Count rate is converted to water content
via suitable calibration curves.
The type of meter available for soils work usually involves placing the
down-hole tool at a discrete depth, taking a count or number of counts, and
then moving the tool by hand to another discrete depth. This process is not
troublesome for shallow soils studies. However, for deeper access wells in
the intermediate zone, an inordinate amount of time would be involved. A mo-
torized unit is available on the market which permits lowering the tool in the
well at a constant rate via a motor drive. Concurrently, internal electronic
components translate pulse rate into water content which is recorded on a
recorder.
Calibration—
The individual neutron moisture logger should be calibrated against sam-
ples of known water content. Such calibration may be done in the field by
taking a count at a given soil depth followed by a determination of the water
content of soil cores from the same depth. Alternatively, calibration stan-
dards can be used, e.g., laboratory soil samples of known water content.
Field Installation of Access Wells-
Access tubes for neutron moisture logging are usually constructed of
seamless steel or aluminum tubes. Plastic wells may cause difficulty in that
hydrogen or chlorine atoms in the tubing may moderate the thermal neutrons,
interfering with soil moisture evaluation. Aluminum wells may deteriorate in
saline wells.
The inside diameter of wells should be as close as possible to the out-
side diameter of the probe. Work by Ralston (1967) showed that for a 100-mc
Am-Be source in a 1.5-inch tool, water content could not be accurately
32
-------
evaluated in wells greater than 4 inches. Brakensiek, 0shorn, and Rawls
(1979) recommend that the radial air gap should not exceed 0.02 inch.
Wells installed for shallow water content monitoring can be easily in-
stalled by successively augering within the tube and driving the tube. Myhre,
Sanford, and Jones (1969) reported on a simple power-driven auger for install-
ing wells to about 5-foot depths.
For deeper wells, it will be necessary to use standard drilling tech-
niques, e.g., core drilling, rotary, etc. During installation by drilling
techniques, it is essential that a tight fit be established between the cavity
and well to minimize the amount of vertical leakage of water. Drilling mud is
not recommended for the drilling operation because it inhibits caving around
the well and also interferes with water content observations. Some workers
have attempted to backfill the gap between the cavity wall and the casing with
fine material (Brackensiek, Osborn, and Rawls, 1979) or by grout. The results
of field studies by Halpenny (personal communication, 1979) showed that bound
water within grout markedly attenuates the flux of fast neutrons from a
source. Consequently, the sensitivity of a logger in detecting water content
changes in the surrounding formations is correspondingly reduced.
Advantages and Limitations of Neutron Moisture Logging--
The principal advantage of neutron moisture logging is that water content
profiles are obtained in-situ without disturbing the soil. Thus, a history of
profiles can be established at the same site during a monitoring program.
Moisture logs clearly show the presence of perched water tables, together with
their growth and dissipation (see Figure 9). Water content changes at a given
depth in a succession of wells may provide clues on lateral flow velocities.
Moisture logs may also be used to estimate unsaturated flux during drainage
(see below).
Certain limitations should be noted. The presence in excessive concen-
trations of other fast neutron moderators (e.g., boron, chlorine) may cause
erroneous water content determinations. Neutron moisture logs indicate only
the water content of soils and nothing about the energetics of soil water.
Water content values must be translated to equivalent soil-water suction val-
ues via a soil-water characteristic curve. Another related point is that wa-
ter may move through a specific subsurface horizon without causing a change in
storage. Consequently, a neutron log may not manifest water movement in this
case.
Nonuniform conditions (e.g., stones or cracks) reduce the reliability of
the method (Gairon and Hadas, 1973). Caution is required in handling the nu-
clear source and equipment is very expensive (Gairon and Hadas, 1973). Since
water content is measured in a soil volume of undetermined size, it is not
possible to relate water content directly to a specific sampling depth (i.e.,
the resolution is poor).
33
-------
Water Content Via Gamma Ray Transmission
A second technique using radioisotopes to indirectly estimate water con-
tent involves gamma ray transmission, the technique discussed earlier for de-
termining soil density. Principles of the method for water content evaluation
are reviewed by Gardner (1965), Bouwer and Jackson (1974), and Reginato and
van Bavel (1964). As stated by Gardner (1965): "The degree to which a beam
of monoenergetic gamma rays is attenuated ... in passing through a soil column
depends upon the overall density of the column. If the density of the soil
less its water content is constant, then changes in attenuation represent
changes in water content."
As indicated in the discussion of the use of the method for determining
bulk density, the basic components for the method include a source of monoen-
ergetic gamma rays such as cesium-137 and a detector such as Nal (TI) scintil-
lation crystal. Accessories include a high-voltage supply, amplifier, sealer,
timer, spectrum analyzer and pulse height analyzer, and a photomultipl ier tube
(Bouwer and Jackson, 1974).
For field usage, two parallel access wells are required, one for the
source and one for the detector (see Figure 11). Both source and detector
must be properly positioned with respect to each other during sampling.
Advantages and Limitations of the Gamma Transmission Method--
According to Brakensiek, Osborn, and Rawls (1979), advantages of the
method include the following:
Measurements ... by gamma-ray attenuation provide a nondestruc-
tive method of determining soil moisture and soil density. The
site sampling can be repeated as often as desired. The mea-
surements can be made vertically as close as one inch. Mea-
surements at one inch from an interface (surface) are valid.
The precision is high and accuracy of the results is excellent.
Errors in measurement can usually be detected and corrected at
time of measurement. No particular health hazard is involved
in this type of measurement provided simple safety precautions
are observed and the radioactive source is shielded when out of
the access tube.
Disadvantages of the method include: changes in bulk density in shrinking and
swelling materials may affect calibration, equipment is expensive, and insta-
bilities in count rate may occur (Reginato and Jackson, 1971; Reginato, 1974).
Also, the need to maintain access wells a constant distance apart is imprac-
tical for deep sampling, particularly in rocky materials.
Water Content Using Tensiometers
Tensiometers are used to measure (negative) soil-water pressures during
unsaturated flow. If caution is exercised, they may be used to estimate
soil-water content via suitable soil-water characteristic curves. In parti-
cular, knowing the soil-water pressure in the field via tensiometer data, a
34
-------
o o o o
o o o
GROUND SURFACE
PREAMPLIFIER
PHOTOMULTIPLIER TUBE
SCINTILLATION CRYSTAL-
Figure 11. Dual probe used to measure water content by gamma ray
attenuation (after Brakensiek, Osborn, and Rawls, 1979)
35
-------
soil-water characteristic curve (e.g., Figure 6) for the field soil could be
consulted to determine the corresponding water content value.
Basically, a tensiometer consists of a porous ceramic cup cemented to a
rigid plastic tube, a small-diameter (spaghetti) tubing leading to a manometer
and terminating in a reservoir of mercury, and a filler plug in the rigid
plastic tube (see Figure 12). Except for the portion of the small-diameter
tubing filled with mercury, the internal volume of the system is completely
filled with water. Alternative designs use bourdon type gages in lieu of mer-
cury manometers.
MANOMETER
FILLING PLUG
MERCURY LEVEL IN MANOMETER
MERCURY RESERVOIR
TENSIOMETER
BODY (PVC)
CERAMIC.
CUP
AIR
SOIL
PARTICLE
WATER
Figure 12. Schematic representation of tensiometer and section
through the ceramic cup (after Nielsen, Biggar, and
Erh, 1973; and Bouwer, 1978).
When properly installed in the soil, the pores in the cup will form a
continuum with the pores in the soil. Water will move either into or out of
the tensiometer system until an equilibrium is attained across the ceramic
cup. The mercury level in the manometer tubing will correspondingly adjust.
From Figure 12, the soil-water pressure head, h, will be (Nielsen, Biggar, and
Erh, 1973):
h = -(12.55 x - y - z)
(7)
36
-------
where x = height of mercury column
y = distance from soil surface to mercury reference level
z = depth below land surface to ceramic cup.
Holmes, Taylor, and Richards (1967) and Brakensiek, Osborn, and Rawls
(1979) noted the precautions to be taken during installation of tensiometers.
The individual units should be filled with deaired water and the cups should
be immersed in water during transport to the field. A hole slightly larger in
diameter than the tensiometer is dug to the desired depth. A slurry is pre-
pared from native soil and poured into the bottom of the hole. The tensiome-
ter cup is then forced into the slurry and the tube backfilled with soil.
Care must be exercised to ensure that surface water does not leak down along
the tube.
Holmes, Taylor, and Richards (1967) pointed out that to ensure minimum
effect of the tensiometer on the soil-water system, the sensitivity of the
unit should be as large as possible. Furthermore, the response lag of the
system should be kept to a minimum, particularly in rapidly changing soil-
water systems.
Bianchi (1967) presented the design of a tensiometer coupled with strain
gages, in lieu of the mercury system, to permit the conversion of soil-water
pressure into the equivalent electrical resistance. His system would allow
deep emplacement of tensiometers and the use of automatic recording equipment.
Watson (1967) presented a design for a tensiometer-pressure transducer system
which would also allow for the recording of soil-water pressure at depth.
Figure 13 illustrates this system. He recommended such a device for recharge
basin studies, and indicated that the unit can also be used for automatic re-
cording of soil-water pressures at remote stations. In lieu of employing one
transducer per unit, Brakensiek, Osborn, and Rawls (1979) described a strain
gage/recorder combination using one transducer to service several tensiometers
in turn.
Advantages and Limitations of Tensiometers for
Measuring Water Content--
The principal advantage of tensiometer units is that they provide in-
place continuous readings of soil-water pressure, which may be translated into
water content values (Gairon and Hadas, 1973). Two limitations of tensiome-
ters are (1) they are useful only in measuring soil-water pressures in excess
of about -0.8 bar, and (2) they are sensitive to temperature. The most seri-
ous problem with the method is that, because of hysteresis, it is necessary to
know whether the soil is wetting or drying when the tensiometer is read. In
particular, during the redistribution of soil water following infiltration,
portions of the profile may be wetting while others are drying. Thus, a ten-
siometer reading could represent the value of a wetting curve, a drying curve,
or one of an infinite number of scanning curves.
37
-------
RUBBER STOPPER.
ELECTRIC CONNECTION.
GASKET>
LOCK WASHER-
CERAMIC.
AIR-PRESSURE CONNECTION
1/8 STD. FLARE FITTING
STAINLESS STEEL TUBING
'7/8" DIA. 3/16" WALL THICKNESS
.GASKET
.REFERENCE PORT
^STATHAM PRESSURE TRANSDUCER
PM 131 TC
• FIBER GASKET
• SENSING DIAPHRAGM
Figure 13. Cross section of tensiometer-pressure transducer
assembly (after Watson, 1967).
38
-------
Water Content Using Electric Resistance Blocks
Electrical resistance blocks, used to measure either soil-water content
or soil-water pressure, consist of electrodes embedded in suitable porous ma-
terial. Plaster of Paris, fiber glass, and nylon cloth have been used. The
principle of operation of these blocks is that water content (or negative
pressure) within the blocks responds to the water content (or suction) of the
soil with which the blocks are in intimate contact (Holmes, Taylor, and Rich-
ards, 1967). Correspondingly, the electrical properties of the block also
change as reflected by measurement of resistance.
Moisture blocks are calibrated in soil from the site at which they are
to be installed. Such calibration involves evaluating resistance readings
against a range of soil-water contents or negative pressures.
Phene, Hoffman, and Rawlins (1971) presented an alternative to the stan-
dard gypsum block comprising a heat source and sensor embedded in a ceramic
body (see Figure 14). This device measures both soil-water content and matric
potential. The basis of the units is that air is a good thermal insulator
with respect to water. With drying and entry of additional air into the me-
dia, the remaining water films become thinner. Consequently, the flow path
for heat conduction increases and a larger temperature gradient is needed to
dissipate heat. Phene, Rawlins, and Hoffman (1971) also discussed laboratory
and field tests with their sensor. Apparently, the accuracy of the sensor was
as good as or better than other techniques used to measure matric potential.
Advantages and Limitations of Gypsum Blocks—
Holmes, Taylor, and Richards (1967) discussed the advantages of resis-
tance blocks indicating that (1) they appear to be best suited for general use
in study of soil-water relations, (2) they are inexpensive, and (3) they can
be calibrated for either suction or water content. Generally, blocks are used
for soil-water pressures less than -0.8 atmosphere.
Phene, Hoffman, and Rawlins (1971) reviewed the problems with gypsum
blocks together with alternatives to circumvent these problems:
In a gypsum block the water content is measured by the elec-
trical resistance between two imbedded electrodes. Several
problems are usually encountered when using gypsum blocks;
first, salinity affects electrical conductivity independently
of water content; second, the gypsum used in attempting to mask
variations in soil salinity eventually dissolves resulting in
an unstable matrix for the sensor; third, contact resistance
between the porous body and the soil can restrict the exchange
of water between the block and the soil; and fourth, the hys-
teresis in the water content-matric potential relation of the
porous body can cause errors in the interpretation of the data
depending on whether the measurements are taken during a drying
or wetting cycle. Elimination of the first problem requires a
measurement of the water content of the porous body that is in-
dependent of salinity. This, in turn, would eliminate the need
39
-------
oGXEDOQB
Figure 14. Thermistor soil cell, meter, and hand auger. Cell
is being placed in a soil cavity (after Soiltest
Inc., 1976).
for the buffer material so that a stable matrix could be used.
Contact resistance between the block and the soil can be mini-
mized by packing the dry soil carefully around the sensor and
then irrigating. Generally, the status of soil water is only
of interest during drying conditions, since after a rain or an
irrigation, the soil water content changes rapidly as the wet-
ting front passes. Thus, hysteresis is of little concern and
only a desorption calibration curve is needed.
Blocks are rather insensitive to moisture changes, in the wet range
(Gairon and Hadas, 1973). Hence, blocks are commonly used as an adjunct to
tensiometers to monitor moisture content values at pressures less than -0.8
atmosphere.
40
-------
Water Characteristic
As discussed by Bouwer (1978), the negative pressure head in the vadose
zone equals the vertical distance above a water table provided vertical flow
is not occurring. Consequently, for uniform conditions, a plot of the change
in volumetric water content with distance above a water table represents the
water characteristic curve of the vadose zone material, again assuming that
there is no vertical flow. In practice, the volumetric water content could be
determined using the neutron moisture logger described previously.
Bouwer (1978) described the nature of characteristic curves for layered
conditions. The negative pressure head still equals the vertical distance
above the water table. However, the water content-head relationship depends
on the soil at the measurement point. In other words, an irregular water con-
tent distribution occurs above the water table, indicating that although head
changes occur continuously with vertical distance, discontinuities may occur
in water content distribution. Bouwer (1978) indicated that certain fine-
textured soils may actually be saturated (i.e., contain perched groundwater
lenses), whereas coarser-textured material above and below may be unsaturated.
Specific Yield, Fillable Porosity, and Specific Retention
A number of techniques are available for estimating the specific yield of
the groundwater zone. If the storage properties of material above and below
the water table are similar, the specific yield value determined for a ground-
water zone could be assumed to approximate the specific yield in the vadose
zone. In actuality, this assumption may be very tenuous because of the marked
variations in lithology noted in the field.
The most common technique for estimating specific yield is to conduct
pumping tests on wells. (Such tests also provide data on the transmissivity,
T, of the aquifer.) Common testing techniques are described by Lohman (1972).
Alternatively, if long-term groundwater withdrawals are known together with
the corresponding head changes, specific yield values of the regional ground-
water system could be estimated (e.g., Matlock and Davis, 1972).
As indicated by Stallman (1967) and Bouwer (1978), specific yield values
determined from short-term pumping tests may underestimate the true value be-
cause of delayed drainage. Instead of relying on results from short-duration
tests, Stallman (1967) suggested that a more realistic estimate of specific
yield could be obtained by observing variations in water content profiles dur-
ing the decline of a water table. In particular, the specific yield could be
calculated by using a sequence of water content profiles to determine the vol-
ume of water drained near the water table during a decline in the water table
(Bouwer, 1978). Meyer (1963) used this technique to estimate temporal varia-
tions of the apparent specific yield near pumping wells.
The use of neutron moisture logs appears to be the most suitable approach
for estimating the specific yield of vadose zone sediments near the water ta-
ble. Such logs could also be used to estimate the specific yield of deposits
in other regions of the vadose zone where perched groundwater bodies are gen-
erated during cyclic recharge events.
41
-------
Neutron moisture logs also afford the means of estimating the finable
porosity of sediments near the water table, i.e., the volume of water placed
into storage per unit rise in water level. As indicated in another section,
because of entrapment of air during the rise of a water table, the fill able
porosity will initially be less than the specific yield.
The specific retention may also be determined from neutron moisture logs
obtained during the recession of a water table. It will be recalled that the
specific retention equals the volume of water retained against the pull of
gravity during drainage.
UTILIZATION OF MONITORING METHODS FOR OBSERVING
STORAGE CAPACITY BENEATH A POLLUTION SOURCE
The methods described in this section are examples of the available tech-
nology for estimating the capacity of a given volume of vadose zone underlying
a pollution source to store water and water-borne pollutants. The example
given in the section entitled "Thickness" showed the maximum (theoretical)
capacity. In light of the discussion on field capacity, however, it is appar-
ent that the actual storage capacity would be less than the maximum because of
drainage. Another example using equation 1 will illustrate this point.
Repeating equation 1:
APD
_
100 D vz.
w
In practice, bulk density values (D^,) required to solve equation 1
could be evaluated using the gamma ray transmission method for shallow depths
and the scattering method for deeper regions. Water content values under
background conditions and at field capacity could be estimated from such al-
ternative methods as neutron moisture logging, gamma ray transmission or scat
tering, tensiometers, and moisture blocks.
Assume again the following conditions: (1) 100-foot thick vadose zone,
(2) bulk density of 8.1 lb/ft3 (1.3 gm/cm3), and (3) water content of dry
vadose zone material = 10 percent of weight. Assume also that the water con-
tent of the material comprising the vadose zone at field capacity is 20 per-
cent by weight.
From equation 1:
dw =
= 13 feet.
For this hypothetical case, 13 acre- feet of water could be theoretically
stored in the vadose zone per acre of surface area when the media is at field
capacity.
42
-------
SECTION 5
MONITORING WATER MOVEMENT IN THE VADOSE ZONE
In Section 4, methods for estimating the capacity of the vadose zone to
store water and water-borne pollutants were reviewed. Although a knowledge of
storage capacity is important, individuals charged with monitoring water-borne
pollutants are more likely to be concerned with the rate of movement through
the vadose zone. In other words, how fast are water and entrained constitu-
ents moving from a surface source to the water table? In this section, the
movement of water alone is examined. Relationships affecting the mobility of
pollutants are reviewed in Section 6.
In practice, it is a fairly simple task to measure the rate of water
movement across a soil-water interface at the land surface. For example,
seepage in a pond can be estimated by determining water level changes and
using appropriate depth-volume-surface area relationships. In contrast, the
process of water movement through the soil and lower vadose zone underlying
the source is of staggering complexity. Elements contributing to the complex-
ity of flow include variations in the state of water saturation and spatial
variations in the physical/hydraulic properties of the vadose zone.
In light of the difficulties in attempting to describe water movement in
the vadose zone, it is not possible to present exact techniques for estimating
transit time of water and water-borne pollutants through this region. How-
ever, a number of indirect methods are available which will be reviewed in
this section. As in the previous section, a review of technical elements is
presented first, together with laboratory techniques. Field methods are dis-
cussed in a later subsection. Finally, a possible arrangement of facilities
for estimating movement in both saturated and unsaturated zones is presented.
TECHNICAL REVIEW
For ease of discussion, the following categorization will be employed:
(1) infiltration at the land surface, (2) unsaturated flow in the vadose zone,
and (3) flow in saturated regions of the vadose zone. This subdivision is
somewhat arbitrary inasmuch as infiltration is generally regarded as an un-
saturated flow process.
Infiltration
Infiltration is the process by which water enters a soil. The maximum
rate at which water enters a soil is the infiltration capacity. The Soil
43
-------
Conservation Service has classified soils on the basis of infiltration rates
and transmission rates as follows (U.S. EPA et al., 1977):
Transmission rate
Class (in/hr)[cm/hr)
Very slow <0.06 <0.15
Slow 0.06 to 0.2 0.15 to 0.51
Moderately slow 0.2 to 0.6 0.51 to 1.5
Moderate 0.6 to 2.0 1.5 to 5.1
Moderately rapid 2.0 to 6.0 5.1 to 15.2
Rapid 6.0 to 20.0 15.2 to 50.8
Very rapid >20.0 >50.8
The infiltration rate of soil wetted continuously is initially high but
decreases steadily with time, approaching an asymptotic value. The asymptotic
value is approximately that of the saturated hydraulic conductivity of the
soil. The physical reason for the typical decrease in infiltration with time
is described by Hi 11 el (1971):
When infiltration takes place into an initially dry soil, the
suction gradients are at first much greater than the gravita-
tional gradient As the water penetrates deeper and the
wetted part of the profile lengthens, the average suction gra-
dient decreases, since the overall difference in pressure head
(between the saturated soil surface and the unwetted soil in-
side the profile) divides itself along an ever-increasing dis-
tance. This trend continues until eventually the suction
gradient in the upper part of the profile becomes negligible,
leaving the constant gravitational gradient as the only remain-
ing force moving water downward in this upper or transmission
zone. Since the gravitational head gradient has the value of
unity ... it follows that the flux tends to approach the hy-
draulic conductivity as a limiting value.
The distribution of water as a function of.depth within a soil during
infiltration is generally called a water content profile. As first described
by Bodman and Coleman (1944), water content profiles may be divided into three
distinct zones: (1) a transmission zone in which the upper few centimeters of
soil may be saturated and below which the soil is very nearly saturated—this
zone continually lengthens during infiltration; (2) a wetting zone exhibiting
rapid changes in water content with depth and time; and (3) the wetting front,
the visible limit of water penetration. Figure 15 shows water content pro-
files during infiltration into a soil which is continuously flooded and into
the same soil when water is applied slowly by sprinkler. Because of this slow
application rate, the water content values in the profile during sprinkling at
all times remain below the values for the ponded case.
Following the cessation of surface flooding during a water-spreading op-
eration, water becomes redistributed in the soil profile. The upper portion
of the profile drains, while the lower portion continues to wet. The physics
of redistribution is complex, particularly because of hysteresis.
44
-------
WATER CONTENT,
X
Q.
O
CO
SOIL SURFACE
SOIL CONTINUOUSLY
PONDED
SOIL WETTED BY
SPRINKLING
Figure 15. Water content profiles of a soil wetted slowly by
sprinkling and a soil continuously ponded (after
Hillel, 1971).
45
-------
For years hydrologists have attempted to model infiltration using (1)
algebraic and empirical equations, (2) approximations of rigorous flow equa-
tions, and (3) equations derived from simplifications of the physical system
(Brakensiek, 1979). The fact that the work still continues is apparent by
examining the papers presented at a recent workshop on infiltration research,
published by the Sciences and Education Administration (1979). The interested
researcher should consult this publication for details.
Of the algebraic/empirical equations for describing infiltration, Braken-
siek (1979) recommends the Green and Ampt (1911) equation and the Philip
(1969) two-parameter equation. The Green and Ampt equation was discussed in
detail by Childs (1969), Philip (1969), Hillel (1971), and Bouwer (1978).
Assumptions of the Green and Ampt approach are (Hillel, 1971): "...
There exists a distinct and precisely definable wetting front, and that the
matrix suction at this wetting front remains effectively constant, regardless
of time and position." (Matrix suction is a measure of the sorptive capacity
of a soil for water.) The Green and Ampt equation has been found to be satis-
factory for describing infiltration into initially dry coarse-textured soils.
As presented by Bouwer (1978), the Green and Ampt equation is as follows:
V. V"*'f'h"- (8)
1 Lf
where V-j = infiltration rate
K = hydraulic conductivity of wetted zone
Hw = depth of water above soil
her = critical pressure head of soil for wetting
Lf = depth of wetting front.
Brakensiek (1979) itemized parameter estimation procedures for the Green and
Ampt equation.
Philip (1969) derived a two-parameter algebraic equation from physical
principles to describe vertical one-dimensional infiltration by solving the
basic partial differential equation for unsaturated flow. Philip's solution
described the time dependence of infiltration in terms of a power series
which, when t is not too large, reduces to:
It = Stl/2 + At; Vi = 1/2 St-1/2 + A (9)
where 1^ = cumulative infiltration
v-j = infiltration rate at time t
46
-------
S = sorptivity
A = a constant.
According to Bouwer (1978), the sorptivity depends on the pore configura-
tion of the soil, the initial water content, and the depth of applied water.
Approaches for estimating the parameters in the Philip two-parameter equations
are summarized by Brakensiek (1979). For short infiltration events, A is
taken to equal K/2, where K is the saturated hydraulic conductivity. For long
infiltration events, A is equal to K (Bouwer, 1978).
The initial infiltration rate of a soil is theoretically infinite (Bou-
wer, 1978). Note that both equations 8 and 9 comply with this requirement.
In the case of equation 8, the value of the initial rate approaches infinity
because the depth to the wetting front is zero (Lf = 0). For equation 9,
the infiltration rate, V-j, at time t = 0 is also infinite because the first
term on the right-hand side overwhelms the second (A) term.
An empirical infiltration curve commonly used by hydrologists is that of
Morton. This equation may be written in the following form:
Vi = Vco+ (V0 - VoJe-Kt (10)
where V-j = infiltration rate at time t
V0 = initial infiltration rate
Voo = final infiltration rate
e = base of natural logarithms
K = a constant depending on soils and vegetation.
Values for K are found from field infiltration data by expressing equa-
tion 10 in logarithmic form. A basic problem with Morton's equation is that
it does not satisfy the theoretical requirements that the initial infiltration
must be of infinite value (Bouwer, 1978, p 256). According to Bouwer, the
equation is best suited to describing infiltration when water is applied by
rain or sprinkling for short periods.
Factors affecting the infiltration capacity (and thus the potential
amount of pollutants entering the lower vadose zone) of a soil include the
following: soil texture, soil structure, initial water content, presence of
shallow impeding layers, entrapped and confined air, biological activity, en-
trained sediment, and salinity of applied water. The effect of soil texture
on infiltration is complex. For example, a clay soil may posses a rather high
sorptivity value but water movement is retarded by the energy losses in the
fine pores. Inasmuch as sorptivity decreases with time, the long-term infil-
tration rate depends on the hydraulic conductivity of a soil. Thus, a sandy
soil should have a greater long-term infiltration rate than a clay soil. Soil
structure has a profound effect on infiltration rate, particularly at the be-
ginning of infiltration. This is, water moves preferentially very rapidly
47
-------
through the larger interpedal cracks at the beginning of infiltration. Later,
as the soil swells closing the cracks, water movement occurs through the soil
blocks. Hillel (1971) pointed out that the long-term infiltration rate of a
highly structured soil approaches the rate of a uniform soil because of the
control imposed by the hydraulic conductivity of the lower soil zone.
The initial water content of a soil affects the infiltration rate because
of the concurrent effect on sorptivity and available porosity. Thus, the
dryer the soil, the greater will be the initial infiltration rate compared to
a wetter soil (see Figure 16). The final or long-term infiltration rate of a
soil, however, is independent of the initial water content. Shallow impeding
layers within the soil zone retard the infiltration rate compared to a non-
stratified soil. As pointed out by Hi 11 el (1971), the effect is the same
whether a fine soil is underlain by a lens of coarser material, or vice-versa.
Air entrapped in the soil pores during infiltration effectively behaves as
solid grains in restricting water movement. Eventually, however, the air may
dissolve in the water, causing an increase in intake rates. Confined air
ahead of the wetting front may also restrict infiltration because of the
buildup in air pressure due to a restricting lens or shallow water table (Wil-
son and Luthin, 1963). In fact, the air pressure may increase to the point
that the air-entry value of the soil is exceeded and air escapes at the
surface.
Fine sediment entrained in the water may clog the surface pores and ef-
fectively create a barrier to the infiltration process. Water movement in the
underlying pores becomes unsaturated and flow occurs primarily in the finer
pore sequences. Microbial activity may also result in a clogging of soil
pores with the byproducts of their metabolic processes, e.g., slimes or gas-
ses. Microorganisms may produce a secondary effect in that the pH of the
overlying water may be increased due to microbial respiration, resulting in a
precipitation of carbonates within the surface pores (Bouwer, personal com-
munication, 1978).
The salinity of the applied water may be a dominant factor in affecting
infiltration for soils containing substantial clay fractions (i.e., soils with
a tendency to shrink and swell). Thus, it is well known among agriculturists
that a high sodium concentration in the applied water relative to calcium and
magnesium may disperse the clays, slowing down intake rates. This effect is
discussed in detail in the section "Saturated Flow in the Vadose Zone."
Unsaturated Flow in the Vadose Zone
The two fundamental mathematical expressions describing the flow of water
in porous media are Darcy's equation and the equation of continuity. For un-
saturated flow, Darcy's equation is written (after Nielsen, Biggar, and Erh,
1973) as:
J = -K(9)VH (11)
where J = specific discharge or flux (volume per unit area per unit time);
also called Darcian flux
48
-------
UJ
tr
z
o
<
INITIAL WATER CONTENT OF A
-------
K(9) = hydraulic conductivity, expressed as a function of water
content, 0
H = total hydraulic head, the sum of soil-water pressure head,
h, and potential head, z
VH = the hydraulic gradient, expressed in vector form.
The specific discharge has the units of velocity. The negative sign is
used by convention because flow occurs in the direction of decreasing hydrau-
lic head.
As expressed by equation 11, the specific discharge is a macroscopic
quantity inasmuch as the cross-sectional area of flow includes both solids and
voids. A more representative depiction of velocity would exclude solids from
the flow path and account for the water content of the void spaces. Conse-
quently, an alternative expression for flow in unsaturated media is as
follows:
(12)
where V is the average linear velocity (Freeze and Cherry, 1979).
As pointed out by Freeze and Cherry (1979), the average linear velocity
is also a macroscopic quantity. The actual or microscopic velocity within
pore sequences would be greater than V because the tortuous flow paths are
greater than the linearized path assumed in defining v. However, note that "v
will always be greater than J for a given soil, permitting a more realistic
estimate of travel time of nonreactive water-borne pollutants.
The second important flow equation is the equation of continuity which
in one dimension (the z direction) is written as:
) (13)
where t is time.
Basically, the hydraulic conductivity is a factor representing the ease
with which water moves through a soil or aquifer. In equation 12, the unsat-
urated hydraulic conductivity is shown to be a function of water content, 6.
However, as discussed in a previous section, 6 depends on the negative pres-
sure head, h. As a consequence of the h-9 relationship, the unsaturated hy-
draulic conductivity is also a function of the negative pressure head. The
relationship between the unsaturated hydraulic conductivity and negative pres-
sure head is shown on Figure 17. The effect of hysteresis is also shown on
the figure.
The important feature of Darcy's equation relative to the movement of
water into the vadose zone beneath a disposal area is that two factors must'
be defined in order to characterize the flow rate. These factors are the
50
-------
UNSATURATED
TENSION SATURATED
rO.03
o
-0.02
o
0
8
(E
Q
-0.01
-300 -200 -100
PRESSURE HEAD, H, (CM OF WATER)
Figure 17. Effect of the negative pressure of soil water on the
unsaturated hydraulic conductivity. The figure also
shows the effect of hysteresis on K (after Freeze
and Cherry, 1979).
hydraulic conductivity and the hydraulic gradient. Thus, the purpose of lab-
oratory and field techniques (discussed below) is to determine K(0), or
K(h), and the hydraulic gradient.
A number of laboratory techniques are available for estimating the rela-
tionship between the unsaturated hydraulic conductivity and pressure head or
water content. At the outset, it should be emphasized that because the hy-
draulic conductivity is dependent on water content (and thus head) in unsatu-
rated soils, it is extremely difficult to measure K(6) and K(h) in the
51
-------
laboratory or field. Thus, for unsaturated soils, the water content, 6,
changes continuously along the flow path and K(0) values also vary continu-
ously. That is, a unique value for the hydraulic conductivity cannot be
determined.
One laboratory technique for determining K(h) as a function of h is the
so-called long-soil column method. Briefly, the method entails introducing
water at the soil surface of a long-soil column at a constant rate such that
unsaturated flow is produced in the soil. The soil at the base of a column
may eventually become saturated but the remainder of the soil will be unsat-
urated. The test is continued until the hydraulic gradient becomes equal to
one or, in other words, 0 is nearly constant throughout the upper region of
the column. Hydraulic heads are measured via tensiometers placed in the col-
umn. When the hydraulic gradient equals one, the unsaturated hydraulic con-
ductivity equals the flow rate divided by the cross-sectioned area from
Darcy's equation. Note that the resultant hydraulic conductivity is related
to the water content in the column at the time of the experiment. The exper-
iment may be repeated at different flow rates and different water content val-
ues to provide a range of K(0) - 6 relationships.
Alternative laboratory methods for determining the unsaturated hydraulic
conductivity at successive steady-state flow rates are presented by Klute
(1965). These methods are suitable for soil cores. One of the most common
laboratory methods, the pressure outflow technique, consists of applying suc-
cessive air pressures to a moist soil section in an appropriate chamber.
Using the air pressures and related water content values, an equation (Kirk-
ham and Powers, 1972, p 274) is solved for corresponding K(0) values. In
contrast to such steady-state methods, Alemi, Nielsen, and Biggar (1976) pre-
sented centrifugal techniques for determining the unsaturated hydraulic con-
ductivity during transient flow conditions.
Indirect methods for estimating K(9) have relied on soil-water charac-
teristics (for example, Millington and Quirk, 1959; and Jackson, Reginato, and
van Bavel, 1965). Paetzold (1979) indicated that a computer program is avail-
able to calculate K(0) from H versus 0 curves.
Saturated Flow in the Vadose Zone
For saturated flow, Darcy's equation is written in the form:
v = -KVH (14)
where v = specific discharge
K = saturated hydraulic conductivity
VH = hydraulic gradient.
To account for flow only within the pore space, Darcy's equation is also
written as:
52
-------
V = -|-VH (15)
H
where V = average linear velocity
St = porosity.
As with unsaturated flow, the average linear velocity is a macroscopic
quantity of lesser magnitude than the actual microscopic velocity in the pore
sequences.
In the context of this report, saturated flow will be considered only in
perched groundwater regions of the vadose zone. The complex flow patterns
within groundwater mounds immediately above a water table will not be
considered.
It is commonly assumed that perched groundwater is always created by a
relatively impermeable layer which impedes vertical water movement. Although
this concept is true, perched groundwater may also develop at the interface
between two layered regions which individually may be quite permeable.
Figure 18, reproduced from Bear, Zaslavsky, and Irmay (1968), illustrates
the requisite condition for perched groundwater formation when a region of
higher permeability overlies a region of lesser permeability in the vadose
zone. For the first case, the flow rate, J, is less than l<2, the hydraulic
conductivity in the lower strata. Flow is unsaturated and water could not be
extracted from the system except by means of a suction cup. In the second
case, J equals l<2 and the soil water pressure is atmospheric. Again, it may
not be possible to obtain a water sample except by means of a suction cup. In
the third case, J is greater than l<2 and a region of positive water pressure
is present at the interface. Water samples could be obtained from a well or
open cavity.
For saturated media, the hydraulic conductivity in Darcy's equation is
not generally expressed as a function of the water content. However, it is
necessary to account for possible variations in the hydraulic conductivity de-
pending on the direction of flow. For example, in layered alluvial deposits,
the hydraulic conductivity in the horizontal direction is generally greater
than in the vertical direction. The condition of having variable K values de-
pending on direction is called anisotropy. In anisotropic media, the hydrau-
lic conductivity is a second rank tensor (Bear, 1972).
Inasmuch as the hydraulic conductivity, K, of a soil is a measure of the
ease of water movement, it follows that K depends greatly on soil texture and
structure. Thus, a sandy soil will have a greater hydraulic conductivity than
a clay soil, even though the latter has a greater porosity. Similarly, a
well-structured soil will permit rapid flow through the cracks. Such a soil
will have a greater hydraulic conductivity than will a poorly-structured soil,
even though the soil type may be the same. Ranges of hydraulic conductivity
values for representative water-bearing materials in California are shown in
Table 3.
53
-------
MORE PERMEABLE
SOIL
K|
LESS PERMEABLE
SOIL
K,
K,
ELEVATION
v\
\
\
45
P/ tf" PRESSURE
HEAD
Figure 18. Pressure head diagram in a transition from a more
permeable to a less permeable layer (after Bear,
Zaslavsky, and Irmay, 1968). For 1, the flow rate
J < l<2; for 2, J = l<2; and for 3, J > K2.
54
-------
TABLE 3. TYPICAL HYDRAULIC CONDUCTIVITY VALUES (after
Cooley, Harsh, and Lewis, 1972).
Hydraulic conductivity
Material (gpd per square foot at 60°F)
Granite
Slate
Dolomite
Hematite
Limestone
Gneiss
Basalt
Tuff
Sandstone
Till
Loess
Beach sand
Dune sand
Clay
Silt
Very fine sand
Fine sand
Medium sand
Coarse sand
Very coarse sand
Very fine gravel
Fine gravel
Medium gravel
Coarse gravel
Very coarse gravel
Cobbles
0.0000009
0.000001
0.00009
0.000002
0.00001
0.0005
0.00004
0.0003
0.003
0.003
1
100
200
0.001
1
10
100
1,000
4,500
6,500
8,000
11,000
16,000
22,000
30,000
Over
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
40,
0.000005
0.000003
0.0002
0.009
0.002
0.05
1
10
30
0.5
30
400
600
1
10
100
1,000
4,500
6,500
8,000
11,000
16,000
22,000
30,000
40,000
000
Although the saturated hydraulic conductivity of a soil is not dependent
on the water content, K is not necessarily a constant. For example, a de-
crease in temperature may dissolve air bubbles entrapped in the soil pores,
causing an increase in the hydraulic conductivity. The hydraulic conductivity
of soils containing clays will also be affected by the cationic constituents
and solute concentrations of applied water. Of particular concern with re-
spect to the cationic composition of the applied water is the relationship of
monovalent sodium to divalent calcium and magnesium. This relationship is
commonly represented by the sodium adsorption ratio:
SAR =
(16)
Ca
++
Mg
++
where Na+, Ca++, and Mg++ are expressed in milliequivalents per liter.
55
-------
SAR values greater than 6 and 9 in an irrigation source are expected to
reduce the hydraulic conductivity of shrinking-swelling soils (Ayers and West-
cot, 1976), thus decreasing the long-term infiltration rate. However, the
ratio of carbonate and bicarbonate levels in irrigation water must be consid-
ered. As stated by Ayers and Westcot: "When drying of the soil occurs be-
tween irrigations, a part of the 003 and HCC^ precipitates as Ca-MgCC^,
thus removing Ca and Mg from the soil water and increasing the relative pro-
portion of Na which would increase the sodium hazard." The older concept of
SAR has been adjusted to accommodate the effect of lime dissolution or deposi-
tion by means of the adjusted SAR:
adj SAR = SAR (l + (8.4 - pHc)) (17)
where Na, Ca, and Mg are obtained from water analyses. The pHc is the theo-
retical pH of water equilibrated with CaC03- In practice, pHc is calculated
using equations given by Ayers and Westcot.
The relationship of exchangeable sodium percentage (ESP), the fraction of
the exchange complex occupied by sodium, to the total salt concentration of
the soil solution also affects the hydraulic conductivity of soils with clays.
According to McNeal (1974), a combination of high ESP and low total salt con-
centration leads to swelling and dispersion of the soil minerals. Swelling
reduces the hydraulic conductivity of the soil. Dispersion may also cause a
migration of clay particles into lower regions of the soil profile, possibly
forming a layer which restricts free water movement.
Quantitative interrelationships between ESP, total salt content of the
soil solution, and hydraulic conductivity were demonstrated on laboratory soil
columns by Quirk and Schofield (1955). In particular, they observed that for
a given ESP, the hydraulic conductivity, K, increased with an increase in to-
tal salt content. The relationships between K, ESP, and salt content for
Pachappa sandy loam (McNeal, 1974) are shown on Figure 19. This figure also
illustrates that a greater and greater total salt content is required to off-
set the effect of sodium on K as ESP increases.
Quirk and Schofield (1955) coined the expression "threshold salt concen-
tration" to represent the level at which the salt content would need to be re-
duced to produce a 10- to 15-percent reduction in K at a given value of ESP.
As pointed out by McNeal (1974), these concepts have applicability in
land disposal operations as well as in irrigated agriculture, particularly if
a sizable fraction of the effluent contains softened water.
Laboratory devices for measuring the hydraulic conductivity of disturbed
or undisturbed soils are generally classified as permeameters. Klute (1965)
describes the construction and operation of common permeameters. Soil cores
taken in the field may be used directly in the permeameter. For research pur-
poses, columns are filled with dried and sieved soils packed to the desired
bulk density. Klute (1965) recommends that water to be applied to the field
soil should be used in conducting the tests.
56
-------
NO DETAIL SHOWN WITHIN
,5% OF LINE ESP=0
SALT CONCENTRATION (MEQ/LITER)
Figure 19. Relationships among hydraulic conductivity, salt concentration,
and ESP for Pachappa sandy loam (after McNeal, 1974).
constant head
are arranged so
Two general types of permeameters that are used are the
type and the falling head type. Generally, the permeameters
that flow is vertically downward through the soils. For the constant head
type of permeameter, water is ponded continuously at a constant distance above
the surface of the soil core. The rate of water movement through the core is
measured via a graduated cylinder and stopwatch. The hydraulic conductivity
is calculated from Darcy's equation. In the falling head method, the soil
core is connected to a standpipe. After saturating the soil, the water level
in the standpipe is allowed to fall. The time for the head to drop through
successive distances is measured and Darcy's equation is used to calculate K.
As Klute (1965) pointed out, the results of the permeameter methods are
of questionable value when applied to a large area. Thus, a large number of
samples would be required to account for the spatial variability in field hy-
draulic conductivity of soils. When samples are also obtained from deeper
regions of the vadose zone for permeameter tests, the problem of spatial
variability assumes an even more three-dimensional character.
Relationships have been developed between hydraulic conductivity and
characteristic grain sizes of soils and aquifers. These relationships require
conducting grain-size analyses of representative samples and employing appro-
priate equations. Details are presented by Freeze and Cherry (1979, pp 350-
352). In general, this approach should be used with caution because of the
difficulty of obtaining representative samples.
57
-------
FIELD METHODS FOR DETERMINING THE RATE OF WATER
MOVEMENT IN THE VADOSE ZONE
Infiltration
Recent reports by the U.S. EPA et al. (1977) and Sidle (1979) present ex-
cellent reviews of methods for estimating long-term infiltration rates. For
disposal operations involving the surface spreading or ponding of wastes, the
two most suitable procedures for determining long-term infiltration are cylin-
der infiltrometers and test basins. Infiltrometers may be either the single-
ring or double-ring type. In the double-ring infiltrometer method, a metal
cylinder from 6 to 14 inches in diameter is driven to a depth of 6 inches into
the soil at the test site. A larger ring, ranging from 16 to 30 inches in
diameter, is placed concentrically around the smaller ring. The areas within
inner and outer rings are flooded and the rate of recession of water level in
the inner ring is measured via a hook or point gage.
By keeping the outer region flooded, flow in the inner region is re-
stricted mainly to the vertical direction. Wastewater to be used during the
full-scale disposal operation should be used during infiltrometer tests. Be-
cause of the spatial variability of soil properties, an inordinate number of
ring tests may be required to ensure that results are within a certain per-
centage of "true" mean values. Even with care, however, experience has shown
that double-ring infiltrometers tend to overestimate the true infiltration
rate because of the divergence of flow due to shallow impeding layers and un-
saturated flow (Bouwer, 1978). Bouwer recommended using a single large cyl-
inder, in lieu of a double-ring inf iltrometer, to offset the problem of
divergence from unsaturated flow.
A superior alternative to inf iltrometers is to install a number of small
test basins at the site. Test basins should be large enough to permit instal-
lation of the desired accessory facilities (see Figure 22). For studies in-
volving an irrigated field, Nielsen, Biggar, and Erh (1973) used basins that
were about 400 square feet. Basins may be constructed by digging a narrow
trench around the desired rectangular area to such a depth that lateral flow
will be minimized. Side boards, possibly covered with plastic sheeting, are
placed in the trench, ensuring that good contact is made with the soil within
the plot and also at the corners. Sufficient free-board should be allowed to
permit a range of flooding heads. During tests, water is metered into the
plot at a discharge rate such that a constant head is maintained. The dis-
charge rate to maintain this head is the application rate. The infiltration
rates at selected times are calculated by dividing the application rate by the
surface area.
Alternatively, the head may be allowed to fall a short distance by shut-
ting down the inflow, and the volume which infiltrates per unit area over a
measured period of time may be converted to infiltration rates.
As with infiltrometers, the same water source to be used during the
actual operation should be applied during the tests. The effect of such
effluent on surface intake rates (e.g., by clogging) may be readily ex-
amined. Similarly, the effect of unfavorable exchangeable sodium
58
-------
percentages and electrolyte concentrations on the hydraulic conductivity can
be quantified.
Because of the spatial variability of soil properties, even on supposedly
homogeneous areas, it is recommended that a number of test basins be located
on the proposed waste disposal site. For example, by arranging the basins in
accordance with recommended statistical procedures (e.g., Steele and Torrie,
I960), results amenable to statistical analyses will be generated for infil-
tration rates and other properties of interest.
A common problem with the test basins and infnitrometers is that shallow
impeding layers may promote lateral movement of water in preference to truly
vertical flow. Lateral flow rates generally exceed vertical rates. Hence,
measured intake rates tend to overestimate the intake rates of a larger dis-
posal area.
If the technique for applying wastewater to the disposal site is sprink-
ler irrigation, it is recommended that infiltration tests be conducted using
rainfall simulators. Such units are described by Bertrand (1965), Sidle
(1979), and Hamon (1979).
In conducting infiltration trials, a record should be kept of the intake
rate as a function of time. Plotting of the results will produce an exponen-
tial relationship between intake and time. The resultant curve will asymp-
totically approach a final value, specifically that of the saturated hydraulic
conductivity.
Unsaturated Flow
The principal unsaturated flow parameters which require quantification at
a waste disposal site following infiltration include the following: (1) the
amount of water moving into the lower vadose zone, (2) the direction of un-
saturated water movement, (3) hydraulic gradients, (4) the unsaturated hy-
draulic conductivity, and (5) flow rates (flux).
Amount of Water Movement Below the Soil Zone—
By means of an appropriate soil moisture accounting system, it may be
possible to estimate the amount of water draining from the soil zone into the
lower vadose zone. Such an accounting system or water balance is based on the
hydrologic equation for the soil-water system. This equation is often written
in the form:
ASM = PPT - PET - RO ' (18)
where ASM = the change in soil moisture storage
PPT = precipitation + irrigation
PET = potential evapotranspiration
RO = runoff.
59
-------
A commonly used water balance method is that developed by Thornthwaite
and Mather (1957). This method, which is essentially a bookkeeping procedure,
accounts for soil moisture additions and extractions on an annual basis. Ba-
sic credit components are the quantities (expressed as depths) of rainfall and
irrigation. The principal losses (debits) from the system are evapotranspira-
tion, runoff, and deep percolation. Monthly temperature and precipitation
data are required. The potential water storage capacity of the soil must be
estimated for the soil depth of interest. Generally, the potential storage
capacity can be taken as the difference in water content (on a volumetric ba-
sis) of the soil at field capacity and that at a selected drained value.
The analyses of Thornthwaite and Mather (1957) generate values of changes
in soil moisture storage. For example, if potential evapotranspiration ex-
ceeds rainfall or other water additions to the system, the depth of water in
storage is decreased. In contrast, when additions exceed the potential water
holding capacity, the excess either runs off or percolates. By measuring
runoff, the amount of percolation is determined by differences. The most dif-
ficult component of the analysis to quantify is the consumptive use, or evapo-
transpiration. A recent report to the American Society of Civil Engineers
(Jensen, 1973) discussed in detail the present state of the art for measuring
evapotranspiration losses.
Fenn, Hanley, and DeGeare (1975) and Mather and Rodriguez (1978) used
variations of the Thornthwaite-Mather method to estimate the potential of in-
filtrating rainwater to move into a landfill, creating leachate. Assumptions
were made on the thickness and storage capacities of soil cover and solid
waste cells, on packing of solid waste, and on transmission characteristics of
water in solid waste. According to Mather and Rodriguez (1978), the advan-
tages of the water budget approach are that it is easy to apply, it provides
quantitative values, and it is based on straightforward assumptions. The
method should find applicability to landfills in humid regions where precipi-
tation amounts are great enough to produce surplus water in the soil layers.
For landfills in arid and semi arid regions protected from sheetflow, surplus
water may not be available for deep percolation into solid waste cells and the
method may find limited utility.
Direction of Water Movement and Hydraulic Gradients--
Methods for monitoring water movement and hydraulic gradients in satu-
rated regions such as perched groundwater cannot be used to monitor unsatur-
ated flow because of the so-called outflow principle: water will not freely
enter a soil cavity unless the soil-water pressure is greater than atmo-
spheric. Consequently, special methods must be used in unsaturated soils.
Of the available methods, the three most commonly used to infer unsaturated
flow are tensiometers, psychrometers, and neutron moisture logging.
Tensiometers--The principles of operation, methods of installation, and
limitations of tensiometers were discussed in the previous section. In char-
acterizing the direction of unsaturated water movement, a battery of tensiom-
eters must be installed with individual cups terminating at successive depths
throughout the region of interest. Using the principles of hydraulics, the
total hydraulic head, H, at a point in a flow system is the sum of the
60
-------
positional (gravitational) head, z, plus the pressure head, h. The positional
head is the distance between the measuring point (e.g., tensiometer cup) and
an arbitrary datum. That is, positional head equals zero at the datum. The
pressure heads are readings taken from the tensiometer manometer system as
shown on Figure 12. Positional heads are positive and pressure heads are neg-
ative during unsaturated flow. Differences in the total head between tensiom-
eter cups may be used to estimate the direction of flow. In particular, flow
occurs in the direction of decreasing total head.
Figure 20 illustrates three hypothetical flow cases in a soil profile
overlying a water table. This example is an adaptation of one presented by
Reeve (1965). In the example, the datum is taken at the base of the aquifer.
Three tensiometers are located in the soil zone above the water table. The
cups are connected to water manometers for illustrative purposes only. In a
real case, the tensiometers would be connected to above-ground mercury manom-
eters as shown on Figure 12.
In the first case, water is not moving upward or downward in the profile,
as evident by calculating the total hydraulic heads. Note that the water lev-
els in the open ends of the three manometers are equal and correspond to the
water table. For unit #1, the total hydraulic head, H, is the sum of the po-
sitional head, z, and the pressure head, -hi. That is, H} = Z]_ - h]^.
But this is simply the distance from the datum to the water table. For unit
#2, H2 = Z2 - h2. Again, this equals the distance from the datum to the
water table. It can be shown that the total hydraulic heat at unit #3 also
equals the distance from the datum to the water table. In order for flow to
occur in the soil between units, a difference in total head must exist. But
H2 - HI = 0 and H3 - H2 = 0. In other words, the system is in equi-
librium. Incidentally, for such a system, measurements in tensiometers afford
a means of defining the water table position.
In the second case, the elevation of the water table is different and
water is flowing vertically downward in the profile. In this example, the
pressure heads are all zero because the level of the meniscus is at the same
level as the cups. For all cups, the positional heads remain equal to the
distance from the datum to the cups. In this case, the total hydraulic head
at each cup is just the value for the positional heads, z]_, zg, and 23.
Inasmuch as the positional head (and thus total head in this case) decreases
downward, flow also occurs in a downward direction.
Finally, in the third case, evaporation is occurring at the soil surface.
By applying the same reasoning as above, it could be shown that the total hy-
draulic head decreases vertically upward in the profile so that upward move-
ment occurs.
It is necessary to install more than one array of tensiometers to detect
horizontal flow. Thus, if individual arrays terminate at varying depths, dif-
ferences in total hydraulic heads between corresponding units in successively
lateral arrays may suggest that horizontal movement is occurring in the un-
saturated state. Because of the heterogeneity in soil properties, however,
definitive conclusions on lateral flow may be tenuous.
61
-------
SOIL
SURFACE
PONDED WATER
o>
ro
WATER
TABLE
DATUM
2 = 0
COVERED
!
V
s
V
s
s
s
,
h3
sd
^ *
s
N
N
S
S
S
s
y
v
\
S
s
V
s
V
\
s
N
s
s
s
s
s
s
s,
V
s
§>
«3
23
1
f
1
^
^
^
A
5
#2
TENSIOMETER
/BULB
h
_j
- ^*
2
^
V
V
s
V
s
V
V
s
s
s
s
V
s
s
s
s,
\
\
^
\
c
p
Z
2
V
s
s
s
s
\
^
*-l
f A
*i
f
s
s
s
s
s
s
; z
s
s
s
s
s
s
s
s
s
s
s
I
1
\
A
s
*
S
N
S
V
\
^
>(
o
s
s
>
o
s
s
s
s
s
s
&
s
s
^
fc
1
s
o
o
o
^
o
V
o
N
o
\
w
Z3
1
1
^
S
s
s
s
s
s
s
\
N
1
/
3SS
ff '
z
2
H
S
1
J^^
4&(
S
^f
y
z
1
r
EVAPOTRANSPIRATION
t t t t t t t t
"2$
h|.
BEDROCK
CASE n \
STATIC EQUILIBRIUM
GRAVEL AQUIFER
CASE * 2
WATER MOVING VERTICALLY
DOWNWARD
CONFINED AQUIFER
CASE * 3
WATER MOVING VERTICALLY
UPWARD
Figure 20. Distribution of hydraulic heads for three unsaturated flow cases (after Reeve, 1965).
-------
An array of tensiometers may be useful in detecting the presence of clog-
ging layers and in manifesting the head loss by such layers (Bouwer, 1978).
For example, in water- spreading operations, it may happen that because of dis-
persion, fine solids may migrate in the profile and filter out at the inter-
face between soil materials of differing texture. A tensiometer above this
layer will gradually show an increase in soil-water pressure, while a unit be-
low the layer will shift to the more negative side.
Time lag in response to a rapidly changing system is also a problem with
tensiometers. The tensiometer design of Watson (1967) involving a pressure
transducer system for measuring soil-water pressure has more rapid response
characteristics than mercury-water manometer systems.
Psychrometers--To measure soil-water pressure under very dry conditions,
the tensiometer unit is no longer applicable because of air-entry problems.
To measure lower negative pressures, therefore, other instrumentation is re-
quired. In recent years, progress has been made in developing the thermocou-
ple psychrometer for this purpose. In-situ pressure measurements down to -100
atmospheres may be possible with these units (Warrick, personal communication,
1980).
The principle of psychrometric measurement of soil -water potential is
given by Rawlins and Dalton (1967) who also diagrammed a representative unit
(see Figure 21). The basic idea in soil -water psychrometry is that a rela-
tionship exists between soil-water potential and the relative humidity of soil
water. This relationship is given by:
where ¥ = soil water potential (matric potential plus osmotic potential)
R = ideal gas constant
T = absolute temperature
n = volume of a mole of water
P/P0 = relative humidity.
In construction, psychrometers consist of a porous bulb comprising a
chamber to sample relative humidity of a soil, a sensitive thermocouple, heat
sink, reference electrode, and associated electronic circuitry. Such units
use the principle of Peltier cooling to lower the temperature of one junction
of the thermocouple below the dewpoint, thereby allowing an evaluation to be
made of the relative humidity.
Earlier units were plagued with problems relating to temperature effects
on measurement of potential. By designing the instruments to meet certain
boundary conditions, however, these problems were overcome (Merrill and Raw-
lins, 1972).
63
-------
ACRYLIC TUBING
EPOXY RESIN
COPPER LEAD WIRES
COPPER HEAT SINKS
TEFLON INSERT-
THERMOCOUPLE WIRE
CERAMIC BULB
ACRYLIC TUBING
COPPER LEAD WIRE
COPPER HEAT SINKS
TEFLON INSERT
THERMOCOUPLE WIRE
CERAMIC BULB
Figure 21. Soil psychrometer (after Raw!ins and Dal ton, 1967).
Operation of the Pel tier-effect psychrometer consists of (Merrill and
Rawlins, 1972):
(1) measuring the electromotive force (e.m.f.) while the mea-
suring junction is dry (dry e.m.f.); (2) passing a current
through the sensor, cooling the measuring junction below the
dewpoint, and thereby wetting it; (3) measuring the e.m.f.
again after an initial, transient phase of the e.m.f. has sub-
sided, but before the e.m.f. has decayed significantly as a re-
sult of depletion of water on the junction (wet e.m.f.); and
(4) again passing current through the thermocouple wires in the
reverse direction to head and dry the measuring junction. This
mode of operation permits any temperature difference between
the measuring junction and the reference junctions to be com-
pensated for by subtraction of the dry e.m.f. from the wet
e.m.f. (A e.m.f.).
Two types of thermocouple psychrometers are available. One unit, in-
stalled in access tubes, consists of positioning psychrometers in porous cups
at the base of tubing. This unit may be withdrawn for recalibration. The
second type of unit, called "sealed-cup" psychrometers by Merrill and Rawlins
(1972), contains the thermocouple unit permanently sealed into a porous
enclosure.
Each unit must be calibrated prior to field installation. Techniques for
such calibration are given by Merrill and Rawlins. Also in field installa-
tions, it is important to account and correct for diurnal changes in tempera-
ture. Installation of a sensitive thermistor is recommended. Merrill and
Rawlins point out that field psychrometers could be tied into a remote sensing
scheme, provided care is exercised in instrumentation.
64
-------
Enfield, Hsieh, and Warrick (1973) used thermocouple psychrometers to es-
timate the direction and rate of water flux in material in a vadose zone with
a water table at 310 feet in a desert environment in Washington. Individual
psychrometric units were placed 3 meters apart within a specially designed
well casing. In addition to the thermocouple psychrometers for measuring the
matric potential, temperature transducers were also installed, providing data
to relate moisture flux due to temperature gradients.
Neutron moisture logging—Although the neutron thermalization technique
is used mainly to determine changes in volumetric water content of materials
in the vadose zone, the method may also be used to infer water movement. In
particular, if a soil-water characteristic curve is available for incremental
depths throughout the vadose zone, it may be possible to relate water content
values to water pressure. Hydraulic head gradients, and therefore flow direc-
tion, could then be inferred. Unfortunately, the accuracy of the method may
not be great enough to detect slight water content changes, particularly in
the dry range.
Wilson and DeCook (1968) and Wilson (1971) used moisture logs from a net-
work of 100-foot deep access wells to infer the rates of lateral movement of
recharge waves in the vadose zone during river recharge and during artificial
recharge. The arrival of such waves was inferred by the change in water con-
tent in a perched mound at about 33 feet.
Flux, Velocity, and Unsaturated Hydraulic Conductivity--
As discussed earlier, one of the prime goals of vadose zone monitoring
is to estimate the flux and velocity of water in this zone. Some of the meth-
ods for estimating flux and velocity are based directly on the equation of
continuity and information on hydraulic conductivity values is not required.
However, other methods are based on Darcy's equation and values of both hy-
draulic conductivity and hydraulic gradients are necessary. In this report,
both classes of methods are referred to as "draining-profile" techniques be-
cause measurements are made during drainage cycles. Other methods reviewed
are: (1) flow meters for direct measurement of flux, (2) use of suction cups
for estimating velocity, (3) tracer techniques, and (4) an approximate method
using water budget data.
It should be emphasized that, by and large, the methods presented may be
suitable for characterizing flux and hydraulic conductivity in soils and may
have limited applicability to deeper regions of the vadose zone. However, it
is important in many waste disposal operations to determine loss in the near-
surface region. For example, during site evaluation of hazardous waste dis-
posal ponds, a series of tests could be conducted to determine a site with
minimal K and J values.
Draining-prof ile methods—The simplest approach for estimating flux at
successive depths in the vadose zone is to obtain a series of water content
profiles during drainage. The method is essentially based on the equation of
continuity, expressed in terms of flux:
65
-------
H • T? , (2°)
or
J = -/If dz (21)
o
To account for all sources and sinks, Bouwer and Jackson (1974) suggested
using the following expression for flux:
Jx = R + I - ET - JZ (||) dz (22)
where J^ = combined flux
R = rainfall
I = irrigation
ET = evapotranspiration.
For a draining profile covered to prevent ET, equation 22 reduces to
equation 21.
An example of hypothetical water content profiles from a field site is
shown on Figure 22 (after Bouwer and Jackson, 1974). Calculation of flux out
of a given depth, z, between successive times, t^ and t2, requires the
solution of equation 21. Inasmuch as the right-hand side of the equation is
simply the shaded area of the sketch divided by the time difference t2 -
t]_, the flux may be found by graphical integration of water content profile
between successive measurement dates.
Water content profiles could be obtained using techniques discussed in
another section, e.g., neutron moisture logging or tensiometers. An advantage
of neutron moisture logging is that profiles can be obtained at considerable
depth, provided access wells are installed carefully to prevent side leakage.
The sequence to obtain water content profiles using tensiometers is as
follows:
1. Install tensiometers at sequential depth intervals through the soil
region of interest.
2. Obtain soil cores from the soil at depths bracketing the tensiometer
cups.
3. Obtain water characteristic curves for each core in order to define
the relationship between water content and soil-water pressures. Inasmuch as
66
-------
0
e-
Figure 22. Hypothetical water content profiles at several
times during drainage of a soil profile (after
Bouwer and Jackson, 1974).
field data will be obtained during drainage, curves must be obtained for dry-
ing cycles.
4. Determine pressure values in tensiometers during drainage of the soil
profile. See Figure 23(a).
5. Convert pressure readings to water content values using the appro-
priate water characteristic curves. See Figure 23(b).
6. Plot water content versus soil depth. See Figure 23(c).
An alternative draining-profile method requires two stages: (1) deter-
mining the hydraulic properties of soil in a number of plots within a field
or disposal area, and (2) using the resultant hydraulic properties and appro-
priate instrumentation to estimate flux at additional field locations. Basi-
cally, the method is an adaptation of a two-step technique presented by LaRue,
Nielsen, and Hagan (1968) for estimating soil-water flux below the root zone
of an irrigated field.
During step 1, a number of test plots are established at random locations
in the area of interest. Each plot is instrumented with a depth-wise sequence
of tensiometers (installed in triplicate at each depth) and possibly an access
tube to permit moisture logging. These units should be installed in the cen-
ter of the plot to minimize the effects of lateral flow. The periphery of
each plot is provided with side walls to facilitate applying a head of water
at the soil surface. Figure 24 shows a possible arrangement for a test plot.
Soil cores obtained during installation of tensiometers are used to construct
water characteristic curves for distinct depth intervals in the profile. The
following procedure is used:
67
-------
T
4
z3+ y- I2.55x
1504 10- 12.55(16.7)
-50 cm
(a)
z I -
23...
0.3-,
0.2-
0.1-
SOIL WATER CHARACTERISTIC
FOR DEPTH INTERVAL z=z,
-100
-200
h
(b)
-300
0.3
WATER CONTENT
PROFILE FOR
TIME t = t
Figure 23. Schematic representation of procedure for converting soil-water
pressure values from tensiometer data to water content values on
on a water content profile for a given time.
68
-------
\
SOIL SURFACE
DATUM
2=0
ACCESS WELL FOR
'NEUTRON MOISTURE
LOGGING
SOIL SURFACE
COVERED WITH
PLASTIC AND
EARTH
Figure 24. Sketch of field plot for determining K and J
showing tensiometers and access well.
69
-------
1. Flood each plot until the reading on the lowermost tensiometer indi-
cates that a maximum degree of saturation has been reached in the surrounding
soil.
2. Following disappearance of free water, cover the soil surface with
plastic and earth to prevent evaporation.
3. Allow the soil to drain, taking regular tensiometer readings and log-
ging the access well.
4. From water content profile, estimate flux out of various depth inter-
vals using the graphical approach discussed above.
5. Calculate hydraulic gradients for the selected depth intervals for
each sequence of tensiometer readings averaged over the time interval.
6. Use the hydraulic gradients and flux values to calculate hydraulic
conductivity from Darcy's equation,
J = -K(6)VH.
7. Plot the K(9) values versus average water content values (9) for
each depth interval. Note that only one K(6) value is obtained for each
time interval.
8. Repeat steps 4 through 7 for additional time steps and obtain K ver-
sus 6 curves for each depth interval.
An example (courtesy, A.W. Warrick) will clarify the above procedure for
the first step of the method of LaRue, Nielsen, and Hagan. Assume that a test
plot has been established with tensiometers installed for measuring water
pressures at discrete depths to 200 cm. An access well has been installed for
obtaining water content profiles by a neutron moisture meter. The soil pro-
file is wetted to 200 cm, following which the surface is covered and water
content values and soil-water pressures are measured. Resultant water content
profiles for 24 hours and 40 hours after the beginning of drainage are shown
on Figure 25.
The area between the curves (shown by cross-hatching) was found by pla-
nimetering to be 2 cm (0.79 inch) of water. Pressure head values measured on
tensiometers at the 120-cm (47.2-inch) and 150-cm (59.1-inch) depths after 24
hours and 40 hours of drainage were as follows:
h1(120 cm; 47.2 in) h2(150 cm; 59.1 in)
t = 24 hours -36 cm (-14.2 in) -29 cm (-11.4 in)
t = 40 hours -40 cm (-15.7 in) -32 cm (-12.6 in).
70
-------
e
O.I
0.2 0.3
50
"00
150
135 cm
Figure 25. Water content profiles in hypothetical
soil column at times t = 24 hours and
t = 40 hours.
The problem is to find a K versus 9 relationship for the 120-cm (47.2-
inch) to 150-cm (59.1-inch) depth interval. Note that only one value can be
found for one time interval; for additional values, profiles for other time
intervals would need to be given. The procedure is as follows:
1. Find the drainage for the time interval t = 24 hours to t = 40 hours.
Assume that we are precisely interested in the 135-cm (53.1-inch) depth
(half-way between 120 cm (47.2 inches) and 150 cm (59.1 inches)). Inas-
much as the area on Figure 25 corresponds to 2 cm (0.79 inch) and the time in-
terval is 40 - 24 = 16 hours, then the Darcian flux is:
J = 0.125 cm/hr.
2. Next find the hydraulic gradient. If the reference is taken at the
level of the lowermost tensiometer and the positive direction is upward, the
hydraulic heads at 120 cm (47.2 inches) and 150 cm (59.1 inches) at 24 hours
are:
= (0 - 29) = -29 cm (-11.4 inches)
Hl20 = (30 - 36) = -6 cm (-2.4 inches).
The gradient at 24 hours is:
VH = (-29 - (-6))/30 = -0.76 (24 hours).
71
-------
Similarly, the hydraulic heads at 120 cm (42.2 inches) and 150 cm (59.1
inches) at 40 hours are:
Hl5Q = (0 - 32) = -32 cm (12.6 inches)
HIZO = (30 - 40) = -10 cm (3.9 inches).
The hydraulic gradient at 40 hours is:
VH = (-32 - (-10))/30 = -0.73 (40 hours).
The above values represent averages over the 120- to 150-cm (47.2- to 59.1-
inch) depth interval or, alternatively, at the 135-cm (53.1-inch) depth.
3. Calculate the hydraulic conductivity from Darcy's law. Using J from
step 1 and the hydraulic gradient at 135 cm (53.1 inches) averaged over the
period 24 to 40 hours:
J = -K(6)VH
where H = -°-762- °-73 = -0.75,
0.125 = -K(6) (-0.75)
or K(0) = 0.167 cm/hr (0.007 in/hr).
This K(6) value corresponds to an average water content value (0.23
cm3/cm3) in the interval 120 cm (47.2 inches) to 150 cm (59.1 inches) dur-
ing the period 24 to 40 hours. The K(6) and 6 values thus provide one
point on a K(6) versus 6 curve for this depth interval. (Alternatively,
K corresponds to an average h value for the four input values; havg = (-36
- 40 - 29 - 32)/4 = -34.2 cm (13.5 inches). The corresponding 0 value coulr1
'on found from a water characteristic curve for the depth interval.)
When the hydraulic properties of the field have been estimated, the pro-
cedure to follow at additional field sites during step 2 of the LaRue, Niel-
sen, and Hagan methods is as follows:
1. Install tensiometers at depth intervals of interest, corresponding
to intervals used on the test plots. Alternatively, to minimize expense, two
tensiometers could be used to bracket the lowermost interval.
2. Measure water pressures during drainage cycles using tensiometers.
Calculate hydraulic gradients.
3. From the water characteristic curve corresponding to the depth inter-
val of interest, convert water pressure value to an average 6 value.
4. Using the K versus 6 curve for the interval, convert 0 values to
K values.
72
-------
5. Use the K and hydraulic gradient values with Darcy's equation to cal-
culate flux.
Nielsen, Biggar, and Erh (1973) evaluated a number of simplified tech-
niques for measuring flux and hydraulic conductivity values in the field. One
of the promising methods is based on the following relationship:
JL = KQ (1 + a l^tL'1) (23)
where K0 = steady-state hydraulic conductivity
a = an empirical factor relating K and 0
J|_ = flux at depth L
t = time of observation.
According to Warrick and Amoozegar-Fard (1980), this equation is based on
the following assumptions:
1. The soil is initially at a uniform water content 00
2. The unsaturated conductivity is given by the equation:
K = K0 exp (a(0 - 00)) (24)
3. A unit hydraulic gradient exists at all times
4. Evaporation and transpiration are negligible.
With the hydraulic gradient equal to unity, the equation of continuity,
expressed by equation 13, may be written in the following form:
(25)
or
L{£-K
-------
9 produces a straight line, equation 24 is assumed to be valid and we may
proceed to use equation 23 to calculate J[_. Incidentally, the graph is also
used to determine the factor a, which is simply the slope of the straight
line.
Nielsen and Biggar (1973) presented a plot of In K(6) versus 6 for
test plot data for Yolo loam, an unusually uniform soil. The scatter of
points was minimal. For a more heterogeneous soil profile, values of K(0)
ranged over four orders of magnitude for a given water content. However, the
flux of water at a given depth was still adequately described using an average
straight line from the In K(9) versus 9 plot. Problems arising from the
spatial variability in soil properties are described in a later section.
It should be noted that equation 26 provides us with yet another alter-
native method for estimating flux. That is, the left-hand side of the equa-
tion is an alternative form for expressing the flux. Thus:
In other words, if the hydraulic gradient is equal to unity in the flow
system, the flux at a depth equals the corresponding hydraulic conductivity
value.
Flux values determined from the draining-profile techniques are converted
to values of average linear velocity by dividing by the appropriate water con-
tent values. For example, if the calculated flux of a soil is 0.125 cm/hr and
the average water content is 0.23, the average linear velocity is:
v = J/6 (27)
= 0.125/0.23 cm/hr
= 0.54 cm/hr.
Direct Measurement of Flux
Attempts have been made in recent years to develop equipment to measure
soil -water flux in the unsaturated state which does not require information on
the hydraulic conductivity.
Two type of flow meters reported by Gary (1973) involve (1) direct flow
measurement, and (2) the displacement of a thermal field by water in motion.
The direct flow unit involves measuring the flow of soil water intercepted by
a porous tube containing a sensitive flow transducer. The second unit entails
measuring very accurately the transfer of a heat pulse in water moving with a
porous cup buried in the soil. Because of the intimate contact between the
soil and porous cup, the water moving in the cup forms a continuum with soil
water. Laboratory calibration curves are prepared to relate output of a sen-
sitive millivolt recorder during imposition of heat pulses to empirically mea-
sured flow rates.
74
-------
For installation into field soils, the porous discs containing either
flow transducers or heat sources are mounted in cylinders which are buried in
the soil. A limitation, therefore, is that flow is measured in disturbed
soils.
According to Dirksen (1974a), the accuracy of the above flow meters can
be improved by:
1. Minimizing convergence or divergence
2. Extending the range of water fluxes and soil types
3. Reducing or preferably eliminating very tedious calibration
procedures
4. Minimizing disturbance of soil during installation and its ef-
fect on original flow pattern.
Dirksen (1974b) discussed the design of a flux meter he developed. A
feature of his unit which circumvents some of the difficulties above is that,
as soil hydraulic conductivity changes, the resistance of the meter is ad-
justed so that the head loss across the meter matches the head loss in the
soil as measured by nearby tensiometers. The fluxes through the meter and
soil then are equal. Dirksen also presented a method for installing his flux
meter into soils to effect a minimum of disturbance.
To date, no data are available on the use of such flow meters in the
lower vadose zone of deep profiles.
Indirect Estimates of Velocity Using Suction Cups
An indirect estimate of vertical velocity in the vadose zone beneath a
surface source entails observing the response of suction cups as the wetting
front moves into the profile. These devices are described in detail in a
later section of this report. Basically, the technique involves periodically
attempting to sample from an array of suction samplers within the vadose zone.
Generally, when the surrounding pore-water system is unsaturated, very little,
if any, sample will be obtained. However, when the wetting front reaches a
particular cup, samples are more readily obtainable. By observing the re-
sponse of depth-wise units, an apparent vertical velocity may be inferred.
Meyer (personal communication, 1978) used this method to follow the wet-
ting front during deep percolation of irrigation water in the San Joaquin Val-
ley. Signer (personal communication, 1979) used a similar approach during
recharge studies in Texas.
Tracers
A direct method for measuring the average linear velocity of water move-
ment in the vadose zone entails determining the depth-wise arrival times of
tracers introduced at the soil surface. A number of alternative tracers are
available including physical tracers such as temperature, ionic constituents
75
-------
such as chloride, organic tracers such as fluorescent dyes and fluorocarbons,
and radioactive tracers such as tritium. Evans, Sammis, and Warrick (1976)
evaluated the relative merits of two tracers, temperature and tritium, for de-
termining flux beneath an irrigated field. Results were compared with flux
values found by using Darcy's equation. In general, it was found that results
using temperature and tritium compared poorly with calculated values. They
pointed out that problems with tritium include a short half-life (12.26 years)
and possible reactions with clay particles.
Frissel et al. (1974) compared 36C1, 6°Co, and tritium for tracing
moisture movement in laboratory soil columns. From observed breakthrough
curves for the tracers, it was determined that for clay and sand soils, trit-
ium adsorption was negligible but that the use of 3^C1 and ^Co ions would
lead to errors because of anion exclusion.
A basic problem with fluorescent tracers is sorption on clays resulting
in errors in estimating flow velocities. Fluorocarbons may have promise, al-
though sorption occurs on organic constituents of a soil (Brown, personal com-
munication, 1979).
Simplified Method for Estimating Velocity
Using Mater Budget Data
For many disposal operations, it may not be possible to use the above re-
fined methods for estimating the velocity of water-borne pollutants in the
vadose zone. An approximate procedure is required to estimate whether pollu-
tants have reached the water table. The following is suggested as one such
approach. The method is based on a variation of equation 1, expressed as:
dvz - r <28>
where dvz = depth of penetration of water below a given soil depth
dw = depth of water applied
6 = volumetric water content at field capacity.
The method assumes vertical piston movement of water beneath the disposal
site with water content values equal to field capacity.
In practice, an estimate of seepage rate is determined using a water bud-
get method such as that of Thornthwaite and Mather (1957), or other simplified
methods. The rate is converted into an equivalent depth of water per unit
time. Average field capacity values for materials in the vadose zone are es-
timated from drillers' logs.
As an example, assume that the depth of water seeping beneath a disposal
facility for 1 year is 5 feet. If the volumetric water content corresponding
to field capacity is 15 percent, the depth of penetration in the vadose zone
is:
76
-------
_ 5.0
vz
= 33 feet.
Thus, the rate of penetration of 5 feet of seepage is 33 feet per year.
If the vadose zone is 100-feet thick, it would require about 3 years for wa-
ter to reach the water table.
Effect of Spatial Variability on Soil Properties
When using field data to estimate s
conductivity, it is important to keep in
eters may exist because of soil heteroge
in soil texture and structure, even with
values of related properties such as wat
conductivity.
The problem of such spatial variabi
sen and Biggar (1973), Coelho (1974), an
area, Nielsen, Biggar, and Erh (1973) fo
were normally distributed with depth and
whereas hydraulic conductivity values we
lowing conclusion is particularly notewo
Even seemingly uniform land are
hydraulic conductivity values.
density, and water content are
tion, methods for measuring wat
tivity, and hydraulic gradients
more accurate than required to
cause of the heterogeneity of t
make predictions over a large
from good to unsatisfactory, de
diction parameter of interest.
Based primarily on the results of s
(1973) and Guma'a (1978), the variablili
ranked as follows by Warrick and Amoozeg
Low variability: (Coefficient
20 percent)
Bulk density
Water content at a zero ten
Medium variability: (Coefficie
percent)
Textures (sand, silt, or cl
Field water content
Water content at specified
15 bars
77
ch properties as flux and hydraulic
mind that variations in these param-
eity. For example, large variations
a given mapping unit, will affect
r content, porosity, and hydraulic
ities was examined in detail by Niel-
Guma'a (1978). For their study
nd that variations in water content
horizontal distance in the field,
e log-normally distributed. The fol-
thy:
s manifest large variations in
Variations in texture, bulk
uch less. For a given loca-
r content, hydraulic conduc-
will yield values that are much
haracterize an entire field be-
e soil. Thus, our ability to
ea from a single plot can range
ending on the particular pre-
udies by Nielsen, Biggar, and Erh
ies in various soil parameters were
r-Fard (1980):
f variability less than
ion
t of variation 20 to 75
y)
ension between 0.1 and
-------
High variability: (Coefficient of variation greater than
100 percent)
Saturated hydraulic conductivity
Unsaturated hydraulic conductivity
Apparent diffusion coefficient
Pore water velocity
Problems arising from the spatial variability of hydraulic properties of
the vadose zone will increase as we proceed deeper in the profile. For exam-
ple, even small lenses may cause an effect on soil-water pressures, hydraulic
conductivity, and flux. Consequently, it may be that the best we can expect
is to estimate flux out of the soil zone of a waste disposal facility.
Saturated Flow in the Vadose Zone
Saturated flow occurs in the vadose zone primarily within perched ground-
water bodies which develop at the interface of regions of varying hydraulic
conductivity (see Figure 18). The principal saturated flow characteristics
which should be quantified for the vadose zone are the following: (1) hy-
draulic gradients, (2) direction of flow, (3) hydraulic conductivity values,
and (4) flow rates.
Hydraulic Gradients and Flow Direction in
Perched Groundwater—
Two basic tools for measuring hydraulic gradients in perched groundwater
are the piezometer and observation well. As will be shown later, these de-
vices are also useful for obtaining water samples for analyses.
Piezometers—Piezometers consist of small-diameter pipes which are
driven, augered, jetted, or otherwise drilled into a saturated zone or a zone
in which saturation is expected. Reeve (1965) discussed in detail common
techniques for installing and cleaning new piezometers. In general, it is
essential that a tight fit be maintained between the outer wall of the piezom-
eter and the surrounding media. For shallow units, piezometers may be in-
stalled by augering and driving with a sledge hammer. Deeper units will
require jetting or use of standard drilling equipment. It may be necessary to
fill the cavity between wells and boreholes with grout to ensure tightness of
fit. As with regular wells, piezometers should be developed by pumping or
bailing to clean and open up the material at the base of the unit. In some
cases, it may be necessary to install piezometers with screened well points to
prevent the upward movement of saturated sediment into the unit.
In many field situations, it is desirable to install a number of piezom-
eters at successive depths within a perched groundwater body. Two alternative
methods of construction are possible: (1) installing a battery of separate
piezometer wells, or (2) installing a cluster of wells within a common bore-
hole. Clustered wells are constructed by installing the individual units in
the borehole at desired depths and surrounding the well points with sand. The
region of the borehole between units is backfilled with bentonite. Fenn et
al. (1977) presented additional details on methods for installing piezometer
clusters together with their advantages and disadvantages.
78
-------
In operation, the piezometer provides a measure of the hydraulic head at
the terminus of the tube. The total heeid consists of the sum of gravity and
pressure heads. The gravity head is referenced to an arbitrary datum, say,
the ground surface. For a nest of piezometers, it is convenient to take the
elevation of the lowermost unit as the datum. The pressure head is the height
of water above the bottom opening of the
Depth to water in piezometer units
sounders, or air lines. By referencing
mapping of the piezometric surface will
provide clues on the direction of water
Although piezometers are generally
be installed into the lower vadose zone,
(1968) describe units which were placed
80-foot thick vadose zone of a recharge
of individual units were based on moisti
Each piezometer contained a screened we'
the outside of the casing and drill hol<
unit.
is measured by chalked tape, electric
piezometric values to a common datum,
be facilitated. Such mapping will
movement.
used in the soil zone, they may also
For example, Wilson and DeCook
into perched groundwater within the
site near Tucson, Arizona. The depths
re logs in a network of access wells.
1 point, and grout was poured between
to inhibit side leakage.
different depths may be used to monitor
An array of piezometers with the erds of individual units terminating at
the vertical movement of water in
perched water tables. As an example (adapted from Reeve, 1965), Figure 26 in-
dicates three hypothetical flow situations in a soil containing three piezom-
eter units beneath a water table. For purposes of calculation, the datum is
selected at the base of the lowermost piezometer. For the first case, that
of perched groundwater above an impermeable layer, vertical flow is not occur-
ring. This may be verified by calculation. The total hydraulic head for the
water levels in all wells is the same and corresponds with the water table.
It will be recalled that water flows in the direction of decreasing total hy-
draulic head. For unit #1, the gravitational head, zj_, is zero and the
pressure head is hi. Similarly, for unit #2, the positional head is Z2
and the pressure head is h2- But \\^ + 42 = nl- Consequently, HI =
\\2 and no flow occurs between units #1 and #2. Similar analysis would show
that flow does not occur between units 12 and #3.
For the second case, water is applied to the surface causing water move-
ment downward in the profiles. The perched water body is underlain by an un-
confined gravel aquifer. In this case,
direction. Water levels in the individual piezometers are at the base of the
unit; that is, pressure heads are all zero. For unit #1, the total hydraulic
head is zero because zi = 0 and hi = 0.
equals 22 and the pressure head = 0. The total hydraulic head at unit #2 is
Z2- Inasmuch as H2 > HI, flow occurs between units 12 and #1. Also,
because the vertical distance between tlie lower ends of units #1 and #2 is
Z2, the hydraulic gradient is unity, i.e.:
- 0
= 1
flow occurs in the vertically downward
79
-------
00
o
z,=o
1
h
1
U
3
•
-3
Ml
3
^
T
1
ti
1
U
i
2
VJI
2
t
V
V
s
V
V
\
s
V
s
s
i
T
h
,
ur
i
i
*IIT
2
i
ur
3
F
jr
3
a
r UN
T
Z2
1
111
2
•
r u
N
i
T
i
F
1
1
1
i
U
L
3
:3
Nl
3
I
T
•
t
Z
I
Uf
2
2
I
yir
2
!
I
r
j
i
i
u
i
N
1
^
1
1
$
^
T
BED ROCK
CASE I
STATIC EQUILIBRIUM
GRAVEL AQUIFER
CASE 2
WATER MOVING VERTICALLY
DOWNWARD
CONFINED AQUIFER
CASE 3
WATER MOVING VERTICALLY
UPWARD
Figure 26. Hydraulic head distribution in three piezometers during three hypothetical flow cases
(after Reeve, 1965).
-------
A similar analysis also shows that flow occurs between units #3 and #2 under a
unit hydraulic gradient.
The third case represents a water table region underlain by a confined
aquifer. At the same time, evaporation is occurring at the land surface. The
depth to water is greater in the shallower units. This is, the pressure head
increases with depth in the profile. An analysis using hydraulic principles
would show that the total hydraulic head decreases with distance above the
datum so that water is moving vertically upward in the system below the water
table. The last two cases demonstrate an important principle. Namely, when
vertical flow occurs, it may be difficult to locate the actual depth to a wa-
ter table in a moving system using only the data from piezometers.
The cases shown on Figure 26 do not include the effects of lateral flow
superimposed on vertical flow. It is known that many field soils exhibit an-
isotropy. That is, the hydraulic conductivity in the horizontal direction may
be many times greater than in the vertical direction. For example, Bouwer and
Jackson (1974) observed the ratio of horizontal to vertical hydraulic conduc-
tivity to be 7:1 for a finely stratified loamy sand. For such soils, it may
be advisable to install more than one array or battery of piezometers in a
lateral direction away from a flooding operation. The basal openings of cor-
responding units in sequential arrays should terminate at the same elevation.
Such an arrangement would permit observing lateral head differences resulting
from horizontal flow.
A basic problem with piezometers is that a lag may occur in the response
of the units to a rapidly changing system and when flow occurs in stratified
material. Bouwer and Jackson (1974) presented a time lag equation for
piezometers:
TL = irr2 YQ/AK Y0 = Trr2/AK (29)
where TL = time lag constant
r = radius of tube
Y0 = distance of sudden rise of water level
A = geometry factor
K = hydraulic conductivity.
The value of A is determined from suitable equations, graphs, or tables.
It may be shown that the basic time lag manifests the time for which 63 per-
cent of the sudden change in actual water level position has been registered
by the water level in the measuring device (Bouwer and Jackson, 1974).
Since T|_ is directly porportional to r, more rapid response may be ob-
tained by reducing the area of the opening of the piezometer. Lissey (1967)
presents a piezometer design based on this principle.
81
-------
Observation wells—An observation well consists of an uncased borehole or
perforated pipe extending from ground surface into a perched water table. For
shallow water tables, observation wells may be installed by hand augers or
simple drilling equipment. For deeper wells (e.g., those extending below the
main water table), it will be necessary to use standard drilling equipment.
Development of cased wells by surging or pumping may be necessary to ensure
free flow of water into and out of perforations.
A principal function of observation wells is to permit observation of wa-
ter table fluctuations. During water-spreading operations, for example, water
level changes in observation wells may manifest the arrival of downward flow-
ing effluent. Water levels are measured by means of chalked tape, electric
sounders, or air lines. Continuous records of water table fluctuations are
possible with the use of automatic water-stage recorders. Water table read-
ings should be referenced to a datum, e.g., the top of the casing, so that
levels from a network of wells can be used to construct isopiestic maps.
Under steady flow conditions, an observation well may perform adequately
in providing data on the location of the water table. However, during tran-
sient vertical flow, reliable estimates are difficult to obtain for reasons
stated by Kirkham (1947):
If there is downward seepage, the bottom of the perforated pipe
acts as a sink and the water table is accordingly depressed
about the pipe; if there is upward seepage, the bottom of the
perforated pipe acts as a source, the water table then mound-
ing up about the pipe.
Saturated Hydraulic Conductivity--
Alternative techniques have been developed for estimating the saturated
hydraulic conductivity of vadose zone materials either in the presence or ab-
sence of a water table. Detailed reviews of these methods were presented by
Bouwer and Jackson (1974) and Bouwer (1978).
Estimates of K within perched groundwater bodies of the vadose zone may
be obtained by pumped hole techniques such as the auger hole method, the pie-
zometer method, tube method, well point method, and multiple-well methods
(Bouwer and Jackson, 1974). The auger hole, piezometer, and well point meth-
ods entail installing a single cavity (cased or uncased) below the water
table, lowering the water level in the cavity by pumping or bailing, and mea-
suring the rate of recovery of the water level. Knowing the geometry of the
cavity and the known or assumed depth to an impermeable layer, suitable equa-
tions, curves, or nomographs may be used to calculate K. In the multiple-well
method, water is pumped from one well or set of wells into other wells until
the difference in water levels between the wells becomes stabilized. Knowing
the geometry of the cavities and flow system, appropriate equations or curves
are used to determine K (Bouwer and Jackson, 1974, pp 612-627). All of the
above methods, except the tube method, determine K primarily in a horizontal
direction.
82
-------
If the perched groundwater region is extensive containing productive
wells, it may be possible to use standard pumping test methods to determine
the transmissivity, T, of the formation. Such methods are reviewed in detail
by Lohman (1972). Knowing the transmissivity of the system and the thickness
of the layer, m, the perching layer, the hydraulic conductivity is estimated
by the relationship T/m = K. A problem may exist in using this method if
leakage beneath the perching layers is substantial.
In many waste disposal operations, it is of value to estimate the satu-
rated hydraulic conductivity in the vadose zone even though a water table is
absent. For soils, the saturated K value indicates the potential intake rate.
Thus, comparison of K values of soils at a number of sites using the methods
discussed below will facilitate locating a disposal operation to minimize deep
seepage. Similarly, a range of K values in deeper strata of the vadose zone
will indicate the location of possible perching layers. In addition, if hy-
draulic gradients are assumed to equal unity, the K values represent an upper
limit of flux through the vadose zone.
A number of field techniques have been developed for estimating K in
shallow soils in the absence of a water table. Determination of the hydraulic
conductivity, K, in the absence of a water table generally involves techniques
which bring the soil at the measuring point to saturation or near-saturation.
Bouwer and Jackson (1974) reviewed five possible methods: (1) the shallow
well pump-in method, (2) the'cylinder permeameter method, (3) the infiltration
gradient technique, (4) the air-entry permeameter technique, and (5) the
double-tube method. The shallow well pump-in method measures K mainly in the
horizontal direction and is suitable for stoney soils. The cylinder permeam-
eter and infiltration gradient techniques measure K in the vertical direction
and are not suitable for stoney soils. The air-entry permeameter measures K
in a vertical direction and with care may be used in stoney soils. An impor-
tant consideration in using these methods is to minimize entrapping air in the
pores of the media. Obviously, occluded air will decrease the value of the
measured K below the true saturated value. For additional details on these
methods, including associated equations and techniques for determining K, the
reader is referred to the review of Bouwer and Jackson (1974).
The above methods are suitable primarily for determining Kin shallow re-
gions of the vadose zone.
Laboratory and field methods are currently available for measuring the
hydraulic conductivity in deeper unsaturated regions of the vadose zone. Lab-
oratory related techniques which have been used to estimate K values from
drill cuttings include permeameter tests and empirical relations between the
grain-size distribution and K. Permeameter tests are not particularly mean-
ingful because of the disturbance caused by the drilling process. In con-
trast, certain empirical relationships might exist between grain-size analyses
and the hydraulic conductivity. Davis and deWiest (1966, p 375) present a
table relating hydraulic conductivity to the dominant size of selected
sediments.
The U.S. Bureau of Reclamation (1977) has developed two alternative field
techniques for estimating the saturated hydraulic conductivity, K, of
83
-------
unconsolidated unsaturated sediments of the vadose zone. One approach entails
pumping water into a borehole at a steady rate such that a uniform water level
is maintained within a basal test section. Knowing the dimensions of the hole
and inlet pipes, the depth of water, and the constant inflow rate, appropriate
equations and curves are consulted to calculate K. For depths less than 40
feet, the hole is cased to the desired depth. A water inlet pipe and a sepa-
rate water level measurement pipe are then placed inside the casing. Gravel
is then poured inside the casing to ensure a gravel pack throughout the test
section. Subsequently, the casing is pulled back and a test is initiated.
For depths greater than 40 feet, it is necessary to use preperforated casing
to isolate the test section. The total open area of the perforations must be
known. An observation pipe is placed in the casing on a 6-inch bed of gravel.
The use of preperf orated casing permits driving the casing in depth-wise in-
crements such that a profile of K values is obtained.
The second method for estimating K is used in the vicinity of a wide-
spread lens of slowly permeable material. The method entails installing an
intake well and a series of observation wells. Water is pumped into the well
at a steady rate and the water level response in the observation wells is re-
corded. Appropriate equations and curves are consulted to calculate K.
Weeks (1978) presented a method for measuring vertical air permeability
values of layered materials in the vadose zone. Basically, the method entails
measuring air-pressure changes in specially constructed piezometers during
barometric pressure changes at the land surface. By coupling pressure-
response data with auxiliary information on air-filled porosity and numerical
solutions of the one-dimensional flow equation, an estimate of the air perme-
ability is obtained. If the material is well-drained and permeable enough
that the Klinkenberg effect (see Glossary) is minimal, air permeability values
may be converted to the corresponding hydraulic conductivity values. The in-
terested reader should consult the report of Weeks (1978) for additional
details.
Flow Rates in Perched Groundwater--
The velocity of water movement in extensive perched groundwater and the
total discharge are frequently of interest. These characteristics are esti-
mated from the following forms of Darcy's equation:
H? - H-i
v = -K ( 2 L l) (30)
and
H9 - H1
Q = _KA ( 2 L l) (31)
where v = specific discharge
K = hydraulic conductivity
H2 = total head in downstream well
84
-------
HI = total head in upstream well
L = distance between wells
A = cross-sectioned area of aquifer normal to flow direction
Q = total discharge.
The hydraulic conductivity, K, of the perched system could be estimated
using the methods discussed above. For an example of the use of these equa-
tions, the reader is referred to Bouwer (1978, pp 46 and 47).
Combination of Methods for Observing Saturated
and Unsaturated Flow—
A combination of methods will be necessary for monitoring the flux of
wastewater from ground surface into a perched groundwater table. Such a sys-
tem is shown on Figure 27. The tensiometers and electrical resistance blocks
could be used to monitor flux, hydraulic gradients, and the direction of water
movement in the unsaturated region. The observation well is useful in esti-
mating the response of the water table to recharge. The piezometer nest is
used to detect the vertical direction of flow beneath the water table.
Again, it must be kept in mind that observation wells and piezometers
will not accurately detect the location of the water table in a fluctuating
system. However, by installing an access well and monitoring water content,
positional changes in the water table may be monitored via a neutron logger.
Although more effective than observation wells, the piezometers in monitoring
water-level response, the neutron logger measures water content changes and
not energy changes.
85
-------
00
en
^TENSIOMETERS
0 D
ELECTRICAL
RESISTANCE
CELLS
-OBSERVATION
WELL
PIEZOMETER-
LAND SURFACE
NEUTRON
LOGGER
^WA
WATER TABLE
—ACCESS WELL
Q_
UJ
Q
Figure 27. Array of devices for monitoring water movement in the vadose zone and perched
groimdwater (after Freeze and Cherry, 1979).
-------
SECTION 6
MONITORING QUALITY CHANGES IN THE VADOSE ZONE
So far in our discussion on monitoring in the vadose zone, we have con-
centrated on the storage and movement of water, the vehicle for transmitting
pollutants. In a quality monitoring program, the major emphasis will be to
monitor the spatial and temporal variations of specific water-borne pollu-
tants. In the ensuing discussion, such pollutants are categorized as follows:
major inorganic chemical constituents, trace chemical constituents, organic
chemical constituents, microbial constituents, and radionucl ides. Physical
constituents, such as temperature and suspended sediment, will not be
considered.
Particular constituents within each of these categories which should be
monitored in the vadose zone will obviously be site-specific, i.e., depend
upon the waste disposal operation. For example, when monitoring agricultural
return flows, specific pollutants may include the nitrogen series, particu-
larly N03-N, phosphate, total dissolved salts (TDS), and pesticides. For
land treatment sites, monitoring may emphasize bacteria, virus, heavy metals,
N03-N, and phosphate, TDS, total organic carbon (TOC), biochemical oxygen
demand (BOD), and chemical oxygen demand (COD). For sanitary landfills, a
range of pollutants in all four categories should be monitored in leachate,
with particular emphasis on heavy metals. Finally, for hazardous waste dis-
posal operations, great emphasis should be placed on monitoring the fate of
specific organic toxins during flow in the vadose zone.
In the following discussion, reactions affecting the movement of pollu-
tants in the vadose zone are briefly discussed and current methods for vadose
zone monitoring are reviewed.
TECHNICAL REVIEW
As with water movement, the present state of knowledge on pollutant move-
ment in the vadose zone is derived from the efforts of specialists working in
soils and deeper geological regions. For example, soil chemists and soil
microbiologists have examined chemical and microbiological interactions, re-
spectively, in soils. Similarly, geochemists have studied low-temperature
chemical reactions in groundwater. Among the references which the interested
reader may wish to consult for greater detail on soils/geochemistry relative
to pollutant movement are the following: (1) Bohn, McNeal, and O'Connor
(1979), a review of the modern concepts of soils chemistry; (2) Alexander
(1961), a text on soil microbiology; (3) Hem (1970), a detailed review of wa-
ter chemistry, including a discussion of chemical constituents in groundwater;
87
-------
and (4) texts by Davis and deWiest (1966), Bouwer (1978), and Freeze and
Cherry (1979) containing chapters on water quality.
Several review papers have also been published in recent years dealing
with interactions between water, pollutants, and vadose zone materials. Such
papers include those of Ellis (1973), Rhoades and Bernstein (1971), Murrmann
and Koutz (1972), McNeal (1974), Pratt et al. (1978), and Chang and Page
(1979). In his review paper, Runnel!s (1976) itemized 11 physical-chemical
processes that may operate in the subsurface to purify liquid wastes. The
specific processes are: dilution, buffering of pH, precipitation by reaction
of wastes with indigenous water or solids, precipitation due to hydrolysis,
removal due to oxidation and reduction, mechanical filtration volatilization
and loss as a gas, biological assimilation or degradation, radioactive decay,
membrane filtration and sorption. Runnel Is (1976) briefly discussed each of
these factors, and the interested reader should consult his paper for details.
Fuller (1977) also presented a review article dealing with the movement of
selected heavy metals in soils. Among the factors which Fuller (1977) exam-
ined relative to the movement of these metals were: soil pH, oxidation-
reduction (eH), surface area of soils, pore-size distribution, organic matter,
concentration of ions or salts, and presence of hydrous oxides. Pratt et al.
(1978) discussed the removal of biological and chemical contaminants by soil
systems. The removal process included filtration, adsorption, decomposition,
ion exchange, oxidation-reduction, chemical complex formation, chemical pre-
cipitation, and other chemical reactions.
In recent years, a number of soil column experiments have been conducted
in the laboratory to identify the factors influencing the mobility of pollu-
tants. The column studies of Fuller (1978) and his associates at the Univer-
sity of Arizona on the mobility of trace contaminants from landfill leachate
are an example of such laboratory studies. Based on statistical analyses of
these laboratory studies, Korte et al. (1976) found that the following factors
were dominant in affecting the movement of trace contaminants in soil\s: soil
texture and surface area, percentage of free oxides, and pH.
Field studies have also been conducted to determine the factors affecting
the movement of pollutants in vadose zones. Apgar and Langmuir (1971) ob-
tained water samples to a depth of about 55 feet from the vadose zone beneath
a landfill in Pennsylvania. (The technique used to obtain these samples was
suction cup lysimeters, described later in this report.) Based on their field
observations, Apgar and Langmuir (1971) concluded that the following factors
were important in attenuating pollutants: dilution and dispersion, oxidation,
chemical precipitation, cation exchange, and anion exchange. During field
studies to determine chemical changes in water during artificial recharge at a
site in Texas, Wood and Signor (1975) examined the following mechanisms: ca-
tion exchange, anion exchange, mineral solution, adsorption, desorption, and
sulfate reduction. They concluded that at their site, ion exchange and de-
sorption were the major mechanisms affecting water quality in the vadose zone.
88
-------
Ion Exchange, Cation Exchange Capacity, and
Sorption/Desorption Effects in Soils
It is generally recognized by soil scientists that among the factors af-
fecting the mobility of the chemical constituents of water in soils, ion ex-
change and sorption/desorption are of prime importance. A brief review of
these factors is presented in this section. For details, the reader should
consult the text of Bohn, McNeal, and O'Connor (1979).
Ion exchange represents an important soil water interaction for altering
the composition of the soil solution. Ion exchange is a function of the ca-
tion exchange capacity (CEC) of a soil, which in turn is related to colloidal
clay minerals, soil organic matter, iron, and aluminum sesquioxides and hy-
drous oxides.
Clays commonly produce an overall net negative change on their exchange
complex. Murrmann and Koutz (1972) pointed out that the cation exchange pro-
perty arises from the need to balance the negative charge on clay micelles to
maintain neutrality. To accomplish this, a swarm of positive ions in the soil
solution becomes associated with the negative charge on the exchange complex.
These ions are mobile and readily exchange with other cations in the soil so-
lution to maintain chemical equilibrium. This process represents cation ex-
change. Representative CEC values range from 100 milliequivalents per 100 gms
(meq/100 gms) for montmorillonite clays to 5 meq/100 gms for kaolonite type
clays.
For most soils, the level of exchangeable cations on the exchange complex
is greater than the amounts in the soil solution. Thus, a plot of ionic con-
centration versus distance from the clay surface (i.e., from the micelle sur-
face) shows a marked drop-off in cation concentration away from the surface,
approaching asymptotically to the level in the soil solution. In contrast,
the concentration of anions in the vicinity of the micelle is lower than the
concentration in the soil solution. In fact, because of this repulsion of
anions, the anion concentration in the solution extracted from clay may be
greater than in the solution originally added (Rose, 1966), a phenomenon known
as negative adsorption.
The ease with which a given ion on the exchange complex exchanges with
ions in solution is, among other things, a function of the valence of the
ion—the smaller the valence of the ion, the more readily is it exchanged.
The ease of replacement of common soil cations is (Rose, 1966): Li > Na >
K > Mg > Ca > Ba > Al.
The organic matter fraction of a soil may contribute substantially to the
CEC. In fact, additions of organic matter through sludge application may im-
prove the CEC of soils with naturally low CEC (Broadbent, 1973). In addition
to contributing a negative charge to the soil exchange complex, organic matter
together with the sesquioxides may contribute a positive charge. Apparently,
this phenomenon is highly pH dependent.
A complex relationship exists between soluble salts in the soil solution,
those on the exchange complex, and salts in the solid (precipitated) phase.
89
-------
The complexity is increased when it is necessary to account for the effects of
soil water. Aspects of this problem will be reviewed later under the discus-
sion of methods to estimate CEC and soluble salts.
In the opinion of Murrmann and Koutz (1972), adsorption of metals onto
the surface of soils is the most important process for removing chemicals from
wastewater. In contrast to ion exchange in which ions retain their mobility,
in adsorption reactions, ions are held so tightly that they become essentially
immobile. Exact mechanisms for adsorption are not clear although covalent
bonding appears to be important (Chang and Page, 1979). Methods to quantify
the extent of adsorption of ions onto solids have been developed. In particu-
lar, adsorption isotherms, such as the Freundlich isotherm and the Langmuir
adsorption isotherm, are commonly used. Ellis (1973) discussed the use of
these isotherms to quantify the adsorption of phosphate, sulfate, boron, and
heavy metals by soils.
Rhoades and Bernstein (1971) indicated that the tremendous sorptive area
of soils, coupled with high cation exchange capacities, influences the fate of
pesticides, radionucl ides, nutrient elements, and other inorganic solutes dur-
ing disposal in soils.
In their artificial recharge studies in Texas, Wood and Signor (1975)
concluded that silica was the only constituent affected by desorption. The
source of the mobilized silica apparently was not dissolution of quartz or
amorphous silica, but desorption of monomeric silica from the aquifer matrix.
They concluded that, "The amount of mobilized silica was very large, and the
process of desorption should be considered in a predictive model of water
quality in areas similar to the one desorbed."
A commonly used measure of the partitioning of ionic species between
solid and liquid phases as a consequence of processes such as adsorption is
the "distribution coefficient," K
-------
where v-,- = pore velocity of adsorbed species
vw = pore velocity of migrating water
9 = volumetric water content
Dj, = bulk density of soil.
This equation is applicable to both saturated and unsaturated media.
Factors affecting the use of equation 33 are: (1) the assumption that reac-
tions are fast and reversible may not be satisfied, (2) the adsorption iso-
therm may not be linear, (3) the distribution coefficient may not be constant
with varying soil-water contents, and (4) concentrations of other ionic spe-
cies may affect the magnitude of K
-------
chemical composition may not be the same as that for solution from the smaller
sequences (Rhoades and Bernstein, 1971).
For surface soils, other factors should be examined in the determination
of soluble salts. For irrigated soils or on land disposal operations subject
to wetting and drying, the water content of soils may range over large values
because of evaporation, so that conversion to the field water content value
may not be meaningful (Pratt, Jones, and Hunsaker, 1972). In this case, the
saturation extract technique is recommended (Rhoades and Bernstein, 1971).
Water content by this method represents about twice that at field capacity.
Therefore, the salt content extracted from the saturated sample is about one-
half the concentration at field capacity. Bower and Wilcox (1965) explain in
detail the procedure to obtain a saturated extract. Briefly, deionized water
is mixed into a weighed oven-dry sample of soil until the soil glistens and no
free water has collected on the soil surface. If this end point is sustained
for an hour or more, the sample is placed in a Buchner funnel. Vacuum is ap-
plied and the filtrate collected for analyses. Total salt concentration of
the extract may be estimated by measuring the specific electrical conductance
and using the relationship (Bower and Wilcox, 1965): salt concentration
(mg/1) = 640 x electrical conductivity, mmho/cm (Ayers and Westcot, 1976).
If detailed information is required on specific ionic constituents in ex-
tracts, chemical analyses using procedures in "Methods of Soil Analyses"
(Black, 1965) may be used. According to Bower and Wilcox (1965). the princi-
pal ions of importance in soils studies are: Ca++, Mg++, K+, Na ,
COf, SOf, CL~, and boron, as well as the nitrogen series.
Cation Exchange Capacity and Exchangeable lons--
The sum of individual exchangeable basis of a soil sample is equal to the
cation exchange capacity. Alternatively, the CEC may be obtained directly
using methods detailed by Chapman (1965). Briefly, the exchangeable cations
in a soil sample are replaced by either ammonium acetate or sodium acetate,
and the amounts of ammonium and sodium ion adsorbed are determined. A problem
may develop with the use of ammonium ion because this ion becomes strongly ad-
sorbed on some clays. Cation exchange capacity is expressed as mi Hi equiva-
lents per 100 gms (meq/100 gms) of sample.
Sorptive Capacity: Specific Surface--
The cation exchange capacity and specific surface together govern the
sorptive characteristics of a soil. Mortland and Kemper (1965) discussed the
principles of adsorption relating to the specific surface of clays and re-
viewed a number of adsorption isotherms. A method for determining specific
surface based on sorption of ethylene glycol was presented.
Column Studies and Batch Tests—
An approximate idea of the attenuating properties of vadose zone materi-
als for specific pollutants can be obtained from laboratory column studies.
In practice, cylindrical columns are packed with vadose zone samples (core
samples could also be used) and subsequently flooded with wastewater from the
92
-------
disposal site. Samples of column effluent are collected and analyzed and the
breakthrough of particular constituents is determined. Advantages of labo-
ratory column tests include the following: (1) the method is simple, (2)
results are related to a specific mass of material, and (3) results may gen-
erally be obtained in a short time. Disadvantages include: (1) if disturbed
samples are used, water movement in the column may differ from in-place flow,
(2) flow along the column walls may occur, and (3) flow rates may differ from
in-place rates.
Batch testing is an alternative method for estimating the attenuation of
water-borne pollutants as a result of interactions in the vadose zone. Batch
tests consist of placing fragmented samples of vadose zone material in flasks
together with a measured aliquot of wastewater. The flasks are shaken via a
mechanical shaker for a given period of time. The fluid is subsequently
drained from the flasks and analyzed for constituents of interest. Advantages
of batch tests include the following: (1) the method is simple and inexpen-
sive, (2) results are related to a specific mass of material, (3) results are
obtained in a short time, and (4) results may be used to prepare adsorption
isotherms or selectivity coefficients in ion exchange reactions (Freeze and
Cherry, 1979).
Disadvantages of batch tests include the following: (1) results may be
affected by sample disturbance, (2) flow conditions differ from those in
place, and (3) samples are exposed to the air, i.e., adsorptive capacity of
oxidized material may differ from reduced material, affecting the transfera-
bility of results to field cases (Freeze and Cherry, 1979).
Attenuation of Specific Constituents in the Vadose Zone
Major Chemical Constituents—
Certain wastewaters such as landfill leachate may contain excessive con-
centrations of the major chemical constituents, including calcium, magnesium,
sodium, nitrate, chloride, sulfate, phosphate, and bicarbonates. The fate of
both calcium and magnesium is dependent upon precipitation and affected by
sorption on the cation exchange complex of clays. Generally, calcium and mag-
nesium will precipitate during reactions with bicarbonate and sulfate. McNeal
(1974) discusses precipitation reactions in detail. Precipitation of carbon-
ates occurs as a result of the concentration of the soil solution, although
the solubility is affected by the partial pressure of C02 in the gas phase
and the salt concentrations of the solution. McNeal (1974) illustrated the
use of the "Langelier saturation index" to predict the approximate amounts of
calcium carbonate precipitating from waters.
Next to precipitation in the carbonate form, the precipitation of calcium
in the form of CaSCvj (gypsum) is important as a process for removing calcium
from solution. It is possible to predict the precipitation of gypsum using
Debye-Huckel Theory.
The solubility of MgC03 is also influenced by the presence of C02,
and free movement within the vadose zone will occur until the pH elevates.
93
-------
Sodium and potassium salts are soluble and mobile unless concentrations are
increased to several thousands of parts per million.
In addition to precipitation, calcium and magnesium mobilities will be
limited by exchange reactions in clays. For this case, calcium and magnesium
will be in competition with sodium for exchange sites. The relative degree of
sodium adsorption is expressed by the adjusted sodium adsorption ratio, adj
SAR, defined by Ayers and Westcot (1976) as:
adj SAR = Na+/((Ca++ + Mg++)/2)l/2 (l + (8.4 - pHc)). (17)
Terms of this equation were defined earlier.
The mobility of nitrogen compounds in vadose zone materials is primarily
related to the oxidation-reduction potential. Thus, if the system is aerobic,
nitrification of organic-N sources occurs readily, producing N03-N as an end
product. Nitrate-N is highly mobile, moving readily with the soil solution
into the lower vadose zone and ultimately into groundwater. Under anaerobic
conditions, nitrification is inhibited and the Nfy-N form predominates.
Ammonium-N is attenuated in soils with clays or organic materials by two mech-
anisms: sorption of the positively charged ammonium ion on the clay-organic
exchange complex, and fixation of the ammonium ion within the crystal lattice
of clay minerals. The latter mechanism is particularly pronounced in clays
with a 2:1 silica-alumina ratio (Nomrrrik, 1965). Ammonia sorption on the ex-
change complex is affected by the presence of other cations, such as calcium
and magnesium which compete for available sites.
In soil systems in which wastewater is applied intermittently, nitrifica-
tion of sorbed Nfty-N occurs during drying cycles (Bouwer, 1978). Conse-
quently, during subsequent flooding cycles, NC^-N may be leached into the
lower vadose zone.
Under anaerobic conditions, nitrogen may be lost through denitrification.
In particular, denitrification occurs quite readily as a result of the activ-
ity of heterotrophic bacteria, which convert NC^-N to volatile gasses such
as N20 and N2 via a number of intermediate compounds. An organic sub-
strata is necessary as an energy source for denitrifying microorganisms.
During controlled field studies, Rolston and Broadbent (1977) noted that de-
nitrification was also affected by soil-water content, soil temperature, and
by the presence or absence of a growing crop in the soil.
The mobilities of sulfate and bicarbonate are linked to reactions with
calcium and magnesium. In addition, sulfate may be sorbed to a minor extent
in the aquifer matrix and be retained by Fe hydrous oxides (Keeney and Wild-
ung, 1977). The latter reaction, however, requires low pH. The presence of
organic acids in leachate migrating into the vadose zone may result in reduc-
ing sulfate mobility. Reduction of sulfate under anaerobic conditions may
lead to the formation of ^S and eventually insoluble sulfides such as ZnS
and Fe$2. An increase in bicarbonate often accompanies an increase in sul-
fate (Wood and Signer, 1975).
94
-------
Factors influencing the solubility of iron include pH, the redox poten-
tial (eH), and the dissolved C02 and sulfur species (Hem, 1970). The re-
duced form, i.e., the ferrous ion, is soluble and mobile. The oxidized form,
or ferric ion, forms relatively insoluble precipitates with sulfur and carbon-
ate species. Hem (1970) presented a pH-Eh diagram which demonstrates the con-
ditions under which iron solubility is very low.
Phosphate retention and mobility were discussed by Keeney and Wildung
(1977). Under acid conditions, phosphorus is sorbed on the surface of Fe and
Al containing minerals. Organic acids may have a local effect on lowering pH
values, promoting the above effect. For alkaline conditions, the sorption of
P on CaCO? or formation of Ca phosphate minerals may occur (Keeney and Wild-
ung, 1977). Phosphate retention on clays and hydrous oxides may also be
important.
Trace Contaminants—
Among the general mechanisms associated with the removal of trace metals,
it appears that adsorption is of prime importance. In reviewing the results
of recent laboratory studies on metal attenuation, Chang and Page (1979)
indicated:
... strong adsorption of trace metal ions occurred at the sur-
face of amorphous iron and manganese oxides and aluminum min-
erals.... Other soil properties such as texture and cation
exchange capacity did not appear to significantly influence
soil adsorption characteristics. Unlike the electrostatic ca-
tion exchange reactions, covalent bonding-induced adsorption
generally is more specific and the reactions are not easily re-
vised by the presence of other cations.... Even trace elements
which usually form anions in aqueous solution may be effec-
tively adsorbed or rapidly converted into insoluble forms.
Murrmann and Koutz (1972) also compared the relative effects of cation
exchange and adsorption for removal of heavy metals in wastewaters. They in-
dicated that in contrast to cation exchange, "... soil has a capacity to re-
tain heavy metals so tightly that they can be replaced only with difficulty."
Based on results of laboratory studies, Korte et al. (1976) developed
figures on the relative mobility of 11 trace contaminants in the 10 most prom-
inent soils orders of the United States. These figures are reproduced on Fig-
ures 28 and 29.
Recently, Fuller (1977) published a comprehensive report on the movement
through soils of the following elements: arsenic, beryllium, cadmium, chro-
mium, copper, cyanide, lead, mercury, selenium, and zinc. The interested
reader should review this report for details. According to Fuller, the prin-
cipal mechanism for attenuating arsenic is adsorption by soil colloids. How-
ever, if waterlogging should occur during migration of leachate, reducing
conditions will favor the mobilization of arsenic. Finally, according to Ful-
ler, "At the low concentrations usually found in wastewaters, landfill leach-
ates, and other aqueous waste streams, As probably will not precipitate in
95
-------
SOILS
MOLOKAI
NICHOLSON
MOHAVECQ
FAN NO
MOHAVE
DAVIDSON
AVA
KALASKA
ANTHONY
WAGRAM
Cu
Pb
LJDV
Be
Zn
/ MOBILITY
MODE
Cd
Ni
:RATE MOBILITY
Hg
HIGH MOBILITY
Figure 28. Mobility of Cu, Pb, Be, Zn, Cd, Ni, and Hg
in 10 soils (after Korte et al., 1976).
SOILS
MOLOKAI
NICHOLSON
DAVIDSON
AVA
FANNO
MOHAVEr
CO
KALASKA
MOHAVE
WAGRAM
ANTHONY
Se
L
^
V As Cr
.OW MOBILITY
MODERATE
MOBILIIY
1IGH MOBILITY
Figure 29. Mobility of Se, V, As, and Cr in 10 soils
(after Korte et al., 1976).
96
-------
soils except possibly as an impurity in phosphorus compounds formed over a
long period of time."
The complex chemistry of selenium is described by Fuller. The behavior
of selenium is closely related to that of sulfur in acid formation and other
properties. Based on experimental studies at the University of Arizona, Ful-
ler concluded that other factors being equal, selenium is less mobile in
acidic than in neutral or alkaline soils.
Regarding the mobility of zinc, Fuller (1977) indicated that Zn2+ forms
slowly soluble precipitates with carbonate, sulfides, silicate, and phosphate
ions. Sulfide values may increase in leachate if waterlogging occurs, promot-
ing precipitation of zinc (and other cationic heavy metals). Zinc is also
strongly sorbed on the exchange complex of soil.
Unlike other members of the halogen group, fluoride compounds tend to be
rather insoluble (Hem, 1970). The mobility of fluoride in leachate may be
limited by the formation of fluorite (CaF2) with a solubility product of
10-10.57 (Hem, 1970). High concentrations of calcium in leachate would tend
to favor fluorite formation in spite of common ion effects.
Reactions of strontium in water are similar to those of calcium. Stron-
tianite, formed by the reaction of strontium and bicarbonate, is slightly less
soluble than calcite (Hem, 1970). In addition, relatively insoluble strontium
sulfate may be formed in sulfate rich waters (Davis and deWiest, 1966). Both
reactions may occur, limiting the mobility of strontium.
Copper appears to be strongly complexed to organic matter (Fuller, 1977);
consequently, copper-organic chelates may form in leachate which may be solu-
ble and mobile. The formation of hydrous oxides of Mn and Fe provides the
main control in the immobilization of copper. Hem (1970) reported that copper
solubility is generally lower in reducing systems than in oxidizing systems,
particularly if reduced sulfur species are present. Reducing or anaerobic
conditions could exist if saturation develops. Reduction of sulfate, present
in leachate, would then lead to precipitation of copper (as well as iron,
zinc, cadmium, lead, and mercury (Fuller, 1977)).
Both nickel and cobalt are strongly adsorbed by iron and manganese oxides
(Hem, 1970). The low solubility of CaCO^ may be an important factor in lim-
iting cobalt concentrations in solutions in which HCO§ concentrations
are high. According to Hem (1970), there are no effective solubility controls
over molybdenum concentrations in water. Consequently, the mobility of the
anionic form, molybdate, will probably be high.
Jurinak and Santillan-Medrano (1974) examined the transport of lead and
cadmium in soils. Processes considered to be important in governing cadmium
and lead movement in soils included the following: precipitation and dissolu-
tion, ion-pair formation, pH flux, cation exchange, and adsorption. Jurinak
and Santillan-Medrano (1974) concluded that the principal mechanism regulating
Pb solubility in noncalcareous soils was precipitation of the forms:
Pb(OH)2 and PbstPO^OH. In calcareous soils, PbCQ^ also precipi-
tates. "Calcareous soils appear to be an excellent sink for Pb^+ ions."
97
-------
Jurinak and Santillan-Medrano found that the solubility of cadmium in
soils is about 100 times greater than lead in the pH range 5 to 9. "The data
suggest that particularly at low concentrations, the adsorption of Cd by the
soils is a more important mechanism in retention than in the case of Pb where
precipitation of slightly soluble compounds/regulates solubility."
Pratt et al. (1978) indicated that in solution, mercury exists in inor-
ganic form as a monovalent or divalent ion and as complex ions or ion pairs.
Regarding mobility of mercury, these authors stated, "The concentrations of
Hg in soil solutions are governed by ionic adsorption by organic and inorganic
materials and by the low solubilities of Hg as phosphate, carbonate, and
sulfide."
Alesii (1976) studied the mobility of simple and complex forms of cya-
nide in soil columns. He observed, "Cyanide as Fe(CN),3- and CN~ in water
were both found to be very mobile in soils. Soil properties, such as low pH,
percentage free iron oxide and Kaolin, chloride, and gibbsite type clay (high
positive charges) tended to increase the attenuation of cyanide. High pH,
presence of free CaCOs (high negative charge), low clay content, and mont-
morillonite clay tended to increase the mobility of the cyanide forms." When
an abundant energy source is available, microorganisms volatilize cyanide into
less harmful forms of nitrogen. However, as noted by Fuller (1977), anaerobic
conditions inhibit the microbial degradation of a cyanide. In a survey of
municipal wastewaters in California, Pratt et al. (1978) found that the maxi-
mum cyanide level was 0.16 mg/1. They concluded that during recharge, cyanide
levels would be reduced well below the recommended limit of 0.2 mg/1.
Organic Constituents--
A problem in specifying the mobility of specific organic pollutants is
that quantitative studies have only recently been reported. One problem is
that analytical procedures to identify organic constituents are still being
developed. "Methods of sampling and analyzing for most organics are in their
relative infancy compared to those used for other pollutants" (U.S. EPA,
1979a).
Two alternative methods which may have applicability in characterizing
organic pollutants during flow in the vadose zone have been proposed by the
U.S. EPA (1979a). One method is quantitative, employing GC/MS (gas chromatog-
raphy/mass spectrometry). The other method provides for "screening" to detect
the presence or absence of organic toxins by GC/MS, followed by extra GC or LC
(liquid chromatography) quantification of identified pollutants.
Recently, Leenheer and Huffman (1976) described the development of the
dissolved organic carbon (DOC) technique for fractionating organics into hy-
drophobic and hydrophilic components using macroreticular resins. The tech-
nique was applied to several natural waters. This technique has advantages
over other methods for concentrating organics, such as activated carbon. For
example, Robertson, Toussaint, and Jerque (1974) reported that only 10 percent
of organics present in groundwater beneath a landfill in Oklahoma were identi-
fied using carbon adsorption followed by carbon chloroform and carbon alcohol
extraction.
98
-------
The principal mechanisms for attenuating organic pollutants are adsorp-
tion and decomposition (Pratt et al., 1978). As an example of the importance
of adsorption, Robertson, Toussaint, and Jorque (1974) observed that polychlo-
rinated biphenyls (PCBs) tend to be strongly adsorbed in soils. Leenheer and
Huffman (1976) indicated that both hydrophobic and hydrophilic organics may be
sorbed by sediment. Chang and Page (1979) discussed adsorption of organic
substances on soils in some detail:
The adsorption of dissolved organic substances usually is com-
plicated by the chemical characteristics (i.e., molecular
weight, structure and the presence of various functional
groups) of each organic compound.... Since many organic sub-
stances tend to form negatively charged colloids in aqueous
solution, the most likely sites for adsorption to take place
would be the positively charged edges of clay minerals. These
organic substances also form organometallic complexes which may
be adsorbed as neutral or positively charged molecules.
Biodegradable organics are decomposed by microorganisms rendering poten-
tially harmful constituents into gases. Davis (1956) reported that microbio-
logists have observed the utilization of hydrocarbons by certain bacteria,
actinomycetes, filamentous fungi, and yeasts. In general, however, hydro-
carbons are not as readily decomposed as carbohydrates, proteins, or fats.
Furthermore, the cyclic hydrocarbons are less susceptible to microbial decom-
position than are the aliphatic hydrocarbons. Chlorination of wastewaters
prior to recharge may destroy both pathogenic organisms and microorganisms re-
sponsible for decomposition (Pratt et al., 1978).
The pH may also be a factor in the mobility of organics. For example,
Leenheer and Huffman (1976) noted the formation of an organic precipitate upon
acidification of a groundwater sample from an oil shale area near Rock
Springs, Wyoming. Raising the pH dissolved the precipitate. The pH of leach-
ate during aerobic decomposition within a sanitary landfill may be low enough
to cause the flocculation of certain organics. In the underlying media, floes
would then be filtered out. However, in alkaline soils, the pH would eventu-
ally elevate to a point in the flow system at which organic floes would
dissolve.
In studies on leaching of spent oil shale, Schmidt-Collerus (1974) noted
that the solubilization of polycondensed organic matter was enhanced by the
presence of seepage water high in TDS. Possibly, the flux of organics in
wastewater would be accelerated in a similar fashion if the initial TDS were
high.
Microorganisms--
According to Gilbert et al. (1976), bacteria are removed at the soil sur-
face by filtration, sedimentation, and adsorption. Virus removal occurs
mainly by adsorption, which increases with decreasing pH. Other factors
listed by Gilbert et al. as important in attenuating bacteria and virus are:
salt concentration, pH, organic matter, soil composition, infiltration rates,
and climatic conditions. Survival and movement of microorganisms within a
99
-------
soil relate to soil moisture content, temperature, pH, nutrient availability,
and antagonisms. The presence of excessive salt levels in wastewater, coupled
with the presence of high levels of toxic substances (e.g., trace contami-
nants, pesticides), may limit the migration of bacteria and virus to a small
sepent of the vadose zone beneath a landfill.
Specific factors affecting the movement of virus in soils are summarized
in Table 4. Chang and Page (1979) cautioned that immobilization of viruses by
soil should not be equated with virus inactivation. "Many adsorbed virus par-
ticles have been demonstrated to be infectious for significant periods of
time. Viruses immobilized by soil adsorption may also become desorbed when
the chemical composition of the percolating wastewater is changed."
Pesticides--
Factors influencing the fate and behavior of pesticides in soil systems
include: (1) chemical decomposition, (2) photochemical decomposition, (3)
microbial decomposition, (4) volatilization, (5) plant or organism uptake, and
(6) adsorption-desorption (Bailey and White, 1970). The last item, adsorp-
tion-desorption, directly or indirectly influences the magnitude of the other
five factors and is considered to be the prime factor governing the interac-
tions between pesticides and soil colloids. Leonard, Bailey, and Swank (1976)
classified organic pesticides as ionic or nonionic. In turn, ionic pesticides
are subclassified as cationic (paraquat, disquat), basic (s-triazones), and
acidic (benzoic acids, phenols, picolinic acid). Nonionic pesticides include
chlorinated hydrocarbons and organophosphates. Cationic pesticides are re-
tained tenaciously on the exchange complex. Changes in soil pH have a pro-
found but complex effect on pesticide forms. For example, a decrease in pH
increases the molecular form of an acidic pesticide but increases the conju-
gate acid form of the base (Leonard, Bailey, and Swank, 1976). These changes
will modify adsorption-desorption characteristics and the mobility of pesti-
cides. Leonard, Bailey, and Swank also reviewed the effect of clay on
sorption-desorption properties and pointed out that organic matter greatly in-
creases the sorptive tendency of soils.
Temperature effects adsorption-desorption of pesticides. For example,
an increase in temperature decreases adsorption and promotes desorption (Leon-
ard, Bailey, and Swank, 1976). Temperature also affects sorption through ef-
fects on solubility and vapor pressure.
Regarding soil moisture effects, Leonard, Bailey, and Swank indicated, "A
decrease in soil moisture (a) causes an increase in concentration per unit
volume, ... and may increase surface acidity, and, thus increases adsorption;
and (b) causes a decrease in competition with water for adsorption sites,
which should increase adsorption." Furthermore, when the pesticide concen-
trates to the solubility product, crystallization will result.
Pesticide degradation will occur in soils. Photodecomposition will re-
duce pesticide levels near the soil surface but may be inconsequential with
depth. Chemical degradation is a complex phenomena, related to pH, redox po-
tential, and surface acidity.
100
-------
TABLE 4. FACTORS THAT INFLUENCE THE MOVEMENT OF VIRUSES IN SOIL
(after U.S. Environmental Protection Agency, U.S. Army
Corps of Engineers, and U.S. Department of Agriculture,
1977).
Factor
Remarks
Rainfall
Soil
composition
Flow rate
Soluble
organics
Cations
Viruses retained near the soil surface may be eluted after a
heavy rainfall because of the establishment of ionic gradients
within the soil column.
Low pH favors virus adsorption;
adsorbed virus.
high pH results in elution of
Viruses are readily adsorbed to clays under appropriate condi-
tions and the higher the clay content of the soil, the greater
the expected removal of virus. Sandy loam soils and other
soils containing organic matter also are favorable for virus
removal. Soils with a low surface area do not achieve good
virus removal.
As the flow rate increases, virus removal declines, but flow
rates as high as 32 ft/d (9.8 m/d) can result in 99.9-percent
virus removal after travel through 8.2 ft (2.5 m) of sandy loam
soil.
Soluble organic matter competes with viruses for adsorption
sites on the soil particles, resulting in decreased virus ad-
sorption or even elution of an already adsorbed virus. Defini-
tive information is still lacking for soil systems.
The presence of cations usually enhances the retention of
viruses by soil.
Microbial decomposition of pesticides depends on such factors as "mi-
crobial population ecology, soil moisture and temperature, organic matter
content, pH, redox potential, pesticide concentration, availability for degra-
dation, and nutrient concentration and availability" (Leonard, Bailey, and
Swank, 1976). In addition to decomposition by microbial activity, pesticide
compounds may be taken up by plants.
Volatilization of pesticides will reduce their concentration in the soil.
Volatilization depends on such factors as temperature and soil moisture, as
well as on the vapor pressure of specific compounds.
Bailey and White (1970) listed the following factors as being most sig-
nificant in governing the leaching and movement of pesticides in soils: ad-
sorption, physical properties of the soil, and climatic conditions. The
effects of adsorption are reviewed above. Regarding soil physical properties,
Bailey and White indicated that pesticides are leached to a greater degree in
101
-------
light-textured soils than in heavier-textured soils. The porosity of soils
may have an effect on diffusion rates of volatile pesticides. Air diffusion
plays a prominent role in the eventual loss of pesticides from soil due to
volatilization. Pore-size distribution of soils affects the rate at which the
water infiltrates and moves through a soil.
For a given soil type, leaching of pesticides is increased with an in-
crease in the amount and frequency of rainfall. The same relationship holds
for irrigation. Evapotranspiration will tend to increase the concentrations
of pesticides at the soil surface.
Radioactive Wastes
The principal mechanisms relied upon for the attenuation of radionuclides
in the vadose zone are ion exchange and adsorption. Consequently, the pres-
ence of clays, organic matter, and hydrous oxides at disposal sites is of par-
amount importance in reducing the potential for groundwater pollution for
radioactive wastes. The "distribution coefficient," K^, is a commonly used
measure of the ability of a solid matrix to retard the movement of radionu-
clides (and other solutes) via sorption effects.
Borg et al. (1976) reviewed the chemical and physical factors affecting
the measured values of the distribution coefficient for radionuclides. They
stated:
Among other variables, mineralogy, particle size, nature of
solution, and chemical nature of radioactive species are impor-
tant.... In general, a decrease in particle size results in
an increase in Kj. Fresh silicate rocks have lower K^'s
than their altered counterparts. For the same reasons, old
fractures absorb more than fresh fractures in a given rock.
The sorption of Cs and Sr is greater than that of Ru and Sb and
is probably related to the fact that the latter two elements
form anionic complexes which do not readily take part in ion
exchange processes. Cs in clay-rich rocks often is sorbed more
than Sr because of lattice shrinkage that traps the larger ion.
Pu forms a positively charged polymer that is highly sorbed in
the pH range of 2 to 8.
Regarding the presence of ions in the soil solution competing with radio-
nuclides for sorptive sites, Borg et al. (1976) cited the work of Nelson
(1959). In particular, Nelson (1959) conducted batch and column experiments
on the sorption of Sr on Hanford soil in the presence of cesium, sodium, bar-
ium, calcium, magnesium, and aluminum. He found that water of hydration was a
significant factor in that cations with the combination of highest valence and
lowest hydrated radii provided the greatest competition with strontium for
sites. The work of Kokotov, Popova, and Urbanyuk (1961) was also cited by
Borg et al. (1976). Kokotov, Popova, and Urbanyuk (1961) measured Kj values
for Sr and Cs in Russian soils. For Sr, the order of effectiveness in lower-
ing the Kd (i.e., increasing the mobility) was Sr2+ > Ca2+ > Mg2+ > K+ > NHf
> Na+. For Cs. the order of effectiveness in lowering Kj values was Cs+ > Rb+
> NH4 > K+ > H+.
102
-------
The effect of soil pH on distribution coefficients of various radionu-
clides was examined by several investigators. The basis for their studies was
that changes in pH affect cation exchange properties of soils and concomi-
tantly Kj values. In particular, increasing the soil pH changes cation se-
lectivity by increasing both the cation exchange capacity and the preference
of the exchange complex for polyvalent over monovalent ions (Bohn, McNeal, and
O'Connor, 1979). Rhodes (1957b) evaluated the effect of pH on the uptake of
the following polyvalent radionuclides: plutonium, cerium, zirconium, yt-
trium, and ruthenium, as well as monovalent cesium and divalent strontium.
The soil used in his tests was from Hanford, Washington. Uptake of the poly-
valent species exhibited a maximum between about pH 4 and pH 8. Above pH 8,
a region of reduced uptake was observed, persisting up to pH 11. Rhodes
(1957b) noted that the effect of increasing pH on uptake of polyvalent radio-
isotopes is complicated by two possible consequences of changing pH: (1)
changing ion species, and (2) formation of polymers or colloids with changing
pH.
For example, in an earlier paper, Rhodes (1957a) concluded that plutonium
forms a positively charged polymer between pH 2 and pH 8 which is highly
sorbed. Apparently, because of a change in the nature of the polymer, sorp-
tion of plutonium decreases above pH 8, reaching a minimum of about pH 12.
In his experiments using monovalent cesium, Rhodes (1957b) observed that
pH had a negligible effect on the uptake of this ion when HC1 was used to
adjust the pH of the solution. However, when NaOH was used to adjust the so-
lution pH, a marked reduction in the distribution coefficient was noted.
Apparently, sodium ions competed with cesium ions for exchange sites. Experi-
mental results with strontium indicated that when the pH was adjusted with
either NaOH or HC1, the distribution coefficient of strontium increased from
about pH 4 to about pH 10. Above pH 10, uptake was reduced by the presence of
large sodium ion concentrations.
Alternative combined forms of radionuclides may have an effect on distri-
bution coefficients and associated relative velocities. For example, Jakubick
(1976) determined that the velocity of Pu02 in soils is about 100 times
faster than the velocity of Pu(1^03)4.
FIELD METHODS FOR MONITORING POLLUTANT
MOVEMENT IN THE VADOSE ZONE
Source monitoring, a primary activity of a monitoring program, is not
covered in this section. Instead, the interested reader is referred to the
comprehensive methodology of Huibregtse and Moser (1976). For convenience,
field monitoring in the vadose zone will be categorized as follows: (1) indi-
rect methods, (2) direct methods for solids sampling, (3) direct methods for
solution sampling in unsaturated media, and (4) direct methods for sampling
saturated regions.
Indirect Methods
Two alternative properties of a wastewater which could be used to esti-
mate pollutant mobility in the vadose zone are temperature and electrical
103
-------
conductivity. Temperature could serve as an indicator of wastewater movement
provided the source were at either an elevated or lowered temperature, and if
the disposal site were underlain by shallow perched groundwater. Insertion of
sensitive thermistors into the soil and groundwater system could possibly mon-
itor the spread of the source. The limitation of this method is that the tem-
perature wave would be dampened by contact with soil and groundwater. That
is, the extent of the plume could be poorly defined.
Resistivity Methods for Soil Salinity—
An indirect property which has been extensively used to characterize soil
salinity and to delineate the area! distribution of shallow pollution plumes
is resistivity, or its inverse property, electrical conductivity. As pointed
out by Rhoades and Halvorson (1977), most soil minerals are insulators. Con-
sequently, electrical conduction in saline soils is mainly through the pore
water containing dissolved electrolytes. Exchangeable cations do not contri-
bute extensively to electrical conduction in nonsodic soils because they are
not present in abundance and they are less mobile than the soluble electro-
lytes. According to Rhoades (1979a), the electrical conductivity of a saline
soil, ECa, depends primarily on the electrical conductivity of the liquid,
ECW; on the volumetric water content, 9; on the tortuousity, T; and on the
extent of surface conductance, ECS. For a given soil, the specific conduc-
tance of a saturation extract, ECe, is uniquely related to ECW. For sim-
plicity, a relationship between ECW and ECa is obtained at a uniform water
content value, 6, in order to standardize 6 and T effects. In practice,
the standard water content is taken to be that at field capacity. The rela-
tionship between ECa and ECe then becomes:
ECa = A ECe + B (34)
where B = ECS
A = slope of the ECa versus ECe line.
The four-electrode method--Rhoades and Halvorson (1977) presented three
methods for determining ECa and three methods for establishing ECe versus
ECa calibrations. One method for measuring soil electrical conductivity in-
situ uses the Wenner four-probe array (see Figure 30). This method is a com-
mon surface resistivity technique used by geophysicists (Zohdy, Eaton, and
Mabey, 1974). Basically, the method entails placing four electrodes into the
soil surface at a constant distance "a" apart. An electrical current is ap-
plied to the outer two electrodes and the resultant electrical resistance is
measured as a potential drop across the inner two electrodes. Resistance is
measured via a resistance meter. The apparent bulk soil conductivity, ECa,
is calculated by the following equation:
ECa ' f^' Tt
where R^ = measured resistance in ohms at temperature t
a = spacing of electrodes
104
-------
RESISTANCE METER
CURRENT
ELECTRODE
C, P| P2 C2
O 9
POTENTIAL
ELECTRODES
SOIL
a
CURRENT
ELECTRODE
C2
AREA OF CONDUCTIVITY MEASUREMENT
Figure 30. Wenner four-probe array and associated area of conductivity
measurement (after Rhoades and Halvorson, 1977).
105
-------
f^ = a factor to adjust the reading of a reference temperature
of 25°C.
ECa is reported in mmhos/cm.
In the absence of layering, the depth of penetration of the electrical
current is about one-third at the outer electrode spacing, y (Rhoades, 1979a).
Similarly, the volume of soil measured by a given "a" spacing in a single ECa
determination is about iry3/g. To increase the volume of soil scanned during a
salinity survey, it is thus a matter of increasing the "a" spacing of the
electrodes. The limitation on "a" spacing, and thus depth, is the degree of
uniformity of texture within the subsoil. A depth of 4 feet appears to be a
reasonable lower limit. Figure 30 depicts the theoretical mass of the soil
scanned by the method.
Rhoades and Halvorson (1977) successfully used the four-probe method to
determine the average bulk soil salinity in soils of the Northern Great
Plains. The four-probe method was also used by Rhoades to detect the presence
of a saline shallow water table by varying the "a" spacing. In two soils ex-
amined by this method, water tables were detected at depths of about 1 meter
(3.3 feet) and 3 meters (9.8 feet). The conductance of the soil solution in
discrete depth intervals can also be obtained by varying the "a" spacing and
using a difference equation presented by Rhoades and Halvorson (1977). They
used this method to delineate salt-affected regions in incremental soil depths
in Montana.
According to Rhoades (1979a), one of the major advantages of the four-
probe method is that because of the high degree of correlation between ECe
and ECa, the method facilitates obtaining representative values of the sa-
linity of typically heterogeneous soils without excessive expenditures of time
and money. Other advantages cited by Rhoades (1979b) are: larger volumes of
the salinity of soil are scanned than when soil samples, suction cups, or
salinity sensors are used; and because the method is fast and simple, it is
particularly well suited for routine salinity monitoring and mapping. A dis-
advantage is that the accuracy of the method decreases when soil layering be-
comes pronounced.
The EC-probe—Rhoades and van Schilfgaarde (1976) developed an "EC-probe"
to detect changes in soil salinity in discrete depth intervals within strati-
fied soils. Figure 31 shows construction features of the unit. The probe
utilizes the Wenner array of four electrodes with a fixed "a" spacing of 2.6
cm (1.02 inches). Salinity can thus be assessed in 15-cm (5.9-inch) depth in-
tervals. The corresponding soil volume scanned with this electrode setting is
about 90 cm3 (5.5 in^). The electrodes in the unit designed by Rhoades
and van Schilfgaarde (1976) were constructed of brass embedded within a lucite
housing. The probe is attached to a shaft with a handle. Wire leads from the
electrodes are brought to the surface through the hollow shaft and handle.
The outer two electrodes are attached to a current source and the inner two
electrodes are attached to a resistance meter.
In operation, a soil cavity is augered to the depth of interest using a
soil auger or sampler of about the same diameter as the probe. The probe is
106
-------
c
k-BLACK ANODIZED ALUMINUM TUBE
BLACK ANODIZED ALUMINUM-
LI 25" ACRYLIC PLASTIC ROD—
1/32" TEFLON
BLACK OXIDE STEEL , BRASS
1° TAPER-ll
BLACK OXIDE STEEL, BRASS
6° TAPER-I
3/ie" x i"STEEL DOWEL PIN
BRASS ELECTRODE
1.125" DELRIN PLASTIC ROD
Figure 31.
Cross section of EC
Halvorson, 1977).
probe (after Rhoades and
placed within the cavity. An electric current is applied to the outer elec-
trodes and the potential drop between the inner electrodes is measured via a
resistance meter. The resistance value is converted to conductance by means
of the cell constant for the probe. The cell constant is determined by sub-
merging the unit in a container filled with solutions of known conductivity.
Individual ECa versus ECe calibrations must be obtained for each of the
soil strata. During operation of the probe in the field, it is important to
measure salinity at a uniform water content value, such as that at field
capacity.
Frequently during salinity studies, it is advantageous to obtain periodic
measurements at the same location. For such cases, Rhoades (1979c) designed
an inexpensive probe which can be left in place within a cavity. "... Im-
planted probes offer certain advantages, like avoiding the need for making new
access holes, remaining in the same position with time, minimizing the
107
-------
complications arising from several access holes being made in the sampling
area " Ostensibly, a number of probes could be clustered within a common
hole.
Four-electrode conductivity cell—The four-electrode conductivity cell is
described in detail by Rhoades et al. (1977). Basically, the cell is con-
structed as follows. Undisturbed soil cores are obtained using a soil-core
sampler with lucite column sections as core inserts. The lucite sections are
removed from the sampler and segmented to form individual cells. Electrodes
are inserted into threaded holes in the cell walls. The ECa of the soil is
subsequently measured using a resistivity meter.
In the unit described by Rhoades et al. (1977), light stainless steel
electrodes are inserted into the soil surface at 45-degree intervals around a
circumference to a fixed depth of 3 mm (0.118 inch). Consequently, any four
neighboring electrodes could be regarded as a Wenner array, the outer two
electrodes being current electrodes and the inner two electrodes being poten-
tial electrodes. By rotating the connectors, eight independent ECa measure-
ments can be made.
A cell constant must be obtained to permit converting resistance values
to conductivity values. In practice, cell constants are obtained by filling
the units with solutions of known conductance and measuring the cell
resistance.
Methods for obtaining ECP versus ECg calibrations—Field measurements
of ECa are converted into corresponding tCe values via calibration rela-
tionships. Rhoades and Halvorson (1977) discussed three calibration methods
in detail. In particular, measurements of conductivity are obtained using the
Wenner four-electrode method, the EC-probe, or the four-electrode cell. Soil
samples are then obtained within the volume of soil measured by the selected
method and laboratory measurements are made to determine the conductivity of
the saturated extract, ECe. A number of soil samples of differing salinity
must be measured in order to obtain a range of ECe versus ECa values. In
addition, the field ECa measurements should be obtained at a uniform water
content, such as that at field capacity.
Earth Resistivity Profiles for Delineating Pollution Plumes—
The electrical resistivity/conductivity method described above is useful
for assessing salinity in the soil zone. The method has also been success-
fully used to monitor salinity in perched groundwater regions of the vadose
zone. Again, the technique is based on the observation that the resistivity
of solute in the pore space of the vadose zone offers less resistivity to cur-
rent flow than dry rocks. By the same token, a highly salinized groundwater
plume offers less resistivity to an applied current than less saline native
groundwater. Consequently, by obtaining a series of resistivity-depth pro-
files in a grid within a waste disposal site, the lateral extent, and possibly
the depth, of the plume could be delineated (Fenn et al., 1977).
Cartwright and his associates observed pollution plumes emanating from
landfills in Illinois using earth resistivity surveying (Cartwright and
108
-------
McComas, 1968; Cartwright and Sherman, 1972). Warner (1969) discussed the use
of the method as a preliminary approach for delineating zones of polluted
groundwater.
The basic approach used for earth resistivity surveys is the same as that
for soil salinity surveys. That is, the Wenner array is used with increasing
"a" spacings in order to observe resistivity changes with depth. According to
Fenn et al. (1977), it is more appropriate to refer to "apparent resistivity"
rather than "resistivity" alone. This term accounts for subsurface heteroge-
neity and is defined as "... the weighted average of the actual resistivities
of the individual subsurface materials or strata within the depth of penetra-
tion at the resistivity measurement" (Fenn et al., 1977).
Two methods are used to interpret the field resistivity data: theoreti-
cal or empirical (Fenn et al., 1977). For the theoretical approach, the field
resistivity values are plotted. "... The resulting curve (is) compared with
sets of master curves developed for numbers of resistivity layers with defi-
nite ratios of resistivity and thickness. By this method, the value of resis-
tivity for each geologic unit as well as its thickness and depth can be
determined." The empirical method entails relating the field resistivity data
to geological information, i.e., the properties of layered sediments. The
success of the empirical approach thus depends greatly on the availability of
geological information, such as texture and the vertical and areal distri-
bution of the layered sediments, and the chemical quality of perched
groundwater.
Stellar and Roux (1975) indicated that the overall success of the resis-
tivity method in delineating plumes depends on the interrelationships of four
factors: (1) the contrast between the conductivities of polluted and natural
groundwater, (2) the depth to the top of the polluted groundwater body, (3)
the thickness of the polluted groundwater body, and (4) lateral variations in
surficial geology. Results of four resistivity surveys were presented by
Stellar and Roux to demonstrate the importance of these interrelationships.
Control data were obtained using geological information from existing wells or
by installing test wells. The case studies included pollution plumes from
industrial waste disposal sites and landfills. The depths to polluted ground-
water ranged from 5 feet to 60 feet below sand surface. The method was
successful in delineating the plume in all except one case. In the one unsuc-
cessful case, it was determined that the following adverse conditions existed:
(1) man-made obstructions were extensive over large portions of the area, (2)
the water table was relatively deep, (3) the difference in conductivity value
of polluted and natural groundwater was minor, and (4) extreme lateral varia-
tions in geology existed above the contaminated plume.
Salinity Sensors—
Another indirect method for in-situ evaluation of soil salinity is the
so-called "salinity sensor." The basis of these devices is the relationship
between specific electrical conductance, ECe, of soil solution and the total
concentration of salts in solution.
109
-------
As described by Richards (1966), the basic idea of the salinity sensor is
that electrodes embedded in porous ceramic, forming a hydraulic continuum with
soil water, can be used to directly measure the specific conductance of the
soil solution. From suitable calibration relations, therefore, specific con-
ductance values can be directly related to the total salt content. The unit
described by Richards comprised a plate about 1-mm (0.039-inch) thick with
platinum electrodes fired in-place on opposing faces. An important feature of
this sensor is that the unit is spring loaded so that good contact with soil
could be ensured.
Because of the strong dependency of specific conductance on temperature,
it is important to accurately measure the temperature of the soil solution.
Richards used a thermistor to provide temperature compensation in his unit.
Oster and Willardson (1971) reviewed problems arising from calibration of
sensors. They also reported on field studies using the unit. Of particular
importance was their observation that sensors should not be used at soil-water
pressures less than -2 atmospheres. Also, they indicated that when sensors
are placed in trenches at field sites, the permeability of the materials in
the backfilled trench tends to be greater than in indigenous soil. During
leaching trials, therefore, the salinity measured with the sensors tended to
be lower than in adjacent soil. Differences were attributed to greater leach-
ing in the trench. However, structural differences were also present, i.e.,
the pore sizes in the trench soil were probably of a different range than in-
digenous soil.
Rhoades (1979b) indicated that a principal advantage of using salinity
sensors compared to soil sampling for assessing salinity is that readings are
taken in the same location. Thus, by installing a network of sensors, a chro-
nological record of in-situ salinity changes can be obtained from a large vol-
ume of soil. Rhoades also indicated that salinity sensors "... are simple,
easily read, and sufficiently accurate for salinity monitoring purposes."
The U.S. Salinity Laboratory Staff (1977) discussed the use of a data ac-
quisition system which automatically obtains electrical conductivity values
from a number of salinity sensors. The units were installed during field
studies on salt movement in irrigation return flow. By means of a computer
program, the system activated stepping switches, read the sensors in sequence,
adjusted the readings for soil temperature, and printed out the calculated EC
values. The resultant data were transmitted by means of a teletype to River-
side, California.
Rhoades (1979b) compared the use of salinity sensors and four-probe units
for assessing soil salinity. He concluded that four-probe units have the fol-
lowing six major advantages over the salinity sensor: (1) they are less sub-
ject to calibration change, (2) they are more durable and less costly, (3)
they do not suffer from time lags in the response to changing salinity, (4)
they can be sized to measure within different soil volumes, (5) they are more
versatile, and (6) they are well suited for mapping and diagnosis, as well as
monitoring.
110
-------
Solids Sampling in the Vadose Zone
An important aspect of a monitoring program for the vadose zone underly-
ing a waste disposal operation is to obtain samples of the solid materials
throughout this region. Solids are defined as soils and underlying geologic
materials. The rationale for solids sampling in the vadose zone was stated by
Dunlap et al. (1977):
Only by analysis of earth solids from the unsaturated zone un-
derlying pollutant-releasing activities can those pollutants
which are moving very slowly toward the water table because of
sorption and/or physical impediment be detected and their rates
of movement and degradation measured. Such pollutants, which
probably include a major proportion of organics and microorga-
nisms, are not likely to be detected in groundwater until the
activities releasing them have been in operation for protracted
periods. Because of their potential for long-term pollution of
groundwater, it is imperative that the behavior of these pol-
lutants in the subsurface be established at the earliest prac-
ticable time.
Dunlap et al. (1977) also listed reasons for sampling from the groundwa-
ter zone ("zone of saturation"). These reasons apply equally well for sampl-
ing from perched groundwater:
Analyses of organic pollutants in solid samples from the zone
of saturation are needed for a realistic evaluation of the to-
tal extent and probable longevity of organic pollution in an
aquifer. Such analyses provide a measure of the quantity of
pollutants which are sorbed on aquifer solids and which are in
equilibrium with, and in essence serve as a reservoir for, pol-
lutants in solution in the adjacent groundwater.
Microbial populations which may be involved in the biologfcal
alteration of pollutants in subsurface formations are likely to
be in such close association with subsurface solids that they
will not be present in well waters in numbers which are quan-
titatively indicative of their presence in the formations;
hence, analyses of subsurface solids are needed for accurate
evaluation of such populations.
Even when the best well construction and groundwater sampling
procedures are used, it is difficult to completely eliminate
the possibility that contaminating surface microbes may be
present in groundwater samples. Solids taken from the interior
of cores carefully obtained from the zone of saturation proba-
bly provide the most authentic samples of aquifer microorga-
nisms that can be obtained.
In this section, the following items are discussed: (1) soils sampling
methods for inorganic chemicals, organic constituents, and microorganisms; (2)
sampling of solids in the lower vadose zone; (3) field analyses; (4) field and
111
-------
laboratory handling of samples; and (5) advantages and disadvantages of solids
sampling methods.
Soils Sampling--
Soil sampling methods used by soil scientists and irrigation and drainage
engineers to evaluate the physical properties of soils are also suitable for
sampling to determine inorganic chemical constituents. Particular methods
which have been used include the following types.of hand augers: screw-type
augers, post-hole augers, barrel augers, and dutch augers (Donnan, 1957).
Another common type of sampler is the split-spoon sampler, which is a barrel-
type auger, one side of which pivots on a hinge. Tube-type samplers have been
used extensively to obtain soil cores. The Veihmeyer sampler is an example of
such a unit. The end of the sampler is bevelled and sharpened to facilitate
insertion into the soil. A drive hammer is used to force the tube to the de-
sired depth. An intact core is removed from the tube. Such a core may show
the distinct breaks in soil layering.
A problem associated with using augers and tube-type samplers is that if
the soil is very dry, the sample tends to fall out of the unit when it is be-
ing withdrawn. "Core catchers" could be used to overcome this problem
(Thomas, personal communication, 1979). Another problem is that if the soil
is wet, some chemical interactions may occur between the soil water and the
metal parts of the sampler. Thus, trace metals could be introduced into the
sample.
The problem of contamination between a soil sample and sampling device
is of particular importance when sampling for organic chemicals and microorga-
nisms. Bordner, Winter, and Scarpino (1978) recommended the following proce-
dure for obtaining soil samples for microbial analysis:
... scrape the top one inch of soil from a square foot area
using a sterile scoop or spoon. If a subsurface sample is de-
sired, use a sterile scoop or spatula to remove the top surface
of one inch or more from a one-foot square area. Use a second
scoop or spoon to take the sample. Place samplings in a ster-
ile one quart screw-cap bottle until it is full. Depending on
the amount of moisture, a one-quart bottle holds 300-800 grams
of soil. Label and tag the bottle carefully and store at 4°C
until analyzed.
Techniques for sampling organic and microbial constituents have been ex-
amined by The Groundwater Research Branch of the Robert S. Kerr Environmental
Research Laboratory, as reported by Dunlap et al. (1977). One method used
with success comprises an auger and dry-tube coring procedure. The method en-
tails augering a hole to the top of the desired sampling depth and forcing a
"dry-tube" core sampler into the sampling region. The core sampler used by
Dunlap et al. (1977) was a steel tube about 18 inches long and 3 inches in
diameter. The core barrel is fitted with a steel drive shoe which is of
slightly smaller inside diameter than the diameter of the barrel. This ar-
rangement facilitates removing the core. As pointed out by Dunlap et al.
(1977), contamination problems are minimized by the auger/dry-tube coring
112
-------
technique primarily because drilling fluid is not used. It would appear that
the method could be modified to include a lucite plastic insert in the core
barrel to minimize contaminating soil samples obtained for trace metal
analysis.
Sampling from the Lower Vadose Zone—
The methods discussed above for obtaining soil samples could, with diffi-
culty, be used to sample from the lower vadose zone. For example, Fenn et al.
(1977) indicated that post-hole augers could be used to sample to depths down
to 80 feet in certain material by using a tripod and pulley. A more conve-
nient faster approach is to employ engine-driven augers and drilling equip-
ment. Among the power-driven sampling units available are continuous-flight
spiral augers (which could be hollow-stemmed), core samplers, bucket augers,
cable-tool drill rigs, and rotary drill rigs. The spiral-type units are not
particularly suitable for sampling because it is not always possible to dis-
tinguish the sampling depth. By using the hollow-stem auger, however, it is
possible to insert a core sampler inside the auger and sample from discrete
depths. Soil sampling machines are available commercially for obtaining
either auger samples or core samples by attaching appropriate tools. These
machines are mounted on trailers with independent engines. Alternatively,
they may be purchased for mounting on pick-up trucks and driven by power take-
offs. Hydraulic controls facilitate drilling or coring operations. Travel-
ling Kelley bars permit sampling at depths greater than 25 feet (courtesy,
Giddings Machine Company).
A bucket auger consists of a large-diameter bucket fitted with cutting
blades (Fenn et al., 1977). The bucket is rotated in the hole until filled.
When brought to the surface, the bucket is dumped and a sample can be easily
obtained. Alternatively, a core is obtained prior to dumping. Rible et al.
(1976) used this method to sample for nitrate distribution in the vadoze zone
underlying irrigated fields in California.
Drill rigs could be used for sampling purposes at the same time that test
wells and monitor wells are being installed at the waste disposal site. The
cable-tool, or percussion, technique entails alternately raising and dropping
a heavy string of tools in the cavity. The drill bit crushes large-diameter
material. Water is commonly added to the borehole and, after removing the
tool string, a bailer is inserted. The fine cuttings are removed by the
bailer. Casing is driven in as drilling progresses. An obvious disadvantage
of this method is that the water added during drilling changes the chemistry
of the native pore water. This problem may be obviated somewhat by analyzing
the water and adjusting the analyses obtained from solids samples.
The hydraulic rotary method utilizes a rotating bit to excavate a bore-
hole and a continuous flow of water to remove drill cutting. Drillers' mud
is added to the circulating water during drilling to hold open the hole. Ob-
viously, solids samples would be badly contaminated by the mud/water mixture
used with this method. Also, the precise depths at which cuttings are ob-
tained are impossible to define.
113
-------
In the air rotary method, air is used instead of water to bring the drill
cuttings to the surface. Thus, the problem of contamination encountered with
the regular hydraulic rotary drilling method is not a factor in air drilling.
There may be difficulty, however, in defining the sampling depth precisely,
i.e., particle-size segregation may occur because of density differences.
The core sampling method of Dunlap et al. (1977) has been used to obtain
solids samples for analyses of organic chemicals and microorganisms in depths
to about 25 feet. In addition to dry samples, solids have also been obtained
from saturated regions. Problems developed, however, when sampling loose sat-
urated material. An alternative approach to overcome such problems which was
considered by Dunlap et al. was to use a hoi low-stem auger in lieu of a regu-
lar auger. Dunlap et al. also described a piston-type sampler for collecting
solids for organic analyses when fluid is used in the drilling process. To
date, the sampler has not been extensively field tested.
Field Analysis of Pore Water-
Whenever possible, pore water should be extracted from solid samples in
the field and analyzed immediately for unstable constituents such as tempera-
ture, pH, eH, and EC (Mooij and Rovers, 1976). When the soils are unsatu-
rated, special techniques are required to obtain samples of the pore water.
Fenn et al. (1977) reviewed a number of alternative methods for extracting
pore water from core samples, including a method by Lusczynski (1961) which
uses CC>2 under pressure to displace the fluid. Another method described by
Fenn et al. is a "hydraulic squeezer," which uses a pressurized piston to
force pore water into a syringe.
Instead of using pore water directly, it is often more convenient to pre-
pare saturated pastes in the field and extract water for specific conductance
measurements. Rhoades and Halvorson (1977) described a convenient extraction
unit for field use. Saturated paste is transferred into a small-diameter fil-
ter funnel made of plastic with a filter pipe in place. Vacuum is provided
by a hand pump and the sample is collected in a reservoir. The EC of the sam-
ple is determined via a portable EC bridge.
Field and Laboratory Preparation of Samples—
Dry solids samples collected for chemical analyses of pore water in the
laboratory should be stored in paper bags. If the samples are wet, it will be
necessary to store the samples in plastic bags. Depending on the intended
analyses, the samples are air-dried in the laboratory. If the analyses are
intended for the nitrogen series, the atmosphere in the handling room should
be free of ammonia.
Samples collected for determining microorganisms and organic pollutants
using the "dry-tube" core sampler require special handling. Dunlap et al.
(1977) described the procedure in detail. Briefly, the core is pressed from
the barrel of the sampler and the first 5 to 8 cm (2 to 3.1 inches) of solids
is removed for analysis of physical/chemical parameters. In order to obviate
microbial contamination of the core sample by contact with the walls of the
sampler, a subs ample is removed from the center of the core (see Figure 32).
114
-------
o
n
•MICROBIAL SAMPLE
1.3 X 15.2 CM
ORGANIC SAMPLE
7.6 X 10.2 CM
MICROBIAL SAMPLE
1.3 X 15.2 CM
-ORGANIC SAMPLE
7.6 X 10.2 CM
Figure 32.
Subs amp! es from a "dry tube" core sampler
(after Dunlap et al., 1977).
115
-------
This is effected by inserting a sterile stainless steel tube into the center
of the core. The subsample is extracted using a sterile rod and placed in
containers. Subsamples for analyses of adenoside triphosphate (ATP) are ob-
tained in a similar manner, i.e., subsampling from the center of the core.
Two 5-cm (2-inch) long subsamples for organic analyses are obtained after
removing the central cores. The subsamples are dropped into aluminum baking
pans and covered with aluminum foil. The pan and subsamples are then placed
into an insulated polystyrene box with liquid nitrogen to quick-freeze the
sample.
Problems With Solids Sampling--
An important limitation of field sampling of solids is that an inordinate
number of samples may be required to ensure that the measured mean value of a
particular parameter falls within a given range of a true value. A principal
reason for the problem is that variations arise because of the natural spatial
variability of soil properties. Such variations were noted by Rible et al.
(1976); Pratt, Warneke, and Nash (1976); and Lund and Pratt (1977) during
studies on nitrate leaching in irrigated fields in California. Pratt, War-
neke, and Nash (1976), for example, showed that five to 21 sampling sites were
needed (eight samples per site) from four fields ranging from 62 to 223 acres
to obtain measured mean nitrate concentrations that were within 30 percent of
the true mean at the 95-percent confidence limit. They also estimated that it
would have been necessary to sample 23 to 161 sites per field to be within 10
percent of the true mean.
In summarizing the utility of solids sampling for nitrate leaching from
irrigated fields, Lund and Pratt (1977) stated:
Monitoring nitrate leaching by soil sampling does not appear
to be feasible for basin-wide studies. As a research tool to
study the components of the nitrogen cycle it has proved use-
ful, but we would not recommend it as a monitoring method.
Rhoades (1979b) described another problem with solids sampling: "Some
changes in soil-water composition occur as soil is removed from its natural
condition, dried, ground, sieved, extracted, etc., hence only relative changes
in composition, not absolute, can be determined from soil sampling."
Solution Sampling in Unsaturated Media
Wells and open cavities cannot be used to collect solution flowing in the
vadose zone under suction (negative pressures). The sampling devices for such
unsaturated media are thus called suction samplers. Three types will be dis-
cussed: (1) ceramic-type samplers, (2) hollow fiber samplers, and (3) mem-
brane filter samplers.
Ceramic-Type Samplers—
There are two types of samplers constructed from ceramic materials: the
suction cup and the filter candle. Both types operate in the same manner.
116
-------
Basically, ceramic-type samplers (also called suction "lysimeters") comprise
the same type of ceramic cups used in tensiometers. When placed in the soil,
the pores in these cups become an extension of the pore space of the soil.
Consequently, the water content of-the soil and cup become equilibrated at the
existing soil-water pressure. By applying a vacuum to the interior of the cup
such that the pressure is slightly less inside the cups than in the soil solu-
tion, flow occurs into the cup. By an appropriate technique, the sample is
pumped to the surface, permitting laboratory determination of the quality of
the soil solution in-situ. Although cups have limitations, at the present
time they appear to be the best tool available for sampling unsaturated media,
particularly in the field.
Suction cups may be subdivided into three categories: (1) vacuum op-
erated soil-water samplers, (2) vacuum-pressure samplers, and (3) vacuum-
pressure samplers with check valves. The soil-water samplers generally
consist of a ceramic cup mounted on the end of a small-diameter PVC tube, sim-
ilar to a tensiometer (see Figure 33). The upper end of the PVC tubing pro-
jects above the soil surface. A rubber stopper and outlet tubing are inserted
PLASTIC TUBE
VACUUM TEST HAND PUMP
VACUUM
COLLECTED SOIL WATER SAMPLE
Figure 33. Soil-water sampler (courtesy, Soil Moisture
Equipment Corporation, 1978).
117
-------
into the upper end. Vacuum is applied to the system and soil water moves into
the cup. To extract a sample, a small-diameter tube is inserted within the
outlet tubing and extended to the base of the cup. The small-diameter tubing
is connected to a sample-collection flask. A vacuum is applied via a hand
vacuum-pressure pump and the sample is sucked into the collection flask.
These units are generally used to sample to depths up to 6 feet from the land
surface. Consequently, they are used primarily to monitor the near-surface
movement of pollutants from land disposal facilities or from irrigation return
flow.
To extract samples from depths greater than the suction lift of water, a
second type of unit is available, the so-called vacuum-pressure lysimeter.
These units were developed by Parizek and Lane (1970) for sampling the deep
movement of pollutants from a land disposal project in Pennsylvania. The de-
sign of the Parizek and Lane sampler is shown on Figure 34. The body tube of
the unit is about 2-feet long, holding about 1 liter (0.26 gallon) of sample.
,2-WAY PUMP
PLASTIC TUBE
AND CLAMP
VACUUM PORT
AND GAUGE
/IK
PLASTIC TUBE
AND CLAMP
TAPE
PRESSURE
VACUUM IN
BENTONITE
3/16-INCH
COPPER TUBE
PLASTIC PIPE
24 INCHES LONG
6-INCH HOLE
WITH TAMPED
SILICA SAND TST
BACKFILL \J
POROUS CUP
BENTONITE
SAMPLE BOTTLE
DISCHARGE TUBE
Figure 34. Vacuum-pressure sampler (after Parizek and Lane, 1970).
118
-------
Two copper lines are forced through a two-hole rubber stopper sealed into a
body tube. One copper line extends to the base of the ceramic cup as shown
and the other terminates a short distance below the rubber stopper. The
larger line connects to a sample bottle and the shorter line connects to a
vacuum-pressure pump. Care must be exercised to ensure that all lines and
connections are sealed.
In operation, a vacuum is applied to the system (the longer tube to the
sample bottle is clamped shut at this time). When sufficient time has been
allowed for the unit to fill with solution, the vacuum is released and the
clamp on the outlet line is opened. Air pressure is then applied to the sys-
tem forcing sample into the collection flask. A basic problem with this unit
is that when air pressure is applied, some of the solution in the cup may be
forced back through the cup into the surrounding pore water system. Conse-
quently, this type of pressure-vacuum system is recommended for depths only
up to about 50 feet below land surface. In addition to the monitoring effort
of Parizek and Lane, these units were used by Apgar and Langmuir (1971) to
sample leachate movement in the vadose zone underlying a sanitary landfill.
Wood (1973) reported on a modified version of the design of Perizek and
Lane. Wood's design is the third suction sampler discussed in this section.
The design of Wood overcomes the main problem of the simple pressure-vacuum
system; namely, that solution is forced out of the cup during application of
pressure. A sketch of the sampler is shown on Figure 35. Basically, the cup
ensemble is divided into lower and upper chambers. The two chambers are iso-
lated except for a connecting tubing with a check valve. A sample delivery
tube extends from the base of the upper chamber to the surface. This tube
also contains a check valve. A second shorter tube terminating at the top of
the sampler is used to deliver vacuum or pressure. In operation, when a vac-
uum is applied to the system, negative pressure extends to the cup through the
open one-way check valve. The second check valve in the delivery tube is
shut. The sample is delivered into the upper chamber, which is about 1 liter
(0.26 gallon) in capacity. To deliver the sample to the surface, the vacuum
is released and pressure (generally of nitrogen gas) is applied to the shorter
tube. The one-way valve to the cup is shut and the one-way valve in the de-
livery tube is opened. Sample is then forced to the surface. High pressures
can be applied with this unit without danger of damaging the cup. Conse-
quently, this sampler can be used to depths of about 150 feet below land sur-
face (Soil Moisture Equipment Corporation, 1978). Wood and Signer (1975) used
this sampler to examine geochemical changes in water during flow in the vadose
zone underlying recharge basins in Texas.
A sampling unit employing a filter candle was described by Duke and Haise
(1973). The unit, described as a "vacuum extractor," is installed below plant
roots. Figure 36 shows an illustrative installation. The unit consists of a
galvanized sheet metal trough open at the top. A porous ceramic candle (12
inches long and 1.27 inches in diameter) is placed into the base of the
trough. A plastic pipe sealed into one end of the candle is connected to a
sample bottle located in a nearby manhole or trench. A small-diameter tube
attached to the other end of the candle is used to rewet the candle as neces-
sary. The trough is filled with soil and placed within a horizontal cavity of
the same dimensions as the trough. The trough and enclosed filter candle are
119
-------
U
1
-VACUUM-AIR PRESSURE LINE
•UPPER CHECK VALVE
-SAMPLE DISCHARGE LINE
,UPPER CHAMBER
-LOWER CHECK VALVE
-TUBING
-LOWER CHAMBER
-SUCTION CUP
Figure 35. "Hi/Pressure-vacuum soil-water sampler" (courtesy,
Soil Moisture Equipment Corporation, 1978).
120
-------
pressed up against the soil via an air pillow or mechanical jack. In opera-
tion, vacuum is applied to the system such that soil water is induced to flow
into the trough and candle at the same rate as in the surrounding soil. The
amount of vacuum is determined from tensiometers. Hoffman et al. (1978) used
this type of sampler to collect samples of irrigation water leaching beneath
the roots of orange trees during return flow studies at Tacna, Arizona (see
Figure 36).
CROSS SECTION
A—A
ADJUSTABLE
VACUUM
SOIL
SOLUTION
DUAL CHAMBE
TRICKLE TUBING
AIR PILLOW
UNDISTURBED
SOIL
SAMPLING.
BOTTLE
SHEET METAL
TROUGH
I m
Figure 36. Facilities for sampling irrigation return flow via filter
candles, for research project at Tacna, Arizona (after
Hoffman et al., 1978).
Sampler preparation, installation, and operation—Laboratory studies on
ceramic cups by Wolff (1967) showed that new cups yield several milligrams per
liter of Ca, Mg, Na, HC03, and SiO^. Consequently, it is recommended that
cups be leached prior to installation. In his studies, Wood (1973) cleaned
new cups as follows: "... the cups were cleaned by letting approximately 1
liter of 8N HC1 seep through them. This acid treatment was followed by allow-
ing 15 to 20 liters of distilled water to seep through and rinse the cups
thoroughly. The cups are adequately rinsed when there is less than a 2% dif-
ference between the specific conductance of the distilled water input and the
output from the cup."
Installation of the shallow vacuum-type samplers is similar to the in-
stallation of tensiometers. A cavity is first augered to the desired depth.
Soil from the base of the hole is wetted to form a slurry (Soil Moisture
121
-------
Equipment Corporation, 1978). The next step is to place the sampler into the
hole, forcing the cup into the slurry. The entire cup should be covered by
the slurry. The hole is then backfilled with soil, tamping to ensure that
side leakage is prevented. An alternative installation method entails placing
bentonite clay at the base of the open hole and placing a small amount of 200-
mesh silica sand above the bentonite. The sampler is placed in the hole and a
6-inch layer of sand is added above the cup. Bentonite clay is then placed on
top of the sand as a seal against side leakage of water. Finally, the hole is
filled with native soil.
The pressure-vacuum type lysimeters, either with or without the check
valves, may be installed singly in a cavity or clustered with several units in
a cavity. Installation of the single units entails using one of the methods
discussed above for the shallow vacuum-type units. An example of a cluster
design is shown on Figure 37. In the illustration, two units are installed in
a hole. The base of the hole should be sealed with bentonite. A layer of
silica sand is then placed on top of the bentonite plug. The lowermost cup is
placed on the silica sand and subsequently surrounded by more sand. Native
soil is backfilled in the cavity to an elevation near that of the next sam-
pler. Bentonite is added as a plug. Silica sand is then added, the second
unit is installed, and backfilling proceeds as above. In some cases, it may
be a good precaution to add a layer of bentonite above the final silica sand
layer to minimize the possibility of side leakage.
The above-ground assembly comprises a pressure-vacuum system and collec-
tion bottles. For a system with several samplers, a manifold assembly is ad-
vantageous. Generally, pressure should be added with a gas such as nitrogen
to preclude oxidation of chemical constituents.
For the shallow soil solution samplers, vacuum could be applied either as
a constant vacuum or a falling vacuum. A constant vacuum is maintained using
a vacuum pump with a preset vacuum. In the falling vacuum type, an initial
vacuum is applied to the system and vacuum gradually decreases as the sample
is drawn into the cups (Hansen and Harris, 1975). Problems may arise in the
operation of the constant vacuum system because of differences in sample col-
lection rate. In particular, some units fill more rapidly than others such
that solution may overflow into adjoining collection flasks. Thus, cross-
contamination may occur. Chow (1977) presented a design of a mercury-pressure
control device to shut off the vacuum to a flask when filled to a preset
level.
Vacuum extractor assemblies comprising sheet metal troughs and filter
candles could be installed by a technique used by the U.S. Salinity Laboratory
Staff (1977) during their irrigation return flow studies at Tacna, Arizona.
In particular, the extractors were installed in a rectangular tunnel formed by
first augering a horizontal hole at the desired depth. Subsequently, a rec-
tangular shaper was forced into the tunnel. The metal troughs containing
filter candles were filled with soil augered from the hole. The extractor as-
sembly was then forced against the ceiling of the tunnel by inflating a Butyl
rubber air pillow. Two tensiometers were installed over each extractor to fa-
cilitate determining the vacuum to the filter candles.
122
-------
/
PRESSURE EVACUATION ACCESS
6
MIN
TWO
g_
||
/'" .-jl ^jL DISCHARGE 1
IL/" VACUUM/PRESSURE
^^ BOTTLE
&%&Ji~j\
" HOLE Jj
IMUM FOR I
LYSI METERS |
1
I
t
u
'1
^
1
I
III
0-
t
J
i
i
«U
i
X*
*Y frnl DISCHARGE 2
• -~
v_x
J
-rr-^^P7—
m=ni NATIVE BACKFILL
1 «[ BENTONITE PLUG
Ti
g. NATIVE BACKFILL
J
•i]] —
H
| SILICA FLOUR
1
| NATIVE BACKFILL
Sir BENTONITE PLUG
ill
jr NATIVE BACKFILL
g (NO LARGE FRAGMENTS
I OR ROCKS)
T
[
-• ii"
f SILICA FLOUR
[
Figure 37. Clustered suction cup lysimeters in a common borehole
(after Hounslow et al., 1978).
123
-------
Problems with ceramic-type samplers—Although the ceramic-type suction
samplers are the best available technique for obtaining pore water samples in-
situ, certain problems in their usage should be noted. One problem is that
samples cannot be obtained over the entire range of soil-water pressures.
That is, a ceramic unit is capable of sampling only to negative soil-water
pressures equal to the air-entry value of the unit. Generally, this is about
-1.0 bar, although units with air-entry values down to -2.0 bars are
available.
In addition to the limitation imposed by the air-entry value of ceramic
samplers, the representativeness of the sample collected is also questionable.
In other words, how closely does the chemical composition of the sample repre-
sent that of pore water? Hansen and Harris (1975), in controlled laboratory
and field tests using ceramic cup samplers, examined the effect of five fac-
tors on sample concentration during extraction: cup intake rate, leaching,
diffusion, sorption, and screening. Screening, or "salt sieving," refers to a
decrease in overall concentration of solutes during flow through a membrane.
Leaching and diffusion were found to be of minor importance. Using phosporus
and nitrate-N as indicators of sorption, they found that nitrate-N was not
sorbed by ceramic cups. This result is in agreement with earlier studies by
Wagner (1962) who noted that N03-N was not sorbed by cups, although sorption
of Nfy-N did occur. Hansen and Harris (1975) found evidence that P was
strongly sorbed. However, they concluded that sorption is serious only when
"... the ceramic sorptive capacity is significantly greater than that of the
adjacent soil."
The intake rate had a profound effect on sample concentration in a system
with varying solute concentration. Thus, fast-rate samplers collected the
largest fraction of their sample at the beginning of a sampling interval. In-
asmuch as the concentration of solute in the pore water system was changing
rapidly, the concentration of sample collected by the sampler was not truly
representative of water draining to the water table.
The fact that ceramic-type samplers extract pore water under low suctions
also affects the representativeness of the sample. England (1974) described
two of the causative factors:
First, numerous experimental and theoretical studies have
clearly shown that the concentration and the composition of the
soil solution are not homogeneous throughout its mass. Ca-
tions, in particular, vary widely in degree of dissociation
from the surface of electronegative colloidal soil particles.
Thus, water drained from large pores at low suctions may have a
chemical 'quality' that is very different from that extracted
from micropores. A point source of suction, such as the porous
cup, samples roughly a sphere, draining different-sized pores
as functions of distance from the point, the amount of applied
suction, the hydraulic conductivity of the medium, and the
soil-water content.
Also, the concentrations of various ions in a soil solution
generally do not vary inversely with the soil-water content.
124
-------
Reitemeier ... and many others have presented evidence that the
total dissolved quantities of some ions increase on dilution
while concurrently, those of other ions may decrease.
In light of these effects, Hansen and Harris (1975) concluded that each
group of pore sizes has a unique volume of drainable water, ion concentration,
and drainage-rate curve. They concluded that in order to collect a soil-water
sample with an ion concentration representative of that draining to the water
table, the rate of sample collection (and thus the incremental volumes) should
correspond to the pore water drainage rate.
The problem of representativeness also arises when ceramic units are used
for solution sampling in highly structured media. As noted by Shaffer, Frit-
ton, and Baker (1979), tfte bulk of the flow occurs through the interped
cracks. Consequently, units which sample pore water from within soil blocks
do not sample from the bulk of the water passing into the lower vadose zone.
In addition to well-structured soils, this physical situation could occur
within spoil piles of coal strip-mined areas.
Field studies by Biggar and Nielsen (1976) illustrated a limitation of
point samplers such as ceramic samplers (and fiber type samplers) for estimat-
ing the mass flux of solute beneath a given soil depth (mass flux of solute is
the mass of solute crossing a unit area per unit time). In particular, be-
cause of the spatial variabil ity of soil properties, the soil water velocity,
and thus soil water flux, may not be normally distributed. In fact, for their
test area, Biggar and Nielsen found velocity to be logarithmically normally
distributed. Consequently, substantial errors could be made in estimating the
mass flux of solute beneath a given soil depth by multiplying average values
of the flux of water by average values of the concentration of the soil solu-
tion. "Hence, we expect point measurement such as the discharge from drain
tiles, solution samples from suction probes and piezometer wells, and exca-
vated soil samples to provide good indications of relative changes in the
amount of solute sufficiently precise to ascertain the amounts of fertilizers
or other solutes leached beyond the recall of roots. Only in case of a thor-
ough analysis of the frequency distribution of such measurements would quanti-
tative results be assured."
Cellulose-Acetate Hollow Fiber Samplers--
Jackson, Brinkley, and Bondietti (1976) described a suction sampler con-
structed of cellulose-acetate hollow fibers. These semi permeable fibers have
been used for dialysis of aqueous solutions, functioning as molecular sieves.
Soil column studies using a bundle of fibers to extract soil solution showed
that the fibers were sufficiently permeable to permit rapid extraction of so-
lution for analysis. Soil solution was extracted at soil-water contents rang-
ing from 50 to 20 percent.
Levin and Jackson (1977) compared ceramic cup samplers and hollow fiber
samplers for collecting soil solution samples from intact soil cores. Their
conclusion was: "... porous cup lysimeters and hollow fibers are viable ex-
traction devices for obtaining soil solution samples for determining EC, Ca,
Mg, and P04-P. Their suitabil ity for N03-N is questionnable." They also
125
-------
concluded that hollow fiber samplers are more suited to laboratory studies,
where ceramic samplers are more useful for field sampling.
Membrane Filter Samplers--
Stevenson (1978) presented the design of a suction sampler using a mem-
brane filter and a glass fiber prefilter mounted in a "Swinnex" type filter
holder. The construction of the unit is shown on Figure 38. The membrane
filters are composed of polycarbonate of cellulose-acetate. The "Swinnex"
filter holders are manufactured by the Millipore Corporation for filtration of
fluids delivered by syringe. A flexible tube is attached to the filter holder
to permit applying a vacuum to the system and for delivering the sample to a
bottle.
SOIL
SAMPLING TUBE
FILTER SUPPORT/BASE
"SWINNEX"
FILTER HOLDER
MEMBRANE FILTER
GLASS FIBER PREFILTER
LASS FIBER "WICK
.'-•-GLASS
i FIBER
COLLECTOR
/ffff//f//////TI///f/t}~/J/7/t
SOIL
Figure 38. Membrane filter sampler (after Stevenson, 1978).
The sampler is placed in a hole dug to a selected depth. Sheets of glass
fiber "collectors" are placed in the bottom of the hole. Next, two or three
smaller glass fiber "wick" discs that fit within the filter holder are placed
in the hole. Subsequently, the filter holder is placed in the hole with the
126
-------
glass fiber prefilter in the holder contacting the "wick" discs. The hole is
then backfilled.
In operation, soil water is drawn into the collector system by capillar-
ity. Subsequently, water flows in the collector sheets toward the glass fiber
wicks as a result of the suction applied to filter holder assembly. The glass
fiber prefilter minimizes clogging of the membrane filter by fine material in
the soil solution.
During field tests with the sampler, it was observed that sampling rates
were a function of soil-water content. That is, sampling rates decreased with
decreasing soil-water content. The "wick and collector" system provided con-
tact with a relatively large area of the soil and a favorable sampling rate
was maintained even when the "collector" became blocked with fine soil. The
basic sampling unit could be used to depths of 4 meters.
Water Sampling from Saturated Regions of the Vadose Zone
Saturated regions of the vadose zone which could be monitored for water
quality mainly comprise perched groundwater occurring at the interface of re-
gions of varying texture. Perched groundwater bodies generally vary greatly
in area! extent and may be either ephemeral or exist for long periods of time.
For convenience, perched groundwater will be arbitrarily subdivided into the
following two categories: (1) shallow perched groundwater within 30 feet of
the surface, and (2) deeper perched groundwater at depths greater than 30 feet
of the surface.
Inasmuch as perched groundwater is in essence saturated groundwater, many
of the techniques apply to monitoring in perched groundwater (e.g., Todd et
al., 1976). A guiding principle for selecting monitoring methods for both the
saturated zone and perched groundwater was aptly stated by Vanhof, Weyer, and
Whitaker (1979):
... for an efficient, long-term operation of an operational
monitoring network, the devices to be used must be simple, rug-
ged, and foolproof. The inside diameter of the installed
structure should allow for the insertion of down-hole tools for
development, pumping, groundwater sampling, hydraulic head mea-
surement, and completion zone maintenance. The installed
equipment should be simple enough to be used by trained but not
educationally skilled personnel.
Shallow Perched Groundwater--
Six alternative methods for obtaining groundwater samples from shallow
perched groundwater will be discussed: (1) tile lines, (2) collection pans
and manifolds, (3) wells, (4) piezometers, (5) multilevel samplers, and (6)
groundwater profile samplers. Following a discussion of each of these meth-
ods, alternative materials for well casing are reviewed and sample extraction
methods are outlined.
127
-------
Sampling Tile Drain Outflow--
If a tile drainage system has been installed to control the elevation of
a perched groundwater table, samples could be collected from the tile outfall.
In some cases, it may be desirable to install commercially available composite
or discrete sampling devices.
Willardson, Meek, and Huber (1973) discussed a "flow-path groundwater
sampler" which enables collection of water in different flow paths around a
tile.
Collection Pans and Manifolds--
Collection pans and manifolds could be installed in regions where tempo-
rary saturation lenses develop during the percolation of water for overlying
surface sources. Generally, pans are installed to permit sampling pore water
through a vertical profile, usually several meters in thickness. In construc-
tion, pan collectors are similar to the vacuum extractors except that a ce-
ramic filter candle is not included. Thus, the media in the trough must be
saturated.
Commonly, the collector pans are connected to a central chamber or trench
containing sample bottles and accessories. McMichael and McKee (1966) used a
large-diameter culvert as the central chamber in their studies on wastewater
reclamation at Whittier Narrows. Parizek and Lane (1970) used a trench to
collect samples from pans in a forest soil during effluent irrigation studies
in Pennsylvania.
The collectors developed by McMichael and McKee (1966) employed pan sam-
plers for the collection of water during periods of saturation in a profile to
about 10 feet. These samplers were conical in shape, 24 inches in diameter
and 9 inches deep. The samplers were installed to various depths at a radial
distance of about 10 feet from the central well. Individual sampling pans
were filled with gravel to prohibit clogging. Each sampler was connected to
the central culvert via tubing. Excavations used for installation of the sam-
plers were carefully backfilled. During sampling, percolate intercepted by
the pans flowed by gravity to the central well into collection flasks.
The trench lysimeter described by Parizek and Lane (1970) used pan lysim-
eters. The largest such lysimeter consisted of a 47-inch wide by 13-foot long
trench excavated to about 20 feet. The sides of the trench were lined with
wood and braced. Sampling pans were installed at 12-inch intervals to a depth
of about 20 feet. The pans, constructed from galvanized metal, were 12 inches
by 18 inches with copper tubing soldered into a raised end to permit sample
drainage. Each unit was driven into the soil a short distance from the side
walls of the lysimeter. While applying effluent during land treatment stud-
ies, samples of percolate intercepted by the pans were collected in flasks
within the lysimeter.
A simple manifold collector was installed within a sanitary landfill by
Wilson and Small (1973) for the collection of leachate (Figure 39). The col-
lector consisted of 20 feet of 2-inch diameter PVC functioning as a horizontal
128
-------
collector manifold connected by a section of PVC pipe to a 4-inch diameter
length of vertical PVC pipe. The unit was installed within the base of a new
landfill cell prior to introducing solid waste into the area. The manifold
was drilled with a large number of openings to permit entrance of leachate.
The manifold was joined to the upright well about 4 inches above the base of
the pipe. The base of the well was sealed so that the bottom 4 inches func-
tioned as a storage reservoir. The manifold drain line was sloped toward the
well to facilitate drainage. Leachate collecting in the reservoir was hand
bailed.
UPRIGHT
WELL
SUM
20-FT LONG SECTION,
PERFORATED PVC PIPE,
THE COLLECTOR MANIFOLD
Figure 39. Leachate collector installed at base of sanitary landfill.
A basic problem with collectors such as those discussed above is that
samples are available only when the soil-water pressure is greater than atmo-
spheric. During drainage, the soil-water system may shift rapidly to the un-
saturated state, prohibiting sample collection.
Wells--
Wells installed to monitor water level fluctuations could also be used
for sample collection. Construction techniques for installing wells were dis-
cussed earlier. However, special care must be exercised to thoroughly develop
129
-------
wells used for water quality monitoring. Development will ensure removal of
drillers' mud and foreign water used in the drilling process. Development
should be continued until sediment-free water is pumped from the well (Ger-
aghty and Miller, 1977). Techniques for well development are thoroughly dis-
cussed in "Manual of Water Well Construction Practices" (U.S. EPA, no date).
Geraghty and Miller also point out that monitoring wells must be correctly
oriented in relation to the flow direction of polluted groundwater emanating
from a source. A suggested practice is to locate at least one well upgradient
of the source, one or more wells on site, and several wells downgradient of
the source.
Observation wells should be unperforated throughout the vadose zone over-
lying the perched groundwater zone. Perforations below the water table should
extend throughout the entire saturated thickness for reasons stated by Ger-
aghty and Miller (1977):
The vertical placement of the monitoring well screen within the
aquifer is ... critical. For example, wells screened near the
top of the zone of saturation may not detect contamination mi-
grating in natural gradient flow paths along the base of an
aquifer. In contrast, lighter-than-water constituents, such as
hydrocarbons, may move along the top of the zone of saturation,
thereby evading detection by monitoring wells screened only in
the bottom portion of an aquifer.
It is generally recognized that wells with fully penetrating perforated
sections provide information on the "bulk" or "average" concentration of pol-
lutants in groundwater as a pumped unit (Pickens and Grisak, 1979). In many
cases, a profile of pollutant distribution is also of importance because, as
pointed out above, a layering of constituents may occur. The sampling units
discussed below (i.e., piezometers, multilevel samplers, and profile samplers)
are useful for determining pollutant stratification. Depth-wise samples can
also be obtained in wells using packer pumps, such as the Casee Sampler.
Packer pumps are discussed in detail by Fenn et al. (1977).
Piezometers--
Piezometer wells used for characterizing the vertical head distribution
in a perched groundwater mass could also be used for water sampling. Water
samples from a piezometer array with individual well points terminating at
different depths will indicate the vertical displacement of indigenous ground-
water with the recharge source. Similarly, if a number of arrays are distri-
buted in a horizontal transect, samples from individual units terminating at
the same elevation will manifest quality changes during the lateral spread of
a pollution plume.
Methods for installing piezometers were discussed in an earlier section.
Units in a given array may either be drilled separately or clustered in a
common borehole. For the individual-type array, each piezometer should be
thoroughly developed and cleaned in accordance with sound well construction
practices. For the clustered units, development of the borehole could be
130
-------
difficult, but at any rate the gravel envelope around each well point should
effectively prevent the entrance of fines into the well.
The location of individual arrays should be based on a thorough knowledge
of the local hydrogeology, such as the slope of the water table, hydraulic
conductivity, and flow velocity. Generally, wells are installed at three lo-
cations: upgradient of the source, on-site, and downgradient of the source.
Multilevel Samplers--
Pickens et al.
shown on Figure 40.
(1978) presented details of a multilevel sampling device,
_.. . .-,_._ ... The sampler is particularly suited for sampling in cohe-
sion! ess deposits with shallow groundwater in which flow occurs predominantly
in the horizontal direction. The unit is installed in a single borehole and
permits groundwater sampling at several levels. Hydraulic head measurements
may also be obtained using a mercury manometer.
MULTI-LEVEL GROUND-WATER SAMPLER
FIELD INSTALLATION
CROSS-SECTION OF
SAMPLING POINT-TYPE A
CROSS-SECTION OF
SAMPLING POINT-TYPE B
END CAP-, MALE a
FEMALE
GROUND
SURFACE
WATER
TABLE
/COUPLINGS
_ PVC
SAMPLER
PIPE
-hCOUPLING
PLING
-7 PVC
/PIPE
SCREEN
ONE HOL
RUBBER
STOPPER
ONE-HOLE
RUBBER
STOPPER
PVC PIPE
ADHESIVE
TAPE
GLASS
WOOL
FIBRE
GLASS
CLOTH
• END CAP
Figure 40. Multilevel sampler (after Pickens et al., 1978).
Basically, the sampler consists of the following components: (1) PVC
pipe, (2) ports or openings at desired incremental depths, (3) screened
131
-------
covering on the openings, and (4) polypropylene tubing sealed into the open-
ings within the pipe. The polypropylene tubing extends to the surface. A
unit may be designed in the field by locating the position of openings using
stratigraphic information from drilling.
The pipe can be installed by common drilling techniques such as hollow-
stem auger and cable-tool. Using the hollow-stem auger, the casing is in-
stalled within the auger when the desired depth is reached and the auger is
withdrawn. For cable-tool construction, it is necessary to drive in a steel
casing to the desired depth, install the PVC pipe within the casing, and then
pull out the casing. For both methods, it is presumed that the deposits sur-
rounding the borehole will collapse around the casing, ostensibly preventing
side leakage. Vanhof, Weyer, and Whitaker (1979) questioned this assumption.
However, in their paper, Pickens et al. indicated that sealing should not be
a problem with cohesionless deposits. For installation in cohesive deposits,
they recommend that bentonite or grout be installed between openings to pre-
vent cross-contamination.
During sampling, a vacuum is applied to the polypropylene tubing and wa-
ter flows from the formation to a collection bottle. The water can be pumped
into an air-free cell for measurement of pH and Eh. Quick delivery from the
aquifer to the cell probably minimizes degassing of the water.
One of the stated advantages of the multilevel sampler is that concentra-
tion profiles can be obtained at a lower cost than by installing piezometer
nests. In a subsequent paper, Pickens and Grisak (1979) present comparative
cost data to substantiate this contention.
Groundwater Profile Sampler--
Hansen and Harris (1974) designed a depth-wise sampler, which they desig-
nated a "groundwater profile sampler." The sampler, shown on Figure 41, con-
sists of a 1.25-inch diameter well point, of optional length, with isolated
chambers containing fiberglass probes. The individual chambers are filled
with sand and separated by caulking compound. Small-diameter tubing provides
surface access to the probes. The positioning of probes is optional, depend-
ing on aquifer materials and desired sampling frequency. In operation, a
vacuum is applied to the tubing, pulling groundwater into individual sampling
flasks. Hansen and Harris recommended that all samples should be extracted
simultaneously and at the same rate to minimize variations in aquifer thick-
ness sampled by the individual probes. Groundwater at depths to 30 feet may
be sampled with this unit.
Well Casing Material —
Two important considerations when designing wells and piezometers as
quality monitoring facilities are casing material and casing diameter. A num-
ber of materials are available for well casing including plain steel, galva-
nized steel, PVC, fiberglass, and Teflon. Selection of material for wells at
a specific site will depend on the constituents being monitored. When moni-
toring for salinity where only the major constituents are being evaluated,
plain steel or galvanized steel could be used. When monitoring for trace
132
-------
1/4 OD TUBING
SAMPLE COLLECTION
FLASKS
WATER TABLE
1/4" CAULKING
HOLES
I 1/4 WELL POINT
SAND MATRIX
CAULKING
V
Figure 41. Groundwater profile sampler (after Hansen
and Harris, 1974).
133
-------
metals, contamination from steel casing could preclude the quantification of
specific constituents, such as iron and zinc. Consequently, casing con-
structed from PVC, fiberglass, or Teflon is advisable. Because of cost, PVC
is commonly used. This material has the disadvantage that it cannot be driven
into place, such as when using the cable-tool method.
A controversy exists over the type of casing material to install when
monitoring for organic pollutants. For example, Dunlap et al. (1977) oppose
the use of PVC pipe, stating:
In some earlier work ... PVC casing was utilized for casing of
sampling wells. This material is relatively inexpensive and
easy to use, but it is less desirable as a casing material than
the Teflon tubing-galvanized pipe combination for two reasons.
First, organic constituents of groundwater may be adsorbed on
the PVC casing. Second, there is evidence that PVC casing may
contribute low levels of organic contaminants to the samples,
such as phthalic acid esters used as plasticizers in PVC manu-
facture and solvent from cements used to join lengths of PVC
tubing.
For their sampling wells, Dunlap et al. used a combination of galvanized
pipe and Teflon tubing. The galvanized pipe extended from ground surface to
about 1 foot above the water table. Teflon tubing was used in the portion of
the well below the water table.
In contrast to the findings of Dunlap et al. regarding interaction of PVC
pipe with organic constituents, Geraghty and Miller (1977) indicated the
following:
PVC pipe was used for all well casing. It is light and easy to
handle and is more inert toward dissolved organic substances
than steel casing. The iron oxide coating that develops on
steel casing has an unpredictable and changeable adsorption
capacity. However, when the adsorption sites on PVC are satu-
rated, water remains in equilibrium with it. Leakage of or-
ganic compounds from PVC is negligible. As a control, samples
of pipe and a cemented joint were submitted to the laboratory
where they were soaked in water and the water was analyzed.
No contaminants were detected.
No general recommendations are available on selecting well casing mate-
rial when monitoring for microorganisms and virus. Conceivably, metal con-
stituents in steel wells could inhibit the growth of microorganisms as
dissolution of the casing occurs. For this reason, PVC wells are probably
more suitable for sample collection when microbial levels are being evaluated.
It is also probably advisable to disinfect the well following well completion.
Chlorine in granular or liquid form is commonly used as a disinfectant (U.S.
EPA, no date). The "Manual of Water Well Construction Practices" gives de-
tails on disinfection practices. Care must be exercised in using chlorine
when excessive levels of organic constituents are present because of the pos-
sible formation of chlorinated hydrocarbons. In addition, the well should
134
-------
be thoroughly pumped after a suitable contact time using sterilized pumping
equipment.
Selection of well casing diameter should be based on the water extraction
method, quantity of sample required, budget limitations, and other uses of the
well. If large quantities of sample are required and a submersible pump is to
be installed, the minimum diameter of the casing should be 6 inches, and 8
inches is preferred (Schmidt, personal communication, 1979). If other extrac-
tion methods such as hand bailers, gas-lift pumps, or peristaltic pumps are
suitable, -smaller-diameter wells could be used. However, many wells are also
used for monitoring water level changes and permanent fixtures such as gas
pressure lines may occupy part of the interior space. For practical reasons,
therefore, the casing diameter should probably not be less than 4 inches.
Sample Extraction Methods--
A number of techniques are available for extracting groundwater from
wells and piezometers. Fenn et al. (1977) reviewed several of these methods
in detail. A brief discussion is presented here of the following methods:
(1) hand bailers, (2) air-lift and gas-lift pumps, (3) suction-lift pumps, (4)
piston pumps, (5) centrifugal pumps, and (6) submersible pumps.
Hand bailers--!n their simplest form, hand bailers consist of a pipe
closed at one end and open at the other. A cable is attached to the open end
to permit raising or lowering the bailer. Weights are added to the bottom of
the bailer to overcome buoyancy. To collect a sample, the bailer is lowered
into a well down to and below the water table. The filled container is then
raised to the surface.
A more elaborate type of bailer is shown on Figure 42. This unit con-
sists of a tube open at both ends but with a one-way check valve at the base.
Basic units, such as the one shown in the figure, use a glass marble as part
of the valve assembly. In operation, the unit is lowered by cable down to and
below the water table. Water flows up and into the sampler body. When the
unit is raised, the glass marble falls into a conical seat effectively plug-
ging the end of the bailer. Bailers may be constructed from available mate-
rials such as steel pipe, PVC tubing, or Teflon tubing. Selection of material
depends on the constituents being examined. The unit shown on the figure was
constructed of Teflon to sample for highly volatile organic pollutants (Dunlap
et al., 1977). The bailer is sterilized by autoclaving when sampling for
microorganisms.
Another type of hand bailer is the "Kemmerer" sampler. These units are
commonly used for sampling in lakes but are also suitable for groundwater sam-
pling. The units basically consist of a cylinder with spring-loaded rubber
stoppers at each end. The unit is lowered to the desired depth below the wa-
ter table and a weighted "messenger" is sent down a cable to spring shut the
stoppers. The sample is thus trapped within the container.
Bailers are not particularly effective for flushing casing prior to sam-
pling unless the volume of water in the casing is minimal.
135
-------
-NICKEL WIRE CABLE
JJ_
l-l/4"O.D.Xl" I.D. TEFLON
EXTRUDED TUBING,
18" TO 36" LONG
•3/4 DIAMETER
GLASS MARBLE
1 DIAMETER TEFLON
EXTRUDED ROD
5/16" DIAMETER HOLE
Figure 42. Hand bailer made of Teflon (after Dunlap et al., 1977).
Air-lift and gas-lift pumps--A basic air-lift pump is illustrated on Fig-
ure 43. As shown, compressed air introduced into a central pipe bubbles up
through the column of water in the well. The specific gravity of the column
of water is consequently reduced and an air-water mixture is discharged from
an eductor pipe. In the case shown, the well casing is the eductor pipe. The
principles of air-lift pumping are discussed in "Ground Water and Wells"
(Johnson, 1966), and by Anderson (1977), and Trescott and Finder (1970). As
pointed out in "Ground Water and Wells," the air-lift method is particularly
effective when the submergence ratio of the air line is about 60 percent. The
submergence ratio is the percentage of the total length of air line that is
below water when pumping. For example, if the air line is 180-feet long and
136
-------
AIR
cz
™ ~~]
PUM
Lll
h
i
• •.
PIN
-J.
• "
• •
w :
.• •
\ "
G
• •
• •
• •
* •
• •
.'.'.
• * • t •
* • • • •
• • • * •
-0
_ 0
V
_o
V
o°
0?
—
X"
oT
^
^^j
^L ^^
T *.*
*0"
—
V?"
« * •
(V
o ..
-^ ::
^!!
» •
D° * *
•
ILr^iIn:
L— LJT':
i
*.'.'.
. * *
:::h
-Ar.
• • •
• • •
fc
nrT
r
• • •
• • •
* * •
m
hm= MAXIMUM HEIGHT TO WHICH THE
AIR-WATER MIXTURE WILL RISE
hw= SUBMERGED LENGTH OF THE AIR LINE
T = SPECIFIC WEIGHT OF THE AIR-WATER MIXTURE
Tw= SPECIFIC WEIGHT OF WATER
J. = POTENTIOMETRIC SURFACE
Figure 43. Schematic representation of simple air-lift pump showing
principle of operation (after Trescott and Pinder, 1970).
137
-------
the water level when pumping is at 74 feet, the submergence rate is (180-74)/
180 or 59 percent. Apparently, favorable pumping rates are still possible
with as little as 30-percent submergence.
Trescott and Pinder (1970) developed a simple air-lift pump for sampling
from small-diameter piezometers. An automobile engine was used to provide air
pressure for the pump. Components of the pump included: a ball check valve,
an air line, a pressure gage, and a header assembly on the well. The check
valve which screws into the spark plug socket of an engine cylinder operates
such that clean air from outside the engine is sucked into the cylinder during
the inlet stroke. The compressed air is then forced into the air line during
the compression stroke. The check valve maintains pressure in the air line.
The header assembly which is fastened to the top of the well casing contains
an inlet opening, through which the air line is inserted, and a discharge
pipe. The discharge rate for air-lift pumps increases with increasing pipe
diameter.
If sufficient air is available and the submergence rate is favorable,
enough water can be extracted by the air-lift method for sampling purposes.
The system could also be used for the initial development of the well and for
flushing prior to sampling.
Gronowski (1979) described a new commercial pump operating on the same
principle as the air-lift pump except that the pumping energy is obtained from
14-ounce propane cylinders. According to Gronowski, "The unit consists of
dual-conductor plastic tubing with a valve assembly on top for attachment to
the propane cylinder. A shaped copper tubing serves as the intake assembly.
Propane injected down one side of the tubing forces water to rise up the other
side to ground surface. Several samples of adequate volume can be collected
with one 14-ounce propane cylinder."
Bianchi, Johnson, and Haskell (1962) designed an alternative air-lift
pump which in effect is an automated hand bailer. As shown on Figure 44, the
unit consists of a cylinder with a flat one-way valve at the base and a two-
hole rubber stopper at the upper end. One nylon tube is inserted through one
of the holes and extended to the lower end of the cylinder. A second nylon
line is inserted into the second opening. In operation, the cylinder is low-
ered below the water table and water flows upward through the valve. Air
pressure is then applied to the shorter line, the one-way valve shuts, and wa-
ter is forced up the longer line to the surface. A tire pump could be used to
apply pressure. Samples have been lifted with this unit from depths greater
than 100 feet.
The simple air-lift pumps introduce air into the sample. Consequently,
extracted samples are unsuitable for determining dissolved gasses and many un-
stable constituents in groundwater. The effect of propane on such constitu-
ents, when using the gas-lift pump, should be examined before the method is
employed in a specific monitoring program.
Suction-lift pumps--The basic principle of suction-lift pumps is that,
when a vacuum is created inside a tube or pipe extending below the water ta-
ble, air pressure on the water outside the tube forces water to rise inside
138
-------
•WIRE BALL
-AIR LINE
-DISCHARGE LINE
-TOP SEAL
PUMP BODY
VALVE GUIDES
TOOT VALVE
.FOOT VALVE
SEAL
•BASE PLATE
•FILTER SCREEN-
Figure 44. Positive action air-lift pump
Johnson, and Haskell, 1962).
(after Bianchi,
the tube (Anderson, 1977). The practical suction lift with such pumps is
about 25 feet. Allison (1971) discussed a simple suction-lift pump for ex-
tracting groundwater from auger holes. The method could also be used for
cased wells. Basically, the unit consists of plastic tubing of sufficient
length to extend below the water table. The upper end of the tube terminates
in a collector flask. A second line from the flask is connected to a hand-
held vacuum pump. By applying vacuum to the system, water is forced into the
flask.
A variation of the suction-lift pump is the so-called peristaltic pump,
or squeeze pump. The pump consists of a rotor equipped with rotating heads at
both ends. The rotor is mounted in a closed housing. Plastic tubing is
placed between the rotor heads and the housing such that when the rotor turns,
the plastic tubing is squeezed by the rotating heads. The tubing is lowered
into a well below the water table and primed with water. When the unit is ac-
tivated, the squeezing action of the heads produces a negative pressure on the
water in the tubing extending into the well. At the same time, a positive
pressure is exerted on the water in the tubing leaving the housing.
An advantage of the peristaltic pump is that water is in contact only
with the tubing. Contamination of the water sample with pump accessories is
thus minimized. Dunlap et al. (1977) used peristaltic pumps for both grab
sampling and continuous extraction of groundwater when sampling for organic
pollutants.
139
-------
A basic problem with peristaltic pumps is that discharge rates are very
low. Also, foreign water is added when priming the unit. The latter problem
could be overcome by using hand-bailed groundwater for priming.
Piston pumps—Piston pumps are a type of positive displacement pump often
used in windmill wells. Basically, the pump consists of a cylinder containing
a plunger or piston which is driven up and down via a rod. The lower end of
the cylinder contains an inlet valve consisting of a one-way check valve. A
one-way check valve is also mounted in the piston. The piston also contains
exterior leathers which contact the walls of the cylinder.
The operation of the pump is described in "Ground Water and Wells":
As the piston moves downward, the intake valve closes at the
instant that the pressure above it exceeds the pressure below
it; and the discharge valve opens when the pressure below it
exceeds that above it. Water thus is trapped in the cylinder
during the downstroke of the piston, then is forced upward into
the discharge pipe on the next upstroke.
During installation, the cylinder is attached to a length of piston rod
inside a length of PVC pipe. The PVC pipe functions as an eductor tube. Suc-
cessive lengths of pipe and rod are added until the cylinder is located at the
desired depth below the water table. The upper end of the pipe/rod assembly
is rigidly attached to the casing. The PVC pipe could be attached to a T-
section, permitting the sample to be discharged into a collector flask. When
the water table is not in excess of about 10 feet, the pump may be operated by
hand. For deeper water tables, a tripod and pulley arrangement may be neces-
sary. Alternatively, a pitcher pump could be attached to the top of the well.
In addition to providing water samples, piston pumps are an effective
method for developing a new well. Also, if the volume of water stored in the
well is not excessive, piston pumps may be used for flushing the casing prior
to sampling. A problem with piston pumps for routine sampling is that the
installation time may be excessive compared to more portable units. An expen-
sive alternative would be to permanently install a separate unit in each moni-
toring well. Another basic problem with piston pumps is that metals from the
cylinder may contaminate pumped samples.
Centrifugal pumps--Engine-driven centrifugal pumps could be used to ex-
tract water samples from lifts of about 15 feet. Depending on well diameter,
such pumps could be used to extract large quantities of samples.
Submersible pumps—Submersible pumps comprise a sealed centrifugal-type
pumping plant including an electric motor which operates below the water ta-
ble. Water is pumped to the surface via delivery tubing.
McMillion and Keeley (1968) described a portable submersible pump suit-
able for sampling in wells greater than 4.5 inches inside diameter. The com-
ponents of the unit are a submersible pump, 300 feet of flexible hose, an
electric winch and spool assembly, and a portable generator. The generator
provides electricity to drive the pump and operate the winch. In operation,
140
-------
the hose is unreeled from the spool assembly to position the pump at the de-
sired depth and the pump is activated. Pumping rates range from 7 to 14 gal-
lons per minute. The assembly is stored in a van during transit between
wells.
Although the submersible pump assembly of McMillion and Keeley is expen-
sive, it is convenient to install and operate within a well, the time for sam-
ple collection is minimal, and the unit permits rapid flushing of stagnant
water in the casing prior to sampling. According to Fenn et al. (1977), if
there are no leaks in the system, the unit could be used to sample for dis-
solved gasses.
Well Flushing Prior to Sampling--
Sound monitoring practice requires that a monitor well or piezometer be
thoroughly flushed prior to sample collection. The rationale is that ground-
water within a passive monitor well tends to be affected by exposure to the
atmosphere and by mixing within the casing. (Passive monitoring is defined
as periodically sampling wells in the path of groundwater flow for changes in
concentration of the constituents of interest (Hart Associates, 1979)).
Consequently, in order to obtain a truly representative sample of water from
the perched region, the water in the casing should be completely displaced.
Fenn et al. (1977) thoroughly reviewed procedures for flushing. The
principal recommendations are cited below:
To safeguard against collecting nonrepresentative stagnant wa-
ter in a sample, the following guidelines and techniques should
be adhered to during sample withdrawal: As a general rule, all
monitoring wells should be pumped or bailed prior to withdraw-
ing a sample. Evacuation of a minimum of one volume of water
in the well casing and preferably three to five volumes is
recommended for a representative sample. In a high-yielding
groundwater formation and where there is no stagnant water in
the well above the screened section, evacuation prior to sample
withdrawal is not as critical. However, in all cases where the
monitoring data are to be used for enforcement actions, evacu-
ation is recommended.
For wells that can be pumped or bailed to dryness with the sam-
pling equipment being used, the well should be evacuated and
allowed to recover prior to sample withdrawal. If the recovery
rate is fairly rapid and time allows, evacuation of more than
one volume of water is preferred.
For high-yielding monitoring wells which cannot be evacuated to
dryness, bailing without prepumping the well is not recom-
mended; there is no absolute safeguard against contaminating
the sample with stagnant water. The following procedures
should be used:
141
-------
(a) The inlet line of the sampling pump should be placed just
below the surface of the well water and three to five volumes
of water pumped at a rate equal to the well's recovery rate.
This provides reasonable assurance that all stagnant water has
been evacuated and that the sample will be representative of
the groundwater body at that time. The sample can then be col-
lected directly from the pump discharge line.
(b) The inlet line of the sampling pump (or the submersible
pump itself) should be placed near the bottom of the screen
section, pumped approximately one well volume of water at the
well's recovery rate, and the sample collected directly from
the discharge line.
A nonrepresentative sample can also result from excessive pre-
pumping of the monitoring well. Stratification of the leachate
concentrations in the groundwater formation may occur, and ex-
cessive pumping can dilute or increase the contaminant concen-
trations from what is representative of the sampling point of
interest.
Techniques for Sampling Perched Groundwater in
the Lower Vadose Zone--
In deep vadose zones, such as those of western alluvial basins, it is not
unusual to encounter one or more regions of perched groundwater in the lower
vadose zone. Water samples from these deeper saturated regions reflect the
quality of water which has moved below the soil zone enroute to the water ta-
ble. In many cases, large quantities of water can be extracted from these
deeper perched groundwater zones. Furthermore, the samples may reflect the
integrated quality of water draining from an extensive portion of the overly-
ing vadose zone. This will be particularly true when the overlying source is
diffuse, such as irrigation return flow. Because of the integrated nature of
the water in these perched regions, samples may be obtained using one or more
wells at a fraction of the cost for installing batteries of suction cups.
Wilson and Schmidt (1979) presented data from a number of case studies on
monitoring perched groundwater. The sampling techniques reviewed below are
discussed in detail in their paper. Two methods are possible for sampling
perched groundwater: (1) cascading water, and (2) special wells. Generally,
perched groundwater cascades into wells through poorly welded joints, cracks,
or perforations exposed as the water table recedes. Figure 45 illustrates a
cross section of a conceptualized well with cascading water. Samples of cas-
cading water could be obtained from abandoned wells in the vicinity of the
source being monitored. Alternatively, cascading water could be obtained from
operating wells whenever pumps are removed for servicing.
Samples of cascading water can be obtained by lowering a bucket to the
desired depth. In sampling for chemical constituents, the sample container
and sample bottles should be rinsed several times during sample collection.
To sample for microorganisms, a sterilized container should be used. Inasmuch
as cascading water is exposed to the atmosphere, some problems may be
142
-------
GROUND SURFACE
PERCHED
WATER LEVEL
PERCHING
LAYER
WATER TABLE
ZONE OF CASCADING
WATER IN AQUIFER
PERFORATED
INTERVAL
BASE OF AQUIFER
Figure 45. Conceptualized cross section of a well showing cascading
water from perched zone (after Wilson and Schmidt, 1979).
experienced in determining pH, dissolved oxygen, and oxidation-reduction
potential.
If wells with cascading water are not present in the vicinity of a
source, it will be necessary to construct special wells (Figure 46). A major
problem will be to locate wells at the proper depth to tap perched water. Al-
ternative methods for locating existing or potential perching regions were
discussed earlier in this report, including drillers' observations, grain size
of drill cuttings, and borehole geophysical methods.
Materials for casing will depend on the constituents of interest in the
monitoring program. Thus, if metals and possibly organic pollutants are of
concern, it may be advisable to use PVC casing. Inasmuch as PVC casing cannot
be driven, the rotary technique will be required for installation when the re-
gion of interest is deep. If the region is not too deep, however, it may be
possible to use the cable-tool method as follows: steel casing is driven into
the borehole as drilling progresses and, when the desired depth is reached,
PVC casing is lowered into the steel casing. The steel casing is then pulled
143
-------
out, leaving the PVC casing in place.
throughout the region of interest.
The PVC casing should be preperforated
GROUND SURFACE.
PERCHED
WATER LEVEL
PERCHING
LAYER
PERCHED ZONE
WATER TABLE
AQUIFER
BASE OF
AQUIFER
PERFORATED
INTERVAL
Figure 46. Conceptualized monitor well used to sample from a
perched zone (after Wilson and Schmidt, 1979).
The ideal casing size for monitoring wells is probably about 8 inches
(Schmidt, personal communication, 1979). This size permits easy installation
and removal of a portable submersible pump for sampling. In addition, a water
level recorder could also be installed within 8-inch casing.
144
-------
When several perching layers occur, a cluster of sampling wells could be
installed within a common borehole. Methods for clustering wells are identi-
cal to those discussed elsewhere for multiple piezometer installations.
Several limitations in sampling perched groundwater should be noted.
Perched zones do not occur everywhere and may not be present at a specific
area of interest. The detection and location of perched groundwater may be
expensive, involving the construction of test wells or requiring the use of
geophysical methods as discussed above. Even when found, some perched ground-
water bodies may be ephemeral. Alternatively, certain perching regions may
be more responsive to recharge from a given source of pollution than other re-
gions. Thus, the location of successful monitoring wells may in part be a
matter of chance.
A problem in sampling cascading wells is that water may enter the casing
from a number of distinct perching regions. Consequently, the water collected
below the uppermost region could be a mixture of water from various depths.
However, such a blend could provide useful information.
145
-------
REFERENCES
Alemi, M.H., D.R. Nielsen, and J.W. Biggar, Determining the Hydraulic Conduc-
tivity of Soil Cores by Centrifugation, Journal Soil Science Soc.
America, 40, 42-218, 1976.
Alesii, B.A., Cyanide Mobility in Soils, unpublished M.S. Thesis, The Univer-
sity of Arizona, Tucson, Arizona, 1976.
Alexander, M., Introduction to Soil Microbiology, John Wiley and Sons, Inc.,
New York, T96T:
Allison, I.E., A Simple Device for Sampling Ground Waters in Auger Holes, Soil
Sci. Soc. Amer. Proc., 35, 844-845, 1971.
Anderson, K.E., editor, Water Well Handbook, Missouri Water Well Drillers
Assocation, Rolla, Missouri, 1977.
Apgar, M.A., and D. Langmuir, Ground-Water Pollution Potential of a Landfill
Above the Water Table, Ground Water, 9(6), 76-93, 1971.
Ayers, R.S., and D.W. Westcot, Water Quality for Agriculture, Irrigation and
Drainage Paper No. 29, Food and Agriculture Organization of the United
Nations, 1976.
Ayers, R.S., and R.L. Branson, editors, Nitrates in the Upper Santa Ana River
Basin in Relation to Groundwater Pollution, California Agric. Exp. Sta-
tion, Bulletin 861, 1973.
Bailey, G.W., and J.L. White, Factors Influencing the Adsorption, Desorption,
and Movement of Pesticides in Soil, in Residue Reviews, Vol. 32,
Springer-Verlag, New York, 1970.
Bear, J., Dynamics of Fluids in Porous Media, American Elsevier, New York,
1972.
Bear, J., D. Zaslavsky, and S. Irmay, Physical Principles of Water Percolation
and Seepage. United National Educational, Scientific and Cultural Organi-
zation, 465 pp, 1968.
Bertrand, A.R., Rate of Water Intake in the Field, in Methods of Soil Analy-
ses, C.A. Black, editor, Agronomy No. 9, 197-208, American Society of
Agronomy, Madison, Wisconsin, 1965.
146
-------
Bianchi, W.C., C.E. Johnson, and E.E. Haskell, A Positive Action Pump for Sam-
pling Small Bore Wells, Soil Sci. Soc. Amer. Proc., 26(1), 86-87, 1962.
Bianchi, W., Measuring Soil Moisture Tension Changes, Agricultural Engineer-
ing, 43(7), 398, 399, 404, 1967.
Biggar, J.W., and D.R. Nielsen, Spatial Variability of the Leaching Character-
istics of a Field Soil, Water Resources Research, 12(1), 78-84, 1976.
Black, C.A., editor, Methods of Soil Analyses, (in two parts), Agronomy No. 9,
American Society of Agronomy, Madison, Wisconsin, 1965.
Blake, 6.R., Bulk Density, in Methods of Soil Analyses, C.A. Black, editor,
Agronomy No. 9, 374-390, American Society of Agronomy, Madison, Wiscon-
sin, 1965.
Bodman, 6.B., and E.A. Coleman, Moisture and Energy Conditions During Down-
ward Entry of Water into Soils, Soil Sci. Soc. Amer. Proc., 8, 116-122,
1944.
Bohn, H.L., B.L. McNeal, and 6.A. O'Connor, Soil Chemistry, Wiley Intersci-
ence, New York, New York, 1979.
Bordner, R., J. Winter, and P. Scarpino, Microbiological Methods for Monitor-
ing the Environment, EPA-600/8-78-017, U.S. Environmental Laboratory, En-
vironmental Monitoring and Support Laboratory, 1978.
Borg, I.Y., R. Stone, H.B. Levy, and L.D. Ramspott, Information Pertinent to
the Migration of Radionuclides in Ground Water at the Nevada Test Site,
Part I: Review and Analysis of Existing Information, Lawrence Livermore
Laboratory, Livermore, California, National Technical Information Ser-
vice, 1976.
Bouwer, H., Groundwater Hydrology, McGraw-Hill Book Co., New York, 480 pp,
1978.
Bouwer, H., and R.D. Jackson, Determining Soil Properties, in Drainage for
Agriculture, J. van Schilfgaarde, editor, Agronomy No. 17, 611-675, Amer-
ican Society of Agronomy, Madison, Wisconsin, 1974.
Bower, C.A., and L.V. Wilcox, Soluble Salts, in Methods of Soil Analyses, C.A.
Black, editor, Agronomy No. 9, 933-951, American Society of Agronomy,
Madison, Wisconsin, 1965.
Brakensiek, D.L., Empirical and Simplified Models of the Infiltration Process,
in Infiltration Research Planning Workshop, Part 1, State of the Art Re-
ports, Sciences and Education Administration, The U.S. Dept. of Agric.,
ARM-NC-4, 1979.
Brakensiek, D.L., H.B. Osborn, and W.J. Rawls, Field Manual for Research in
Agricultural Hydrology, Agricultural Handbook No. 224, Sciences and Edu-
cation Administration, The U.S. Dept. of Agric., 1979.
147
-------
Broadbent, F.E., Organics, in Proceedings of the Joint Conference on Recycling
Municipal Sludges and Effluents on Land, sponsored by The Environmental
Protection Agency, The U.S. Dept. of Agric., and The Nat. Assoc. of State
Univ. and Land-Grant Colleges, Champaign, Illinois, July 9-13, 1973.
Cartwright, K., and F.B. Sherman, Electrical Earth Resistivity Surveying in
Landfill Investigations, Proceedings 10th Annual Engineering and Soils
Engineering Symposium, Moscow, Idaho, 77-92, 1972.
Cartwright, K., and M.R. McComas, Geophysical Surveys in the Vicinity of Sani-
tary Landfills in Northeastern Illinois, Ground Water, 6(5), 23-30, 1968.
Cary, J.W., Soil Water Flowmeters with Thermocouple Outputs, Soil Sci. Soc.
Amer. Proc.. 37, 176-181, 1973.
Chang, A.C., and A.L. Page, Hydraulic and Water Quality Considerations of Ar-
tificial Groundwater Recharge, in Proceedings of the Twelfth Biennial
Conference on Ground Water, Sacramento, September 20-21, 1979, California
Water Resources Center, The University of California, Davis, Report No.
45, pp 88-99, 1979.
Chapman, H.D., Cation Exchange Capacity, in Methods of Soil Analyses. C.A.
Black, editor, Agronomy No. 9, 891-900, American Society of Agronomy,
Madison, Wisconsin, 1965.
Childs, E.C., An Introduction to the Physical Basis of Soil Water Phenomena,
Wiley Interscience, New York, New York, 1969.
Chow, T.L., A Porous Cup Soil-Water Sampler with Volume Control, Soil Science,
124, 173-176, 1977.
Coelho, M.A., Spatial Variability of Water Related Soil Physical Properties,
unpublished M.S. Thesis, The University of Arizona, Tucson, Arizona,
1974.
Cooley, R.L., J.F. Harsh, and D.C. Lewis, Principles of Ground-Water Hydrol-
ogy, Hydrologic Engineering Methods for Water Resources Development, Vol.
10, The Hydrologic Engineering Center, U.S. Army Corps of Engineers,
Davis, California, 1972.
Corey. J.C., S.F. Peterson, and M.A. Wakat, Measurement of Attenuation of
13/Cs and 241^ Gamma Rays for Soil Density and Water Content Deter-
minations, Soil Sci. Soc. Amer. Proc., 35, 215-219, 1971.
Davis, J.B., Microbial Decomposition of Hydrocarbons, Industrial and Engi-
neering Chemistry, 48(9), 1444-1448, 1956.
Davis, S.N., and R.J.M. deWiest, Hydrogeology, John Wiley and Sons, Inc., New
York, 1966.
148
-------
Day, P.R., G.H. Bolt, and D.M. Anderson, Nature of Soil Water, in Irrigation
of Agricultural Lands, R.M. Hagan, H.R. Haise, and T.W. Edminster, edi-
tors, Agronomy No. 9, 193-208, American Society of Agronomy, Madison,
Wisconsin, 1967.
Dirksen, C., Water Flux Measurements in Unsaturated Soils, in Flow-Its Mea-
surement and Control in Science and Industry. Vol. 1, 479-486, Inst. Soc.
Amer., Pittsburgh, Pennsylvania, 1974a.
Dirksen, C., Field Test of Soil Water Flux Meters, Trans. Amer. Soc. Agric.
Eng.. 17(6), 1038-1042, 1974b.
Donnan, W.W., Field Methods, in Drainage of Agricultural Lands. J.N. Luthin,
editor, American Society of Agronomy, Madison, Wisconsin, 1957.
Duke, H.R., and H.R. Haise, Vacuum Extractors to Assess Deep Percolation
Losses and Chemical Constituents of Soil Water, Soil Sci. Soc. Amer.
Proc., 37, 963-964, 1973.
Dunlap, W.J., and J.F. McNabb, Subsurface Biological Activity in Relation to
Ground Water Pollution, EPA-660/2-73-014, U.S. Enviornmental Protection
Agency, Corvallis, Oregon, 1973.
Dunlap, W.J., J.F. McNabb, M.R. Scalf, and R.L. Cosby, Sampling for Organic
Chemicals and Microorganisms in the Subsurface, EPA-600/2-77-176, U.S.
Environmental Protection Agency, 1977.
Ellis, B.G., The Soil as a Chemical Filter, in Recycling Treated Municipal
Waste Water and Sludge Through Forest and Cropland, W.E. Sopper and L.T.
Kardos, editors, The Pennsylvania State University Press, 1973.
Enfield, C.G., J.J.C. Hsieh, and A.W. Warrick, Evaluation of Water Flux Above
a Deep Water Table Using Thermocouple Phsychrometers, Soil Sci. Soc.
Amer. Proc., 37(6), 968-970, 1973.
England, C.B., Comments of "A Technique Using Porous Cups for Water Sampling
at Any Depth in the Unsaturated Zone" by Warren W. Wood, Water Resources
Research, 10(6), 1049, 1974.
Estes, J.E., D.S. Simonett, L.R. Tinney, C.E. Ezra, B. Bowman, and M. Roberts,
Remote Sensing Detection of Perched Water Tables, Technical Completion
Report, Contribution No. 175. California Water Resources Center, The
University of California, Davis, California, 1978.
Evans, D.D., T.W. Sammis, A.W. Warrick, Transient Movement of Water and Sol-
utes in Unsaturated Soil Systems, Phase II, Project Completion Report,
OWRT Project No. B040-ARIZ, Department of Hydrology and Water Resources,
The University of Arizona, Tucson, Arizona, 1976.
149
-------
Everett, L.G., editor, Groundwater Quality Monitoring of Western Coal Strip
Mining: Identification and Priority Ranking of Potential Pollution
Sources. EPA-60Q/7-79-024, U.S. Environmental Protection Agency. Las
Vegas, Nevada, 1979.
Everett, L.G., K.D. Schmidt, R.M. Tinlin, and O.K. Todd, Monitoring Groundwa-
ter Quality: Methods and Costs. EPA-600/4-76-023, U.S. Environmental
Protection Agency, Environmental Monitoring and Support Laboratory, Las
Vegas, Nevada, 1976.
Fenn, D., E. Cocozza, J. Isbister, 0. Briads, B. Yare, and P. Roux, Procedures
Manual for Ground Water Monitoring at Solid Waste Disposal Facilities.
EPA/530/SW-611, U.S. Environmental Protection Agency, 1977.
Fenn, D.G., K.J. Hanley, and T.V. DeGeare, Use of the Water Balance Method for
Predicting Leachate Generation from SoTTd Waste Disposal Sites. EPA/530/
SW-618, U.S. Environmental Protection Agency, Office of Solid Waste,
1975.
Freeze, R.A., The Mechanism of Natural Ground-Water Recharge; 1 - One-Dimen-
sional Vertical, Unsteady, Unsaturated Flow Above a Discharging Ground-
Water Flow System, Water Resources Research. 5(1), 153-171, 1969.
Freeze, R.A., and J.A. Cherry, Groundwater, Prentice-Hall Inc., Englewood
Cliffs, New Jersey, 1979.
Frissel, M.G., P. Poelstra. K. Harmsen, and G.H. Bolt, Tracing Soil-Moisture
Migration with 36C1, 60Co and Tritium, in Isotope and Radiation Tech-
niques in Soil Physics and Irrigation Studies 1973. International Atomic
Energy Agency, Vienna, 1974.
Fuller, W.H., Movement of N-Selected Metals, Asbestos, and Cyanide in Soil:
Application to Waste Disposal Problems. EPA-600/2-77-020. U.S. Environ-
mental Protection Agency, 1977.
Fuller, W.H., Investigation of Landfill Leachate Pollutant Attenuation by
Soils, EPA-600/2-78-158, U.S. Environmental Protection Agency, Cincin-
nati, Ohio, 1978.
Gairon, S., and A. Hadas, Measurement of the Water Status in Soils, in Arid
Zone Irrigation, B. Yaron, E. Danfois, and Y. Vaadia, editors, 215-226,
Springer-Verlag, New York, 1973.
Gardner, W.H., Water Content, in Methods of Soil Analyses, C.A. Black, editor,
Agronomy No. 9, 82-125, American Society of Agronomy, Madison, Wisconsin,
1965.
Geraghty & Miller, Inc., The Prevalence of Subsurface Migration of Hazardous
Chemical Substances at Selected Industrial Waste Land Disposal Sites,
EPA/530/SW-634, U.S. Environmental Protection Agency, 1977.
150
-------
Gilbert, R.G., C.P. Gerba.'R.C. Rice, H. Bouwer, C. Wallis, and J.L. Melnick,
Virus and Bacteria Removal from Wastewater by Land Treatment, Applied and
Environmental Microbiology, 32(3), 333-338, 1976.
F T .---—- r v- -r -.- --~r -- "** N
Green, W.H., and G.A. Ampt, Studies on Soil Physics: I. Flow of Air and Water
Through Soils, Journal of Agricultural Research, 4, 1-24, 1911.
Gronowski, R., Innovations Simplify Shallow Well Monitoring, Water Well Jour-
nal, 33(7), 57, 1979.
Guma'a, G.S., Spatial Variability of In-Situ Available Water, unpublished
Ph.D. Dissertation, The University of Arizona, Tucson, Arizona, 1978.
Hall, W.A., Theoretical Aspects of Water Spreading, Amer. Soc. Ag. Eng.,
36(6), 394-397, 1955.
Hamon, W.R., Infiltrometer Using Simulated Rainfall for Infiltration Re-
search, in Infiltration Research Planning Workshop, Part 1, State of the
Art Reports, Sciences and Education Administration, The U.S. Dept. of
Agric., ARM-NC-4, 1979.
Hansen, E.A., and A.R. Harris, A Groundwater Profile Sampler, Water Resources
Research, 10(2), 375, 1974.
Hansen, E.A., and A.R. Harris, Validity of Soil-Water Samples Collected With
Porous Ceramic Cups, Soil Sci. Soc. Amer. Proc., 39, 528-536, 1975.
Hantush, M.A., Growth and Decay of Ground-Water Mounds in Response to Uniform
Percolation, Water Resources Research, 3(1), 277-234, 1967.
Hart, Fred C., Associates, Draft Environmental Impact Statement on the Pro-
posed Guidelines for the Landfill Disposal of Solid Waste, U.S. Environ-
mental Protection Agency, Office of Solid Waste, 1979.
Hem, J.D., Study and Interpretation of the Chemical Characteristics of Natural
Water, U.S. Geological Survey Water Supply Paper, 1473, 1970.
Hi!lei, D., Soil and Water Physical Principles and Processes, Academic Press,
New York, 1971.
Hoffman, G.J., C. Dirksen, R.D. Ingvalson, E.V. Maas, J.D. Oster, S.L. Raw-
Tins, J.D. Rhoades, and J. van Schilfgaarde, Minimizing Salt in Drain Wa-
ter by Irrigation Management, in Agricultural Water Management, 1, 233-
252, Elsevier Scientific Publishing Company, Amsterdam, 1978.
Holmes, J.W., S.A. Taylor, and S.J. Richards, Measurement of Soil Water, in
Irrigation of Agricultural Lands, R.M. Hagan, H.R. Haise, and T.W. Ed-
minster, editors. Agronomy No. 9, 275-298, American Society of Agronomy,
Madison, Wisconsin, 1967.
151
-------
Hounslow, A., J. Fitzpatrick, L. Cerrillo, and M. Freeland, Overburden Miner-
alogy as Related to Ground-Water Chemical Changes in Coal Strip Mining,
EPA-600/7-78-156, U.S. Environmental Protection Agency, Ada, Oklahoma,
1978.
Huibregtse, K.R., and J.H. Moser, Handbook for Sampling and Sample Preserva-
tion of Water and Wastewater. EPA-600/4-76-049, U.S. Environmental Pro-
tection Agency, Qua! ity Assurance Branch, Cincinnati, Ohio, 1976.
Jackson, D.R., F.S. Brinkley, and E.A. Bondietti, Extraction of Soil Water
Using Cellulose-Acetate Hollow Fibers, Journal Soil Science Soc. America,
40, 327-329, 1976.
Jackson, R.D., R.J. Reginato, and C.H.M. van Bavel, Comparison of Measured and
Calculated Hydraulic Conductivities of Unsaturated Soils, Water Resources
Research, 1, 375-379, 1965.
Jakubick, A.T., Migration of Plutonium in Natural Soils, in Transuranium Nuc-
lides in the Environment, International Atomic Energy Agency, Vienna,
T97_
Jensen, M.E., Consumptive Use of Water and Irrigation Water Requirements,
American Society of Civil Engineers, 1973.
Johnson, E.E., Inc. Ground Water and Wells, E.E. Johnson, Inc., St. Paul, Min-
nesota, 1966.
Jurinak, J.J., and J. Santill an-Medrano, The Chemistry and Transport of Lead
and Cadmium in Soils, Research Report No. 18, Agric. Exp. Station, Utah
State University, 1974.
Keeney, D.R., and R.E. Wildung, Chemical Properties of Soils, in Soils for
Management of Organic Wastes and Waste Water. Soil Sci. Soc. Amer., Amer-
ican Soc. of Agron., Crop Science Soc. of America, Madison, Wisconsin,
1977.
Keys, W.S., The Application of Radiation Logs to Ground-Water Hydrology, in
International Atomic Energy Agency Symposium on Use of Isotopes in Hy-
drology. 477-488, 1967.
Keys, W.S., and L.M. MacCary, Application of Borehole Geophysics to Water Re-
sources Investigations, in Technical Techniques of Water-Resources Inves-
tigations of the U.S. Geological Survey, Book 2, 126 pp, 1971.
Kirkham, D., Studies of Hillside Seepage in the Iowa Drift Area, Soil Sci.
Soc. Amer. Proc.. 12, 73-80, 1947.
Kirkham, D., and W.L. Powers, Advanced Soil Physics, Wiley Interscience, New
York, New York, 1972.
152
-------
Klute, A., Laboratory Measurement of Hydraulic Conductivity of Unsaturated
Soil, in Methods of Soil Analyses, C.A. Black, editor, Agronomy No. 9,
253-261, American Society of Agronomy, Madison, Wisconsin, 1965.
Klute, A., The Movement of Water in Unsaturated Soils, in The Progress of Hy-
drology, Proc. First Int. Seminar for the Hydrology Prof., The National
Science Foundation Science Seminar, Dept. of Civil Eng., The University
of Illinois, July 13-25, 1969.
Kokotov, Y.A., R.F. Popova, and A.P. Urbanyuk, Sorption of Long-Lived Fission
Products on Soils and Clay Minerals, Radiokhimiya. 3(2), 199-206, 1961.
Korte, N.E., J. Skopp, W.H. Fuller, E.E. Niebla, and B.A. Alesii, Trace Ele-
ment Movement in Soils: Influence of Soil Physical and Chemical Proper-
ties, Soil Science, 122(5), 350-359, 1976.
Kraijenhoff van deLeur, D.A., Some Effects of the Unsaturated Zone on Non-
steady Free-Surface Groundwater Flow as Studied in a Scaled Granular
Model, J. Geophys. Res.. 67(11), 4347-4362, 1962.
LaRue, M.E., D.R. Nielsen, and R.M. Hagan, Soil Water Flux Below a Rye grass
Root Zone, Agronomy Journal, 60, 625-629, 1968.
Leenheer, J.A., and E.W.D. Huffman, Jr., Classification of Organic Solutes in
Water by Using Macroreticular Resins, Journal Research U.S. Geol. Survey,
4(6), 737-751, 1976.
LeGrand, H.E., System for Evaluating the Contamination Potential of Some Waste
Sites, Journal American Water Works Association. 56(8), 959-974, 1964.
Leonard, R.A., G.W. Bailey, and R.R. Swank, Transport, Detoxification, Fate,
and Effects of Pesticides in Soil and Water, in Land Application of Soils
Material, Soil Conservation Society of America, 1976.
Levin, M.J., and D.R. Jackson, A Comparison of In-Situ Extractors for Sampling
Soil Water, Journal Soil Science Soc. America. 41, 535-536, 1977.
Lissey, A., The Use of Reducers to Increase the Sensitivity of Piezometers,
Journal of Hydrology. 5(2), 197-205, 1967.
Lohman, S.W., Ground-Water Hydraulics, U.S. Geological Survey Professional
Paper No. 708, 1972.
Lund, L.J., and P.F. Pratt, Variability of Nitrate Leaching Within Defined
Management Units, in Proceedings of the National Conference on Irrigation
Return Flow Quality Management, Colorado State University, 1977.
Lusczynski, N.J., Filter-Press Method of Extracting Water Samples for Chlo-
ride Analysis, U.S. Geological Survey Water Supply Paper, 1544-A, 1961.
Luthin, J.M., editor, Drainage of Agricultural Lands, American Society of
Agronomy, Madison, Wisconsin, 1957.
153
-------
Luthin, J.M., and P.R. Day, Lateral Flow Above a Sloping Water Table, Soil
Sci. Soc. Amer. Proc., 19, 406-410, 1955.
Mann, J.F., Wastewaters in the Vadose Zone of Arid Regions: Hydrologic In-
teractions, Ground Water, 14(6), 367-372, 1976.
Mather, J.R., and P.A. Rodriguez, The Use of the Water Budget in Evaluating
Leaching Through Solid Waste Landfills, Final Report on OWRT Project No.
A-040-DEL, Water Resources Center, The University of Deleware, Newark,
Deleware, 1978.
Matlock, W.6., G.C.A. Morin, and J.E. Posedly, Well Cuttings Analysis in
Ground-Water Resources Evaluation, Ground Water. 14(5), 272-277, 1976.
Matlock, W.G., and P.R. Davis, Groundwater in the Santa Cruz Valley, Technical
Bulletin 194, Agric. Exp. Station, The University of Arizona, Tucson,
Arizona, 1972.
McMichael, F.C., and J.E. McKee, Wastewater Reclamation at Whittier Narrows.
Publication No. 33, State of California, Water Quality, 1966.
McMillion, L.G., and J.W. Keeley, Sampling Equipment for Groundwater Investi-
gations, Ground Water, 6(2), 9-11, 1968.
McNeal, B.L., Soil Salts and Their Effect on Water Movement, in Drainage for
Agriculture, J. van Schilfgaarde, editor, Agronomy No. 17, 409-433, Amer-
ican Society of Agronomy, Madison, Wisconsin, 1974.
McWhorter, D.B., and J.A. Brookman, Pit Recharge Influenced by Subsurface
Spreading, Ground Water, 10(5), 6-11, 1972.
Meinzer, Oscar E., Ground Water, in Hydrology, Oscar E. Meinzer, editor, 385-
477, Dover Publications, Inc., New York, 1942.
Merrill, S.D., and S.C. Rawlins, Field Measurement of Soil Water Potential
with Thermocouple Psychrometers, Soil Science. 113(2), 102-109, 1972.
Meyer, W.R., Use of a Neutron Moisture Probe to Determine the Storage Coeffi-
cient of an Unconfined Aquifer, U.S. Geological Survey, Professional
Paper 450-E, 174-176, 1963.
Miller, R.H., The Soil as a Biological Filter, in Recycling Treated Municipal
Waste Water and Sludge Through Forest and Cropland, W.E. Sopper and L.T.
Kardos, editors, The Pennsylvania State University Press, 1973.
Millington, R.J., and J.P. Quirk, Permeability of Porous Media, Nature, 183,
387-388, 1959.
Mooij, H., and F.A. Rovers, Recommended Groundwater and Soil Sampling Proce-
dures, Environmental Protection Service, Report EPS-4-EC, Canada, 1976.
154
-------
Mortland, M.M., and W.D. Kemper, Specific Surface, in Methods of Soil Analy-
ses, C.A. Black, editor, Agronomy No. 9, 532-543, American Society of
Agronomy, Madison, Wisconsin, 1965.
Murrmann, R.P., and F.R. Koutz, Role of Soil Chemical Processes in Reclamation
of Wastewater Applied to Land, in Wastewater Management by Disposal on
the Land, U.S. Army Corps of Engineers, Cold Regions Research and Eng.
Lab. Specialty Report 171, 1972.
Myhre, D.L., J.O. Sanford, and W.F. Jones, Apparatus and Techniques for In-
stalling Access Tubes in Soil Profiles to Measure Soil Water, Soil
Science. 108(4), 296-299, 1969.
Nelson, J.L., Soil Column Studies with Radiostrontium: I. Effects of Tempera-
ture of Species of Accompanying Ion. H.W. 62035, Hanford Atomic Products
Operation, Richland, Washington, 1959.
Nielsen, D.R., and J.W. Biggar, Analyzing Soil Water and Solute Movement Under
Field Conditions, in Soil-Moisture and Irrigation Studies II, Interna-
tional Atomic Energy Agency, 1973.
Nielsen, D.R., W. Biggar, and K.T. Erh, Spatial Variability of Field Measured
Soil-Water Properties, Hilgardia. 43(7), 215-260, 1973.
Nommik, H., Ammonium Fixation and Other Reactions Involving a Nonenzymatic Im-
mobilization of Mineral Nitrogen in Soil, in Soil Nitrogen, W.V. Bar-
thai emew and F.E. Clark, editors, Agronomy No. 10, 19657
Norris, S.E., The Use of Gamma Logs in Determining the Character of Unconsoli-
dated Sediments and Well Construction Features, Ground Water, 10(6), 14-
21, 1972.
Ongerth, H.J., D.P. Spath, J. Crook, and A.F. Greenberg, Public Health, As-
pects of Organics in Water, Journal American Water Works Association, 65,
495, 1973.
Oster, J.D., and L.S. Willardson, Reliability of Salinity Sensors for the Man-
agement of Soil Salinity, Agronomy Journal, 63, 695-698, 1971.
Paetzold, R.F., Measurement of Soil Physical Properties, in Infiltration Re-
search Planning Workshop, Part 1, State of the Art Reports, Sciences and
Education Administration, The U.S. Dept. of Agric., ARM-NC-4, 1979.
Parizek, R.R., L.J. Kardos, W.E. Sopper, E.A. Myers, D.E. Dairs, M.A. Farrell,
and J.B. Nesbitt, Haste Water Renovation and Conservation, Pennsylvania
State Studies. No. 23. The Pennsylvania State University Press, 1967.
Parizek, R.R., and B.E. Lane, Soil-Water Sampling Using Pan and Deep Pressure-
Vacuum Lysimeters, Journal of Hydrology. 11 1-21, 1970.
Pavoni, J.L., D.J. Hagerty, and R.E. Lee, Water Resources Bulletin. 8(6),
1091-1107, 1972.
155
-------
Phene, C.J., G.J. Hoffman, and S.L. Rawlins, Measuring Soil Matric Potential
In-Situ by Sensing Heat Dissipation Within a Porous Body: I Theory and
Sensor Construction, Soil Sci. Soc. Amer. Proc.. 35, 27-33, 1971.
Phene, C.J., S.L. Rawlins, and G.J. Hoffman, Measuring Soil Matric Potential
In-Situ by Sensing Heat Dissipation Within a Porous Body: II Experimen-
tal Results, Soil Sci. Soc. Amer. Proc.,35. 225-229, 1971.
Philip, J.R., Theory of Infiltration, in Advances in Hydroscience. 5, 216-269,
1969.
Pickens, J.F., and G.E. Grisak, Reply to the Preceding Discussion of Vanhof et
al. at "A Multilevel Device for Groundwater Sampling and Piezometric Mon-
itoring", Ground Water, 17(4), 393-397, 1979.
Pickens, J.R., J.A. Cherry, G.E. Grisak, W.F. Merritt, and G.A. Risto, A Mul-
tilevel Device for Ground-Water Sampling and Piezometric Monitoring,
Ground Water, 16(5), 322-327, 1978.
Pratt, P.P., W.W. Jones, and V.E. Hunsaker, Nitrate in Deep Soil Profiles in
Relation to Fertilizer Rates and Leaching Volume, Jour. Envir. Quality,
1(1), 97-102, 1972.
Pratt, P.F., J.E. Warneke, and P.A. Nash, Sampling the Unsaturated Zone in Ir-
rigated Field Plots, Journal Soil Science Soc. America. 40, 277-279,
1976.
Pratt, P.P., A.C. Chang, J.P. Martin, A.L. Page, and C.F. Kleine, Removal of
Biological and Chemical Contaminants by Soil Systems in Association with
Ground Water Recharge by Spreading or Injection of Treated Municipal
Waste Water, in Health Aspects of Wastewater Recharge, A State-of-the-Art
Review, Water Information Center, Huntington, New York, 1978.
Quirk, J.P., and R.K. Schofield, The Effect of Electrolyte Concentration on
Soil Permeability, J. Soil Science. 6, 1963-1978, 1955.
Ralston, D.R., Influences of Water Well Design on Neutron Logging, unpublished
M.S. Thesis, The University of Arizona, Tucson, Arizona, 1967.
Rawlins, S.L., and F.N. Dalton, Psychrometric Measurement of Soil Water Poten-
tial Without Precise Temperature Control, Soil Sci. Soc. Amer. Proc., 31,
297-301, 1967.
Reeve, R.C., Hydraulic Head, in Methods of Soil Analyses, C.A. Black, editor,
Agronomy No. 9, 180-196, American Society of Agronomy, Madison, Wiscon-
sin, 1965.
Reginato, R.J., and C.H.M. van Bavel, Soil Water Measurement with Gamma Atten-
uation, Soil Sci. Soc. Amer. Proc.. 28(6), 721-724, 1964.
Reginato, R.J., Count Rate Instability in Gamma Ray Transmission Equipment,
Soil Sci. Soc. Amer. Proc., 38, 156-157, 1974.
156
-------
Reginato, R.J., and R.D. Jackson, Field Measurement of Soil-Water Content by
Gamma-Ray Transmission Compensated for Temperature Fluctuations, Soil
Sci. Soc. Amer. Proc.. 35, 529-533, 1971.
Reitemeier, R.F., Effect of Moisture Content on the Dissolved and Exchangeable
Ions of Soils and Arid Regions, Soil Science, 61(3), 195-214, 1946.
Rhoades, J.D., and L. Bernstein, Chemical, Physical and Biological Character-
istics of Irrigation and Soil Water, in Water and Water Pollution Hand-
book, Vol. 1, L.L. Ciaccio, editor, 141-222, Marcel Dekker, Inc., New
York, 1971.
Rhoades, J.D., and J. van Schilfgaarde, An Electrical Conductivity Probe for
Determining Soil Salinity, Soil Sci. Soc. Amer. Proc.. 40, 647-651, 1976.
Rhoades, J.D., and A.D. Halvorson, Electrical Conductivity Methods for Detect-
ing and Delineating Saline Seeps and Measuring Salinity in Northern Great
Plains Soils, ARS W-42. The U.S. Dept. of Agric.. 1977.
Rhoades, J.D., M.T. Kaddah, A.D. Halvorson, and J.R. Prather, Establishing
Soil Electrical Conductivity-Sal inity Calibrations Using Four-Electrode
Cells Containing Undisturbed Soil Cores, Soil Science. 123, 137-141,
1977.
Rhoades, J.D., Monitoring Soil Salinity: A Review of Methods, in Establish-
ment of Water Quality Monitoring Programs, American Water Resources Asso-
ciation, 1979a.
Rhoades, J.D., Sal inity Management and Monitoring, in Proceedings of the
Twelfth Biennial Conference on Ground Water. Sacramento, September 20-21,
1979, California Water Resources Center, The University of California,
Davis, 1979b.
Rhoades, J.D., Inexpensive Four-Electrode Probe for Monitoring Soil Salinity,
Journal Soil Science Soc. America, (in press), 1979c.
Rhodes, D.W., Adsorption of Plutonium by Soil, Soil Science, 84, 465-472,
1957a.
Rhodes, D.W., The Effect of pH on the Uptake of Radioactive Isotopes from So-
lution by a Soil, Soil Sci. Soc. Amer. Proc.. 21, 389-392, 1957b.
Rible, J.M., P.A. Nash, P.F. Pratt, and L.H. Lund, Sampling the Unsaturated
Zone of Irrigated Lands for Reliable Estimates of Nitrate Concentrations,
Journal Soil Science Soc. America, 40, 566-570, 1976.
Richards, L.A., A Soil Salinity Sensor of Improved Design, Soil Sci. Soc.
Amer. Proc., 30, 333-337, 1966.
Richards, L.A., editor, Diagnosis and Improvement of Saline and Alkali Soils.
Agric. Handbook 60, The U.S. Dept. of Agric., 1954.
157
-------
Richards, L.A., Physical Condition of Water in Soils, in Methods of Soil
Analyses, C.A. Black, editor, Agronomy No. 9, 128-152, American Society
of Agronomy, Madison, Wisconsin, 1965.
Robertson, J.M., C.R. Toussaint, and M.A. Jerque, Organic Compounds Entering
Ground Water from a Landfill, EPA-660/2-74-077, U.S. Environmental Pro-
tection Agency, 1974.
Rolston, D.E., and F.E. Broadbent, Field Measurement of Denitrification, EPA-
600/2-77-233, U.S. Environmental Protection Agency, 1977.
Rose, C.W., Agricultural Physics, Pergamon Press, 1966.
Runnells, D.D., Wastewaters in the Vadose Zone of Arid Regions: Geochemical
Interactions, Ground Water. 14(6), 374-385, 1976.
Schmidt, K.D., Protection of Ground Water from Nonpoint Sources of Pollution,
in Drinking Water Quality Enhancement Through Source Protection, R.B.
Pojasek, editor, Ann Arbor Science Pub., Inc., Ann Arbor, Michigan, 1977.
Schmidt-Collerus, J.J., The Disposal and Environmental Effects of Carbonaceous
Solid Wastes from Commercial Oil Shale Operations, University of Denver
Research Institute, Denver, Colorado, 1974.
Schwille, F., Petroleum Contamination of the Subsoil - A Hydrological Problem,
in The Joint Problems of Oil and Water Industries, P. Hepple, editor,
Elsevier, Amsterdam, 1967.
Sciences and Education Administration, Infiltration Research Planning Work-
shop, Part 1, State of the Art Reports. The U.S. Dept. of Agric., ARM-
NC-4, 1979.
Scott, V.H., and J.C. Scalmanini, Water Wells and Pumps: Their Design, Con-
struction, Operation and Maintenance, Bulletin 1889, Division of Agricul-
tural Sciences, The University of California, 1978.
Shaffer, K.A., D.D. Fritton, and D.E. Baker, Drainage Water Sampling in a Wet,
Dual-Pore Soil System, Jour. Envir. Quality, 8(2), 241-246, 1979.
Shuval, H.E., and N. Gruener, Health Considerations in Renovating Wastewater
for Domestic Use, Environmental Science and Tech., 7, 600, 1973.
Sidle, R.C., Infiltration Measurements and Soil Hydraulics Characteristics, in
Infiltration Research Planning Workshop, Part 1, State of the Art Re-
ports, Sciences and Education Administration, The U.S. Dept. of Agric.,
ARM-NC-4, 1979.
Silka, L.R., and T.L. Swearingen, A Manual for Evaluating Contamination Poten-
tial of Surface Impoundments, EPA-570/9-78-003, Ground Water Protection
Branch, Office of Drinking Water, U.S. Environmental Protection Agency,
1978.
158
-------
Simonson, R.W., What Soils Are, in Soil. The Yearbook of Agriculture. The U.S.
Dept. of Agric., 1957.
Slawson, 6.C., editor, Groundwater Quality Monitoring of Western Oil Shale De-
velopment: Identification and Priority Ranking of Potential Pollution
Sources. EPA-600/7-79-023. U.S. Environmental Protection Agency. Las
Vegas, Nevada, 1979.
Soil Moisture Equipment Corporation, Operating Instructions for the Model 1900
Soil Water Sampler, Santa Barbara, California, 1978.
Soil test, Inc., Soil Testing Equipment, General Catalog, 1976.
Sopper, W.E., and L.T. Kardos, editors, Recycling Treated Municipal Waste
Water and Sludges Through Forest and Cropland. The Pennsylvania State
University Press, 1973.
Stallman, R.W., Flow in the Zone of Aeration, in Advances in Hydroscience.
Vol. 4, Ven Te Chow, editor, Academic Press, New York, 1967.
Steele, R.6.D., and J.H. Torrie, Principles and Procedures of Statistics.
McGraw-Hill Book Co., New York, 1960.
Stevenson, C.D., Simple Apparatus for Monitoring Land Disposal Systems by Sam-
pling Percolating Soil Waters, Environmental Science and Engineering,
12(3), 329-331, 1978.
Stollar, R.L., and P. Roux, Earth Resistivity Surveys - A Method for Determin-
ing Ground-Water Contamination, Ground Water. 13(2), 145-150, 1975.
Taylor, G.S., and J.N. Luthin, The Use of Electronic Computers to Solve Sub-
surface Drainage Problems, Water Resources Research. 5(1), 144-152, 1969.
Thomas, G.W., and R.E. Phillips, Consequences of Water Movement in Macropores,
Jour. Envir. Quality. 8(2), 149-152, 1979.
Thomas, R.E., The Soil as a Physical Filter, in Recycling Treated Municipal
Wastewater and Sludge Through Forest and Cropland. W.E. Sopper and L.T.
Kardos, editors, The Pennsylvania State University Press, 1973.
Thornthwaite, C.W., and J.R. Mather, Instructions and Tables for Computing Po-
tential Evapotranspiration and Water Balance, Drexel Institute of Tech-
nology Laboratory of Climatology, Publications in Climatology, Volume X,
No. 3, Centerton, New Jersey, 1957.
Todd, O.K., R.M. Tinlin, K.D. Schmidt, and L.G. Everett, Monitoring Groundwa-
ter Quality; Monitoring Methodology, EPA-600/4-76-026, U.S. Environmen-
tal Protection Agency, Environmental Monitoring and Support Laboratory,
Las Vegas, Nevada, 1976.
Topp, G.C., and W. Zebchuck, The Determination of Soil-Water Desorption
Curves for Soil Cores, Canadian Journal of Soil Sci.. 59, 19-26, 1979.
159
-------
Trescott, P.O., and G.F. Pinder, Air Pump for Small-Diameter Piezometers,
Ground Water. 8(3), 10-15, 1970.
U.S. Bureau of Reclamation, The U.S. Dept. of the Interior, Ground Water Man-
ual, A Water Resources Technical Publication, 1977.
U.S. Environmental Protection Agency, Hazardous Waste Guidelines and Regula-
tions, Federal Register. 43(243), Part IV, 58946-59027, 1978.
U.S. Environmental Protection Agency, Draft Application Forms for EPA Pro-
grams, Federal Register. 44(116), Part III, 34346-34416, 1979a.
U.S. Environmental Protection Agency, Landfill Disposal of Solid Waste, Pro-
posed Guidelines, Federal Register, 44(59), Part II, 18138-18148, 1979b.
U.S. Environmental Protection Agency, Manual of Water Well Construction Prac-
tices, EPA-570/9-75-001, Office of Water Supply, no date.
U.S. Environmental Protection Agency, U.S. Army Corps of Engineers, and The .
U.S. Dept. of Agric., Process Design Manual for Land Treatment of Munici-
pal Wastewater. EPA-625/1-77-008 (COE EM/110-1-501), 1977.
U.S. Salinity Laboratory Staff, Minimizing Salt in Return Flow Through Irri-
gation Management. EPA-600/2-77-134, U.S. Environmental Protection
Agency, Ada, Oklahoma, 1977.
van Bavel, C.H.M., Neutron Scattering Measurement of Soil Moisture: Develop-
ment and Current Status, Proc. Int. Symp. Humidity and Moisture, Washing-
ton, D.C., 171-184, 1963.
Vanhof, J.A., K.U. Weyer, and S.H. Whitaker, Discussion of "A Multilevel De-
vice for Ground-Water Sampling and Piezometric Monitoring by J.F. Pick-
ens, J.A. Cherry, G.E. Grisak, W.F. Merritt, and B.A. Rizto", Ground
Water, 17(4), 391-393, 1979.
van Schilfgaarde, J., Theory of Flow to Drains, in Advances in Hydroscience,
6, 43-106, 1970.
Vomicil, J.A., Porosity, in Methods of Soil Analyses. C.A. Black, editor,
Agronomy No. 9, 299-314, American Society of Agronomy, Madison, Wiscon-
sin, 1965.
Wagner, G.H., Use of Ceramic Cups to Sample Soil Water Within the Profile,
Soil Science, 94, 379-386, 1962.
Walker, W.H., Where Have All the Toxic Chemicals Gone?, Ground Water, 11(2),
11-20, 1973.
Warner, D.L., Preliminary Field Studies Using Earth Resistivity Measurements
for Delineating Zones of Contaminated Ground Water, Ground Water. 7(1),
9-16, 1969.
160
-------
Warrick, A.W., and A. Amoozegar-Fard, Area Prediction of Water and Solute Flux
in the Unsaturated Zone, Final Report on Grant No. R-804-751-010, U.S.
Environmental Protection Agency, (in press), 1980.
Watson, K.K., A Recording Field Tensiometer with Rapid Response Characteris-
tics, Journal of Hydrology, 5, 33-39, 1967.
Weeks, E.P., Field Determination of Vertical Permeability to Air in the Un-
saturated Zone, U.S. Geological Survey Professional Paper No. 1051, 1978.
Wheeler, M.L., Moisture and Solute Transport in Porous Media, in Actinides in
the Environment, A.M. Friedman, editor, ACS/Symposium Series 35, American
Chemical Society, 1976.
Willardson, L.S., B.D. Meek, and M.J. Huber, A Flow Path Ground Water Sampler,
Soil Sci. Soc. Amer. Proc., 36, 965-966, 1973.
Wilson, L.G., Observations on Water Content Changes in Stratified Sediments
During Pit Recharge, Ground Water. 9(3), 29-40, 1971.
Wilson, L.G., and J.N. Luthin, Effect of Air Flow Ahead of the Wetting Front
on Infiltration, Soil Science. 96, 136-143, 1963.
Wilson, L.G., and K.J. DeCook, Field Observations on Changes in the Subsurface
Water Regime During Influent Seepage in the Santa Cruz River, Water Re-
sources Research, 4(6), 1219-1234, 1968.
Wilson, L.G., P.S. Osborne, and D.J. Percious, Dilution of an Industrial Waste
Effluent with River Water in the Vadpse Region During Pit Recharge. Paper
No. 68-727, Presented at the Winter Meetings, American Society of Agri-
cultural Engineers, 1968.
Wilson, L.G., and K.D. Schmidt, Monitoring Perched Ground Water in the Vadose
Zone, in Establishment of Water Quality Monitoring Programs, American
Water Resources Association, 1979.
Wilson, L.G., and G.G. Small, Pollution Potential of a Sanitary Landfill Near
Tucson, Hydraulic Engineering and the Environment, Proceedings 21st An-
nual Hydraulics Division Specialty Conference, American Society of Civil
Engineers, 1973.
Winograd, I.J., Radioactive Waste Storage in the Arid Zone, EOS, Transactions
American Geophysical Union. 55(10), 884-894, 1974.
Wolff, R.G., Weathering Woodstock Granite Near Baltimore, Maryland, American
Journal of Science. 265, 106-117, 1967.
Wood, W.W., A Technique Using Porous Cups for Water Sampling at Any Depth in
the Unsaturated Zone, Water Resources Research. 9(2), 486-488, 1973.
161
-------
Wood, W.W., and D.C. Signor, Geochemical Factors Affecting Artificial Recharge
in the Unsaturated Zone, Trans. Amer. Soc. of Agric. Eng.. 18(4), 677-
683, 1975.
Zohdy, A.A.R., G.P. Eaton, and D.R. Mabey, Application of Surface Geophysics
to Ground-Water Investigations, in Techniques of Water Resources Investi-
gations of The United States Geological Survey. Chapter Dl, Book 2, Col-
lection of Environmental Data, 1974.
162
-------
APPENDIX A
CONVERSION FACTORS
U.S. Customary to SI (Metric)
U.S. customary
Name
acre
acre- foot
cubic foot
cubic feet per second
degrees Fahrenheit
feet per second
foot (feet)
gallon(s)
gallons per acre per
day
gallons per day
gallons per minute
horsepower
inch(es)
inches per hour
unit
Abbreviation
acre
acre- ft
ft3
ft3/s
°F
ft/s
ft
gal
gal/ acre* d
gal/d
gal/min
hp
in.
in./hr
Multiplier
0.405
1,234
28.32
0.0283
28.32
0.555(°F-32)
0.305
0.305
3.785
9.353
4.381 x lO-5
0.0631
0.746
2.54
2.54
Symbol
ha
m3
L
L/s
°C
m/s
m
L
L/ha'd
L/s
L/s
kW
cm
cm/h
SI
Name
hectare
cubic metre
litre
cubic metre
litres per second
degrees Celsius
metres per second
metre(s)
litre(s)
litres per hectare
per day
litres per second
litres per second
kilowatt
centimetre (s)
centimeters per
mile
mi
1.609
hour
km kilometre
(continued)
163
-------
CONVERSION FACTORS (continued)
U.S. customary
Name
miles per hour
million gallons
million gallons per
acre
million gallons per
day
parts per million
pound(s)
pounds per acre per
day
pounds per square
inch
square foot
square inch
square mile
unit
Abbreviation
mi/h
Mgal
Mgal/acre
Mgal/d
ppm
Ib
lb/acre«d
lb/in.2
ft2
in.2
mi 2
Multiplier
0.45
3.785
3,785
8,353
43.8
0.044
1
0.454
453.6
1.12
0.069
0.69
0.0929
6.452
2.590
259
Symbol
m/s
'ML
m3
m3/ha
L/s
m3/s
mg/1
kg
g
kg/ha»d
kg/cm2
N/cm2
m2
cm2
km2
ha
SI
Name
metres per second
megalitres (litres
x 106)
cubic metres
cubic metres per
hectare
litres per second
cubic metres per
second
milligrams per
litre
kilogram(s)
gram(s)
kilograms per
hectare per day
kilograms per
square centimetre
Newtons per square
centimetre
square metre
square centimetre
square kilometre
hectare
164
-------
GLOSSARY
adjusted sodium adsorption ratio: A modified form of the sodium adsorption
ratio which accounts for the interrelationships among sodium, calcium,
and magnesium ions in an applied water or matrix solution, and for the
tendency for calcium and magnesium ions to precipitate or dissolve in the
presence of carbonate and bicarbonate.
adsorption isotherm: A graphical representation of the amount of adsorbate
(solute) adsorbed by an adsorbent as a function of the equilibrium con-
centration of adsorbate (Bohn, NcNeal, and O'Connor, 1979).
anisotropy: A porous system is anisotropic with respect to the hydraulic con-
ductivity (or other properties) if values of the hydraulic conductivity
are not equal in all directions away from a given point.
antecedent water: Water stored in a soil prior to infiltration of applied
water.
average linear velocity: The flux (specific discharge) divided by the water
filled porosity; accounts for variations in the cross-sectional areas
through which flow occurs in a porous system.
capillary fringe: The basal region of the vadose zone comprising sediments
that are saturated, or nearly saturated, near the water table, gradually
decreasing in water content with increasing elevation above the water
table.
cascading water: Perched groundwater which enters a well casing via cracks or
uncovered perforations, trickling or pouring down the inside of the
casing.
cation exchange capacity: The total capacity of a porous system to adsorb ca-
tions from a solution.
clay micelle: An individual clay particle.
distribution coefficient: A representation of the partitioning of solids be-
tween liquid and solid phases in a porous system (Freeze and Cherry,
1979).
165
-------
electrical conductivity: The inverse of the electrical resistivity of a solu-
tion containing electrolytes; measured via a wheatstone bridge and a
standardized cell containing platinum electrodes, exactly 1 cm^ in sur-
face area, located a distance of 1 cm apart.
field capacity (specific retention): The amount of water retained by a soil
(or aquifer) against the force of gravity.
fillable porosity: The volume of water that an unconfined aquifer stores dur-
ing a unit rise in water table per unit surface area.
flux: The volume of water crossing a unit area of porous material per unit
time.
gravitational head: The component of total hydraulic head relating to the po-
sition of a given mass of soil water relative to an arbitrary datum.
groundwater recharge: The flux of water across a water table.
groundwater zone: The hydrogeological region underlying a water table.
hydraulic conductivity: The proportionality factor in Darcy's equation; gen-
erally, a measure of the ease with which water moves through a porous
system.
hydraulic gradient: The change in total hydraulic head of water per unit dis-
tance of flow.
infiltration capacity: The maximum rate at which water enters a soil.
isotropy: A porous system is isotropic with respect to hydraulic conductivity
(or other properties) if values of the hydraulic conductivity are equal
in all directions away from a point.
Klinkenberg effect: An effect which accounts for the greater permeability of
a porous system to air movement than water movement; in particular, the
velocity of air in the immediate vicinity of grains is finite because of
slippage; in contrast, the velocity of water near grains is zero.
matric potential: The energy required to extract water from a soil against
the capillary and adsorptive forces of the soil matrix.
matric suction: For an isothermal soil system, matric suction is the pressure
difference across a membrane separating soil solution, in-place, from the
same in bulk (Hi!lei, 1971).
matrix: The solid framework of a porous system.
osmotic potential: The component of the total soil-water potential associated
with dissolved ions.
166
-------
perched groundwater: A saturated groundwater body in the vadose zone, com-
monly developed at the interface between regions of varying texture.
percolation: The movement of water through the vadose zone, in contrast to
infiltration at the land surface and recharge across a water table.
porosity: The fraction of a given volume of a porous system not occupied by
solid grains.
pressure head: The head of water at a point in a porous system, negative for
unsaturated systems, positive for saturated systems; quantitatively, it
is the water pressure divided by the specific weight of water.
saturation extract: A solution extracted from a solids sample wetted with
distilled water until a prescribed end point has been reached.
sodium adsorption ratio: A relationship between the concentration of sodium
in an applied water or matrix solution and the concentrations of calcium
and magnesium.
soil bulk density: The mass of dry soil in a volume of bulk soil.
soil peds: A soil structural unit, such as a crumb, prism, granule, or
block.
soil-water characteristic curve: A graphical representation of the change in
water content of a soil or porous media with changing soil-water
pressure.
soil-water pressure: The pressure on the water in a soil-water system, as
meaured by a piezometer for a saturated soil, or by a tensiometer for an
unsaturated soil.
specific discharge: See flux.
specific yield: The volume of water that an unconfined aquifer releases from
storage per unit surface area of aquifer per unit drop in the water table
elevation.
tortuosity: A general term used to describe variations in the flow paths of
water in a porous system caused by the presence of solid grains; quanti-
tatively, it is the ratio of the distance travelled by a mass of water
flowing through pores of a system to the length of the system (Hi 11 el,
1971).
total hydraulic head: The sum of gravitational head and pressure head in wa-
ter within a porous system.
total soil-water potential: The sum of the energy-related components of a
soil-water system; i.e., the sum of the gravitational, matric, and os-
motic components.
167
-------
vadose zone: The hydrogeological region extending from the soil surface to
the top of the principal water table.
water content: The amount of water stored within a porous matrix, expressed
on either a volumetric (volume per unit volume) or mass (mass per unit
mass) of solid.
water content profile: A log of the change in water content with depth in a
profile through the vadose zone.
168
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/7-80-134
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
MONITORING IN THE VADOSE ZONE:
Elements and Methods
A review of Technical
5. REPORT DATE
June 1980
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
L-rG. Wilson
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
General Electric Company--TEMPO
Center for Advanced Studies
Santa Barbara, California 93102
10. PROGRAM ELEMENT NO.
INE833
11. CONTRACT/GRANT NO.
V-Q591-NALX
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
U.S. Environmental Protection Agency-Las Vegas, Nevada
Office of Research and Development
Environmental Monitoring Systems Laboratory
Las Vegas, Nevada 89114
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report covers the topics of 0) principles of pollutant movement in
the vadose zone (zone of aeration or unsaturated zone), (2) basic chemical
reactions of fluids in the zone, (3) state-of-the-art monitoring techniques,
and (4) the relative advantages and disadvantages of the different monitoring
techniques. Recent intense concern over hazardous waste disposal has indicated
the need for instruction on how to monitor in the vadose zone and to identify
the potential gains from the limitations of the methods available for
monitoring in this zone. This report provides technical information needed by
regulating agencies and industrial concerns in dealing with waste disposal
problems. In addition, the basis for future research is provided through
identification of present monitoring limitations.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COS AT I Field/Group
Ground water
Water quality
Water wells
Aquifers
Hydrogeology
Sanitary landfills
Monitoring
Soil moisture
Leachate
Unsaturated zone
Vadose zone
44G
48E,F,G
68C,D
18, DISTRIBUTION STATEMENT
RELEASE TO THE PUBLIC
19. SECURITY CLASS (ThisReport)'
UNCLASSIFIED
21. NO. OF PAGES
184
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
------- |